December 2015 Volume LII No. 2

Transcription

Volume LII No. 2
THE PHILIPPINE REVIEW OF ECONOMICS
December 2015
Volume LII No. 2
ISSN 1655-1516
December 2015
The Philippine Review
of Economics
FESTSCHRIFT FOR HAL C. HILL
Editor-in-Chief
EMMANUEL S. DE DIOS
Editorial Advisory Board
RAUL V. FABELLA
HAL C. HILL
CHARLES Y. HORIOKA
KIAN GUAN LIM
ROBERTO S. MARIANO
JOHN VINCENT C. NYE
GERARDO P. SICAT
Maria Socorro Gochoco-Bautista
Arsenio M. Balisacan
Guest Co-Editors
Urban-rural income and wage
gaps in the Philippines
American private direct
investment in the Philippines
after independence
Gerardo P. Sicat
Singapore’s five decades of
development
Chia Siow Yue
The Philippines: on the road to
being an emerging economy
Development progeria
JEFFREY G. WILLIAMSON
Malaysia-China trade
Associate Editors
DANTE B. CANLAS
RAMON L. CLARETE
LAWRENCE B. DACUYCUY
FRANCISCO G. DAKILA JR.
CRISTINA C. DAVID
JONNA P. ESTUDILLO
MARIA S. FLORO
GILBERTO M. LLANTO
ANICETO C. ORBETA
ERNESTO M. PERNIA
STELLA LUZ A. QUIMBO
Managing Editor
HONLANI RUTH J. RABE
Karl Kendrick Chua, Louie
Limkin, John Nye, and
Jeffrey G. Williamson
Class participation in politics in
Southeast Asia
The effect of trade policy
on firm productivity in Thai
manufacturing
Households’ Access to
Financial Services
The ASEAN business cycle and
China’s slowdown
Goods trade liberalization
under the ASEAN Economic
Community
Just how good is
unemployment as a measure
of welfare? A note
Dante B. Canlas
Sarah Lynne S. Daway
and Raul V. Fabella
Tham Siew Yean and
Andrew Kam Jia Yi
Joseph J. Capuno
Archanun Kohpaiboon and
Juthathip Jongwanich
Gilberto M. Llanto
Bhanupong Nidhiprabha
Ramon L. Clarete and
Philip Arnold P. Tuaño
Emmanuel S. de Dios and
Katrina Dinglasan
A joint publication of the
University of the Philippines
School of Economics
and the
Philippine Economic Society
P R E The Philippine Review of Economics
A joint publication of the UP School of Economics (UPSE)
and the Philippine Economic Society (PES)
EDITOR-IN-CHIEF
Emmanuel S. de Dios
UP SCHOOL OF ECONOMICS
EDITORIAL ADVISORY BOARD
Raul V. Fabella
Hal C. Hill
AUSTRALIAN NATIONAL UNIVERSITY
Charles Y. Horioka
ASIAN GROWTH RESEARCH INSTITUTE
(KITAKYUSHU)
Kian Guan Lim
SINGAPORE MANAGEMENT UNIVERSITY
Roberto S. Mariano
UNIVERSITY OF PENNSYLVANIA
John Vincent C. Nye
GEORGE MASON UNIVERSITY
Gerardo P. Sicat
Jeffrey G. Williamson
HARVARD UNIVERSITY
ASSOCIATE EDITORS
Dante B. Canlas
Ramon L. Clarete
Lawrence B. Dacuycuy
DE LA SALLE UNIVERSITY
Francisco G. Dakila Jr.
BANGKO SENTRAL NG PILIPINAS
Cristina C. David
PHILIPPINE INSTITUTE
FOR DEVELOPMENT STUDIES
Jonna P. Estudillo
NATIONAL GRADUATE INSTITUTE
FOR POLICY STUDIES (TOKYO)
Maria S. Floro
AMERICAN UNIVERSITY (WASHINGTON D.C.)
Gilberto M. Llanto
Aniceto C. Orbeta
Ernesto M. Pernia
Stella Luz A. Quimbo
MANAGING EDITOR
Honlani Ruth J. Rabe
COPY EDITOR
Dinna Dayao
Aims and Scope: The Philippine Review of Economics (pre)
invites theoretical and empirical articles on economics and
economic development. Papers on the Philippines, Asian and
other developing economies are especially welcome. Book
reviews will also be considered.
The pre is published jointly by the up School of Economics
and the Philippine Economic Society. Its contents are indexed
in the Journal of Economic Literature, EconLit, and RePEc.
pre’s readership includes economists and other social scientists
in academe, business, government, and development research
institutions.
Publication Information: The pre (issn 1655-1516) is a
peer-reviewed journal published every June and December of
each year. A searchable database of published articles and their
abstracts is available at the pre website (http://pre.econ.upd.edu.
ph).
Subscription Information:
Subscription correspondence may be sent to the following
addresses:
• [email protected]
• pssc Central Subscription Service,
psscenter, Commonwealth Avenue, 1101, Diliman, Quezon
City, Philippines.
P.O. Box 205, up Post Office, Diliman, Quezon City,
Philippines 1101
PHONE: 922-9627, FAX: 924-4178/926-5179
Submissions: Authors may submit their manuscripts to addresses
below:
• [email protected] or [email protected]
• The Editor, The Philippine Review of Economics, Rm 237,
School of Economics, University of the Philippines, Diliman,
Quezon City, 1101.
Manuscripts must be written in English and in ms Word format.
All graphs and tables must be in Excel format. Submission of a
manuscript shall be understood by the pre as indicating that the
manuscript is not under consideration for publication in other
journals. All submissions must include the title of the paper,
author information, an abstract of no more than 150 words, and a
list of 3–4 keywords. Complete guidelines can be viewed in pre’s
website.
Copyright: The Philippine Review of Economics is protected
by Philippine copyright laws. Articles appearing herein may
be reproduced for personal use but not for mass circulation. To
reprint an article from pre, permission from the editor must be
sought.
The Philippine Economic Society
Founded 1961
2015 OFFICERS AND
BOARD OF DIRECTORS
PRESIDENT
Victor Abola
The Philippine Economic Society (PES) was established
in August 1962 as a nonstock, nonprofit professional
organization of economists.
UNIVERSITY OF ASIA AND THE PACIFIC
VICE PRESIDENT Rosemarie Edillon
NATIONAL ECONOMIC AND DEVELOPMENT AGENCY
SECRETARY Majah-Leah Ravago
UNIVERSITY OF THE PHILIPPINES DILIMAN
TREASURER Ronald Mendoza
ASIAN INSTITUTE OF MANAGEMENT
MEMBERS
Anthony Abad
TA TRADE ADVISORY GROUP
Gil Beltran
DEPARTMENT OF FINANCE
Roehlano Briones
Over the years, the PES has served as one of the strongest
networks of economists in the academe, government, and
business sector.
Recognized in the international community of professional
economic associations and a founding member of the
Federation of ASEAN Economic Associations (FAEA), the
PES continuously provides a venue for open and free
discussions of a wide range of policy issues through its
conferences and symposia.
Through its journal, the Philippine Review of Economics
(PRE), which is jointly published with the UP School of
Economics, the Society performs a major role in improving
the standard of economic research in the country and in
disseminating new research findings.
PHILIPPINE INSTITUTE FOR DEVELOPMENT STUDIES
Jose Camacho
UNIVERSITY OF THE PHILIPPINES LOS BAÑOS
Lawrence Dacuycuy
DE LA SALLE UNIVERSITY
Diwa Guinigundo
BANGKO SENTRAL NG PILIPINAS
Peter Lee U
UNIVERSITY OF ASIA AND THE PACIFIC
Emilio Neri
BANK OF THE PHILIPPINE ISLANDS
IMMEDIATE PAST PRESIDENT
Stella Luz Quimbo
UNIVERSITY OF THE PHILIPPINES
PHILIPPINE REVIEW OF ECONOMICS
Emmanuel de Dios
EDITOR
COUNCIL OF ADVISERS Romeo Bernardo
LAZARO, BERNARDO, TIU AND ASSOCIATES
Raul Fabella
Cielito Habito
ATENEO DE MANILA UNIVERSITY
Ernest Leung
Solita Collas-Monsod
Cesar Virata
C. VIRATA AND ASSOCIATES
At present the society enjoys the membership of some
800 economists and professionals from the academe,
government, and private sector.
• Lifetime Membership – Any regular member
who pays the lifetime membership dues shall be
granted lifetime membership and shall have the
rights, privileges, and responsibilities of a regular
member, except for the payment of the annual
dues.
• Regular Membership – Limited to individuals 21
years of age or older, who have obtained at least
a bachelor’s degree in economics, or who, in the
opinion of the Board of Directors, have shown
sufficient familiarity and understanding of the
science of economics to warrant admission to
the Society. Candidates who have been accepted
shall become members of the Society only upon
payment of annual dues for the current year.
• Junior Membership – This is reserved for fulltime college or graduate students majoring in
economics. Affiliation for junior membership is
coursed through the Junior Philippine Economic
Society (JPES).
For more information, visit: www.phileconsociety.org.
The Philippine Review
of Economics
Volume LII No. 2
December 2015
ISSN 1655-1516
iiiForeword
1
Urban-rural income and wage gaps in the Philippines:
measurement error, unequal endowments,
or factor market failure?
Karl Kendrick Chua, Louie Limkin, John Nye,
and Jeffrey G. Williamson
22
American private direct investment in the Philippines after
independence
Gerardo P. Sicat
38
Singapore’s five decades of development: lessons and future
directions
Chia Siow Yue
65 The Philippines: on the road to being an emerging economy
Dante B. Canlas
84
Development progeria: the role of institutions
and the exchange rate
Sarah Lynne S. Daway and Raul V. Fabella
100
Malaysia-China trade: old and new routes
Tham Siew Yean and Andrew Kam Jia Yi
118
Class participation in politics in Southeast Asia
Joseph J. Capuno
ii
Contents
143
The effect of trade policy on firm productivity in Thai
manufacturing
Archanun Kohpaiboon and Juthathip Jongwanich
170 Households’ Access to Financial Services:
Some Evidence from Survey Data
Gilberto M. Llanto
192
The ASEAN business cycle and China’s slowdown
210
Goods trade liberalization under the ASEAN Economic
Community: effects on the Philippine economy
Ramon L. Clarete and Philip Arnold P. Tuaño
234
Just how good is unemployment as a measure of welfare?
A note
Emmanuel S. de Dios and Katrina Dinglasan
Foreword
Hal C. Hill retired from the Australian
National University in 2014, where he
was H.W. Arndt Professor of Southeast
Asian Economies, capping a long
and distinguished academic career.
This Festschrift is a small tribute to a
distinguished colleague and friend. The
papers were contributed by an impressive
and select group of scholars, some of
whom are his former students but all of
whom, in any case, are his close friends.
Hal has devoted his entire career to
understanding and promoting economic
development in Southeast Asia. He
stands in the front ranks of experts in
the economics and political economy of
many countries in the region, especially
Indonesia and the Philippines, the subjects of his research over many decades.
He has mentored scores of the region’s scholars, and policy makers, including
many who currently hold important positions in government. He has guided and
generously promoted the careers of young researchers.
Hal’s work is especially valued in the Philippines—his first overseas foray
after graduate school was as a Visiting Assistant Professor of Economics at the
University of the Philippines School of Economics from May 1980 until May
1982—where his critical analysis of difficult issues and unprejudiced attitude
have made him a scholar trusted by all sides.
Both co-editors of this Festschrift have had the honor and pleasure of working
closely with Hal over the years. As part of his wide-ranging work, Hal co-edited
two volumes on the Philippines with Arsi Balisacan, the first being a wellreceived survey of long-term trends and major sectors in the Philippine economy
[Balisacan and Hill 2003]. This volume was significant in bringing together the
work and wisdom of both an earlier and a more recent generation of development
scholars, providing an updated analysis until the threshold of the current century.
An important sequel to that venture in “transnational scholarship” was Balisacan
and Hill [2007], which tackled subnational issues of regional development and
iv
Foreword
decentralization. Typical of Hal’s influence here, as elsewhere, was his insistence
on a comparative perspective, i.e., on the need to examine local experience in the
light of what other countries have undergone, thus avoiding the solipsism that
tempts all country specialists. In 2013, he organized and co-edited a volume with
Corina Gochoco-Bautista [Hill and Gochoco-Bautista 2013] which discusses
the challenges Asia faces in sustaining growth by examining Asia’s growth
performance relative to its growth dynamics, the development of key sectors
and institutions, savings and investment behavior, and aspects of country-level
idiosyncrasies and regional cooperation and integration. Ever active and relevant,
Hal is even now among those helping the Philippine government craft a longterm vision for Philippine growth and development dubbed Filipino 2040.
An initiative of the National Economic and Development Authority in
collaboration with the country’s multilateral development partners, the business
sector, and the academe, Filipino 2040 aims to identify the policy and institutional
imperatives necessary to realize the vision of a high quality of life for all Filipinos.
On a personal level, Hal has ever been the gracious host whenever any of us
would visit Canberra. He would open a nice bottle (or maybe two or three) from
his wine cellar. On his visits to the Philippines, he has generously plied the dean’s
cellar with some of the best bottles in his collection. Corina once gave his name
as a reference when she applied for a visa to Australia. The embassy tried to verify
whether she in fact knew him by asking her to list his birth date! Hal very kindly
provided the information.
Hal is certain to remain active and to continue his distinguished scholarship
for many more years to come. In the meantime, this is one way to thank him
for his friendship and to convey to him the high regard and appreciation of the
economics profession.
References
Balisacan, A. and H. Hill, eds. [2003] The Philippine economy: development,
policies, and challenges. Oxford: Oxford University Press.
Balisacan, A. and H. Hill, eds. [2007] The dynamics of regional development:
the Philippines in East Asia. Cheltenham, U.K.: Edward Elgar and Asian
Development Bank.
Hill, H. and M.S. Gochoco-Bautista [2013] Asia rising-growth and resilience in
an uncertain global economy. Cheltenham, U.K.: Edward Elgar and Asian
Development Bank.
PRE
The Philippine Review of Economics
Vol. LII No. 2, December 2015 pp. 1-21
Urban-rural income and wage gaps in the Philippines:
measurement error, unequal endowments, or factor
market failure?
Karl Kendrick Chua*, Louie Limkin*, John Nye**,
and Jeffrey G. Williamson***
Income inequality is higher in the Philippines than in most of
its Asian neighbors, and spatial inequality accounts for a fairly
large share of it. There is little evidence of labor market failure
in the Philippines since, when properly measured, wage gaps by
skill level are modest. Unequal endowments account for most
of the urban-rural income gaps. That is, individual attributes
of workers and households explain the majority of the urbanrural gaps, and schooling, skill, and experience are the three
individual characteristics that matter most. Provincial variables,
like typhoon incidence, government corruption, school
crowding, and access to health facilities, matter far less. Workers
born in the cities and immigrants to the cities invest more in
human capital than do rural workers. However, this paper cannot
tell us how much of that is due to better human-capital-building
infrastructure supply in the cities and how much is due to higher
urban demand for that infrastructure.
JEL classification: D3, J3, O3, R1
Keywords: regional inequality, wage gaps, migration, the Philippines
1. Introduction
Large urban-rural wage and income gaps have been one of the most-studied
phenomena in the development literature.1 At the same time, high rates of labor
migration from rural to urban employment and from agriculture to industry and
modern service employment have been notable characteristics of the transition
Weber [1899], Hatton and Williamson [1992], Yankow [2000], Gould [2007], Sicular et al. [2007], and
Young [2013].
1
2
Chua, Limkin, Nye, and Williamson:
Urban-rural income and wage gaps in the Philippines
from pre-industrial to modern development. The coexistence of rapid emigration
from poor and low-wage rural areas with persistent urban-rural income gaps
might seem a puzzle, until one realizes that rapid semi-skilled and skilled labor
demand growth and job creation in urban areas might explain both.
The Philippines, despite starting above its Southeast Asian neighbors’
development levels in the first half of the twentieth century and even in the
1950s, has underperformed since the 1970s, especially in industry (de Dios
and Williamson [2014]; World Bank [2013]). It also has relatively high income
inequality, rural-urban income gaps, and regional inequality.2 Policy is said
to have favored Luzon, and, in particular, Metro Manila and its surrounding
regions. Although labor market forces should have led to a closing of the
urban-rural gap and a diminution in regional inequality, they have not done so,
perhaps, some argue, due to impediments to migration, such as minimum wage
legislation, regularization policy, labor unions, urban immigration restrictions,
land zoning rules, and other market interventions.
This paper does find some evidence of labor market failure, but it is very
modest. In general equilibrium theory, land, labor, and capital markets all matter
in accounting for spatial income gaps. But what about the spatial labor market
itself? Are there big urban-rural wage gaps? If so, are they explained by policy
intervention, weak institutions, or rather by rural behavior, lack of language
skills, poor rural schools, fear of religious persecution (a poor Muslim south and
a rich Christian north), or some overwhelming demographic glut in rural areas
[Williamson 2013]?
Understanding the determinants of income inequality both within and
between regions has been and will be essential to identifying the sources of
poor Philippine performance and to designing policy that will promote more
rapid equitable development. The last major study of spatial income inequality
in the Philippines was undertaken by Balisacan and Fuwa [2004]. In their
study, the authors concluded that income inequality had undergone little change
from 1985 to 2000. They also noted that income inequality was higher in the
Philippines than in most of its Asian neighbors, although it is not as high as in
Latin America, with which the Philippines is often compared. Furthermore, they
concluded that while spatial inequality accounted for a large share of overall
income inequality in the Philippines, that share seemed to be diminishing.
Finally, they reported that income per capita was slowly converging across
provinces, in contrast with many other developing economies.
For example, see Berry [1978] and Balisacan and Fuwa [2004].
2
The Philippine Review of Economics, Volume LII No. 2, December 2015
3
This paper extends the Balisacan and Fuwa assessment forwards one decade.
It also improves the data base and estimates the sources of the urban-rural gap
using Fields and Oaxaca decomposition techniques. We explore the roles of
measurement error, unequal endowments, and labor market failure, plus the
likely impact of capital and land market failure. Thus, we ask: How much of the
nominal wage gap disappears when proper site-specific prices—especially those
of non-tradable services—are used for rural areas? How much of the measured
average wage gap disappears when it is computed for the same occupations and
skills, rather than for some average wage? Are the gaps different for skilled and
unskilled workers, perhaps reflecting different responses to wage differentials?3
These are the questions that should be answered in making a judgment
about labor market failure. If, instead, the focus is on the determinants of
spatial income inequality, then different questions are relevant. Indeed, do
household wage incomes behave differently than do wage rates by occupation
and skill, perhaps due to household labor participation rate or employment
rate differences between rural and urban locations? Do household total income
gaps behave differently than wage income gaps, perhaps due to unobserved
positive individual attributes (migrants self-select), and/or due to differences in
investment income opportunities by location, and/or due to remittances from
family abroad? Most importantly, are those large rural-urban income gaps
explained mostly by wage gaps or mostly by endowment gaps?
2. Philippine urban-rural gaps, total inequality, and regional inequality: a
comparative assessment
Based on total household income, inequality in the Philippines improved
very slowly between 2000 and 2012 (Figure 1).4 The trends are broadly the same
whether we use the Gini coefficient or Theil’s T Statistic. However, we will use
Theil’s T Statistic in what follows since it is easily decomposable while the Gini
is not.
Table 1 reports country urban-rural gap measures for three statistics: the
average wage; the average household wage income; and the average household
total income. The gap is measured as the difference between urban and rural,
divided by urban: thus, the percentage by which urban exceeds rural. The sample
is small in any of those categories, and only the Philippines reports all three.
The idea here is that skilled workers are more likely than poorer, unskilled workers to have the financial
resources to invest in a move.
4
Philippine household income data are available every three years. Thus, to update the assessment by
Balisacan and Fuwa, our analysis uses data from 2002 to 2012.
3
4
0.60
0.50
0.40
0.30
2000
2003
2006
Gini coeff.
2009
2012
Theil T
Source: NSO, WB staff estimates
FIGURE 1. Measures of inequality
TABLE 1. Cross-country comparison of wage and income gaps
Country/Area
Philippines
Year
2011/12
r-u gap
Data used
34
Individual wage
44
HH total income
54
HH wage income
Nepal
Bangladesh
EU25
Indonesia
India
2003
2000
2003
2000
2005
4
9
12
16
23
Individual wage
Individual wage
HH total income
Individual wage
Individual wage
US
2011
24
HH total income
Vietnam
India
Malaysia
Cambodia
China
1998
2005
2004
1999
2002
26
31
53
56
58
Individual wage
Individual wage
HH total income
HH wage income
Per capita income
China
2011
81
HH wage income
Source
Labor Force Survey (2011)
Family Income and Expenditure
Survey (2012)
Family Income and Expenditure
Survey (2012)
Hertz et al. (2008)
Hertz et al. (2008)
Cameron et al. (2008)
Hertz et al. (2008)
Karan an Selvaraj (2008)
Bureau of Labor Statistics, US
Department of Labor
Hertz et al. (2008)
Karan an Selvaraj (2008)
Ragayah (2008)
Morris (2007)
Sicular et al. (2007)
National Bureau of Statistics of
China
5
But the Philippine results are revealing. The wage gap, 34 percent, is the
smallest of the three. It reflects three influences: the fact that urban jobs are more
skilled; that the urban cost of living is higher; and that there may be some labor
market failure by specific job category. The household total wage income gap, 54
percent, is much bigger. It reflects the fact that urban immigrants self-select, that
low-wage (high-wage) earners cluster in low-income (high-income) households,
and that urban workers have higher hours worked per month. Surprisingly,
the household total income gap, 44 percent, is lower than the household wage
income gap. This cannot be explained by remittance rates since the share of
remittance income to total household income was roughly the same for rural and
urban households. In 2012, they were 17 percent of total rural household income
and 15 percent of urban household income. Most importantly, the wage gap is
significantly less than both income gaps.
Compared with its Asian neighbors, wage gaps in the Philippines are high:
that 34 percent is the highest in Table 1; the reported wage gap is much higher
than in Nepal (4 percent), Bangladesh (9 percent), and Indonesia (16 percent);
and it is even higher than Vietnam (26 percent) and India (23-31 percent). The
Philippine household wage income gap is only exceeded by China (81 versus
54 percent). Finally, the Philippine household total income gap of 44 percent
is closest to Malaysia (53 percent), and both are much bigger than that of the
United States (24 percent) and the European Union (12 percent).
60
55
50
45
40
35
30
2000
urban-rural
2003
2006
Luzon-VisMin
2009
2012
PHL-ARMM
FIGURE 2. Income gaps (in percent)
Given the country’s heterogeneous island geography with multiple languages,
ethnic groups, and religions, one might expect that spatial income gaps
6
(Figure 2) would account for a large share of the high Philippine inequality.
But do they? Table 2 reports some country measures where total inequality is
decomposed into between regions and within regions. The canonical case of
Italy—with its persistent “North-South” or Mezzogiorno problem—records 13
percent of total inequality due to regional gaps. But the figure for the Philippines
is higher, 20-24 percent in 2000, although it dropped to 11-14 percent in 2012.
The figures are smaller for better integrated and richer countries, but they are
bigger for China (38 percent) and Vietnam (36 percent).5 Although it is not
shown here, the urban versus rural between share in total Philippine inequality is
a bit more than 10 percent in 2012, which is almost equal to the regional between
share of 11 percent.
TABLE 2. Within and between income inequality: the Philippines
compared with other countries
Country/Area
Year
Within
region/
province
Between
regions/provinces
Source
Philippines
2011
80/76
20/24
Family Income and
Expenditure Survey (2012)
Philippines
2012
89/86
11/14
Family Income and
Expenditure Survey (2012)
Indonesia
China
Vietnam
Indonesia
Mozambique
Indonesia
Italy
Russia
India
Czech
Republic
Canada
Poland
1997
1998
1999
1999
1997
1990
995
1995
1983
50
62
64
78
80
83
87
90
94
50
38
36
22
20
17
13
10
6
Akita (2003)
Akita (2003)
Minot et al. (2003)
Tadjoeddin (2003)
Simler and Nhate (2005)
Tadjoeddin (2003)
Forster et al. (2005)
Mishra and Parikh (1992)
1996
97
3
1997
1999
98
98
2
2
Gray et al. (2003)
We exclude Indonesia from these comparisons since it has very different estimates for within and between
province inequalities from two different sources (Akita [2003]; Tadjoeddin [2003]) and even for the same
source [Akita 2003].
5
7
Thus, while urban and regional income gaps certainly contribute to inequality
in the Philippines, it is the variance of household incomes within urban and within
rural areas and regions that matters most (80-89 percent). A study by Alwyn
Young [2013] used consumption data from country Demographic and Health
Surveys to show that the urban-rural gap explains a much higher share of total
consumption inequality, on average 40 percent in his low-income sample, and the
Philippines was about in the middle of his 65-country pack [Young 2013:1748].6
3. Gap measurement errors
3.1. Getting the prices right
Since migrants respond to real wages and since living costs are lower in rural
areas within all provinces and lower still in rural areas in the poorest provinces,
some effort must be made to get the prices right. Typically, either the urban-rural
gap estimates are nominal, or they are deflated only by the prices of tradable
commodities (e.g. they exclude non-traded services which are priced low in rural
areas since nominal wages are also low), and when deflated, they use provincial
capital prices rather than rural prices. Since rural cost of living indices or consumer
price indices are not available from official Philippine sources, we compute them
using a modified version of the provincial consumer price indices reported by the
National Statistics Office. We consider the official provincial consumer price indices
as urban since the underlying price data are mainly collected from urban areas.
Our rural consumer price indices are computed using five consumption
categories: agricultural food; other food; tradable goods and services7; semitradable goods and services; and non-tradable goods and services.8 Agricultural
food prices are taken from the Bureau of Agricultural Statistics, which provides
farm gate and retail prices of key commodities. Only rice has complete price data
at the provincial level so it is used as a proxy for agricultural food, and rural prices
are calculated as the average of the farm gate and retail price. The agricultural
food price index in the initial year (2001) is constructed as the average of the farm
gate and retail price divided by retail price. Succeeding years follow trends in the
farm gate and retail price. The prices of other food, tradables, and semi-tradable
goods and services follow the provincial consumer price indices. Non-tradable
goods and services prices are assumed to follow the movement of nominal rural
wages. The initial year price of non-tradables is computed as the ratio of rural to
urban nominal wages multiplied by the urban price of these non-tradables. Rural
non-tradable price movements in succeeding years follow the growth in rural
It should be noted that Young’s consumption measure is based on ownership of durables and housing
quality [Young 2013:1734], a measure that would exaggerate gaps.
7
For the purposes of this analysis, tradables include alcoholic beverages, tobacco, clothing, footwear,
furnishings, household equipment, and communication.
8
It is assumed that 30 percent of semi-tradables are not traded.
6
8
nominal wages. This method ensures that the prices of food and non-tradable
goods and services are cheaper in rural areas.
65
60
55
50
45
40
35
30
25
20
2000
2003
2006
nominal rural-urban
2009
2012
real rural-urban
FIGURE 3. Rural-urban income gaps (nominal and real in percent)
As expected, the urban-rural income gap falls when these improved rural site
prices are used instead of provincial-capital prices (Figure 3). Over the 20002012 period, the urban-rural real income gap is 17 percent lower than the nominal
gap, and higher urban prices account for around 6.1 percent of the urban-rural
nominal income gap. That is, the average urban-rural cost of living difference/
average urban-rural nominal income gap = 17 percent/52 percent = 33 percent.
The Oaxaca decomposition offers another way to assess how much cost of living
differences account for the urban-rural gap. Using individual household income
data for 2000-2012, the decomposition tells us that about P38,000, or 34 percent
of the P110,000 difference, is explained by cost of living differences (Table 3), a
figure very close to the 33 percent using descriptive statistics.
TABLE 3. How much of the urban-rural income gap is explained by
the cost of living differences?
Explanatory variable: prices
Average HH income
Urban
229,999.3
Rural
120,342.8
Difference
109,656.5
Decomposition
Explained
37,734.9
Unexplained
71,921.6
Percent Explained
34.4
9
3.2. Getting the wages right
What happens to the 34 percent nominal urban-rural wage gap when average
wages (mean wage in urban barangays versus mean wage in rural barangays) are
broken down by skilled or unskilled wage categories? The urban-rural nominal
wage gap for unskilled labor is 31 percent; the real wage gap is 6 percent. The
urban-rural nominal wage gap for skilled labor is 17 percent; the real wage gap
is nil. Both figures are below the 34 percent gap for the average wage (Figures 4
and 5).9 Presumably, the measured real wage gap would decline even further if
we could disaggregate jobs in greater detail. In short, the reported nominal urbanrural wage gap falls significantly when prices and wages are measured more
accurately. Also, note that the skilled labor urban-rural wage gap declined over
the past few years, while the unskilled wage gap remained constant (Figure 6).
One reason for the recent decline in the skilled wage gap could be the dispersion
of business process outsourcing firms from Manila to secondary cities like Cebu,
Bacolod, Iloilo, and Davao.10
700
600
P per day
500
400
300
86 or 31%
200
100
277
}
190
0
urban
rural
Source: Labor Force Survey
FIGURE 4. Average nominal rural and urban wages in 2011
(skilled, in pesos per day)
Note that we predicted the lower real wage gap for skilled workers, appealing to the notion that they were
better able to finance the move.
10
In 2007, 82 percent of the full-time business process outsourcing workers were located in Manila.
However, the share fell to 75 percent in 2013.
9
10
700
600
75 or 12%
P per day
500
}
400
300
629
554
200
100
0
urban
rural
FIGURE 5. Average nominal rural and urban wages in 2011 (skilled)
50
45
40
35
30
25
20
15
10
5
0
2001 2002 2003 2004 2001 2006 2007 2008 2009 2010
unskilled gap
skilled gap
FIGURE 6. Urban-rural skilled and unskilled ages gaps.
(Average nominal rural and urban wages)
Finally, we can find no evidence of a Todaro effect in the Philippines.11 That
is, underemployment plays no role in accounting for these wage gaps. Wages
adjusted for underemployment—the daily wage divided by 8, multiplied by
Strictly speaking, the Todaro effect [1971] means that rural wages would be closer to urban wages if we
took (allegedly higher) urban unemployment into account. Here, we use underemployment to adjust wages
and loosely refer to it as the Todaro effect.
11
11
normal hours worked per day (only for underemployed workers working less
than 8 hours a day)—differ from unadjusted wages only by about 0.5 percent. Of
course, it’s possible that the explanation for the small difference is that workers
already factor in underemployment when reporting their daily wages during the
survey interview. Since the Todaro-adjusted wage is not statistically different
from the unadjusted wage, the latter is used in all the regressions that follow.
4. What explains the Philippine wage gap?
As we have seen, urban-rural wage gaps diminish greatly when computed by
skill and properly deflated by location-specific cost-of-living. What accounts for
gaps when properly measured?
This section offers a more explicit statistical explanation of the wage gap
using the Oaxaca and Fields decomposition methods. The period covered is 20012009. The unit of observation goes down to the individual and household level,
and it is grouped by urban or rural location, which is in turn based on barangay
characteristics. Individual attributes like education, age, and wages by occupation
or skill are available for all years. We also use provincial variables, although
they are available only triennially (i.e., 2003, 2006, and 2009). Our main data
sources are the quarterly Labor Force Survey and the triennially Family Income
and Expenditure Survey. Education, health amenities, and outcome variables
are sourced from government administrative data. Poverty and Gini statistics
are computed from the Family Income and Expenditure Survey. Infrastructure,
political, and geographic data are sourced from the Asia Pacific Policy Center
provincial data set. The Oaxaca decomposition uses the estimated coefficients
from two separate regressions—one for rural barangays and one for urban (and
for other group categories, respectively)—to account for the difference between
the average urban-rural wage gaps (unadjusted by our new rural prices). The
difference is broken down into the explained and unexplained components.
There is reason to believe that the unexplained residuals can in large part be
attributed to undocumented (positive) attributes of city immigrants [Young 2013].
The regressions are pooled Ordinary Least Squares. Multi-collinearity often
results when all the provincial-level variables are used at the same time, but the
regressions exploited below are limited to those that are free of multi-collinearity.
To better understand the contribution of these variables to the wage gap,
we implement the Oaxaca and Fields decompositions. As noted above, the
regressions underlying the decompositions are pooled Ordinary Least Squares,
where year and region are added as controls, and the results are presented in
Table 4.12 As expected, wages are higher on average for urban, skilled, and large
Pooled Ordinary Least Squares with year controls, and year and occupation controls, were also run. The
results are broadly similar to the regressions used below. The regressions using occupation controls yield
higher R2.
12
12
industry workers compared to their counterparts. Higher education, age, large firm
employment, road density, and the regional minimum wage are also associated
with higher wages. The last column in Table 4 uses the Fields decomposition
index (the share of the total variance attributable to each explanatory variable in
percent) to assess the importance of each variable. The most important variables
are skilled work (23.3 percent), large industry employment (3 percent), college
education (50.9 percent), age as a proxy for experience (11.5 percent), and
provincial road density (10.4 percent).
TABLE 4. Explaining nominal wages: individual pooled data – Ordinary Least
Squares regressions
Dependent variable: ln(wage)
(2)
(3)
(4)
(4) Fields
Decomposition
0.0544*** 0.0514*** 0.0600*** 0.0500***
1.2
0.4186*** 0.4189*** 0.4185*** 0.4188***
23.3
0.1577*** 0.1577*** 0.1581*** 0.1576***
9.8
(1)
Urban
Skilled
Large industry
Sector of employment
Industry
Services
Individual Characteristics
Educational attainment
Elementary undergraduate
Elementary graduate
High school undergraduate
High school graduate
College undergraduate
College graduate
Age
Age2
Provincial characteristics
Student to room ratio
Number of doctors
Remittance to HH income ratio
Domestic remittance to HH inc ratio
Distance from Manila
Gini coefficient
Poverty incidence
LGU income (from local sources)
Total road density
Access to electricity
Access to safe water
Political dynasty index
Conflict incidence
Frequency of typhoons
Prices
Min wage
Constant
0.2501*** 0.2512*** 0.2529*** 0.2497***
0.0186*** 0.0195*** 0.0204*** 0.0185***
0.1424***
0.1786***
0.2493***
0.3419***
0.5033***
0.8291***
0.0417***
-0.0004***
0.1406***
0.1765***
0.2466***
0.3392***
0.5004***
0.8262***
0.0418***
-0.0004***
-.29167*** -1.8014***
0 0.0002**
-0.0012*
-0.0019*
0 -0.0000*
-0.1418***
-0.0021***
0.0076***
-0.0006
0.2057***
0.0004*
-0.0002**
0
0
0
-0.0043*** -0.0044***
0.0021
-0.0023
-0.0001
0.0001
0.0006*** 0.0011***
3.3875*** 3.5585***
0.1404***
0.1764***
0.2468***
0.3395***
0.5007***
0.8266***
0.0418***
-0.0004***
3.0
0.5
0.1419***
0.1780***
0.2483***
0.3409***
0.5022***
0.8280***
0.0417***
-0.0004***
-2.9
-3.1
-3.6
-3.0
4.0
46.9
29.9
-18.4
-2.5462*** -2.2050***
0.0001
0.0001
-0.0017**
-0.0035***
-0.0000*
0
-0.1136***
-0.0015***
0.0059**
0.0022
0.1559***
0.0014***
0
-0.0001
-0.0002
0
0
-0.0051*** -0.0040***
-0.0034
0.002
-0.0002
0.0001
0.0011*** 0.0007***
3.2935*** 3.5501***
0.9
-9.7
-0.4
1.9
0.5
2.1
-0.4
10.4
1.7
0.0
0.0
0.0
0.3
1.0
4.1
(1)
R2
Observations
Controls
Time
Region
0.54
272150
Y
Y
13
Dependent variable: ln(wage)
(2)
(3)
(4)
(4) Fields
Decomposition
0.54
0.54
0.54
272150
272150
272150
Y
Y
Y
Y
Y
Y
In addition, bivariate regressions of the log nominal wage with provincial level
variables reveal that only the following are significant and have the expected signs:
student to classroom ratio; grade 6 test scores; percentage of children who are fully
immunized; distance of provincial capital from Manila; Gini coefficient; poverty
incidence; local government revenue; road density; proportion of households
with electricity; access to safe water; political dynasty; conflict; and number
of typhoons per year. When these provincial-level environmental attributes are
added to the regression equation reported in Table 4, students per classroom (a
school quality proxy), remittances, distance of provincial capital from Manila,
the Gini coefficient, poverty incidence, local government income, road density,
and incidence of conflict are all significant and have the expected signs. However,
except for road density, their explanatory power is very small since their presence
raises only slightly and each Fields statistic is small. These results suggest that
initial location conditions (e.g., poverty, inequality, typhoon incidence, size of the
provincial economy, distance from Manila), location amenities (e.g., classroom
size, road infrastructure, health facilities)13, and location remittance experience all
influence barangay wage levels, but not by much.
Most importantly, after controlling for all these provincial and individual
attributes, urban location has little else to offer (1.2 percent Fields effect).
What contribute most to the urban-rural average wage gap are the individual
characteristics relevant to high-wage firms, like schooling, skills, and experience.14
However, if provincial characteristics influence individual characteristics, then
these decompositions understate the importance of location-specific humancapital-building infrastructure.
The Oaxaca decomposition in Table 5 confirms these results. Five groups are
defined for the wage decompositions: urban-rural for all workers; urban-rural
for unskilled workers only; skilled-unskilled15; formal-informal16; and workers
See Albouy [2008] for a comprehensive assessment of the impact of amenities and quality of life on
wages across metropolitan areas in developed countries.
14
How skills determine city competitiveness is explored at length in Glaeser and Mare [2001].
15
Occupation codes from the Philippine Standard Occupational Classification Code and their respective
average wages were used to determine which occupations are skilled or unskilled.
16
Here we follow the definition in the 2013 Philippine Development Report, “Creating More and Better
Jobs”, where those working in agriculture, transport, communications, wholesale trade, and retail trade are
all treated as informal workers.
13
14
in large and small industries.17 Table 5 reports those underlying regressions: the
variables included in the decomposition explain 93 percent of the urban-rural
wage gap. These variables also explain 46 percent of the skilled-unskilled wage
gap, 77 percent of the formal-informal wage gap, and 68 percent of the largesmall industry wage gap.
TABLE 5. Individual pooled data: Oaxaca decomposition of wages for 5 groups
Urban rural
Educ only
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Age only
Educ and
age only
Provincial
variables only
40.4
41.8
54.1
58.4
32.8
84.0
Educ and
age only
Provincial
variables only
129.9
151.5
45.4
9.6
271.1
3.4
Educ and
age only
Provincial
variables only
68.0
34.3
68.5
18.2
81.1
17.6
279.1
157.9
91.2
84.9
6.3
93.1
45.5
45.7
49.9
5.3
85.9
5.8
Skilled-Unskilled
Educ only
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Age only
469.1
165.4
263.7
129.1
151.5
45.5
116.1
167.5
40.9
-4.8
269.4
5.0
Formal-Informal
Educ only
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Age only
278.7
178.4
102.3
78.7
23.6
78.9
58.4
43.9
57.1
12.2
90.1
12.0
The distinction between small and large industry is used as a proxy for labor policy, in which large firms
are more likely to be affected by minimum wage and contractual terms. The distinction is derived using the
2008 Informal Sector Survey. Industries in which informal workers comprise at least 90 percent are defined
as small.
17
15
Urban-Rural Unskilled
Educ only
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Age only
Educ and
age only
Provincial
variables only
23.3
36.4
39.0
40.4
10.3
82.7
Educ and
age only
Provincial
variables only
104.7
62.9
68.5
13.7
153.9
8.2
210.5
150.9
58.7
58.2
1.5
97.4
20.1
39.7
33.6
3.4
56.3
5.7
Large-Small Industry
Educ only
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Age only
331.6
164.0
167.6
114.5
53.1
66.3
96.9
70.7
5.8
10.5
157.1
6.3
5. What explains the Philippine income gap?
Finally, we apply the Oaxaca decomposition to total household income from
2000 to 2009 using the same regression model and variables that we did for wages.
To repeat our earlier finding, about P38,000 of the P110,000 urban-rural nominal
income gap, or 34 percent, is explained by cost of living differences. Table 6 deals
with the real income gap. Since the individual and provincial variables exhibit
much the same statistical impact on income gaps as they do on wage gaps, we
do not report their detail here in the text except to repeat that education and
experience are the individual variables that dominate: by themselves they explain
71 percent of the household total income gap.
However, one income gap finding does emerge which contrasts dramatically
with the wage gap findings: provincial environmental variables have a much
bigger impact—50 percent of household income gaps are explained by provincial
variables alone. This result strongly suggests that provincial variables have a
powerful effect on labor participation rates, skill mix, occupation mix, property
incomes, and remittances; they have little impact on wage rates themselves.
16
TABLE 6. Oaxaca Decomposition of urban-rural
real total household income gaps
All
Average wage
Urban
Rural
Difference
Deocmposition
Explained
Unexplained
Percent Explained
Educ
only
Age only
Educ and Provincial
age only variables only
178,896.0
115,311.8
63,584.2
66,985.1
0.0
100.0
41,782.9
21,801.4
65.7
0.0
63,584.2
0.0
45,083.2
18,501.0
70.9
31,577.9
32,006.3
49.7
6. Understanding spatial wage and income gaps: thinking in
general equilibrium
Income inequality is higher in the Philippines than in most of its Asian
neighbors, and spatial inequality accounts for a fairly large share of it. So what
accounts for the large urban-rural wage and income gaps in the Philippines?
Individual attributes of workers and households explain the majority of the gaps,
and schooling, skill, and experience are the three individual characteristics that
matter most. Provincial variables like typhoon incidence, government corruption,
school crowding, and access to health facilities have only a modest impact on
urban-rural wage gaps, but they have a powerful impact on urban-rural income
gaps. Workers born in the cities and immigrants to the cities invest more in human
capital than do rural workers, but this paper cannot tell us why. That is, it cannot
say how much of this is due to better human-capital-building infrastructure
supply and how much is due to individual demand for that infrastructure, and, if
the latter, how much is due to market incentives.
Many questions remain unanswered. First, we have no direct evidence that
explains why the urban-rural skilled wage gap fell substantially in the last few
years of our sample, whereas the equivalent gap for unskilled wages remained
essentially constant. To the extent that this has been driven primarily by more
abundant skills in Metro Manila and contiguous regions, it implies that firms
have found the location attractive. But it also suggests that opportunities have
emerged for businesses to expand in rural areas where their added demand for
skilled workers has helped close the wage gap. An investigation into the specific
changes in policies, markets, or institutions that led to this result might suggest
strategies that would further improve labor market integration between urban and
rural areas, thus reducing spatial inequality.
Second, we do not yet know why secondary cities have relatively low skilled
wages (Figure 7). This could be due to a shortage of investment opportunities for
17
skill-intensive firms in secondary cities. If so, we need to know more about the
causes of the shortage.
700
}
500
PHP per day
}
159 or
23%
600
112 or
16%
400
648.1
300
549.4
495.2
200
100
0
NCR
11 other metros
elsewhere
FIGURE 7. Average nominal wages in 2011(skilled)
Third, we have shown that unemployment and underemployment account
for very little of the urban-rural income and wage gaps, either for skilled
labor, unskilled labor, or the average of the two. This is puzzling given that
unemployment rates rise with education in the Philippines (Figure 8). We
have also shown that those large urban-rural income gaps are driven mostly by
education and experience. So why doesn’t an excess supply of educated workers
in the cities lead skill-intensive firms to expand employment, thus driving down
both the income and the wage gaps? Are skilled jobs rationed with their rates of
pay sticky downwards?
Fourth, the relative stability of the unskilled urban-rural gap over the last
decade must be explained by one or both of two forces: unskilled workers may
not be exploiting the gap with much vigor; and firms are not exploiting it with
much vigor either. Do constraints on worker mobility matter, such as poverty
making investment in the move difficult, high minimum wages, labor regulations,
or union restrictions? Are there barriers in capital markets that do not make it
profitable for firms to invest outside well-established urban areas? And what
about land markets?
18
FIGURE 8. Unemployment by educational attainment
Joseph Capuno [2010] reported strong evidence of systematic spatial clustering
of land values in the Philippines. That is, land prices tend to converge only within
certain regions, especially in those areas of major population density such as the
Metro Manila region. Capuno [2012:20] could only find evidence of statistically
significant positive spatial autocorrelation for Metro Manila and three neighboring
provinces; otherwise, he could observe no significant correlations in land prices.
Moreover, he [2012:21] could find no evidence of spatial autocorrelation when
considering the role of distance in land price integration forces.
It seems to us that the lack of wage and land price convergence across regions
in the Philippines is not caused primarily by distance or geography but rather
by the fact that different jurisdictions continue to have significant administrative
autonomy, which raises barriers to market integration. These institutions are
consistent with continued spatial segmentation of land and capital markets. Lack
of labor market integration might, at least partially, be explained by mobility
restrictions on firms, but also by constraints on land use.
Given the judicial and administrative constraints on the spatial mobility
of firms and their land use, Philippine workers do an excellent job of sorting
themselves to take advantage of the spatial opportunities. The fact that education,
age, experience, and other individual characteristics remove much of the observed
urban-rural wage gaps suggests fairly efficient labor market adjustment. The fact
that firms do not locate or expand in areas with cheaper land and labor, or that
trade does not promote national integration, seems to suggest that capital and
19
land market failure offer a better way to understand persistent urban-rural income
gaps than does labor market failure. The persistence of spatial income gaps in the
Philippines has much more to do with unequal spatial endowments than labor
market failure. And the fact that individual endowments explain so much of
spatial inequality in the Philippines suggests that local political economy forces
are critical since they determine the quality of the education-health infrastructure.
Given the large role that spatial infrastructure inequalities have in promoting the
rural-urban gap, government programs to ameliorate these differences should
come in the following areas:
• Promotion of the development of secondary urban centers outside of the
national capital region;
• Improvement of the investment climate, which may constrain companies from
expanding investment in existing urban areas that would be able to attract
more rural unskilled workers and also retain many of the skilled workers who
would otherwise work abroad as overseas Filipino workers;
• Improvement of infrastructure and transportation between areas identified as
weakly integrated despite geographical proximity;
• Reconsideration of the current rules on the limited and highly politicized
rezoning of agricultural lands to commercial so as to permit more investment;
• Relaxation of labor market rules that might hamper expansion of employment
in existing businesses; and most importantly
• Promotion by the central government of programs that might serve to level the
education-health infrastructure playing field between rural and urban areas.
*International Bank for Reconstruction and Development
**George Mason University and
the Higher School of Economics in Moscow
***Harvard University, the University of the Philippines,
and the University of Wisconsin
The authors gratefully acknowledge comments by and help with the data from Arsi
Balisacan, Roger van den Brink, and Timothy Hatton. Most importantly, we thank Hal
Hill for his years of support of research on the Philippines, including the research for
this paper.
20
References
Akita, T. [2003] “Decomposing regional income inequality in China and
Indonesia using two-stage nested Theil decomposition method”, The Annals of
Regional Science 37(1); 55-77.
Albouy, D. [2008] “Are big cities bad places to live? Estimating quality of life
across metropolitan areas”, nber Working Paper 14472, National Bureau of
Economic Research.
Balisacan, A.M. [1999] “Reconstructing poverty profiles in the Philippines”,
University of the Philippines School of Economics Discussion Paper, 9915.
Balisacan, A.M. and N. Fuwa [2004] “Changes in spatial income inequality in the
Philippines: an exploratory analysis”, Research Paper 2004/34, unu-wider,
United Nations University.
Berry, A. [1978] “Income and consumption distribution trends in the Philippines,
1950-70”, Review of Income and Wealth 24(3): 313-331.
Cameron, S., T. Merridew, and F. Pichler [2006] First European quality of life
survey: urban-rural differences. Luxembourg: Office for Official Publications
of the European Communities.
Capuno, J.J. [2012] “Agglomeration and sub-regional disparities under
decentralization: evidence of spatial clustering of land values in the
Philippines”, Review of Urban and Regional Development 24(3): 106-120.
de Dios, E.S. and J.G. Williamson [2014] “Has the Philippine forever lost its
chance at industrialization?”, Philippine Review of Economics 51(2): 47-66.
Förster, M. and D.E.M. Mira [2005] “Income distribution and poverty in oecd
countries in the second half of the 1990s”, oecd Social, Employment and
Migration Working Papers 22:1-79.
Glaeser, E. and D. Mare [2001] “Cities and skills”, Journal of Labor Economics
19:316.
Gould, E.D. [2007] “Cities, workers, and wages: a structural analysis of the urban
wage premium”, The Review of Economic Studies 74(2): 477-506.
Gray, D., J.A. Mills, and S. Zandvakili [2003] “Statistical analysis of inequality
with decompositions: the Canadian experience”, Empirical Economics 28(2):
291-302.
Hatton, T.J. and J.G. Williamson [1992] “What explains wage gaps between farm
and city? Exploring the Todaro model with American evidence, 1890-1941”,
Economic Development and Cultural Change 40(2): 267-294.
Hertz, T., P. Winters, A. P. de la O Campos, E. J. Quiñones, B. Davis, and A.
Zezza [2008] “Wage inequality in international perspective: effects of
location, sector, and gender”, esa Working Paper 8/08. Rome, Italy: Food and
Agriculture Organization of the United Nations, Agricultural and Development
Economics Division.
Karan, A. K. and S. Selvaraj [2008] “Trends in wages and earnings in India:
increasing wage differentials in a segmented labour market”, Asia-Pacific
21
Working Paper Series, International Labour Organisation Sub-regional Office
for South Asia, New Delhi.
Minot, N., B. Baulch, and M. Epprecht. [2003] Poverty and inequality in Vietnam:
spatial patterns and geographic determinants. Washington, dc: International
Food Policy Research Institute.
Mishra, P. and A. Parikh [1992] “Household consumer expenditure inequalities
in India: a decomposition analysis”, Review of Income and Wealth 38(2): 225236.
Morris, E. [2007] Promoting employment in Cambodia: Analysis and options.
International Labour Organization, ilo Subregional Office for East Asia.
National Statistics Office (nso) [2012] Labor Force Survey (2001-2011) data file.
National Statistics Office (nso) [2012] Family Income and Expenditure Survey
(2003, 2006, and 2009) data file.
Novotný, J. [2007] “On the measurement of regional inequality: does spatial
dimension of income inequality matter?”, The Annals of Regional Science
41(3): 563-580.
Ragayah, H.M.Z. [2008] “Income Inequality in Malaysia”, Asian Economic
Policy Review 3(1): 114-132.
Sicular, T., Y. Ximing, B. Gustafsson, and L. Shi [2007] “The urban-rural income
gap and inequality in China”, Review of Income and Wealth 53(1): 93-126.
Simler, K.R. and V. Nhate [2005] “Poverty, inequality, and geographic targeting:
evidence from small-area estimates in Mozambique”, International Food
Policy Research Institute, fcnd Discussion Paper, 192.
Tadjoeddin, M.Z., W.I. Suharyo, and S. Mishra [2003] “Aspiration to inequality:
regional disparity and centre-regional conflicts in Indonesia”, In uni/wider
Project Conference on Spatial Inequality in Asia, Tokyo, Japan (pages 28-29).
Todaro, M.P. [1971] “Income expectations, rural-urban migration, and
employment in Africa”, International Labour Review 104(5).
Weber, A.F. [1899] The growth of cities in the nineteenth century: A study in
statistics (Vol. 11). Published for Columbia University by Macmillan.
Williamson, J. G. [2013] “Demographic dividends revisited”, Asian Development
Review 30(2): 1-25.
World Bank (wb) [2013] Philippine development report: Creating more and
better jobs. Philippines: Asia-Pacific region.
Yankow, J.J. [2006] “Why do cities pay more? An empirical examination of some
competing theories of the urban wage premium”, Journal of Urban Economics
60(2): 139-161.
Young, A. [2013] “Inequality, the urban-rural gap, and migration”, Quarterly
Journal of Economics 128(4): 1727-1785.
PRE
American private direct investment in the Philippines
after independence
Gerardo P. Sicat
The Philippines has lagged behind East and Southeast Asian
economic growth successes in the course of contemporary times.
A historical perspective is presented to explain this lag. When
the country entered the definite road to independence from
American rule in 1934, it adopted a political Constitution that
limited foreign capital to a minority role in three industries – land,
public utilities, and the exploitation of natural resources. During
the colonial period and after political independence in 1946,
American foreign direct investment in the country was dominant
and large. This continued after independence as US citizens
were granted equal rights of Filipinos and American businesses
received war damage payments to rehabilitate their investments.
This was further helped by bilateral preferential trade with the
United States during twenty-eight years of economic adjustment.
In actual practice, the restrictive provisions in the Constitution
invited many virulent forms of nationalistic economic policies
of exclusions of foreigners in other lines of economic activity
and led to inward-looking economic development policies.
These policies discouraged American foreign direct investments
causing disinvestment in key areas. The restrictive economic
provisions in the Constitution continue to this day even as the
economy is opening more to world markets.
JEL classification: P45, D9
Keywords: Philippine economic history and development, foreign direct investment,
American foreign direct investment, Philippine-American economic
relations
In Hal Hill’s work on the Philippine economy, he has called attention to the
following conundrum: Why does the Philippines, despite its many good attributes,
lag behind in economic development when compared to East Asian high economic
23
achievers?1 In this essay, I hark back to what I have called the “original sin”,
the biggest long-term mistake in our nation-building: the restrictive economic
provisions in the Philippine Constitution.2
It is not as much a mystery. When we take a historical perspective, we link
developments to the question of why American and other foreign direct investments
have not come in sufficient volume to help push the growth of industry, incomes,
employment and productivity. Those restrictive provisions led to protective
and inward-looking policies that characterized subsequent policies promoting
industrialization and economic development. In that respect, they served as a lever
for bad policies that moved the country in many wrong directions.
In turn, because those provisions further strengthened the vested interest in
the continuation of these policies, their reform could not take hold. Today, the
problem of amending these provisions in the Constitution remains as a principal
problem hindering still future progress.
Before the Second World War, United States private industrial capital in the
Philippines was dominant. The public social and economic infrastructure built by
the colonial administration was extensive and growing. The country was the envy
of early economic development in East Asia among colonized economies. When
the grant of full political independence in 1946 happened immediately in the
aftermath of a ruinous war of occupation by Japan, the country was economically
devastated. However, the rehabilitation from that devastation was relatively quick.
Total us direct investments3 in the world at yearend in 2012 is estimated at
us$4.45 trillion. Three-quarters (74 percent) of this total is located in highly
advanced countries. Only 14 percent of American direct investments is located
in Asia, principally also in the developed countries of the region. In this context,
direct investments to the Philippines were hardly a speck in the numbers. A recent
work summarizing data on flows of foreign direct investments (including us direct
investments) going to Southeast Asian countries from 1970 to 2011 reports that
Singapore, Thailand, Indonesia, and Malaysia have been the largest receivers of
foreign direct investments. It also observes that the Philippines has consistently
received a relatively small amount of these investment flows over the same period.4
To those drawn into the discussion on investment and economic policy in the
Philippines, the relative insignificance of foreign direct investment numbers is all
too familiar.
Hal Hill expounded on this point in seminars he delivered at the University of the Philippines School
of Economics. His work on the Philippines is sometimes indistinguishable with that of Arsenio Balisacan
because they have worked together in major pieces.
2
The author tried to elaborate on this point in several short essays found in Sicat [2013:186-187]. In a
presentation on the topic of restrictive economic provisions on foreign capital before an annual meeting of
the Philippine Economic Society circa 2004, he called these provisions the “ original sin” of development
policy.
3
See Jackson [2013].
4
See Sjöholm [2013].
1
24
Sicat: American private direct investment
in the Philippines after independence
1. The triad of us-Philippine relations
The Philippines was one of the most devastated economies at the end of World
War II. War confiscations and conscriptions, market disruptions, inflationary spiral
especially towards the closing months, destruction due to the battle for liberation—
all these led to output and physical capital losses that were substantial. The total
output of the country at the end of the war in 1945, was reduced to 30 percent of
the level of the pre-war gdp.5
Before the start of the World War II in the Pacific, the preparations for
Philippine independence were in full swing. In 1934, the us Congress passed
the law that spelled the road toward full political independence. A year later, the
political Constitution (which had built into it the restrictive economic provisions on
foreign investments) that was crafted by an elected national convention composed
of Filipino leaders was approved by the American president as required. Hence,
in 1936, the ten-year transitional Philippine Commonwealth took office. The
Commonwealth stage was a semi-independent political status under the American
dominion before the grant of full independence on July 4, 1946 as the Republic of
the Philippines. This path toward independence was interrupted by a destructive
war of occupation by Japan covering four years.
Economic rehabilitation from the war proceeded rapidly during the next
decade after independence, starting in 1945 when Japan’s occupation ended. Aside
from a sustained volume of American military expenditures early in this period,
three major actions propped up the economy. Such measures would involve deep
economic entanglements of the Philippines with its former colonial ruler. These
actions—the triad of relations—were to influence future events strongly.
First was a trade and economic adjustment to wean the country from its
dependence on the us economy, an outcome of the long American occupation of
the country. The us Congress enacted the Philippine Trade Act of 1946 (us Public
Law 371, 79th Congress) to amend its trade and tariff policies to accommodate
the new parameters that would be contained in the commercial trade agreement
with its former colony. The law granted the Philippines a period of declining trade
preferences and fixed quotas on certain products like sugar, coconut oil, and other
major products that would end in 1973. In the early discussions of this adjustment
period before the war, the proposed adjustment period was shorter, twenty years
being the longest. The war destruction in part led to this extension of the period of
adjustment to twenty-eight years.
Second was the payment of war damage transfers6 to assist in speeding up the
rehabilitation of the domestic economy. The war damage payments were embedded
See Sicat [2015:191-204].
The author uses the word “transfers” in contrast to war damage payments as reparations, which connotes
monetary retribution as payment made by a defeated party after the war, as in the Japanese reparations. The
us considered the payment of war damage claims (which had to be formally declared first) as assistance to
help the recovery effort, more in the context of the Marshall Plan for Europe after the war.
5
6
25
in the law called the Philippine Rehabilitation Act (Public Law 370, 79th Congress),
which contained the war damage payments and the mechanisms for delivering the
program. This companion law on Philippine independence appropriated a large
sum of money—us$520 million—in two parts: us$120 million to help finance
and restore public buildings, roads, and social infrastructure; and us$400 million
to pay war damage compensation to assist in the rehabilitation of the economy that
heavily suffered during the war.
Finally, the American government required that before these new provisions were
to take effect, the new Philippine republic had to give Americans the same rights
as Filipino citizens. This meant in polite treaty language, “national treatment”, a
common feature of commercial agreements between states. Essentially, this meant
protection from confiscatory national policies and guarantee from discriminatory
policies. The restrictive provisions in fact were directed at foreign control of the
economy, which was dominated by American investments at the time. President
Franklin Delano Roosevelt approved the Philippine Constitution as prelude to the
process leading to Philippine independence early in his first term in office at the
bottom of the Great Depression in 1934. Faced with more pressing economic issues
at home, he approved the restrictive provisions in the Philippine Constitution. It
therefore would fall on Harry S. Truman (the successor to the presidency upon
Roosevelt’s death in 1944) to exact parity rights as a condition to protect American
investor interests after Philippine independence. For the Philippines to yield to
this demand, a constitutional amendment had to be secured by plebiscite. The
plebiscite was made and the government of Manuel A. Roxas, the first president of
the Republic, triumphed in his sponsorship of the yes vote.7
The 1935 Constitution limited foreign capital participation in three economic
markets—utilities, land, and exploitation of natural resources—to a maximum of
40 percent, meaning that such industries must have 60 percent Filipino control of
the capital equity. This was a key intent of the “nationalistic” provisions of the
Philippine Constitution when it was enacted in 1935 in reference to corporate
activities. Had the basic law of the land simply remained silent on this issue, it
would still have been possible to enact nationalistic provisions, but they would
not need to be cast in solid stone to accommodate changes in national moods
and perception.
Thus, a package of three important economic actions was needed before the
road toward full economic rehabilitation and adjustment of the newly independent
republic could proceed. To nationalistic Filipino leaders, the grant of equal
rights to Americans was economic blackmail. To the Americans, these actions
were pragmatic elements to help a former colony recover as well as to protect
the rights of its citizens who continued to have vital interest in the Philippines.
To the incumbent, newly elected leaders of the Philippine government (led by
The parity rights amendment was held on March 11, 1947 with 79 percent of voters casting approval.
7
26
President Manuel A. Roxas, who was then the newly inaugurated first president
of the republic), this was a realistic compromise to ease the way for the newly
independent country.
Another element of relations with the us at the inception of the Philippine
republic was the military bases agreement. This could be considered the fourth
element of the post-independence conditions with the former colonial master. The
us was given a military foothold for 99 years to operate military bases in the
country. Two major military bases were principally included in this agreement,
the Clark Air Base and the Subic Naval Base, in addition to a number of military
installations that were installed during the long American colonial period.
From an economic viewpoint, this agreement helped to provide for the country
an important American military shield. A steady flow of earnings of foreign
exchange resulted from us military operational spending in the bases throughout
the post-war period when the military bases were stationed in the country. From a
defense standpoint, the newly independent country was kept from undertaking a
heavy domestic expenditure on defense matters. Defense expenditure was a major
expense among newly independent countries.
2. The war damage payments by citizenship of recipients
The war damage payments made to American citizens give a clue to the
amount of American private capital in the Philippines. The losses were therefore
compensated at the original value of acquisition. When the payments were made,
the replacement cost of capital was already three times the cost of capital before
1940. They were disbursed in pesos at the fixed exchange rate of 2 pesos to
a dollar.
About two-thirds of the war damage compensation was to pay war damage
to the private sector on the basis of claims of war damage filed before the Joint
Philippine-American War Damage Commission8 (henceforth referred to as the
Commission). The first job of the Commission was to pay as much of the small
claims first. Approximately one million claimants, out of a total of 1.25 million
claimants, filed claims for us$500 or less. According to the final report9 of the
Commission, these claimants were paid a total of us$131,766,435 as of end
of December 1950, or, converting the amount at the 2 pesos to one dollar fixed
exchange rate, P263,532,870. The total amount of these small payments amounted
to 32.9 percent of the us$400 million that was appropriated for private payments.
It was safe to assume that all the small payments were received by Filipino citizens.
The Commission had three members: two Americans and a Filipino, with one of the American
commissioners serving as chairman. The activities of the Commission and their actions with respect to
implementing its mandate are explained in the main study.
9
See Philippine War Damage Commission [1951].
8
27
The question of how much war damage compensation was received by
Americans meant that that the citizenship of the recipients had to be known. The law
on the war damage payment was only to qualified American and Filipino citizens
and citizens of other nations that were not enemies, to include also corporations
owned by such citizens. The report of awards did not track the citizenship of
the awardees. In order to make an estimate of awards by citizenship, an effort is
made to classify the citizens that filed claims and who were issued approvals for
those claims.
The task was to classify by citizenship of awardees of the “large” payments,
that is, those exceeding us$500 and above. To get a distribution by citizenship of
all these awards, I had to tag each of the recipients—individuals or enterprises—
with the appropriate citizenship: Filipino, American, Chinese, and the like. The
greatest area of doubt was in determining whether some recipients were Filipino
or American citizens.
The individual recipients of war damage awards could be fairly identified
by their names, although this was not fully accurate. The names of business
establishments were another matter. Most of the establishments were corporations.
From the names of these recipients, I tried to form some basis for the citizenships by
identifying each recipient with a citizenship label. The classification by citizenship
was essentially educated guesswork to determine overall magnitudes that yielded
approximate orders of magnitudes.
I labeled familiar Filipino names “Filipino” by citizenship. American names,
and those I associate with American enterprises, I labeled “American”. Chinese
names were easy to identify when they used proper names. Uncertainty was
greatest in identifying the citizenship of enterprises, for which I could only add
the label “American or Filipino”. These recipients were separated from those
tagged more definitely as either American or Filipino. In all these labeling efforts,
acceptable errors were likely to be made.
The corporate names of the firms were more challenging. There was little to
go by that could give the citizenship of the ownership of corporate enterprises.
Sometimes the corporate names gave a clue. This was clear enough for some
businesses, as when the corporate names were emblazoned with Chinese names
or titles. Most of the time, an educated guess sufficed, for instance, from studies
listing industries and establishments10, from personal and imperfect memory
growing older, and from introspection on the business of the corporation and the
likely citizenship of the owners.
The study of Kunio Yoshihara [1985] on Philippine industrialization used some names of Chinese,
American, and Filipino enterprises during the 1960s. Many of these firms existed before the war. A more
assiduous study would have been to use an establishment set of names from the Securities and Exchange
Commission in an earlier period or some listing from the Bureau of Commerce and Industry. I had no
resources for hiring research assistants, and neither did I have the luxury of time to study this aspect of the
problem. In any case, relative accuracy, not definite accuracy, was what I was after.
10
28
American investments were heavily in sugar centrals, mining, logging and
timber exploitation; in public utilities such as power generation and distribution,
transportation, both land, sea and air; in domestic and international trading; in
manufacturing operations that processed agricultural products; in construction and
metallic fabrication; and in banking and finance. They were typically the larger
industrial enterprises operating in the country, although there were also some
Filipino-owned industrial enterprises in similar fields.
There were many instances when it was difficult to tag the enterprise definitely,
in which case they were identified as “Filipino or American”. Exact judgment
depended on subjective judgment, a dubious but necessary method in order to
establish a reasonable measure of allocation in the absence of total information.
For a norm, that probability could be 50-50. An optimistic judgment that the
group of “American or Filipino” recipients thus tagged contained more Americans
represents “60 percent likelihood of being American”. A less optimistic
assessment on the opposite side could be one in which 60 percent likelihood of
being Filipino”. I use these subjective judgments in providing a “high” and “low”
estimate of American citizen recipients.
Finally, there were awardees whose citizenship did not belong to the three
above-mentioned. These awards were made principally to religious institutions
that received war damage compensation. I labeled most of these “Other” citizens.
Catholic Church institutions could have been grouped under “Filipino”, but they
fall under the hierarchic control of the Vatican. Therefore, all the war damage
payments to archdioceses, bishoprics, and foreign Catholic religious institutions
that received awards were classified as “Other citizens”.
3. Summary of war damage payments by citizenship
Table 1 presents a comprehensive table. Part A is the raw table showing
the citizenships of the large recipients of war damage as approved by the Joint
Philippine-American Commission. The citizenships of recipients separated those
grouped under American, Filipino, “American or Filipino”, Chinese, and Other.
Part B of the table uses the “high” scenario in which American citizens received
47 percent of the large payments made by the US government. In this estimate,
Filipinos received 42 percent. These numbers are seen in the last column under
Part B. The ratio of awards to the amount of claims (seen in the second to last
column) is slightly higher for American awardees, probably not out of bias but by
virtue of the greater reliability of claims made by larger claimants—corporations
with much better record keeping.
Part C of the table shows the “low” scenario result. From the viewpoint of the
immediate impact on the economic rehabilitation efforts, the issue of who received
more of the war damage payments between Filipino and American citizens did
not matter as much. However, this exercise is helpful in identifying the amount of
private American capital in place in the country before the war.
29
There were no reliable estimates of American private investments that I could
find before the war. An estimate due to the American Chamber of Commerce in
the Philippines was P537 million, of which P200 million was in sugar, coconut,
and abaca plantations, mining, and lumber operations. (In us currency, this was
us$268 million at the fixed rate of two pesos to a dollar.11 ) Another estimate of
us$160 million in direct investments and us$36 million held in Philippine bonds
and securities (or nearly us$200 million) was due to a presentation made by the
us National Foreign Trade Association president before a hearing on Philippine
independence issues.12 These were very sizable magnitudes of investment in an
economy that was undergoing steady economic development before the war.
TABLE 1. Summary of US war damage payments (“large awards”),
by citizenship of awardees
Citizenship of awardees
Value of claims
submitted in
thousand 1940
pesos
Amount of war
Ratio of award
damage approved to claims
in thousand 1940
pesos
A. Raw estimates
Filipino
203.02
87.41
43.1%
American
166.46
85.00
51.1%
American or Filipino ("AF")
161.77
85.90
53.1%
Chinese
26.87
7.67
28.5%
Others
50.50
203.02
46.7%
TOTAL
608.63
289.56
47.6%
B. High estimates (Americans 60% of "AF")
Filipino
267.73
121.77
45.5%
American
263.53
136.54
51.8%
Chinese
26.87
7.67
28.5%
Others
50.50
23.59
46.7%
TOTAL
608.63
289.56
47.6%
C. Low estimates (Americas 40% of "AF")
Filipino
300.08
138.95
46.3%
American
231.17
119.36
51.6%
Chinese
26.87
7.67
28.5%
Others
50.50
23.59
46.7%
TOTAL
608.63
289.56
47.6%
D. Estimated high estimates calibrated to 2015 US Dollar (based on Part B estimates)
Filipino
4,015.92
1,826.52
2,007.96
American
3,952.95
2,048.05
1,976.47
Chinese
403.08
115.00
201.54
Others
757.53
353.84
378.77
TOTAL
9,129.48
4,343.40
4,564.74
Percentage
distribution of
awards to total
30.2%
29.4%
29.7%
2.6%
8.1%
100.0%
42.1%
47.2%
2.6%
8.1%
100.0%
48.0%
41.2%
2.6%
8.1%
100.0%
913.26
1,024.02
57.50
176.92
2,171.70
Source: From basic table of war damage payments. G.P. Sicat, “US War Damage Payments Income Flows After
Independence and Philippine Economic Development,” UP School of Economics, forthcoming, 2016.
11
12
See Hartendorp [1953].
See United States [1938].
30
My objective in presenting these estimates is to show that for the economic
rehabilitation during the early years of the Philippine republic, a sizable
amount of American capital was returned to their proper owners. Since the war
damage payments were linked with the rehabilitation of the damaged economic
establishments, this return of American capital also signified the continuing
commitment to the future of the Philippine economy. The war damage payments
also helped to identify orders of magnitude that had a direct connection to the
amount of capital in place before the war.
The Commission in its final report indicated that replacement cost of damaged
capital at the time of payment was three times more than before the war, even
though the approved payments represented only about 52.5 percent of the
submitted claims. When replacement cost was therefore taken into account, the
recipients effectively received only around 20 percent of the value of replacement
costs. The award of war damage payments could therefore act as supplying merely
a significant stimulant for the rehabilitation process. On this point, the report of the
commission said: “They permitted many industries and other activities to attract
new capital or obtain credit to start rebuilding the industrial commercial, and
other agricultural enterprises which provide employment and otherwise benefit
the Philippine economy.”13 Some slippage from this objective or procedure might
have occurred, but most of the payments of war damage were made after proof of
rehabilitation work was made available.
In other words, although the war damage payment might have constituted a
fifth of the total investment requirement to rehabilitate an enterprise, in effect it
performed a multiplier impact on the rehabilitation effort. Such an amount of new
payments would have implied an expenditure multiplier close to 5, but it surely
was much less since the payments would have triggered a high bill of imports of
materials and equipment not available in the domestic economy. The country’s
productive capacity had been rendered very low due to the ravaged productive
capacity after the war.
Finally, a point to be made here is that the amount of money transferred to
private pockets at that critical time was very significant in relative value even
when viewed today. Using the price multiplier for the value of the US dollar in
1940 compared to the much inflated 2015 us dollar, here are some results.14 The
us$520 million valued at the 1940 us dollar (which was allotted for public and
private war damage) is roughly the equivalent of us$7.8 billion in 2015 dollars.
See Philippine War Damage Commission [1951]. The report further stressed that “before second and final
payments were made on claims for which combined payments would amount to us$25,000 or more, the
Commission … communicated with the claimants to make certain that the first payments had been expended
in accordance with their sworn statements and that, second payments would also be used for rehabilitation.”
14
I use an inflator-multiplier of 15 to compare the pre-war us$ to the current us$ for 2015, taking into
account the changes in prices in 1940 to the present in the Philippines. In the paper calculating the war
damage from the Japanese occupation (see footnote 6), I used the same inflator. In this context, these
calculations are relatively conservative.
13
31
Since the amount was converted at the fixed exchange rate of 2 pesos to one dollar,
the equivalent of this sum in 1940 Philippine pesos was P15.6 billion. The amount
of us$400 million which was appropriated for private war damage compensation,
using the same approximations, meant us$6 billion in 2015 us dollars. In the
peso values of 1940, this amounted to P12 billion. (A proper adjustment would
of course also be reckoned in the much depreciated pesos of 2015, which I do
not show.) By whatever measures, these numbers are very significant, especially
as they were made for a newly independent country with a population size of 19
million. (In 2015, the Philippine population is nearing 100 million.)
Part D of Table 1 shows the same calculations for the amounts of payments
made by citizenship of the awardees.15 Following the discussion for the “high”
case estimates of American awardees, we specifically track the amounts paid
(shown in last column of Part D). American citizens received approximately
us$1,024 million, or around us$1 billion in US dollar values of 2015. The total
amount of payments for all awardees was US$2.171 billion measured in 2015 us
dollars. (The peso values of the awards that are calibrated to 2015 us dollars but
are converted at the 2 to 1 dollar exchange rate of the 1940 pesos are shown in
column 3 of Part D.)
4. Three reasons
From the viewpoint of the newly independent nation, the trade adjustment
of twenty-eight years brought about a stable relationship in trade and economic
relations with the strongest economy in the world at an opportune time of
global growth after World War II. Further, parity rights, which gave Americans
the same opportunities as Filipinos, promised to continue attracting American
investments. At the same time, the extraordinary war damage payments provided
new income flows essential for rehabilitation. In particular, as it pertained to
American investments, the income flows awarded to Americans were linked to the
rehabilitation of their damaged businesses in the country.
This question then arises: Why has American capital not come to the country
in droves after the war and during the long interim period to the end of special
relations in 1973? Three reasons come to mind why this has transpired:
• First, the restrictive provisions in the Constitution abetted other policies that
further relegated the role of foreign capital in other parts of the economy,
thereby rendering the development process more difficult and perilous;
• Second, the evolution of American firms in the country reacted to the timebound adjustment process more as a signal to divest rather than to be more fully
engaged in the economy; and
Since I am interested only in understanding the orders of magnitudes for the war damage payments then, I
simply return to the 1940 fixed exchange rate of the us$ to the peso. The average exchange rate for the pesos
to the dollar in 2015 is P46 per us$1. The contemporary pesos would be very much depreciated.
15
32
• Third, inward-looking development strategies were implemented by
industrialization efforts based on import substitution. To implement it the
government undertook nationalist policies that equated promotion of industry
with the exclusion of foreign economic participation.
4.1. Restrictive policies
The restrictive economic provisions were designed to spell out the boundaries
of foreign investment participation in certain economic sectors, implying that
those not so mentioned were to welcome foreign capital. In its place, the legal
language of restrictive provisions provoked measures of economic nationalism
that encouraged exclusion or limited participation rather than suggest the attraction
of foreign capital to speed up development in the rest of the economy. In practice
therefore, the provisions incited other restrictive economic policies.
A wider network of economic policies limiting foreign investments in other
sectors ensued during the early years of political independence. Commercial
banking was closed to foreign participation, except for those already allowed
before. The licenses for new banks were issued only to Filipino capital even
under low capitalization requirements. For many years until the 1970s, no new
foreign banks were allowed to enter the economy, except the four foreign banks
that were in operation at the time of independence. It would take the 1990s,
when more openness to the participation of foreign banks in the country would
be allowed.
When import and foreign exchange controls were instituted in 1949, the
allocation decision became one of administrative selection rather than market
solution.16 In an environment of economic controls, selectivity implied that
resource allocation happened through regulation and administration rather than
through the market. In this climate of policy, industrial and agricultural projects
were the result of selection criteria in which nationality was a critical element in
the choice.
A further consequence of the restrictive economic provisions was that they
encouraged policies that were discriminatory over foreign economic interests.
In 1955, retail trade nationalization was passed by Congress. The new law
placed restrictions that excluded foreigners from the retail distribution sector of
the economy. The measure was designed to encourage Filipino retailers to take
over the small sari-sari corner store from the Chinese merchants. But the retail
distribution network extended to big stores too, and this led to the exclusion of
foreigners from the retail distribution industry.
See Golay [1961] for a longer discussion of nationalistic measures during the first two decades of
independence.
16
33
The rhetoric of industrial and economic policy in the 1950s was “Filipino First”,
echoing the demands of those who wanted more protection for Filipino enterprises
and the exclusion of foreigners from the nation’s economic development. Though
such rhetoric was removed from the official language in the investment promotion
programs, the industrial and investment priority schemes for new industrial
projects emphasized that priority areas of investments were reserved for Filipino
enterprises under the Board of Investment. Only industries classified as “pioneer”
were listed as investment areas where foreign direct investments would be allowed
up to 100 percent equity from foreign enterprises. In many East Asian countries,
especially at the beginning of their industrial promotion efforts, businesses
that were 100 percent owned by foreigners were typically allowed to receive
government incentives in order to attract new foreign capital.
4.2. The divestment of American firms
Before the war, American investments were strong across the spectrum of
economic sectors. Specifically, they were dominant in agriculture (sugar, coconut
processing, abaca, and other plantation agriculture), natural resources extraction,
especially mining and forestry, and in public utilities. Targeted constitutionally
by the restrictions on foreign ownership, the grant of parity rights meant that they
could remain in American hands even after independence. However, the future
prospects for these investments in spite of parity rights became bleak in the face of
nationalistic regulations and policies.
The strength of the economic rehabilitation efforts after the war emanated partly
from the participation of American companies that were awarded war damage
payments. Agricultural enterprises, especially in sugar and coconut industries,
were able to recover easily and enabled export earnings to begin to recover. These
were big contributors to the export sector. The same story applied to the mining
companies that resumed their operations after the war. The re-entry of mining
firms into the production picture helped the mining industry recover.
The public utilities were among the biggest recipients of war damage
payments and their restoration of service was quick. Meralco—the electricity
retail distribution monopoly in the Manila area—was one of the biggest recipients
of war damage. The company was able to undertake resumption of service and
to re-expand destroyed facilities. The telephone and telecommunications sector
re-established services. These utilities were most exposed to dealings with the
local regulators.
By the late 1960s, American companies in the country began to feel the pinch.
Nationalistic economic policies and regulations foreshadowed future difficulties
for them, and so it was better to move ahead of 1974. Even if parity rights gave
American investors some comfort as they faced the future, it was a kind of
restless comfort. As the period of special economic relations was coming to a
close at the historical time line of 1974, the end of parity rights was written on
34
the wall. Eventually, the public utilities saw the future as bleak for full control
of management, and they sold to partners toward the years before the end of
parity rights.
During this period then, the incentive to reinvest and to expand existing
capacities began to weaken as more restrictive economic measures were added
to those already in place. Thus, the stream of divestments began. Some American
investors sold the land they owned to other Filipinos. Companies that had a desire
to remain in business accepted the implications on land policies. They created a
land company that was owned by their workers’ provident funds to which they
sold the land. In turn, the land company that was owned by at least 60 percent by
Filipinos leased back the land for the company’s use. This roundabout method of
dealing with the land question made investors essentially temporarily satisfied, but
it added complications to the company’s structure and, hence, its decision making.
Even this solution did not please those that wanted complete divestiture. In the
latter case, the local partners often had the inside track and benefited handily. The
lack of a strong market to buy up these assets meant that the local partners had the
upper hand and bought the assets at favorable prices.
Some American investments found business opportunities in the domestic
economy, and they ventured further into purely local markets. New investments
took place behind the logic of import substitution, but they reflected small-scale
commitments of new capital since the domestic market was limited. Formerly
imported goods had better economic logic being assembled for the domestic
market so as not to lose the market. This was the case for many industrial products
that thrived in food, cosmetics, pharmaceuticals, and energy refining, where the
buyers were home consumers. The end of parity rights was not a problem for
these industries for they had a comfortable position participating in the domestic
market. The same domestic protection given to local industries was accorded
these investments. American companies that had an early foothold before with
imports of these goods established their manufacturing and processing plants in
order to keep their domestic market share. This industrial foothold anticipated a
future in the country’s growing market. But these industries were relatively small
in themselves because they simply undertook production for the local market like
most of the industries promoted by the protectionist industrial policies.
Plantation agriculture in old lines of agriculture continued in part because of
the early economic foundation laid by export industries destined for the American
market. This was the case for the quota-based industries, such as sugar and the
traditional coconut-based processing industries, and in the case of the pineapple
industries, which had natural markets in other countries as well. The banana
industry was such a case, with markets intended for third countries like Japan.
However, the banana industry—which started originally as a proposal to set up an
investment by the United Fruit company—got shot down. It was resuscitated when
domestic interests took over production while they allowed marketing functions
35
to the American investors. Dole Plantation got an earlier foothold, with large land
areas made available through government lease at the beginning. Later expansions
led to innovations in contract farming with local landowners. What has thrived
well was the profitable contract farming arrangements with local planters, which
involves other technical and other production tie-ups and marketing agreements.
In general, because of land reform issues which limited future land use to smaller
plots of land, even plantation agriculture became almost a closed field for
investment expansion.
It was only after the late 1970s that more enthusiastic American capital
began locating in new investment areas. These new investments arose from the
globalization of the international division of labor in the electronics industry.
Philippine entry into this lucrative opening for foreign direct investments in
general was the establishment of export processing zones. These export processing
zones enabled industrial locators to speed up export or import processing, the
acquisition of land for factories, the isolation from much bureaucratic red tape
and control measures that are often associated with tight, inward-looking policies.
By the mid-1990s, the investment attraction framework within export processing
zones accelerated further and, finally, the Philippines is able to again attract new
flows of foreign direct investments, including American capital. In short, it opened
up industry for the participation of a wider range of foreign direct investments
without being hampered by the restrictive economic policies of an inward-looking
industrial framework. In this, American capital had found a way to come back
more strongly.
Despite the improved capacity to attract foreign capital-engaged electronic
parts assembly, the country lost its competitiveness in expanding the garmentmaking industries. In this case, the economic policies that continued to hamper the
growth of this industry were tied up with labor market policies, especially those
linked with minimum wage setting. American investment in garment making was
a strong presence very early during independence and should have expanded more.
But the import substituting framework for industrial growth favored the highly
protected textile industries and disfavored the importation of raw materials from
the world for the garment makers. And then, rising domestic wages for formal
employment in organized manufacturing further edged out industries engaged in
exports that were heavily labor-intensive.
4.3. Inward-looking policies
The inward-looking approach promoted measures to replace imported goods
rather than an industrialization that was based on comparative advantage. This
program was a hollow strategy. Most of the promoted industries using domestic
capital were mainly import-dependent operations that assembled imported raw
material inputs. With little domestic feedback on production and lacking scale in
production size, these new industries simply produced profits for their investors
36
while raising the cost of the products to Filipino consumers. The opportunities
that were built on tariff protection and other forms of subsidies to domestic
industrialists also worked for import-substituting American investors.
The protectionist industrial policies emanated from the early import and
exchange controls and by tariff protection. These policies created a large group of
domestic industries that depended mainly on protection. As a result, future efforts
to reform the framework to allow for amendments that opened the industries to
more competition encountered strong opposition.
The special trade preferences, for instance, heightened the economic isolation
of the country in international trade. The special preferences made it difficult for
the country to multilateralize its approach to international economic issues. The
special concessions to trade under the us-Philippine bilateral tariff preferences
and quotas for specific industries (sugar and coconut industries) created the excuse
to resist multilateral trade policies. Thus, the country did not join the General
Agreement of Tariffs and Trade.17 Industries that benefited from tariff preferences
were more concerned about enjoying those preferences. Of course, the General
Agreement of Tariffs and Trade was for multilateral trade and not bilateral
preferential trade. And it would be many decades later, in 1994, at the birth of the
World Trade Organization, the successor to the General Agreement of Tariffs and
Trade, before it could join fully the world’s trade rules.
In one sense, the Philippines was uniquely blessed with special preferential
treatment with the United States for that market represented an infinite potential
for economic growth, if the trade opportunities were fully exploited for domestic
reasons. Under the revised Laurel-Langley Agreement, the terms of tariff
preferences were even better than those originally under the original agreement.
But the import-substitution climate of domestic industrial protection and extreme
nationalistic measures influenced by the original restrictive economic provisions
of the Constitution swamped all reasons for exploiting an open trading system.
Without the blessings of a similar preferential trade agreement with the United
States, the economies of South Korea, Taiwan, Hong Kong, and Singapore (and
later East Asian countries)—by simply espousing a more open foreign direct
investment policy for themselves—were able to reap major benefits by attracting
heavily American direct investments which manufactured labor-intensive consumer
products for the us markets. Because of the unlimited growth prospects that were
opened for their exports to the rich American market, these countries realized their
industrial revolutions much sooner and more quickly than the Philippines.
University of the Philippines School of Economics
The multilateralization of trade commitments would only come about with the entry of the Philippines
into the World Trade Organization. The Philippines had participated in the negotiation of the later rounds of
multilateral tariff negotiations, beginning with the Tokyo Round and the Uruguay Round that ultimately led
to the formation of the World Trade Organization in 1994.
17
37
The author is Professor Emeritus at the University of the Philippines School of Economics.
This paper is part of his longer study of the US war damage payments to the newly
independent Republic of the Philippines after World War II. Collection of data arose out
of an unplanned research effort in 2003 when the author found the individual records
involving the recipients of war damage payments in the official minutes of meetings of
the Joint Philippine-US War Damage Commission that were lodged in the US National
Archives in Washington D.C. The festschrift for Hal Hill was the immediate incentive for
him to restart the study. The author uses segments from this study to restate an old theme
concerning the missed opportunities arising out of Philippine-US economic relations.
References
Balisacan, A. and H. Hill [2003] “Introduction”, in A. Balisacan and H. Hill, eds.,
The Philippine economy: development, policies and challenges. Quezon City:
Ateneo de Manila University Press.
Golay, F.H. [1961] The Philippines: public policy and national development. New
York: Cornell University Press.
Hartendorp, A.V.H. [1953] History of industry and trade of the Philippines.
Manila: American Chamber of Commerce.
Jackson, J.K. [2013] D.C. direct investment abroad: trends and current issues.
Washington D.C.: Congressional Research Service,
Philippine War Damage Commission [1951] “Rehabilitation of the Philippines:
final and ninth semiannual report of the United States Philippine War Damage
Commission, Manila, Philippines, March 31, 1951.”
Sicat, G.P. [2013] Weighing in on the Philippine economy and social pderogress.
Mandaluyong: Anvil Publishing.
Sicat, G.P. [2015] “The Philippine economy during the Japanese occupation,
1941-1945”, in M. Boldorf and O. Tetsuji, eds., Economies under occupation:
the hegemony of Nazi Germany and imperial Japan in World War II. London &
New York: Routledge. 191-204.
Sjöholm, F. [2013] “Foreign direct investments in Southeast Asia”, Working Paper
Series 987, Research Institute of Industrial Economics.
united states (US) [1938] Joint preparatory committee on Philippine affairs,
report of May 20, 1938. Washington D.C.: Government Printing Office.
Yoshihara, K. [1985] Philippine industrialization: foreign and domestic capital.
Singapore: Oxford University Press.
PRE
Singapore’s five decades of development:
lessons and future directions
Chia Siow Yue
This paper looks at Singapore’s transition from a Third World
economy at the time of political independence in August 1965
to a First World economy five decades later. It is a unique
city-state with a dearth of land and natural resources and
pursuing a government-led, mnc (multinational corporation)led, and export-led development strategy. Government and
governance have been outstanding, and the city-state overcame
its constraints by adopting open trade and investment policies,
making the region and the world its economic hinterland. Five
decades later, while facing the same demographic and land
constraints, it has to overcome the challenges of transitioning
into a private-enterprise-driven and innovation-driven economy.
JEL classification: F1, F2, H1, J1, O1, O2
Keywords: Singapore development model, economic policy, social policy,
city-state economy
1. Introduction
Singapore gained its political independence on August 9, 1965 and celebrates
its 50 years of nationhood in 2015 with much introspection as well as aspirations
in going forward. Notwithstanding its small nation status, its rapid transition
from a Third World to a First World economy has attracted worldwide attention
from the scholarly, policymaking, and business communities.
In 1965, the economic and political challenges facing Singapore were more
challenging than those faced by many newly independent developing economies.
Its city-state economy was small, with land area of only 580 square kilometers
and a population of less than 2 million; there was a dearth of natural resources,
including crucial water and energy. Political independence had cut off its
Malay Peninsula hinterland and its two economic pillars—entrepôt trade and
British military base—faced dismal prospects. There was high unemployment
39
and budget deficits, militant trade unions, and ethnic unrests. Many thought
Singapore was a “basket case”.
However, there were also some positive initial conditions. Singapore inherited
the British legal system and use of the English language in government and
business and has a wel1-developed seaport, commercial and financial expertise,
a highly monetized economy, and no backward agricultural sector. Its per capita
income was only second only to Japan in Asia. Furthermore, as with other Asian
newly industrializing economies at the time, Singapore also benefited from
the growth of global production networks and supply chains that spearheaded
its industrialization. Above all, Singapore was lucky to have a visionary and
non-corrupt political leadership and a pragmatic and effective economic
management team.
Going forward, the overall economic challenge for Singapore is how to
successfully transition to an innovation-driven economy and meet the rising
aspirations for better quality of life and social inclusiveness. While there are
many constraints and challenges, Singapore also has a foundation of high-quality
institutions and is educationally and financially in much better shape than five
decades ago. Following the September 2015 General Elections, the government
is establishing a Futures Committee to explore the pathways for the next phase of
Singapore’s economic development.
Section 2 of this paper analyzes the policy and performance record of the
past five decades. Section 3 examines the domestic and external constraints,
challenges, and opportunities in going forward. Section 4 concludes.
2. Performance record and policies pursued
2.1. Transition from Third World to First World economy
This transition from 1965 took five decades. In the first two decades, the
political leadership undertook the daunting task of nation building, economic
restructuring, and forging social cohesion. It adopted a development model
characterized by being government-led, foreign mnc-led and export-led.
The government built a world-class, non-corrupt public service with recruitment
and promotion based on meritocracy; reformed the labor movement resulting in
industrial peace and a flexible labor market; and built low-cost public housing
to remove urban slums and provided home ownership for the masses. Singapore
faced threats to its traditional economic pillars of entrepôt trade and servicing the
British military base, and the government played a proactive role in jump-starting
the economy through industrialization. The very small domestic market and
policy lessons from failed import substitution experiments elsewhere led to the
choice of an export-led industrial strategy. The lack of manufacturing expertise
and the daunting task of transforming trading entrepreneurs into industrial
entrepreneurs capable of international marketing led to reliance on foreign mncs.
This would enable Singapore to access foreign investment capital, technology and
managerial expertise, and integrate into global production networks and supply
40
Chia: Singapore’s five decades of development:
chains. Where mncs seemed hesitant to enter, government-linked-corporations
were formed to fill the void.
As a labor shortage emerged by the late 1970s, Singapore’s industrial strategy
evolved towards higher-skill and higher-value added manufacturing and services.
In more recent decades, the government has steered the economy through the
1991 Strategic Economic Plan, the 2003 Economic Review Committee Report,
and the 2010 Economic Strategies Committee Report. In particular, the 2010
report developed strategies for Singapore to achieve sustained and inclusive
growth to 2020.
• Economic growth targets: Annual gdp growth of 3-5 percent and annual
productivity growth of 2-3 percent. Economic growth to be driven by
productivity in the face of growing labor shortage and rising real wages.
• Human skills: Singaporeans with the skills, expertise, and flair to take on
higher quality and higher- level jobs across the whole range of occupations.
• Singapore companies: Deeper base of globally competitive Singapore
companies with leadership in market niches across Asia; a key base for global
companies to grow and manage their pan-Asian operations and for Asian
enterprises that are expanding internationally.
• Innovation: Vibrant climate of innovation, seeking commercial success
through design, new products, and services, and tapping knowledge from a
broader base of public and private sector r&d.
• Leading global city with its unique characteristics.
In the May 2011 General Elections, the ruling People’s Action Party lost
a significant percentage of the popular vote, due to public dissatisfaction over
the rising cost of living and housing prices; stagnating wages for the lowincome workers and widening income gap; and crowding out by foreign labor
of public housing, public transportation, and public healthcare facilities. In the
interim years to the August 2015 General Elections, the government attempted
to address many of these grievances through pumping up public housing, public
transportation, and public healthcare facilities, increasing subsidies and transfers
to the low income and the elderly, improving the wages of low-income workers,
and slowing down the inflow of foreign workers. In the recent August 2015
General Elections, the People’s Action Party government was returned to office
with a much higher percentage of the popular vote, reflecting satisfaction with
the ability of the government to deliver on its promises.
2.1.1 Growth and global competitiveness
High gdp growth was achieved for a sustained period, with full employment
and low inflation delivering rising incomes and living standards. Economic growth
averaged 7.5 percent annually in the period 1965-2014, to reach a per capita gdp
of S$71,318 (us$56,284) by 2014. Annual gdp growth moderated from 7.6
41
percent in the decade 1990-1999 to 4.9 percent in the decade 2000-2009 as the
Singapore economy matured and moved towards its technological frontier and
as the economy was buffeted by many cyclical shocks. For 2011-2014, annual
economic growth further slowed to 3.6 percent. Average labor productivity growth
slowed from 3.4 percent in the 1990s to less than 1 percent in recent years.
The 2015 Global Competitiveness Report ranked Singapore second behind
Switzerland in overall competitiveness among 144 countries. Singapore ranks high
on basic requirements, institutions, infrastructure, macroeconomic environment,
health and primary education, efficiency enhancers, higher education and
training, goods market efficiency, labor market efficiency, and financial market
development. Singapore scored less well on the technological front in terms of
technological readiness and innovation and sophistication.
2.1.2. Structural change
Singapore’s current economic structure is shown in Table 1.
TABLE 1. Singapore’s changing economic structure
2000
2014
S$ million
Total GDP, current market prices
2000
2014
Percentage distribution
162,584
390,089
100.0
100.0
52,434
92,055
32.3
23.6
40,699
67,817
25.0
17.4
Construction
8,863
18,961
5.5
4.9
Utilities
2,719
5,148
1.7
1.3
Services producing sectors
93,154
259,448
57.3
66.5
Wholesale and retail trade
20,405
64,440
12.6
16.5
Transport and storage
15,464
25,359
9.5
6.5
Accommodation and food services
3,521
8,161
2.2
2.1
Information and communications
5,697
14,915
3.5
3.8
Financial services
15,748
46,026
9.7
11.8
Business services
16,725
58,168
10.3
14.9
5,759
17,018
3.5
4.4
11,238
21,568
6.9
5.5
Goods producing sectors
Manufacturing
Ownership of dwellings
Taxes on products
Source: Yearbook of Statistics Singapore, 2011 and 2015
In 1965, the leading industries were domestic market oriented (printing and
publishing, and food and beverage manufacturing). Industrial development was
export-led and foreign mnc-led. By the 1980s, the leading industries comprised
petroleum refining, marine transport equipment, and electronic products and
components; Singapore was well integrated into regional production networks
and global supply chains. By the 2000s, the leading industries were electronic
and optical products, pharmaceutical and biological products, marine transport
equipment, and machinery and equipment.
42
Singapore also grew as a regional services hub in finance and sea and
air transport, linking Southeast Asia to the rest of the world. Its competitive
advantages are a strategic geographical location; well-developed physical
and telecommunications infrastructure; expertise in commerce, finance and
infrastructure management; well-educated and English-speaking workforce;
conducive legal environment; minimal restrictions on right of establishment;
favorable tax regime; absence of controls on capital flows and foreign exchange
transactions; and political, social and economic stability. It serves as the regional
headquarters of many American, European, and Japanese mncs. Emerging
service activities include healthcare, education, and creative industries.
2.1.3. Social well-being
Singapore ranked 9th in the undp’s 2014 Human Development Index
with improvements in recent decades in life expectancy at birth, expected
years of schooling, mean years of schooling, and Gross National Income
per capita. The Boston Consulting Group’s 2015 study ranks Singapore 10th
in overall well-being, the only non-European nation to make the top 10.
Singapore was found to have made particularly strong progress in education and
governance, but it lagged in economic stability and income equality.
Singapore’s Gini coefficient rose to 0.478 by 2012, before it fell to
0.464 in 2014. After adjusting for government transfers and taxes, it
moderated from 0.432 in 2012 to 0.412 in 2014, still high by Organisation
for Economic Co-operation and Development standards. The ratio of
household income from work per household member at the top and bottom
percentiles rose to 9.64 in 2008 before declining to 8.92 in 2014 before
government transfers and taxes and to 6.02 after. Also in 2014, the median
monthly resident household income from work per household member was
S$2,380, with 8.2 percent of resident households earning under S$2,000.
Various government measures introduced in recent budgets include increasing
social transfers, especially to the working poor through the Workfare Income
Supplement scheme and the Workfare Training Support scheme. The Workfare
Income Supplement scheme provides support for low-wage workers and
encourages them to work by supplementing their work income and Central
Provident Fund (cpf) savings and encourages them to enhance their skills and
employability. The Workfare Training Support scheme encourages employers to
send their older low-wage workers for training.
However, with rising costs of living, low-income households are pressured
to make ends meet. And with expectations of continuing volatile growth and
retrenchments in times of economic recession and corporate restructuring,
the absence of unemployment assistance means unmitigated hardships for the
unemployed. Measures to help the elderly, especially those without adequate
social safety nets, include the Pioneer General healthcare subsidies and the Silver
43
Support scheme. Measures have also been introduced to equalize education and
training opportunities to increase inter-generational social mobility.
2.2. Effective role of the state in development
Government intervention following political independence in 1965
was extensive and went beyond the provision of public and merit goods
as the public sector also played a leading role as investor and catalyst for
Singapore’s development.
International reports such as the World Economic Forum’s Global
Competitiveness Report, Transparency International’s Corruption Perceptions
Index, and the World Bank’s Ease of Doing Business Index have ranked Singapore
highly in quality of economic management and lack of government corruption.
The single-tier city-state government, a dominant-party legislature, and
close integration of government ministries and agencies make for cohesive topdown decision-making and swift implementation of policies and measures.
Public sector manpower recruitment is based on meritocracy, and public sector
corruption is well controlled. Statutory boards were created to provide added
administrative and financial flexibility, and many were given significant autonomy
to carry out their tasks. Government-linked-corporations were established to lead
the development of strategic sectors such as banking, telecoms, air and maritime
transport development, industrial estates, and industry clusters. Pragmatism
is manifested in government willingness to continually calibrate the balance
between state intervention and market discipline. The leadership and policymakers
continuously interact with international organizations, foreign governments and
leading business leaders and thinkers around the world to establish friendly ties
and identify emerging technological, economic and business trends globally and
regionally so as to plan appropriate and timely policy responses.
Prudent fiscal and monetary policies were pursued. A low tax regime
encouraged investments, savings, and the work ethics, while tax evasion is
severely punished and budgetary expenditures were kept in check by eschewing
the welfare state. There is zero external public debt and a large accumulation of
forex reserves.
Innovative institutions were encouraged to tackle unique problems that
Singapore faced early on. For example, the institutions to promote industry,
investment, and export include the following:
• The Economic Development Board was established in 1961 to spearhead
Singapore’s industrialization and foreign direct investment drive. Its
International Advisory Council includes global heads of mncs to advise on
international and regional strategies. Its network of overseas offices in North
America, Western Europe and Asia target sectors, activities and firms for
investment promotion. It functions as a one-stop investment center, assisting
44
investors to reach commercial production in Singapore in a matter of weeks.
It focuses not only on pre-investment promotion but also on post-investment
services, keeping existing investors satisfied so that they are encouraged to
stay, reinvest and expand. Since 1993 the Economic Development Board also
led in promoting outward investment drive by Singaporean firms.
• spring Singapore promotes Singaporean enterprises through assistance
in financing, capability and management development, technology and
innovation, and accessing new markets. As the national standards and
accreditation body, it develops and promotes internationally recognized
standards and quality assurance infrastructure that build trust in Singapore
enterprises, products, and services.
• International Enterprise Singapore promotes the overseas growth of
Singapore-based enterprises and international trade. With a global network,
it offers services to help enterprises export, develop business capabilities,
find overseas partners, and enter new markets. At the same time, it works to
position Singapore as a base for foreign businesses to expand into the region
in partnership with Singapore-based companies.
The National Wages Council fosters tripartism, industrial peace, and wage
flexibility. It was formed in 1972 as a tripartite body (representing employers,
trade unions and government) when a tight labor market led to concerns
over possible rise in industrial disputes and excessive wage demands. It
forges national consensus on wages and wage-related matters and its annual
recommendations served as guidelines for negotiation between employers and
unions. Following the 1985-1986 recession, the National Wages Council shifted
from issuing quantitative to qualitative guidelines to enable more flexibility
in wage negotiations and introduced a flexible wage system with variable
components linked to company performance. Following on the Asian financial
crisis of 1997-1998, further wage flexibility was introduced in the form of a
monthly variable component to enable companies to respond more readily to a
volatile business environment.
cpf provides home ownership and a social safety net. It is a compulsory
self-funded scheme inherited from Singapore’s colonial past, initially intended
to meet Singaporeans’ retirement needs but subsequently expanded to include
housing and healthcare needs. The employer and employee contributions to
the cpf have been tinkered with from time to time to reflect cyclical upturns
and downturns in business costs and profitability. With expanded use of cpf
funds, the amount left for old age retirement has been seriously eroded, raising
concerns over its adequacy to meet the retirement needs of a rapidly ageing
population.
The Housing and Development Board provides public housing for the masses.
It was established in 1960. By 2014, 82 percent of Singapore’s population was
45
living in Housing and Development Board-built flats, with 80 percent home
ownership. Public housing has transformed the physical and social landscape
of Singapore. It has enabled slum removal and efficient town planning in landscarce Singapore as well as social and ethnic integration. From time to time,
the affordability of public housing became a hot political issue. While rising
property values benefit the existing homeowners, there was growing concern
among young couples whose home ownership aspirations were checked by rising
prices and a long waiting period. In recent years, a slate of new policies has
been introduced to improve affordability and availability, particularly for young
couples and low-income families.
2.3. Outward- and forward-looking sectoral policies
2.3.1. Industrial development
The industrialization strategy is export-led and fdi-led. Export orientation
was necessitated by Singapore’s small domestic market. A strategy of trade
openness enables export production according to comparative advantage with
exploitation of economies of scale and agglomeration, while zero import tariffs
on capital and intermediate goods help keep production costs competitive
and those on consumer goods benefit consumers through lower retail prices.
Singapore is a key node in the region’s production and distribution networks.
High trade-dependence, however, also makes Singapore vulnerable to changes
in the external environment and necessitates nimble economic management to
moderate the effects of transmitted business cycles.
Singapore’s export manufacturing necessitated dependence on fdi. The citystate lacked industrial or crafts traditions, and transforming trading entrepreneurs
into industrial entrepreneurs with technological and international marketing
capabilities was time consuming. fdi enabled Singapore to have access to
foreign technology as well as industrial capital, management, marketing and
integration into global production networks and supply chains. Attracting fdi
became a policy priority. As most developing countries were then strongly
influenced by the dependency school and eschewed the role of mnc, the firstmover advantage enabled Singapore to successfully leapfrog into export
manufacturing in the late 1960s.
With the emergence of labor shortages by the late 1970s, Singapore’s industrial
strategy shifted towards higher skill and value added manufacturing. The
Manufacturing 2000 program targeted manufacturing at not less than 25 percent
of gdp and called for the development of industry clusters, aimed at upgrading
capabilities across the entire value chain. The Industry 21 initiative identified
electronics, chemicals, engineering, life sciences, education and healthcare,
headquarters, communications and media, and logistics as industry clusters to be
nurtured. This transformation of industry-mix by 2010 is seen in Table 2.
46
TABLE 2. Singapore’s industry clusters, 2010
Total manufacturing
Electronics
Semiconductors
Computer peripherals
Data storage
Information communications
and consumer electronics
Other electronic modules
and components
Chemicals
Petroleum
Petrochemicals
Specialty chemicals
Others
Biomedical manufacturing
Pharmaceuticals
Medical technology
Precision engineering
Machinery and systems
Precision modules and components
Transport engineering
Marine and offshore engineering
Aerospace
Land
General manufacturing industries
Printing
Food, beverages, tobacco
Miscellaneous industries
270,494.7 56,863.8
94,193.6 17,865.0
57,029.4 11,904.3
8,966.9 2,235.6
14,487.2 1,977.5
11,375.6 1,111.7
2,334.6
419,963
80,466
40,662
10,659
13,360
8,149
Employment
Value added
Output
Value added
per worker,
S$000
Employment
number
Value added,
S$ million
Output, S$
million
Percentage
distribution
135 100.0 100.0 100.0
222 34.8 31.4 19.2
293 21.1 20.9
9.7
210
3.3
3.9
2.5
148
5.4
3.5
3.2
136
4.2
2.0
1.9
635.9
7,636
83
0.9
1.1
1.8
80,506.5 6,081.4
41,627.9 1,234.1
29,991.0 1,972.0
6,794.5 2,248.4
2,093.0
627.0
23,253.5 11,126.7
19,668.0 9,700.4
3,585.5 1,426.3
26,546.2 7,620.3
14,073.9 4,186.1
12,472.3 3,434.1
25,058.7 8,339.4
16,030.6 4,875.7
7,207.7 2,826.5
1,820.4
637.3
20,936.2 5,831.0
2,656.7 1,340.3
7,283.3 1,868.1
10,996.2 2,622.7
23,826
3,535
4,735
9,846
5,710
13,738
5,369
8,369
92,323
38,814
53,509
115,478
91,140
18,213
6,125
94,132
17,406
26,500
50,226
255
349
416
228
110
810
1807
170
83
108
64
72
54
155
104.1
62
77
71
52
29.8
15.4
11.1
2.5
0.8
8.6
7.3
1.3
9.8
5.2
4.6
9.3
5.9
2.7
0.7
7.7
1.0
2.7
4.1
10.7
2.2
3.5
4.0
1.1
19.6
17.1
2.5
13.4
7.4
6.0
14.7
8.6
5.0
1.1
10.3
2.4
3.3
4.6
5.7
0.8
1.1
2.3
1.4
3.3
1.3
2.0
22.0
9.2
12.7
27.5
21.7
4.3
1.5
22.4
4.1
6.3
12.0
Source: Yearbook of Statistics Singapore, 2011
Industrial estates and business parks ensure the efficient use of scarce land
resources, enable businesses to enjoy the economies of scale and agglomeration
from locational clustering, and allow speedy start-up of commercial operations
with the ready availability of land and industrial amenities. The Jurong Industrial
Estate was established in the late 1960s and is the forerunner of many smaller
and specialized industrial estates, business parks, and science parks. The latest
developments are the Jurong Island petrochemical cluster and the Biopolis
biotech cluster.
47
2.3.2. Services hub development
The role of services in the Singapore economy rose to reach 66.5 percent
of gdp by 2014, with leading services in trade, business services and
financial services.
Blueprints to develop the Singapore services sector were first outlined in the
1991 Strategic Economic Plan, which emphasized trade, transport, logistics,
telecoms, and financial services. This was followed by the 2003 Economic
Review Committee Report, which has the following recommendations:
• Position Singapore as world-class education and healthcare hub, with
emphasis on attracting a strong cluster of education centers with worldclass universities, executive learning centers, corporate training centers, and
distance learning providers.
• Build the information communications and media cluster into a global hub
in Asia for the digital economy offering a wide range of initiatives and
developments in ict, media, e-commerce, and the Internet. Parallel with
efforts to strengthen the telecommunications infrastructure is the active
promotion of software development, Internet builders, application service
providers, portals, and intermediaries.
• Build on Singapore’s reputation as the premier location for mncs to attract
them to base regional and business headquarters.
• Nurture a logistics and supply-chain management cluster to develop
Singapore into a leading Asia- Pacific integrated logistics hub and to build
up supply chain capabilities by getting logistics players, cargo airlines,
value-added distributors, and manufacturers to locate their supply chain
centers for Asia in Singapore. These world-class logistics services enhance
manufacturers’ global supply chains. A free trade zone logistics park has been
established at Singapore Changi Airport, and a chemical logistics hub set up
on Jurong Island.
The 2010 Economic Strategies Committee Report further recommended that
Singapore carry out the following:
• Grow manufacturing-related services such as headquarter-related activities,
r&d, Intellectual Property (ip) management, and product lifecycle management
by capitalizing on the convergence of manufacturing with services.
• Strengthen Singapore’s position as a leading global-Asia financial and
business hub that connects the global and Asian business community.
• Establish Singapore as the leading business hub in Asia with strengths in
risk management and trading, asset management and private banking, loan
syndication, project and infrastructural financing, and capital raising as
well as information communications technology, accounting, legal, and
consulting services.
48
• Seek to seamlessly integrate physical trade with related services such as
trade finance, risk management, supply chain management, certification,
and distribution.
• In maritime and aviation areas, enhance activities in insurance, financing,
legal, and arbitration services.
• Develop Singapore as a leading consumer business center with clustered
companies in marketing, branding, consumer research, and market
intelligence, capitalizing on Singapore’s Asian cultural affinity, and develop
Singapore into the pan-Asian location of choice.
• Establish Singapore as the location for future-ready urban solutions, by
leveraging on Singapore’s own future urban needs and track record in urban
planning, focusing on areas such as urban mobility/smart transportation,
energy efficiency and management, renewable energy, and water and waste
management. Such initiatives can also foster collaboration among Singaporebased companies to provide solutions that could be scaled up and exported.
2.4. Global and regional trade and investment integration
Singapore has been able to leverage on its strategic geographic location with
infrastructure development, financial, and commercial expertise, and a free
trade policy to become trading, transportation, and financial hubs. Its growth
performance to date is due in no small part to the pursuit of economic openness.
Table 3 shows Singapore’s economic openness through trade in goods and
services and inward and outward direct investment.
TABLE 3. Singapore’s economic openness
Total merchandise trade (S$ million)
Total merchandise exports (S$ million)
Domestic exports (S$ million)
Re-exports (S$ million)
Domestic/total exports (percentage share)
Total merchandise imports (S$ million)
Merchandise trade/GDP ratio (%)
World Trade Organization Most Favored Nation average tariff
(%)
Total services trade (S$ million)
Exports of services (S$ million)
Imports of services (S$ million)
Services trade/GDP ratio (%)\
Inward FDI stock (S$ million)
Inward FDI/GDP ratio (%)
Outward FDI stock (S$ million)
Outward FDI/GDP ratio (%)
Outward/Inward FDI ratio (%)
2000
470,001
237,826
135,938
101,888
57
188,142
292
2010
902,063
478,841
248,610
230,231
52
423,222
297
2014
982,701.9
518,922.7
273,492.1
245,430.6
53
463,779.1
252
101,125
49,213
51,912
63
1999
276,819
92,720
33
0
284,252
152,929
131,323
97
2010
625,780.4
194
425,207.7
132
68
0
357,298.1
177,935
179,362.5
92
2014
853,339.5
219
531,691.1
136
62
49
2.4.1. The pursuit of free trade agreements (ftas)
Singapore’s trade in goods and services amounted to over 300 percent of
gdp, and the import contents of consumption, production, and exports are
extremely high. It is a key node in the global and regional production networks
in the electronics and machinery industries. Top export destinations are asean
(particularly Malaysia and Indonesia), eu, us, as well as Northeast Asia (China,
Japan, Hong Kong, Taiwan, South Korea). The major import sources are asean,
eu, us, China, and Japan.
Table 4 shows that domestically produced exports comprise largely of refined
petroleum and petroleum products and technology-intensive products, such as
chemical and biotech products and high-end electronics. Exports of services
include financial, maritime, aviation, logistics, and tourism services, while
education and health services are also emerging exports. Singapore also serves
as an entrepôt of Southeast Asia, providing efficient and timely shipment services
for a sizeable portion of the region’s trade in commodities and manufactures. Its
advantages are its container port (second in Asia after Shanghai), frequency of
shipping and airline routes, logistics, financial services, and other distribution
services. As a result, shipment costs to destinations in the Americas and Europe
are often cheaper via Singapore than via national ports in the region.
TABLE 4. Singapore’s domestic exports, 2014
Value (S$ million)
Percentage share
Food, beverages, and tobacco
6,522
2.38
Crude materials
2,074
0.76
106,986
39.12
Petroleum and products
Animal and vegetable oils
217
0.08
Chemicals and chemical products
49,383
18.06
Machinery and equipment
73,567
26.90
Electronics
44,059
16.11
Integrated circuits and parts
24,784
9.06
Personal computers and parts
9,313
3.41
Diodes and transistors
3,412
1.25
Disk drives
1,910
0.70
Telecoms equipment
1,454
0.53
29,508
10.79
9,689
3.54
Non-electronic machinery, equipment
Professional, scientific, optical equipment
Others
Total domestic exports
25,054
9.16
273,492
100.00
50
Singapore has been prolific in signing and implementing ftas with countries
and regions across the world. These ftas benefit both ftas and foreign companies
in Singapore who are goods exporters, service providers, and investors, providing
them with improved market access and investment protection.
There is little domestic resistance to ftas, as businesses are accustomed to
import competition. Services liberalization leads to inflow of foreign service
providers, which improves the efficiency, range, and quality of services available
to Singapore’s goods producers and consumers. There are also indirect benefits
as trade and investment expansion leads to higher economic growth, job creation,
and spin-offs for domestic enterprises. National and preferential treatment and
investment protection measures in ftas also facilitate more Singapore-based
companies, especially small and medium-sized enterprises (smes), to venture
abroad, particularly to asean countries.
2.4.2. Inward and outward direct investment integration
From the 1960s to the 1970s, inward fdi helped to close the domestic savinginvestment gap and finance net imports of goods and services. By the mid-1980s,
Singapore had become a net capital exporter, and inward fdi contributed to
technological and managerial know-how and integration into regional production
networks and global supply chains. Inward fdi contributed importantly to
Singapore’s increasing export sophistication, particularly in electronics,
petrochemical, chemical, and pharmaceutical and financial sectors. Wholly foreign
and joint venture firms are much more export-oriented than domestic firms.1
Singapore’s technological development can be divided into four phases.2 First,
from the early 1960s to mid-1970s, there were few innovation links between
foreign mncs and the rest of the Singapore economy. Second, from the mid1970s to late-1980s, there was rapid growth of local technological development
within mncs and development of local supporting industries. Third, from the
late-1980s to late-1990s, there was rapid expansion of applied r&d by foreign
mncs, local firms, and public r&d institutes. Lastly, since the late-1990s, there
has been emerging emphasis on high-tech startups and basic r&d development.
However, Singapore’s ability to innovate and pioneer new technologies still lags
behind the world frontier.
Singapore began the outward investment drive in 1993. On the push side, a
maturing Singapore economy facing severe land and labor constraints and rising
costs needed to develop “an external wing” to sustain its growth performance. On
the pull side, East Asia has become the world’s most dynamic economic region and
This is in part due to the fact that domestic firms acted as local suppliers to Singapore-based mncs rather
than exporting directly.
2
See Wong [2003].
1
51
offers Singapore investors abundant natural resources, low-cost labor, and rapidly
expanding markets. Additionally, Singaporean domestic enterprises with limited
outward investment experience and managerial resources would find it easier to
regionalize due to geographic proximity and cultural and linguistic familiarity,
which help reduce information and transaction costs and need for management
oversight. Furthermore, Singapore government agencies have developed close
relations and extensive networks in the region, and its regional projects, information
networks, and contacts could facilitate investments by the Singapore private sector.
Singapore’s ftas also help facilitate overseas investments.
Table 5 shows inward and outward fdi stock. Inward fdi stock rose rapidly
to reach S$853.3 billion by 2013 or 226 percent of gdp. Singapore played host
to thousands of foreign mncs. Reinvested earnings and expansion investments
accounted for a growing share of fdi inflows. Manufacturing’s sectoral share
has been falling, and fdi is much larger in services. Major sources of fdi are
the advanced industrial economies, with eu, Japan, and the us accounting for a
major share. Over the past decade, the eu’s share has risen rapidly, particularly
from the Netherlands and the uk, while the Japan and us shares have shrunk.
Outward fdi stock grew to S$531.7 billion by 2013, or 141 percent of gdp, and
investments were mostly in services, particularly financial and insurance services.
The outward investment in manufacturing reflects the deindustrialization in
Singapore due to land and labor constraints. The outward push has also responded
to the strengthening Singapore dollar and improved investment climate in the
asean region, China, and India. Investments in overseas financial activities
reflect Singapore’s growing role as an international and regional financial center.
In the post-1997 Asian financial crisis, cash-rich Singapore firms were busy
acquiring regional business assets as the crisis countries relaxed their mergers
and acquisitions restrictions. Singapore’s outward direct investment (odi)
traditionally has a strong Asian bias, but the emphasis has declined from a 75.9
percent share in 1981 to 54.4 percent in 2013. China’s share grew rapidly from
less than 3 percent in 1985 to 19.4 percent in 2013, mostly taking place after
1992, almost equal to asean’s share of 19.8 percent.
2.4.3. Spatial connectivity
To serve Singapore’s highly trade-dependent manufacturing and services
activities, the city-state invested heavily in a wide range of infrastructure including
transportation, logistics, and telecoms systems. To maintain its competitive edge,
there is constant upgrading and expansion of infrastructure. Singapore’s worldclass seaports and airports provide crucial links with the region and the world.
Singapore also has a well-developed information and communications technology
infrastructure that contributes to its international competitiveness as a knowledgebased economy.
52
TABLE 5. Singapore’s stock of inward and outward foreign direct investment,
by industry
Inward FDI
Total
1999
2010
2013
S$ million
1999
2010
2013
170,820.8
625,780.4
853,339.5
100.00
100.00
100.00
58,139.6
133,590.7
150,791.1
34.04
21.35
17.67
1,505.1
1,468.3
3,227.6
0.88
0.23
0.38
26,008.7
108,721.5
146,273.6
15.23
17.37
17.14
Accommodation and
food services
1,993.7
3,811.8
3,976.5
1.17
0.61
0.47
5,958.6
36,793.8
36,858.2
3.49
5.88
4.32
873.7
6,418.4
8,595.1
0.51
1.03
1.01
64,647.4
270,176.8
293,621.8
37.85
43.17
34.41
Real estate activities
5,620.7
20,083.1
29,560.5
3.29
3.21
3.46
Professional, technical, and
administrative services
5,674.9
35,173.7
40,578.7
3.32
5.62
4.76
398.4
9,542.4
14,166.5
0.23
1.52
1.66
Manufacturing
Construction
Information and
communications
Financial and
insurance services
Others
Outward FDI
Total
92,719.9
425,207.7
531,691.1
100.00
100.00
100.00
Manufacturing
22,869.5
88,635.7
109,943.8
24.67
20.85
20.68
Construction
797.4
1,342.3
2,238.9
0.86
0.32
0.42
5,921.3
26,571.0
45,725.0
6.39
6.25
8.60
Accommodation and
food services
1,692.4
3,798.6
4,442.2
1.83
0.89
0.84
3,408.7
10,363.6
13,730.8
3.68
2.44
2.58
Information and
communications
2,257.8
17,958.4
21,935.5
2.44
4.22
4.13
44,717.5
206,204.1
210,207.9
48.23
48.49
39.54
Real estate activities
6,869.4
35,390.9
44,215.7
7.41
8.32
8.32
Professional, technical, and
administrative services
2,737.0
7,617.2
10,205.3
2.95
1.79
1.92
Others
1,448.8
27,326.0
34,480.4
1.56
6.43
6.49
Financial and
insurance services
2.5. Education and skills levels of the workforce
Although Singapore has four official languages (English, Mandarin, Malay,
and Tamil), English is the language of administration and business and the
medium of instruction in tertiary institutions.
In the early decades of industrialization, vocational and skills training in
vocational institutes and polytechnics equipped the workforce with industrial
skills. The government also leveraged on the training capabilities of mncs
53
and their home governments by setting up joint training institutes to provide
industrial skills in demand. Singapore was able to minimize the mismatch
between skills supplied and skills needed. As the economy advanced towards a
knowledge-based-economy, the educational level and skills-set required changed
correspondingly.
The educational attainment of the resident labor force has improved rapidly
in the past decade, as the less educated retired and the new labor force entrants
are much better educated. High-skilled professionals, managers, executives,
and technicians have grown more rapidly (with 53.1 percent share in 2014).
The proportion of university degree holders in the workforce correspondingly
increased from less than 2.4 percent of the labor force in the 1970s to 31 percent
by 2013.
Tertiary education has expanded rapidly since the mid-1980s. By 2015, of
the age cohort that started formal education, 45 percent gets into polytechnics
and 30 percent into universities, making a total of 75 percent with tertiary
education. Table 6 shows percentage distribution of polytechnic and university
enrolment by course. The lion’s share found in engineering sciences and business
administration. In more recent decades, the government realigned its focus and
revised the school curriculum to concentrate on developing students’ creativity
and critical thinking.
Singapore has also been improving on its educational quality. The World
Economic Forum competitiveness rankings for Singapore shows that its
educational institutions are among the best in the world. The Organisation for
Economic Co-operation and Development Program on International Student
Assessment shows Singapore students in top rankings in mathematics, science,
and reading. Quacquarelli Symonds ranks Singapore’s public universities among
the top in East Asia.
2.6. Prudent policies and macro-economic stability
Table 7 shows Singapore’s strong financial resources as evidenced from the
government’s strong budgetary position (reflecting prudent monetary and fiscal
policies over the years), the household sector’s assets including the cpf savings
of individuals, and the strong net international position with no government
external debt and high official foreign reserves. In addition, Singapore is one of
the few countries with an aaa sovereign credit rating by Standard and Poor.
54
TABLE 6. Enrolment in polytechnic and university courses
University first degree courses
Education
Applied arts
Humanities and social sciences
Mass communications
Accountancy
Business and administration
Law
Natural, physical, and mathematical sciences
Medicine, dentistry, health sciences
Information technology
Architecture and building
Engineering sciences
Services
Total
Polytechnic diploma courses
Education
Applied arts
Humanities and social sciences
Mass communications
Business and administration
Legal studies
Science and related technologies
Health sciences
Information technology
Architecture and building
Engineering sciences
Services
Total
1998
630
5,354
474
2,351
3,834
629
2,952
1,203
1,515
1,177
12,975
33,092
1,378
595
10,731
306
1,324
2,093
5,263
1,859
29,139
489
53,180
2010
Number
1,841
1,076
9,694
675
3,203
6,426
1,423
7,640
2,553
3,207
1,786
17,294
299
57,117
174
1,192
275
481
4,696
125
1,070
1,999
3,568
539
7,743
352
22,214
2013
1,296
1,262
10,642
702
4,042
6,476
1,491
7,761
3,081
3,888
1,884
17,426
312
60,263
274
1,805
538
606
5,695
159
1,408
2,504
3,420
752
7,453
483
25,097
1998
2010
2013
1.90
3.22
2.15
1.88
2.09
16.18
16.97
17.66
1.43
1.18
1.16
7.10
5.61
6.71
11.59
11.25
10.75
1.90
2.49
2.47
8.92
13.38
12.88
3.64
4.47
5.11
4.58
5.61
6.45
3.56
3.13
3.13
39.21
30.28
28.92
0.52
0.52
100.00 100.00 100.00
0.78
1.09
2.59
5.37
7.19
1.24
2.14
1.12
2.17
2.41
20.18
21.14
22.69
0.58
0.56
0.63
2.49
4.82
5.61
3.94
9.00
9.98
9.90
16.06
13.63
3.50
2.43
3.00
54.79
34.86
29.70
0.92
1.58
1.92
100.00 100.00 100.00
TABLE 7. Singapore’s financial indicators
Government sector
Government operating revenue, S$ million
Government total expenditure, S$ million
Government overall budget surplus, S$ million
Government domestic public debt, S$ million
Household sector
Household net worth
Central Provident Fund due to members
International position
Net international investment position, S$ million
Government external public debt, S$ million
Official foreign reserves, S$ million
Official foreign reserves, US$ million
2010
46,060
45,338
980
321,182
1,191,581
185,888
651,406
288,954
225,754
2014
57,020
51,728
4,998
387,251
1,467,398
275,364
710,039
340,438
256,860
55
3. Going forward: challenges and opportunities
While Singapore has generally been a successful economic development story
over the past five decades, it also inherits a legacy of constraints and challenges
going forward. There are two more immediate economic challenges: first, the
economy is on a declining path of sustainable growth, and this is in the face
of aspirations for higher living standards and social security and an uncertain
and sluggish global and regional economic environment; and second, there is a
need to push for growth driven by innovation and local enterprise. In the more
distant future, Singapore has to worry about the impact of climate change with an
accompanying rising sea level and how this will impact on a small island nation.
3.1. Slower growth rate and productivity growth
In January 2015, the Singapore prime minister announced that the targeted
growth rate going forward over the next several years has to be reduced
to 2-3 percent instead of the 3-5 percent target set by the 2010 Economic
Strategies Committee.
The Singapore economy has matured and its growth trajectory will mirror
that of other small advanced economies. Domestic constraints of land and labor
and a rapidly ageing population have resulted in high costs and deteriorating
competitiveness in the absence of rapid improvements in productivity and a
vibrant private entrepreneurial sector. Furthermore, the external environment has
become less benign and accommodating: regional economies are becoming more
competitive and challenging Singapore’s hub status, and there is slowing demand
for Singapore’s goods and services by the advanced economies and by China as
they go through periods of slower growth themselves. However, Singapore has
the advantage of a solid foundation of high-quality institutions, well-educated
workforce, and ample financial resources to pull through.
Productivity growth performance has not met expectations and requires a
reinforced push to transform business operations and to intensify skills training.
Labor productivity growth fell from an annual 3.4 percent in the 1990s to 1.1
percent in 2000-2009 and 0.3 percent in 2009-2014. One culprit for the poor
productivity performance was the large influx of low-skilled labor in the past
decade, which disincentivize business upgrading, particularly in the construction
and retail services sectors. In addition to putting a cap on foreign labor inflows in
order to force businesses to upgrade, the government has also several initiatives
including tax benefits, grants, and training subsidies to help companies and
workers invest in productivity, innovate, and deepen skills and expertise. In
particular, SkillsFuture is launched as a national movement to enable Singaporeans
to learn new skills for new jobs as the economy continues to restructure and meet
the challenges of globalization and disruptive technological change.
56
3.2. Widening income gap and inadequate social safety net
The Singapore government has long eschewed Western style social welfare and
preferred to focus on providing education, housing, and healthcare, to equalize
opportunities. But the income gap has widened in the past decade, necessitating
remedial measures in recent years as it could undermine social cohesion.
While obviously more needs to be done to raise the living standards and wellbeing of low-income groups and the elderly poor, the government also needs
to ensure that its rising social expenditures remain fiscally sustainable. It has
traditionally exercised fiscal prudence so that there is overall no net public debt or
external debt.
Some years ago, the goods and services tax was introduced to ensure adequate
tax revenues without raising personal and corporate income taxes, which could
disincentivize work and investment. However, the 2015 budget increased the top
marginal income tax rate from 20 percent to 22 percent explicitly to provide funds
for increased government social expenditures. More taxes on the rich and usercharge fees, as well as further measures to equalize opportunities to education,
training, and remunerative employment, can be expected.
3.3. Demographic and labor constraints
Singapore’s natural population growth has slowed dramatically with the total
fertility rate (tfr) declining below replacement rate for decades, which together
with rising life expectancy, has resulted in a rapidly ageing population and elderly
healthcare costs. Population of working age citizens (20-64 years) peaked in 2005,
but the elderly aged 65 and above is growing rapidly. The ratio of working age to
elderly has declined from 13.5 in 1970 to 4.8 currently and will plummet to 2.1 by
2030. Government policy efforts to reverse tfr had produced no sustained result
and stood at 1.3 percent in 2014. The ageing and shrinking population means a
shrinking labor force and a less vibrant and innovative society.
Population and labor force growths have been heavily dependent on
immigration, with foreigners accounting for one-third of the labor force in recent
years. Going forward, labor force growth can come from increasing the female
and the elderly labor force participation rates and hiring foreign workers.
Female labor force participation has been rising (58.4 percent in 2014) and
could be raised further, as it has not reached the levels of many western societies.
However, higher female labor force participation could impact negatively on
the tfr. It could also be argued that better working conditions for working
mothers—such as flexible hours of employment, more family-friendly policies
by employers, more reliable and affordable childcare facilities, ease of re-entry
married women into the workforce—could increase both female labor force
participation and tfr.
57
With rising life expectancy (82.8 years in 2014), Singaporeans are being
encouraged to postpone retirement. In any event, with cpf balances inadequate
to fund most retirement needs, working longer is a necessity for many. The
government has gradually lifted the official retirement age from age 55 to 60;
it is currently age 65. To incentivize the private sector to employ older workers,
the employer’s cpf contribution rates for older workers have been reduced. The
government is also encouraging movement away from a seniority-based wage
system to a productivity-based wage system to encourage the employment and
re-employment of older workers.
To meet the growing shortage of skills, particularly in the transition to an
innovation-based economy, tertiary education and skills training have been
ramped up with expansion of universities, polytechnics, and various skills training
institutes and continuing changes in educational curriculum and pedagogy. To
reduce the demand for labor, there should be faster and wider adoption of labor
saving production technologies and practices.
An open policy towards foreign labor has enabled Singapore to overcome the
labor constraint since the late 1970s. The large influx in recent years is unsustainable
because of physical and social limits, with vocal concerns on foreigners crowding
out locals for jobs, housing, education, health and transportation services, and
recreational space. Too heavy a dependence on foreign workers has also contributed
to the poor productivity performance of the Singapore economy in the past decade,
particularly in the construction and some service sectors.
For an innovation-driven economy, Singapore needs a large and expanding
pool of foreign human talent, even though local talent is growing with the rapid
expansion and revamp of tertiary education and training institutes. Foreign talent
is being recruited through liberalized immigration policies, easing requirements
for permanent residence and citizenship, offer of scholarships and research
fellowships at Singapore tertiary and research institutions, recruitment missions
by universities and government agencies to the main centers of learning abroad,
and improving the living and cultural attractions and tax regime of Singapore for
foreign expatriates.
While the inflow of foreign talent surged, there were even bigger inflows of
low-skilled workers into labor-intensive manufacturing, services, and construction
sectors. Additionally, the Singaporean workforce is becoming increasingly better
educated and eschewing unattractive and low-paying jobs. Government policy
aims at moderating employer demand for foreign workers through the use of work
permits, foreign worker levies, and dependency ceilings, and there have been
increases in these levies from 2010 to make foreign workers more expensive for
employers. The government also announced that foreigners will continue to make
up the current one-third of Singapore’s workforce in the next 10 years, with some
flexibility to take account of booms and recessions. The large presence of foreign
labor, both professionals and low-skilled employees, was a hot political issue in
the May 2011 General Elections. Since then, the government has been further
58
challenged to fine-tune the balance between the needs of businesses for foreign
professionals and workers, particularly in an economic upturn, and the spatial and
social limits of allowing a continuing influx. Cutting off ready access to foreign
workers will exert pressure on businesses to adopt labor-saving technologies,
operations, and practices and will speed up economic restructuring.
3.4. Land and natural resource constraints
Land reclamation is costly and reaching a limit so Singapore would need to
make more efficient and innovative use of its land and space. Additionally, there
is the longer-term threat of climate change and rising sea level, impacting on
Singapore’s coastlands and resulting in alternative sea routes bypassing the Straits
of Malacca.
Going forward, Singapore needs to do the following:
• Enhance land and space productivity. To gain the greatest benefits from its
limited land, Singapore has to create more space upwards and underground
and be more creative in urban planning and design to ensure that Singapore
remains a livable and business-friendly city. Sectoral policy towards
production of goods and services will have to increasingly focus on value
added to economize on land and space demands, such as high-value added and
clean manufacturing and high-end tourism, healthcare, and education markets.
This will also ensure Singapore’s continuing competitiveness vis-à-vis other
less advanced and lower-cost asean countries.
• Enhance water planning and development. Many parts of Asia will face
shortage of water as populations, cities, and agricultural and industrial needs
grow. Singapore has historically imported most of its water supplies from
neighboring Malaysia. However, over the years, Singapore has developed
alternative water resources, including collection of rain water, recycling of
waste water, and desalination of sea water. The city-state will become selfsufficient in water by the time the second water agreement with Malaysia
expires in 2061. In fact, water technology and management has become a new
Singapore growth industry and export of services.
• Enhance the region and the world as Singapore’s hinterland: Fortunately,
Singapore is neither “landlocked” nor “sea-locked”. Instead, it is favorably
located in the dynamic Asian region. Continuing improvement of physical
(land, sea, and air) and information technology connectivity will minimize the
cost for cross-border movement of goods and people and facilitate access to
ideas and technologies. Support for global and regional free trade will help
overcome the dearth of land and natural resources and achieve economies of
scale and scope to overcome its small market size. Singapore must continue to
support trade liberalization in the World Trade Organization and regional and
bilateral ftas.
59
• Meet the rising costs of energy as well as reduce its carbon emissions. As an
energy-deficient economy, Singapore has resorted to using the price signal
in energy pricing, adopting clean technologies and energy-efficient usages
in factories, offices, and homes, and designing of energy-efficient buildings.
At the business level, Singapore should promote industries and services and
processes that are less energy-intensive and/or use less of fossil energy. At
the household level, consumers should be incentivized to save on energy use,
such as installing energy-efficient household equipment. At the transportation
level, Singapore pioneered efforts to restrict private car ownership and use
and heavily taxed fossil-fuel consumption. Exploring energy alternatives
include the nuclear option, which will be especially challenging for a densely
populated city-state.
3.5. Weak local enterprises
High-income Singapore faces the challenge of weak local enterprises and
entrepreneurship. Unlike South Korea and Taiwan, Singapore in the earlier phases
of development and industrialization failed to nurture local private enterprise
and chose to depend on foreign mncs and government-linked-corporations. By
the mid-1980s, local enterprises were confronted with higher cost structures
exemplified by rising land and space costs, labor shortages, and rising wage
costs. Also, major public sector infrastructure projects were tendered out to large
European, Japanese, Korean, and prc firms, creating a vicious cycle of local firms
lacking the experience and track record to bid for them domestically or regionally.
The 1988 sme Master Plan marked the first coordinated national effort to
upgrade local smes and to promote domestic entrepreneurship. Since then,
a plethora of sme assistance schemes have been hatched and implemented.3
Furthermore, the sme 21 Report outlined the need to nurture innovative high
growth world-class smes able to compete in the global marketplace; enhance
sme productivity; and create a knowledge-based pro-enterprise environment.
A multi-agency sme 21 Implementation Committee was formed. The
Technopreneurship 21 initiative was announced in April 1999 specifically to
boost development of technopreneurs. Another Master Plan, named sme21, was
created to take Singapore smes into the 21st century. As a result of the two sme
These include the following: the Local Industry Upgrading Programme which helps local enterprises
achieve greater efficiency through the transfer of management skills and technological know-how from
mncs and large local companies; the Economic Development Board-Joint Venture Matching Service to
help smes seek strategic alliance with foreign organizations for growth; the Business Development Scheme
to encourage smes to seek business opportunities overseas for marketing arrangements, technological
tie-ups, and other business partnerships; the Small Industry Technical Assistance Scheme to help smes
improve their productivity and technological standards through grants to defray part of the costs of approved
upgrading projects; and the Promising Local Enterprise Programme that aims to build 100 local companies
with at least S$100 million sales turnover in ten years.
3
60
Master Plans, Singapore has a spectrum of some 63 assistance schemes to assist
local enterprises.
Government agencies involved in smes include the following: International
Enterprise Singapore, a statutory board responsible for taking smes overseas;
spring Singapore, a first stop for all smes, which are then directed to the relevant
agencies; A*Star that fosters scientific research and the exploitation of technology
through incubator units; Jurong Town Corporation that provides industrial space
and has incubator space as well; Information Development Authority which
develops, promotes, and regulates the information technology and telecoms market
as well as assists the adoption of online and e-commerce technology by smes. In
addition to these government agencies, there are private sector institutions, tertiary
institutions, and non-government organizations, which smes can access.
Notwithstanding these multi-pronged efforts, the World Economic Forum
2011-2012 Competitiveness Report continues to show poor rankings for
Singapore in the following: local supplier quantity; local supplier quality; and
control of international distribution. Due to the dominance of foreign mncs and
dominance of the services sector in the economy, there are few internationally
well-known Singaporean manufacturing brand names. unctad’s list of top
firms from developing countries has few Singaporean names and they are mainly
services firms or government-linked-corporations such as Singapore Airlines,
Singapore Telecommunications, Keppel Corporation, Sembawang Corporation,
Singapore Changi Airport, and the Maritime and Port Authority of Singapore.
The actual and effective assistance provided to smes paled in comparison to
that offered to mncs and government-linked-corporations. Furthermore, local
smes lack a sizeable domestic market to test their ideas, processes, and products
and to grow before expanding abroad.
Many schemes to promote and facilitate local entrepreneurship and sme
development and availability of venture capital have been introduced in recent
years. There is also a noticeable trend of young university and polytechnic
graduates entering directly into business ventures instead of taking up jobs with
the public sector and foreign mncs. So there is hope for Singapore with more
homegrown entrepreneurs.
3.6. Shifting to an innovation-driven economy
According to the World Economic Forum, Switzerland, Singapore, and
Sweden are the top three in the Global Competitiveness Index rankings in
2011-2012.
• Switzerland tops the Global Competitiveness Index rankings; its most
notable strengths are in innovation, technological readiness, and labor
market efficiency. Switzerland’s scientific research institutions are among
the world’s best, and the strong academia-business collaboration together
with high corporate r&d expenditures and strong ip protection ensure the
61
commercialization of research into marketable products and processes. Its
rate of patenting ranks 7th worldwide. Productivity is enhanced by a business
sector and a population that are proactive in adapting the latest technologies.
Competitiveness is also buttressed by excellent infrastructure, well-functioning
goods markets, and highly developed financial markets.
• Sweden, like Switzerland, has been placing significant emphasis on
innovation-led growth. The quality of its public and private institutions are
tops in efficiency, transparency, and ethics. Goods and financial markets are
very efficient. Combined with a strong focus on education and training and
high level of technological absorption, Sweden has developed a sophisticated
business culture and is a global leading innovator.
• Singapore is 2nd in Global Competitiveness Index rankings (after Switzerland)
and top among Asian economies. While Singapore is tops in rankings on
institutions, efficiency of its goods and labor markets, financial market
development, world-class infrastructure, and education, it lags in innovation
and business sophistication and have considerable catching up to do. Singapore
needs to encourage stronger adoption of the latest technologies and measures
that support the sophistication of its companies.
Singapore has to transition rapidly from being an investment-driven to an
innovation-driven economy. At the investment-driven stage, the key drivers are a
strong business environment, openness to trade and investment, well-developed
physical infrastructure, legal and regulatory framework, and a solid skill base.
But at the innovation-driven stage, the key drivers are a strong innovation system,
ip protection, innovative and techno-savvy entrepreneurs, and sophisticated
domestic producers and consumers.
Before the 1990s, r&d activities in Singapore remained low for a number
of reasons: industrial growth in the 1960s and 1970s depended more on cost
efficiency than on innovative capability; the mnc subsidiaries in Singapore had
ready access to the processes and technologies from their overseas parents, and
they generally preferred to conduct r&d in their home base; and there were no
performance requirements to conduct local r&d or partner with local institutions
in r&d. Singapore also lacked a critical mass of scientists and researchers to
provide a stimulating research environment for both mnc and local enterprise
r&d activities.
Recognizing the uphill task, the Ministerial Committee on Research and
Development in 2005 reviewed the r&d strategies and directions for Singapore.
The committee recommended that national r&d efforts should be driven by 5 key
strategic thrusts.4
This is excerpted from a speech by Dr. Tony Tan, outgoing Singapore deputy prime minister and incoming
chairman of the National Research Foundation, on August 11, 2005.
4
62
• Provide more resources for r&d. Singapore’s gross expenditure on research
and development was 2.15 percent in 2003, which lags significantly behind
other leading innovation countries. Singapore must intensify r&d efforts and
achieve a gross expenditure on research and development of at least 3 percent
of gdp within the next 5 years.
• Need to focus on a small number of strategic areas to develop a critical
mass of research capabilities in industries where it can be economically
competitive. Existing key clusters are electronics, chemicals, marine
engineering, and biomedical sciences. A vibrant research environment will
help identify emerging growth areas, such as in environment and water
technologies and interactive and digital media.
• Need for balance between investigator-led and mission-oriented research in
selected strategic areas. Basic investigator-led research is broadly aligned
with the long-term strategic interests of Singapore, and the Ministry of
Education’s Academic Research Fund was raised from S$550 million for
fy2001-2005 to S$1.05 billion for fy2006-2010. Mission-oriented research
would be closely integrated with industry development and investment
promotion strategies, and the budget for A*Star was raised from S$4 billion
for fy2001-2005 to S$5.4 billion for fy2006-2010.
• Encourage more private sector r&d. Incentive packages would be reviewed to
better attract more global r&d centers and activities to Singapore, supported
by a high quality support framework, including a strong base of scientific
and research manpower and sophisticated IP protection regulations. The aim
is to have two-thirds of r&d, mainly development, performed by the private
sector, and one-third of r&d, mainly research, performed by the public
sector agencies.
• Strengthen the nexus between r&d and business. Universities and research
institutes must improve on their ability to commercialize their research
results and have closer collaboration with industry. There is also a need to
promote technology innovation in local enterprises through stronger cofunding frameworks between public and private sectors. In particular,
polytechnics with strong applied research and downstream capabilities and
industry networks could be encouraged to link with industry associations to
collaborate on r&d initiatives.
The government elevated r&d into a national priority by setting up the
Research, Innovation, and Enterprise Council and the National Research
Foundation. The council announced that it plans to spend S$16.1 billion over
2011-2015 on research, innovation, and enterprise, as compared to an allocation
of S$13.55 billion in 2006-2010. Research and innovation will have to underpin
the competitiveness of Singapore’s industries, catalyze new growth areas,
and transform the economy. Increasingly, intellectual capital will be critical to
63
Singapore’s next phase of economic development. More will be done to facilitate
collaboration between industry and public research institutions to foster greater
commercialization of r&d. A larger portion of r&d funding will be awarded
on a competitive basis to projects that will strengthen capability and contribute
to economic and social outcomes. The National Research Foundation has been
established with key responsibilities to implement the strategic thrusts and to
fund longer-term research in strategic areas. The foundation is provided with
funding of S$5 billion for fy2006-2010. This, together with the increased budget
for A*Star and the Academic Research Fund, will more than double the total
public sector r&d budget from just under S$5 billion for fy2001-2005 to almost
S$12 billion for fy2006-2010. Singapore aims to increase gross expenditure on
research and development to 3.5 percent of gdp by 2015 through greater private
sector r&d activity. The foundation’s research priorities are in environmental
and water technologies; biomedical sciences; and interactive and digital media.
Despite the quantum leap, Singapore’s r&d efforts remain modest in absolute
terms, reflecting the small size of the economy, its strong services orientation,
and weak core of large manufacturing corporations. A challenge for Singapore’s
r&d ambitions is that it lacks the breadth and depth of talent to compete against
the bigger and more advanced economies of the us, eu, Japan, and Russia as
well as the emerging economic powers of China, India, and South Korea.
4. Conclusion
In the past five decades, Singapore’s non-corrupt and visionary political
leadership has brought about political stability and social cohesion, and, together
with pragmatic and efficient economic management, has enabled Singapore
to achieve well beyond the economic potential of a small city-state that lacks
natural resources. However, in the process, Singapore’s development depended
heavily on foreign mncs and neglected the nurturing of domestic enterprises. Its
heavy dependence on foreign labor delayed the expansion of tertiary education
and disincentivized industrial upgrading as businesses had ready access to a large
pool of low-wage labor from neighboring countries. Critics have also criticized
Singapore as being a “nanny state” with its extensive and intrusive role in the
economy and society.
Singapore is a now a high-income country and is on the innovation-driven
stage of economic development. Going forward, Singapore has to overcome
domestic constraints and challenges as well as operate in a more challenging
regional and global environment. The government will have to hold back on
its dominant role and let the private sector and civil society grow and become
more creative and risk taking. Given the best educational opportunities in the
world, and a liberalizing trade and investment environment, and with government
playing a less intrusive but more supportive role, the next phase of Singapore’s
64
economic development should be more private-led and local-enterprise led.
Whether the Singapore city-state will continue to excel over the next few decades
remains to be seen.
Singapore Institute of International Affairs
References
Chia, N.C. [2015] “Adding a basic pillar to the Central Provident Fund system: an
actuarial analysis”, Singapore Economic Review Special Issue: a fifty year
retrospective on the Singapore economy 60(3): 1530037-1 - 1530037-26.
Chia, S.Y. [2015] “Globalisation and regionalisation: Singapore’s trade and FDI”,
Singapore Economic Review Special Issue: A Fifty-Year Retrospective on the
Singapore Economy 60(3): 1530034-1 - 1530034-23.
Chia, S.Y. [2012] “Singapore”, unpublished background monograph prepared for
the ASEAN 2030 project coordinated and funded by the Asian Development
Bank Institute.
Chia, S.Y. [2011] “Inward and outward fdi and the restructuring of the
Singapore economy”, in S. Chalongphob, Y.C. Park and S.J. Kang, eds.,
Foreign Direct Investments in Asia. uk: Routledge.
Economic Strategies Committee (esc) [2010] Report of the Economic Strategies
Committee: highly skilled people, innovation economy, distinctive global city.
Phang, S.Y. [2015] “Singapore’s housing policies: responding to the challenges
of economic transition”, Singapore Economic Review Special Issue: a fifty
year retrospective on the Singapore economy 60(3): 1530036-1-1530034-25.
Singapore Department of Statistics [various years] Yearbook of statistics
Singapore.
Singapore Department of Statistics [2014] Key household income trends 2014.
Singapore Ministry of Trade and Industry [2003] Economic review committee
report.
Spring Singapore. Assistance programmes for smes. Downloaded from
website: http://www.spring.gov.sg/
Transparency International (ti) Global corruption report. Downloaded from
website: https://www.transparency.org/research/gcr/
World Bank (wb) [2015] Doing business 2015 report. Downloaded from
website: http://www.worldbank.org/
World Economic Forum (wef) [2010] The global competitiveness report 20102011. Downloaded from website: http://www.weforum.org/reports/globalcompetitiveness-report-2010-2011-0
PRE
The Philippines: on the road to being
an emerging economy
Dante B. Canlas
In the 1980s, the Philippines was viewed as a failure in terms of
its goal to industrialize. But in the past few years of the current
century, the country’s economic prospects improved and the
country is predicted to be joining the next group of “breakout
nations”. The paper looks at the short-run macroeconomic
policy reforms and long-term growth-oriented policies since
1986 that have contributed to improved growth performance
and bright economic prospects. A responsible budget deficitreduction program and investments in factors that support longrun growth, such as human capital, have been helpful. Moving
forward, stabilization policy has gotten more challenging in an
environment of mobile international capital, flexible exchange
rates, and “quantitative easing”. In the long run, policies
conducive to technological progress are essential.
JEL classification: E32, O11
Keywords: growth, convergence, stabilization, technological progress
1. Introduction
Robert E. Lucas, Jr. [1993] in his article entitled “Making a Miracle”,
cites the Philippines as an example of a country that failed to transform itself
into an economic miracle. Lucas uses the latter phrase to describe a process of
productivity growth and industrial transformation associated with profound
improvements in living standards of the country’s population. Lucas has in mind
the transformation that at the time had been occurring in South Korea, Taiwan,
Hong Kong, and Singapore in East and Southeast Asia—these economies had
witnessed a growth rate in real per capita income of 5 to 7 percent each year, in
about three decades from 1960. South Korea and Taiwan were transformed from
predominantly agricultural economies. Today, they are often held up as models
of successful industrialization. Hong Kong and Taiwan had no agricultural sector
66
Canlas: The Philippines: on the road to being an emerging economy
to begin with and relied primarily on export-led growth. I refer to a country
showing signs of undergoing such a transformation as an emerging economy.
About two decades later, Ruchir Sharma [2012], in his book, Breakout
nations: in search of the next economic miracles, includes the Philippines in
his list of developing countries likely to become an economic miracle in the
second decade of the 21st century. The main basis for inclusion in the list hinges
on exhibiting “growth that surpasses expectations”. This brings to mind the socalled bric countries—namely, Brazil, Russia, India, and China—which some
observers before Sharma had tagged as breakout nations or emerging economies.
What did the Philippines do to merit significantly improved economic
prospects? Is the country’s real per capita income showing a tendency towards
convergence, i.e., catching up with the established economic miracles? There are
many possible answers. This paper examines the macroeconomic and growthoriented policies that underpin the significant improvements in the growth
prospects of the Philippines.
A quick look at the growth performance of the Philippine economy over the
past three decades, juxtaposed with two other middle-income Southeast Asian
economies—namely, Thailand and Indonesia—provides some insights into the
positive assessment that the Philippine economy is currently enjoying. Table 1
shows data spanning three decades that were accessed from the World Bank’s
World development indicators. This period witnessed three economic crises
in the Philippines: the debt default of the country in 1983; the Asian financial
crisis of 1997; and the global financial crisis of 2008. In the 1980s, the average
real per capita income each year of the Philippines declined 0.8 percent,
compared to growth rates of 6 percent in Thailand and 4.46 percent in Indonesia.
The Philippines recovered in the 1990s with an average annual growth rate of
0.58 percent. The recovery gained strength in 2001-2010, growing 2.52 percent
each year.
TABLE 1. Real per capita income and average annual growth rate
Country
Philippines
1980 real
per capita
income, y,
(in 2000 US$)
Growth rate
of y, 19811990 (in %)
Growth rate
of y, 19912000 (in %)
Growth rate
of y, 20012010 (in %)
2010 real per
capita y,
(in 2000 US$)
1,088
(0.80)
0.58
2.53
1,375
Thailand
772
6.0
3.64
3.26
2,774
Indonesia
356
4.46
3.15
3.52
1,034
Source: World Bank, World development indicators, accessed August 2011
67
In 1983, the Philippines experienced a balance-of-payments crisis that
forced the government to declare a moratorium on foreign-debt servicing. To
overcome its liquidity problems, the national government arranged a standby loan
agreement with the International Monetary Fund (imf). The latter’s preferred
financial programming techniques and conditionality lending practices anchored
on tight fiscal and monetary policies resulted in a deep recession in 1984-1985.
Real gdp growth declined 11 percent during the two-year period. The economy
recovered in 1986, which gathered strength up to 1989. However, in view of the
deep recession in 1984-1985, the 1980s proved to be a lost decade for the country.
Political shocks in 1990, including attempted coups d’état against the
President of the Republic and natural disasters, caused economic activities to
decline in 1991. A new Philippine President was elected in 1992, but an electric
power shortage in 1992-1993 kept the economy weak, and it did not recover until
1994. The growth was sustained up to 1997, despite the emergence that year of
the Asian financial crisis that was triggered by the devaluation of Thailand’s baht.
The growth rate of real gdp slowed down, ending flat in 1998. A new president
was elected in 1998, but lack of fiscal discipline under the new administration
caused the budget deficits of the national government to balloon, which caused
interest rates to rise and growth to slow down. Still, given the strength of the
recovery in 1994-1997, a moderate growth rate in per capita income was posted
in the 1990s.
Entering the first decade of the 21st century, several shocks—both economic
and non-economic—again battered the economy. In 2000, allegations of the
President’s involvement in an illegal numbers game led to the Philippine
Senate impeaching the President, resulting in his eventual unseating in 2001.
Nonetheless, the economy posted a positive growth rate in real gdp that year.
Then the September 11 terrorist attack took place, an event that led the us federal
government to declare war against terrorism. That war had serious repercussions
on the Philippine economy, in view of the conflicts with Muslim rebels in the
southern part of the country. In any event, the economy proved sufficiently
resilient by posting an annual average growth rate in real per capita gdp of 2.53
percent over the period 2001-2010.
The economic growth in 2001 ushered in the longest recovery in the past
few decades. Policymakers responsible for the economy managed to temper the
output volatility, keeping business fluctuations moderate. The economy dodged
a recession in 2009, with real gdp managing to grow 0.1 percent, in spite of the
2008 global financial crisis that was triggered by the collapse of the sub-prime
housing loan market in the us. Still, the economy failed to realize its capacity
output, as real gdp grew only by an average of 4 to 4.5 percent each year. The
failure to reach full potential is often attributed to poor governance, mainly an
inability to fight corruption in high places during the period 2001-2010.
68
In 2010, real gdp grew a strong 7.6 percent. This moderated to 3.6 percent
in 2011. In 2012, real gdp recovered and grew 6.6 percent, the highest among
Southeast Asian economies. Real gdp grew 7.1 percent in 2013 and 6.1 percent
in 2014. Clearly, the country’s growth performance improved considerably over
the period 2012-2014, with real gdp growth averaging 6.6 percent each year.
The remarkable growth performance of the economy in the past three years is
one reason for the improved assessments of the country’s macroeconomic
performance in the future. Many observers of the Philippine economy attribute
the high-growth rate to a combination of market reliance and good governance
under the incumbent administration.
For a developing economy to catch up with more developed ones, its real per
capita income must exhibit growth on a sustained basis. To provide a quantitative
tone, in a country where real per capita gdp grows 6.9 percent each year, a
doubling of real per capita gdp every 10 years can be expected. Several factors,
both economic and non-economic, matter for growth of real per capita income in
the long run [Barro 1994]. Among the economic factors, it is important in the short
run to avoid unwanted business fluctuations through appropriate macroeconomic
policies [Fischer 1993]. In the long run, it is vital for the economy to overcome
the usual tendency towards diminishing marginal productivity in both factors of
production: labor and capital. As the new endogenous growth theory emphasizes,
such growth is rooted in capital accumulation, whereby capital is defined broadly
to include not only physical, but also human, technological, and social overhead
capital (Romer [1986, 1990]; Lucas [1988]).
For the Philippines, having experienced recurrent boom-and-bust cycles
that forced the government to tap standby lending facilities of the imf on a
recurrent basis, it is useful to revisit the macroeconomic policy reforms that have
contributed to an economic recovery and laid the foundation for that recovery to
gather strength. Meanwhile, among the long-run factors, it is widely recognized
that human capital investment, especially education, is vital to economic growth.
Acting in combination with other forms of capital, human capital can lead to
increasing returns ([Romer 1986]; [Barro and Sala-i-Martin 1990]). Moreover,
since education is a core value of the Filipino people that is enshrined in the
Philippine Constitution of 1987, an investigation of the role of education in the
growth process is warranted.
To begin with, this paper examines the various short-run macroeconomic
policy reforms that the Philippines has pursued to minimize unwanted business
fluctuations. Then, in relation to growth over the long run, the paper assesses
the role that human capital accumulation, focused on education, has played.
There is no question about government support for basic education. But in view
of the important contribution of higher education to technological progress, the
major source of long-run growth, public policy needs to pay more attention to it
[Canlas 2015].
69
Moreover, since economic policy reforms dating back to 1986 have sought to
integrate the Philippines with the rest of the world’s economy, the growth effects
of foreign-trade policy reforms are useful to investigate. For instance, since
1986, political administrations have in succession pursued trade and investment
liberalization, a policy measure that can be relied on to trigger growth over the
long run.
The second section discusses the macroeconomic policy reforms designed to
minimize unwanted business fluctuations. The third section looks at key aspects
of capital accumulation that have long-run growth effects, focusing on human
capital. The fourth section concludes the paper.
2. Macroeconomic policy reforms
Macroeconomic polices are intended to provide a stable and predictable
environment for the private sector. Minimizing risk and uncertainty is vital in
enabling households and enterprises to realize their consumption and investment
plans, elements of aggregate demand that support growth.
Both fiscal and monetary policies have long been recognized as constraints
to sustained growth in the Philippines, a point that was stressed in Canlas, Khan,
and Zhuang [2008] and reiterated by the National Economic and Development
Authority [2010]. Fiscal policy is often singled out as the most binding constraint
to the country’s sustained growth, particularly, the relatively low tax effort,
defined as the proportion of tax revenues to gdp. Fiscal and monetary policies,
however, are closely related, and are commonly viewed in tandem.
Fiscal policy encompasses measures that affect the size of the public debt,
including taxation and government spending. Monetary policy spans policies
affecting the composition of that debt between money and non-interest-bearing
liabilities of the government. When the monetary authority, the Bangko Sentral
ng Pilipinas (bsp), accommodates by lending to the government to finance the
latter’s budget deficit, money supply rises, which constitutes the non interestbearing liabilities of the government. However, when the monetary authority
does not accommodate, then the government through the National Treasury
issues interest-bearing debt claims. In the short run, the choice of financing
a deficit may entail differential impacts on aggregate output and the general
price level.
2.1. Fiscal policy
Any deficit in the national government budget has to be financed. When the
government borrows to finance the deficit, the public debt rises, resulting in heavier
debt servicing. The fiscal history of the Philippines indicates persistent deficits in the
national government budget. In the 1970s, for instance, the Philippine government
pursued an expansionary fiscal policy in an attempt to ward off the recessionary
70
effects of the two oil price shocks in 1974 and in 1979. The Philippines entered the
1980s with a large public debt, largely foreign. When interest rates rose on a global
basis in the early 1980s as a result of disinflation moves by the US Federal Reserve
Board, the government found debt servicing quite burdensome, eventually forcing
the government to default on foreign-debt servicing.
Following the debt moratorium and tapping of a standby loan from the imf,
the government resorted to fiscal and monetary tightening. That ushered in the
recession of 1984-1985, the longest in the postwar economic history of the
Philippines, in which real gdp declined 11 percent during the two-year period.
The fiscal position of the government has long been at the core of the
liquidity crises repeatedly faced by the Philippines. In a small open economy like
the Philippines, the current-account deficit is identically the same as the budget
deficit of the public sector and the savings-investment gap of the private sector.
Table 2 shows such a decomposition of the current-account deficit three years
before the government declared a debt moratorium. It is shown that the budget
deficit of the government as a percentage of gdp, after a near-balance in 1980,
rose to 4.8 percent in 1981 and ballooned to 5.3 percent in 1982. It moderated to
3.3 percent following the declaration of a debt moratorium in 1983.
TABLE 2. Current account, consolidated public sector account,
and savings-investment gap as percentage of GDP
Year
Current
account
Consolidated public
sector account
Savingsinvestment gap
1980
(5.9)
0.46
(6.4)
1981
(6.2)
(4.8)
(1.4)
1982
(9.0)
(5.3)
(3.7)
1983
(8.3)
(3.3)
(5.0)
Sources of basic data: Bangko Sentral ng Pilipinas for current account; Department of
Finance and Department of Budget and Management for consolidated public sector
account. Savings-investment gap is calculated as a residual.
The problem posed by the budget deficit of the national government returned
in 1990, when the budget deficit reached 4.8 percent of gdp. Fiscal policy
was subsequently tightened, forcing the consolidated public sector account to
moderate to 2.1 percent of gdp in 1991. Real gdp growth slowed down to 0.5
percent that year.
A large government budget deficit as a percentage of gdp may also be an
offshoot of weak economic growth. The tax system has automatic stabilizers that
lose their vitality in an economy that’s slowing down. The stabilizers include
revenues from personal income and corporate income taxes. This was evident in
2009, when the budget deficit of the national government swelled again to 3.9
per cent of gdp as a result mainly of the economic slowdown that emerged in the
71
aftermath of the 2008 global financial crisis. And so in 2010, the fiscal position
of the government was still deemed as the most binding constraint to growth in
the Philippines, triggering earnest calls even in government for fiscal reforms.
It is well recognized that government spending at all levels has to be reined
in, while protecting core values in education, health, and infrastructure. But more
importantly, it is widely agreed that insofar as raising the tax effort is concerned,
there is still plenty of room for improvement. Table 3 shows figures on the tax
effort (total tax revenues divided by gdp for the given year) for selected years.
In the years prior to the 1983 foreign debt crisis, the tax effort was declining and
averaging only about 10 percent each year. In 1986, a series of measures aimed
at improving tax collection and tax administration started. From about 10.6
percent in 1985, the tax effort improved, peaking at 17 percent in 1997 before it
weakened again.
TABLE 3. Tax effort (in %), selected years
Year
1980
1981
1982
1985
1990
1995
1997
Tax effort
11.4
10.3
9.9
10.6
14.1
16.3
17.0
Note: “Tax effort” is defined as total tax collection divided by GDP.
Source of basic data: National Statistical Coordination Board, Philippine statistical yearbook.
In 1986, following the restoration of democratic political institutions, the
administration under the leadership of President Corazon Aquino instituted a
tax reform package. One major motivation for these reforms stemmed from the
recurrent liquidity crises of the past that were rooted in the recurrent budget
deficits of the government. In addition, the introduction of foreign-trade policy
reforms gave rise likewise to the need to strengthen the internal tax system
[Canlas 2000]. The Aquino administration started an import liberalization and
tariff reduction program. As border or customs tariffs and trade taxes declined,
strengthening the internal tax system became an imperative, particularly,
personal and corporate income taxes and indirect taxes. Since reliance on
foreign-trade taxes was being reduced, internal tax collection had to increase to
compensate for the foregone border taxes. Table 4 shows the declining long-term
shares of customs tariffs to total taxes and the increasing shares of internal taxes.
TABLE 4. Share of foreign-trade and internal taxes (in %)
Tax type
1980
1985
1990
1995
1997
Internal
56.9
69.4
68.5
67.7
76.4
Customs
38.0
27.8
30.2
31.4
23.0
Others
5.1
2.8
1.3
0.9
0.6
Total
100
100
100
100
100
Source: National Statistical Coordination Board, Philippine statistical yearbook
72
In 1987, President Aquino issued an Executive Order that introduced a
10-percent value-added tax on all sales transactions not explicitly excluded from
the first Executive Order on value-added tax. In addition, a Republic Act was
enacted that amended the National Internal Revenue Code. The latter introduced
a “simplified net income taxation” for the self-employed and professionals in an
attempt to widen the tax base, accompanied by a reduction in marginal tax rates.
The Ramos administration, which succeeded Aquino’s in 1992, resumed the
tax-reform program. It expanded the coverage of the value-added tax law. In
1998, the end of the Ramos term, a Republic Act called the Comprehensive Tax
Reform Program was enacted. The program further sought to expand the tax base
while reducing marginal tax rates. One outcome of the reform measures covering
tax policy and administration during the Ramos years was a significant increase
in the tax effort. In 1997, the tax effort peaked at 17 percent, enabling the national
government to post a budget surplus for the first time since the late 1970s.
In 1998, the Estrada administration took over the reins of government. The
budget deficit of the national government expanded as spending rose without
any accompanying tax measure. Mr. Estrada was unseated in 2001, but the
imperative to raise tax revenues re-emerged.
The Macapagal-Arroyo administration assumed power in 2001. The tax
effort weakened during this administration, down from 16.9 percent prior to the
1997 Asian financial crisis to 12.4 percent in 2004. And so pressure mounted to
enhance tax collections. In response, the administration caused the enactment of
a Republic Act called the Reformed Value-added Tax, which expanded coverage
and raised the rate to 12 percent from 10 percent in 2007.
In 2010, the tax effort remained a major challenge to the administration of
President Benigno S. Aquino. The latter had shepherded the enactment of a law
in 2012 that restructured taxes on cigarettes and alcoholic beverages and indexed
them to inflation. Accompanied by disciplined government spending, the budget
deficit of the national government declined to about 2 percent of gdp in 2012.
Moving forward, the tax reform program will continue to rely a great deal
on internal, rather than, border taxes. For direct taxation, this strategy calls for
strengthening collection and administration of personal income and of corporate
income taxes. A noticeable trend is the growing share of indirect taxes to total
tax revenue mainly on account of improvements in value-added tax collection.
With the share of indirect taxes increasing, questions about tax incidence
generally crop up. Indirect taxes are considered regressive since the burden on
low-income taxpayers is disproportionately large. As a proportion of earned
income, poor people pay a larger amount in taxes. There is little that can be
done to reduce such burden, except to make government spending progressive,
that is, low-income taxpayers should benefit more from any additional unit of
government spending. In line with making government spending progressive,
the PNoy administration has prioritized some spending programs in the national
73
government budget for low-income households, such as, the “conditional cash
transfer” program for the poor.
2.2. Monetary policy
At this point, there is wide agreement among economists that monetary policy
should be based on rules rather than discretion [Kydland and Prescott 1977]. The
main reason for rules is to minimize the uncertainty and volatility that emanate
from discretionary monetary policy. In doing so, monetary policy minimizes the
risk from various sources that the private sector confronts.
Monetary policy rules in the Philippines have been evolving and undergoing
refinements over time [Canlas 2012]. In the early 1970s, following the decision
to allow the world’s major currencies to go on a generalized float against one
another, the monetary authority followed suit by adopting a flexible exchange
rate system. However, the central bank as monetary authority reserved the right
to intervene in foreign-exchange markets to keep the exchange value of the
Philippine peso against the us dollar within a targeted band. The central bank
observed exchange-rate targeting as a monetary policy rule.
Under exchange-rate targeting, in which the central bank tried to keep the
nominal exchange rate within a narrow band, the system in place resembled a
fixed exchange-rate system [Canlas 2002]. When the central bank embarked
on an expansionary monetary policy, the country’s inflation rate exceeded that
of the us, thereby making imports from the latter attractive. As the demand for
imports increased, the demand for us dollars likewise rose. When the central bank
accommodated, the official foreign reserve assets that it held declined. The erosion
continued amid non-stop inflationary policy. The situation eventually triggered a
speculative attack against the Philippine peso, thereby accelerating the erosion of
official foreign reserves. When the latter fell to a critically low level, the economy
experienced serious liquidity problems, resulting in a balance-of-payments crisis
and an abandonment of narrow exchange-rate targeting.
The central bank then allowed the exchange rate to move within a wider band.
At the same time, it changed its policy stance and embraced monetary-aggregate
targeting as monetary policy rule. One factor that influenced the adoption of such
a monetary policy rule stemmed from the fact that in the 1990s, the Philippines
was still tied to an imf program. In this context, the conduct of fiscal and
monetary policies was based on financial programming techniques that the imf
imposed on client countries experiencing balance-of-payments crises.
From the standpoint of the imf, a balance-of-payments crisis stems from
an excess of aggregate demand over aggregate supply. Demand-management
techniques are thus deemed essential. Inflationary monetary policy is to be
curtailed. Money demand and supply need to balance. Disinflationary monetary
policy means money supply growth has to decline. In the short run, however,
disinflation results in adverse output effects.
74
Disinflation and recession occurred in the early 1980s, after the Philippine
government declared a moratorium on foreign debt servicing and it tapped
a standby loan from the imf for the needed liquidity. Sudden tightening of
monetary (and fiscal) policy led to the recession of 1984-1985.
Monetary-aggregate targeting had its downside. Demand for money could
be unstable at times, making such targeting ineffective. In 2002, the new bsp as
monetary authority adopted inflation targeting.1 Under this monetary policy rule,
the bsp announces a target inflation rate and uses its array of monetary policy
tools to achieve the inflation target. When, for instance, the actual inflation rate
exceeds the target, the monetary authority may cause interest rates to rise to
tighten money supply and curb inflation. It is important for the bsp to maintain
constant communication with the public so that its intentions are properly
communicated and its messages shorn of uncertainty to the extent possible.
Inflation targeting requires that the bsp has both policy and instrument
independence. Policy independence means that the fiscal authority does not
influence bsp policy making. By instrument independence, the bsp must be
equipped with enough tools of its own to achieve its inflation targets. In line with
this independence, an institutional reform in 1993 created an independent bsp.
One of the important provisions of the law calls for a monetary board; majority
of the board’s members are appointed by the Philippine president from the
private sector. Prior to this reform, members of the president’s cabinet who sat in
an ex-officio capacity dominated the monetary board.
A prior issue is this: Is inflation a monetary phenomenon? Friedman said that
“inflation is always and everywhere a monetary phenomenon”. With a quantity
theory of money in mind, if the central bank increases the money supply from
a position of balance, then the real money stock exceeds the demand for it. To
restore balance, the general price level must rise, which means that inflation rate,
defined as the percentage change in the general price level, must rise.
Canlas [1992] tested a quantity theoretic model of inflation in the Philippines.
The theoretical model implies a one-for-one impact of money growth on the
inflation rate. Some econometric techniques—including first-order differencing
for stationarity, finding the optimal lag length, and Granger causality tests—are
put to work. Time-series data are used. For the period 1973-1990, the regression
results show that money growth has a contemporaneous positive effect on the
inflation rate, but it is less than one-for-one. The optimal lag length of the effect
of money growth on inflation using Hsiao’s test [1981] is shown to be one year;
in subsequent regressions, however, the effect of money growth lagged one-year
is insignificant. Overall, there is empirical support for an inflation model based
on the quantity theory of money.
For an exposition on the intellectual development of inflation targeting, see Svensson (1999a, 1999b).
1
75
Since the adoption of inflation targeting as a monetary policy rule in 2002,
the general price performance of the Philippine economy has improved
tremendously. Over the period 2005-2012, the inflation rate averaged only 5
percent each year, except in 2009, when relative-price shocks emanating from
food and energy products intervened and caused the inflation rate to increase to 8
percent that year. Since then, the inflation rate has again moderated to an average
of about 4 percent each year.
Monetary policy can be trusted to promote growth insofar as it dampens
inflationary expectations. As the latter is dissipated, interest rates decline
with salutary effects on investment. An increase in investment is expected
to support growth. However, an increase in physical capital is subject to
diminishing marginal productivity, which in the long run dampens growth. The
challenge, therefore, is to overcome this tendency towards diminishing marginal
productivity in order to generate long-run growth.
3. Towards long-run growth
Proper conduct of macroeconomic policies can be trusted to minimize
unwanted business fluctuation, a phenomenon to which all market-oriented
economies are exposed. Having stabilized the economy, the associated major
challenge is ushering in self-sustaining growth. This section looks at factors that
contribute to long-run growth.
3.1. Investing in human capital
A major force that propels growth in the long run emanates from the
accumulation of human capital, which is commonly associated with the
knowledge and skills, or efficiency units, which people bring to the workplace.
Given diminishing marginal productivity of physical capital, accumulation of
the latter alone is not sufficient to yield positive growth in the long run. The
model of Solow-Swan [1957] emphasized the importance of investment for
growth.2 It is insightful in terms of accounting for growth in a given country,
but it is not designed to account for growth convergence in a large cross-section
of heterogeneous countries. Modern economic growth theory, or what is often
referred to as endogenous growth, has thus extended the Solow growth model in
trying to account for long-run growth in a number of developed countries and to
account for convergence (or lack of it) in a large sample of countries.3
Human capital accumulation takes various forms, including investments
in education, training, health, and nutrition. Health investments enable people
See Canlas [2003] for tests of the Solow model in the Philippine context.
For an exposition on the origins and key doctrines of endogenous growth, see Romer [2000].
2
3
76
to develop their physical skills, such as manual dexterity and visual acuity.4
However, it is education and training that are popularly linked to human capital
investments. Various levels of education, for instance, are credited with helping
raise literacy and cognitive skills and equipping people with knowledge and
scientific skills. Moreover, human capital investment is associated with other
aspects of household behavior, including desired family size and labor force
participation (Becker, Murphy, and Tamura [1990]; Tamura [2006]). As a result,
many countries, especially those that have succeeded in achieving growth in real
gdp over a long-run period, have made it a point to invest in quality education at
all levels.
In empirical studies that seek to determine the sources of growth, much of
the growth stems from total factor productivity (tfp). The latter refers to the
efficiency in use of all factors of production, e.g., labor and capital. The role
of education in enhancing tfp is well understood and widely accepted in the
economics profession. It seems clear that education and training equip people
with efficiency units that are useful in the workplace [Becker 1961, 1964]. On
the job, people accumulate more skills, which further raise their human capital,
as Mincer [1962] emphasized.
At the same time, however, family members’ uses of time include non-market
activities—such as attending to the education, health, and nutritional needs of
children—which lead to the accumulation of human capital that eventually gets
transmitted within the household across generations [Becker 1965]. In countries
that are able to produce across time highly trained and educated workers who
can master production techniques that continuously emerge in a modernizing
economy, aggregate production efficiency is increased in the long run, with
positive effects on tfp.
Following Solow [1957], Denison [1962], and Jorgenson and Griliches
[1967], the growth rate of output is decomposed into the sum of the growth of
labor and capital plus tfp, which is taken as a residual. Early empirical studies
tended to show that much of the growth rate of output stems from growth of tfp.
The latter is viewed as technological progress, which emanates from a variety
of factors, including research and development (r&d) or knowledge production,
and realization of scale economies. Technological progress follows as a matter of
course from knowledge production, which is human-capital intensive and has the
dimensions of a public good [Grossman and Helpman 1989]. That is, knowledge
generated by one firm may spill over to all firms in the industry. And so as tfp
improves, the tendency towards diminishing marginal productivity is overcome,
resulting in positive growth rates of output over a long period of time.
For empirical tests supportive of the positive growth effects of education on output growth using an
endogenous growth model, see Benhabib and Spiegel [1998].
4
77
Attempts to decompose the sources of output growth in the Philippines
shows that in the 1980s and the 1990s, much of the growth of output was due
to increases in labor and capital, while tfp contributed only a small proportion,
as shown in Table 5. In the 1980s, tfp was a minus 1.62 percent. This turned
positive 0.25 percent in the 1990s, accounting for 9.9 percent of the growth rate
of real tfp. Over the period 2001-2006, tfp gained strength and increased to 2.4
percent, which is about 50 percent of real gdp growth rate.
TABLE 5. Total Factor Productivity in the Philippines
Period
Growth of capital
Growth of labor
Total Factor Productivity
1981-1990
2.05
1.37
(1.62)
1991-2000
1.77
0.87
0.25
2001-2006
1.12
1.24
2.41
Source: Table 2.9 in Canlas, Cham, and Zhuang [2009]
To account for the improvements in tfp, it is insightful to note the trends in
educational attainment in the Philippines and its impacts on the skill composition
of the labor force. A rising share of skilled workers normally contributes to
tfp improvements.
3.2. Investing in education
Recognizing the significance of education in the process of economic growth
and development, the Philippine Constitution of 1987 provides for allocating
the biggest proportion of the expenditure program of the national government
budget to education. The national government has long been the major provider
of education in the Philippines, particularly, basic education, which consists
of elementary and secondary education levels. The latter are provided without
out-of-pocket costs to parents who enroll their children of school age in
public schools.
At the tertiary education level, the national government runs a network of
state colleges and universities, wherein tuition fees are charged, but at subsidized
rates. The rest of the demand for tertiary education is met through private
colleges and universities.
Moreover, the government supports skills training through a system of private
post-secondary schools offering technical and vocational education and training.
These private institutions co-exist with government-run institutes.
Table 6 shows the distribution of workers by educational attainment. It is seen
that the proportion of unskilled workers (those with only elementary schooling)
has over the past two decades been declining. Meanwhile, the share of semiskilled to skilled workers (those with at secondary and tertiary schooling) has
been rising.
78
TABLE 6. Employment by educational attainment (in %)
Education
1991
2001
2006
Elementary school graduate
24.2
19.2
16.8
High school graduate
18.1
23.2
24.9
Tertiary and above
10.9
13.2
14.4
Source of basic data: National Statistics Office, Philippine labor force survey,
various issues
Investments in education and training transmit growth through another
mechanism. By narrowing the differentials between earnings, income inequality
is reduced. Persistent income inequality tends to be deleterious to growth in
so far as such inequality raises political pressure to implement redistributive
programs that are financed by raising taxes. Raising taxes on personal and
business incomes tends to erode the marginal productivity of labor and capital
that can be privately captured, with dampening effects on capital accumulation,
both human and physical, and on growth. However, by reducing the proportion
of poor individuals and households, political pressure for redistribution may
be eased.
3.3. Technological progress, trade policy, and growth
It is widely known that growth comes not just by producing increasing
quantities of the same good, but also by producing new products of everincreasing quality [Stokey 1995]. The growth of Apple Company, for example,
comes not from producing the same first-generation Mac Book or iPad but
from making newer and more powerful versions of these products. Moreover,
further growth is obtained not by selling in the same market but by accessing
new markets for an expanding array of new products. Trading countries differ
in their endowments of knowledge capital. The developed countries with large
endowments of knowledge capital do a good deal of r&d and are thus able to
produce and trade a wide range of new products [Grossman and Helpman 1989].
Over time, developing countries that adopt open trade regimes may acquire
production knowledge about some of these new products through imitation,
with salutary effects on output growth. Clearly, technological progress and trade
policies matter a lot for growth.
Trade in intermediate products is one of the fastest growing segments of
international trade. These intermediate products are normally the product of
investments in r&d; they lie at the core of an increasing number of new products
of higher quality. In the crop sector of agriculture, for example, the discovery of
high-yielding seed varieties has significantly raised the yields per hectare of rice
and corn. In information technology, the microchip revolution has enabled the
manufacture of supercomputers with huge memories and storage capacity and
79
allowed various industries in the service trades, such as banking and finance, and
retail and wholesale trade, to exponentially increase their productivity.
The rise of intermediate products emerging from r&d is taking place in a
system involving separable and multi-stage production processes. This has given
rise to outsourcing. Firms in developed countries subcontract to firms in less
developed countries the labor-intensive stages of production, for instance. But
this can also be done under a vertical-integration arrangement if the firm from
the developed country finds it more efficient to do so.
Considering that r&d yields knowledge that spills over to entities beyond
the r&d proponents themselves, there is a public policy concern that it may be
under-produced. Third parties are able to capture the returns from r&d but the
original r&d proponents may not be fully compensated. People tend to invest
only up to the marginal rate of returns that they can appropriate. If the private
returns fall short of the social rate, then there is room for some form of subsidies
to avoid under-investment in knowledge production.
3.4. Philippine trade policy reforms
In 1986, the Philippine government ushered in some trade policy reforms
aimed at integrating the economy with the global economy and make local firms
internationally competitive. To kick-start this reform process, the government
liberalized imports and embarked on a tariff-reduction program. Later on, it also
liberalized entry of foreign direct investments.
In the 1950s, trade policy was based on import substitution. The government
encouraged domestic firms to manufacture goods that at the time were being
imported in large amounts, ostensibly to save on scarce foreign exchange and to
“learn by doing”. Local firms were protected from competing imports through
high import tariffs and quantitative restrictions, using infant-industry arguments.
The firms, however, relied on imported capital equipment and intermediate
products. Since they were selling their products solely in the domestic market,
the goal of saving on foreign exchange was not achieved. Moreover, since the
Philippines was starting with a low human capital base, new products that could
be exported did not emerge. The foreign exchange earnings of the country were
derived largely from exports of agricultural and mineral products. Since these
foreign exchange earnings were not adequate to finance the import needs of
the import-substituting industries, the country experienced recurrent liquidity
problems, accompanied by the collapse of a fixed peso-us dollar exchange rate.
Eventually, reliance on import substitution declined. Import liberalization and
tariff reduction combined to shrink the size of import-substituting industries.
In 1994, the Philippine acceded to the World Trade Organization, which
was anchored on a “most-favored nation principle”. Under this principle,
tariffs and other commercial policies extended to one member country of the
World Trade Organization cannot be withheld from other member countries. In
80
addition, the Philippines has actively participated in regional preferential trading
arrangements. One of these is the asean Free Trade Area-Comprehensive
Effective Preferential Tariff, which envisions non-tariff barriers among member
countries, starting with manufactured products, and later on, extended to trade
in agricultural goods and in services. The arrangement has evolved: In 2015,
asean is committed to instituting a single-market arrangement.
Today the Philippines is a small open economy that is integrated with the rest
of the world economy through trade in commodities, securities, and national
currencies. Greater international labor mobility is also envisioned. The country
adopts a flexible exchange-rate system and allows international mobility of
capital. It has significantly reduced its public debt as a proportion of gdp, and its
sovereign debt papers are nearing investment-grade status.
4. Concluding remarks
Several development observers regard the Philippines at this point as
an emerging economy that is poised to join the Southeast and East Asian
economies that are now considered newly industrializing. This optimism derives
largely from the structural reform program that has been pursued by political
administrations in succession since 1986. The policy reforms may be broadly
classified into the following: short-term macroeconomic policies designed for
stable growth; and long-term structural policy reforms aimed at sustained and
broad-based growth.
The importance of short-run stabilization may be appreciated by taking a
historical perspective. For nearly four decades since the 1950s, growth in the
Philippines was hampered by inappropriate macroeconomic policies that resulted
in unwanted volatility and business fluctuations. Political administrations since
1986 have in succession tried to address these concerns. Beginning in 1986,
following the dismantling of martial-law rule and the restoration of democratic
political institutions, the administration exerted efforts to strengthen the fiscal
position of the national government by instituting a responsible deficit-reduction
program. Monetary policy was made independent with the enactment of a
law that established the bsp as a monetary authority independent of the fiscal
authority. These measures have ushered in economic recoveries from balance-ofpayments crises and gdp declines, while putting a lid on inflation.
Moreover, attention has been paid to growth-enhancing reforms, such
as increased investments in capital broadly defined to include human and
knowledge capital, in an effort to achieve long-run growth. In addition, the
government pursued trade liberalization policies that are anchored on import
liberalization and tariff reduction. The government also liberalized foreign direct
investments, which allowed a hundred percent ownership in several industry
areas; restrictions or limits on foreign ownership were explicit in the so-called
negative lists that over time are being shortened. The privatization of government
81
firms that previously monopolized some industry sectors—such as oil and
petroleum, commercial banking, and telecommunications—was pursued.
The economy has responded positively to these policy reforms. In 2001,
the economy recovered from a political shock and posted a gdp growth that
has been sustained in spite of the intervention of shocks like the September 11
terrorist attack, the global financial crisis of 2008, and the Eurozone debt crisis
of 2012. The economy has achieved sufficient resilience as it became integrated
with the global economy through trade in commodities, securities, and national
monies. Entering the second decade of the 21st century, the economy is expected
to achieve a sufficiently high growth rate in real gdp with stable prices amid
continuing challenges like the weak recovery in the US, the debt problems of
some Eurozone countries, and the growth slowdown in China.
The current administration is restoring credible leadership at the top of the
political ladder, and this is serving the economy well. Not only is the current
political leadership committed to market-friendly policies; it is likewise
aggressive in addressing concerns about good governance, including fighting
corruption and strengthening the legal and judicial system. The latter is vital
to making the infrastructure program of the government at both the national
and local levels succeed. Since the program is anchored on public-private
partnership, contractual performance and sound adjudication in case of
contractual disputes are critical. Investors find these policy commitments quite
appealing, which are necessary conditions to becoming an economic miracle.
Philippine National Oil Company
Alternative Fuels Corporation
References
Barro, R. [1994] “Economic growth and convergence”, cs Press for International
Center for Economic Growth, Occasional Paper no. 46.
Barro, R. and X. Sala-i-Martin [1990] Economic growth. New York: McGrawHill, Inc.
Becker, G. [1964] Human capital: a theoretical and empirical analysis. New
York: Columbia University Press for the National Bureau of Economic
Research.
Becker, G. [1965] “A theory of the allocation of time”, Economic Journal
75:493-517.
Becker, G., K. Murphy, and R. Tamura [1990] “Human capital, fertility, and
economic growth”, Journal of Political Economy 98:S12-S37.
Benhabib, J. and M. Spiegel [1994] “The role of human capital in economic
development: evidence from aggregate cross-country data”, Journal of
82
Monetary Economics 34:143-173.
Canlas, D. [1992] “Inflation in a low-income country: tests based on the quantity
theory of money”, University of the Philippines School of Economics
Discussion Paper no. 9209.
Canlas, D. [1998] “Tax reforms in the Philippines”, Asian Tax Reforms: Issues
and Results, proceedings of an international symposium, Tokyo: Hitotsubashi
University.
Canlas, D. [2002] “Monetary policy and exchange rate in the Philippines”,
Augustine H.H. Tan, ed., Monetary and financial management in Asia in the
21st century. New Jersey: World Scientific.
Canlas, D. [2003] “Economic growth in the Philippines: theory and evidence”,
Journal of Asian Economics 14:759-769.
Canlas, D. [2012] “Business fluctuations and monetary policy rules in the
Philippines: with lessons from the 1984-1985 contraction”, presentation
delivered as the Bangko Sentral ng Pilipinas Sterling Chair in Monetary and
Banking Economics.
Canlas, D., R. Cham, and J. Zhuang [2008] “Development performance and
policy”, in D. Canlas, M.E. Khan, and J. Zhuang eds. chapter 2.
Canlas, D., M.E. Khan, and J. Zhuang eds. [2008] Diagnosing the Philippine
economy: toward inclusive growth. London: Anthem Press.
Canlas, D. [2015] “Investing in human capital for inclusive growth: focus on
higher education”, paper presented on September 22, 2015 in observance of
Development Policy Research Month, Philippine Institute for Development
Studies, Metro Manila.
Denison, E. [1962] “Sources of growth in the United States and the alternatives
before us”, Committee for Economic Development, Supplement Paper no. 13.
Fischer, S. [1993] “Does macroeconomic policy matter? Evidence from
developing countries”, CS Press for International Center for Economic
Growth, Occasional Paper no. 27.
Grossman, G. and E. Helpman [1989] “Product development and international
trade”, Journal of Political Economy 97:1261-1283.
Hsiao, C. [1981] “Autoregressive modeling and money-income causality”,
Journal of Monetary Economics, 85-106.
Jorgenson, D. and Z. Griliches [1967] “The explanation of productivity change”,
Review of Economic Studies 34:249-280.
Kydland, F. and E. Prescott [1977] “Rules rather than discretion: the
inconsistency of optimal plans”, Journal of Political Economy 473-491.
Lucas, R. Jr. [1988] “On the mechanics of economic development”, Journal of
Monetary Economics 22:3-42.
Lucas, R. Jr. [1993] “Making a miracle”, Econometrica, 61:251-271.
Mincer, J. [1962] “On-the-job training: costs, returns, and some implications”,
Journal of Political Economy 70:50-77.
83
National Economic and Development Authority (neda) [2010] A strategic
framework and action plan for inclusive growth. Metro Manila.
Romer, P. [1986] “Increasing returns and long-run growth”, Journal of Political
Economy 94:1002-1037.
Romer, P. [1990] “Human capital and growth: theory and evidence”, CarnegieRochester conference series on public policy 32:251-286.
Romer, P. [2000] “The origins of endogenous growth”, Journal of Economic
Perspectives 8:3-22.
Sharma, R. [2012] Breakout nations: in pursuit of the next economic miracles.
New York: W.W. Norton and Co.
Svensson, L. [1999a] “Inflation targeting as a monetary policy rule”, Journal of
Monetary Economics 607-654.
Svensson, L. [1999b] “Inflation targeting: some extensions”, Scandinavian
Journal of Economics 337-361.
Solow, R. [1957] “A contribution to the theory of economic growth”, Quarterly
Journal of Economics 70:65-94.
Stokey, N. [1995] “r&d and economic growth”, Review of Economic Studies
62:469-489.
Swan, T. [1957] “Economic growth and capital accumulation”, Economic Record
32:334-361.
Tamura, R. [2006] “Human capital and economic development”, Journal of
Development Economics 79:26-72.
PRE
Development progeria: the role of institutions
and the exchange rate
Sarah Lynne S. Daway* and Raul V. Fabella*
Convergence is more the exception than the rule in the
development landscape. As a possible explanation, we posit
development progeria: the phenomenon where a low-income
country exhibits the industrial share dynamics of high-income
mature economies where the Non-traded Goods Sector outgrows
the Traded Goods Sector and the share of the non-traded goods
sector outstrips the share of the traded goods sector. We argue
that this seems to be the case of the Philippines in the last
25 years.
We then inquire into the drivers of this phenomenon. One
possibility is the Rodrik hypothesis: that market and institutional
distortions hamstring the Tradable goods sector more than
they do the Non-tradable goods sector. The other possibility is
the exchange rate policy being favorable or unfavorable to the
Tradable goods sector. Using cross-country data for countries
with per capita income of us$10,000 or less, we show that these
two factors cannot be rejected as drivers of development progeria.
JEL classification: 014, 043, F31
Keywords: Development progeria, institutions, real exchange rate,
low-income economies
1. Introduction
1.1. What is development progeria?
Development progeria is the phenomenon where a low-income economy
exhibits the industry share dynamics considered normal in advanced high-income
economies: that is, where the share of the modern Tradable goods sector falls
while that of the Non-tradable goods rises in the course of development. In
medicine, progeria is a genetic malfunction where afflicted six-year olds exhibit
85
the physical characteristics of sixty-year-olds. Like advanced mature economies
that are characterized by slow growth, economies afflicted by development
progeria also grow slowly and thus have poor prospects for catching up with
mature economies. In contrast, economies on the non-progeriac or convergent
trajectory experience an extended period of growing industrial share of Tradables
coincident with the retreat of Non-Tradables share in total output before they
finally graduate to mature economy status.
1.2. The Philippines in the past 25 years
For the most part of the quarter-century following the 1986 overthrow of the
authoritarian regime of Ferdinand Marcos, the Philippines has chafed under the
moniker “sick man of Asia.” It sank to the bottom of the asean 5 in per capita
growth, rate of poverty reduction, and investment rate [Usui 2011]. Figure 1 shows
the share trajectories of Agriculture, Manufacturing, Industry, and Services.
60.0
50.0
Shares (%)
40.0
Agriculture
Manufacturing
30.0
Industry
Services
20.0
10.0
0.0
’86 ’87 ’88 ’89 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05
Source of basic idea: World Bank Development Index
FIGURE 1. Trajectory of industry sector shares in GDP: Philippines, 1986-2009
It is clear that over the 25-year period from 1986-2009 the share of the
Services sector gained continually while the shares of Industry and Manufacturing
stagnated or lost out. To compare the experience with those of select neighbors
and mature economies in the same period, Figure 2 below, which is taken
from Fabella and Fabella [2010], presents the change in the percentage share
of the Manufacturing, Industry, and Services sectors in the period 1986-1996
by countries.
86
Daway and Fabella: Development progeria:
the role of institutions and the exchange rate
10.0
8.0
4.0
Manufacturing
2.0
Industry
0.0
Ch
i
A
US
ut
h
Ph
ilip
Ko
pi
re
a
ne
s
ala
M
ail
Th
So
-6.0
In
do
4.0
sia
an
d
ys
-2.0
na
ia
Services
ne
Change in % shares
6.0
G
er
m
an
y
-8.0
Source of basic idea: World Bank Development Index
FIGURE 2. Change in percentage shares, 1986-1996
Starting with the mature economies of Germany and the United States, the
Services sector share rose while the shares of both Industry and Manufacturing
for both periods fell. These are archetypes of late mature economy trajectory.
South Korea, a new Organisation for Economic Co-operation and Development
economy, saw its Services sector and Manufacturing shares rise but still with a
slight increase in Industry share. Indonesia and Thailand saw their Service sector
shares falling in the face of rapidly rising Industry and Manufacturing shares in
both periods. This is the archetypal catch-up trajectory. Malaysia exhibited a rising
Service sector share in this period but with rising Industry and Manufacturing
shares. China exhibited the same pattern.
Figure 3 is taken from Fabella and Fabella [2010] and Fabella [2013]. It shows
the trajectories for the second period (1996-2009).
Malaysia reverts somewhat to the archetypal catch-up trajectory with a
falling Service sector share to go with a rising Industry but falling Manufactures
share. China continues the pattern where both the Industry and Manufacturing
sectors gain to go with a rising Service sector share. It is an economy firing on
all cylinders.
87
8
6
Manufacturing
2
Industry
Ch
i
ia
ne
s
ys
pi
ala
re
a
ilip
US
er
m
an
y
So
G
In
do
ne
sia
ut
A
h
Ko
Ph
4
M
ail
an
d
-2
-6.0
Services
na
0
Th
Change in % shares
4
Sources: Fabella and Fabella [2010] and Fabella [2013] (Transactions of the NAST)
FIGURE 3. Change in percentage industry shares, 1996-2009
Finally, we turn our attention to the Philippines. The trajectory exhibited by
the Philippine industrial structure in the last quarter-century mimics that of late
mature economies: in both periods, Industry and Manufacturing shares declined
while Services share rose. By the end of the second period, the Services sector
share stood at the 55.4 percent, a feature common to Organisation for Economic
Co-operation and Development and mature economies. Premature economic
ageing accompanied its journey throughout the 25-year period. This also means
that its prospect for convergence with mature economies is poor.
The question we confront is this: How does development progeria get
engendered? We explore how the growth of the Tradable goods and the Nontradable goods sectors are differentially affected by the exchange rate and the
quality of governance. In the next section, we formally generate the relationship.
2. The model
2.1. A small open economy
We consider a model of the small open economy with two sectors, the tradable
sector T and the non-tradable sector N, each using two factors, labor L and capital
K, where K is a specific factor of production and thus is not mobile across sectors.
We assume full employment of labor, L = LT + LN, where LT is the labor used in T
and LN is used in N. Suppressing K, we can write the production technology of T
and N respectively as:
88
T = Af(LT), f ' > 0 and f " < 0,
N = Bg(LN), g' > 0 and g" < 0,
where A > 0 and B > 0 are positive Hicks-neutral technical parameters. The
equilibrium allocation of L (we assume an undistorted labor market) is given by
the first order conditions:
Af'(LT) = w,
Bg'(L – LT) = w,
which combines to give Af'(LT) = Bg'(L – LT). The latter can be solved for the
equilibrium LT* (and thus LN*). Since progeria is at its core about differential
growth rates, we are interested initially in how T and N will grow as L grows.
Totally differentiating and solving for (dLT*/dL), we get:
(dLT*/dL) = [Bg"]/[Af" + Bg"] > 0.
The algebraic structure of (dLT*/dL) is identical to that of the “power of the
contract” in contract theory, where the efficient allocation of risk dictates that
more risk is shouldered by the agent who suffers least from risk. If g" = 0, or the
marginal productivity of labor in N is constant, but f" ≠ 0, all of the additional
labor will be efficiently absorbed in N. Thus, the sector which suffers the least
fall in marginal productivity from additional hiring should employ most of the
additional labor at equilibrium.
2.2. The power of T
The “power of T” is the power of the tradable goods sector T to efficiently
employ the increase in labor endowment. It is given by (dLT*/dL).
Note that the power of T is the inverse of the capacity of N to efficiently absorb
the additional labor. Thus, additional labor will be efficiently absorbed by that
sector where the wage rate falls less with additional supply.
2.3. Market or institutional distortions and development progeria
It is a known fact that least-developed countries (ldcs) are generally beset
by a myriad of institutional and market distortions. Rodrik [2008] argued and
showed evidence that these market and institutional failures on the supply side
adversely affect T and N differentially, T worse than N. We assume here that T and
N are affected through the technical parameters A and B, respectively.
89
Let the universe of supply side market and institutional failures (but excluding
the labor market) be summarized by a single parameter D > 0 and let A(D) and
B(D) represent the relationship between the Hicks-neutral parameters and D. The
higher is D the smaller are both A and B so that A' <0 and B' < 0. It is understood
that D is lowest in mature developed economies.
While D adversely affects both T and N, T is more adversely affected by D
than is N. We define this differential response: The Rodrik Differential Response
Condition: Suppose D* and D** are two levels of D, D* < D**, then
A(D*)/B(D*) > A(D**)/B(D**)
or in continuous terms,
–[A'(D)/A(D)] > – [B'(D)/B(D].
In other words, the Hicks-neutral productivity of T is pulled down more than
that of N by a rise in D. In terms of the production frontier, a higher D shifts the
production frontier inwards throughout, but a drop in the intercept in the T axis is
more than the drop in the N axis.
We are interested in how a rise in D ceteris paribus affects the composition of
output at equilibrium. The following is shown in another paper [Fabella 2015].
Lemma 1: At equilibrium output basket, a rise in D ceteris paribus will reduce the
labor used in T,
that is, (dLT*/dD) < 0, if the Rodrik Differential Response Condition holds.
Therefore, the share of T in total output will fall and that of N will rise as D
rises. We now turn our attention to the response of relative growth rates of sectoral
outputs when D rises ceteris paribus. In the same paper by Fabella [2015], the
following is further shown.
Lemma 2: (dT/T) – (dN/N) < 0 as D rises ceteris paribus, if the Rodrik Differential
Response Condition holds.
The growth rate of T decreases more than that of N when D rises. When
local market and institutional distortions are large (D high), the share of
Tradables will, over time, be lower and the share of Non-tradables higher. The
more distorted are the market and institutional environments, the more likely is
development progeria.
90
When local market and institutional distortions are large (D high), the share
of tradables will over time be lower and the share of Non-tradables higher than
otherwise similar countries but with lower D. The more distorted is the local
market and institutional environment, the more likely is development progeria
relative to another country otherwise identical. The hypothesis corresponding to
the above follows:
Hypothesis 1: The better is the governance quality in the economy, the higher the
share of the Tradable goods sector in total gdp.
3. The exchange rate and development progeria
Yet another factor that possibly contributes to development progeria is the
value of the domestic currency. Rodrik [2008] has shown that overvaluation is
bad for the growth of low-income countries. Though we will not directly tackle
overvaluation in this paper, we are interested in the impact of the value of the
domestic currency on the industry shares. The production frontier of this simple
open economy given D* is
T(N; D*) = A(D*)fL – g-1[N/B(D*)]
where g-1(.) is the inverse function of g(.) assumed to exist and is convex and
increasing (that is, concave and increasing g(.) is assumed to be a one-to-one
map). The first derivative of T(N; D*) with respect to N is
TN(N; D*)= A(D*)f '[–g-1'/B(D*)] < 0.
The production frontier is well-behaved, i.e., downward sloping and strictly
concave to the origin. Being small and open, the economy is best off producing at
the point in the frontier characterized by
–TN(N; D*) = EPT/PN
where PT is the price of a unit of tradables in foreign currency determined in the
world, and PN is the price of a unit of non-tradables in domestic currency (say,
the Philippine peso) determined in the domestic market, and E is the amount of
domestic currency that exchanges for a unit of foreign currency. EPT/PN is known
as the real exchange rate and is effectively the price of tradables in terms of nontradables in domestic currency units. The higher is (EPT/PN), the lower is the nontradable sector output in the output mix of the economy or the larger is T’s share
in total gross domestic product. The opposite result happens when (EPT/PN) is
lowered ( a real appreciation of the domestic currency).
Hypothesis 2: The higher is (EPT/PN), the higher is the share of T in gdp.
91
4. Estimation results
We employ Blundell and Bond’s [1998] and Windmeijer’s [2005] two-step
system-generalized- method-of-moments (system-gmm) procedure to estimate
the equation below:
yit = βχχit + βZZit + μit ,(1)
where yit is the ratio of the value-added of the tradable sector to gdp in country
i in period t. χit is a vector containing predetermined and endogenous regressors,
which may include the lagged values of the dependent variable; Zit is a vector
of strictly exogenous regressors; and μit is the error term containing the fixedindividual effects.
The main advantage of the two-step system-gmm procedure is that it enables
one to account for the endogeneity of the regressors by allowing one to use the
lagged values of both the dependent and independent variables as instruments. It
also treats the Nickell bias, which is ubiquitous in macro-panel datasets with large
n (cross-section length) and small t (number of periods). Moreover, Windmeijer’s
two-step correction procedure generates more precise and more efficient
estimates, mitigating the finite-sample bias. Furthermore, we favor the two-step
system-gmm procedure over Arellano and Bond’s [1991] difference-gmm model,
as the former is more appropriate for dealing with variables that are or close to
“random walk,” which most, if not all macroeconomic variables are purported
to be. Lastly, information loss due to differencing in unbalanced panel datasets
is less severe under the two-step system-gmm model than under the differencegmm model.
To estimate Equation 1, we employ the ratio of annual manufacturing
value added to gdp as a proxy for the ratio to gdp of the tradable goods
sector, considering the fact that most manufactured goods are tradable and that
manufacturing value added is more readily available in cross-country datasets.
The XX vector comprises the following: lags of manufacturing growth; the
annual growth of the services sector value-added, which proxies for the nontradable sector; the International Country Risk Guide (icrg) index, which is
a measure of the quality of institutions; the ratio of purchasing power parity
conversion factor to the market exchange rate, which substitutes for the real
exchange rate, with a higher ratio implying a decline in the competitiveness of
locally produced goods, thereby hurting the tradable goods sector1; the average
The ratio of purchasing power parity conversion factor to market exchange rate measures the amount of
1
us dollars required in the local economy to purchase the same basket of goods and services that a dollar
can purchase in the United States. Thus, a higher ratio signifies that the same basket of goods and services
becomes relatively more expensive in the domestic economy. We prefer this series over the real effective
exchange rate, as the latter has considerably less observations.
92
tariff rate on manufactured goods, which is a measure of distortion in the
tradable goods sector; and fixed capital formation as a percentage of gdp, which
is purported in theory as the main driver of output growth. We also include the
“power of N,” which is analogous to the power of T and is computed as follows:
ΔLN / ΔL = [(ΔLN / LN) / (ΔL / L)] (LN / L),(2)
which is the ratio of the growth rate of employment in the services sector (the
non-tradable sector) to the growth rate of total employment multiplied by the
ratio of employment in services to total employment. As with the power of T, the
power of N is a measure of the absorptive capacity of the services sector or of its
ability to effectively employ the increase in labor supply. As this is negatively
related to the power of T, we then expect it to negatively affect the manufacturing
to gdp ratio.
The Z vector consists of a measure of tropical land area to account for countryspecific characteristics and of period dummies. This is in line with the “Tropical
Paradox,” which is based on the observation that tropical countries closer to the
tropical zone have lower per capita incomes than countries in more temperate
climates. The period dummies are included to control for time-related shocks.
Except for the icrg index, the rest of the data are downloaded from the
World Development Indicators website. The dataset is an unbalanced panel of 51
developing countries2 with real GNIs per capita of at most US$10,000, spanning
six periods from 1984 to 2013.3 Each period is an average of five years, as is
typical in cross-country growth regressions, to minimize the impact of short-run
fluctuations on the estimated parameters.4
Table 1 below presents the regression results and the instrumentation details.5
With 51 instruments created using the second to fifth lags of the dependent and
the appropriate independent variables, the estimated model passes the requisite
diagnostic tests, i.e., the Arellano-Bond test of ar(2) in first differences and the
Hansen J-tests of over-identifying restrictions. As a further test of the model’s
validity, we also checked that the estimated coefficient of the lagged value of
manufacturing growth lies between the Ordinary Least Squares regression and
fixed-effects estimates. Indeed, the Ordinary Least Squares estimate is 1.19,
which is greater than the two-step system-gmm estimate of 1.18, which is, in
turn, greater than the fixed-effects estimate of 0.60 (see Table A2 in the appendix).
See the appendix for a list of countries included.
We begin with 1984 due to data limitations.
4
Due to the unbalanced nature of the panel, the effective number of periods is only 2.47 on average.
5
The summary statistics of the regression variables are in Table A1 in the appendix.
2
3
93
TABLE 1. Two-step system GMM results
Dependent variable: Manufacturing-to-GDP ratio
Variable
Coefficient
Standard
error
t-statistic p-value 95% confidence
interval
Manufacturing-to-GDP ratio (-1)
1.18
0.03
35.47
0.00
1.11
1.25
-0.33
0.03
-10.15
0.00
-0.39
-0.26
Power of N (-1)
-0.06
0.01
-4.71
0.00
-0.09
-0.04
Services growth rate
-0.13
0.04
-3.34
0.00
-0.20
-0.05
Fixed capital formation as percentage
of GDP
0.09
0.03
3.18
0.00
0.03
0.14
ICRG index
0.05
0.02
2.74
0.01
0.01
0.08
Exchange rate (purchasing power
parity-adjusted)
-0.08
0.01
-7.49
0.00
-0.10
-0.06
Tropical area
-0.51
0.33
-1.58
0.12
-1.17
0.14
Period 3
1.31
1.03
1.27
0.21
-0.76
3.39
Period 4
1.52
0.91
1.67
0.10
-0.31
3.35
Period 5
0.92
0.92
1.01
0.32
-0.92
2.77
Period 6
1.04
0.85
1.22
0.23
-0.68
2.76
Number of observations: 119
Number of groups: 51
Number of instruments: 51
Diagnostic tests
Arellano-Bond test for AR(1) in first differences: Prob > z = 0.006
Hansen J-test of over-identifying restrictions: Prob > chi2 = 0.766
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: Prob > chi2 = 0.527
Difference (null H = exogenous): Prob > chi2 = 0.824
GMM instruments: lagged manufacturing-GDP ratio, 2nd-5th lags
GMM instruments: lagged power of N, lagged services growth, exchange rate, 2nd lag
GMM instruments: fixed capital formation (percentage of GDP), lagged ICRG, 2nd-5th lags
IV instruments: Period 1, Period 2, Period 3, Period 4, Period 5, Period 6, eq(level)
The results in Table 1 provide evidence for the genesis of development
progeria: the tradable sector, i.e., manufacturing, is adversely affected by an
expansion of the non-tradable sector, i.e., services. The power of N is statistically
significant and has the expected negative effect on manufacturing growth.
Moreover, a higher growth rate of the services sector stunts the manufacturing
sector, as is observed in development progeriacs.
94
In line with the theoretical results, icrg, the real exchange rate proxy and
the average tariff rate on manufactured goods have the expected effects on
manufacturing growth. icrg has a positive significant coefficient, which is
consistent with the theoretical result that good institutional quality has a strong
impact of spurring the tradable sector. The coefficient of the real exchange rate
proxy also has a negative significant effect: A higher ratio of the purchasing power
parity conversion factor to market exchange rate implies that a given basket of
goods and services becomes more expensive in the domestic economy than in the
United States, undermining the competitiveness of the tradable sector.
However, the coefficient of “tropical area” is not significant, although it has
the requisite negative sign. This is in line with Rodrik et al. [2002], who find that
the effect of geography on incomes per capita in a sample of 140 countries tends
to vanish once the quality of institutions is controlled for.
As a robustness check, we employ an alternative measure of the power of
N: the ratio of employment in services to total employment in the services and
manufacturing sectors. The expected sign is still negative, as relative employment
expansion in the services sector, which stands in for the non-tradable sector,
would be to the detriment of manufacturing’s own expansion prospects.
Indeed, Table 2 below shows that all the regressors are still significant and
have the expected signs, as in the previous model.6 In particular, the coefficient
of the alternative definition of the power of N is still negative and significant,
albeit much higher in magnitude than that of the previous definition. Moreover,
the model passes all the diagnostic checks even with 56 instruments generated
using the 2nd to 4th lagged values of the pertinent dependent variables.7
TABLE 2. Two-step system GMM results with alternative definition
of the power of N
t-statistic
p-value
95%
confidence
interval
Alternative power of N
Alternative power of N (-1)
Fixed capital formation as percentage of GDP
ICRG index
Standard
error
Variable
Coefficient
0.84
-0.13
-0.17
0.20
-0.06
0.08
0.03
0.03
0.03
0.03
0.03
0.02
0.01
0.01
30.20
-3.78
-5.53
5.79
-2.46
6.25
2.87
0.00
0.00
0.00
0.00
0.02
0.00
0.01
0.79
-0.20
-0.23
0.13
-0.11
0.06
0.01
0.90
-0.06
-0.11
0.26
-0.01
0.11
0.04
The summary statistics of the regression variables are in Table A3 in the appendix.
A caveat, however, is that the coefficient of lagged manufacturing-to-GDP ratio does not fall within the
interval bounded by the Ordinary Least Squares regression and fixed estimates. The Ordinary Least Squares
estimate is 0.78, while the fixed effects estimate is 0.38.
6
7
95
Exchange rate (purchasing power parityadjusted)
-0.04 0.01 -2.97 0.01 -0.06
Average manufacturing tariff
-0.03 0.01 -2.55 0.01 -0.05
Period 3
7.70 2.78
2.77 0.01
2.11
Period 4
9.15 2.71
3.37 0.00
3.69
Period 5
7.75 2.88
2.69 0.01
1.97
Period 6
7.20 2.90
2.49 0.02
1.37
Number of observations: 127
Number of groups: 48
Number of instruments: 56
Diagnostic tests
Hansen J-test of overid. restrictions: Prob > chi2 = 0.774
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
GMM instruments: manuf’g-GDP ratio, alternative power of N, 3rd-4th lags
GMM instruments: services growth rate, 3rd-4th lags
GMM instruments: lagged ICRG, fixed capital formation (percentage of GDP), average
manufacturing tariff, 2nd-3rd lags
IV instruments: Period 1, Period 2, Period 3, Period 4, Period 5, Period 6, eq(level)
interval
95%
confidence
p-value
t-statistic
Standard
error
Variable
Coefficient
-0.01
-0.01
13.28
14.60
13.54
13.02
5. Summary
In this paper, we first define development progeria as a phenomenon where
the industrial share dynamics in low-income countries mimic the industrial
share dynamics in mature high-income economies. This means that the Nontraded goods sector share is growing faster than the Traded Sector share in poor
countries. By contrast, the industrial share dynamics of economies in the catch-up
or convergent trajectory has the Traded goods sector share first outstripping the
Non-traded sector share before shifting to the share dynamics of the mature highincome economies once a high income level is attained.
Development progeria entails slow overall growth for these economies and
a thin chance of a catching up. We argue that the Philippines seems to be so
96
afflicted in the 25 years after the fall of the Marcos regime. We then inquire into
the possible contributors to development progeria. We focus on two factors: the
market and institutional distortions in developing countries which, according to
Rodrik [2008] pull down the Traded goods sector more than they do the Nontraded goods sector growth and shares; and the impact of the value of the local
currency which, when weak, tend to favor the Tradable goods sector and, when
strong, the Non-traded goods sector. We use icrg as proxy for the first and the
purchasing power parity conversion factor to market exchange rate as proxy of
the second.
We show, using the cross-country data from the World Bank Development
Index for countries with per capita income us$10,000 or less and using
Manufacturing as proxy for the Tradable sector and Services for the Non-traded
goods sector, that those two factors after controlling for other factors cannot
be rejected as contributors to development progeria. The better the governance
environment and the more favorable the exchange rate to market rate, the higher
is the share of tradable goods sector share. The variable “power of N” which is the
capacity of the Non-tradable sector to absorb labor and the growth of the Nontradable Sector have each a negative and significant effect on the share of the
tradable sector. Likewise, the higher is the investment rate (proxied by the ratio
of Gross Fixed Capital Formation to gdp), the higher is the share of T in gdp.
That the variable “tropical area” is not significant is consistent with the Rodrik et
al. (2002) result that when institutions and policies are introduced, the equatorial
paradox loses its bite.
*University of the Philippines School of Economics
References
Arellano, M. and S. Bond [1991] “Some tests of specification for panel data:
Monte Carlo evidence and an application to employment equations”, The
Review of Economic Studies 58 (April): 277-297.
Blundell, R. and S. Bond [1998] “Initial conditions and moment restrictions in
dynamic panel data models”, Journal of Econometrics 87: 115-143.
Fabella, R. [2013] “Development progeria: malady and remedy”, Transactions
35(1).
Fabella, R. [2015] “Development progeria: the algebra”, unpublished working
paper. University of the Philippines School of Economics.
Fabella, R. and M.C. Fabella [2012] “Development progeria?”, Chapter 4 in
Looking Back, Moving Forward: 25 Years of BusinessWorld. Quezon City:
BusinessWorld Publishing Corp.
Rodrik, D. [2008] “The real exchange rate and economic growth”, Brookings
97
Papers on Economic Activity. Fall 2008. Available at http://www.brookings.
edu/~/media/projects/bpea/fall-2008/2008b_bpea_rodrik.pdf.
Rodrik, D., A. Subramanian, and F. Trebbi [2002] “Institutions rule: the primacy
of institutions over geography and integration in economic development”,
nber working paper 9305.
Roodman, D. [2006] “How to do xtabond2: An introduction to difference and
system gmm in Stata”, Center for Global Development working paper 103.
Usui, N. [2011] “Transforming the Philippine economy: ‘walking on two legs’”,
adb Economics Working Paper Series No. 252. Available at http://poseidon01.
ssrn.com/delivery.php.
Windmeijer, F. [2005] “A finite sample correction for the variance of linear
efficient two-step gmm estimators”, Journal of Econometrics 126 (May): 2551.
APPENDIX: List of countries included in the regressions
51 countries in Table 1: Albania, Armenia, Azerbaijan, Burkina-Faso,
Bangladesh, Belarus, Bolivia, Brazil, Botswana, Chile, China, Cameroon,
Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, Egypt, El Salvador,
Guatemala, Guyana, Honduras, Indonesia, India, Jamaica, Kazakhstan, Latvia,
Madagascar, Malaysia, Mexico, Mongolia, Morocco, Namibia, Nicaragua,
Pakistan, Panama, Peru, Senegal, Sierra Leone, South Africa, Sri Lanka,
Suriname, Thailand, Trinidad and Tobago, Turkey, Uganda, Uruguay, Venezuela,
Vietnam, Zambia, Zimbabwe
48 countries in Table 2: Albania, Armenia, Azerbaijan, Burkina-Faso,
Bangladesh, Belarus, Bolivia, Brazil, Botswana, Chile, China, Cameroon,
Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, Egypt, El Salvador,
Guatemala, Honduras, Indonesia, India, Jamaica, Jordan, Kazakhstan, Latvia,
Madagascar, Malaysia, Mexico, Mongolia, Morocco, Namibia, Nicaragua,
Pakistan, Panama, Peru, Paraguay, South Africa, Sri Lanka, Suriname, Thailand,
Trinidad and Tobago, Turkey, Uganda, Uruguay, Venezuela, Vietnam
98
TABLE A1. Summary statistics of regression variables in Table 1
Variable
Number of
Mean
observations
Standard Minimum
deviation
Maximum
Manufacturing-to-GDP ratio
119
16.81
6.32
4.85
34.93
119
17.69
6.28
6.20
34.93
119
18.64
6.28
5.95
37.97
Power of N (-1)
119
-0.28
3.79
-38.46
9.73
119
5.20
3.12
-2.03
18.86
Fixed capital formation as percentage of GDP
119
23.00
5.56
10.81
39.37
ICRG index
119
64.99
14.92
-2.83
82.97
119
42.32
14.12
17.17
81.47
Tropical area
119
0.66
0.44
0.00
1.00
Period 1
119
0.00
0.00
0.00
0.00
Period 2
119
0.00
0.00
0.00
0.00
Period 3
119
0.17
0.38
0.00
1.00
Period 4
119
0.23
0.42
0.00
1.00
Period 5
119
0.29
0.45
0.00
1.00
Period 6
119
0.32
0.47
0.00
1.00
TABLE A2. Ordinary Least Squares and fixed effects estimation results
Ordinary Least Squares
Coefficient
p-value
Fixed effects
Coefficient
p-value
1.19
0.00
0.60
0.00
-0.34
0.00
-0.33
0.00
-0.02
0.70
-0.01
0.80
-0.17
0.01
0.11
0.18
Fixed capital formation as percentage of GDP (-1)
0.08
0.03
-0.08
0.17
ICRG index (-1)
0.02
0.18
-0.10
0.19
-0.05
0.00
-0.09
0.03
0.03
0.95
Exchange rate (purchasing power parity-adjusted)
Tropical area
Period 1
0.00
Period 2
0.00
Period 3
0.23
0.67
0.33
Period 4
0.86
0.10
0.00
Period 5
0.30
0.50
-0.73
0.16
Period 6
0.00
-0.62
0.33
Constant
1.66
24.09
0.00
0.00
0.00
0.29
0.45
R-squared:
Within
0.41
0.62
Between
0.97
0.43
Overall
0.93
N
119
0.43
119
99
TABLE A3. Summary statistics of regression variables in Table 2
Variable
Number of
observations
Mean
Standard
deviation
Minimum
Maximum
Manufacturing-to-GDP ratio
127
17.41
6.38
4.85
34.93
127
18.40
6.56
6.24
37.97
Alternative power of N
127
70.28
6.61
52.16
82.56
127
69.06
6.99
47.12
83.87
127
67.91
6.80
44.57
82.14
127
5.09
3.01
-2.03
18.86
Fixed capital formation as percentage of
GDP (-1)
127
22.76
6.19
9.67
45.16
ICRG (-1)
127
65.43
14.51
-2.83
82.97
127
41.74
14.14
17.17
81.47
127
9.02
9.33
0.56
86.87
Period 1
127
0.00
0.00
0.00
0.00
Period 2
127
0.00
0.00
0.00
0.00
Period 3
127
0.18
0.39
0.00
1.00
Period 4
127
0.23
0.42
0.00
1.00
Period 5
127
0.31
0.46
0.00
1.00
Period 6
127
0.28
0.45
0.00
1.00
PRE
Malaysia-China trade: old and new routes
Tham Siew Yean* and Andrew Kam Jia Yi*
The opening up of China and its subsequent rapid integration
into the global economy provided opportunities for Malaysia to
increase trade with China through their respective participation
in global value chain activities. This paper aims to examine the
nature of Malaysia-China trade and the contributory factors to this
bilateral trade pattern and to assess the prospects and challenges
for this bilateral trade, with the new Silk Road initiatives.
The main findings of this paper show increased information
and communications technology and palm oil trade between the
two countries. But the trade pattern also indicates Malaysia’s
trade in electronics and palm oil with China is facing either
stagnating or declining revealed comparative advantage. Instead,
China’s rise has also raised severe competitive pressures as
each country tries to shift up the global value chain. Internal
weaknesses also contributed to the weakening competitiveness
of Malaysia’s electrical and electronics sector so that the share
of this sector’s manufacturing and exports fell substantially,
while the share of resource based manufacturing activities
gained ground.
Malaysia is trying to tap the promises of the new Silk Road
initiative for enhancing its trade and investment relations with
China. However, realizing this potential faces several hurdles,
such as the viability of this initiative and internal challenges
within China and Malaysia. The most urgent of the three
challenges is for Malaysia to overcome its domestic challenges,
since without the requisite domestic changes, no investor,
including China, will venture into the country.
JEL classification: F10, F14, N75
Keywords: Malaysia, China, trade, Silk Road, electronics, palm oil
101
1. Introduction
Malaysia’s strategic geographic location in Southeast Asia has facilitated its
trade between the East and West since the time of the Malacca Sultanate in the
14th and 15th century a.d. Malacca’s importance as a center for entrepôt trade
along this old trade route led to close relations between the Malacca Sultanate
and several countries, including China. This in turn paved the way for the
establishment of Malacca-China diplomatic relations and the arrival of traders
from China to Malacca. The subsequent waves of migration of Chinese workers to
Malaysia and a growing Chinese population led to a significant share of Chinese
people in Malaysia’s total population (38 percent)1 in 1957.
Although bilateral relations between Malaysia and China waxed and waned
over time, Malaysia became the first member country of the Association of
Southeast Asian Nations (asean) to normalize relationships with China in
1974, thereby laying the foundation for cordial bilateral relations, including trade
relations, between the two countries. Not surprisingly, Malaysia’s first free trade
agreement is the asean-China Free Trade Agreement, which was signed in 2002.
The growing importance of China as a trade partner of Malaysia is manifested in
the upward trend of exports and imports between the two countries since 1990,
with the exception of a dip during the years of the 2008-2009 global financial
crisis. In fact, China became Malaysia’s largest trading partner since 2010. Within
asean, Malaysia has been China’s largest trading partner since 2008. It is also
China’s third-largest trading partner in Asia, after Japan and South Korea, despite
the relatively small size of Malaysia’s economy. Both sides have pledged to aim
for a trade volume of us$160 billion by 2017.2
Since China’s reforms in 1978 and the opening up of its economy in the
1990s, China has grown to become the second-largest economy in the world
and has joined the upper middle income economies in 2012. Demand from
China, including the demand from its rising middle-income group, has and will
undoubtedly continue to influence global and regional trade. China’s increasing
economic size and grand plans, as manifested in the new Silk Road or the “One
Belt, One Road” development strategy and the Asian Infrastructure Investment
Bank in 2014, and its trade ambition may offer a window of opportunity for
Malaysia to deepen its economic relations with China in the coming years. This
is especially important at this juncture of Malaysia’s economic development
as domestic growth is faltering in the midst of restructuring problems, political
unrest, and global uncertainties.
The objectives of this paper are twofold. First, we seek to examine the nature
of Malaysia-China trade and the contributory factors to this bilateral trade pattern.
This figure is an estimate by Hirschman [1980].
This figure is an estimate by Sta Maria [2014].
1
2
102
Tham Siew and Kam: Malaysia-China trade: old and new routes
Second, we assess the prospects and challenges for Malaysia-China trade, given
the new Silk Road initiatives. The paper is organized as follows. In the second
section, we examine two important sectors in the bilateral trade between Malaysia
and China. In the third section, we look into the prospects and challenges of the
new routes. In the final section, we summarize the main findings of this paper.
2. Two major sectors
Despite the importance of Malaysia in China’s trade, the bilateral relations
between these two countries are not well explored, except as part of asean’s trade
with China (see, for example, Mendoza et al. [2015]). The only recent study that
examined the bilateral trade relations is that of Lee [2014] which examined it from
China’s perspectives, namely within the context China’s economic engagement
with Southeast Asia. Lee’s focus is on the electrical and electronics sector where
he finds increasing competition between the two countries due to weaknesses in
Malaysia’s economy and China’s move up the manufacturing value chain.
While the bilateral trade relation is very much focused on electrical and
electronics, palm oil constitutes another important export from Malaysia to China.
China is Malaysia’s largest export destination for palm oil products, taking up 20.4
percent of Malaysia’s global exports. It is therefore pertinent to include this sector
in analyzing the trade links between Malaysia and China.
The paper will focus on information and communications technology (ict)
and palm oil trade between Malaysia and China, given the importance of these
two sectors in the bilateral trade as well as Malaysia’s policy focus in these
two sectors.3
2.1. Information and communications technology
Within manufacturing, the ict4 sector stands out as the main contributor to
the bilateral trade between the two countries. Table 1 shows that ict goods (on
average) contribute approximately half of Malaysia’s total manufacturing exports
and import to/from China from 2000 to 2013. The bulk of Malaysia’s exports
of ict goods are in electronic components. In 2013, electronic components
comprise 48 percent of total manufacturing exports to China. The second-largest
share in ict exports is contributed by computer and peripherals, averaging at
around 19 percent for the period shown. However, its export share has fallen by
more than half from a peak of 28 percent in 2001 to 11 percent in 2013. Imports
from China are mainly in computer and peripherals (average of 21 percent for
the period shown in Table 1) while imports of electronics component contribute
3
See Tham and Kam [2015] for a summary on the information communications technology policies in
Malaysia as a driver for the economy since the Seventh Malaysia Plan (1996-2000) onwards.
4
ict product classification is based on oecd [2011].
103
an average of 14 percent for the period shown. By 2013, the share of imports
of electronics components and computer and peripherals are 20 percent and 10
percent, respectively.
Table 1 also indicates a diverging pattern between the import and export of
Malaysia’s ict goods with China. The share of ict imports in total manufacturing
imports can be seen to decrease steadily from 2003 to 2013, excluding the year
of the 2009 global financial crisis. On the other hand, exports show a converse
pattern with the share of ict exports to China falling before the global financial
crisis from 2001 to 2007 and increasing after the global financial crisis to 67
percent in 2010 before falling slightly annually to 61 percent in 2013.
The share of electronics exports fell from 19 percent in 2000 to 12 percent
in 2003, after which it fluctuated from 11 percent in 2004 to 14 percent in 2007.
However, its share increased steadily from 37 percent in 2010 to 48 percent in
2013. This is accompanied by an increasing share of electronics components in
total manufacturing imports from 2010 to 2013. Notably, although Malaysia has an
overall trade deficit in manufacturing exports with China, Table 1 shows a surplus
in Malaysia’s trade in overall ict and in the electronics sub-sector after 2008.
TABLE 1. Information communications technology (ict) import and export of
Malaysia to China, 2000-2013 (share of manufacturing export, percentage)
0.03
-0.13
-1.25
-2.18
-3.37
-4.30
-4.40
-3.96
-3.24
2.06
4.46
4.60
2.01
0.51
10.0
12.0
4.0
2.0
3.0
3.0
3.0
2.0
3.0
5.0
5.0
5.0
5.0
5.0
5.0
7.0
15.0
30.0
34.0
30.0
30.0
27.0
25.0
23.0
24.0
15.0
12.0
10.0
10.0
21.0
7.0
6.0
4.0
4.0
3.0
2.0
2.0
2.0
2.0
2.0
2.0
1.0
1.0
1.0
3.0
10.0
14.0
14.0
13.0
14.0
16.0
14.0
13.0
10.0
16.0
12.0
12.0
18.0
20.0
14.0
7.0
5.0
4.0
4.0
6.0
5.0
3.0
3.0
3.0
3.0
6.0
6.0
4.0
4.0
4.0
41.0
52.0
55.0
57.0
56.0
56.0
50.0
46.0
41.0
50.0
41.0
37.0
38.0
39.0
47.0
Source: Authors’ calculations using OECD 2011 classifications. See Appendix 1.
Trade balance,
electronics (US$
billions) (X-M)
55.0
54.0
44.0
32.0
34.0
35.0
34.0
36.0
46.0
71.0
67.0
65.0
62.0
61.0
50.0
Total ICT as share of
manufacturing import
Trade balance, overall
ICT (US$ billions)
(X-M)
5.0
4.0
2.0
3.0
4.0
3.0
2.0
1.0
1.0
1.0
1.0
1.0
1.0
0.0
2.0
Miscellaneous
Total ICT as share of
manufacturing export
19.0
16.0
18.0
12.0
11.0
14.0
11.0
13.0
12.0
52.0
37.0
44.0
44.0
48.0
25.0
Electronics
components
Miscellaneous
2.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Consumer electronics
Electronics
components
22.0
28.0
21.0
15.0
17.0
16.0
18.0
19.0
29.0
16.0
26.0
18.0
15.0
11.0
19.0
Computer and
peripherals
Consumer electronics
7.0
3.0
1.0
0.0
1.0
2.0
2.0
2.0
3.0
1.0
1.0
1.0
1.0
1.0
2.0
Communication
equipment
Computer and
peripherals
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Average
Imports
Communication
equipment
Exports
-0.02
-0.13
-0.20
-0.43
-0.91
-1.20
-1.36
-1.09
-0.99
4.60
4.18
5.78
3.80
3.38
104
The concentration of electronics exports to China can be attributed to
China’s rapid integration into the global value chain in this sector since opening
its economy to foreign direct investment in the 1990s (Lee [2014]; Kong et al.
[2015]). Both Malaysia and China used foreign direct investment policies to
attract foreign multinational companies in the electrical and electronics sector for
the development of their respective manufacturing sector (Xing [2013]; Tham and
Kam [2015]) and are linked to each other’s production through the global value
chains in this sector.
It can be seen that Malaysia’s exports of electronics goods is more concentrated
in parts and components, rather than finished goods (Figure 1). However, although
exports of parts and components increased rapidly before the global financial
crisis, its increase has slowed down after the crisis. For example, in the four years
before the global financial crisis, exports of parts and components to China grew
by 156.6 percent but subsequently grew by only 18 percent from 2009 to 2013,
due in part to the global slowdown after the global financial crisis.
12
10
Parts and
components
8
6
4
2
Finished
goods
13
12
20
11
20
10
20
09
20
08
20
07
20
20
06
05
20
20
04
03
20
20
02
01
20
20
20
00
0
Note: Parts and components and finished goods are aggregated based on Aldaba [2015].
Source: Calculated from COMTRADE
FIGURE 1. Malaysia’s electronics exports to China by parts and components
and finished goods, us$ billions
Bilateral revealed comparative advantage5 in Figure 2 shows Malaysia having
a comparative disadvantage in overall electronics exports to China prior to 2005.
After 2005, the revealed comparative advantage for overall electronics increased
steadily to a value that is above one, barring the year of the global financial crisis.
5
Bilateral rca = (Exports of product Xi from Malaysia to China/All manufacturing exports of Malaysia
to China) / (All exports of Xi from Malaysia/All manufacturing exports of Malaysia). A value above one
represents comparative advantage of product Xi in China’s manufacturing sector.
105
The revealed comparative advantage for all electronics components is above one,
while the converse holds true for finished goods. This suggests that the electronics
sector in Malaysia may have undergone some form of industrial upgrading in
order to compete in the Chinese manufacturing market.
The emergence of wafer fabrication and r&d centers is an indicator of
functional upgrading in a country. According to Shan and Rasiah [2015], foreign
multinational companies have established four chip design centers, one r&d
support facility, five wafer fabrication plants, and 28 assembly and test plants,
while national firms have two wafer fabrication plants and seven assembly and test
plants (Table 2). They attribute the “shift to wafer fabrication” to the investment
grants offered to foreign firms. Therefore, a few firms such as Infineon, osram,
and on Semiconductor have embarked on wafer fabrication in Malaysia since
2005. China is also rapidly upgrading, as shown in Table 2. The race to move
up the global value chain in both countries has led to a slight stagnation in the
revealed comparative advantage for parts and components electronics observed
after the global financial crisis (Figure 2).
1.8
1.6
1.4
1.2
All parts and
components
electronics
1.0
0.8
Overall
electronics
0.6
0.4
All finished
electronics
0.2
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
0
Source: Calculated from COMTRADE
FIGURE 2. Revealed comparative advantage for electronics exports to China,
2000-2013
106
TABLE 2. Integrated circuit firms facing upgrading in Malaysia, China,
and Vietnam, 2011
Chip design
Support R&D
Wafer
fabrication
Assembly and
test
R&D
Chip design
Support R&D
Wafer
fabrication
Assembly and
test
Foreign companies
R&D
National companies
China
1
3*
2#
25
34
0
11
8
6
58
148*#
Malaysia
0
0
0
2
6
0
4
1
5
25
43
Vietnam
0
1
0
1
0
0
3
1
0
2
8
Countries
Total
Notes: Firms are defined by the registration status, and hence, some subsidiaries of the same firm are counted
more than once.
* Denotes firms having both R&D and chip design in the same registration premises;
# Denotes firms having both R&D and other supportive R&D in the same premises.
Source: Shan and Rasiah [2015]
2.2. Palm oil
There are two main sub-sectors dominating palm oil exports to China, reports
the Malaysian Palm Oil Board: palm oil, palm oil kernel, and finished products;
and oleochemicals and biodiesel. In Figure 3a, it can be seen that the largest
contributor to exports is palm oil, palm oil kernel, and finished products. Figure 3a
also shows that exports of both products have increased substantially since 2000.
This is essentially contributed by the increasing demand from China, which does
not produce any palm oil. The nation uses the oil in its food processing industry,
especially in the production of instant noodles, as palm oil is relatively cheaper
than soybean oil and rapeseed oil. Figure 3b shows the trend in the exports of
palm oil products—excluding palm oil, palm oil kernel, and finished products—
and oleochemicals and biodiesel due to the significant difference in value for the
latter two products. Soaps and surfactants show increasing trade value since 2008,
as shown in Figure 3b.
In terms of imports, oleochemicals and biodiesel comprise the largest import
value from China (Figure 3c). Figure 3c further shows that oleochemical and
biodiesel imports have increased sharply since 2012, amounting to over US$770
million in 2014. Other palm oil imports from China are also increasing gradually
over the years as well (Figure 3d).
One important distinction between manufacturing and the palm oil exports is
the trade balance. While Malaysia has a trade deficit in the manufacturing sector,
trade in palm oil products has always been at a surplus since 2000. Figure 4a
shows the surplus from Malaysia’s palm oil trade with China increasing steadily
until 2011, after which the surplus narrowed slightly. ict and electronics trade
also show trade surpluses only from 2008 to 2013. Figure 4b shows that all
107
subsectors of palm oil have generated trade surpluses for Malaysia since 2000.
This surplus is led by palm oil, palm oil kernel, and finished products, followed
by oleochemicals and biodiesel.
b) Exports to China, US$ millions
a) Exports to China, US$ millions
5000
120
4500
4000
100
3500
14
20
12
20
20
20
20
20
20
20
20
10
0
08
20
0
06
40
1000
500
04
1500
02
60
00
2500
2000
0
20 0
0
20 1
0
20 2
0
20 3
0
20 4
0
20 5
0
20 6
0
20 7
0
20 8
0
20 9
1
20 0
1
20 1
1
20 2
1
20 3
14
80
3000
Palm oil, palm oil kernel,
and finished products
Palm kernel cake
Oleochemicals and
biodiesel
Soaps and surfactant
c) Imports from China, US$ millions
Palm kernel cake
Other palm products
20
900
d) Imports from China, US$ millions
18
800
16
700
14
12
600
500
10
20
0
20 0
0
20 1
0
20 2
0
20 3
0
20 4
0
20 5
0
20 6
0
20 7
0
20 8
0
20 9
1
20 0
11
20
1
20 2
1
20 3
14
20
14
12
20
20
20
0
20
0
20
0
20
10
0
08
2
0
20
06
4
100
4
6
200
2
8
300
0
400
Palm kernel cake
Oleochemicals and
biodiesel
Palm kernel cake
Other palm products
Source: Computed from COMTRADE
FIGURE 3. Palm oil trade between Malaysia and China, US$ millions
108
a) Trade surplus across industries, US$ millions
8000
6000
Electronics
4000
2000
0
ICT
-2000
-4000
Palm oil
-6000
-8000
Manufacturing
-10000
13
12
20
11
20
10
20
09
20
08
20
07
20
06
20
05
20
04
20
03
20
02
20
01
20
20
20
00
-12000
b) Trade surplus by palm oil products, US$ millions
5000
4500
4000
3500
Palm kernel cake
3000
2500
2000
Oleochemicals and biodiesel
1500
1000
500
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
0
Other palm products
Note: * “Electronics” is based on consumer electronic equipment and electronic components in Appendix 1.
FIGURE 4. Trade surpluses/deficits, US$ millions
However, Figure 5 provides some cautionary notes on the performance of
Malaysia’s palm oil exports to China. Since 2000, the revealed comparative
advantage for overall palm oil has been declining, where in 2013, it is considered
as a product with marginal comparative advantage in China (slightly above one).
The declining trend is striking as it holds true for all the product groups—except
for palm kernel cake, which does not show a comparative advantage for the period
shown. Even though the main palm oil products (palm oil, palm oil kernel, and
109
finished products) still have a comparative advantage in the Chinese market, the
second most traded product (oleochemicals and biodiesel) has lost its advantage
since 2005. Other palm products have lost their comparative advantage since
2007, while soaps and surfactants became comparatively disadvantaged in 2013.
4.00
3.50
Overall
3.00
2.50
2.00
Palm kernel cake
1.50
Oleochemicals and biodiesel
1.00
0.50
13
12
Other palm products
20
11
20
10
20
09
20
08
20
07
20
06
20
05
20
04
20
03
20
02
20
01
20
20
20
00
0.00
FIGURE 5. Revealed comparative advantage for palm oil products, 2000-2013
The drop in comparative advantage for Malaysia’s palm oil exports may be
due to excess supply as the rise in global demand and high commodity price
had encouraged production to escalate. The average annual growth in palm oil
production from 1993-2012 was 7.3 percent; the same figure for palm kernel oil
production in the same period was 6.3 percent [Basiron 2013]. Total world supply
outstripped world demand for every year shown in Table 3, except for 2009.
In 2013, for example, the growth of palm oil production in both Indonesia and
Malaysia in 2013 surpassed palm oil demand and resulted in a high inventory and
a price decline.
TABLE 3. Palm oil supply and demand, 2005-2012 (million tonnes)
2005
2006
2007
2008
2009
2010
2011
2012
Supply
33.4
36.0
37.7
42.6
45.3
45.9
50.1
51.2
Demand
32.7
35.4
37.4
41.3
45.4
46.5
48.7
51.0
Source: European Delegation to Malaysia, 2012
110
In addition, there was also increasing competition from Indonesia as it changed
its export duty structure to favor production and export of processed palm oil
[European Delegation to Malaysia 2012]. Hence its export duty on crude palm oil
increased while the export duty for processed palm oil products went down from
25 percent to 13 percent. This posed considerable competition to Malaysia as 68.8
percent of its export revenue is from the processed product segment.
3. Prospects and challenges
The new Silk Road was announced by the Chinese Premier in 2013 during a
visit to Kazakhstan. This was subsequently renamed “One Belt, One Road” as
it encompasses two routes: a land trade route linking China with Central Asia,
Russia, and Europe; and a second route or the Maritime Silk Route, stretching
from Fujian on China’s coast, through the Malacca Straits, around the horn of
Africa, and ending in Venice. This second route is of particular relevance to the
maritime countries in Southeast Asia, including Malaysia.
The “One Belt, One Road” initiative promises to enhance trade links along the
area covered by this initiative. This is estimated to include about 50 percent of the
world’s gdp and approximately the same for global trade [Bloomberg Brief 2015].
The use of the Asian Infrastructure Investment Bank, also proposed by China,
to fund the infrastructure needs along these routes envisages a reduction in
transportation costs that should stimulate more trade. This is backed by a us$40billion Silk Road Fund and a promise of us$50 billion for the Asian Infrastructure
Investment Bank and the proposed new brics New Development Bank by the
Chinese government.
Prior to the announcement of the “One Belt, One Road” initiative, the
Malaysian government has been actively courting closer economic cooperation
with China. In 2012, the China-Malaysia Qinzhou Industrial Park was
established. Then followed the establishment of the Malaysia-China Kuantan
Industrial Park in 2013, which started operations in the same year. Sta Maria,
the Secretary General of the Ministry of International Trade and Industry in her
keynote address at the Symposium on the 40th Anniversary of China Malaysia
Diplomatic Relations on May 22, 2014, noted that this is the first time a twin park
model is introduced anywhere in the world, serving complementary roles with
the same principles. The Chinese Central Government has approved a rmb2.4
billion development fund for Qinzhou Industrial Park, over and above the rmb1
billion pledged by the Government of Guangxi, where Qinzhou Industrial Park
is located. In turn, the Malaysian Federal and State Government have allocated
rm700 million to Malaysia-China Kuantan Industrial Park [Sta Maria 2014].
While Qinzhou Industrial Park will focus on food processing, biotechnology,
and information technology, Malaysia-China Kuantan Industrial Park will host
businesses engaged in steel manufacturing, aluminium processing, and palm
oil refinery. This initiative is in line with Malaysia’s ambitions to use this park
111
to increase Chinese investment in Malaysia [Khor 2013]. China has also been
upgrading the Kuantan port with Beibu Gulf Holding (Hong Kong) Co. Ltd given
a 38 percent equity share in a consortium that received a 30-year concession to
manage, operate and develop the port as part of its move to secure more seaports
[Pitlo III 2015]. Port Klang and Singapore’s port are key ports in the Maritime Silk Route
since the Straits of Malacca is a strategic passageway from China to the West.
Port Klang is reportedly moving towards establishing sister-port relationships
with seven ports in China—namely, Shanghai, Ningbo, Tianjin, Guangzhou,
Xiamen, Dalian, and Fuzho—in order to enhance its status in the new
Silk Road.
Malaysia also signed the Five Year Programme for Economic and Trade
Cooperation with China in October 2013. The agreement aims to enhance
bilateral cooperation in agriculture, energy, and mineral resources, information
and telecommunications, manufacturing, infrastructure and engineering,
tourism, logistics, and retailing. This program aims to engage the private
sector to stimulate investment and trade, thereby signaling again Malaysia’s
intentions to attract more Chinese investment into the country.
Economic cooperation has recently expanded into the financial sector. Last
year, the two countries’ central banks agreed to establish a yuan clearing bank in
Kuala Lumpur [Chew 2014]. This is part of China’s quest to internationalize the
renminbi, and Malaysia is the second country in Southeast Asia, after Singapore,
to host a yuan clearing bank.
Apart from these bilateral initiatives, Malaysia is also part of the aseanChina Free Trade Agreement as well as the ongoing negotiations for a Regional
Comprehensive Economic Partnership that includes the free trade agreement
partners of asean, including China.
All these initiatives indicate Malaysia’s strategic interests to tap China’s
ambitious plans for catalyzing its own economic performance. This is especially
important for Malaysia in view of the sharp fall in private investment after the
Asian financial crisis and the moderation in its growth performance. Nevertheless,
the prospects of tapping the economic potential of the “One Belt, One Road”
initiative face numerous challenges, including domestic challenges within China
and Malaysia. Ghosh [2015] points out the need for cooperation and partnership
of the countries along the route for the maritime silk route to succeed, which is
not likely to be forthcoming since opaqueness of the plan in general tends to breed
suspicion rather than trust. More importantly, it remains to be seen as to how the
proposed route can coexist with the existing system of maritime cooperation.
3.1. China shifts to slower growth
China has been growing at an average rate of 10 percent per annum for the
last three decades. Clearly, a slowdown in growth, estimated to be seven percent
112
for 2015, is inevitable. Since the rapid growth in the past has generated excess
production capacity, dissipated its low-wage advantage, increased resource
constraints, and led to considerable environmental damage, China will have to
manage an internal transition to a slower—but more sustainable growth rate.
Specifically, China needs to rebalance its economy from export-led to domestic
consumption led. This will shift China from being an export platform to a market
provider so that its trade surplus will also decrease. Although the most recent
International Monetary Fund review indicates an increasing role in domestic
consumption driving growth in China and a reduction in the current account
surplus, it remains to be seen if this rebalancing can be sustained in the longer
term [imf 2015].
Given that past overinvestment has contributed to the rising demand of China
for commodities and the subsequent commodity price boom over the last decade,
the shift to lower growth and reduced demand for commodities is expected to
have significant spillovers on primary commodity-producing economies like
Malaysia. Palm oil prices are trending downwards, and this can affect Malaysia’s
future trade surplus, which has also been trending downwards after the global
financial crisis.
3.2. Malaysia restructures its economy
Economic restructuring is the key economic problem in Malaysia. Within
manufacturing, resource-based industries have been the largest growth drivers of
this sector from 2002-2012 due to the high commodity prices during this period
[Rizwan et al. 2014]. The sector grew at 6.2 percent on a compounded annualized
growth rate during this period, compared to the 1.5 percent growth in the
electrical and electronics sector. It became the largest manufacturing sub-sector
after 2005. Likewise, exports of resource-based industries grew at 10.3 percent on
a compounded annualized growth rate basis during the same period. Its share in
total manufacturing exports grew from 30 percent in 2002 to 51 percent in 2012,
while exports of electrical and electronics fell from 70 percent to 49 percent over
the same period.
Both electrical and electronics and palm oil industries face multiple challenges
in their respective structural transformation. Deepening the electrical and
electronics sector from low value added to high valued-added activities has been
emphasized since the Second Industrial Master Plan (1996-2005). The plan
advocated the use of a cluster concept for propelling manufacturing development
forward to achieve economies of agglomeration. Unfortunately, the plan was
deemed to be poorly executed [Danaraj 2011]. Since then, there are further efforts
to shift the electrical and electronics sector towards higher value-added activities,
but as noted in Tham [2015], a critical mass of local firms in the electronics sector
has yet to be achieved (see Table 2). Although small, incremental changes have
taken place within this sector, including the emergence of a few local design
113
houses as indicated earlier, they lack the critical mass to be catalytic in the desired
shift. Thus, Malaysia has yet to make critical leap up the value chain, as observed
in the Taiwan and South Korea’s respective electrical and electronics sector.
Instead, Malaysia’s electrical and electronics sector continues to be dogged with
a shortage of talents that is exacerbated by brain drain and inflows of less skilled
workers. Weak industry-university linkages and inadequate r&d spending of 1.13
percent of its gdp, compared to China’s 1.98 percent of its gdp, in 2012 are other
contributory factors to the slow deepening of the electrical and electronics sector.
The palm oil sector is currently more concentrated in upstream and midstream
activities rather than downstream. Similarly, palm oil export is highly biased
toward the upstream and midstream segments, with a total contribution to the
overall industry at 74 percent, while downstream contribution came in low at 17
percent in 2013 [Khoo 2014]. The industry faces multi-faceted challenges such
as limited land, stagnating yields, dependency on foreign labor, trade wars from
producers of other oils, increasing competition from Indonesia, and environmental
concerns. Therefore, upstream activities need to enhance their productivity. The
government is pushing for a shift to the more lucrative downstream segment, but
this will require more private investment that is holding off largely due to the
negative perceptions associated with numerous unresolved financial scandals that
have emerged of late. In addition, there are ongoing leadership struggles within
parties and nationally, while public acceptance of a one-party dominant political
system continues to weaken [Alagappa 2015]. Fundamentally, there is a critical
need for the country to politically shift from the “well-trodden but increasingly
contested bases of race and religion” [Algappa 2015:72] discourse that is
emphasized by the dominant party of the country.
While the government is also pushing for a shift towards services, exports in
the country are still dependent on manufacturing. A shift to services will also face
the same talent shortage challenges as in the manufacturing sector, if knowledgeintensive services are the focus. It will also face the same lack of investor
confidence due to the ongoing political deadlock. Economic restructuring will
be further derailed if the current political challenges are not surmounted in the
near future.
4. Conclusion
The opening up of China and its subsequent rapid integration into the global
economy provided opportunities for other emerging countries to have increased
trade with the nation through their respective participation in global value chain
activities. This is best reflected in Malaysia’s ict trade with China. Similarly,
China’s rise increased the demand for natural resources like palm oil, thereby
contributing to the high commodity prices in the last decade and the growing
importance of palm oil trade between Malaysia and China. But the bilateral
trade patterns shown in this paper indicate that Malaysia is facing stagnating or
114
declining revealed comparative advantage in its exports of electronics and palm
oil to China. This does not augur well for the future of Malaysia-China trade
relations, from Malaysia’s perspective.
Instead, China’s rise also poses severe challenges to Malaysia through
competitive pressures as each country tries to shift up the global value chain.
Internal weaknesses also contributed to the weakening competitiveness of
Malaysia’s electrical and electronics sector so that the share of this sector in
manufacturing and exports fell substantially, while the share of resource based
manufacturing activities gained ground.
Malaysia is trying to tap the promises of the “One Belt, One Road” initiative
for enhancing its trade and investment relations with China. Realizing this
potential faces several hurdles, including the viability of this initiative and internal
challenges within China and Malaysia. The most urgent of the three challenges for
Malaysia is to overcome its domestic challenges; without the requisite domestic
changes, no investor, whether domestic or foreign, will venture into the country.
Overcoming domestic economic challenges, however, will require Malaysia’s
current political development to move beyond its racial and religious paradigms
towards a more inclusive paradigm that embraces multicultural diversity and
religious tolerance.
*Institute of Malaysian and International Studies,
Universiti Kebangsaan Malaysia
The authors thank Professor Hal Hill for his many years of mentoring that has helped each
of them to grow as professional economists. They hope Prof. Hill will enjoy reading this
paper on one of his pet issues.
References
Alagappa, M. [2015] “Asia’s next great challenge: to balance growth and political
development”, Global Asia 10(3), Fall: 68-78.
Aldaba, R.M. [2015] “The Philippines in the electronics global value chain,”
paper presented at the Second eia Workshop on “Leveling up Southeast
Asia’s value added”, April 24 to 26, 2015, Okura Hotel, Japan.
Basiron, Y. [2013] “Malaysian supply and demand of palm oil: challenges and
opportunities until 2020”, http://www.mpoc.org.my/upload/Oil-WorldOutlook-Conference-2013-Hamburg.pdf. Accessed 15 October 2015.
Bloomberg Brief [2015] “One Belt, One Road: assessing the economic impact
of China’s new Silk Road”, http://www.bloombergbriefs.com/content/uploads/
sites/2/2015/07/SC_062615-obor.pdf. Accessed 15 October 2015.
Chew, Elffie [2014] “Malaysia’s central bank works with pboc to start yuan
clearing”, http://www.bloomberg.com/news/articles/2014-11-10/malaysia-s-
115
central-bank-works-with-pboc-to-start-yuan-clearing. Accessed 15 October
2015.
Danaraj, N. [2011] “Technology policy: a critical appraisal”, in Malaysia: policies
& issues in economic development. Kuala Lumpur: Institute of Strategic and
International Studies.
European Delegation to Malaysia [2012] “Malaysia – the Malaysian palm oil
sector
overview”,
http://www.ice.gov.it/paesi/asia/malaysia/upload/173/
Palm%20Oil_overview_2012.pdf. Accessed 15 October 2015.
Ghosh, P.K. [2015] “Daunting realities: territorial disputes and shipping
challenges to China’s maritime Silk Road”, Global Asia 10(3), Fall: 48-52.
International Monetary Fund (imf) [2015] People’s Republic of China 2015
Article IV consultation. imf Country Report No. 15/234. https://www.imf.
org/external/pubs/ft/scr/2015/cr15234.pdf. Accessed 15 October 2015.
Khoo, D. [2014] “Palm oil industry needs to invest more in downstream
activities”, http://www.thestar.com.my/Business/Business-News/2014/06/21/
Call-to-venture-downstream-Palm-oil-industry-needs-to-invest-more-indownstream-activities-to-reduc/?style=biz. Accessed 15 October 2015.
Khor, Y.L. [2013] “The significance of China-Malaysia industrial parks”, Institute
of South East Asian Studies Perspective No. 317. Singapore: iseas.
Kong, X.X., M. Zhang, and S.C. Ramu [2015] “China’s semiconductor industry
in global value chains”, Economic Research Institute for asean and East Asia
(ERIA) Discussion Paper No. 2015-15. Jakarta.
Lee, J. [2014] “China’s economic engagement with Southeast Asia: Malaysia”,
Trends in Southeast Asia, No.1. iseas: Singapore.
Hirschman, C. [1980] “Demographic trends in peninsular Malaysia: 1947-1975”,
Population and Development Review 6(1): 103-125.
Malaysian Palm Oil Board (mpob) [2013] Revision of customs tariff codes for oil
palm products & schedule of tariff rates. Kuala Lumpur.
Mendoza, R., K. Chua, and M. Melchor [2015] “An analysis of the evolving
asean-China trade linkages”, Rizalino S. Navarro Policy Center for
Competitiveness Working Paper 15-002. Manila: Asian Institute of
Management Policy Center.
Organisation for Economic Co-operation and Development (oecd) [2011] oecd
guide to measuring the information society. Paris.
Pitlo III, L.B. [2015] “asean’s connectivity and China’s ‘One Belt, One Road’”,
http://thediplomat.com/2015/03/asean-connectivity-and-chinas-one-belt-oneroad/. Accessed 15 October 2015.
Rizwan, M.H.R., K.T.F. Wong, and R.S.L. Cho [2014] “Agglomeration in
practice: the Malaysian experience in diversifying manufacturing”, http://
www.bnm.gov.my/documents/conference_vol/2014_Econs_Research/
Paper_6_The_Diversification_of_Manufacturing_in_Malaysia.pdf. Accessed
15 October 2015.
116
Shan, X.S. and R. Rasiah [2015] “Sticky spots on slippery slopes: the development
of the integrated circuits industry in emerging East Asia”, http://www.globelics.
org/publication/sticky-spots-on-slippery-slopes-the-development-of-theintegrated-circuits-industry-in-emerging-east-asia/. Accessed 15 October 2015.
Sta Maria, R. [2014] “Looking into the future of China-Malaysia relationship”,
http://my.china-embassy.org/eng/sbgx/t1158491.htm. Accessed 15 October 2015.
Tham, S.Y. and A.J.Y. Kam [2015] “Trade in value-added: case of Malaysia”,
www.eria.org/eria-dp-2015-58.pdf. Accessed 15 October 2015.
Tham, S.Y. [2015] “Diversification and industrial policies in Malaysia”,
Development and modern industrial policy in practice: issues and country
experiences. Cheltenham: Asian Development Bank and Edward Elgar.
Chapter 12 in J. Felipe ed..
The Star [2015] “Malaysia-China in sister port pact”, http://www.thestar.com.my/
News/Nation/2015/08/01/Msia-and-China-in-sister-port-pact-pka-chief-Tieup-will-boost-trade-and-tourism/. Accessed 15 October 2015.
Xing, Y. [2013] “Rethinking the success of China’s high-tech exports”, in
W. Gungwu and Z. Yongnian, eds., China: development and governance.
Singapore: World Scientific. Chapter 23.
117
APPENDIX 1. Information communications technology codes (HS 1996)
Computers
and peripheral
equipment
Communication
equipment
Miscellaneous
Consumer electronic
equipment
Electronic
components
844351
851711
852390
851810
852190
852330
847050
851719
852410
851821
852210
852460
847110
851730
852491
851822
852290
853400
847130
851750
852499
851829
852530
854011
847141
851780
852910
851830
852540
854012
847149
851790
852990
851840
852712
854020
847150
852510
854381
851850
852713
854040
847160
852520
901320
851890
852719
854050
847170
852790
851910
852721
854060
847180
853110
851921
852729
854071
847190
851929
852731
854072
847220
851931
852732
854079
847290
851939
852739
854081
847330
851940
852812
854089
847350
851992
852813
854091
851721
851993
852821
854099
851722
851999
852822
854110
900911
852010
852830
854121
900912
852020
950410
854129
852032
854130
852033
854140
852039
854150
852090
854160
852110
854190
852190
854212
852210
854213
854214
854219
854230
854240
854290
854890
Source: Organisation for Economic Co-operation and Development, 2011
PRE
Class participation in politics in Southeast Asia
Joseph J. Capuno
Using the dataset from the Asian Barometer Survey conducted in
2010-2012, we determine the variations in political participation
across socioeconomic classes in five Southeast Asian countries
and apply Oaxaca decomposition method to explain the
variations. In general, we find high rates of voting participation
across classes in Indonesia, Malaysia, the Philippines, and
Thailand. In Singapore, barely half of the sample voted in
previous elections. In Indonesia, Thailand, and Malaysia,
the middle classes distinguish themselves from the rest by
participating in other political activities, including contacting
officials or the news media, joining others to voice out or
directly address their common concerns, and attending rallies or
demonstrations. Furthermore, interclass differences in political
participation is due more to divergence in mean characteristics in
Thailand, but they are due more to the heterogeneous effects of
these characteristics in Indonesia and Malaysia. Relative to the
middle classes in these three countries, those in the Philippines
and Singapore each appears politically disengaged.
JEL classification: D72, Z10, 057
Keywords: political participation, middle class, Oaxaca decomposition,
Southeast Asia
1. Introduction
Are the middle classes more politically active or engaged than other
socioeconomic classes in democratic Southeast Asian countries? There is
theoretical basis to expect them to be so, which some casual observations seem to
confirm. Aristotle first advanced the idea that the middle class is crucially linked
with democracy, i.e., a large prosperous middle class mediate between the rich and
the poor, and thereby creates the structural foundation for democratic processes
[Glassman 1995]. Building on this idea, Lipset [1959] hypothesized a country’s
119
democratization—more evidently, perhaps, among the middle classes—as it
develops. Also, Moore [1967] famously claimed that there can be no democracy
without the middle class. Enunciating the casual link between middle class and
democratization, Acemoglu and Robinson [2006] posit that “… a strong and large
middle class may aid democratization because it is less in favor of radical policies
than the poor. Hence, if the rich are convinced that democracy is controlled by
the interest of the middle class agents, they have less to fear from democracy
and are less inclined to use repression to avoid it.” This idea finds support in a
study where democratic institutional reforms are found likelier to be in countries
with sizable middle classes [Loayza, Rigolini, and Llorente 2012].1 Arguably, the
middle classes help bring about such reforms by espousing democratic views and
values and then acting on them when needed.
The last two decades or so in Southeast Asia witnessed a range of participation
of the middle classes in major events. In Thailand, the Bangkok middle classes
have participated in political rallies that led to the ouster of the popularly elected
Thaksin Shinawatra as prime minister in 2006. Thaksin’s sister, Yingluck
Shinawatra, while also popularly elected as prime minister in 2011, nonetheless
also faced massive anti-government protests in 2013 that culminated in her
removal from office in 2014 by the Constitutional Court on corruption charges.
Even in post-Soeharto Indonesia, corruption in the highest level of government
continues to agitate various groups, including the middle classes that, together
with the poor population, elected Joko Widodo for president in 2014. In the
Philippines, the urban middle classes organized the so-called Million People
March in August 2013 that brought in more than 100,000 Filipinos to Luneta
Park to rally against corruption and pork barrel funds. Before that, they joined in
the street rallies in edsa that culminated in the removal from office of President
Joseph Estrada in early 2001. In Singapore and Malaysia, the same political
parties continue to remain in power, although Malaysia’s United Malays National
Organization is facing a growing opposition led by its former member Anwar
Ibrahim, whose supporters include the youth, the ethnic Chinese, and the urban
residents.
Notwithstanding the importance of these political activities, do they really
distinguish the middle classes from the rest? Several previous studies on Southeast
Asia (e.g., Robison, David, and Goodman [1996]; Embong [2001a]; Kimura
[2003]) noted that even in earlier periods, the middle classes—loosely referring
to those living in major urban areas, are professionals, are white-collar workers2,
or are college educated—engaged in political rallies, protest demonstrations, and
For recent studies linking the middle classes, economic development and democracy, see, for example,
Gerring et al. [2005], Banerjee and Duflo [2008], Chun, Hasan and Ulubasoglu [2011], and Birdsall [2015].
2
The studies in Embong [2001a] use the Erikson-Goldthorpe scheme in classifying people into social
classes on employment status and occupational classifications.
1
120
Capuno: Class participation in politics in Southeast Asia
election debates and campaigns to demand reforms or change in government.
Rather than leading in these political activities, the middle classes, however, were
merely following, and they were more sporadic than regular in their engagements.
Their participation was prodded more by a general or popular sentiment (say
against a repressive regime) and less by pure class consciousness or interest.
Two reasons are suggested for their apparently restrained role. Firstly, they
also benefitted from the economic policies promulgated by the very regimes that
they wanted to reform or change. Secondly, they continue to espouse traditional
values that emphasize respect for authority and primacy of group welfare, just
like the other Asian societies. Indeed, it was Singapore’s Lee Kwan Yew who
famously argued that Asians can achieve economic growth without giving up
traditional values, unlike Western societies (Emerson [1995]; Fukuyama [1995]).
Yet, it is hard to determine from the aforementioned studies whether the middle
classes as a group are not unlike the rest. One the one hand, the socioeconomic
classes are not well identified; much of the findings is based on the observed
actions of known members of the middle class who may be extremist themselves.
Second, there are other forms of political activities—like voting in elections or
contacting officials that are less public or dramatic than others—wherein the
socioeconomic classes may differentiate themselves. We thus re-investigate the
issue of class participation in politics in five Southeast Asian countries.
Indonesia, Malaysia, the Philippines, Singapore, and Thailand are selected for
two reasons. Firstly, all five have the survey data required for the method adopted
here. Of the countries in Southeast Asia, only these five, Vietnam, and Cambodia
were included in the third wave of the Asian Barometer Survey conducted in
2010-2012. Being in the same region, they all share a common geography, sociocultural heritage, and history, which facilitate comparison. Secondly, they vary
enough in terms of economic development and politics that can be explored to
answer our main research question. The World Bank classifies Indonesia and the
Philippines as lower-middle countries and Thailand and Malaysia as upper-middle
income countries. Like most Asian developing countries, the four developing
Southeast Asian countries also experienced in recent years reductions in poverty
rates and expansion of the middle-income sub-population [Asian Development
Bank 2010]. All four have positive polity iv scores in 2010, indicating they
are more democratic than autocratic.3 Of the five, only Singapore is considered
a high-income country, but it also has a negative polity iv score and, perhaps,
a more entrenched Confucian culture. Arguably, Singapore may provide the
benchmark for the group.
Following the lead of previous studies on the Asian values thesis that used
survey data (e.g., Chang, Zhu, and Pak [2007]; Chang, Chu, and Diamond
polity iv is developed by the polity iv Project. http://www.systemspeace.org/polity/polity4htm.
Accessed 30 July 2015.
3
121
[2012]), we use for this paper the dataset from the third round of the Asian
Barometer Survey. Instead of using the self-reported income status or occupation,
we construct wealth/asset indices and then delineate the nationally representative
sample into three socioeconomic classes. We then compare the three classes in
terms of proportions that participate in certain political activities. The differences
in proportions are further decomposed using the Oaxaca method into two
parts: one part that is due to differences in the observed characteristics (like
age, education, urbanity, occupation) of the classes; and the other part that is
due to heterogeneous effects of the same characteristics across classes.4 Unlike
in previous studies, our approach here allows us to measure and decompose
the class differences in their political participation and to compare the results
across countries.
In general, we find high rates of voting participation across classes in
Indonesia, Malaysia, the Philippines, and Thailand. In Singapore, barely half of
the sample voted in previous elections. In Indonesia, Thailand, and Malaysia,
the middle classes distinguish themselves from the rest by participating in other
political activities, including contacting officials or the news media, joining
others to voice out or directly address their common concerns, and attending
rallies or demonstrations. Interclass differences in political participation is due
more to divergence in mean characteristics in Thailand, but they are due more
to the heterogeneous effects of these characteristics in Indonesia and Malaysia.
Relative to the middle classes in these three countries, those in the Philippines
and Singapore each appears politically disengaged.
The rest of the paper is organized as follows. Section 2 discusses the data used
in the analysis.5 Section 3 details the methods used in the analysis. The results are
then presented in Section 4. The concluding section discusses the implications of
the findings.
2. Data
2.1. Survey data
The data used here is obtained from the third wave of the Asian Barometer
Survey6 jointly conducted by Taiwan-based Academia Sinica and the National
Taiwan University. The third-wave surveys were conducted in 2010-2012 in nine
East Asian and Southeast Asian countries, including Indonesia, Malaysia, the
Philippines, Singapore, and Thailand. Local institutions administered the countrylevel surveys and used the same survey instrument and sampling design to collect
comparable data. In each country, a nationally representative, random sample of
Dow [2009] used the same method to account for the differences in political views between genders.
Sections 2 and 3 draw heavily from Capuno [2015].
6
The website is www.asianbarometer.org.
4
5
122
voting-age population served as interview respondents. The sample consisted
of 1,550 for Indonesia, 1,214 for Malaysia, 1,200 for the Philippines, 1,000 for
Singapore, and 1,512 for Thailand.7
The survey questionnaire used has several modules designed to capture,
among others, the respondent’s participation in elections and other political
activities and his or her demographic and socioeconomic characteristics (e.g.,
family composition, self-reported household income level, and social status).
Additionally, the enumerator also records his or her own observations about the
respondent and the community characteristics in the interview sites.
2.2. Indicators of political participation
We constructed ten binary indicators of political participation. The first
4 indicators pertain to electoral participation. These indicators are voted_
last election (= 1 if the respondent voted in the most recent national election,
parliamentary, or presidential; 0 otherwise), voted_most elections (= 1 if voted
in every or most past elections, 0 otherwise), member_political organization
(= 1 if member of a political party or of an organization that supports candidates,
0 otherwise), and support_candidate (= 1 if tried to persuade others to vote for a
certain candidate or party or do anything else to help out or work for a party or
candidate running in the election, 0 otherwise). Note that the valid samples for the
above indicators exclude those who were not eligible to vote or to participate in
the past elections, possibly because they were too young then.
The next three indicators pertain to personal initiatives to draw the attention
of authorities presumably to their concerns. The indicators of personal
engagement are contacted_official (= 1 if contacted elected officials or legislative
representatives or officials at higher level at least once in the past three years,
0 otherwise), contacted_other leaders (= 1 if contacted traditional leaders/
community leaders or other influential people outside the government at least
once in the past three years, 0 otherwise), and contacted_media (= 1 if contacted
news media more than once in the past three years, 0 otherwise). Note that it
is possible that those who responded did the implied political activities together
with others, but that is not asked in the survey.
The last three indicators pertain to collective actions other than electoral
participations to directly address a common concern or to call public attention.
The three indicators are participated_with others (= 1 if got together with others
to try to resolve local problems or to raise an issue or sign a petition at least once
in the past three years, 0 otherwise), attended_rally (= 1 if attended a campaign
The sample sizes were set to allow for a minimum confidence interval of ±3 percent at the 95 percent
probability. The sample sizes are standard in similar surveys (e.g., World Values Survey, EuroBarometer
Survey).
7
123
meeting or rally in the last national election, 0 otherwise), and attended_
demonstration (= 1 if attended a demonstration or protest march more than once
in the past three years, 0 otherwise). Note that the samples for attended_rally
exclude those who responded “not applicable” to the relevant survey question,
possibly again because they were too young.
No single indicator is a definitive measure of political engagement, but
altogether they will indicate the extent. Our participation indicators are similar to
those in Wu and Lee [2012] and Verba, Nie, and Kim [1978]. Table 1 presents the
variables used in this paper and their definitions.
TABLE 1. Variable definitions
Variable name
Political participation
Voted_last election
Voted_most elections
Member_political
organization
Supported_candidate
Contacted_officials
Contacted_other leaders
Contacted_media
Participated_with others
Attended_rally
Attended_demonstration
Economic class
Lower class
Middle class
Upper class
Other characteristics
Male
Age
Age2
Some college
Finished college
Employed
Catholic
Definition
= 1 if voted in the last election, 0 otherwise
= 1 if voted in all or most past elections, 0 otherwise
= 1 if member of a political party or an organization that supports
a candidate, 0 otherwise
= 1 if tried to persuade others to vote for a certain candidate
or party or do anything else to help out or work for a party or
candidate running in the election, 0 otherwise
= 1 if contacted elected officials or legislative representatives
or officials at higher level at least once in the past three years, 0
otherwise
= 1 if contacted traditional leaders/community leaders or other
influential people outside the government at least once in the
past three years, 0 otherwise
= 1 if contacted news media more than once in the past three
years, 0 otherwise
= 1 if got together with others to try to resolve local problems or
to raise an issue or sign a petition at least once in the past three
years, 0 otherwise
= 1 if attended a campaign meeting or rally in the last national
election, 0 otherwise
= 1 if attended a demonstration or protest march more than once
in the past three years, 0 otherwise
= 1 if belongs to wealth quintile 1 or 2, 0 otherwise
= 1 if belongs to wealth quintile 3 or 4, 0 otherwise
= 1 if belongs to wealth quintile 5, 0 otherwise
= 1 if male, 0 otherwise
Age in years
Square of age
= 1 if has at least some college education, 0 otherwise
= 1 if finished college, 0 otherwise
= 1 if employed, 0 otherwise
= 1 if religion is Roman Catholic, 0 otherwise
124
Variable name
Buddhist
Islam
In union
Household size
Urban
Mega_majorcity
Access index
Definition
= 1 if religion is Buddhism, 0 otherwise
= 1 if religion is Islam, 0 otherwise
= 1 if married or living in with a partner, 0 otherwise
Number of household members
= 1 if residing in urban area, 0 otherwise
= 1 if living in capital city, megacity, regional center, or major city,
0 otherwise
= 100 x {[paved road + public transport + post office + school +
police + sewerage + clinic + cellular signal + recreation facility +
church + town hall + market stall]/12}
2.3. Socioeconomic classes
Following the method in Filmer and Pritchett [2001], we classified into
economic classes the samples in each country based on wealth indices
constructed, which are based on the reported household amenities and assets. The
lists of amenities and assets vary slightly across countries. First, we developed
binary indicators of the household’s connection to electricity and piped water and
ownership of motor vehicle (car or jeep), tractor, television set, cable television,
motorcycle, telephone, mobile phone, bicycle, radio, pumping set, refrigerator,
camera, and livestock (goat and cow). Next, we took the mean and standard
deviation of each indicator, and then we derived the first principal component of
the vector containing these indicators. These statistics are then used to generate
a score for each household asset or amenity. The asset scores are then added up
to derive the household’s overall wealth index. Ranked from lowest to highest
wealth index, the households are then grouped into quintile.
The first two quintiles together are classified as the lower class, the third
and fourth quintiles together as the middle class, and the fifth quintile as the
upper class. By construction, the lower class may include households with
incomes above the poverty threshold but who may become poor due to sudden
unemployment, illness, calamity, or other adverse shocks. As well, the middle
class possibly includes some wealthy households whose fortunes are still modest
to be considered part of the economic elite. By construction, the last group is
the wealthiest and the smallest in size (only 10 percent of the sample). This is
consistent with the observation that the rich are few and rarely participate in such
surveys.
For the Philippines, Thailand, or Indonesia, the first principal component
accounts for 21 percent of the within-country variations in the household
amenities and assets (and livestock). The corresponding proportions are 18
percent in Singapore and 15 percent in Malaysia. To further validate the wealth
quintiles, we correlated them with the self-declared household income levels. In
the surveys, the respondents were asked to identify where they are at in a fiverung income ladder, where the first rung is the “lowest level” and the fifth and last
125
rung is the “highest level”. The resulting correlation coefficients are 63 percent
for Thailand, 55 percent for Indonesia, 47 percent for the Philippines, 46 percent
for Malaysia, and 25 percent for Singapore.8
Finally, we also compared the proportions of white-collar workers in the lower
class, middle class, and upper class in each country. The expectation is that most,
if not all, white-collar workers would belong in the middle and upper classes.9
The proportions of white-collar workers among the employed are as follows:
in the Philippines, 34 percent in the lower class, 57 percent in the middle class,
and 66 percent in the upper class; in Thailand, 12 percent in the lower class, 29
percent in the middle class, and 72 percent in the upper class; in Indonesia, 17
percent in the lower class, 41 percent in the middle class, and 66 percent in the
upper class; in Singapore: 76 percent in the lower class, 82 percent in the middle
class, and 90 percent in the upper class; and in Malaysia, 40 percent in the lower
class, 63 percent in the middle class, and 70 percent in the upper class.
Overall, our wealth quintiles and the economic classes derived from them
correlate fairly well with other indicators of social classes. The correlation is
relatively lower in Singapore, arguably because of the limited list of amenities
and assets contained in the Asian Barometer Survey questionnaire. Furthermore,
the choices are too commonplace to sharply distinguish Singaporeans by wealth
status. Since most Singaporeans are well off, and all live in urban areas, they
are likely to own all the assets in the list, save perhaps the livestock and farm
equipment.
2.4. Covariates
Table 1 shows the variables used to further characterize the sample respondents.
The individual-level characteristics are gender (male = 1, 0 otherwise), age (in
years), attainment of at least some tertiary-level of formal schooling (at least
some college), employment status (employed), married or living-in with partner
(in-union), and religion (Catholic, Islam, Buddhist). The household-level
characteristics include household size and residence in an urban area, capital
city, megacity, or regional center. Both age and its squared term (age2) and
household size are continuous variables; the rest of the aforementioned variables
are dichotomous variables.
The correlation estimates are for samples that reported their household income level and excludes those
with missing responses or replied not to know their income levels.
9
The classification of the respondents into “white-collar” worker or otherwise is based on the scheme
proposed by Erikson and Goldthorpe [1992]. In this paper, “white-collar” workers include “professionals,
large enterprise employers, managers (with 1-10 or more than 10 subordinates), and routine clerical and
sales workers.” However this classification can be applied only to respondents who, at the time of the survey,
were currently employed and declared their main occupation.
8
126
We also constructed an access index based on the reported presence of certain
public facilities or amenities within the respondent’s community or immediate
environment. In developing the index, first we developed binary indicators of
the presence of paved roads leading to the respondent’s abode, and of access
to a public transportation, post office, school, police station, public sewerage
facility, clinic, cellular phone signal, recreation facility, church, town hall, and
marketplace. Each indicator has a value of 1 if the reference public amenity or
facility is present and 0 if otherwise. Then we summed up the values of these
indicators, divided the total by 12, and finally multiplied the resulting quotient by
100 to derive the household’s access index score. The access index score ranges
from 0 (lowest) to 100 (highest). A high score implies that the household has
fairly easy access to sources of information or facilities for mobility or social
interactions.
Annex 1 shows the mean and standard deviations of the covariates for each
country. In each country, about half of the sample is male, and the mean age
is between 41 to 47 years. Between 9 percent in Indonesia and 48 percent in
Singapore attained some college education. In the Philippines, about 84 percent are
Catholic. In Indonesia, 87 percent practice Islam, while the figure is 57 percent in
Malaysia. The proportion of Buddhists is 94 percent in Thailand and 40 percent in
Singapore. In each country, at least 48 percent claims to be employed and at least
68 percent are married or living-in with a partner (in union). The mean household
size ranges from around 4 (Singapore) to nearly 5.7 (Malaysia). Less than half of
the respondents in Indonesia, Malaysia, and Thailand live in urban areas. A huge
majority of the sample in the Philippines resides in megacities or major cities,
while less than 25 percent of the samples in Indonesia or Malaysia do so. (Note
that this indicator cannot be constructed with the survey data for Thailand and
Singapore.) Finally, the mean access index is lowest in the Philippines at 65.10
percent and highest in Singapore at 98.08 percent. As expected, around forty
percent of the sample in each country belongs to the lower class and another forty
percent to the middle class; the rest belongs to the upper class.
3. Methods
3.1. Test of proportions
For each country sample, we apply tests of proportions to determine if the
middle class differs from other economic classes in terms of political participation.
Basically, the test ascertains if the middle class and another class have statistically
equal proportions of members who, say, voted in the last election. If the
proportions are not equal, we then proceed to decompose the difference using the
Oaxaca method.
127
3.2. Oaxaca decomposition method
First introduced in Blinder [1973] and Oaxaca [1973], the original Oaxaca
decomposition method (also called the Blinder-Oaxaca decomposition method)
was used to decompose the sources of wage differentials between sexes, where
wages are a linear function of gender and other control variables. In this paper,
we follow the extension of the original method to non-linear functions made by
Sinning, Hanh, and Bauer [2008].
To fix ideas, consider an outcome Y being a nonlinear function of the
characteristics of members of group g. Y could be a dichotomous indicator of
political participation (e.g., 1 = Yes, 0 otherwise) and is related by a logistic
function, say, to a vector of observable characteristics X (that contains, for
example, gender, employment, and educational attainment). Let the average
outcome for group M be denoted by YM. Let the average outcome for group L
be denoted by YL. The difference in the average outcomes can be decomposed as
follows:
YM - YL = {Eβ (YiM| XiM) - Eβ (YiL| XiL) } + {Eβ (YiL| XiL) - Eβ (YiL| XiL)}
L
L
M
L
+ {[Eβ (YiM| XiM) - Eβ (YiM| XiM)] - [Eβ (YiL| XiL) - Eβ (YiL| XiL)]},
M
L
M
L
where Eβg (Yig|Xig) refers to the conditional expectation of Yig, and Eβg (Yig|Xih) refers
to the conditional expectation of Yih evaluated at the parameter vector βg, with
g,h= (M,L) and g ≠ h.
To elaborate, the difference in the mean outcomes of groups M and L can be
decomposed into three parts, as represented by the three terms in parentheses in
the right-hand side of the equation above.
The first part, {Eβ (YiM| XiM) - Eβ (YiL| XiL) }, is due to the differences in the
L
L
average characteristics of the two groups. Specifically, it denotes the change in
the mean outcome of group L if its members attain the same characteristics as
those in group M, but multiplied by the coefficients of group L.
The second part, {Eβ (YiL| XiL) - Eβ (YiL| XiL)}, is due to differences in the
M
L
coefficients of the two groups. It denotes the change in the mean outcome of
group L if its coefficients are set equal to that of group M, but evaluated at the
same average characteristics of group L.
The last term, {[Eβ (YiM| XiM) - Eβ (YiM| XiM)] - [Eβ (YiL| XiL) - Eβ (YiL| XiL)]},
M
L
M
L
derives from the interaction of the differences in the mean characteristics and
in the coefficients of the two groups. In our empirical implementation of this
decomposition, the reference group is the middle class (M) and the comparison
group (L) could be other households not in the middle class (“other class” means
low class and upper class), those that are in the low class alone, or those that are
in the upper class alone.
128
To illustrate the insight that can be gained with this method, consider one
characteristic, say, college education, and one outcome, say, voted in the last
election. Suppose the middle class has a higher proportion of members than the
lower class who voted. The decomposition allows us to determine how much of
the difference in proportions would be reduced if the lower class achieved the
same average level of education as the middle class. This is important since if by
doing so the differences are completely or substantially eliminated then the case
for subsidized college education of the underprivileged is strengthened, in this
case the higher education will inculcate in them the virtue and value of exercising
their rights and duties to vote.
However, if the coefficients are significantly different, the decomposition
method will also indicate that the same college education provided to the poor
will not make them exercise their voting rights as much as the middle class, which
renders such an education policy inadequate for the purpose. The differences in
coefficients reflect the heterogeneous effects of the same intervention targeted
to the two groups. These may arise out of idiosyncrasies in personal (or class)
experiences or expectation, or these may be conditioned by the social context
or culture to which members of the same class belong. The decomposition
method used here provides an estimate and the statistical significance of the
heterogeneous effects, but it does not identify the reasons for the differential
impacts. Notwithstanding this limitation, distinguishing the differences in mean
characteristics from the differences in their effects to explain the variations in
political participation across classes is still relevant to policy. Since at least some
observable characteristics (like education and employment) are directly influenced
by government intervention, the decomposition method can shed light on the
effectiveness in terms of shaping a national political consensus of government
policies and programs, on the one hand, and of possible (yet still unidentified)
psychosocial and cultural factors, on the other.
Since the dependent variables are all dichotomous variables, we use
probit regression models to estimate the differences in the probability of an
outcome (measuring political participation) conditional on a list of covariates
(characteristics). To carry out non-linear Oaxaca decomposition analysis, we use
the Stata module nldecompose [Sinning, Hahn, and Bauer 2008]. Note that in
both the tests of proportions and Oaxaca decompositions, the reference group is
the middle class. Thus, a negative (positive) difference in proportions indicates a
lower (higher) proportion of the middle class than either the lower class or upper.
129
4. Results
4.1. Test of proportions
Table 2 shows the results of the tests of proportions for Indonesia, the
Philippines, and Thailand. In the left panel, we find that, overall, a high proportion
(90 percent) of Indonesians exercised their right to vote, but a significantly lower
percentage had other forms of political engagements. Only a fourth of them
reported to have joined a rally or a campaign meeting, and about one in ten joined
forces with others to address a collective issue.
Nonetheless, there are indications that the Indonesian middle class is different
from the other classes. When compared to the lower class, they appear to be
more active in being members of political organizations, in supporting political
candidates or parties, in contacting local officials or the news media, or in joining
others to address a common concern. When compared to the rich, however, they
appear less inclined to contact officials or the news media or to team up with
others to resolve a common concern.
In the middle panel, we find that nearly 80 percent of the Filipinos voted in
past elections, and about 23 percent of them reported to have joined political
rallies. Less than five percent reported to have participated in other types of
political activities. Generally, members of the Filipino middle class do not appear
to be more or less politically active or engaged than the lower class or upper class.
They appear to be distinct from the lower class only in that a lower percentage of
them have joined forces with others to address a common problem.
Like Indonesians, most Thais exercised their right to vote (right panel). Unlike
in Indonesia and the Philippines, nearly half of the Thais claimed to have joined
political campaigns or rallies during the last national election. Less than 10
percent of Thais reported to have participated in other forms of political activities.
When compared to the lower class, members of the middle class are less likely
to vote or to attend a rally, but they are more likely to contact the news media or
participate with others. When compared to the rich, the middle class is likelier to
vote, to support a political party or a candidate, or to attend a rally.
Contacted_other
leaders
Contacted_
officials
Supported_
candidate
Member_political
organization
Voted_most
elections
Voted_last
election
Indicator
Middle
Overall
Middle
class
proportion class
vs.
vs.
lower
others
class
0.903 -0.005 -0.020
[0.0296] (0.015) (0.016)
N=1550
0.923 -0.008 -0.015
[0.266] (0.014) (0.015)
N=1528
0.070 0.030** 0.035**
[0.256] (0.014) (0.015)
N=1550
0.084
0.016 0.026*
[0.277] (0.015) (0.015)
N=1550
0.015
0.001 0.013**
[0.124] (0.006) (0.006)
N=1550
0.046
0.006
0.018
[0.209] (0.011) (0.011)
N=1550
Middle
Overall
class
proportion
vs.
upper
class
0.026
0.798
(0.023)
[0.401]
N=1032
0.006
0.786
(0.020)
[0.410]
N=1127
0.019
0.022
(0.019)
[0.146]
N=1200
-0.006
0.119
(0.021)
[0.324]
N=1032
-0.023**
0.026
(0.012)
[0.159]
N=1200
-0.020
0.028
(0.017)
[0.166]
N=1200
Difference in proportions
Indonesia
Philippines
Thailand
Difference in
proportions
Middle Middle
Middle
Middle
Overall
Middle
class
class
class
class
proportion Middle
class
vs.
vs.
class vs. vs.
vs.
vs.
lower upper
others
lower
upper
others
class
class
class
class
0.024
0.027
0.018
0.913
0.003 -0.033** 0.072***
(0.025) (0.028) (0.035)
[0.283]
(0.015)
(0.015)
(0.024)
N=1509
0.014
0.025 -0.011
0.921
0.003 -0.033** 0.076***
(0.025) (0.027) (0.033)
[0.268]
(0.014)
(0.014)
(0.023)
N=1498
0.003
0.008 -0.008
0.006
-0.002
-0.001
-0.008
(0.009) (0.009) (0.014)
[0.077]
(0.004)
(0.004)
(0.007)
N=1512
0.013
0.023 -0.009
0.081
0.014
-0.001 0.037**
(0.021) (0.022) (0.030)
[0.273]
(0.015)
(0.016)
(0.017)
N=1503
-0.011 -0.008 -0.017
0.034
0.014
0.016
0.007
(0.009) (0.009) (0.014)
[0.181]
(0.010)
(0.010)
(0.013)
N=1509
0.006
0.005
0.010
0.040
-0.002
0.003
-0.016
(0.010) (0.011) (0.013)
[0.200]
(0.010)
(0.011)
(0.015)
N=1509
TABLE 2. Test of proportions: Indonesia, the Philippines, and Thailand
130
Middle
Overall
Middle
class
proportion class
vs.
vs.
lower
others
class
0.019 -0.002 0.013**
[0.138] (0.007) (0.006)
N=1550
0.126
0.008 0.034*
[0.332] (0.017) (0.018)
N=1550
0.248 -0.004 -0.025
[0.432] (0.022) (0.025)
N=1550
0.021
0.004
0.007
[0.142] (0.008) (0.008)
N=1550
Middle
Overall
class
proportion
vs.
upper
class
-0.035***
0.030
(0.014)
[0.171]
N=1200
-0.048*
0.063
(0.026)
[0.244]
N=1200
0.041
0.226
(0.029)
[0.418]
N=1032
-0.004
0.029
(0.011)
[0.168]
N=1200
Philippines
Thailand
Difference in
proportions
Middle Middle
Middle
Middle
Overall
Middle
class
class
class
class
proportion Middle
class
vs.
vs.
class vs. vs.
vs.
vs.
lower upper
others
lower
upper
others
class
class
class
class
0.010
0.009
0.014
0.043
0.013
0.020*
-0.004
(0.010) (0.011) (0.013)
[0.202]
(0.011)
(0.011)
(0.016)
N=1503
-0.023 -0.029* -0.009
0.082
0.014
0.034**
-0.022
(0.014) (0.016) (0.019)
[0.275]
(0.015)
(0.015)
(0.021)
N=1510
0.010
0.023 -0.019
0.499 -0.078*** -0.175*** 0.112***
(0.027) (0.029) (0.038)
[0.500]
(0.026)
(0.028)
(0.034)
N=1507
-0.009 -0.008 -0.013
0.051
0.002
-0.002
0.004
(0.010) (0.010) (0.014)
[0.220]
(0.013)
(0.014)
(0.017)
N=1509
Notes: Figures in brackets are standard deviations. Figures in parentheses are standard errors. N means number of observations.
*p<0.10
**p<0.05
***p<0.01
Attended_
demonstration
Attended_rally
Participated_with
others
Contacted_
media
Indicator
Indonesia
TABLE 2. Test of proportions: Indonesia, the Philippines, and Thailand
131
132
Table 3 shows the corresponding results for Singapore and Malaysia. Relative
to Indonesians, Filipinos and Thais, Singaporeans in general appear to be less
politically engaged. Less than half of them voted in most of the past elections, and
only 56 percent of them voted in the last election. Less than eight percent of them
joined in rallies, and less than four percent joined any other political activity.
Moreover, a lower proportion of the middle class than the lower class voted in
most elections or contacted any government official. Also, a lower percentage of
them than the upper class supported a political party or candidate.
The overall pattern of political participation of Malaysians is closer to that of
the Indonesians, Filipinos, or Thais than to Singaporeans. About three in every
four Malaysians exercised their right to vote. About one in ten was a member of
political parties or organizations or supported political candidates. Nearly two in
ten contacted government officials or joined others to address a common concern.
About three in ten attended political rallies. When compared to the lower class,
the middle class has a greater percentage of members who belong to political
organizations or contacted officials, but less of them attended demonstrations.
When compared to the upper class, the middle class appears less likely to contact
traditional leaders or community leaders or to contact news media.
TABLE 3. Test of proportions: Singapore and Malaysia
Indicator
Voted_last election
Voted_most
elections
Member_political
organization
Supported_
candidate
Contacted_officials
Contacted_other
leaders
Contacted_media
Singapore
Middle Middle
Middle
Overall
class
class
class
proportion
vs.
vs.
vs.
lower
upper
others
class
class
0.554
-0.022 -0.017 -0.035
[0.497]
(0.034) (0.038) (0.047)
N=855
0.484
-0.059* -0.070* -0.034
[0.500]
(0.034) (0.037) (0.048)
N=855
0.012
0.0003 0.002
-0.004
[0.109]
(0.007) (0.007) (0.011)
N=1000
0.012
-0.013* -0.011 -0.016*
[0.110]
(0.006) (0.007) (0.011)
N=979
0.02
-0.013 -0.018* -0.004
[0.140]
(0.008) (0.01)
(0.011)
N=1000
0.008
-0.005 -0.008 -0.0004
[0.089]
(0.005) (0.007) (0.006)
N=1000
0.022
0
0.002
-0.004
[0.147]
(0.009) (0.010) (0.014)
N=1000
Malaysia
Middle
Middle
Overall
Middle
class
class
proportion class vs. vs.
vs.
others
lower
upper
class
class
0.770
0.008
0.014
0.004
[0.421]
(0.027)
(0.029)
(0.037)
N=1026
0.746
-0.013
-0.001
-0.033
[0.436]
(0.028)
(0.030)
(0.037)
N=1026
0.100
0.032*
0.042**
0.012
[0.300]
(0.019)
(0.019)
(0.025)
N=1214
0.105
-0.015
-0.012
-0.019
[0.306]
(0.018)
(0.019)
(0.024)
N=1214
0.288
0.023
0.062**
-0.054
[0.453]
(0.027)
(0.028)
(0.037)
N=1214
0.181
-0.024
0.002
-0.075**
[0.385]
(0.022)
(0.024)
(0.032)
N=1214
0.060
-0.013
-0.001
-0.038**
[0.238]
(0.014)
(0.014)
(0.021)
N=1214
Indicator
Participated_with
others
Attended_rally
Attended_
demonstration
Singapore
Middle Middle
Middle
Overall
class
class
class
proportion
vs.
vs.
vs.
lower
upper
others
class
class
0.038
-0.019 -0.016 -0.026
[0.191]
(0.012) (0.013) (0.018)
N=1000
0.076
0.014
0.021
-0.002
[0.265]
(0.017) (0.019) (0.025)
N=973
0.01
-0.004 -0.0002 -0.0008
[0.010]
(0.006) (0.007) (0.009)
N=1000
133
Malaysia
Middle
Middle
Overall
Middle
class
class
proportion class vs. vs.
vs.
others
lower
upper
class
class
0.167
-0.015
-0.014
-0.014
[0.373]
(0.022)
(0.024)
(0.029)
N=1214
0.316
-0.003
-0.029
0.052
[0.465]
(0.027)
(0.030)
(0.035)
N=1214
0.030
-0.030*** -0.036*** -0.016
[0.172]
(0.009)
(0.011)
(0.012)
N=1214
Notes: Figures in brackets are standard deviations. Figures in parentheses are standard errors. N means
number of observations.
*p<0.10
**p<0.05
***p<0.01
In summary, there is generally a high level of participation in elections among
Southeast Asian people. A small but significant minority in most countries attends
rallies, contacts officials or the news media, or joins others to voice out or directly
address their common concerns. In these political activities, the middle classes
distinguish themselves from the rest. The factors that make them distinct are
identified below.
4.2. Oaxaca decomposition10
Table 4 shows for Indonesia, the Philippines, and Thailand the results of
the Oaxaca decomposition of the sources of the class differences in political
participation. Note that the decomposition is done only for those participation
indicators where class differences are statistically significant. In the left panel,
we see that, in Indonesia, the differences between the middle class and the
lower class are not due to the dissimilarities in their average characteristics
but rather to the relatively bigger effects of the same characteristics on the
middle class. The results show that the differences in coefficients account for
at least three percentage points of the total gap in proportions of the middle
class and lower class that reported to be members of political organizations
or supported political candidates. Furthermore, the differences in coefficients
The results reported are based on a set of 32 unweighted regression runs, one pair for each indicator
of political participation where proportions of the middle classes and another class (lower classes, upper
classes) differ. Sampling weights are not used to make the results comparable with those obtained in the tests
of proportions, which do not allow weights. The detailed regression results are not reported here to save on
space, but they are available from the author upon request.
10
134
account for at least a percentage point of the total gap in proportions that
contacted officials or the news media. The decomposition reveals, however,
that the differences between the middle and upper classes are only apparent.
Neither the differences in average characteristics, their differential impact on
the likelihood of political participation, or the interaction of these two factors
appears to be statistically significant.11
TABLE 4. Oaxaca Decomposition of political participation:
Indonesia, the Philippines, and Thailand
Indicator
Voted_last
election
Characteristics
Indonesia
Middle Middle
Middle
class
class
class
vs.
vs.
vs.
lower
upper
others
class
class
Philippines
Middle
Middle Middle
class
class
class vs.
vs.
vs.
lower
upper
others class
class
Middle
class
vs.
other
Interaction
Coefficients
Interaction
-0.009
(0.008)
0.027*
(0.015)
0.012
(0.009)
-0.038***
(0.009)
0.042*
(0.025)
-0.036*
(0.020)
0.038*
(0.020)
-0.002
(0.020)
0.039
(0.026)
-0.039***
(0.009)
-0.005
(0.022)
0.007
(0.016)
0.068***
(0.023)
0.028
(0.026)
-0.014
(0.03)
Coefficients
Member_political
organization
Characteristics
Middle
class vs.
upper
class
Coefficients
Interaction
Voted_most
elections
Characteristics
Thailand
Middle
class
vs.
lower
class
-0.005
(0.012)
0.030*
(0.016)
0.010
(0.015)
Note that the decomposition method applied here works only when the two classes have the same set
of characteristics. When a characteristic is unique to or highly prevalent in a class, it is dropped by the
statistical method used here. This is the reason why a number of decomposition results in Tables 4 and 5 are
statistically insignificant. It is possible the omitted variable may be the one that drives the class differences.
11
Indicator
Supported_
candidate
Characteristics
Indonesia
Middle Middle
Middle
class
class
class
vs.
vs.
vs.
lower
upper
others
class
class
Interaction
-0.001
(0.004)
0.013**
(0.006)
0.001
(0.004)
Coefficients
Interaction
Participated_with others
Characteristics
Coefficients
Interaction
Attended_rally
Characteristics
Coefficients
Interaction
Middle
class
vs.
other
Thailand
Middle
class
vs.
lower
class
0.002
(0.002)
0.012***
(0.005)
-0.001
(0.003)
-0.010
(0.011)
0.032
(0.020)
0.013
(0.014)
Middle
class vs.
upper
class
0.005
(0.017)
0.005
(0.026)
0.021
(0.011)
Coefficients
Interaction
Contacted_
media
Characteristics
Middle
Middle Middle
class
class
class vs.
vs.
vs.
lower
upper
others class
class
0.008
(0.009)
0.036**
(0.017)
-0.017
Coefficients
Contacted_
officials
Characteristics
Philippines
135
-0.008
(0.010)
-0.014
(0.014)
-0.001
(0.011)
-0.015
(0.010)
-0.019
(0.012)
-0.001
(0.012)
0.015
(0.011)
-0.040
(0.030)
-0.022
(0.023)
(0.024)
0.016***
(0.005)
-0.006
(0.024)
0.008
(0.019)
-0.001
(0.007)
-0.043***
(0.016)
0.015
(0.012)
-0.002
(0.009)
0.030*
(0.017)
0
(0.013)
0.004
(0.01)
-0.100***
(0.026)
0.009
(0.01)
Figures in parentheses are bootstrap standard errors estimated with 100 replications.
*p<0.10
**p<0.05
***p<0.001
-0.046*** 0.103***
(0.013)
(0.031)
-0.131*** 0.002
(0.027)
(0.042)
-0.011
0.004
(0.02)
(0.036)
136
In the case of the Philippines, we also find that it is the heterogeneous
effects rather than the dissimilarities in average characteristics that distinguish
the middle class from the lower class in terms of likelihood to team up with
others to address a common concern. Unlike in the Indonesia, however, the
effects are lower for the middle class than the lower class.
In the case of Thailand, it is the dissimilarities in mean characteristics
that make the middle class less likely than the lower class to vote in elections
and more likely than the upper class to vote or to attend rallies. We also find
that the interaction of the differences in the average characteristics and in
the coefficients explains why the middle class is less likely than the lower
class to vote in the previous election. The heterogeneous effects of average
characteristics also explain why the middle class appears more likely than
the lower class to team up with others to solve a common problem but less
likely than the lower class to attend rallies. Finally, the apparent discrepancies
in proportions of middle class and upper class that supported political
candidates are neither due to divergence in mean characteristics or in their
differential effects.
Table 5 shows the results of the decomposition analysis for Singapore
and Malaysia. In the case of Singapore, it is the differential impacts of the
characteristics that make the middle class less likely than the rest to support a
candidate or than the lower class to contact any government official. In addition,
we find that the interaction of the differences in average characteristics and the
differences in coefficients also explain why the middle class is likelier than the
lower class to contact government officials The decomposition results indicate
that it is not the differences in mean characteristics, the differences in the impacts
of the characteristics, or the interaction of these two differences that account for
why the middle class and the lower class are not equally likely to vote, or why the
middle class and the upper class are not equally likely to support a candidate.
TABLE 5. Oaxaca Decomposition of political participation:
Singapore and Malaysia
Indicator
Voted_most elections
Characteristics
Coefficients
Interaction
Singapore
Malaysia
Middle
Middle
Middle
Middle
Middle
class vs.
class vs.
class vs. class vs.
class vs.
lower class upper class others
lower class upper class
-0.012
-0.028
(0.017)
(0.020)
-0.052
-0.037
(0.032)
(0.034)
0.005
-0.005
(0.017)
(0.020)
Middle
class vs.
others
Indicator
Member_political
organization
Characteristics
Singapore
Middle
Middle
Middle
Middle
class vs. class vs.
class vs.
class vs.
others
lower class upper class others
Coefficients
Interaction
Supported_candidate
Characteristics
Coefficients
Interaction
Contacted_officials
Characteristics
-0.002
(0.005)
-0.012**
(0.005)
0.001
(0.005)
Coefficients
Interaction
Contacted_other leaders
Characteristics
-0.001
(0.003)
-0.032***
(0.012)
0.015**
(0.007)
-0.004
(0.009)
-0.011
(0.008)
-0.001
(0.009)
-0.002
(0.009)
0.035**
(0.018)
-0.001
(0.009)
Malaysia
Middle
Middle
class vs.
class vs.
lower class upper class
-0.010
(0.011)
0.054***
(0.017)
-0.003
(0.014)
-0.005
(0.011)
-0.065*
(0.035)
-0.006
(0.021)
0.006
(0.007)
-0.052**
(0.023)
0.008
(0.015)
Coefficients
Interaction
Attended_demonstration
Characteristics
Coefficients
Interaction
0.017
(0.020)
0.059*
(0.033)
-0.013
(0.026)
Coefficients
Interaction
Contacted_media
Characteristics
137
0.001
(0.004)
-0.029***
(0.008)
-0.001
(0.006)
0.001
(0.003)
-0.043***
(0.013)
0.006
(0.008)
Figures in parentheses are bootstrap standard errors estimated with 100 replications.
*p<0.10
**p<0.05
***p<0.001
138
Finally, in the case of Malaysia, we find that it is more the heterogeneous
effects of characteristics that distinguish the middle class from the other classes.
In particular, this divergence makes the middle class more likely than the lower
class to become members of political organizations or to contact officials.
However, the dissimilarities in coefficients make the middle class less probable
than the upper class to contact traditional or community leaders or less likely than
the lower class to attend demonstrations.
In sum, we find evidence that the differences in class participation in politics
are explained by differences in mean characteristics and in the effects of these
characteristics. However, it is not always the case that raising the average
characteristics of the lower class to match that of the middle class will make them
as politically engaged as the latter; in fact, there are instances when doing so
will make them less engaged. Part of the reason is that the same characteristics
may lead to lower or negative incremental effects on their probability of
political participation.
5. Discussion and conclusion
Here we find evidence in five Southeast Asian countries that, generally,
the middle classes are like other socioeconomic classes in terms of
voting participation, but they are unlike the rest in other forms of political
engagement. We also find that the middle classes are not always more
politically active than other classes. Moreover, if the characteristics that seem
to make them active were adopted by or transplanted in others, they may yield a
different, even negative, effect on others. We also find the middle-class activism
varies in both extent and form across countries.
In democracies, elections are an important institutional mechanism by which
the people hold their leaders accountable. As the pivotal group of voters, the
middle class is also expected to participate more in this activity than others.
Our results show that voting participation is equally high across socioeconomic
classes in Indonesia, the Philippines, Thailand, and Malaysia. In Thailand, the
voting participation rate of the middle class is slightly lower than that of the
lower class, but it is higher than that of the upper class. However, the middle
classes are distinct in their involvement in other election-related activities, such as
membership in political organizations in Indonesia and Malaysia or supporting a
political candidate or party in Indonesia and Thailand.
Democracy also thrives in other forms of political activism. In Indonesia,
Thailand, and Malaysia, the middle classes further distinguish themselves, in
terms of contacting government officials or news media, participating with others,
and attending rallies and demonstration. In Indonesia and Thailand, the differential
participation rates across classes is explained more by the heterogeneous effects
of average characteristics, whereas in Malaysia it is more due to the inequalities
139
in mean characteristics. These results suggest political activism may be more
directly influenced by policies in Indonesia and Thailand than in Malaysia.
Relative to the three countries, the Philippines and Singapore each appear
distinct. In contrast to the Indonesian, Thai, or Malaysian middle classes, the
Filipino middle class seem unremarkable. Its rates of participation in all electionrelated activities or in most of the other political activities are not different from
any other class. Where it diverges is only in terms of participating with others,
showing an even lower proportion than the lower class. Even its high voting rate
does not distinguish it from other classes, although the overall voting rates among
the poor and, perhaps, even of the middle class in the Philippines could be driven
by clientilist relations. That is, perhaps a large segment of the population votes for
politicians in exchange for money or personal favors.12
In Singapore, barely half of the sample reported to have voted in the past
elections. Moreover, the level of engagement in other political activities of the
middle class is generally lower than either the lower class or the upper class.
Their lower participation rates could imply that they are resigned to any attempt
to change or reform the government. Alternatively, it could mean that they are
“co-opted” through the government programs (like public housing, for example).
Investigating the issue further will have implications on the claim made that
Singapore follows a different Asian style of democracy.13
Our results suggest two directions for future research. One direction is to
apply the same method on the same or similar data but to explore other interclass
differences in their political or social orientations or activities. For example, it may
be investigated if the middle class eschews traditional values and adopts modern
values more than the lower class or upper class. The results of this investigation
will shed light on the role of the middle classes in the democratization of Asian
countries. The other direction is extending the decomposition method to identify
the critical characteristics that matter more and in developing methods to identify
the cognitive, social, or contextual factors that lead to heterogeneous effects.
University of the Philippines School of Economics
The author gratefully acknowledges the Freiburg Institute for Advanced Studies - Albert
Lüdwig University, which hosted the author and funded initial work on this research,
with support from the People Programme (Marie Curie Actions) of the European
Union’s Seventh Framework Programme (FP7/2007-2013) under REA grant agreement
no. [609305]. He also appreciates the assistance and the data from the Asian Barometer
Project Office. The views expressed in this paper are the author’s own.
Thanks to Prof. Corina Gochoco-Bautista for this interpretation of the results.
In another study [Capuno 2015], the Singaporean middle class, however, appears more pro-democracy in
their views and values than other socioeconomic classes.
12
13
140
References
Acemoglu, D. and J.A. Robinson [2006] Economic origins of dictatorships
and democracy. Cambridge, uk: Cambridge University Press.
Asian Development Bank (adb) [2010] “The rise of Asia’s middle class”,
in Key indicators for Asia and the Pacific 2010. Mandaluyong City,
Philippines: Asian Development Bank.
Banerjee, A.V. and E. Duflo [2008] “What is middle class about the middle
classes around the world?”, Journal of Economic Perspectives 22(2): 3-28.
Birdsall, N. [2015] “Does the rise of the middle class lock in good government
in the developing world?”, http://www.cgdev.org/publication/does-rise-middleclass-lock-good-governemnt-developing-world. Accessed 30 April 2015.
Blinder, A.S. [1973] “Wage discrimination: reduced form and structural
estimates”, Journal of Human Resources 8: 895-909.
Capuno, J.J. [2015] “An Oaxaca-decomposition analysis of popular support
for democracy in Southeast Asian countries: are the middle classes
different from the rest?”, draft paper. University of the Philippines School
of Economics, Diliman, Quezon City.
Chang, Y., Y. Zhu, and C. Pak [2007] “Authoritarian nostalgia in Asia”,
Journal of Democracy 18(3): 66-80.
Chang, Y., Y. Chu, and L. Diamond [2012] “A longitudinal and comparative
analysis of citizens’ orientations toward democracy and their evaluation
of the overall performance of the democratic regime in East Asia”, Asian
Barometer Working Paper Series No. 53. National Taiwan University and
Academia Sinica, Taipei, Taiwan.
Chun, N., R. Hasan, and M. Ulubasoglu [2011] “The role of the middle
class in economic development: what do cross-country data show?”, adb
Economics Working Paper Series No. 245. Mandaluyong City, Philippines:
Asian Development Bank.
Dow, J.K. [2009] “Gender differences in political knowledge: distinguishing
characteristics-based and returns-based differences”, Political Behavior
31: 117-136.
Embong, A.R. [2001a] Southeast Asian middle classes: prospects for social
change and democratisation. Bangi, Malaysia: Penerbit Universiti
Kebangsaan Malaysia.
Emerson, D. [1995] “Singapore and the ‘Asian values’ debate”, Journal of
Democracy 6: 95-105.
Erikson, R. and J.H. Goldthorpe [1992] The constant flux: a study of class
mobility in industrial societies. Oxford: Clarendon Press.
Filmer, D. and L. Pritchett [2001] “Estimating wealth effects without
expenditure data—or tears: an application to educational enrollment in
states of India”, Demography 38(1): 155-132.
Fukuyama, F.D. [1995] “Confucianism and democracy”, Journal of
141
Democracy 6(2): 20-33.
Gerring, J., P. Bond, W. Barndt, and C. Moreno [2005] “Democracy and
growth: A historical perspective”, World Politics 57(3): 323-364.
Glassman, R.M. [1995] The middle class and democracy in socio-historical
perspective. Leiden, The Netherlands: Brill Publishers.
Hewison, K. [1999] “Political space in Southeast Asia: ‘Asian style’ and other
democracies”, Democratization 6(1): 224-245.
Kimura, M. [2003] “The emergence of the middle classes and political change
in the Philippines”, The Developing Economies 41(2): 264-84.
Lipset, S.M. [1959] “Some social requisites of democracy: economic
development and political legitimacy”, American Political Science Review
53(1): 69-105.
Loayza, N., J. Rigolini, and G. Llorente [2012] “Do middle classes bring
about institutional reforms?”, Economics Letters 116: 440-444.
Moore, B. [1967] Social origins of dictatorship and democracy.
Harnondsworth: Allen Lane, The Penguin Press.
Oaxaca, R. [1973] “Male-female wage differentials in urban labor markets”,
International Economic Review 14: 693-709.
Robison, R. and D.S.G. Goodman [1996] The new rich in Asia: mobile
phones, McDonald’s, and middle-class revolution. New York: Routledge.
Verba, S., N. Nie, and J.O. Kim [1978] Participation and political equality.
New York: Comparative Politics.
Welzel, C. [2011] “The Asian values thesis revisited: evidence from the World
Values Surveys”, Japanese Journal of Political Science 12(1): 1-31.
Wu, C. and F. Lee [2012] “A comparative analysis of the wealth divide and the
issue of political inclusion”, Asian Barometer Working Paper Series No.
60, Asian Barometer Project Office, National Taiwan University, Taipei,
Taiwan.
Male
Age
Age2
Some college
Employed
Catholic
Buddhist
Islam
In union
Household size
Urban
Mega_majorcity
Access index
Lower class
Middle class
Upper class
Variables
0.87
0.84
4.7
0.30
0.22
68.9
0.41
0.39
0.19
0.34
0.37
2.03
0.46
0.41
20.2
0.49
0.49
0.40
Indonesia
(N=1550)
Mean Standard
deviation
0.50
0.50
41.8
14.1
1946.6 1293.5
0.09
0.29
0.65
0.48
0.74
4.91
0.68
0.84
65.1
0.43
0.39
0.19
0.44
2.27
0.47
0.37
21.9
0.49
0.49
0.39
Philippines
(N=1200)
Mean Standard
deviation
0.50
0.50
40.9
15.4
1906.8 1404.5
0.27
0.45
0.49
0.50
0.84
0.36
0.44
2.16
0.43
14.9
0.49
0.49
0.39
0.75
4.64
0.25
76.78
0.40
0.39
0.19
98.08
0.40
0.41
0.19
0.68
4.12
7.30
0.49
0.49
0.39
0.47
1.43
0.49
0.40
0.94
0.24
Singapore
(N=1000)
Mean Standard
deviation
0.50
0.50
41.5
14.5
1931.8 1370
0.36
0.48
0.61
0.49
Thailand
(N=1512)
Mean Standard
deviation
0.48
0.50
46.9
14.7
2418.5 1466.2
0.13
0.34
0.78
0.42
0.57
0.70
5.65
0.44
0.14
85.69
0.40
0.40
0.20
0.49
0.46
3.16
0.50
0.34
18.12
0.49
0.49
0.40
Malaysia
(N=1214)
Mean Standard
deviation
0.50
0.50
41.4
15.2
1945.1 1362.9
0.16
0.36
0.53
0.50
ANNEX 1. Means and standard deviations of the explanatory variables
142
PRE
The effect of trade policy on firm productivity
in Thai manufacturing
Archanun Kohpaiboon* and Juthathip Jongwanich*
The paper examines the effect of trade policy on firm productivity
using two recent industrial censuses of Thai manufacturing (i.e.,
2006 and 2011). Trade policy and global participation are treated
as two different variables in our analysis. Controlling for firms’
global participation, which is defined as export-sale ratio and the
extent to which raw materials are imported, our study finds that
trade liberalization could induce firms to commit to activities
that improve productivity. The effective rate of protection, where
output and input tariffs are taken into consideration together,
matters in improving firm productivity. Thus, it would be risky to
continue tariff reform by focusing solely on a reduction in input
tariffs while leaving output tariffs untouched. In fact, both input
and output tariffs must be taken into consideration to neutralize
incentives in trade policy reform.
JEL classification: F10, N75
Keywords: trade policy, Thai manufacturing, effective rate of protection, productivity
and developing countries
1. Introduction
While tariffs in general went down substantially in the past two decades after
the establishment of the World Trade Organization in 1995, much remains to be
done in developing countries. In particular, exceptions of liberalization schedules
(i.e., tariff peaks) are often found together with an escalating tariff structure,
where tariff rates are escalating from raw materials to finished products. This
results in nominal protection underestimating actual/effective protection. It is
done with a hope that maintaining such cross-border measures would give more
time for firms to improve their international competitiveness and survive in the
more intense competitive environment.
144
Kohpaiboon and Jongwanich: The effect of trade policy
on firm productivity in Thai manufacturing
Whether the hope materializes or not is an empirical research question. Until
the new millennium, there were a number of empirical studies examining the
relationship between trade liberalization and firm performance. These studies can
be categorized into two main groups. The first examines the relationship between
trade openness and output growth through cross-countries econometric analyses
at the aggregate level.1 The second is based on case studies of specific countries.2
The main finding is in favor of the hypothesis that trade protection retards firm
performance and, eventually, medium- to long-term growth.
However, both groups of studies are subject to shortcomings. In the first group,
the link between trade policy and firm performance is weak and, to a certain
extent, treated as a black box. They are subject to serious econometric problems
in terms of endogeneity and measurement errors. As argued in Levine and Renelt
[1992] and Sala-i-Martin [1997], cross-country econometric analyses are tenuous
at best. In the second group, there are policy insights, but their case studies
include only a handful of countries and their analytical tools vary significantly
across studies. This makes it difficult to generalize the findings of these studies.
From the late 1990s, the proliferation of plant-level data available in
many countries allows researchers to re-visit and mitigate some of the above
shortcomings. Where the effect of trade policy on firm performance is concerned,
there are at least two aspects relevant in policy circles of developing countries.
The first aspect is related to self-selection hypothesis.3 The positive
relationship found between trade liberalization and performance could be due
to a self-selection process, in which firms that enter export markets are already
more productive than non-exporters before they ever actually begin to export. The
self-selection hypothesis suggests that the nature of trade policy and firm market
orientation must be treated as two separate explanatory variables in the analysis.
While both could be important to productivity, they are two different things. The
former refers to the policy environment, while the latter is a firm decision.
The second aspect of research in this field is to examine channels through
which trade policy affects firm performance (Amiti and Konings [2007]; Melitz
and Redding [2012]). In particular, Amiti and Konings [2007] argue possible
different effects of input and output tariffs due to different operating channels.
Lower output tariffs can increase productivity by inducing tougher competition,
whereas cheaper imported inputs can raise productivity via learning, variety, and
quality effects.
See the literature reviews of this group of studies in Edwards [1993].
Two influential works include the Bhagwati-Krueger project for the National Bureau of Economic
Research in the 1970s and the Papageorgious-Michalely-Choksi study for the World Bank in the 1980s.
3
See the literature review in Lopez [2005].
1
2
145
However, a number of case studies4 point to the fact that introduction of new
technologies does not guarantee that productivity increases instantaneously. Most
likely, a long process of learning and mastery of skills may be required to reach
high levels of productivity. Actually, a long-term commitment and real resources
are required to have substantial effects on productivity. This can be influenced
by the policy environment such as trade policy. As argued in these case studies,
firms receiving the effective rate of protection (erp) tend to be “unresponsive
to improved technological capability”. It does not matter whether effective
protection is mainly driven by output or input tariffs. Hence, a decomposition of
the erp into output and input tariffs might not be appropriate.
Against this backdrop, the paper aims to examine the determinants of firm
productivity with emphasis on the effects of trade policy. The 2006 and 2011
industrial censuses are used in our analysis. Our proposed study has at least three
contributions to the existing literature:
• First, we carefully distinguish the possible effects between trade (export and
import) and trade policy (e.g. cross-border protection) on productivity. The
latter has important policy implications for trade policy reform.
• Second, we examine whether the effect of trade policy varies across firm types
by introducing the interaction term between firms’ specific and trade policy
variable. The former focuses market orientation (whether firms participate in
the global market) and input sourcing (whether firms import raw materials or
intermediates from abroad).
• Third, the effect of output and input tariffs are re-examined as opposed to a
case when using the erp (both input and output tariffs combined). Whether
the effects of output tariffs are actually less than that of input tariffs could have
a substantial policy impact as policymakers in developing countries prefer a
reduction in tariffs on inputs rather than on output.
Thailand is chosen for this study since trade policy reform remains a
challenging issue for policymakers. In particular, efforts to streamline tariff rates
to 3 rates (0-1 percent for raw materials, 5 percent for intermediates, and 10
percent for finished products) are at best far from complete. There is almost a fifth
of tariff lines subject to a 20 percent tariff rate or higher. In addition, by design,
the tariff structure in Thailand is cascading, so that nominal protection tends to
underestimate the effective rate. Hence, effective protection seems to vary across
industries. This allows us to test the effect of protection on firm productivity.
The remainder of this paper is organized as follows. Section 2 presents
the analytical framework. Section 3 briefly discusses trade policy and firm
See the cases of Keesing [1983], Keesing and Lall [1992], Westphal et al. [1979, 1984], Aw and Batra
[1998], Wortzel and Wortzel [1981], Hobday [1995], Pietrobelli [1998], Pack and Saggi [1997], and Nelson
and Pack [1999].
4
146
productivity in Thai manufacturing. Section 4 discusses the model, while section
5 presents the data set used in this study. Section 6 empirically assesses the effects
of trade policy on firm productivity. The final section provides some conclusions
and policy implications.
2. The analytical framework
Gains from trade in terms of output growth/productivity have been studied
for several decades. Beginning with the standard neoclassical trade model,
Ricardian comparative advantage model, and/or Hecksher-Ohlin-based
comparative advantage, gains from trade are derived from resource reallocation
from sectors the country has a comparative disadvantage in to those sectors in
which it has comparative advantage. A country’s comparative advantage is driven
by technology, or resource endowment, or both. Note that under these model
settings, trade could lead to output expansion, but it does not have an impact on
rate of economic growth. In the 1980s, the source of gains from trade shifted to
intra-industry trade. In this model setting, consumers enjoy a variety of products
that are close but are not perfect substitutes. As a result, the market is fragmented
in response to consumer wants and needs, and differentiated goods are produced
under increasing return to scale and traded (Krugman [1979]; Helpman and
Krugman [1985]). Fragmentation of the market results in a struggle by firms to
attain adequate production volumes to cover fixed costs so that trade enlarges
market size.
Note that all the models mentioned above are still based on a representative
(homogenous) firm. They cannot explain well why in a given industry only
some and not all firms export [Greenaway et al. 2004]. So the assumption
of a representative firm is relaxed and the literature on firm heterogeneity has
grown (Melitz [2003]; Bernard et al. [2003]). The firm heterogeneity literature
argues that even within a narrowly defined industry, some firms are much larger,
more productive, and more profitable than the others [Melitz and Trefler 2012].
International trade makes better-performing firms expand their products into
larger markets, while resources are re-allocated from less productive firms into
productive ones. This, therefore, leads to the improvement of industry efficiency.
Gains from trade are also examined in the economic growth literature. For
example, in extensions of neoclassical models, including the Solow-Swan model
and the Ramsey growth (optimal-saving) model, trade liberalization increases
output level in the economy but not the rate of growth. The rate of growth in
these models depends on growth of input and the rate of technological progress
(Baldwin [1992]; Srinivasan and Bhagwati [1980]). As argued in the HarrodDomar model, trade could generate positive growth effects under a circumstance,
where the marginal product of capital is bounded by some positive number
(Srinivasan [1999]; Lopez [2005]).
147
The effect of international trade on productivity and on economic growth
has been highlighted in the new or endogenous growth theory (Young [1991];
Rivera-Batiz and Romer [1991]; Pack [1994]). Nonetheless, the effect of trade
liberalization on productivity and growth in developing countries is still unclear.
On the one hand, Rivera-Batiz and Romer [1991] argue that international
integration helps to improve technological progress by expanding the size of
the market, which helps to expand innovation activity. In addition, cross-border
technological spillovers could be created from integration so that economic
growth/productivity is promoted. This mechanism occurs in both developing and
developed countries. On the other hand, Young [1991] argues that the positive
effect of trade on growth would be far less in developing countries. While trade
could positively affect labor productivity through learning by doing process,
the gains would be larger for developed countries. These countries produce and
trade products that still need learning by doing activity. But the products that are
usually manufactured in developing countries require less skill and are not much
involved with the learning process.
During the 1990s, studies examining the effect of trade on output growth were
based on the endogenous growth framework. They can be categorized into two
main groups. The first examines the relationship between trade openness and
output growth through cross-countries econometric analyses at the aggregate
level. The second is based on case studies of specific countries.
In the first group, (real) output growth, especially that at the aggregate level, is
used to represent firm performance alongside different measures of trade openness
(Edwards [1993]; Sachs and Warner [1995]; Lopez [2005]). In some studies, trade
measures are instrumented to redress a possible endogeneity problem (Frankel
and Romer [1999]; Alcala and Ciccone [2004]; Noguer and Siscart [2005]).5
Most of the empirical studies find a positive relationship between trade openness
and economic growth.
The second group analyzes the relationship based on case studies of specific
countries. Either a survey or an econometric analysis at the aggregate level is used
(World Bank [1991]; Kohpaiboon [2003]) . As in the cross-countries analysis,
most studies find a positive relationship between trade openness and growth.
However, both groups are subject to shortcomings, which make the link between
trade policy and firm performance remain a black box to a certain extent.6 In
For example, Frankel and Romer [1999], Rodriguez and Rodrik [2001], and Irwin and Tervio [2002]
use geography as an instrumental variable in examining the relationship between trade liberalization and
economic growth.
6
Note that some scholars (Edwards [1993]; Srinivasan and Bhagwati [2001] ) argue that case studies such
as those done by the National Bureau of Economic Research [1978] or World Bank [1991] could provide
good evidence about the effect of trade on growth. However, drawing conclusions about the role of trade
from case studies is still difficult, since they include only a handful of countries while differences in firm/
plant characteristics are not taken into account [Lopez 2005].
5
148
particular, by using aggregate data either cross-country or within a country,
heterogeneity at the firm and industry levels is not taken into account. This tends
to mix countries and industries with very different characteristics (Berry [1992];
Tybout [1996]; Lopez [2005]). As mentioned earlier, allocation within an industry
is crucial in generating efficiency within an industry and in contributing to a
country’s economic growth.
As plant/firm-level panel data become increasingly available in many
countries, researchers have re-visited the trade-growth nexus. Shortcomings,
especially econometric problems, were better addressed. Using the micro data set,
a new hypothesis is formed relating to the effects of trade liberalization on firms’
performance. The positive relationship found between trade liberalization and
performance is explained by two competing theses.
First is the self-selection thesis, in which firms entering the export markets
are already more productive than non-exporters. Hence the positive relationship
would not be directly related to trade liberalization. Also, the nature of trade
policy and firms’ market orientation must be treated as two separate explanatory
variables in the analysis. They are two different things. The former refers to the
policy environment, whereas the latter is a firm’s decision.
The second thesis is the learning-by-exporting thesis. In this thesis, firms
participating in foreign markets are more likely to experience productivity gains
as opposed to non-exporters, as the former would receive new information about
technological progress, product designs, and quality of goods from their foreign
exposure (Grossman and Helpman [1991]; Aw and Hwang [1995]; Bernard and
Wagner [1997]).7
When policy implications are concerned, both these groups of studies argue in
favor of liberalization of trading partners instead of home countries. Reducing trade
barriers between or among trading partners and policies affecting the profitability
of being an exporter in the home country would induce more productivity
improvement in the home country. Interestingly, Melitz and Trefler [2012] argue
for another possible channel where liberalization in the home country could
yield productivity gains that may help exporters to import inputs/technology at
cheaper costs, which induce higher productivity improvements. Trefler [2004] and
Lileeva and Trefler [2010] also argue that liberalization at home could improve the
country’s overall productivity through better resource reallocation.
Empirical studies (Amiti and Konings [2007]; Melitz and Redding [2012])
further analyze the channels through which trade policy affects firm performance.
In particular, Amiti and Konings [2007] argue possible different effects of
input and output tariffs due to differences in operating channels. Two possible
See possible explanations of self-selection—for example, sunk costs, imported technology, and increased
R&D—in Keesing and Lall [1992], Bernard and Jensen [2004], and Lopez [2005].
7
149
explanations are provided. First, it occurs when locally manufactured and
imported intermediates are not close substitutes. Any change in input tariffs would
have a significant effect only on the firms that actually import them. For those
who use locally manufactured intermediates, such change would not have any
significant effects on their behavior. Second, as shown in the firm heterogeneity
literature, switching market orientation between the domestic market and exports
is costly. Hence, changes in output tariffs might not have any significant impact
on those who already export. They just continue in business due to the presence of
sunk and fixed costs within the export business.
Both circumstances above are often observed in a country that is long
engaged with the world and is pursuing a dual-objective trade policy, where
being reluctant to lower tariffs is associated with the introduction of input tariff
exemption schemes to promote export-oriented activities. Interestingly, based on
Indonesian plant level data from 1991 to 2001, Amiti and Konings [2007] show
that the effects of input tariffs are more favorable than that of output tariffs in
generating firms’ productivity improvement. However, some studies8 argue that
firms’ productivity improvement is a long process of learning and mastery of
skills so that both input and output tariffs facing by a firm should be considered
together in affecting firms’ performance.
3. Trade policy and firm productivity in Thai manufacturing
In Thailand, a tariff is the core measure in conducting trade policy. Non-tariff
measures have been occasionally used in a narrow range of products, mainly in
certain sensitive agricultural products such as soybean, palm seed, silk, and milk.
Like other developing countries, a high tariff level associated with an escalating
tariff structure was used to promote industrialization from the 1960s to the mid1980s. From 1983 to 1995, tariff levels remained virtually unchanged with few
exceptions, whereas the effort to promote Thailand as an export platform for
multinationals was done through the introduction of various tariff exemption
schemes (Kohpaiboon [2006]; Kohpaiboon and Jongwanich [2007]).
As part of its commitments under the World Trade Organization, a
comprehensive plan for tariff reduction and rationalization was proposed in
1990 and implemented in 1995 and 1997. Maximum tariffs were reduced from
100 percent to 30 percent. Tariffs were significantly lowered on some 4,000
items (at the 6-digit HS level) or 75 percent of total tariff lines. By the end of
the 1990s, the tariff bands were reduced from 39 to six (0, 1, 5, 10, 20, and 30
percent). Nonetheless, there were numerous exceptions whose tariff rates exceed
30 percent. Tariff restructuring has received renewed emphasis as an essential
part of the overall economic reforms aimed at strengthening efficiency and
See details in footnote 6.
8
150
competitiveness (Warr [2000]; World Trade Organization [1999]) . The Thai
government introduced tariff cuts, commencing in June 2003 (implemented in
October 2003), followed by a four-year period of tariff reduction from 2004 to
2008. In 2010, there are around 900 items involved in tariff reduction process,
covering a wide range of manufacturing intermediates such as rubber and articles
thereof (HS40), glass and glassware (HS70), knitted fabrics (HS60), other base
metals (HS81), woven fabrics (HS58), articles of stone (HS68), man-made
staple fiber (HS55), wadding yarns (HS56), cotton (HS52), and miscellaneous
vegetable preparations (HS21) [Kohpaiboon and Jongwanich 2007].
Table 1 presents the average of most-favored-nation tariffs of Thailand and
selected Asian economies in 2010. It is clear that average tariffs in Thailand
are relatively high compared with other middle-income countries in the region.
Interestingly, the weighted average was lower than the unweighted one, implying
that tariffs imposed on certain products are redundant. In general, agricultural
products are subject to higher tariffs than manufacturing products.
TABLE 1. Weighted average of most-favored-nation tariff rate of RCEP
members during 2010-2012
Country (year)
Unweighted
Thailand (2011)
8.7
5.0
9.0
4.9
Viet Nam (2010)
9.8
12.2
24.4
10.7
Singapore (2011)
Weighted
Agricultural
products
Non-agricultural
products
0
0
0
0
Philippines (2011)
6.2
12.2
23.2
10.4
Myanmar (2011)
5.6
6.6
12.6
4.9
Malaysia (2012)
5.3
6.7
8.7
6.5
Indonesia (2012)
6.6
9.8
1.8
11.1
Laos (2008)
9.7
13.6
19.3
12.6
Brunei (2011)
2.5
1.7
0
2.6
10.9
12.0
14.7
11.1
Australia (2011)
2.8
3.8
1.6
3.9
New Zealand (2011)
2.0
2.7
1.9
2.8
China (2010)
9.9
8.6
21.5
7.4
India (2012)
13.3
9.4
48.6
7.7
Japan (2011)
3.0
2.1
7.0
1.3
11.2
9.6
34.1
5.6
Cambodia (2012)
South Korea (2011)
Source: Authors’ calculations using most-favored-nation tariff rates from the World Trade Organization
Table 2 presents the distribution of tariff lines in Thailand over the past two
decades. Clearly, the distribution changed as a result of the comprehensive tariff
reform in the mid-1990s. During the pre-1997 period, more than a quarter of
total HS6 tariff lines had tariff rates in the category of more than 30 percent. The
bracket of 15-30 percent also accounted for 30 percent of total tariff lines. After
1997, there was a dramatic shift of tariff lines to lower brackets. For example,
more than 50 percent of tariff lines are in the 0-10 percent bracket. This is followed
151
by 3.9 percent in the 10.1-15 percent bracket and 21.4 percent in the 15.1-20
percent brackets. The share of tariffs above 20 percent dropped from 40 percent
in the pre-1997 period to around 20 percent during the post-1997 period. The
2003 tariff reduction plan marginally changed the distribution of tariff lines.
Tariff reductions in 2003 have basically involved shifting the tariff lines from the
16-20 percent bracket to a lower bracket, with little impact on those belonging
to the above 20-percent brackets. The proposed changes from the next two
years seem to follow the same pattern, while changes proposed for 2006-2008
seem negligible.
TABLE 2. Share of 4-digit HS categories of applied tariff rates in Thailand,
1989–2008
Tariff bands
1989
1995
2002
2003
2.5
2.6
5.6
5.7
6.0
0.1 - 5
14.4
17.3
33.3
37.7
48.8
5.1 - 10
14.2
17.6
14.1
14.2
14.8
10.1 - 15
12.7
3.2
3.9
4.5
3.6
15.1 - 20
15.4
16.4
21.4
17.9
8.4
20.1 - 30
15.8
16
13.8
14.3
12.7
30 - 100
25
26.8
7.8
5.8
5.7
0
2004 to 2008
Source: Data for 1989 and 1995 from World Trade Organization. Data for 2002 to 2008 are from the
authors’ compilation from Official Document provided by Ministry of Finance.
Table 3 presents statistical indicators of how firms engage globalization,
i.e., export-sales ratio and ratio of raw material imports to total materials used,
both measured in percent. They are reported in terms of average, maximum, and
minimum. They are classified into 5 categories across the erp figures. Patterns
observed in Table 3 tend to be in line with the theoretical postulations of the firm
heterogeneity literature. Engaging exports incurs fixed and sunk costs. Therefore,
in a given industry, regardless the erp figure, only some firms export. There is
a vast difference between maximum and minimum values in all erp categories.
In theory, industries subject to high and positive erp tend to sell their products
locally in order to reap economic rents induced by protection. This pattern is, to a
certain extent, found in Table 3. The mean value of the export-sales ratio of firms
located in industries with around zero erp and negative erp is higher than those
experiencing positive erp. Interestingly, such a pattern is not clearly observed
when using a raw material import criterion. The mean value of percent of raw
material imports to total used swings up and down across the erp categories.
Industries subject to around zero erp exhibit the highest raw material import
ratio, followed by those in highly positive erp. Firms in highly negative erp and
moderate positive erp have the same figure of raw material import ratio. The
unclear pattern is due to the extent to which domestic and imported raw materials
are substituted varies across industries instead of protection.
152
TABLE 3. Market orientation and raw material sourcing behavior of Thai
manufacturing firms in 2011
Export-sale ratio (percent of total sales)
Mean
Max
Min
Highly Negative ERP (<-10%)
9.1
88.8
0.0
Moderately Negative ERP (-2% to 10%)
8.3
85.5
0.0
Around Zero ERP (-2% to 2%)
9.0
81.3
0.0
Moderate Positive ERP (2% to 10%)
7.1
79.6
0.0
Highly Positive ERP (>10%)
8.8
85.9
0.0
Raw material import as a percentage of total raw materials used
Highly Negative ERP (<-10%)
8.6
90.3
0.0
Moderately Negative ERP (-2% to 10%)
6.8
81.3
0.0
Around Zero ERP (-2% to 2%)
9.5
87.9
0.0
Moderate Positive ERP (2% to 10%)
8.6
86.4
0.0
Highly Positive ERP (>10%)
9.3
83.6
0.3
Source: Authors’ compilation from the 2011 industrial census
4. The model
The model used here starts with trans-log production function of the firm.
The plant’s value added is a function of two primary inputs (labor and capital),
their squared terms, and their interaction. Labor is further disaggregated into
production (PLij) and nonproduction (NLij) workers to capture their difference in
contributing to firm productivity. Blue-collar workers are regarded as the former
and white-collar ones are the latter. Over and above, a set of firm- and industryspecifics as well as the trade policy variables are included as controlling variables
as expressed in Equation 1.
ln VAij = β0 + β1 ln Kij + β2 (ln Kij)2 + β3 ln PLij + β4 ln NLij + β5 (ln PLij)2
+ β6 (ln NLij)2 + β7 ln PLij * β8 ln NLij * ln Kij + γ1 FSij + γ2 ISij
+ γ3 tradepolicyj + εij (1)
where VAij = Value added of firm i in industry j
Kij = Capital used by firm i in industry j
PLij = Production workers employed by firm i in industry j
NLij = Non-production workers employed by firm i in industry j
FSij = Firm-specific characteristics of firm i in industry j
ISij = Industry-specific characteristics of industry j
tradepolicyij = The nature of trade policy of industry j
153
Firm-specific characteristics (FSij) include market orientation, foreign
ownership (ownij), and R&D investment (RDij). In this study, two aspects of
market orientation are captured: the export-sales ratio; and how much finished
products are exported as a percentage of total sales (mktij). Both are introduced in
the model. International competition would make firms alert to any productivity
improvement and eventually enhance firm productivity so that the coefficient
associated with mktij is expected to be positive. The second aspect of market
orientation is the extent to which imported raw materials are used as a percentage
of total raw materials (rawmij). Firms that import raw materials would benefit from
technology embodied in them, thus improving their productivity. The coefficient
associated rawmij is also expected to be positive.
The consensus in the foreign direct investment literature [Caves 2009]
suggests that foreign firms (ownij) are generally more productive than indigenous
counterparts so ownij is expected to be positive. ownij is measured by firms’ foreign
equity (percent) share. RDij, measured here by the firm’s research, planning, and
development expenditure to total sales, would raise firm productivity so that the
coefficient associated is expected to be positive.
Three industry-specific factors are controlled in our analysis. The first is the
extent to which an industry engages into a global production network. This can
have an implication on productivity as reviewed in part 2, under gains from intraindustry trade. Ideally, details at the firm level (e.g., whether firms are actually
engaged in multinational enterprises' (mnes) production sharing and whether
they import tailor-made raw materials for specific customers, etc.) are needed.
Unfortunately, such details at the firm level are not available within the Thai
dataset. To overcome the unavailability of perfect measures of global production
sharing, two alternative proxies are used in this study. The first two proxies are
shares of parts and component in total imports (gpn1j) and total trade (gpn2j) as
reflected in Equations 3 and 4:
(gpn1j) = P&C Importsj / Total Importsj(3)
(gpn2j) = P&C trade (import + export) / Total Trade(4)
The higher the share, the more important the global production sharing is to the
industry. The parts list is the result of a careful disaggregation of trade data based
on the Revision 3 of the Standard International Trade Classification extracted
from the un Comtrade database. It is important to note that the un Comtrade
database does not provide for the construction of data series covering the entire
range of fragmentation-based trade. The parts list used here is from that developed
in Athukorala and Kohpaiboon [2009].9 To convert Standard International Trade
The use of lists of parts in the Board Economics Classification 42 and 53 is a point of departure. Note
9
154
Classification to International Standard Industrial Classification (isic) , standard
concordance is applied.
The second industry-specific factor is producer concentration (CRj). Its effect
on productivity is ambiguous. Industries with high barriers to entry are likely to be
concentrated and are often capital- and/or skill-intensive. Hence, this could make
firms less responsive to any technological improvement so it negatively affects
productivity (negative sign). On the other hand, as argued in the well-known
creative destruction thesis by Schumpeter, a highly concentrated industry would
give firms incentive to innovate. If so, the coefficient associated with producer
concentration could be positive. Producer concentration is measured by the sum
of the sales share of the top-4 firms in total.
tradepolicyj is introduced to examine the study’s main hypothesis. Two
alternatives of trade policy are used in this study. Effective rate of protection
(ERPj) is used as the first measure of trade policy due to its theoretical superiority.
In theory, policy-induced incentives from cross-border protection measures like
tariffs would not be different, regardless of the incentives generated by either
input or output tariffs. Hence, it would be theoretically superior to employ an
effective rate of protection (ERP), instead of separating input and output tariffs.
n
n
ERPj = (tj – �akj tk) / (1 – �akj)
k=1
(6)
k=1
where tj = Tariff on outputs on industry j
tk = Tariff on inputs k
akj = A value share of inputs k used in finished products on industry j
To examine whether the effect of trade policy varies across firms, the
interaction term between firm specific and trade policy variable is introduced;
ERPj*mktij and ERPj *rawnij are introduced. The former implies that giving
protection to an industry of interest, the effects could vary according to the
extent to which firms export their products to the world. Similarly, in the latter,
the effect of protection on firm productivity could depend on how much a firm
is integrated globally through importing raw materials and intermediates. In
addition, the interaction term between the erp and ownership (ERPj * ownij) is
introduced because a foreign firm might behave differently under different trade
that the parts in Board Economics Classification 211 are not included as they are primary products that
are usually classified as traditional, rather than fragmented-intermediates. The additional lists of parts are
included based on firm interviews reported in Kohpaiboon [2010]. Data on trade in parts are separately
listed under the commodity classes of machinery and transport equipment (Standard International Trade
Classification 7) and miscellaneous manufacturing (Standard International Trade Classification 8) and are
based on firm interviews elaborated in Kohpaiboon [2010]. The list of parts and components is available
on request.
155
policy environments (known as Bhagwati’s hypothesis).10 Trade liberalization
could provide an incentive for foreign firms to behave productively. By contrast,
rent-seeking behavior of foreign firms, which is more likely to occur under trade
restriction, could retard overall productivity improvement.
As discussed in Section 2, input and output tariffs could have different impacts
on firm productivity improvement, hence, inputtariffi and outputtariffi , are
separately introduced as alternative measures of trade policy here. That is,
ERPj + ERPj * mktij + ERPj * rawnij ;
n
n
n
k=1
k=1
k=1
�tj – �akj tk� + �tj – �akj tk� * mktij + �tj – �akj tk�* rawnij (7)
where outputtariffj = Tariff on outputs of industry j (tj)
n
inputtariffj = The weighted average of input tariff from k = 1,…,n. ��akj tk�
k=1
Note that to mitigate any possible endogeneity problem from these industryspecific factors, all of them are lagged.
All in all, the empirical model to be estimated is as follows:
ln VAij = β0 + β1 ln Kij + β2 (ln Kij)2 + β3 ln PLij + β4 ln NLij + β5 (ln PLij)2
+ β6 (ln NLij)2 + β7 ln PLij * β8 ln NLij * ln Kij + γ1 ownij + γ2 R&Dij
+ γ3 rawnij + γ4 mktij + λ1 CRj, t–j + λ2 gpnj, t–j + λ4 tradepolicyj, t–j
tradepolicyj * mktij + φ2tradepolicyj, t–j * rawnij + εij (8)
where ln VAij = Value added of firm i in industry j (in natural log)
ln Kij = Capital used by firm i in industry j (in natural log)
PLij = Production workers employed by firm i in industry j
NLij = Non-production workers employed by firm i in industry j
tradepolicyj, t–j = Lag variable of trade policy measured alternatively by
1. Effective rate of protection (erp) n
2.Outputtariffi (tj) and Inputtariffi (tj) � �akj tk�
k=1
mktij = Market orientation of firm i of industry j measured by a percentage
of export to total sales
rawmij = Input sourcing of firm i of industry j measured by a percentage of
imported raw materials and intermediates to total inputs
ownij = Ownership of firm i of industry j measured by a share of
foreign owners in total capital
10
See the discussion in Kohpaiboon [2006].
156
R&Dij = A share of R&D expenditure as a percent of total sales of firm i of
industry j
CRj, t–j = Producer concentration ratio of industry j at time t-j
gpnj, t–j = The degree of industry involved in the global production networks
of industry j at time t-j, measured by two alternatives:
gpn1j, t–j = The share of parts and components imports to total import
at the 4-digit isic
gpn2j, t–j = The share of parts and components trade (export+import)
to total trade at the 4-digit isic
5. Data set and cleaning procedure
The data set suitable for the current purpose is a long-panel data of
establishments in Thai manufacturing, covering before and after major trade
reform. Unfortunately, such a data set is not available in the country. So far
Thailand has three industrial censuses—1996, 2006, and 2011—all of which are
cross-sectional in nature. These three censuses are not able to formulate as a panel
data set as the identification number used in each census is assigned differently.
In particular, a given identification number of two different censuses does not
necessarily refer to the same firm.
The latest census (2011) contains 98,482 observations. Out of the total,
71,387 observations are self-employed (zero record of paid workers) or microenterprises (less than or equal to 10 workers). Given the current research focus,
we exclude these self-employed and micro-enterprises. Hence, the remaining
observations number 27,095. Similar to what occurred in the censuses in 1996
and 2006, there are many duplicate samples in which at least two observations
report the same value in most of variables. To identify the duplicated
observations, the criterion is if samples report identical values of 7 key variables,
they are treated as duplicated samples. The 7 key variables include total workers,
female workers, initial fixed asset, ending fixed asset, registered capital, sale
value, and input values. In this case, we count only one firm. According to this
criterion, there are 4,418 duplicated samples to be removed. The remaining
observations number 22,677.
Next, we drop observations reporting unrealistic values of the key variables.
They include negative value added, low value added (less than 10,000 baht), and
low fixed assets (less than 10,000 baht). Finally, 8 industries that either serve
niches in the domestic market (e.g. processing of nuclear fuel, manufacture of
weapons and ammunition), in the service sector (e.g. building and repairing of
ships, manufacture of aircraft and spacecraft, and recycling), or are explicitly
reserved for local enterprises (e.g. manufacture of ovens, furnaces and furnace
burners, manufacture of coke oven products) are excluded. All in all, 13,593
observations remain. Summary statistics and correlation of variables are shown
in Tables 4 and 5.
157
TABLE 4. Data summary statistics
Variables
Mean
Min
Max
VAij
15.88
2.40
9.21
25.26
PLij
0.99
1.39
0
8.88
NLij
3.67
1.13
0
9.50
15.84
2.32
9.21
26.32
ownij
4.19
17.21
0
100
mktij
7.430
21.90
0
100
6.27
18.53
0
100
R&Dij
-11.87
6.85
-13.82
20.51
ERPj
0.05
0.17
-0.58
0.60
cr4j,t-j
0.45
0.09
0.32
0.65
GPN1 j,t-j
0.04
0.12
0
1
GPN2 j,t-j
0.03
0.11
0
1
Kij
rawmij
Std. Dev.
Note: All variables are in logarithm, with the exception of ownership, market-oriented,
imported raw material, trade policy, concentration ratio, and production network.
Source: Authors’ calculations
6. Results
Initially, the equations are estimated using the ordinary least squares method
while paying attention to the possible presence of heterogeneity and outliers. Due
to the nature of cross-sectional data, it is likely that outliers could have an impact
on the estimated parameters and lead to misleading inferences. Therefore, careful
treatment of outliers is needed. Cook’s Distance13 is used to identify suspected
outliers. The intra-class correlation or the clustered data, based on industry level,
is tested (Table 6). The results show a low level of the correlation (0.267).
Tables 7 and 8 present estimation results where trade policy is measured
by the erp and tariffs of output and inputs are separately introduced. Column
A in both tables is based on gpn1, whereas Column B is based on gpn2. The
overall results from both tables are largely similar. The estimation results are not
sensitive to choices of gpn. Hence, the following result interpretation will be
discussed, based on these two tables. Coefficients corresponding to the interaction
term between nonproduction workers and capital as well as the squared terms
of two types of workers are statistically significant, suggesting that the translog production function fits the data well, relative to the more restrictive CobbDouglas one. The statistical difference of coefficients associated with production
and nonproduction workers supports the hypothesis that quality of labor matters
in determining firm productivity. The higher the number of white collar workers
employed by firms, the greater the productivity improvement expected, all other
things remaining constant.
0.61
0.54
0.81
0.23
0.29
0.25
0.23
-0.01
0.08
0.14
0.02
0.14
0.13
NLij
Kij
ownij
mktij
rawmij
R&Dij
ERPj
Output tariff
Input tariff
cr4j,t-j
GPN1 j,t-j
GPN2 j,t-j
0.13
0.12
-0.01
0.10
0.06
0.003
0.20
0.22
0.33
0.22
0.57
0.46
1.00
PLij
1.00
PLij
VAij
VAij
Variables
0.09
0.10
0.01
0.10
0.07
0.01
0.19
0.23
0.24
0.23
0.49
1.00
NLij
0.10
0.10
0.06
0.12
0.08
-0.01
0.21
0.23
0.25
0.21
1.00
Kij
0.11
0.11
-0.004
0.04
0.07
0.08
0.05
0.34
0.37
1.00
ownij
0.05
0.08
0.02
0.06
0.09
0.02
0.11
0.36
1.00
mktij
0.08
0.07
0.02
0.00
0.04
0.05
0.11
1.00
rawmij
0.73
1.00
0.03
0.03
-0.02
0.20
0.14
0.06
0.21
0.17
0.04
0.32
1.00
ERPj Output
tariff
0.03 -0.19
0.02
-0.02
1.00
R&Dij
TABLE 5. Correlation matrix
0.24
0.24
-0.05
1.00
Input
tariff
0.14
0.13
1.00
cr4j,t-j
0.92
1.00
1.00
GPN1 j,t-j GPN2 j,t-j
158
TABLE 6. Intra-group correlation
Number of obs = 13593
R-squared = 0.26
Source
Between isic_obs
Within isic_obs
Total
SS
22120.752
62257.048
84377.8
Df
60
13532
13592
MS
368.679
4.601
6.21
F
80.13
Intra-class correlation
Asy. S.E.
95% Confidence Interval
0.267
0.056
0.16
Prob>F
0.000
0.37
Estimated SD of isic_obs effect
1.29
Estimated SD within isic_obs
2.14
Est. reliability of a isic_obs mean (evaluated at n = 217.59)
0.98
Source: Authors’ estimates
TABLE 7. Productivity determinants based on the ERP and 2011 census
Variables
Intercept
PLij
NLij
Kij
PLij*Kij
NLij*Kij
PLij2
NLij2
Kij2
ownij
ownij* ERPj, t-j
mktij
rawmij
R&Dij
ERPj, t-j
ERPj, t-j *mktij
ERPj, t-j *rawmij
cr4j, t-j
GPN1j, t-j
GPN2j, t-j
# obs
Ad-R
F-stat
Column A
Coefficient
t-statistics
2.17*
4.27
2.27*
23.86
0.41*
3.51
0.82*
-0.12*
0.0007
0.04*
0.003
-0.003
0.003*
0.001
0.002*
0.002*
0.013*
-0.30*
-0.01**
0.003
-0.35*
0.64*
12.18
-20.46
0.07
6.36
0.28
-1.22
3.55
0.18
3.89
3.05
8.13
-3.83
-1.97
0.77
-3.05
6.95
13593
0.73
1717 (p-value = 0.00)
Column B
Coefficient
t-statistics
2.12*
4.18
2.27*
23.85
0.41*
3.50
0.82*
-0.12*
0.0007
0.04*
0.003
-0.004
0.003*
0.001
0.002*
0.002*
0.013*
-0.29*
-0.006**
0.003
-0.32*
12.26
-20.42
0.08
6.24
0.30
-1.29
3.65
0.18
4.09
2.96
8.16
-3.78
-2.13
0.83
-2.79
0.55*
5.08
13593
0.73
1711 (p-value = 0.00)
Notes:
* statistically significant at 1 percent ** statistically significant at 5 percent
An increase in the ERP reflects higher trade protection.
159
160
TABLE 8. Productivity determinants based on ERP decomposition
and 2011 census
Variables
Column A
Coefficient
Column B
t-statistics Coefficient t-statistics
Intercept
2.17*
4.26
2.12*
4.17
PLij
2.26*
23.74
2.26*
23.74
NLij
0.41*
3.53
0.42*
3.53
Kij
PLij*Kij
0.80*
11.86
0.80*
11.91
-0.12*
-20.37
-0.12*
-20.34
NLij*Kij
0.00
0.00
0.00
0.01
PLij2
0.04*
6.39
0.04*
6.29
NL
0.004
0.39
0.004
0.39
Kij2
-0.003
-1.01
-0.003
-1.06
2
ij
ownij
0.003*
4.12
0.003*
4.23
mktij
0.004*
2.53
0.004*
2.62
rawmij
0.006*
4.12
0.006*
4.05
R&Dij
0.013*
outputtariffj
-0.27
inputtariffj
8.22
0.13*
8.25
-1.02
-0.27
-1.00
3.02*
4.90
3.26*
5.30
-0.02*
-2.16
-0.02*
-2.33
Inputtariffj*mktij
-0.01
-0.30
-0.007
-0.27
Outputtariffj*rawmij
-0.004
-0.30
-0.003
-0.25
Inputtariffj*rawmij
-0.09*
-2.79
-0.09*
-2.76
cr4j, t-j
-0.32*
-2.77
-0.29
-2.48
0.54*
5.78
Outputtariffj*mktij
GPN1j, t-j
GPN2j, t-j
0.43*
# obs
13593
Ad-R
F-stat
3.89
13593
0.73
0.73
1572 (p-value = 0.00)
1568 (p-value = 0.00)
Notes:
* statistically significant at 1 percent Source: Authors’ estimates
** statistically significant at 5 percent
In both tables, the coefficients corresponding to ownij , mktij , and rawmij turn
out to be positive and significantly different from zero at 5 percent. This finding
is in line with previous studies. That is, foreign firms tend to be more productive
than indigenous ones, all other things remaining constant. Meanwhile, whether
domestic or foreign, the firms that engaged in international business (either
exporting their products, or importing raw materials, or both) tend to be more
productive than those strictly engaged in local markets. Similarly, everything else
being equal, the positive sign of R&D suggests firms spending more on R&D
tend to have higher value added.
161
For industry-specific factors, our study finds a negative and statistically
significance effect of CR4 at the 1 percent. The negative sign suggests that
industries with high barriers to entry or are concentrated tend to make firms
less responsive to any technological improvement. Both gpn1 and gpn2 are
positive and significant at 1 percent, confirming the robustness of the finding
that participating in global production networks could result in higher firm
productivity improvement. This finding is consistently with Kohpaiboon and
Jongwanich [2014] that Thai firms participating in global production networks
go beyond simple assembly, and a statistically significant wage premium in the
industries engaged in such networks is found.
Regarding the effects of trade policy, the coefficient corresponding to erp
turns out to be negative and statistically significant (Table 7). All other things
unchanged, firms operating under a regime of higher cross-border protection
have lower productivity. In other words, protection can retard the process of
productivity improvement. This finding is consistent with the findings of previous
studies. The negative effect of the erp on productivity tends to be higher on
exporting firms, as suggested by its statistical significance of the interaction
term between the erp and mktij . When the erp is decomposed into output and
input tariffs, the coefficient associated with Outputtariffi attains the theoretically
expected sign, but it is not statistically significant (Table 8).
The positive and statistical significance of Inputtariffi must be interpreted with
care. First, this finding is in line with that in Table 7, i.e., we found a negative
effect of protection on productivity as discussed above. Consider the erp formula
expressed in Equation 6 above The negative coefficient associated with the erp
will result in a negative coefficient on output tariffs and a positive coefficient
on input tariffs. This suggests that for a given level of output tariffs, lowering
input tariffs would simply increase effective protection to producers. Thus, the
productivity of firms would decline, when output tariff maintained at the same
level. This finding would be highly relevant to policymakers in developing
countries, where policymakers emphasize input tariff reduction while expressing
reluctance to lower output tariffs.
As revealed intensively in a number of case studies as discussed above, the
greater protection granted producers makes them unresponsive to any productivity
improvement activities, including the long process of learning and mastery of
skills, which require long-term commitment and real resources. This finding is
in line with that in Kohpaiboon and Jongwanich [2007]. In particular, when firms
lobby for protection, they consider both input and output tariffs together to attain
the expected effective protection.
Second, the net effect on productivity remains ambiguous as the interaction
term between Inputtariffi and rawmij turns out to be negative and statistically
significant. When firms import raw materials at an amount greater than 33.6
percent, an increase in input tariffs would have a net negative effect of productivity,
162
all other things equal. In addition, domestically manufactured and imported
intermediates in Thai manufacturing are not close substitutes. Any change in input
tariffs will have a significant effect only on the firms that actually import. For
those who use domestically manufactured intermediates, such a change would not
impact those relying entirely on domestic raw materials.
With regards to output tariffs, the negative effect would occur only in exporting
firms. The coefficient corresponding to the interaction term between and is
negative and statistical significant. For a country like Thailand, which has been
long engaged in international markets, firms already make their decisions taking
market orientation and where to sell their products. In a case where output is
subject to higher tariff rates than inputs are, and various tariff exemption schemes
are available, firms choose to either export or sell domestically.11 These firms are
operating in different environments. Exporting firms are more productive than
domestic-oriented ones as the former usually face more intense competition from
the world. Granted protection could keep allow domestic-oriented firms to remain
in business, produce products serving local niches, and compete with the former
on primary inputs like labor. This is especially true for Thailand, where the labor
market has tightened in recent years. This would inflate wages to a certain extent
and unevenly affect these two groups of firms. It is the exporting firms that are
adversely affected by the inflated wage as their output price is given by the world.
To a certain extent, domestic-oriented firms would pass inflated wages on to
output prices.
A reduction in output tariffs could generate a tougher competitive environment
in domestic markets. The less productive firms that are likely to be purely oriented
to the domestic market may be forced out of business. For exporting firms, such
a reduction in output tariffs would not have any direct effect as they sell at the
world price. Instead, the reduction in output tariffs would lower the inflated wage
and relocate workers from less productive and more domestically oriented firms
to more productive and export-oriented ones (i.e., resource reallocation).
Another interesting finding is that interaction term between ownership and
trade policy (both the erp and disaggregated one) is statistically insignificant
(Tables 9 and 10, respectively). This would reflect the dominant role of exportoriented and efficiency-seeking foreign direct investment, which is motivated
by strengthening global competitiveness. These foreign firms tend to be eligible
for tariff exemption schemes so that their behavior would not be altered by
granted protection.
It would be costly for a firm to sell to both domestic and foreign markets simultaneously as they must
deal with administrative complications—such as how much output is to be sold locally, how to refund the
portion of input tariffs paid—as well as cumbersome tariff exemption schemes. This is especially true for
small and medium firms.
11
163
TABLE 9. Productivity determinants based on ERP decomposition, interaction
with ownership and 2011 census
Variables
Intercept
PLij
NLij
Kij
PLij*Kij
NLij*Kij
PLij2
NLij2
Kij2
ownij
ownij*outputtariffj
ownij *inputtariffj
mktij
rawmij
R&Dij
outputtariffj
inputtariffj
Outputtariffj*mktij
Inputtariffj*mktij
Inputtariffj*rawmij
cr4j, t-j
GPN1j, t-j
GPN2j, t-j
# obs
Ad-R
F-stat
Column A
Column B
Coefficient
t-statistics
Coefficient
t-statistics
2.16*
2.26*
0.42*
0.80*
-0.12*
-0.0002
0.04*
0.004
-0.003
0.001
-0.009
0.06
0.004*
0.007*
0.013*
-0.27
-2.98*
-0.02**
-0.02
4.24
23.79
3.54
11.89
-20.41
-0.02
6.37
0.41
-1.02
0.46
-0.57
1.58
2.78
4.32
8.24
-1.01
-4.83
-1.99
-0.68
2.11*
2.26*
0.42*
0.81*
-0.12*
-0.0001
0.04*
0.005
-0.003
0.001
-0.009
0.06
0.005*
0.007*
0.013*
-0.27
-3.21*
-0.02**
-0.02
4.15
23.79
3.55
11.93
-20.38
-0.01
6.26
0.42
-1.07
0.41
-0.56
1.68
2.90
4.27
8.28
-0.99
5.22
-2.16
-0.67
-0.003
-0.098*
-0.33*
0.54*
-0.22
-3.06
-2.82
5.71
-0.002
-0.098*
-0.29*
-0.17
-3.06
-2.54
0.42*
13593
0.73
1447 (p-value = 0.00)
Notes:
3.84
13593
0.73
1444 (p-value = 0.00)
As a robustness check, the empirical model (Equation 8) is re-estimated by
using the previous industrial census (2006). Tables 10 to 12 correspond to Tables
7 to 9, respectively, but they use the 2006 census. To a certain extent, the results
are in line with what are found in the recent census, with a few exceptions of
statistical insignificance in some coefficients. The main finding of Tables 10 to
12 supports the crucial role of trade liberalization on productivity improvement.
Despite being smaller in magnitude, the coefficient associated with the erp
is negative and statistically significant (Table 10). The difference is that the
coefficient corresponding to the interaction term ERPj*mktij in Table 10 is not
164
statistically significant. It becomes statistically significant when the 2011 census is
used. The difference in the results between the 2006 and 2011 industrial censuses
could be due to the differences in labor market conditions. Another difference is
that all interaction terms with input and output tariffs turn out to be statistically
insignificant, although the coefficients corresponding to them attain theoretical
expected sign when 2006 census data are used.
TABLE 10. Productivity determinants based on ERP and 2006 census
Variables
Column A
Coefficient
Column B
t-statistics
Coefficient
t-statistics
Intercept
6.24*
19.89
6.22*
19.79
PLij
1.77*
29.84
1.77*
29.89
NLij
1.69*
34.75
1.69*
34.78
Kij
0.02
0.49
0.03
0.55
PLij*Kij
-0.09*
-23.78
-0.09*
-23.82
NLij*Kij
-0.10*
-31.68
-0.10*
-31.70
PLij2
0.05*
14.08
0.06*
14.13
NLij2
0.06*
18.44
0.06*
18.47
K
2
ij
ownij
ownij* ERPj, t-j
mktij
0.03*
15.61
0.03*
15.56
0.002*
4.02
0.003*
4.24
0.01
0.73
0.009
0.64
0.001**
1.82
0.001*
2.06
rawmij
0.004*
7.91
0.004*
7.87
R&Dij
0.01*
10.28
0.01*
10.22
ERPj, t-j
-0.15*
-2.17
-0.14*
-2.20
-0.0001
-0.08
-0.0005
-0.23
ERPj, t-j *rawmij
-0.45
-1.56
-0.42
-1.46
cr4j, t-j
0.19*
2.77
0.21*
2.97
GPN1j, t-j
0.52*
6.20
ERPj, t-j *mktij
GPN2j, t-j
# obs
Ad-R
F-stat
0.41*
15564
3.93
15564
0.76
0.76
2731 (p-value = 0.00)
2739 (p-value = 0.00)
Notes:
* statistically significant at 1 percent ** statistically significant at 5 percent
An increase in the ERP reflects higher trade protection.
TABLE 11. Productivity determinants based on ERP decomposition
and 2006 census
Variables
Column A
Coefficient
Column B
t-statistics
Coefficient
t-statistics
Intercept
6.31*
20.07
6.28*
19.98
PLij
1.76*
29.70
1.77*
29.74
34.59
NLij
1.69*
34.56
1.69*
-0.0004
-0.01
0.001
0.01
PLij*Kij
-0.09*
-23.74
-0.09*
-23.76
NLij*Kij
Kij
-0.10*
-31.57
-0.10*
-31.59
2
ij
PL
0.06*
14.18
0.06*
14.21
NLij2
0.06*
18.49
0.07*
18.50
Kij2
0.03*
15.86
0.03*
15.84
ownij
0.003*
6.44
0.003*
6.64
mktij
0.003*
2.26
0.003*
2.43
rawmij
0.005*
4.31
0.005*
4.14
R&Dij
0.01*
10.15
0.01*
10.08
outputtariffj
-0.26
-1.09
-0.26
-1.09
inputtariffj
Outputtariffj*mktij
Inputtariffj*mktij
3.29*
5.45
3.53*
5.86
-0.0002
-0.03
-0.002
-0.23
-0.03
-1.10
-0.04
-1.12
-0.009
-0.90
-0.008
-0.81
Inputtariffj*rawmij
-0.02
-0.91
-0.02
-0.77
cr4j, t-j
0.23*
3.35
0.25*
3.61
GPN1j, t-j
0.43*
5.02
GPN2j, t-j
0.30*
# obs
15564
Ad-R
F-stat
2.80
15564
0.76
0.76
2486 (p-value = 0.00)
2493 (p-value = 0.00)
Notes:
165
166
TABLE 12. Productivity determinants based on ERP decomposition,
interaction with ownership and 2006 census
Variables
Column A
Coefficient
Column B
t-statistics
Coefficient
t-statistics
Intercept
6.31*
20.07
6.28*
19.97
PLij
1.76*
29.70
1.77*
29.74
NLij
1.69*
34.56
1.69*
34.61
Kij
-0.002
-0.03
-0.0004
-0.01
PLij*Kij
-0.09*
-23.71
-0.09*
-23.74
NLij*Kij
-0.10*
-31.51
-0.10*
-31.59
PLij2
0.06*
14.13
0.06*
14.17
NLij2
0.06*
18.42
0.07*
18.44
Kij2
0.03*
15.86
0.03*
15.84
0.004*
2.50
0.003*
2.42
ownij*outputtariffj
0.009
0.70
0.008
0.65
ownij *inputtariffj
-0.02
-0.76
-0.01
-0.56
mktij
0.003*
2.09
0.003*
2.28
rawmij
0.005*
4.35
0.005*
4.21
R&Dij
0.01*
10.12
0.01*
10.05
-1.09
ownij
outputtariffj
-0.26
-1.06
-0.26
inputtariffj
-3.30*
-5.46
3.53*
5.86
Outputtariffj*mktij
-0.001
-0.19
-0.003
-0.38
Inputtariffj*mktij
-0.04
-1.17
-0.04
-1.13
-0.01
-1.02
-0.009
-0.91
Inputtariffj*rawmij)
-0.02
-0.76
-0.02
-0.67
cr4j, t-j
0.23*
3.33
0.25*
3.59
GPN1j, t-j
0.43*
5.01
GPN2j, t-j
0.30*
# obs
15564
Ad-R
F-stat
2.77
15564
0.76
0.76
2281 (p-value = 0.00)
2287 (p-value = 0.00)
Notes:
7. Conclusions and policy recommendations
The paper examines the effect of trade policy on firm productivity by using two
recent industrial censuses of Thai manufacturing (2006 and 2011). The translogproduction function is employed to avoid imposing any coefficient restrictions.
Trade policy and global participation are treated as two different variables in
our analysis. Foreign firms tend to be more productive than indigenous ones,
all other things remaining equal. The firms, whether domestic or foreign, that
engaged in global markets tend to be more productive than those strictly engaged
in local markets. As expected, firms spending more on R&D tend to have higher
productivity. Participating in the global production network could result in firm
productivity improvement.
167
While controlling for firms’ global participation, defined as export-sale ratio,
and the extent to which raw materials are imported, our study finds that trade
liberalization could induce firms to commit to productivity-improving activities.
The key finding that is different from previous studies is that when it comes
to the decision to commit to productivity improvement, the effective rate of
protection, where output and input tariffs are taken into consideration together,
matters. Focusing solely on lowering input tariffs while leaving output tariffs
untouched could retard the overall productivity improvement. This finding would
be highly relevant for policymakers in developing countries, where policymakers
generally emphasize input tariff reductions while expressing reluctance to lower
output tariff.
Two policy inferences can be made from our study. First, our study supports
global integration as this could promote productivity enhancement. Second, it
would be risky to continue tariff reform by focusing solely on input tariffs while
leaving output tariffs untouched. In fact, both input and output tariffs must be
taken into consideration in neutralizing incentives in trade policy reform.
*Thammasat University, Thailand
References
Alcala, F. and A. Ciccone [2004] “Trade and productivity”, Quarterly Journal of
Economics 119: 613-46.
Amiti, M. and J. Konings [2007] “Trade liberalization, intermediate inputs, and
productivity: evidence from Indonesia”, American Economic Review 97(5):
1611-1638.
Aw, B.Y. and A.R. Hwang [1995] “Productivity and the export market: a firmlevel analysis”, Journal of Development Economics 47: 313-332.
Aw, B.Y. and G. Batra [1998] “Technological capability and firm efficiency in
Taiwan (China)”, World Bank Economic Review 12: 59-79.
Athukorala, P. and A. Kohpaiboon [2009] “Intra-regional trade in East Asia: the
decoupling fallacy, crisis, and policy challenge”, adbi Working Paper 177,
Tokyo: Asian Development Bank Institute.
Baldwin, R. [1992] “Measurable dynamic gains from trade”, Journal of Political
Economy 100: 162-74.
Barry, R.A. [1992] “Firm (or plant) size in the analysis of trade and development”,
in G.K. Helleiner, ed., Trade policy, industrialization, and development: new
perspectives. Oxford: Clarendon Press.
Bernard, A.B., E.J. Jonathan, J. Bradford, and S. Kortum [2003] “Plants and
productivity in international trade”, American Economic Review 93(4): 1268-90.
168
Caves, R. [2009] Multinational enterprise and economic analysis. 3rd edition.
Cambridge University Press: Cambridge.
Edwards, S. [1993] “Openness, trade liberalization, and growth in developing
countries”, Journal of Economic Survey 31: 1358-1393.
Frankel, J. and D. Romer [1999] “Does trade cause growth?”, American Economic
Review 89: 379-99.
Hobday, M. [1995] “East Asian latecomer firms: learning the technology of
electronics”, World Development 23:1171-1193.
Helpman, E. and P.R. Krugman [1985] Market structure and foreign trade. mit
Press.
Keesing, D.B. [1983] “Linking up to distant markets: south to north exports of
manufactured consumer goods”, American Economic Review 73: 338-42.
Keesing, D.B. and S. Lall [1992] “Marketing manufactured exports from
developing countries: learning sequences and public support”, in G.K.
Helleiner, ed., Trade policy, industrialization, and development: new
perspectives. Oxford: Clarendon Press.
Kohpaiboon, A. [2006] Multinational enterprises and industrial transformation:
evidence from Thailand. Cheltenham, uk: Edward Elgar.
Kohpaiboon, A. [2010] Product fragmentation phenomenon, production networks
of multinationals and implication on Thai manufacturing (in Thai). Bangkok:
Misterkopy.
Kohpaiboon, A. and J. Jongwanich [2007) “Determinants of protection in Thai
manufacturing”, Economic Papers 26(3): 276-94.
Kohpaiboon, A. and J. Jongwanich [2014] “Global production sharing and wage
premiums: evidence from the Thai manufacturing sector”, Asian Development
Review 31(2): 141-164.
Krugman, P. [1979] “Increasing returns, monopolistic competition, and
international trade”, Journal of International Economics 9: 469-79.
Melitz, M.J. [2003] “The impact of trade on intra-industry reallocations and
aggregate industry productivity”, Econometrica 71(6): 1695-1725.
Melitz, M.J. and S. J. Redding [2012] “Heterogenous firms and trade”, nber
Working Paper 18652, National Bureau of Economic Research, Massachusetts.
Melitz, M.J. and D. Trefler [2012] “Gains from trade when firms matter”, Journal
of Economic Perspectives 26(2): 91-118.
Noguer, M. and M. Siscart [2005] “Trade raises income: a precise and robust
result”, Journal of International Economics 65: 447-60.
Levine, R. and D. Renelt [1992] “A sensitivity analysis of cross-country growth
regressions”, American Economic Review 82: 942-63.
Lopez, R.A. [2005] “Trade and growth: reconciling the macroeconomic and
microeconomic evidence”, Journal of Economic Surveys 19(4): 623-648.
Pack, H. [1994] “Endogenous growth theory: intellectual appeal and empirical
shortcomings”, Journal of Economic Perspectives 8: 55-72.
Pack, H. and K. Saggi [1997] “Inflows of foreign technology and indigenous
169
technological development”, Review of Development Economics 1: 81-98.
Pietrobelli, C. [1998] Industry competitiveness and technological capabilities in
Chile: a new tiger from Latin America? New York: St. Martin’s Press.
Nelson, R. and H. Pack [1999] “The Asian miracle and modern growth theory”,
Economic Journal 109: 416-436.
Rivera-Batiz, L. and P.M. Romer [1991] “Economic integration and endogenous
growth”, Quarterly Journal of Economics 106: 531-56.
Sala-i-Martin, X. [1997] “I just ran two million regressions”, American Economic
Review 87: 178-183.
Sachs, J.D. and A. Warner [1995] “Economic reform and the process of global
integration”, Brookings Papers on Economic Activity 1: 1-118.
Srinivasan, T.N. [1999] “Trade orientation, trade liberalization and economic
growth”, in G.R. Saxonhouse and T.N. Srinivasan, eds., Development, duality,
and the international economic regime: essays in honor of Gustav Ranis. Ann
Arbor, Michigan: University of Michigan Press.
Srinivasan, T.N. and J.N. Bhagwati [1980] “Trade and welfare in a steady-state”,
in J.S. Chipman and C.P. Kindleberger, eds., Flexible exchange rates and the
balance of payments: essays in memory of Egon Sohmen. Amsterdam: NorthHolland.
Srinivasan, T.N. and J.N. Bhagwati [2001] “Trade, development and growth:
Graham memorial lecture”, Essays in International Economics 225: 1-32.
Tybout, J.R. [1996] “Heterogeneity and productivity growth: assessing the
evidence”, in M.J. Roberts and J.R. Tybout, eds., Industrial evolution in
developing countries: micro patterns of turnover, productivity, and market
structure. New York: Oxford University Press.
Warr, P.G. [2000] “Thailand’s post-crisis trade policies: the 1999 wto review”,
World Economy 23(9): 1215-1236.
Westphal, L., Y. Rhee, and G. Pursell [1979] “Foreign influences on Korean
industrial development”, Oxford Bulletin of Economics and Statistics 41: 359388.
Westphal, L., Y. Rhee, and G. Pursell [1984] “Sources of technological capability
in South Korea”, in M. Fransman and K. King, eds., Technological capability
in the Third World. London: Macmillan.
World Bank (WB) [1991] World development report: the challenge of
development. New York: Oxford University Press.
Wortzel, L.H. and H.V. Wortzel [1981] “Export marketing strategies for nic and
ldc-based firms”, Columbia Journal of World Business 16: 51-60.
World Trade Organization (wto) [1999] Thailand: trade policy review. Geneva.
Young, A. [1991] “Learning by doing and the dynamic effects of international
trade”, Quarterly Journal of Economics 106: 369-405.
PRE
Households’ access to financial services:
some evidence from survey data
Gilberto M. Llanto
Many studies look at financial inclusion from the supply side.
The discussion in those studies revolves around the different
types of financial services being developed to provide the
excluded segment of the population with access to such services
and the evolving regulatory frameworks supporting those
innovative financial services.
This paper views financial inclusion from the perspective
of households who use financial services and asks what factors
determine access to financial services. It provides a quantitative
estimation of the factors affecting household decision to
participate in the formal financial markets and the impact of the
utilization of financial services on household incomes. It uses
micro-data from the Annual Poverty Indicators Survey in the
estimation.
The empirical findings provide useful information for
designing policies and interventions to foster inclusive finance. It
points to financial education of households as a key intervention
in financial inclusion strategies.
JEL classification: D14, G21
Keywords: financial inclusion, inclusive finance, access to banking services,
financial education
1. Introduction
Financial inclusion presently occupies center stage in global discussions
of development interventions, and it has drawn the attention of policy makers,
regulators, financial service providers, other stakeholders, and even the support
of the nobility who are concerned with the negative impact on households of the
171
inaccessibility of financial services.1 The importance given to financial inclusion
globally is motivated by the belief that financial inclusion is important for
inclusive growth and poverty reduction. Empirical studies tend to provide some
evidence of the beneficial impacts of financial inclusion at the macroeconomic,
household, and firm level. For this reason, the Global Partnership for Financial
Inclusion was established by the G20 as the main implementing mechanism of
the G20 Financial Inclusion Action Plan. In a recent report on financial inclusion,
the World Bank points out that at the country level about two-thirds of regulatory
and supervisory agencies in many countries are now working on ways to enhance
financial inclusion, while some 50 countries have set formal targets and goals
for financial inclusion [World Bank 2014]. Financial inclusion is an important
strategy for inclusive growth in the 2011-2016 Philippine Development Plan;
the Bangko Sentral ng Pilipinas, taking the lead in expanding the accessibility
of financial services, created the Inclusive Finance Advocacy Staff to work with
various stakeholders in achieving the objectives of financial inclusion.
The Bangko Sentral ng Pilipinas (bsp) defines financial inclusion as “a
state wherein there is effective access to a wide range of financial services for
all Filipinos” [bsp 2013:1]. This follows the standard definition of financial
inclusion in the literature. The Consultative Group to Assist the Poor [2011]
defines financial inclusion as a state in which all working-age adults, including
those currently excluded by the financial system, have effective access to a
range of financial services provided by formal financial institutions: credit,
savings (including current account), payments, and insurance. Effective access
involves convenient and responsible service delivery at a cost affordable to the
customer and sustainable for the provider, while “financially excluded” refers to
those who do not have access to or are underserved by formal financial services
[cgap 2011]. Effective access requires that financial services are appropriately
designed, of good quality, relevant for actual use, and beneficial to the target
market [Llanto 2015:1].
Access to financial services satisfies economic agents’ demand for
consumption smoothing, productive investments, and ways to help them cope
with exogenous shocks, e.g., catastrophic risk. However, a large segment of
the global population, especially poor households and micro-enterprises in
developing countries, has been financially excluded. According to the latest
World Bank estimates, half of the world’s adult population—more than 2.5
billion people—do not have an account at a formal financial institution. Globally,
about 50 percent of adults have one or more bank accounts, and a nearly equal
Queen Maxima of the Netherlands, un Secretary General’s Special Advocate for Inclusive Finance for
Development, lent her presence and support to the launch of the Philippines’ National Strategy for Financial
Inclusion on July 1, 2015 at the Philippine International Convention Center.
1
172
Llanto: Households’ Access to Financial Services:
share are unbanked. In 2011, adults who were banked included the 9 percent
of adults who received loans and the 22 percent of adults who saved through
financial institutions [World Bank 2014:1-2]. In the Philippines, the 2009
Consumer Finance Survey of the bsp found that 8 in 10 Filipino households did
not have a deposit account, only 10.5 percent of adults in the country had a loan
from a formal financial institution, and 93 percent of those without any deposit
account said they did not have enough money for bank deposits [bsp 2012].
To address these issues, innovative financial services intended to address
the problem of financial inclusion—e.g., mobile money, branchless banking,
and e-money—are in varying stages of development and utilization in many
developing countries. For example, in the Philippines low-income households
are primarily using mobile money to send and receive domestic remittances:
on average sending us$57 and receiving us$48 [Pickens 2009]. In Malawi,
Opportunity Bank took two years to develop m-banking service, which was
launched in 2010. Also in 2010, m-pesa and Equity Bank in Kenya announced
the a low-cost, low-entry microsavings account called M-Kesho. The objective
is to convert the majority of m-pesa’s 9.4 million users into account holders at
Equity Bank. There are further plans to offer microinsurance and microloans to
account holders [Kumar, McKay, and Rotman 2010]. Such innovative financial
services will require supportive regulatory frameworks. In this area, Peru and
Philippines have been cited as being more advanced than other developing
countries in the development of such frameworks. According to the bsp [2015],
the Philippines ranked first in Asia and top three in the world in 2014 in terms of
having a conducive environment for financial inclusion based on the Economist
Intelligence Unit maiden survey on financial inclusion environments globally.
The literature has documented a positive relationship between finance and
economic development at the macro level. It has been pointed out that the lack
of access to financial services could lead to a poverty trap and to an increase
in the inequality gap (Cámara and Tuesta [2015]; Beck, Demirguc-Kunt, and
Levnie [2007])2, that inequality decreases as financial markets deepen [Clarke,
Xu and Zou 2006], and that, in the case of India, an all-inclusive financial system
would facilitate the process of human development by addressing the basic
distortions in the level of human development [Kuri and Laha 2011]. Among
others, research at the household level revolves around the link between financial
inclusion and reduction of poverty rates (Honohan [2008]; Park and Mercado
[2015]) and improvements of household welfare with an important function
assigned to financial services as a tool for consumption smoothing and social
protection as in the case of micro-insurance.
The literature on the impact of financial inclusion at the macro and household level is well summarized in
Cámara and Tuesta [2015]. A more copious literature is in World Bank [2014].
2
173
If financial exclusion could have deleterious effects as explained in the
literature, it is important to understand why households, especially poor
households and microenterprises, fail to access financial services. Many studies
on financial inclusion look at it from the supply side. The discussion in those
studies revolves around the different types of financial services being developed
to provide the excluded segment of the population with access to such services
and the evolving regulatory frameworks supporting those innovative financial
services. Financial inclusion is not the same as providing access to financial
services although certainly the first step towards the goal of financial inclusion
is to make those financial services very accessible to the excluded. Financial
products and services could be accessible to the population, but utilization of
such financial services could be low. Thus, there would be a large segment of the
population that will continue to be financially excluded despite the accessibility
of financial services. Financial inclusion is about providing access to financial
services and the excluded households’ and firms’ utilization of those services.
This paper takes it from the perspective of users of financial services, that
is, the households. What prevents those households from accessing financial
services? What influences their decision to use or not to use financial services?
Several demand-side factors have effectively excluded poor households from
accessing and using financial services. There is a range of factors that prevent
access and utilization: socio-economic and cultural factors; the lack of formal
identification needed to satisfy the “know your client” policy imposed on banks
by the regulator; low levels of financial literacy in addition to the absence
of appropriate consumer protection mechanisms (Alliance for Financial
Inclusion [2010]; Llanto [2015]) and lack of awareness of available services;
inappropriateness of certain services to the needs of the low-income sectors;
and the risks of dealing with poor customers [escap 2014]. It is important to
understand the socio-economic characteristics conditioning the use of financial
services by households, which enable such households to smoothen income
cycles generated by unexpected shocks or discontinuous income flows [Cámara
and Tuesta 2015].
Using micro-data from the Annual Poverty Indicators Survey, the paper
provides a quantitative estimation of factors affecting household decision to
participate in the formal financial markets and the impact of utilization of
financial services on household incomes. It also shows that vulnerable groups—
comprised of women, rural dwellers, and young people—find it most difficult to
access banking services.
The empirical findings provide useful information for designing policies
and interventions to foster inclusive finance. It points to financial education of
households as a key intervention in financial inclusion strategies.
174
The paper is organized as follows. Section 2 provides an overview of the
Philippine financial sector and the state of financial inclusion in the country, with
a focus on the critical role of an enabling environment in promoting inclusive
finance. Section 3 discusses the methodology and data used in the empirical
estimation. Section 4 analyzes the empirical findings, and the concluding section
provides some recommendations for fostering financial inclusion and comments
on further research on the subject.
2. Current status of financial inclusion
2.1. Brief profile of the financial sector3
In 2014, the Philippine banking system remained strong and stable despite
external challenges—especially in international capital markets—with
continuing growth in resources, deposit liabilities, and loans. The total resources
of the whole banking system increased by 11.8 percent to more than P11
trillion from P10.3 trillion in the preceding year, 2013. This can be attributed
to growth in loans, financial assets, and equity investments (Table 1). Total
deposits of banks rose to P8.52 trillion in the same period, a 12 percent yearon-year increase from end-December 2013. The number of banking institution
head offices decreased to 648 as of end-December 2014 from the previous year’s
673 head offices, signifying a consolidation of banks and the closure of weaker
banks in the sector (Table 2). The number and types of banking offices are also
shown in Table 2. The notable information here is the growth at 11 percent of
micro-banking offices of mostly microfinance-oriented banks. Together with
microfinance- oriented branches, they cater to the lower-income groups and are
the access points that are accessible to the excluded segments of the population.
Asset quality indicators also improved with the decline of the banking
system’s gross non-performing loan ratio from 2.8 percent as of end-December
2013 to 2.3 percent as of end-December 2014. Likewise, net non-performing
loans were reduced. Capital adequacy ratios remained above the international
standards imposed under the Basel III framework, which became effective
on January 1, 2014. As of end-September 2014, the capital adequacy ratios of
universal and commercial banks stood at 17 percent, while overall the capital
adequacy ratio for all types of banks stood at 16.7 percent at end-September
2014 (Table 1).
This paragraph was drawn from Llanto [2015].
3
175
TABLE 1. Resources, deposits, and loans outstanding, all banks,
December 2014
All
Universal Commercial
Banks
Thrift
Banks
Rural
Banks
1,920
2,608
Number of Banks
Total number of Banks
10,361
6,330
Head offices
648
51
69
543
Other Offices
9,713
6,279
1,851
2,065
Resources (P billion)
11,128
10,398
916
208
Deposits Liabilities (P billion) billion)
8,522
7,680
696
144
Loans Outstanding (P billion)
5,532
4,822
571
138
GNPL to Total Loans (%)
2.3
2.3
4.4
11.9
NNPL to Total Loans (%)
0.6
0.6
1.95
5.9
16.7*
17.0**
Capital Adequacy Ratio (%)
* As of end-June 2014; ** As of end-September 2014
Source: Bangko Sentral ng Pilipinas.
TABLE 2. Number of banking offices, by type, 2013 and 2014
TOTAL
Head Offices
2013
2014
9,935
10,361
Growth Rate (%)
4.3
673
648
-3.7
Branches/Other Offices
9,262
9,713
4.9
Regular Branch
8,077
8,442
4.5
98
99
1.0
Regular Other Banking Office (ROBO)
420
448
6.7
Microbanking Office (MBO)
465
517
11.2
Extension Office (EO)
166
176
6.0
Representative Office
15
13
-13.3
Remittance Desk Office
16
14
-12.5
Marketing Office
2
2
0.0
Sub-Branch
2
1
-50.0
Limited Purpose Branch
1
1
0.0
Micro-finance Oriented Branch
Sources: Bangko Sentral ng Pilipinas
The regional distribution of banking offices in Table 3 gives a rough idea
of the spatial distribution of access to banking facilities. A finer distribution by
municipalities and cities is available at the bsp web site. The population residing
in richer regions, which have more banking facilities, have easier access to
financial services. Those residing in poorer regions with fewer banking facilities
do not have this advantage. Cities/municipalities/provinces/regions with a higher
bank density have more financially included individuals in their respective
176
populations than those with lower bank density. The leading regions from this
perspective are the National Capital Region, Regions III and IV-A in Luzon, and
Regions VI and VII in the Visayas. The regions in Mindanao have fewer banking
offices and presumably have a bigger share of the financially excluded segments
of the population.
TABLE 3. Regional distribution of banking offices, 2010-2015p
Growth
20122013
Growth
20132014
10,410
5.4%
4.4%
100.0
3,299
4.9%
4.3%
31.7
155
156
1.4%
3.3%
1.5
436
456
466
5.6%
4.6%
4.4
286
310
329
337
8.4%
6.1%
3.2
940
975
998
1,033
1,050
2.4%
3.5%
10.0
1,314
1,350
1,406
1,509
1,566
1,575
7.3%
3.8%
15.2
184
189
206
220
234
236
6.8%
6.4%
2.3
Region V
271
281
315
353
380
383
12.1%
7.6%
3.7
Region VI
515
531
543
572
600
604
5.3%
4.9%
5.8
Region VII
580
584
627
653
683
688
4.1%
4.6%
6.6
Region VIII
165
172
174
183
186
187
5.2%
1.6%
1.8
Region IX
171
181
190
199
201
202
4.7%
1.0%
1.9
Region X
320
328
328
347
368
372
5.8%
6.1%
3.6
Region XI
333
338
355
388
396
400
9.3%
2.1%
3.8
Region XII
190
192
195
197
214
216
1.0%
8.6%
2.1
Caraga
193
199
201
207
218
218
3.0%
5.3%
2.1
ARMM
19
19
20
21
21
21
5.0%
0.0%
0.2
2010
2011
2012
2013
2014
Philippines
8,843
9,015
9,375
9,884
10,315
NCR
2,876
2,892
2,993
3,141
3,275
CAR
138
146
148
150
Region I
403
401
413
Region II
257
272
Region III
914
Region IV-A
Region IV-B
2015*
Percent
Share (%)
2014
Note: * - as of March 2015
Source: BSP Statistics
2.2. Financial inclusion
The latest data on financial inclusion are those reported by the bsp from
data gathered through the National Baseline Survey on Financial Inclusion [bsp
2015]. During the launch of the National Strategy for Financial Inclusion on July
1, 2015, the bsp reported the following statistics: 25 percent of Filipino adults
have never saved; 32 percent used to save; and only 43 percent presently have
savings. Of those with savings, only 32 percent save in banks, while 68 percent
keep their savings at home. Around 65 percent of unbanked adults cited lack of
money as the main reason for not having a bank account. About 47 percent of
adults have outstanding loans. The main source of borrowing is informal: 62
percent borrow from family, relatives, or friends; while 10 percent borrow from
informal lenders. About 44 percent of adults sent or received money, while 42
177
percent made payments. Only 3.2 percent of adults have a microinsurance
coverage [bsp 2015].
There are several salient findings of the National Baseline Survey on Financial
Inclusion. In terms of access, the bsp [2015] reports that among the available
access points, Filipino adults are most aware of banks (98.3 percent), pawnshops
(95.7 percent), and automated teller machines (93.5 percent). Seemingly, there is
a relatively low awareness of other access point, e.g. microfinance nongovernment
organizations (30.5 percent), e-money agents (25.6 percent), and non-stock
savings and loan associations (13.6 percent). As of end-December 2014, 36
percent of municipalities do not have a banking office. While the physical network
of banks and atms continues to experience sustained growth, there are disparities
in the regional distribution of access points.
In terms of usage, most of those who are aware of the access points had
also conducted transactions using the previously mentioned top access points.
Nonetheless, only 5 out of 10 Filipino adults have conducted transactions with
banks. Also, there are certain access points which are more frequently used than
others depending on the geographic location (i.e. island group) of the user. For
instance, adults in Mindanao tend to transact more often with cooperatives and
microfinance nongovernment organizations; in Visayas, the majority of adults
have transacted with non-stock savings and loan associations and pawnshops.
Significant usage disparity was also evident between users in urban rural areas
[bsp 2015].
Meanwhile, the proportion of Filipino adults who save remains small at 43.2
percent; while 32.3 percent of the respondents used to save in the past, and the
remaining (24.5 percent) have never experienced saving money. It is worth noting
that the majority (7 out of 10 adults or 68.3 percent) prefer to save their money at
home; 32.7 percent save through banks; and the remaining through other financial
institutions and informal savings group. It seems that such behavior among
most of the Filipino adults stems from the main reasons for saving, i.e. to use in
case of emergencies (63.8 percent); for future expenses on food (55.6 percent);
and education (47.4 percent). This may imply that a significant percentage of
Filipinos would rather forgo the interest income from savings deposits in banks in
exchange for easier access to savings, that is, keeping cash at home. Some of the
other reasons cited for not saving in the banks were lack of money (65 percent),
limited knowledge and capability to manage an account (16.8 percent), cost (11.2
percent), proximity of the banks (7.6 percent), and failure to meet documentary
requirements (4.6 percent), among others. In terms of loans, 47.1 percent of adults
borrow money of whom 61.9 percent borrow from family, relatives, or friends,
and 10.1 percent borrow from informal lenders. Among the main considerations
for borrowing are interest rate and loan amount. On insurance, most are aware of
health and life insurance. Results showed that the most common reasons for not
enrolling in life, health, or accident insurance are lack of money and perception
of high cost.
178
More than half who have accessed banks and automated teller machines
are only somewhat satisfied with their transactions. This is most common in
automated teller machines, cooperatives, and microfinance nongovernment
organizations. Finally, on welfare, the results of the survey indicated that 86
percent of the Filipino adults perceive access to financial products and services is
important, while 88 percent believed that it is beneficial to them. Also, majority
of the adults believe that saving, borrowing, and insurance are important to them
as well. Nonetheless only half of the (potential) borrowers would want to borrow
from financial institutions. The findings of the first National Baseline Survey on
Financial Inclusion indicate that much work has to be done to achieve financial
inclusion in the country. This includes extensive information dissemination,
financial education especially for poor households, and development of financial
products and services that are responsive to the financial needs of the excluded
segment of the population.
It is good to benchmark financial inclusion levels in the Philippines relative
to other countries in order to provide a good perspective and understanding of
the local situation. Data are sourced from the 2014 Global Findex. Account
penetration in the Philippines was lower than all of the countries within the
peer group (asean), except in Vietnam and Cambodia (Figure 1). Account
penetration in the Philippines was lower than all of the countries within the peer
group (asean) except in Vietnam and Cambodia (Figure 1). The percentage of
adults who have savings in a financial institution was lower than all countries
within the peer group except Cambodia. The Philippines and Viet Nam have
more or less the same percentage of adults with savings in a financial institution.
(Figure 2). Surprisingly, the percentage of adults with loans obtained from a
financial institution was lowest in the Philippines (Figure 3).
3. Data and methodology
The methodology for estimation is as follows. Using the Heckman selection
model estimation, the likelihood of availing of loans regardless of the source of
loans—that is, loans from formal institutions or informal lenders—was first tested,
followed by the likelihood of getting a formal loan (access to formal credit), as
a series of Probit models for households. Both procedures used Full-Maximum
Likelihood estimation. Here the observations were limited to those households
with access to formal loans. The Heckman selection model estimation addressed
the sample selection problem that could arise from the use of samples that include
those that did not avail of loans. The other reason is that it will be interesting
to find out what factors matter for household access to loans from formal and
informal sources, and to loans from a formal financial institution, e.g., a bank.
179
World
East Asia and the Pacific
Singapore
Malaysia
Thailand
Indonesia
Philippines
Vietnam
Cambodia
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
Source: Global Findex, 2014
FIGURE 1. Percentage of adults with a formal account
World
Singapore
Malaysia
Thailand
Indonesia
Philippines
Vietnam
Cambodia
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
FIGURE 2. Percentage of adults with savings in a formal financial institution
180
World
Singapore
Malaysia
Thailand
Indonesia
Philippines
Vietnam
Cambodia
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
FIGURE 3. Percentage of adults with loans from a formal financial institution
The second step is to test whether or not financial inclusion helps improve
household income using a two-stage instrumental variable approach to take
care of the possible endogeneity problem between financial inclusion (proxied
by access to formal credit) and household income. Using the number of formal
lending institutions (i.e., universal, thrift, and rural banks) as instrument, the
Two-Stage Least Squares estimation was employed upon satisfaction of the weak
identification and heteroskedasticity tests.
3.1. Heckman selection model
The first step is to test which factors significantly influence a household’s
decision to access financial services (proxied by access to formal credit). To take
into account the endogenous borrowing decisions of households, the Heckman
selection model, which is specified as follows, is employed:
q = xβ+ u (1)
Dloan = I(zδ + v > 0)
(2)
Equation 1 estimates the probability of a household using formal financial
services; that probability is determined by a set of exogenous variables included
in vector x. β is a vector of parameters while u is a normally distributed error term
181
with mean 0 and variance 1 and is assumed to be independent from x. Equation 2
is the first-stage equation—or the so-called “selection” equation—that estimates
the probability of a household availing of a loan (regardless of type), conditional
on a set of exogenous variables included in vector z. Dloan is a dummy variable
taking the value of 1 if the household availed of a loan and 0 otherwise. I(zδ + v
> 0) is an indicator function that takes the value of 1 if the inequality inside the
function holds and 0 otherwise. z contains the same set of variables as in x plus an
instrumental variable, which is an exclusion restriction. δ is a vector of parameters
and v is assumed to have a standard normal distribution and is independent from z.
Essentially, Equation 1 is estimated when q is observed or when Dloan = 1.
Thus, taking the expectation of Equation 1, conditional on z and Dloan = 1, with u
and v being jointly normal with mean 0, the conditional mean can be written as
follows:
E(y | z, Dloan = 1) = xβ + [ρ*λ(zδ)]
(3)
where λ(c) = Φ(c) / Φ(c) is the inverse Mills ratio, which is the ratio of the
standard normal probability density function to the standard normal cumulative
density function, and ρ is the coefficient. In Equation 3, the inverse Mills ratio is
evaluated at zδ.
3.2. Instrumental variable regression model
The second step is to test whether financial inclusion (proxied by access to
credit) is a significant factor affecting household income. To address the potential
endogeneity between access to credit and household income, the two-stage
instrumental variable regression model is estimated, with the number of banks in
a province as the instrument. The specification of the two equations are as follows:
ln(y) = wθ + xβ + ε (4)
w = zα + τ (5)
Equation 4 is the outcome equation that estimates the natural logarithm of
household per capita income, ln(y), conditional on the credit variable, w, and
a vector of x, which includes household head profile, household composition
and location. θ and β are the coefficient of the credit variable and a vector of
coefficients of the other explanatory variables, respectively, while ε is the error
term that is assumed to have a standard normal distribution and is independent
from the explanatory variables. Equation 5 is the credit equation that measures
access to credit. Like in Equation 2, z contains the same set of variables as in x
plus an instrumental variable, which is the number of banks in a province. The
number of banks is assumed to have a significant effect on the credit variable but
182
has no direct effect on the outcome variable or per capita income of a household.
α is a vector of parameters and τ is assumed to have a standard normal distribution
and is independent from z.
The estimations of factors conditioning the likelihood of using financial
services by households and the impact of financial inclusion on household
income used micro-data from the 2013 apis. The survey provides rich data on the
households’ socio-economic characteristics and other aspects of the household
economy,e.g., access to formal and informal loans. A chief limitation is the
absence of information on types of financial products or services used or accessed
by households, e.g., use of savings accounts, remittances services, and the like.
Hence, in this paper, access to formal credit was used as proxy for access to
various types of financial services.
Data used for the estimation come from the apis, a nationwide survey that
collects information on poverty-related indicators, such as those pertaining to
the socioeconomic and living conditions of households and their members, their
access to government programs, and the impact of economic crisis, among others.
This particular sample survey is used for the estimation of the country’s poverty
statistics during years when the Family Income and Expenditure Survey has not
been conducted.4 Although the apis does not collect comprehensive information
on household income and expenditure compared to what the Family Income and
Expenditure Survey does, nevertheless it gathers information on a number of nonincome indicators that are important in poverty monitoring and assessment. The
apis is one of the nationally representative sample surveys in the country that
gathers household-level information on both income and credit availment for the
operation of economic activities of households (psa [2015]).
The 2013 apis has a sample size of 10,864 households, which is relatively
lower than the sample size of the earlier rounds. In addition, this survey was the
first round of the apis that included the income module of the Family Income and
Expenditure Survey. The 2013 apis was conducted by the Philippine Statistics
Authority in July 2013 and was funded by the Department of Budget and
Management (Balamban et al. [2013]).
4. Estimation results
Several regression analyses were done, and the following provided the best
empirical results. Annex A shows the variables used and their definition.
The Family Income and Expenditure Survey is conducted every three years.
4
183
4.1. Heckman selection model estimation5
The first-stage equation looked at the characteristics of households having
access to a loan, regardless of the source of loans. The first-stage regression
results show what matters to households in getting a loan, regardless of the source
of loans (Table 4). The age of the household head matters in accessing loans.
Bigger family size and a high dependency ratio lead household heads to borrow,
while being employed is also a significant factor. The presence or availability of
banks does not necessarily matter in household decisions to borrow at this stage.
The source of loans could be informal lenders, which, as the literature shows, are
mostly the source of loans for poor households.
The second-stage equation looked at the characteristics of households which
had access to formal credit. In the second-stage regression, shifting to a formal
loan source (a bank) household decision to use financial services is positively
and significantly correlated with family size as well as sex, age, marital status,
and educational attainment of the household head (Table 4). The dependency
ratio, measured as the number of dependents below 15 years old, exerts a negative
and significant influence on the decision to use financial services. The bigger
family size means there are more members in the households, and they may not
necessarily be dependents.
Primary and secondary education and more so with tertiary education help
household heads in deciding to access loans from formal financial institutions.
Data tabulations reveal that there are more households with heads having tertiary
level education who availed of formal loans than the household heads with primary
and secondary education. The results seem to indicate that level of education—
more properly, tertiary education—is a significant factor in household decisions
to access formal loans. Similar results by Honohan and King [2012] indicate
that income and education are key demand side determinants of access to formal
banking. On the other hand, level of education does not matter to households in
accessing loans, regardless of source.
Poor households located in the National Capital Region do not necessarily
approach a formal financial institution, such as a bank, to borrow money. In the
first-stage regression, the location of households does not matter in accessing
loans, regardless of the source of loans.
Using the Full-Maximum Likelihood (ml) estimation; at the second stage where access to formal credit is
the dependent variable, the observations are limited to those with access to such credit.
5
184
TABLE 4. Estimated models on access to any type of loans
and on access to formal loans
First-stage equation:
Dependent variable: Access to loans
Regressor
Estimate
Constant
-1.1210
***
Second-stage equation:
Dependent variable: Access to formal credit
Regressor
Estimate
Constant
-0.0922
Natural logarithm of number
of banks
-0.0600
***
Profile of household
head
Sex
(0.0118)
Age
0.0251
0.0142
**
Square of age
-0.0000
***
Married
At least elementary
graduate
0.0636
(0.0433)
At least elementary graduate
0.0257
At least high school
graduate
-0.0140
At least post-secondary/
college graduate
-0.0300
0.2390
***
(0.0437)
0.0552
***
(0.0074)
Dependency ratio
0.2200
***
(0.0737)
Within National Capital Region -0.0410
(0.0498)
0.0805
***
0.1904
***
0.3784
***
Employed
0.0358
(0.0329)
Household size
0.0147
***
(0.0051)
Dependency ratio
-0.1909
***
(0.0501)
Location
Location
***
Household composition
Household Composition
Household size
0.0878
(0.0319)
(0.0488)
Employed
***
(0.0245)
(0.0395)
At least post-secondary/
college graduate
-0.0001
(0.0232)
(0.0381)
At least high school graduate
***
(0.0299)
(0.0001)
Married
0.0144
(0.0000)
(0.0064)
Square of age
***
(0.0041)
(0.0440)
Age
-0.0861
(0.0314)
Profile of household head
Sex
(0.1151)
(0.1681)
Within National Capital
Region
-0.1532
***
(0.0281)
Notes: Figures in parentheses are analytical standard errors;
* significant at 10%
** significant at 5%
*** significant at 1%
4.2. Two-stage instrumental variable estimation
Turning now to the impact of financial inclusion on household income,
Tables 5 and 6 show the results of a two-stage instrumental variable estimation.
In the first-stage regression, a model on access to credit was estimated with the
number of banks in a province as instrument. Among the borrowing households,
the availability of banks and level of education (secondary and tertiary) are
significant positive factors in accessing formal credit, while those with a large
185
family size and high dependency ratio do not access formal credit. Intuitively,
the demand for formal credit tends to increase with the number of formal lending
institutions in the area. While this is true, an increase in the number of banks
does not necessarily stimulate household borrowing. The presence of banks is a
distinct advantage but not a sufficient condition for access to credit. Poorer and
less educated households may decide not to transact with a bank for a variety of
reasons, e.g., lack of information or familiarity with banking procedures, high
transaction cost.
The finding on the education variable is consistent with the findings of
Honohan [2008] and Park and Mercado [2015], who noted that primary education
completion and literacy rates do not have a significant effect on the level
of financial inclusion in developing Asia. It seems that it takes more than just
primary education for households to be able to use financial services.
A high dependency ratio acts as a barrier to financial inclusion. This echoes the
finding of Park and Mercado [2015] that higher age dependency ratio significantly
reduces financial inclusion. This is because a larger segment of the population
is either too young or above the retirement age, which impedes their access to
financial services as they do not earn income.
Although not significant, the signs of the coefficients indicate that households
with very young and very old heads either have higher probability of not availing
of loans, regardless of source, or of accessing formal credit. The seemingly
counterintuitive results on the employment variable is explained as follows. An
inspection of the data showed that as households move from a situation where all
the loan sources are informal lenders (loan2=0, Annex A) to where households do
not borrow at all (loan2=2, Annex A), the proportion of households with employed
heads increases slightly from 88.6 percent to 89.9 percent and then drops
significantly to 79.4 percent. On the other hand, the proportion of households
with heads who were not employed decreases slightly from 11.4 percent to 10.1
percent and then increases substantially to 20.6 percent.
In the second-stage regression, Table 6 shows the results on the impact of
financial inclusion on household income. Financial inclusion (here proxied by
access to credit) has a positive and significant impact on household income.
Higher-income households may not necessarily be availing themselves of formal
loans because they most likely have demand for other financial services, such as
savings with a formal institution, payment services with a bank, insurance, and
others. The financial inclusion survey of the bsp revealed that the proportion of
adults who keep their money in banks is significantly higher in classes A, B, and
C (around 71 percent) than in class D (32.7 percent) and class E (17.2 percent).
Among those who borrow, higher-income households tend to have higher access
to formal credit. Education at any level—whether primary, secondary, or tertiary—
similarly has a positive and significant influence, and the magnitude of the impact
increases with educational level. Apparently, the households with more educated
heads tend to have higher-paying jobs relative to those with less educated heads.
186
TABLE 5. Estimated model on access to credit (first-stage regression)
First-stage equation:
Dependent variable: Access to credit
Regressor
Constant
Estimate
1.6328
***
(0.0882)
Natural logarithm of number of banks
0.0284
***
(0.0068)
Sex
-0.0323
Age
-0.0023
(0.0222)
(0.0032)
Square of age
0.0000
(0.0000)
Married
-0.0135
(0.0224)
0.0125
(0.0223)
0.0668
***
(0.0228)
At least post-secondary/college graduate
0.1197
***
(0.0255)
Employed
-0.0981
***
(0.0208)
Household Composition
Household size
-0.0240
Dependency ratio
-0.1714
***
(0.0043)
***
(0.0415)
Location
Within National Capital Region
-0.0234
(0.0269)
Notes: Figures in parentheses are robust standard errors;
* significant at 10% ** significant at 5% *** significant at 1%
The location variable, the National Capital Region, which can be broadly
interpreted as an urban location variable, is also a significant and positive
determinant of household income. Jobs seem to be more available in dense urban
settings than in the rural areas.
187
The age of the household head seems to matter also on the level of household
income. The sign of the coefficient of the squared age variable implies that
household income increases with the age of the household head up to a certain
point and then decreases thereafter. Very young household heads are at the start
of their career, and it is reasonable to assume that they are earning relatively less
compared to the older ones. On the other hand, very old household heads are
no longer working and earning. Both of these cases result in a lower per capita
income of a household.
Meanwhile, a higher dependency ratio has a significant negative correlation
with household income for obvious reasons. Households with married and/or
employed heads have greater chances of improving household incomes.
TABLE 6. Estimated model on the impact on household income
(second-stage regression)
Second-stage equation:
Dependent variable: Natural logarithm of per capita income
Regressor
Constant
Access to credit
Sex
Age
Square of age
Married
At least post-secondary/college graduate
Employed
Household composition
Household size
Dependency ratio
Location
Within National Capital Region
Estimate
4.4746
(1.3101)
3.0266
(0.7345)
-0.0413
(0.0724)
0.0179
(0.0100)
-0.0002
(0.0001)
0.1209
(0.0691)
0.2004
(0.0701)
0.4113
(0.0912)
0.9286
(0.1246)
0.1919
(0.0994)
-0.0089
(0.0223)
-0.3362
(0.1856)
0.3119
(0.0729)
***
***
*
**
*
***
***
***
*
*
***
Notes: Figures in parentheses are robust standard errors;
* significant at 10% ** significant at 5% *** significant at 1%
188
Current policy discussions hold that financial inclusion is important for
inclusive growth and poverty reduction. Theoretical and empirical studies have
looked at the beneficial impacts of financial inclusion at the macroeconomic,
household, and firm level. The present study paid special attention to households,
especially poor households, to get a better understanding of what factors drive
financial inclusion at this level.
The empirical findings showed what factors significantly influence household
decision to access financial services. The findings clearly showed robust and
significant correlation between household decision to use financial services
(proxied by access to formal credit, due to data limitations) on the one hand,
and the age of the household head, marital status, family size, and educational
attainment of the household head, on the other. Poor households with more
dependents (those below 15 years of age) tend not to use financial services. This
probably has to do with weaker incomes or the financial capacity of households
with too may mouths to feed to repay loans. There is a need to underscore the
importance of a higher level of education (secondary and tertiary) as a positive
and significant factor in household decisions to access and use formal financial
services. On the other hand, households may be able to access informal loans
regardless of the level of education of the household head.
The empirical findings from a two-stage instrumental variable estimation
supports the hypothesis that financial inclusion improves household income.
Financial inclusion (here proxied by access to formal credit) has a positive and
significant impact on household income. The empirical findings of the paper lead
to certain policy implications.
First, expanding access to and use of financial services by low-income
households/individuals may have a positive effect on household/individual
welfare. Access and use of financial services are an important tool of households
for consumption smoothing, making productive investments, and coping with
catastrophic risks. Based on evidence of a strong correlation between financial
inclusion and higher household income, there is a need for policies and
interventions that reduce barriers to financial inclusion.
Second, a key measure to address the financial exclusion of poor households
is financial education. Education at any level—primary, secondary, or tertiary—
builds human capacity and has a positive and significant influence on household
income. This is an important finding in view of the fondness of some politicians
to use credit subsidies to address the problem of lack of access and utilization
of financial services by poor households, small farmers, and similar economic
agents, e.g., microenterprises. It is good to pay attention as well to non-financial
189
factors, such as education, to equip such small economic agents with the capacity
to access and use financial services.
Philippine Institute for Development Studies (PIDS)
The author acknowledges the invaluable research assistance provided by Christian Mina. The
views and opinions in this paper are solely the author’s and do not in any way reflect those of
the Philippine Institute for Development Studies.
References
Alliance for Financial Inclusion (afi) [2010] The afi survey on financial
inclusion policy in developing countries: preliminary findings. Kuala Lumpur.
Balamban, B.B., M.B. Addawe, and M.G.G. Darunday [2013] “Official poverty
statistics and poverty reduction programs of the Philippines”, http://adbi.adb.
org/files/2013.05.08.cpp.sess4.3.poverty.reduction.philippines.pdf. Accessed
21 October 2015.
Bangko Sentral ng Pilipinas (bsp) [2014] State of financial inclusion in the
Philippines, 2013 data updates. Manila: Bangko Sentral ng Pilipinas.
Bangko Sentral ng Pilipinas (bsp) [2015] National strategy for financial inclusion.
Manila: Bangko Sentral ng Pilipinas.
Beck, T., A. Demirguc-Kunt, and R. Levine [2007] “Finance, inequality, and the
poor”, Journal of Economic Growth 12: 27-49.
Cámara, N. and D. Tuesta [2015] “Factors that matter for financial inclusion:
evidence from Peru”, Aestimatio the ied International Journal of Finance, 10:
8-29.
Clarke, G.R.G., L.C. Xu, and H. Zou [2006] “Finance and inequality: what do the
data tell us?”, Southern Economic Journal 72(3): 578-596.
Consultative Group to Assist the Poor (cgap) [2011] “Global standard-setting
bodies and financial inclusion for the poor: toward proportionate standards and
guidance”, a white paper on behalf of the Global Partnership for Financial
Inclusion.
Economic and Social Commission for Asia and the Pacific (escap) [2014]
“Sustainable development financing: perspectives from Asia and the Pacific”,
paper prepared for the Regional Outreach of the Intergovernmental Committee
of Experts on Sustainable Development Financing for the Asia- Pacific Region,
10-11 June 2014.
Honohan, P. [2008] “Cross-country variation in household access to financial
services”, Journal of Banking and Finance 32: 2493–2500.
190
Honohan, P. and M. King [2012] “Cause and effect of financial access: crosscountry evidence from the Finscope Surveys”, iiis Discussion Paper No. 399,
Trinity College, Dublin, May.
Kumar, K., C. McKay, and S. Rotman [2010] “Microfinance and mobile banking:
the story so far”, Focus Note 62. Washington, d.c.: cgap.
Kuri, P.K. and A. Laha [2011] “Financial inclusion and human development in
India: an inter-state analysis”, Indian Journal of Human Development 5(1):
61-77.
Llanto, G.M. [2015] “Financial inclusion, education, and regulation in the
Philippines”, adbi Working Paper 541. Tokyo: Asian Development Bank
Institute.
Park, C.Y. and R. Mercado, Jr. [2015] “Financial inclusion, poverty and income
inequality in developing Asia”, adb Economics Working Paper Series No.
426 January.
Philippine Statistics Authority (psa) [2015] “Technical notes on the Annual
Poverty Indicators Survey”, https://psa.gov.ph/content/technical-notes-annualpoverty-indicator-survey-apis. Accessed 21 October 2015.
Pickens, M. [2009] Window on the unbanked: mobile money in the Philippines.
World Bank: Washington d.c.
Verbeek, M. [2008] A guide to modern econometrics. 3rd edition. West Sussex,
England: John Wiley and Sons Ltd.
Wooldridge, J.M. [2002] Econometric analysis of cross section and panel data.
Cambridge, Massachusetts: MIT Press.
World Bank (WB) [2014] Financial inclusion: global financial development
report. Washington, d.c.
191
ANNEX 1. Definition of variables for Heckman selection odel
ln_pcinc = natural logarithm of per capita income of the household
avail_loan = 1 if the household availed of a loan during the past 6 months,
regardless of source; 0, otherwise
formal_loan = 1 if the household availed of a loan from during the past 6 months
and at least one of the sources is a formal lending institution; 0, otherwise
sex = 1 if household head is male; 0 if female
age = age of household head
agesq = squared age of household head
educ1 = 1 if head attained at most elementary undergraduate (base category)
educ2 = 1 if head attained at least elementary graduate and at most high school
undergraduate
educ3 = 1 if head attained at least post-secondary or college graduate
married = 1 if household head is married; 0, otherwise
employed = 1 if household head is employed; 0, otherwise
fsize = number of members in the household
dep_ratio = proportion of household members aged below 15
ncr = 1 if the household is located within the National Capital Region; 0 if located
outside ncr. In the absence of an urbanity variable, this variable was used to
represent the location variable.
imr = inverse Mills ratio, or ratio of the standard normal probability density
function to the standard normal cumulative distribution function of the
predicted value of avail_loan; addition of this in the model as a regressor
addresses sample selection bias
ln_banks = natural logarithm of the number of banks within a province
(instrument)
_cons = constant term
The following variables were used in the two stage instrumental variable
estimation, in addition to those listed above.
ln_pcinc = natural logarithm of per capita income of the household
loan2 = 2 if the household did not avail of a loan during the past 6 months; 1
if the household availed of a loan and at least one of the sources is a formal
lending institution; 0 if the household availed of a loan and all of the sources
are informal lenders6
The definition of the loan variable was based on the following observations from the data set:
the mean per capita income of households that did not avail of a loan was P32,696.79;
the mean per capita income of households that availed of a loan and at least one of the loans was
sourced from a formal lending institution was P30,009.64; and
the mean per capita income of households that availed of a loan and all loans were sourced from
informal lender(s) was P18,318.37.
6
PRE
The ASEAN business cycle and China’s slowdown
The economies in the Association of Southeast Asian Nations
(asean) have been integrated through increasing trade relations
intensified by tariff reductions and increasing openness to
foreign direct investment. Rising volume of networks trade has
deepened interconnectedness between asean and China, a
recent growth locomotive in the world. asean business cycle is
shaped by volatile China’s trade volume. As China’s expansion
slows down, adverse consequences on asean economies have
become more pronounced. The extent of the damage depends on
each member of asean’s trade exposure and China dependency.
This paper identifies the most vulnerable asean economies
to China’s business cycle. The slowing down of the Chinese
economy would undoubtedly result in a decline in long-term
growth of some asean economies, unless appropriate policy
responses can be implemented.
JEL classification: E32, F43
Keywords: ASEAN Growth, Business Cycles, China’s Slowdown
1. Introduction
Hill [2014] raises an important question on whether there is a Southeast Asia
development model. The 10 countries of the Association of Southeast Asian
Nations (asean) have become more integrated over time. Hill argues that large
socio-economic and institutional disparities among asean members imply that
the asean development model does not exist yet. Nevertheless, convergence in
growth rates and social development are taking place. In time, through spillover
and demonstration effects from asean’s leaders to its followers, we might find
such coherent regional development strategies. According to Hill and GochocoBautista [2013], there is policy consensus on the desirability of conservative
fiscal policy, except during the crisis, and on the independence of central banks.
However, rethinking of an export-led growth strategy should be for those countries
with highly open and trade-dependent countries.
193
China, the world’s second-largest economy, has experienced a drop in economic
growth rate from 10 percent in 2010-2011 to 7 percent in the first two quarters
of 2015. The economic slowdown and the depreciation of the yuan would have
significant negative consequences on some asean economies. Through vertical
integration, Foreign Direct Investment (fdi) has made asean vulnerable to
fluctuations in world trade volume. The asean business cycles are synchronized
through regional trade linkages. On the financial flows, synchronization of
asean-5 stock market has been well documented. Teng et al. [2013] report that
stock markets in traditional asean countries are well aligned with economic
activities in developed economies rather than those in emerging economies. Also,
they do not react to external shocks originating in other asean markets. Each
stock market in asean-5 responds differently in terms of direction and degree
towards changing world economic conditions. In August 2015, the stock market
crash in China led to fears of a market meltdown.1 The “China factor” is significant
in real and financial sectors to the rest of the world, from commodity prices,
exchange rates, and stock prices.
Because asean economies have some degree of heterogeneity, some asean
countries would suffer from the slowing growth of China more than others.
Differences in the level of trade integration, export commodity concentration, and
export market dependency would lead to dissimilar impact of China’s new normal
growth path. Furthermore, the quality of asean institutions differs greatly from
authoritarian to democratic regimes. Hill and Jongwanich [2014] note that the
determinants of outward fdi from China and other emerging East Asian economies
are consistent with standard economic theory. In addition, the desire for natural
resource security and exceptionally high domestic savings rates are particularly
important factors for fdi in the region. fdi from China is, therefore, another
important transmission channel of shocks from China’s economic fluctuations to
asean economies, which have abundant natural resources.
The quality of institutions can affect the attractiveness of fdi in asean [Masron
and Nor 2013]. The heterogeneity among asean economies implies differences
in the effectiveness and efficiency of macroeconomic policy in response to China’s
economic slowdown. China rising has caused a shift in global trade patterns, with
China dominating Western markets at the expense of asean countries [Napoli
2014]. Nevertheless, China’s dominance does not appear to have had a significant
negative effect on growth rates for asean gdp, exports, or fdi stocks. While
China is crowding asean out of developed markets, increased Chinese demand
for asean imports has more than offset this effect. Now the impact of China’s
slowdown has become more pronounced in export sectors in the United States,
Europe, and Japan. Since asean’s export sectors have market exposures to those
China’s Shanghai Composite index rose by 150 percent (year-on-year) in June 2015. The sharp fall in
August 2015 represented a 40 percent fall from the June peak.
1
194
Bhanupong: The ASEAN business cycle and China’s slowdown
markets, although they do not directly rely on China’s imports, eventually they
will be adversely affected by the slowdown in China.
This paper examines how the slowdown in China’s economic growth affects
asean economies. If China’s rebalancing policy means increasing consumption
and reducing investment expenditure, will the link between China and the asean
business cycle become weaker? Which asean economies will be most affected
by China’s new normal growth?
The rest of this paper is organized as follows. Section 2 discusses key aspects
of the China factor that exerts impact on the asean business cycle. Section 3
examines the implication of exchange rate realignment. Section 4 analyzes the
impact of China’s slowdown on asean economies. Section 5 presents simulation
results from a vector autoregressive model, which contains China’s imports growth,
asean business cycle movements, and asean members’ economic activity. The
last section provides concluding remarks.
2. The China factor and the ASEAN business cycle
asean economies experienced similar shocks from the Asian financial crisis
in the period 1997-1998 to the global financial crisis in the period 2008-2009.
The spectacular falls in output were followed by V-shaped recoveries. The reason
behind this co-movement of output can be traced back to similar patterns of
exports and investment cycles in asean economies. Athukorala and Hill [2010]
point out that dependence on fragmentation-based international specialization is
proportionately higher in asean than in North America and Europe. The rapid
integration of China into regional production networks does not crowd out asean
members’ opportunities for international specialization.
Not only are their output growth paths related, inflation, interest rates, and stock
market performance tend to move together in tandem. Since the Asian financial
crisis, the saving-investment gaps in asean economies have widened as the
current account deficit in the pre-Asian financial crisis turned around into surplus.
The loss of China’s competitiveness is shown by the appreciation of the real
effective exchange rate, which weighs down on China’s exports. The export-led
growth policy of China has been characterized by an artificially weak exchange
rate and subsidized credit by public-own banking system. China has welcomed
fdi inflows in order to increase the strength of its industrial development, which
has contributed to enormous capacity. When the world could no longer support
the growth of China’s exports, China’s exports collapsed spectacularly after the
V-shaped recovery in 2010 (Figure 1). Consequently, imports also declined as a
result of network trade. The global financial crisis caused the sharp export decline
in 2009, but the sharp recovery in 2010 was not sustained. Since then both exports
and imports started falling rapidly. Overcapacity among exporting firms would
eventually lead to non-performing loans of banks that provided subsidized credit
to the Chinese export sector.
195
Source: Federal Reserve Bank economic data
FIGURE 1. Growth rate of China’ s merchandise trade
In July 2015, China devalued the yuan after its exports declined by almost 9
percent from the same time in 2014. The sharp fall was due to a strengthening US
dollar and rising wages. China’s devaluation was an attempt to compensate for the
strength of the yuan as it has been pegged to the dollar. Total trade declined by 8
percent for the first ten months of 2015. This dramatic economic slowdown has
led to a substantial export shortfall in asean countries.
China’s major sources of imports are Japan and the United States. Both
Malaysia and Thailand are also important sources of China’s imports, ranking
among the top 10 countries. To gauge the impact of China’s imports on asean
economies (Figure 2), we compare the movement of business cycle in asean
economies (right axis) with China’s imports (left axis). The asean business cycle
is obtained from the first component of principal components of asean’s gdp
growth rates between 1990 and 2014. There is a striking similarity between the
upturns and downturns of the two magnitudes. A long-run relationship between the
two series is confirmed by the Johansen cointegration test. The Granger causality
test indicates that the direction runs from asean business cycle to China’s import
growth. Booming economic activity in asean can predict the rising demand
for imports from China. On the contrary, there is no evidence of cointegration
between asean business cycle and the United States’ imports. Indeed, asean
business cycle is synchronized with China’s economic activity, i.e., trade and
output growth. After the establishment of asean Economic Community in
2016, with diversity in population and economic sizes, income levels, and natural
resources, it is expected that synchronization of economic activity among asean
members would be more intensified through increasing trade intensity within the
asean Economic Community and with China.
196
China’s imports (50%)
ASEAN output cycle (right axis)
Sources: Asian Development Bank data and the author’s calculation
FIGURE 2. The ASEAN business cycle and China’s imports
Another explanation of the business cycle synchronization between
asean and China is related to the performance of the Japanese economy. The
quantitative easing policy of the Bank of Japan, through its record us$665 billion
annual asset-buying scheme, has had only limited impact on output on growth
and inflation. Japan has suffered as exports to key trading partner China slumped.
Japanese consumption spending is also sluggish as a result of an increase in sales
tax in 2014. The deterioration in the global economic outlook, including the
developments in China, would delay Japanese companies in expanding business
investment and raising wages. If the Japanese economy is still in recession and
deflationary mode, the demand for Japanese exports from China will also decline,
implying a reduction in imported parts and components from asean countries.
The synchronized business cycle is related to credit cycles and exchange rates.
Mohan and Nandwa [2009] find evidence of cointegration among the asean
interest rates, and the direction of these interest rates is affected by China’s. This
evidence points to the linkage between asean-5 and China. The synchronization
can also be explained by the exchange rate channel. asean countries tend to avoid
currency appreciation when the dollar trended downward. China also tries to peg
the yuan to the dollar. Thus trade between asean and China can be conducted
within an environment of stable exchange rates. Tang [2014] shows that, because
of the region’s production networks, increasing intraregional exchange rate
volatility among Asian economies leads to a decline in intraregional trade, in
particular among asean economies.2 Ong and Habibullah [2012] find evidence
of a continuous macroeconomic interdependence between asean-5 and China,
which would lead to a successful asean-China economic cooperation.
Intraregional trade within asean amounted to only 24 percent, compared to 69 percent among the
European Community.
2
197
The macroeconomic interdependence between asean and China can be
demonstrated by co-movements of price levels in the two regions. Employing the
principal component of asean inflation rates in 10 member countries, we can
obtain the proxy of inflation cycles, as shown in Figure 3. When compared with
China’s, it is obvious that price stability between China and asean is related.
The long-run relationship between China and asean exist, with the causal
relationship running from China’s inflation to asean’s.3 The global food crisis
in 2008 led China to ban exports of rice and sent waves of panic to Asia. The
inflation rate temporary skyrocketed because of food inflation, followed by export
collapse caused by the global financial crisis.
Principal
component of
ASEAN inflation
China inflation
FIGURE 3. Inflation synchronization: China and ASEAN
Source: Asian Development Bank data and the author’s calculation
Tham and Kam [2014] employ a gravity model that captures the effects of
trade in parts and components as well as final manufactured goods from asean
to China. Devadason [2009] also finds evidence that China’s integration in the
asean region increases the size of the asean export market, rather than reducing
asean’s export expansion. His result is in line with the fact that, although China
has become an important export destination and a source of imports for individual
asean 5 countries, there is no reduction in intra-asean 5 trade. From the
literature, we can conclude that trade seems to be the most important channel
explaining the synchronization between China and asean business cycle.
The long-run relationship is confirmed by Johansen cointegration test. The hypothesis that China’s
inflation rate does not Granger- cause the inflation cycle in asean is rejected at 0.1 percent statistically
significant level.
3
198
3. Implications on exchange rates
The quantitative easing monetary policy of the Fed led to capital inflows
to emerging markets, as investors were looking for higher rates of return from
portfolio investment. After 2009, the Chinese yuan, the Thai baht, the Philippine
peso, and the Malaysian ringgit considerably appreciated against the us dollar. In
terms of international competitiveness, the real effective exchange rates of these
currencies had appreciated substantially since 2010. The yuan strengthened by
30 percent, the peso by 20 percent, and the baht by 10 percent at the end of 2014
(Figure 4). The loss of international competitiveness has compounded economic
problems of export-led growth asean economies.
Source: Federal Reserve Bank economic data
FIGURE 4. Real effective exchange rates
Hooy et al. [2015] find that the yuan real exchange rate has a significant
positive impact on asean’s total exports to China. Parts and components exports
are sensitive to the yuan depreciation, because of the recent production relocations
of multinational corporations from the asean region to China and Vietnam.
The huge capital inflows after the quantitative easing operations prompted the
asean central banks to buy dollars and to accumulate international reserves so as
to prevent currency appreciations. However, the trend of appreciation of asean
currencies turned to a rapid depreciation in 2015 (Figure 5). By 2015, the US
dollar gained strength as the us economy rebounded and the fear of the Fed’s
interest rate hike led to fear of capital outflows from the asean region.
199
Source: The Economist
FIGURE 5. Asian currency depreciation against the US dollar,
October 2015 (% year-on-year)
By November 2015, the Indonesian central bank’s reserves fell by us$1
billion to us$100.7 billion. The decline was due to rising costs to service the
government’s foreign debt and the use of reserves to stabilize the Indonesian rupiah,
which depreciated by 10.7 percent (year-on-year) in October 2015 (Figure 5).
The large depreciation corresponds to a high inflation rate of about 6.4 percent
during the same corresponding period. The gdp growth rates in Indonesia and
Malaysia slowed down to around 4.8 percent in 2015; while Indonesia experienced
current account deficit, Malaysia still maintained current account surplus. The
steep depreciation of the ringgit must have been related to capital outflows and
political risks.
Vietnam is another country that faces inflation at a higher rate than other asean
members. The Vietnamese dong has depreciated gradually in line with the inflation
level. However, Vietnam experienced strong economic growth in 2015, helped by
robust export growth. The dong depreciated by 6.5 percent in 2015. The gain in
Vietnam’s exports to the United States was at the expense of Thailand, which is
losing world market share in electrical and electronic equipment, apparel, and
garments. This is partly due to rising wages and the appreciation of the baht from
2010 to 2014. Thailand also experienced currency appreciation, although the Bank
of Thailand does not intervene in the foreign exchange markets as much as it used
to. As a result, the baht depreciated by 9.2 percent (year-on-year) in October 2015.
Vietnam imports input materials mostly from China to serve its local
production. Therefore, as the Chinese yuan has depreciated, Vietnam is able to
buy materials at lower prices, thus enabling it to cut production costs and raise
Vietnam’s competitiveness. Furthermore, the price decreases in the world market,
200
caused by the weaker demand from China, will also help Vietnamese businesses
cut production costs.
Vietnam still has cheap labor, which can be a main factor in attracting fdi
away from China, which is experiencing rising wages and economic slowdown.
Another factor that attracts fdi to Vietnam is the benefit Vietnam would obtain
from joining the Trans Pacific Partnership Agreement (TPP). Vietnam is the
country that would benefit the most from the Trans Pacific Partnership agreement
as it has the lowest wages among the 12 members.
Since China surprised world markets by devaluating the yuan by 2 percent
in August 2015, net capital outflows have reached us$200 billion. The People’s
Bank of China is reported to have spent us$229 billion in foreign exchange
intervention to prop up the yuan in the third quarter of 2015. There has been
speculation that the Fed interest rate hike is imminent as the us economy has
been approaching the full employment level in the last quarter of 2015. While
the economic recovery in the Eurozone is still protracted and fragile, due to a
decline from Chinese imports, the European Central Bank would still maintain
easy monetary policy. As a result, the gap between the interest rates in the United
States and in the Eurozone will widen in 2016. The dollar may continue to
appreciate. Because of the yuan peg with the dollar, there would be speculation
of another round of yuan devaluation. In the future, we would observe the similar
pattern of Asian currency depreciation as shown in Figure 5.
4. Trade exposure and dependency of China
Trade exposure to external shocks can be measured by the value of total trade
volume relative to gdp. In general, the ratio can be thought of as a measure of the
degree of openness, reflecting trade liberalization over time. The increasing trend
of trade openness also implies increasing risk exposure to external shocks, such
as terms of trade and volume of trade shocks. Small countries tend to have higher
trade volume relative to gdp, while large countries with large domestic markets
would tend to have lower trade exposure to trade shocks from the rest of the
world. The high trade exposure ratio bodes well for degree of trade integration,
which can serve as an economic growth driver.
Hill and Menon [2011] point out the high degree of vulnerability of the
Cambodian economy, which arises from a narrow economic base, a pre-crisis
asset price boom, a fragile financial system, and limited defensive economic
policy instruments. After extremely rapid economic growth during 2000-2007, it
experienced a sharp growth collapse in 2008-2009. The economy has begun to
rebound since early 2010, and the crisis episode has provided the government
with an opportunity to conduct policy reform by diversifying the economy and
creating the preconditions for de-dollarization. From Figure 6, between 2010 and
2014, the average ratio of the volume of Cambodia’s trade to gdp almost reached
100 percent. It is evident that, despite having the highest dollarization level in
201
Si
ng
ap
or
e
m
a
si
ay
et
na
Vi
al
M
ai
la
nd
a
Th
bo
di
am
al
am
ss
ru
Da
Br
un
ei
C
PD
R
s
La
pi
o
ne
na
hi
ilip
Ph
C
si
ne
do
In
M
ya
nm
ar
a
the region, dollarization has contributed to a stable price level and favorable
environment to trade.
Source: Asian Development Bank
FIGURE 6. Average degree of trade exposure, 2010-2014
Since China is a major trading partner of Myanmar, and bilateral trade with
China accounts for 50 percent of total trade, China’s economic growth and
imports slowdown would adversely affect Myanmar’s exports. Despite low trade
exposure (21.4 percent), the falling price of rice and nationwide floods would
compound Myanmar’s economic problems. The gdp growth of Myanmar is
highly correlated with China’s total import growth rate (Figure 7), but not as high
as Cambodia’s.
According to Bird and Hill [2010], the past two decades of reform in
Lao People’s Democratic Republic (pdr) have been largely successful, with
accelerating growth despite unfavorable initial conditions such as the country’s
being landlocked and having weak institutions and ill-defined property rights.
Neighborhood effects have obviously been supportive in Lao pdr’s case, but their
importance should not be overstated. Lao pdr’s gdp growth rate is independent
of China’s economic activity (Figure 7). Since the major trading partner of Lao
pdr is Thailand, its output growth is related more to Thailand’s gdp growth
through export channel.
The Philippine growth rate in 2015 also slowed down compared to the high
rate in 2014. Similar to Indonesia and Lao pdr, the Philippine economy would
not severely affected by China’s slowdown as much as countries which have
high overall trade exposure and high dependency on China. The slump in export
demand in the Philippines cannot be compensated by currency depreciation of 4.5
percent in 2015. The depreciation was caused by expectations of capital outflows
in anticipation of the Fed’s rate hike.
o
La
si
a
s
do
ne
ne
pi
ilip
In
m
Ph
nd
et
na
Vi
ila
si
ay
al
Th
a
ei
M
M
Br
un
or
e
ng
ap
bo
di
Si
am
C
ar
nm
ya
a
a
202
Source: Author’s calculations based on Asian Development Bank data
FIGURE 7. GDP growth correlation with China import growth
TABLE 1. China dependency and ASEAN trade exposure
China dependency
High trade exposure
(% Trade/GDP)
Low trade
exposure
(% Trade/GDP)
High
(Growth correlation:
above 0.35)
Brunei (0.42)
Cambodia (0.51)
Malaysia (0.39)
Singapore (0.47)
Thailand (0.39)
Myanmar (0.44)
Low
(Growth correlation:
below 0.35)
Vietnam (0.3)
Indonesia (0.19)
Lao PDR (0.01)
Philippines (0.2)
Sources: Figures 6 and 7
Malaysia is a country with highly open economy and with export-growth
driven strategy. Hill et al. [2012] attribute Malaysia’s significant slowdown in
investment and growth since the late 1990s to some specific institutional and
political problems, rather than the middle-income trap. Thus major policy reforms
are required to meet challenges to growth acceleration. In 2014, Malaysia’s output
growth rate is highly related to China’s economic growth. The country has high
trade exposure and high dependency on China. Malaysia has suffered from losing
market shares in declining sectors: electrical and electronic equipment, vegetable
oil, and rubber. Indonesia lost its market shares in declining sectors such as rubber,
mineral fuels, and palm oil. In the same development, Thailand’s market shares
in the growing sector of rice and seafood have deteriorated. Furthermore, to make
matters worse, Thailand also lost the market share in a declining sector: rubber.
203
The only export sector that seems to be gaining market share is the automobile
and parts sector. Due to rising wages, Vietnam has been replacing Thailand’s
position as a major exporter of footwear, garment, and seafood. Viewed in this
light, asean countries have certain degree of complementarity and competition
in the area of labor-intensive industry.
The asean economies that depend on commodity exports suffer from
the declines in export demand and commodity export prices. The extent of the
damage is related to the reliance and concentration of commodity exports. Palm
oil, rubber, and rice are primary commodity exports of many asean countries.
They have experienced a declining trend since 2010 (Figure 8). The unfavorable
terms of trade would continue as long as the crude oil price still remains below
us$80 per barrel.4
As Asia’s only major net oil exporter and the world’s second-biggest producer
of palm oil, Malaysia’s economic growth has been negatively affected by the weak
oil price. The Malaysian government encountered a decline in fiscal revenues.
The government derives about 22 percent from energy related business. The
political scandal over funds transferred also led to the weakening of the ringgit. A
collapse in commodity prices, caused by China’s slowdown in export demand for
raw materials, has affected the asean community.
Palm oil
Rubber
Rice
Source: Mundi Index
FIGURE 8. Declining commodity prices
Crude oil prices are positively related to the price of natural rubber, which competes directly with
synthetic rubber, a byproduct of the oil refinery industry.
4
204
In summary, China’s slowdown has reduced the fiscal space of asean
countries, thereby impairing the governments’ ability to use fiscal policy to
counteract the downturn of economic cycles. When the Fed starts normalizing
monetary policy, an increase in interest rates would raise household debts in
emerging economies. Rapid increases in household debts to gdp have made it
difficult for asean monetary authorities to implement effective monetary policies
to stimulate the economy after export collapses. Poverty and indebtedness have
been a major problem for farmers since the collapse of primary commodity prices.
As long as commodity prices remain weak, the inclusive development goals of
poverty eradication and improved income distribution will be a challenge. China’s
slowdown has a far-reaching impact on inclusive growth.
5. Impacts of China’s slowdown
We have established that there is a strong causal relationship between China
and asean economic activity. In this section, we examined the impacts of the
fluctuations on business cycles in asean and China. An unrestricted vector
autoregressive model is constructed for this purpose by utilizing the annual data
on growth rates in asean economies. The data from 1990 to 2014 are obtained
from the Asian Development Bank. As discussed earlier, the first principal
component of asean growth rate is employed as a proxy for the economic
activity of asean economies as a whole (Figure 2). By utilizing two-period lags,
impulse response functions from shocks in asean business cycle and China’s
import growth are obtained. The accumulated responses of asean output growth
rates are reported. In the first group, we select high-trade exposure countries
(Table 1), namely Malaysia, Thailand, Cambodia, and Vietnam. The results
reported in Figure 8 confirm the earlier conjecture that countries with high trade
openness are likely to benefit more from China’s rapid expansion than countries
with lower degrees of trade exposure. Thailand’s growth rates, followed by
Cambodia, respond most strongly among the four countries in the first simulation
group. On the other hand, during China’s slowdown, Thailand and Cambodia
would bear the most burdens of adjustments in terms of growth slowdown. Except
for Malaysia, the impact of China’s economic activity is more prominent than
economic activity within asean community (as captured by impulse responses
to the asean business cycle). For Malaysia, the impact of asean expansion
is more pronounced than the China factor. This is because the trade intensity
between Malaysia and other asean countries, namely Singapore, is higher than
Malaysia’s trade intensity with China. In the case of Thailand, Cambodia, and
Vietnam, the accumulated responses of growth of these economies from asean
economic activity die off within five years, while China’s impacts on growth
rates in high-trade exposure countries are sustained more than a decade. These
findings underline the importance of China factor in contributing to long-term
output growth in Thailand, Cambodia, and Vietnam since China’s accession to the
205
World Trade Organization. On the other hand, China’s new normal growth pattern
would have a detrimental impact on the long-term growth of countries with high
international trade exposure and high dependency on Chinese economy. Hill and
Menon [2013] observe that despite Cambodia’s weak institutions and a legacy
of history and small size that limit government’s policy space, Cambodia has
achieved high economic growth. Through large public and private capital inflows,
economic openness, and reasonably prudent macroeconomic management, the
Cambodian economy grows rapidly in a dynamic, integrating neighborhood of
the asean region. The empirical finding in this paper underlines the importance
of the China factor in generating high growth for low-income asean countries
since they started engaging in trade liberalization. With the new growth target set
for 7 percent in the China’s new development plan for the next five years, heavy
reliance on Chinese expansion can be detrimental to their long-term growth, as
they cannot depend on China’s strength.
For countries with low trade exposure to external shocks, the impact of China
slowdown will be minimal. Figure 10 illustrates that, except for Myanmar, the
China factor is less important than the asean factor; economic activity in the
asean region produces stronger growth impact. Indonesia, the Philippines, and
Lao pdr benefit less from China rising than the expansion of asean economies.
Lao pdr’s growth rate is hardly related to China’s growth. Distance matters
in trading activity. Lao pdr trades more with Thailand than China. Although
Myanmar has low international trade exposure, its gdp growth is highly
associated with China’s. Figure 10 demonstrates that Myanmar’s growth increases
substantially with rising economic activity in China, whereas it does not benefit
much from increasing asean economic activity. This conclusion may change
after the investment and trade impacts of the asean Economic Community
become more pronounced in the next decade.
Indonesia, which has a relatively low trade exposure, due to its large size, has
less influence from the China factor. Basri and Hill [2011] analyze Indonesian
growth dynamics, in particular the V-shaped recoveries after the Asian financial
crisis and global financial crisis. The years 1997-1998 were regarded as a
watershed in the country’s economic history and political economy. The growth
of Indonesian economy has never matched that of high-growth East Asian
Economies. In Figure 10, Indonesia and the Philippines have not benefited much
from China rising, although the two countries’ growth rates respond vigorously in
line with expansion of asean economic activity.5
See Hill [2015] for a discussion on political economy of policy reforms in Indonesia and the Philippines.
5
206
China’s new normal growth implies that the expansion of asean economies
would slow down. China’s slowdown has reverberated around the world. Japan
was also in recession again in the third quarter of 2015. This is Japan’s fifth
recession in seven years. Some asean economies would not be affected, unless
public investment can be implemented effectively to induce growth via enhancing
effect of public infrastructure, which can lead to crowding in effect on private
and consumption expenditures. asean countries are seemingly homogenous,
yet each country has heterogeneity in trade exposure, export markets, and export
products concentration. When China’s growth declines, it would affect asean
economies differently. Countries which have large trade exposure and are highly
dependent on China would suffer the most. On the other hand, countries which
have low trade exposure and rely less on China’s economic activity would suffer
the least from China’s growth slowdown.
The adverse impact of China’s slowdown would be felt less in Cambodia,
Lao pdr, Myanmar, and Vietnam. In particular, developing rapid transportation
network in the region can compensate for the declining volume of trade with
China through increasing trade among asean countries, which would be
enhanced by reduced transportation costs. In addition, fdi outflows from
asean-5 into Cambodia, Lao pdr, Myanmar, and Vietnam can offset declining
export demand from China. Indonesia and the Philippines would be the least to
suffer from China’s slowdown. Market size and age structure of population can
ensure dynamism in investment and consumption expenditures. Thailand would
be the laggard in the region unless a democratic government can be restored.
Authoritarian rules reflect extractive institutions that create risks, and uncertainty
can drive out investment and lower long-term growth.
In the scenario of global growth slowdown, currency depreciation, and rising
interest rates, the main challenge in asean will be household and business
sector debts that would eventually impair the quality of bank loans and prolong
economic recovery. The road to economic recovery will be hard unless the asean
community can increase their intra-regional trade through improving regional
connectedness through massive infrastructure investment to compensate for the
declining China’s demand for exports.
Thammasat University
FIGURE 9. Impacts of high trade exposure on ASEAN economies
FIGURE 10. Impacts of low trade exposure on ASEAN economies
207
208
References
Athukorala, P. and H. Hill [2010] “Asian trade: long-term patterns and key policy
issues”, Asian-Pacific Economic Literature 24(2): 52-82.
Basri, M.C. and H. Hill [2011] “Indonesian growth dynamics”, Asian Economic
Policy Review 6(1): 90-107.
Bird, K. and H. Hill [2010] “Tiny, poor, land-locked, indebted, but growing:
lessons for late reforming transition economies from Laos”, Oxford
Development Studies 38 (2): 117-143.
Devadason, E.S. [2011] “Reorganization of intra-asean 5 trade flows: the ‘China
factor’”, Asian Economic Journal 25 (2): 129-149. Hill, H. [2015] “The political economy of policy reform: insights from Southeast
Asia”, in I. Coxhead, ed., Routledge handbook of Southeast Asian economics.
Oxford and New York: Routledge. 327-344.
Hill, H. [2014] “Is there a Southeast Asian development model?”, Malaysian
Journal of Economic Studies 51: 89-111.
Hill, H. and J. Jongwanich [2014] “Emerging East Asian economies as foreign
investors: an analytical survey”, Singapore Economic Review 59(3): 1-26.
Hill, H. and M.S. Gochoco-Bautista [2013] “Perspectives and issues”, in H. Hill
and M.S. Gochoco-Bautista, eds., Asia rising: growth and resilience in an
uncertain global economy. Cheltenham, U.K. and Northampton, Mass.: Elgar;
Manila: Asian Development Bank. 3-45.
Hill, H. and J. Menon [2013] “Cambodia: rapid growth with weak institutions”,
Asian Economic Policy Review 8(1): 46-65.
Hill, H., T.S. Yean, and R.H.M. Zin [2012] “Malaysia: a success story stuck in the
middle?” World Economy 35(12): 1687-1711.
Hill, H. and J. Menon [2011] “Reducing vulnerability in transition economies:
crises and adjustment in Cambodia”, asean Economic Bulletin 28(2): 134-159.
Hooy, C.W., S-H. Law, and C. Tze-Haw [2015] “The impact of the renminbi
real exchange rate on asean disaggregated exports to China”, Economic
Modelling 47: 253-259.
Ishaq, M. and M.A.U. Rehman [2013] “Surmounting the individual: establishing
a common currency in Asia--a case study of East Asian economies”, Global
Economy Journal 13(1): 63-88.
Masron, T.A. and E. Nor [2013] “fdi in asean-8: does institutional quality
matter?” Applied Economics Letters 20(1-3): 186-189. Mohan, R. and B. Nandwa [2009] “Examining interest rate linkages among
asean-5, China and India”, asean Economic Bulletin 26(2): 174-179. Napoli, C. [2014] “China’s economic rise: implications for asean trade
flows”, Journal of Southeast Asian Economies 31 (3): 345-360.
Ong, H.-B. and M.S. Habibullah [2012] “Is China compatible with asean-5?
a gradual cointegration analysis”, Journal of Economic Studies 39(3-4): 356367. The Philippine Review of Economics, Volume LII No. 2, December 2015
209
Tang, H.C. [2014] “Exchange rate volatility and intra-Asia trade: evidence by
type of goods”, World Economy 37(2): 335-352. Teng, K.T., S.H. Yen, and S.Y. Chua [2013] “The synchronization of asean-5
stock markets with the growth rate cycles of selected emerging and developed
Economies”, The Journal of Applied Economic Research 7(1): 1-28. Tham, S.Y. and A.J.Y. Kam [2014] “Re-examining the impact of ACFTA
on asean’s exports of manufactured goods to China”, Asian Economic
Papers 13(3): 63-82. PRE
Goods trade liberalization under the ASEAN Economic
Community: effects on the Philippine economy
Ramon L. Clarete* and Philip Arnold P. Tuaño**
The economy-wide effects of the asean Free Trade Area (afta) on
the Philippine economy are computed using the Global Trade Analysis
Project (gtap) model of the world economy. Of the 40 industries
representing the Philippine economy that were simulated to assess the
impacts of afta on sectoral output, 24 industries declined. However,
the order of magnitudes of the percentage declines is low, except for
rice, whose output decreased by about 4.5 percent.
Notwithstanding the contraction of production in the majority of
industries, the country comes out a net gainer in aggregate output by
around 1.4 percent in total gross domestic product. This implies that,
overall, the Philippines is slightly better off with the preferential trade
liberalization, with an equivalent variation gain of us$237.4 million.
Considerable movement of workers across industries is observed.
Reductions in skilled worker employment resulted in 31 industries,
and 35 industries do the same in the case of unskilled labor. There is
no change in unemployment due to the fact that the empirical model
assumes full employment of productive factors. Given the fact that
there is a change in employment patterns during trade liberalization
episodes, it is important to assess the empirical relationship between
trade liberalization and unemployment which remains difficult to
pin down.
The empirical literature in developing countries shows that the
employment response of trade reforms is dependent on infrastructure,
trade facilitation measures, and other policies. Therefore, policies
related to providing information and new market opportunities,
organizing value chains, facilitating coordinated investments along
the chain, clustering of related small and medium enterprises to boost
external economies, and providing information to employers and
workers of job opportunities, including training, are important.
JEL classification: F150
Keywords: ASEAN Economic Community, ASEAN Free Trade Area, trade
liberalization
211
1. Introduction
This paper examines the economy-wide effects of goods trade liberalization
in the Association for Southeast Asian Nations (asean) region on the Philippine
economy using the Global Trade Analysis Project (gtap) model of the world
economy. Goods trade liberalization is a key reform of asean leaders under the
asean Economic Community (aec). By transforming the region into a single
market and production base, its leaders seek to make the region more competitive,
attract more direct investments, generate more jobs, and increase productivity,
trade, and per capita income growth.
However, as in every trade liberalization episode, many view the removal
of tariff barriers as a process resulting in cheaper and ostensibly better quality
imported products, resulting in the loss of domestic jobs. On the other hand, trade
reforms are also seen as creating new markets for domestic products and creating
more employment opportunities.
The two effects on employment are likely to happen in the process. Jobs are
lost in some industries of the economy, and they are created in other industries.
This study also provides a brief assessment of the suitability of general
equilibrium economic models in estimating the net impact on jobs of goods trade
liberalization under aec.
Hill [2003:232] and Clarete [2006] noted that one of the development puzzles
in the Philippine economy is that while trade reforms resulted in marginally better
growth levels in the 1990s (at least compared to the previous decade), it did not
trigger improvements in manufacturing employment. Several explanations that
were explored include supply side constraints in the economy and the comparative
lack of demand.
This article is divided into six parts. In the following section, the key policy
reforms, including preferential trade liberalization, in aec are discussed. The
third section reviews the methodology used in computing the effects of tariff
reforms. The results of the ex-ante analysis are in the fourth section of the
paper. Considering that the ex-ante analysis abstracts from structural rigidities
of the economy, the fifth section takes up the adjustment process to freer trade;
the empirical and analytical assessments of the relationship between trade
liberalization and employment effects are examined. Key observations and
recommendations are provided in the concluding section.
2. Goods trade reforms in AEC
This section provides the context of goods trade liberalization in asean. Even
before the 2003 aec declaration at Bali, asean leaders had already declared the
region to be a free trade area. aec prohibited non-tariff barriers, introduced rules
of origin and other non-tariff measures, and provided trade facilitation measures.
212
Clarete and Tuaño: Goods trade liberalization under the ASEAN
Economic Community: effects on the Philippine economy
2.1. The ASEAN Free Trade Area
The agreement that created the asean Free Trade Area (afta), which came
into force in 1992, was signed during the period when the world “exploded” with
free trade agreements [Soloaga and Winters 2001]. According to the World Trade
Organization [2011], there hardly were any preferential trade agreements created
between the 1950s and 1980s. It was in the 1990s that the number surged to about
70 agreements, and then it rose to 300 agreements in 2010.
Under the agreement that launched afta, member countries were required
to reduce the trade taxes imposed on goods imported from their fellow member
countries. Brunei, Indonesia, Malaysia, the Philippines, Singapore, and Thailand,
collectively known as the asean-6, initially comprised the asean membership.
Cambodia, Lao People’s Democratic Republic (or Lao pdr), Myanmar, and
Vietnam, known as the clmv countries, subsequently joined the regional bloc.
This Common Effective Preferential Tariff (cept) Scheme, which afta called
the system of tariffs that would be imposed on the products imported from the
other asean member countries, ensured that the tariff rates would be brought
down at zero to five percent rates, in a time frame of eight years from the signing
of the agreement.
The rules of origin were also agreed upon in the trade agreement. This meant
that the goods, excepting for some commodities, would be subject to preferential
trade and should have a local content of at least 40 percent of the freight on board.
It was also agreed that quantitative restrictions would be removed on goods
whose tariffs would be reduced and all other non-tariff measures related to these
products would be reduced in a five-year period.
With the entry of the clmv countries, the free trade agreement was reaffirmed
in November 1999 when asean agreed to implement a reduction of tariffs to
zero by 2010 for the asean-6 and by 2015 for clmv countries. Products in the
priority sectors, especially in the manufactured goods sectors, would be eliminated
by 2007 for the asean-6 and by 2012 for the clmv, while all tariffs would be
eliminated by 2015 for the asean-6 and by 2018 for the clmv countries.
2.2. Leap to AEC
The transformation of afta into an “economic community” took place in
2003 with the Bali declaration. asean leaders adopted the vision of creating
a “single market and production base” in order to make the regional economy
“stable, prosperous, and highly competitive” [asean 1997] as one of the three
pillars of regional cooperation, including security, sociocultural integration, and
economic integration, as agreed upon in the asean Concord (Bali Concord II).
During the asean Summit in Cebu in January 2007, which developed the
aec Blueprint, the deadline for integration was brought forward from 2020
to 2015. Because of this, the asean Trade in Goods Agreement (atiga) was
213
signed in 2009, which formalized the tariff agreements that have been made in
the past and improved the transparency and predictability of changes in tariffs.
atiga also emphasized trade facilitation measures with a work plan to be put in
place from 2009 to the start of aec in 2015.
2.3. Goods trade reforms under AEC
atiga also emphasized trade facilitation measures with a work plan to be
put in place from 2009 to the start of aec in 2015. Besides tariff reduction, the
following are the significant provisions of atiga and the ancillary agreements:
elimination of non-tariff barriers [De Dios 2007]; continuous reform of the
rules of origin rules; implementation of harmonized trade facilitation processes;
establishment of an asean single window to expedite clearance processes for
trade flows; and harmonization of standards and technical barriers to trade.
2.3.1. Tariff reforms
The afta agreement came into force in 1992, which aimed, among other
purposes, to reduce intra-regional import tariffs to no more than five percent
under the cept Scheme. However, the scheme allowed members to draw up their
respective Inclusion List, Sensitive List, Highly Sensitive List, General Exclusion1
List, and tariff reduction schedules. The new member states—Cambodia, Lao
pdr, Myanmar, and Vietnam (or clmv)—were given flexibility in terms of a
longer implementation period than the asean-6.
In 1995, asean leaders decided to accelerate the cept process by moving
the completion date from 2008 to 2003. The tariff rates on sensitive imported
products from the region were eventually phased into the cept process. Member
states were legally bound to reduce the tariff rates on these products to no more
than 5 percent. For asean-6, the target year of completion of the cept process
for Sensitive List imports was 2010 and 2015 for clmv states.
The 2007 aec Blueprint affirmed the agreements of member states regarding
the parameters of the preferential tariff reforms, i.e. the target rates, deadlines, and
schedules of tariff reduction. Nonetheless, the blueprint emphasized the urgency
of these reforms and desirability of minimizing the cases of departure from the
agreed tariff reduction schedules. The blueprint extended the cept Scheme to
2015 for asean-6 and up to 2018 for clmv states to cover for the integration of
Sensitive List imports into the cept Scheme, as well as to set the final tariff rates
on Highly Sensitive List products.
General Exclusion List products are permanently exempted from the tariff reduction process to uphold
national security, public morals, and public health, as well as protect the environment and articles of artistic,
historic, or archaeological value (asean Secretariat, 1999).
1
214
As part of its commitments to asean, in December 2009, the Philippines
passed Executive Order No. 8502 which removed tariffs on imports from asean,
except for products in the Sensitive List and Highly Sensitive List. Approximately
94 percent of tariff lines in the country’s Tariff Reduction Schedule is already set
to 0 percent.
Agricultural tariffs remain high in the Philippines. The average tariff of
dutiable agricultural and fisheries products in 2013 is 13.01 percent, which is set
to fall to 10.23 percent by 2015. This is due to the feature of the country’s cept
that reduced the tariffs of its Sensitive List imports down to only 5 percent, which
is an acceptable ending rate of the cept. The Philippine Sensitive List includes
swine, poultry, cassava, sweet potatoes, corn, grain sorghum, and sugar for the
detailed list. Except for sugar, the tariff rates of products in the Sensitive List
are already down to 5 percent since 2010. Rice is the only item in the Philippine
Highly Sensitive List. Based on the Tariff Reduction Schedule, rice and sugar
tariffs go down to 35 percent and 5 percent respectively in 2015.
The reduction and elimination of tariffs in asean has been considered
successful. As of 2010, intra-asean tariff rates were virtually zero in the
asean-6, and 2.6 percent was the mean preferential tariff rate under atiga in the
newer clmv member states. By 2015, rates on 98 to 100 percent of all tariff lines
are expected to be in the 0-5 percent range. Table 1 shows the percentage of tariff
lines with 0 percent tariff as of 2013 for each of the asean countries. According
to asean Secretariat, 99.85 percent of regional tariffs of asean-6 member states
are zero, while the corresponding number for newer asean members was 69
percent. About 30 percent of the goods in the clmv currently have rates greater
than zero, while slightly over a third of a percent is observed for the asean-6.
The residual share of tariff lines is for those not offered for preferential reduction,
and these tariffs apply to products in the General Exclusion List. The General
Exclusion List tariff lines for the Philippines is nearly a third of slightly over a
quarter of 1 percent and 0.45 percent or the asean-6 states.
2.3.2. Elimination of non-tariff trade barriers
Quantitative restrictions to imports, whether explicit or not, are non-tariff trade
barriers (ntbs) and are thus prohibited under atiga. Under the aec Blueprint,
all member states are to commit not to expand the number of ntbs they maintain
and to roll back any that they currently implement. In the interest of promoting
transparency, the asean member states commit to notify the asean Secretariat
“Modifying the Rates of Duty on Certain Imported Articles as Provided Under the Tariff and Customs
Code of 1978, As Amended in Order to Implement the Commitment to Eliminate the Tariff Rates on the
Remaining Products in the Inclusion List in Year 2010 Under the Common Effective Preferential Tariff
(CEPT) Scheme for the ASEAN Free Trade Area (AFTA)/ASEAN Trade in Goods Agreement (ATIGA)”
2
215
of their respective use of ntbs. The secretariat has to maintain an effective
surveillance mechanism to monitor the implementation of ntbs and non-tariff
measures. Member states are legally bound to eliminate all ntbs by 2010 for
asean 5, and by 2012 for the Philippines, although this had already changed
because of the June 2014 waiver.3 The clmv states have until 2018 to remove the
ntbs that they maintain.
TABLE 1. Percentage of tariff lines at 0 percent
in the ATIGA tariff schedule of 2013
Country
Brunei
Indonesia
Malaysia
Philippines2
Singapore
Thailand
ASEAN-6
Cambodia
Lao People’s Democratic Republic
Myanmar
Vietnam
CLMV
Percentage of tariff lines (%)
0%
Greater than 0%
Other1
99.27
0.73
98.87
0.17
0.96
98.74
0.59
0.66
98.62
1.11
0.27
100.00
99.85
0.15
99.20
0.35
0.45
40.77
59.23
78.73
20.36
0.91
79.66
19.69
0.65
72.24
25.77
1.99
68.88
30.20
0.92
Source: ASEAN Secretariat [undated]
2.3.3. Trade facilitation
With tariff rates and ntbs eliminated in the region, the agenda for closer
integration in goods trade is increasingly focused on the harmonization and
efficient administration of the various regulations and para-taxes that affect
international trade.
asean [2011] determined the major ntbs in the region. The first category
comprises customs surcharges. In a survey conducted, asean reported that about
70 percent of the ntbs implemented are customs surcharges. Technical measures,
another category, make up 14 percent, followed by product characteristics
requirement, 10 percent. The remaining 6 percent of observed ntbs included
charges other than those collected by customs, state trading and single channel
rules, marketing requirements, and technical regulations.
The Philippines presently has a waiver from the World Trade Organization to continue to maintain the
country’s quantitative restrictions on rice imports in the form of the monopoly of rice imports under the
National Food Authority. The 2012 deadline for the Philippines to eliminate its non-tariff barriers reflected
the country’s original plan not to extend its quantitative restrictions for rice after it would have expired in
2012. The government changed plans and decided to apply for a waiver from the World Trade Organization.
3
216
Non-tariff measures regulate the flow of trade to act on observed problems that
international trade may inadvertently cause. An important category is the set of
sanitary and phyto-sanitary standards and regulations on traded plant and animalbased products. These ensure that imported foods, feeds, planting materials,
and breeding stocks are safe for consumption or for use by the public, or that
their entry into the country does not introduce diseases and pests to native plants
and animals.
In non-agricultural products, governments require that imports meet their
respective standards and conform to existing regulations. These standards
and regulations mitigate the risk of substandard imports that may jeopardize
public interest.
Because of the preferential nature of trade liberalization under asean, rules
of origin have become part of the free trade area rules to determine eligibility
and to mitigate risk of trade deflection. atiga states that the administration of
rules of origin be kept simple and continuously responsive to the requirements of
private businesses in the region, particularly with respect to promoting regional
and global value chains.
The aec Blueprint has also called for the modernization of customs procedures
and regional integration of national customs administration, which are set in the
asean cargo clearance and customs declaration document [asean 2007].
Another important contribution to facilitating trade is the harmonization of
standards, technical regulations, and conformity assessment procedures. The aec
Blueprint calls on member states to work together to develop and implement such
harmonized systems of standards and technical regulations based on international
best practices. Trade facilitation programs, required by atiga, are intended
to avoid or mitigate any adverse effects of administering non-tariff measures,
whether it is before the goods arrive or are exported, during cargo clearance at
customs, or after cargoes are cleared.
Another area that aec is focusing on is improving transparency of regulations.
aec calls for the establishment of a region-wide asean Trade Repository of
existing trade regulations, where information on regulations of member states is
accessible to stakeholders.
The aec Blueprint provides for measures that reduce trade costs. It calls on
member states for collective action through the asean Secretariat to undertake
studies on trade and to implement a comprehensive work program in order to
simplify, harmonize, and standardize trade licensing and cargo clearance
processes and procedures.
3. Structure of the model
A computable general equilibrium model of the global economy is used in
analyzing the impact of aec tariff reforms. The model allows the tracking of the
economy-wide effects of changes in asean tariff protection in a way that takes
217
into consideration their effects on global markets of products and in turn their
feedback to its component national economies.
3.1. Production structure
The model comprises 40 production sectors and 23 countries/regions of the
world. Table 2 lists and classifies the production sectors of the model by broader
sector categories that they belong to, namely agriculture, natural resources,
manufacturing, and services. There are nine primary agricultural production
activities. Forestry, mining and oil, and gas are three natural resource extraction
activities. There are 19 manufacturing activities and nine services industries.
Each of the 40 sectors produces only one unique product. The production
activity for each sector is modeled by a nested Leontief-constant elasticity
of substitution production function, i.e., there are several input-to-output
transformation processes at various levels of the production activity (see Figure 1).
At the top level, the activity’s output is a nested function of two composites,
one for intermediate inputs and the other for primary inputs or value added. The
former in turn depends upon another set of composites, called Armington [1969]
goods, with each good or service made up of a locally produced input and its
imported equivalent, which are substitutable subject to a constant elasticity of
substitution parameter.
Value added is a function of primary factors of which there are four in the
model: labor; skilled labor; land; and capital. These inputs are constant elasticity
of substitution-substitutable among themselves, i.e., they substitute with each
subject to a constant elasticity of substitution parameter. All primary factors are
free to be moved from one industry to another. However, land is an input in only
agricultural production activities, and it is freely movable among the latter.
The outputs are used for final consumption, exports, and intermediate inputs
into Armington composites. Both final and intermediate uses of the product are
sales to the domestic markets. Output is split between domestic sales and exports
through a constant elasticity of transformation parameter.
3.2. Demand structure
Twenty-three countries or regions represent the global economy (see Table 4).
Fifteen of these regions are in Asia, with Southeast Asia having the most number of
individual countries represented. Of the 10 asean member states, Cambodia and
Lao People’s Democratic Republic are grouped as one region, while Myanmar,
Brunei Darussalam, and East Timor make up the rest of Southeast Asia.
218
TABLE 2. The production sectors of the model
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Paddy rice
Cereals
Oil seeds
Sugarcane and beets
Vegetables and fruits
Other crops
Other animal products
Cattle
Fishery
Forestry
Mining
Oil and gas
Meat preparations
Dairy
Vegetable oils
Processed rice
Milled sugar
Other food products
Beverages and tobacco
Textile and garments
Abbreviations
A: agriculture
A
A
A
A
A
A
A
A
A
NR
NR
NR
M
M
M
M
M
M
M
M
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40
Wood products
Paper and publishing
Leather
Chemicals, rubber, and plastic
Petroleum and coal
Non-metal mineral products
Metal products
Machinery products
Electrical products
Transport equipment
Other manufacturing
Construction
Fuel, electricity, and water
Transport services
Trade
Communications
Financial intermediary
Public administration, education, and health
Real estate and commercial services
Rest of services
NR: natural resources M: manufacturing
Local sales
M
M
M
M
M
M
M
M
M
M
M
S
S
S
S
S
S
S
S
S
S: services
Export
Output
Intermediate input composite
AG1
LP1 MP1
AG2
LP2 MP
2
Notes
LP: local product
AG40
Value added
Low
skilled
labor
High
skilled
labor
Land
LP40 MP40
MP: imported product
AG: Armington good
FIGURE 1. Structure of production activities of the model
Capital
219
TABLE 3. The geographical regions of the model
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Indonesia
Cambodia and Lao People’s
Democratic Republic
Malaysia
Philippines
Singapore
Thailand
Vietnam
Rest of Southeast Asia
China
Hong Kong and Taiwan
Japan
South Korea
Abbreviations
SEA: Southeast Asia
SEA
SEA
13.
14.
India
Rest of South Asia
SA
SA
SEA
SEA
SEA
SEA
SEA
SEA
EA
EA
EA
EA
15.
16.
17.
18.
19.
20.
21.
22.
23.
Australia and New Zealand
Canada
United States of America
Brazil
Rest of North and South Americas
European Union 27
Middle East and North Africa
Rest of Africa
Rest of the World
ANZ
A
A
A
A
EU
MENA
Af
ROW
EA: East Asia
SA: South Asia
A: Americas
Each region has a representative private household and a government. Both are
treated in the model as maximizing a utility function subject a budget constraint,
with the latter having a Cobb-Douglas utility function.
The representative regional consumer draws its income from its ownership of
the four primary factors of the model, and income transfers from the government
and the rest of the world. After deducting the taxes it pays to the government,
the consumer then apportions the disposable income to savings and consumption.
The government’s income comes from taxes collected and spends it on local
and imported products. Savings in the model are pooled by a global financial
intermediary, and are allocated to the various regions based on their respective
investment financing requirements.
3.3. Equilibrium conditions
The computable general equilibrium modeling structure of gtap entails the
latter observing accounting relationships that correspond to the various conditions
that define when the global and regional economies are in a state of equilibrium
or balance, utilizing a regional household and a global bank.
The regional household receives and allocates income and expenditure flows.
It receives the incomes of the primary factors in consideration of the use of
the respective services of these factors in producing goods and services. It also
receives taxes, paid by the private household and the government when they buy
goods and services, of which there are two types. Internal income and indirect tax
revenues are those from taxes on local purchases, and trade taxes are those from
imports and exports of goods and services. The border taxes are the revenues from
import tariffs, export taxes, and other related tax measures collected at the border.
220
The regional household spends all its income on two income transfers and
savings. One of the two transfers goes to the private household, and the second
is given to the government. These transfers then figure out as the respective
budgetary constraints of the two when they maximize their respective utility
functions as discussed above. The savings that the regional household generates
goes to a global bank. The bank pools the savings from all regional households of
the model and allocates the total across various competing uses of the savings by
the regions. The manner how this is allocated is that all regions with competing
uses of savings face the same terms of use.
3.4. Global Trade Analysis Project data and model
The data used in the study comes from version 8. 1 of the gtap data [in
Narayanan, Bardi, and McDougall 2012]. The gtap 8.1 data set comprises a total
of 137 countries and 57 sectors. The Center for Global Trade Analysis at Purdue
University maintains and regularly updates the data set since 1993 [Hertel 1999].
The baseline year is 2007.
This study used the gtap model in computing the impacts of aec tariff
reforms. It is a multi-regional and multi-sectoral model, and it is used in
counterfactual analysis such as on the possible implications of the policy changes
(Hertel [1999]; Brockmeier [2001]). The gtap model is set up as one that can
be solved using gempack [Harrison and Pearson 2002]. gempack solves for
percentage changes of the economic variables following a change of policies.4
4. Economic effects of AFTA
4.1. Production effects
Table 5 portrays the changes of production outputs, measured in billions of
pesos, for each of the 40 production sectors of the gtap model. These changes
reflect how the resources of the economy get to be reallocated in the economy as
a result of lower preferential tariff protection in asean.
With lower protection, Philippine industries lose their domestic markets
to imported substitutes and thus reduce their outputs, as shown in the Table 4.
Altogether, 24 industries, or more than half of the total, are observed to
contract production.
The order of magnitudes is in the range from less than a percent to up to 4.5
percent. These are rather low changes, except for rice, whose output decreased by
about 4.5 percent. This is despite the fact that rice continues to have the highest
asean tariff rate of 35 percent.
An alternative solution technique is to solve for the counterfactual values of the economic variables of the
model. For this, modelers use the GAMS software, and a version of the model in GAMS was developed
by Rutherford [2005].
4
221
TABLE 4. Effects of AFTA on production,
by industry, Philippines* (in billions of pesos)
Sectors
Base
value
Paddy rice
142.36
Change
Sectors
%
Base
value
Change
%
-4.58 Wood products
86.28
Cereal
56.64
-0.40 Paper and publishing
96.28
0.07
Oil seeds
35.84
-1.02 Leather
30.28
-0.94
Sugar cane and beets
25.48
-1.83 Chemicals, rubber, and plastic
140.72
Other crops
43.96
227.96
-0.18 Petroleum and coal
15.13 Non-metal mineral products
-1.63
327.6
-0.86
260.56
0.30
87.28
0.28
0.17 Metal products
341.76
0.90
0.35 Machinery products
438.24
-1.36
Cattle
28.12
Fishery
180.08
1912.04
-1.56
Forestry
14.52
-1.15 Transport equipment
246.32
16.54
Mining
92.52
-0.92 Other manufacturing
117.32
-0.62
388.4
1.29
388.4
0.00
585.64
0.08
Oil and gas
Meat preparations
13
278.56
Dairy
52.56
Vegetable oils
83.52
Processed rice
141.28
Milled sugar
43.8
-0.03 Electrical products
-0.91 Construction
0.06 Fuel, electricity, and water
0.45 Transport services
-1.82 Trade
1232.68
0.14
-4.78 Communications
213.92
-0.12
-1.95 Financial intermediary
396.48
0.11
812.72
-0.05
724.72
-0.32
217.48
-0.16
11,344.60
1.41
Other food preparations
430.64
Public administration, education, and
-0.17
health
130
4.62 Real estate and commercial services
278.56
-0.97 Rest of services
All sectors
Note: Shaded entries are those for industries with increases of their respective outputs.
Source: Authors’ computations The economic resources they lost went to the 15 industries.5 Transport
equipment increased the greatest with an increase of 16.54 percent. Although
its component production activities have very low base values, the other crops
industry increased its output by 15.13 percent. Beverages and tobacco comes
second with growth of 4.62 percent. The overall production of the economy went
up by 1.41 percent. In the non-agricultural industries and services, transport
equipment and construction show significant expansion of their outputs.
The Philippines, despite the output contraction of 24 or more than half of
its industries, comes out a net gainer in aggregate output. With a base value of
aggregate output at P11.3 trillion, the country comes out of this tariff liberalization
process as gaining by 1.41 percent. The gains more than outweigh the losses in
aggregate production in the economy.
No change in output is observed for the fuel, electricity, and water industry.
5
222
4.2. Employment effects
Given the changes in output, considerable movement of workers across
industries is observed from the results (see Table 5).6 Thirty-one (31) out of forty
industries contract employment of skilled workers, and 35 industries do the same
in the case of unskilled labor. In agriculture and natural resources, ten out of
12 industries lay off skilled workers. Rice paddy, sugar cane, and oil seeds cut
back on skilled labor employment by at least 1 percent. The other crops industry
appears to absorb those displaced with an expansion of hiring at the rate of 16.22
percent.
Thirteen (13) out of 20 manufacturing industries give up skilled workers to
other industries. In the case of low-skilled labor, 15 industries are observed to do
the same. The industries with the largest layoff rates are agriculture-based, namely
milled rice, milled sugar, and vegetable oils. Six and seven industries in the case
of skilled and low-skilled workers, respectively, follow, whose layoff rates are
between a percent and two, except for the wood products industry that cut back
on low-skilled workers employment by 2.13 percent. The rest of manufacturing
cut back on employment by less than a percent.
The manufacturing industries that increase employment of skilled workers
are the following: dairy; beverages and tobacco; petroleum and coal; non-metal
mineral products; metal products; transport equipment; and construction. The
same industries increase as well their employment of low-skilled workers, except
for petroleum and coal, non-metallic mineral products, and dairy industries,
which shed off jobs albeit at less than a percent.
All eight services industries lay off workers, but the rates of change are all
less than a percent. This is due to the fact that there are services industries that
contracted output; these include communication, public administration, real
estate and commercial services, and the rest of services. These output cutbacks
may explain the reduction in employment reported in Table 5. However, the
remaining three services industries expanded production, and fuel, electricity and
water utility neither contracted nor expanded their output. But all four services
industries gave up workers despite the lack of output contraction.
In Table 6, entries colored green show the industries that increase employment of workers. Under the
allocation columns, shares that are less than a percent are not shaded. Those industries having shares
between 1 and less than 4 are shaded light red, while light green is used to shade entries with shares of at
least 4 percent.
6
223
TABLE 5. Employment changes and allocation of labor by industry, baseline and
simulated levels due to AFTA tariff reforms (in %)
Employment changes
Industry
Paddy rice
Cereal
Oil seeds
Other crops
Cattle
Fishery
Forestry
Mining
Oil and gas
Meat preparations
Dairy
Vegetable oils
Processed rice
Milled sugar
Wood products
Leather
Petroleum and coal
Metal products
Machinery products
Electrical products
Transport equipment
Other manufacturing
Construction
Transport services
Trade
Communications
Public administration, education,
and health
Real estate and commercial
services
Rest of services
Total
Skilled Low-skilled
workers
workers
-5.29
-5.46
-0.77
-0.95
-1.44
-1.63
-2.32
-2.51
-0.53
-0.71
16.22
16.01
-0.15
-0.34
0.04
-0.15
-0.35
-0.54
-1.46
-1.64
-1.24
-1.41
-1.45
-1.64
-0.16
-0.44
0.23
-0.06
-2.04
-2.32
-4.99
-5.26
-2.17
-2.45
-0.39
-0.67
4.38
4.07
-1.18
-1.47
-1.83
-2.13
-0.15
-0.45
-1.15
-1.44
-1.08
-1.37
0.09
-0.21
0.06
-0.24
0.69
0.39
-1.57
-1.87
-1.77
-2.07
16.28
15.91
-0.83
-1.11
1.10
0.78
-0.23
-0.53
-0.10
-0.45
-0.04
-0.39
-0.35
-0.64
-0.12
-0.42
Share of sector in employment
Skilled workers
Low-skilled workers
Baseline
AFTA
Baseline
AFTA
0.004
0.029
0.011
0.003
0.052
0.035
0.021
0.034
0.006
0.013
0.234
0.310
0.167
0.114
0.035
0.020
0.037
0.659
0.279
0.479
0.364
1.214
0.089
2.173
0.110
0.479
1.613
3.957
0.908
1.768
0.377
4.957
1.434
2.337
7.507
2.289
9.481
0.004
0.029
0.011
0.003
0.052
0.041
0.021
0.034
0.006
0.013
0.231
0.305
0.167
0.115
0.035
0.019
0.036
0.656
0.291
0.473
0.357
1.212
0.088
2.150
0.110
0.479
1.624
3.895
0.892
2.056
0.374
5.011
1.430
2.335
7.504
2.281
9.470
0.251
0.566
0.239
0.076
1.687
0.625
0.623
0.693
0.233
0.336
0.665
0.660
0.516
0.322
0.103
0.065
0.093
1.371
0.568
1.505
0.974
1.721
0.291
2.670
0.170
1.096
3.356
4.257
1.170
2.575
0.891
10.189
1.468
4.807
14.894
1.389
4.861
0.237
0.561
0.235
0.074
1.675
0.725
0.621
0.692
0.232
0.330
0.655
0.649
0.514
0.322
0.101
0.061
0.091
1.362
0.591
1.483
0.954
1.713
0.287
2.634
0.169
1.094
3.369
4.177
1.146
2.985
0.881
10.269
1.460
4.786
14.836
1.380
4.841
-0.30
-0.59
36.753
36.643
20.660
20.537
-0.55
-0.84
15.348
15.264
8.930
8.855
-0.39
-0.68
4.300
100.000
4.283
100.000
2.432
100.000
2.416
100.000
Note: Shaded entries are those for industries with increases of their respective outputs.
Source: Authors’ computations
224
120
100
80
Industry
19.80
0.75
AFNR
20.04
0.75
Servces
33.91
60
40
34.20
6.65
79.45
79.21
Base level,
skilled labor
AFTA level,
skilled labor
20
59.44
6.69
59.11
0
Base level,
low skilled labor
AFTA level,
low skilled labor
Note: Agriculture, fishery and natural resources (ANFR)
Source: Table 5
FIGURE 2. Allocation of skilled and low-skilled labor, base and simulated levels
due to AFTA tariff rates (%)
4.3. Effects on trade
Table 6 shows the changes to the country’s exports and imports that afta
may bring about. In agriculture, fishery and natural resources, the top exporting
industries are vegetables and fruits, mining, other crops, and fisheries industries.
Altogether, the agriculture and fisheries industries come out with a positive
net gain in exports at us$183.64 million. The country’s exports in vegetables
and fruits decline by nearly a percent. Other crops industry registered to have the
highest expansion rate at 280.8 percent. The base value is relatively significant,
but this industry is a collection of several crops not elsewhere specified, each of
which may have relatively low levels of exports.
The three natural resources industries suffer cuts of their respective export
levels, except for oil and gas. This sector’s gain is inadequate to offset export
losses in mining and forestry. Mining is a far significant exporter than oil and
gas. The loss of natural resources exports amounts to us$10.7 million, mostly
in mining.
It’s in the manufacturing sectors that the country gains relatively the most
in exports, altogether nearly us$405.4 million. Nine out of 20 industries are
observed to have reduced their exports. The biggest exporter, electrical products,
with base export value of more than us$41.8 billion, experience a 1.59 percent
decline of its exports.
The gainers more than offset the losses of the three large export performers.
Automotive spare parts, under transport equipment, have a significant gain, 47.87
percent, and their base export value is about us$2 billion. Although it expands
only by a percent, metallic products industry has a large base export value of
us$3 billion.
225
TABLE 6. Effects of AFTA on Philippine exports and imports,
by industry* (in million US$)
Industry
Paddy rice
Cereal
Oil seeds
Other crops
Cattle
Fishery
Forestry
Mining
Oil and gas
Meat preparations
Dairy
Vegetable oils
Processed rice
Milled sugar
Wood products
Leather
Petroleum and coal
Metal products
Machinery products
Electrical products
Transport equipment
Other manufacturing
Construction
Transport services
Trade
Communications
Public administration, education, and health
Real estate and commercial services
Rest of services
All Sectors
Base export
values
0.0
1.8
1.5
0.0
792.0
68.0
11.4
0.0
121.0
3.8
1,295.0
0.4
56.9
146.0
768.0
21.9
90.4
1,043.0
165.0
2,648.0
1,112.0
228.0
151.0
1,687.0
812.0
298.0
3,028.0
5,415.0
41,858.0
2,094.0
527.0
113.0
141.0
3,160.0
651.0
579.0
349.0
367.0
Change
(%)
9.09
0.4
-0.59
0.5
-0.81
280.79
-1.01
-3.52
-0.63
-2.39
-0.82
2.9
9.82
7.47
-1.22
3.96
1.58
0.86
76.79
-1.11
-1.38
3.1
-0.27
-0.48
4.23
0.5
1.0
-1.55
-1.59
47.87
-2.07
-1.46
-2.53
-0.43
-1.83
-1.98
-2.26
-2.16
Base import
value
0.0
1.9
1.5
0.0
943.0
73.2
11.9
0.0
136.0
4.2
1,682.0
0.4
58.0
159.0
808.0
21.9
98.2
1,138.0
172.0
2,792.0
1,237.0
252.0
163.0
1,833.0
860.0
350.0
3,119.0
5,613.0
42,331.0
2,190.0
556.0
113.0
141.0
3,160.0
651.0
579.0
349.0
367.0
Change
(%)
-6.77
0.32
0.47
-0.55
1.26
15.29
0.58
1.92
3.3
0.15
1
0.34
2.16
1.19
2.02
20.62
36.97
2.8
1.35
0.97
3.8
1.42
2.13
0.99
0.88
2.95
2.14
1.74
-0.96
6.76
2.28
2.05
1.26
0.46
1.08
1.06
1.28
1.15
2,653.0
470.0
72,927.1
-2.29
2,653.0
0.96
-1.89
0.62
470.0
75,088.2
0.08
0.32
Note: Shaded entries are for industries that expanded their levels of trade due to AFTA tariff reforms.
226
All eight services industries have reduced exports, and together lose about
us$126 million. Transport services and real estate and commercial services
industries are among the country’s largest services exporters. The former declines
by less than half of a percent, while the latter’s exports falls by 2.29 percent. Five
out of the eight services sectors increase their export value by at least 2 percent.
Summing up, the country gains us$452.36 million. The manufacturing industries
take the lead in the country’s exports, followed by agriculture and fisheries.
Beverages and...
Other food...
Milled sugar
Processed rice
Vegetable oils
Dairy
Meat preparations
Oil and gas
Mining
Forestry
Fishery
Cattle
Other crops
Sugar cane and beet
Oil seeds
Cereal
Paddy rice
4.74
-0.3
0.2
0.7
Rest of services
Real estate and ...
Public...
Communications
Trade
Transport services
Fuel, electricity, and...
Construction
Other manufacturing
Transport equipment
Electrical products
Machinery products
Metal products
Non-metal mineral
Petroleum and coal
Chemicals, rubber
Leather
Wood products
1.2
-0.3 -0.1 0.1
FIGURE 3. Effects of AFTA on product prices (in %)
0.60
0.40
0.20
0.00
-0.20
-0.40
Land
Skilled
Low
skilled
Capital
FIGURE 4. Effect on returns to primary factors (in %)
0.3 0.5
0.7
227
In contrast, the country’s import bill rises by us$241.26 million. The majority
of the import activities expanded their activities, except for paddy rice, sugar
cane, and electrical products. Electrical products, metallic products, machinery,
textiles and garments, and transport equipment are the country’s largest importers
among the manufacturing industries. The reduction in imports in electrical
products is the largest, and this result appears consistent with lower exports
coming out of this industry. The overall exports increased by about .62 percent,
while imports expanded by .32 percent. The country has a base trade deficit of
us$2,161 million. In the simulated trade levels, the deficit is observed to go down
by us$211 million.
4.4. Effects on prices
Figures 3 and 4 show the changes of the prices of goods and returns to the
primary resources used in production. In Figure 4, all goods, except 5, gain
increases in prices. That is, simulating the preferential tariff reforms results in
pushing up world prices, albeit at different rates. The highest rate of increase, 4.74
percent, is observed for other crops. The order of magnitudes of price increases is
up to about half of a percent.
Figure 4 shows the changes in returns to owners of resources used as primary
inputs in the various production activities of the economy. Wages of both skilled
and low-skilled workers rise by nearly half of a percent. Land rents, however,
go down by about a fourth of a percent. Land is mostly used in agriculture and
particularly in rice.
4.5. Overall economic effects
The changes in the country’s gross domestic product and economic welfare
are shown in Table 8, which provides information on the gross domestic product
at constant prices which is the total value added generated by an economy. This
is presented in million US dollars as the changes in other countries in the gtap
model are likewise shown for ease of comparison with that of the Philippines.
As expected, asean member states are better off with afta. The percentage
changes are small, and this reflects the fact that the base year of gtap is 2007
which is already just about the end of the cept process.
Is the Philippines better off with the preferential trade liberalization in goods
under aec? The equivalent variation, which measures the amount of money a
country is willing to pay in order to be as well-off as having afta under aec,
in the last column of Table 7 is used to indicate the economic well-being of the
country. Singapore tops the gainers, with us$2.4 billion, followed by Malaysia
and the Philippines. The Philippines has an equivalent variation of us$237.4
million. With this figure, the reforms marginally benefit the country.
228
TABLE 7. Changes in real GDP and economic well-being,
by country (in US$ million)
Country
Real GDP
432,103.19
0.019
Economic
welfare (in
equivalent Country
variation
of income)
64.87
India
12,644.03
0.194
182.27
Malaysia
186,642.11
0.145
430.37
Philippines
144,070.47
0.052
237.41
Singapore
176,759.66
0.037
2,389.27
Thailand
247,109.83
0.135
210.79
Vietnam
68,435.25
0.022
24.31
Rest of
Southeast Asia
28,601.50
0.172
76.41
China
3,494,058.00 -0.002
Base
Indonesia
Cambodia and
Lao People’s
Democratic
Republic
Hong Kong and
600,833.13
Chinese Taipei
Japan
%
-421.73
Real GDP
-0.008
Economic
welfare (in
equivalent
variation
of income)
-328.49
-0.002
-32.80
995,227.63
-0.002
-101.61
1,424,062.63
0.000
-15.45
Base
1,232,816.38
Rest of South
266,235.88
Asia
Australia and
New Zealand
Canada
United States
of America
Brazil
Rest of North
and South
Americas
European
Union 27
Middle East
and North
Africa
%
14,061,782.00 0.000
-601.56
1,365,983.25
-0.001
-47.33
2,470,987.75
0.001
36.62
17,003,710.00 -0.001
-767.54
2,530,414.00
-0.001
183.00
-0.003
-191.83
Rest of Africa
879,130.94
-0.001
-16.45
4,377,944.00 -0.004
-827.83
Rest of the
World
2,782,553.00
0.002
111.00
Source: Authors’ computations
* Shaded entries refer to industries with positive gains in real GDP and equivalent variation of income.
5. Adjustment process
This section identifies several issues in the country’s adjustment process to
determine the complimentary policies and programs of the Philippine government
to fully realize the economic benefits of goods trade integration. It basically
asks the question: What effects may the Philippines expect if all the goods trade
liberalization reforms under aec are fully implemented without regard as to when
those effects are going to take place?
Does asean integration help the Philippines create jobs? As with every
trade liberalization process, local workers, farmers, and business people view
the economic integration undertaken through aec to be a process whereby local
products will be substituted out in the local markets by cheaper and ostensibly
better quality imported products. This implies loss of domestic jobs, travel to look
for jobs overseas, and adjustment costs.
However, local exporters see the economic integration in asean as
providing them new markets for their products and creating more and new jobs
for the country. With the talk about aec supporting the formation of regional
229
value chains, local exporters view this process as bringing them into a more
cooperative—rather than competitive—process.
The two effects on employment are likely to happen in the process. Jobs are
lost in some industries of the economy and created in others. As conventionally
structured, general equilibrium models are full employment models. As such,
no net jobs are created or lost by trade liberalization. What can be observed is a
change in the allocation of the workforce across the various industries as shown in
Table 6. Thus the model is not useful in discovering if trade liberalization creates
more than destroys jobs in the economy. Therefore, an examination of other
factors is important; the rest of the discussion looks at the literature on this topic.
5.1. Trade liberalization and employment: empirical assessments
The empirical relationship between trade liberalization and employment has
been difficult to pin down. In the early 2000s, a series of case studies, done under
the auspices of the International Labor Organization, on the impact of trade
liberalization on manufacturing employment in selected emerging economies
failed to yield a clear picture [Lee 2005]. In the three Asian emerging economies
covered by the case studies (i.e. China, India, and Malaysia), the expansion of
trade led to increases in employment in manufacturing, employing greater number
of low- compared to high-skilled workers. The driver was the increased growth
of export-oriented manufacturing industries, which happened to be relatively
labor intensive compared to import-competing industries. However, jobs were not
altogether lost in the latter despite the competition with imports.
However, these positive impacts of growth on employment failed to transpire
in the Latin American countries in the sample, such as Brazil and Mexico.
Manufacturing employment has either stagnated or declined. Low-skilled workers
tended to lose jobs. The results were attributed “to unfavourable initial conditions
(e.g., extremely unequal distribution of assets), problems of macroeconomic
management and overdependence on external resources, but more work is
required to develop adequate insights” [Lee 2005:8].
5.2. Recent assessments
More recent assessments of the relationship between trade liberalization and
employment take off from the trade model that features heterogeneous firms,
differentiated products, trade costs to export [Melitz 2003], and labor matching
with equilibrium unemployment [Mortensen and Pissarides 1999].
Helpman and Itskhoki [2010] developed such a trade model for two countries.
They showed analytically that the country with lower labor market search costs
gains proportionately more than the other. Reducing trade barriers between the
two countries may raise unemployment, defined as the excess of workers seeking
work and the available job vacancies that are filled up.
230
The state of the country’s labor market institutions has the important role in
determining whether trade liberalization raises or reduces unemployment. When
labor market frictions are high, trade reforms raise the rate of unemployment, but
they lower the unemployment rate under a more flexible labor market.
5.3. Market transaction costs
A related gap in analyzing the effects of trade on employment incorporates
product market transaction costs. Allen [1991] suggested classifying transaction
costs into those related to exchange of assets and those associated with defining
and enforcing property rights. The former arise “because parties to exchanges
must find one another, communicate and exchange information” and must need
“to inspect and measure goods to be transferred, draw up contracts, consult with
lawyers or other experts and transfer title” [Stavins 1995:134]. Lack of logistics
infrastructure, communications and banking facilities, and other common
services constrains the capability of producers to take advantage of known
Vmarket opportunities.
The latter are costs related to establishing and maintaining property rights,
which are needed in organizing and keeping cooperative business relationships.
Property rights transaction costs have the potential of dampening the investments
needed to facilitate adjustments to reforms.7 Exports require investments, which
in turn depend upon the investment climate of the country.
The omission of transaction costs in market models may help explain the
deviation between the simulated effects of economic policy reforms and the
observed secondary data. An ex-post assessment of the effects of unilateral trade
reforms in the Philippines yielded results that did not meet the findings from an
analysis of the reforms using a computable general equilibrium model without
transaction costs8 [Clarete 2006].
5.4. Baseline unemployment
An interesting scenario is if the trade liberalization occurs when there is
starting labor unemployment in the economy. Suppose first that labor or export
product market transaction costs are weak. In this scenario, trade liberalization
will generate new jobs. Without transaction costs, the export industries can absorb
the unemployed by exporting to the rest of the world as much products as its
Allen [2001] regards Coase [1937] as the one who raised the importance of the transaction costs in
establishing and enforcing property rights. Interestingly, in about the same period, Hicks [1935] wrote
about “the cost of transferring assets from one form to another,” referring to the “neoclassical” meaning of
transaction costs.
8
The simulation of the ex-ante effects of trade liberalization was conducted by Habito and Cororaton
[2000] using a 50-sector computable general equilibrium of the Philippine economy.
7
231
resources permit. In this scenario, regional trade liberalization is a remedy to
labor unemployment.
However, it is very likely that the realistic scenario is one that combines
baseline unemployment of labor in an economy and significant export product and
labor market transaction costs. It is not surprising to expect that jobs may be lost
more than created because of short-run lack of capacity of the export industries
to take advantage of new market opportunities. Trade liberalization in this case
could worsen the baseline unemployment instead of alleviating it. But such is not
the only scenario. If transaction costs are weaker compared to the willingness of
workers to accept a lower wage just so that they can get employment, then trade
liberalization can generate new jobs.
6. Concluding observations and policy recommendations
The economy-wide effects of the goods trade liberalization in asean on the
Philippine economy are computed using the gtap model of the world economy.
Goods trade liberalization is a key reform made by asean leaders under the
asean Economic Community, comprising the free trade area tariff reforms,
prohibition of non-tariff trade barriers, and trade facilitation. Altogether, 24
of the 40 industries representing the Philippine economy reduce production.
The economic resources they lost went to the 15 industries that expanded their
respective production. Notwithstanding the contraction of production in 24
industries, the country comes out a net gainer in aggregate output. With a base
value of aggregate output at P11.3 trillion, the country comes out of this tariff
liberalization process as gaining by 1.41 percent. Exports and imports both
increase in value, but there is a positive trade balance as goods and services
outflow outweigh inflows.
Considerable movement of workers across industries is observed. Reductions
in skilled worker employment resulted in 31 industries, and 35 industries do the
same in the case of unskilled labor. The industries with the largest layoff rates
are agriculture-based, namely, milled rice, milled sugar, and vegetable oils.
Manufacturing industries come next. Thirteen to fifteen out of 20 manufacturing
industries let go of their workers to other industries. While all services industries
lay off skilled or low-skilled workers, the layoff rates are all less than a percent.
Despite the loss of jobs, the other industries that expanded production absorbed
the laid-off workers.
The study looked at also the important topic of how the economy adjusts
towards the expected changes as a result of goods trade liberalization in asean.
Does asean integration help the Philippines create jobs? As conventionally
structured, general equilibrium models, like the one used in this study, are full
employment models. As such, no net jobs are created or lost by trade liberalization.
One needs to adjust the model to take into account the possible unemployment in
the economy, a task yet to be done.
232
The analysis of labor market adjustment in this paper to freer trade indicates
that workers are likely to get unemployed for some time, as the export-oriented
activity tends to adjust at a slower rate than imports. The adjustment period can be
shortened with appropriate investment in adjustment facilitation programs. These
may aim to provide information to the private business sector on new market
opportunities, organizing value chains, facilitating coordinated investments
along the chain, clustering of related small and medium enterprises to boost
external economies, and providing information to employers and workers of job
opportunities, including training.
**Ateneo de Manila University
References
Allen, D. W. [1991] “What are transaction costs?”, Research in Law and
Economics 14: 1-18.
Armington, P.S. [1969] “A theory of demand for products distinguished by place
of production”, IMF Staff Papers 16(2): 179-201.
asean Secretariat [undated] “Annex 2: tariff schedules”, http://www.asean.org/
news/item/annex-2-tariff-schedules. Accessed 1 June 2015.
asean Secretariat [2007] asean Cargo Clearance and customs clearance
document. Jakarta: asean Secretariat.
asean Secretariat [1999] asean Economic Community Blueprint. Jakarta:
asean Secretariat.
asean Secretariat [2011] asean on non-trade barriers. Jakarta: asean
Secretariat.
Brockmeier, M. [2001] “A graphical exposition of the gtap model”, gtap
Technical Paper No. 08. Purdue University.
Clarete, R. [2006] “Ex-post effects of trade liberalization in the Philippines”, in
S. Laird and F. de Cordoba, eds., Coping with trade reforms: a developingcountry perspective on the wto industrial tariff negotiations. New York:
Palgrave Macmillan.
Coase, R. [1937] “The nature of the firm”, Economica 4 (November): 386-405.
De Dios, L. [2007] “Non-tariff barriers to trade in the asean priority goods
sectors”, in D. Hew Wei-Yen, ed., Brick by brick: the building of an asean
Economic Community. Singapore: Institute of Southeast Asian Studies.
Habito, C. and C. Cororaton [2000] “wto and the Philippine economy: an
empirical and analytical assessment of post-wto trade reforms in the
Philippines”, a report prepared for the usaid/Philippines agile Project.
Harrison, W.J. and K.R. Pearson [2002] gempack User Documentation: Release
233
8.0. Melbourne, Australia: Center of Policy Studies and impact Project.
Helpman, E. and O. Itskhoki [2007]. “Labor market rigidities, trade and
unemployment”, nber Working Paper 13365. Cambridge, Mass.
Hertel, T. [2001] “Global applied General Equilibrium analysis using the gtap
framework”, gtap Working Paper No. 66. Center for Global Trade Analysis,
Purdue University.
Hertel, T. [1999] Global trade analysis: modeling and applications. New York:
Cambridge University Press.
Hicks, J. [1935] “A suggestion for simplifying the theory of money”, Economica
2(1): 1-19.
Hill, H. [2003] “Industry”, in A. Balisacan and H. Hill, eds., The Philippine
economy: development, policies and challenges. Quezon City: Ateneo de
Manila University Press.
International Labor Organization (ilo) and Asian Development Bank (adb)
[2014] asean Community 2015: Managing integration for better jobs and
shared prosperity. Bangkok, Thailand.
Laird, S. and F. de Cordoba [2006] Coping with trade reforms: a developingcountry perspective on the wto industrial tariff negotiations. New York:
Palgrave Macmillan.
Lee, E. [2005] “Trade liberalization and employment”, DESA Working Paper No.
5 st/esa/2005/dwp/5. http://www.un.org/esa/desa/papers.
Levinsohn, J. [1999] “Employment responses to international liberalization in
Chile”, Journal of International Economies 47: 321-344.
Melitz, M.J. [2003] “The impact of trade on intra-industry reallocations and
aggregate industry productivity”, Econometrica 71(6): 1695-1725.
Mesquita, M. and S. Najberg [2000] “Trade liberalization in Brazil: creating or
exporting jobs?”, Journal of Development Studies 30(3) February: 78-100.
Mortensen, D. T. and C.A. Pissarides [1998] “Technological progress, job creation
and job destruction”, Review of Economic Dynamics 1(4): 733-753.
Narayanan, B.G., A. Aguiar, and R. McDougall [2012] Global trade, assistance
and production: the gtap 8 database. Purdue: Center for Global Trade
Analysis Project, Purdue University.
Rutherford, T. [2005] “gtap6 in gams: the dataset and static model,” prepared for
workshop on applied General Equilibrium modeling for trade policy analysis
in Russia and the CIS, Moscow, December 1-9.
Soloaga, I. and L. A. Winters [2001] “Regionalism in the nineties: what effect on
trade?”, The North American Journal of Economics and Finance 12(1): 1-29.
Stavins, R.N. [1995] “Transaction costs and tradable permits”, Journal of
Environmental Economics and Management 29(2): 133-148.
World Trade Organization [2011] World trade report 2011. Geneva: World Trade
Organization.
PRE
Just how good is unemployment as a measure
of welfare? A note
Emmanuel S. de Dios* and Katrina Dinglasan*
Governments are rightly concerned with employment generation
to make growth inclusive. The use of the open unemployment
rate to measure success, however, may be misplaced. In a
developing country like the Philippines, with a large informal
sector and in the absence of unemployment insurance, open
unemployment is primarily a middle-class phenomenon: the
unemployed are not predominantly poor, and the poor are not
predominantly unemployed. Measures of productivity and shifts
of labor across sectors may contain more information and be more
welfare-relevant.
JEL classification: J21, I32, O15
Keywords: unemployment, underemployment, labor force, welfare, poverty,
development economics
1. Introduction
The Philippine government, like many others, has placed employment
generation at the center of its objective of “inclusive growth” [Philippine
Development Plan 2010]. Employment generation is also regarded as the principal
tool for halving poverty incidence—as committed under the earlier Millennium
Development Goals. This has naturally focused the attention of policy-makers
and the public on the unemployment rate—the headcount of the unemployed
as a proportion of the labor force—as a measure of success or failure of the
government’s performance in achieving inclusive growth.
This note, however, cautions against an uncritical use of the unemployment
rate as a measure of welfare or of inclusion. The reason is that in the specific
conditions of a developing country—particularly one with a large informal sector
and a poorly developed social insurance system—unemployment correlates only
very poorly with poverty. The proposition may be put most starkly as follows:
235
most of the unemployed are not poor; and most of the are not unemployed.
None of this is new: what is merely curious is how hard-won lessons from the
earlier period of development economics have been so casually forgotten.1 Writing
almost forty years ago, Gunnar Myrdal [1968:961-1027], among others, took
great pains to question the uncritical appropriation of unemployment concepts
derived from developed-country experience—or their use as a significant guide
to policy in less developed economies. In industrial-country contexts, he argued,
open unemployment represented a readily available labor reserve that could be
readily mobilized through the management of aggregate demand. In developing,
and especially agricultural countries, by contrast,
...[T]he readily available labor supply [i.e., that which is
measured by open unemployment—esd] represents only a very
small proportion of the real waste of labor. A massive waste of
labor—whether because labor is not utilized at all, or is utilized
for only parts of the year, month, week, and day, or is utilized in
an almost useless way, that is at a low level of productivity—is
one of the obvious facts of economic life in the region. In the
present context, the important point is that little of this slack in
the labor force can be taken up by turning on the tap of aggregate
demand. Underutilization of labor vastly exceeds the supply
that could be mobilized by expansion in monetary demand
[Myrdal 1968:999].
Myrdal drew the conclusion that open unemployment rates understated the true
extent of the problem in many economies where agriculture predominates, and
moreover that real labor underutilization was unlikely to be addressable through
short-term Keynesian aggregate measures: “...‘[F]ull employment’ is a distant
goal and not one which can be reached. The limited scope of organized markets,
among other things, makes aggregative measurement of the underutilization of
labor far less possible” [Myrdal 1968:1001]. Open unemployment rates, in short,
are poor welfare measures for many less-developed countries.
2. Most of the unemployed are not poor
The continuing relevance of critiques such as Myrdal’s can be demonstrated
empirically. We proceed by using the merged files of the (Philippine Statistics
Part of this may be due to the decline in interest in development issues per se among (especially North
American) academic departments, as Krugman later [1994] noted. Hal Hill, with other fellow scholars of
the region, has been among the faithful remnants who stayed the course until the ultimate vindication of
development economics.
1
236
de Dios and Dinglasan: Just how good is unemployment
as a measure of welfare? A note
Authority, psa2) Labor Force Survey and the Family Income and Expenditure
Survey of 2009 to examine the poverty status of the various sections of the labor
force L, namely, the unemployed U and the employed N, with the latter consisting
of the fully employed F and the underemployed D. We have L = U + N = U + (F +
D). The most salient results are shown in Table 1.
TABLE 1. Poverty in the labor force, 2009
Poverty
Incidence (%)
Number of poor
Share in poor
population %
Unemployed
17.01
485,009
3.72
Employed
22.80
8,202,347
62.92
Fully employed
19.37
5,511,609
42.28
Underemployed
35.76
2,690,738
20.64
Labor force
22.38
8,687,356
66.64
Not in the labor force
20.35
4,348,001
33.36
Total
21.66
13,035,357
100.00
Of whom:
Source: Computed from PSA 2009 Labor Force Survey and Family Income and Expenditure Survey
It will be immediately evident from Table 1 that poverty incidence is actually
lowest among the unemployed. Of some 2.85 million unemployed persons in
2009, only 17 percent—less than half a million—were classified as being poor.
This should be compared with the much higher poverty incidence of 36 percent
among the underemployed—who, it should be remembered, are among those
regarded as already employed. Indeed, compared to the unemployed, poverty was
even slightly higher among those who were fully employed (19 percent). Poverty
among the unemployed was also significantly less than the national average in
that year, namely, 22 percent.
Therefore, in terms of the simplest welfare measure—poverty incidence—the
unemployed are paradoxically the best-off group in the population; somewhat
worse is the situation of the fully employed, followed closely by people not in the
labor force. By far, the worst-off are the underemployed.
The other half of the statement is also true, namely, the majority of poor people
in the country are not among the unemployed but rather among the employed.
This is also seen in Table 1, which shows that of the 13 million persons officially
classified as poor in 2009, less than four percent were unemployed. Most of the
poor are in fact employed—indeed 42 percent of them are even fully employed,
while 21 percent are underemployed.
Formerly the National Statistics Office
2
237
3. Unemployment is mainly a middle-class phenomenon
Table 2 shows rates of unemployment among various income groups of the
population, ranging from the poorest twenty percent (Quintile 1) to the richest
(Quintile 5). Open unemployment is lowest among the poorest fifth of the
population, where it is only 5.1 percent. It then rises steeply to between 7 and 9
percent among the middle classes (Quintiles 2-4) before dropping slightly among
the richest. As a result, more than two-thirds of all the unemployed are from the
second to fourth quintiles, while only 15 percent of the unemployed are from the
poorest 20 percent of the population.
TABLE 2. Unemployment across income quintiles, 2009
Unemployment rate (%)
Share of unemployed (%)
Q1
Q2
Q3
Q4
5.1
7.2
8.5
9.1
Q5
6.9
14.9
20.1
23.2
24.0
17.7
Source: Computed from NSO data
(Q1 = poorest income quintile; Q5 = richest income quintile)
The same conclusion is drawn when one looks at educational attainment among
the unemployed (Table 3). Almost half of the employed have not completed a
secondary education.
TABLE 3. Educational attainment among the unemployed and the employed, 2009
Unemployed
Employed
Share difference
No education
0.64
1.82
1.17
Incomplete primary
7.26
15.74
8.48
Complete primary
7.60
15.55
7.95
Incomplete secondary
13.81
13.74
(0.06)
Complete secondary
33.09
25.76
(7.33)
Incomplete college
19.09
13.00
(6.09)
Complete college
18.49
14.21
(4.28)
0.02
0.18
0.16
Complete postgrad
Source: Computed from PSA 2009 Labor Force Survey and Family Income and Expenditure Survey
The bottom line is that unemployment in the Philippine case is primarily
a problem of the middle class. It is a phenomenon that is bound to assume
increasing social significance as the country progresses. For the present, however,
it merely implies that fighting poverty and battling open unemployment are two
different things.
The weak correlation between poverty and unemployment will surprise some,
since it flies against mental pictures formed in the context of industrial economies.
238
For the u.s., for example, a table similar to Table 1 can be computed. Table 4
shows the large difference in poverty incidence as between the unemployed (28
percent) and among the employed (7 percent). The same table also shows the
large gap in unemployment rates as between the poor and the nonpoor in the labor
force (i.e., 26 percent and 6 percent, respectively).
TABLE 4. Poverty incidence and employment status in the u.s., 2012
(in thousands)
Poor
All persons
Nonpoor
Total
Poverty
incidence (%)
46,496
264,152
310,648
15.0
Unemployed
3,367
8,802
12,169
27.7
Employed
9,587
133,006
142,593
6.7
12,954
141,808
154,762
8.4
26.0
6.2
7.9
Labor force
Memorandum (%) Unemployment rate
Source: Computed from the (U.S.) Current Population Survey 2012.
Notes: “Unemployed” includes those who have just been laid off and those who are looking for work;
“employed” includes those at work and those who have a job but are not at work; the labor force excludes
those in the military.
The close association is further seen when one relates the period of
unemployment with poverty. Poverty incidence was only 2.9 percent among
full-time workers, but it was 16.6 percent among those who worked less than
a full-time year [Nichols 2013]. Household evidence in the u.s. also shows
poverty incidence rising with longer spells of unemployment. u.s. data for 2010
show that poverty incidence was 13 percent among people who experienced no
unemployment, but it was 19 percent among those unemployed for 1-28 weeks
and as high as 30 percent for those unemployed for 27 weeks or more [Nichols
and Callan 2013]. Econometrically, Hoynes, Page, and Stevens [2006] find the
unemployment rate to be one of the labor-market opportunity variables that affect
the incidence of poverty at the aggregate level. By contrast, no such relationship
has, to our knowledge, been established in the Philippines. What appears to have
been established instead is a relationship between measures of unemployment
and subjective measures of household satisfaction with government performance
[Mapa et al. 2013]. This actually jibes with our interpretation of unemployment
as a middle-class phenomenon.3 It is, after all, the middle class that performs
a vital role of forming and influencing national political opinion (e.g., through
As an unemployment variable, Mapa et al. [2013] use self-reported “joblessness” as found in the public
opinion polls of the Social Weather Stations. This differs from the official definition in some respects,
notably the reference being to current idleness rather than to a reference week. Like the official statistic,
however, there is “no outstanding correlation” between those who self-report as poor and the self-reported
jobless (Personal communication with Mahar Mangahas).
3
239
media). One should not be surprised, therefore, if a phenomenon affecting them
(unemployment) should figure in the more general opinions of government, even
if it does not necessarily affect the greater majority of the poor.
The reason for the discrepancy in the welfare significance of unemployment as
between poor and rich countries is as follows: Under standard statistical definitions,
being unemployed requires one not to have worked even a single hour during the
past week, to have actively sought work, and to be available for work. In richer
societies, unemployment insurance, welfare benefits, and other transfers typically
kick in when one is out of work. Such a system allows unemployed people to devote
time to job search and still sustain themselves. Since unemployment and welfare
benefits typically pay less than the average wage, people who are unemployed are
counted close to or even below the poverty line. Finally, the fact that part-time jobs
may pay less than welfare and unemployment benefits discourages unemployed
people from accepting such jobs (possibly forfeiting or reducing their benefits)
and tend to keep them fully unemployed. This explains the closer relationship
between unemployment and poverty in those cases.
In the Philippines, as in many other poor countries, however, two things
stand out: (a) there is no system of unemployment or welfare benefits; and (b) an
informal sector exists which is easy to enter and exit owing to low skill demands
and low productivity. The first removes the feasibility for the poor to devote
themselves to full-time job search, since there is no means to support themselves
in the process. At the same time, a large informal sector beckons that is easy to
enter and to exit. Easy entry into low-productivity, low-wage jobs will suffice to
remove one from the ranks of the unemployed, but will hardly ameliorate poverty.
As the old development adage goes, “The poor cannot afford to be unemployed.”
Indeed, the fact of their employment is a sign not of improvement in their welfare,
but of their lack of choice.
By contrast, it is people who are better able to support themselves through
a spell of job search who will be found among the openly unemployed. These
will be those who can rely on personal savings, or who come from families with
sufficient means, who have better access to social networks, or people with some
education and who, therefore, have better job prospects—or all of these—in short
the middle class. For this reason, an unemployed person is more than 80 percent
likely to be non-poor.
4. The poverty impact of falling unemployment
How is a change in the unemployment rate related to a change in poverty
incidence? An answer in purely accounting terms can be provided as follows. Let
P be the number of poor persons in the labor force L and p = P/L the (headcount)
poverty incidence in that category. Then, using the fact that L = U + N = U + (F
+ D), where U, N, F, and D are defined as before, and letting Pk , k = U, F, D,
be the poverty headcounts among the unemployed, the fully employed, and the
240
underemployed, respectively, we obtain:
p = (PU + PF + PD) / L
= (αUU + αFF + αDD) / L
= αU(U/L) + αF(F/L) + αD(D/L)(1)
where the αk, are rates of poverty incidence among k = U, F, D. We note that the
unemployment rate u = U/L = (1 – N/L) so that N/L = (1 – u); F/L = (N/L) – (D/L);
and the underemployment rate d = D/N, so that D/L = (D/N)(N/L) = d(1 – u).
Substituting these into the last identity of (1) above yields:
p = αUu + αF(1 – u) – αF (1 – u) d + αD(1 – u) d
= αF + u (αU – αF) + (1 – u) d (αD – αF )(2)
This last expression relates overall poverty incidence in the labor force with
poverty incidence in its various categories. The association between poverty
incidence and a change in the unemployment rate can then be approximated as.
∂p/∂u = (αU – αF) – d(αD – αF )(3)
If using Table 1 we substitute into (3) the values αU = 0.17, αF = 0.19, αD =
0.34, and d = 0.19 we obtain a value of: –0.00326. This is remarkable not only
for its small magnitude but more importantly its sign. It suggests not only that an
increase in the unemployment rate has little effect on poverty, but that a higher
unemployment rate might indeed even improve it!
It is plainly wrong, of course, to interpret this to mean that poverty incidence
could actually be reduced by increasing the rate of unemployment. It merely
reflects the accounting identity that (with a fixed labor force) the ranks of the
unemployed can fall only by drawing away from the employed. Given the
existing rates of poverty, however, the random unemployed person is even less
likely to be poor than her employed counterpart; so a move from unemployment
to employment can be an ambiguous matter.
More constructively, one might look for conditions under which the expression
in (3) is positive—i.e., where an increase (decrease) in unemployment is likely to
increase (reduce) poverty. The sufficient condition is given by
αU > (1 – d) αF + dαD (4)
As is readily evident, this says that poverty among the unemployed must be
worse than average poverty among the employed (the weight being represented by
d): a reduction in unemployment is more likely to reduce poverty if poverty among
241
the fully employed is far less among the unemployed and the underemployed, and
the lower is the rate of underemployment.
The pathology of the Philippine case (which may also be true of other
developing countries), however, is that αF > aU, αD > αF so that condition (4)
is impossible to fulfill for any d in the interval [0, 1]. This explains the
perverse result.
At any rate, this simple exercise does focus attention on the key problem,
which is the poverty incidence among those who are employed, particularly those
who are fully employed. Somewhat paradoxically, in order for unemployment
reduction to imply poverty reduction, poverty must be reduced among those who
are already employed.
5. Poverty among the employed
For more detail, we can disaggregate the employed by sector as well as by
their poverty status (Table 5). Most of the employed poor can be found in sectors
where informal employment relations predominate and which are notorious for
low-productivity jobs. The most prominent is agriculture, which alone already
accounts for almost two-thirds of the employed poor. Other sectors that serve as
major collecting pools for the employed poor are wholesale and retail trade (think
vendors and hawkers); private household services (e.g., domestic help); informal
sector manufacturing (e.g., sweatshops and small household businesses); and
transport (e.g., jeepney drivers, tricycles, kuliglig, and pedicabs).
Agriculture is also the sector with the highest incidence of poverty (44
percent) among those it employs. Poverty among people engaged in the mining
sector is also extremely high (42 percent), although the poor in that industry
are only a small percentage of the total poor. The high incidence of poverty in
mining doubtless also reflects the desperate conditions of the informal mining
sector, as exemplified by the small-scale mining operations in Compostela
Valley. This example also illustrates the duality of conditions existing in many
important economic sectors. There will in many cases be a wide gulf in scale,
skills, productivity, and pay as between informal and formal employment even
in the same sector: e.g., high- v. low- productivity manufacturing; high- v. lowproductivity services; high and low productivity mining; and so on. As a result,
simple classification of the employed according to industries will not be a reliable
guide to their welfare status. Small exceptions to this are sectors such as finance,
education, and utilities, where poverty is low in both incidence and extent.
242
TABLE 5. Poverty among the employed:
distribution and incidence by sector, 2009
Distribution (%)
Incidence (%)
Agriculture and fishing
63.7
44.3
Mining and quarrying
1.0
41.5
Manufacturing
5.1
13.8
Electricity, gas, water
0.1
2.7
Construction
4.6
19.1
Wholesale and retail services
10.1
11.7
Hotels
0.9
6.8
Transport
4.5
13.7
Financial
0.0
1.0
Real estate
0.4
3.1
Public administration
2.1
9.4
Education
0.2
1.6
Health and social services
0.2
4.1
Other community services
1.8
15.5
Private households
5.2
20.1
Source: Computed from NSO data
6. The takeaway for policy
The foregoing has merely sought to demonstrate how employment status
can be a poor guide to policy. For government (and its critics) to use open
unemployment—especially by itself—as a measure of failure or success is to
completely miss the mark and underestimate the development task at hand. An
undue focus on unemployment could induce policy-makers, for example, to
mistakenly engage in large-scale emergency job-creation schemes financed by
public spending. Such stopgap schemes are likely to have adverse budgetary
consequences without making a real dent on poverty, since all they would do is
transfer people who are already employed in low-productivity jobs to similar lowproductivity jobs—except underwritten now by government.
It has been suggested that perhaps the extent of unemployment taken together
with underemployment might provide a better measure for policy makers to track.
What has been called a “job misery index” (see, e.g., Mapa et al. [2013]) takes
the unemployed and underemployed together as a proportion of the labor force.4
Such a statistic is an improvement over the simple unemployment rate, especially
considering how poverty is markedly higher among the underemployed. But it
A convenient expression involving only rates for computing (U + D)/L is u + (1 – u)d, where u and d are
the unemployment and underemployment rates, respectively.
4
243
unfortunately still falls short of the mark. First, the unemployed are markedly
different as a group from the underemployed, so that adding the two is something
of a statistical pastiche, since the former are predominantly from the middle class,
while more of the underemployed are from the poor. Second, even the job misery
index neglects the fact that far more of the poor are to be found among the fully
employed. If job “misery” and dissatisfaction plague even the fully employed,
then real job misery would have to include virtually the entire labor force, which
threatens to render the concept meaningless.
The mismeasure can impart a wrong sense of the scale of the problem of
employment and its relation to poverty. A recent World Bank development report,
for example, rightly focuses on the problem of providing “good jobs—meaning
jobs that raise real wages and bring people out of poverty”. But its assessment
of the scale of the task is hampered by an inability to sort out the most crucial
welfare aspects of the problem. It defines the “jobs challenge” as one of providing
jobs to “around 10 million Filipinos who were either unemployed (three million)
or underemployed (seven million) in 2012, and to around 1.15 million potential
entrants to the labor force every year…In addition, better jobs need to be provided
to another 21 million Filipinos who are informally employed. All in all informal
workers comprise about 75 percent of total employment” [World Bank 2013:5].
(Emphasis supplied).
As already demonstrated, to regard unemployment plus underemployment as
the target is certainly too narrow. But to lump all the unemployed plus all the
informally unemployed as the problem is also certainly too broad, since that would
comprise some 77 percent of the entire labor force.5 This certainly exaggerates
the welfare problem, since poverty in the entire labor force is no more than 22
percent (Table 1). The result is that no clear focus is achieved.
More importantly, the policies required to address unemployment are
vastly different from those needed to solve low-productivity employment, so
that lumping the two together makes little sense. The former requires mainly
improving the workings of labor markets and the matching of expectations as
between qualified job seekers and employers—so physical and virtual job fairs,
information given to parents and students regarding career options, measures
facilitating labor mobility, and perhaps temporary unemployment benefits for
people between jobs are effective policies to lower the open unemployment rate.
But these measures are obviously unlikely to reduce poverty. Solving the
poverty problem ultimately means raising the productivity and incomes of people
who are already employed. Again it is worth noting how Myrdal’s early lucidity on
Here we accept the World Bank’s estimate that 75 percent of the employed are in the informal sector
(which, it is important to note, does not necessarily make them poor). If unemployment and employment
rates are approximately 0.93 and 0.07, respectively, then the informally employed plus the unemployed are
0.77 (= (0.75)(0.93) + 0.07) as a proportion of the labor force, as stated in the main text.
5
244
these issues has been forgotten. He noted that productivity in the labor force could
be broken into three components: participation, duration, and efficiency. Letting
L, N, H, and Q represent the labor force, employment, hours worked, and output,
respectively, one has Q/L = (N/L)(H/N)(Q/H). “Broadly speaking,” he wrote, “the
modern approach has been preoccupied with the first of these ratios, and then only
in a partial and unrealistically biased way, whereas all three are essential to an
understanding of labor utilization in South Asia” [Myrdal 1968:1016]. Myrdal’s
complaint was that the concern for unemployment related only to N/L, i.e., its
complement (1 – N/L). Since his time, statistical authorities have generated data,
albeit partially, relating to H/N in the form of visible underemployment rates. It is
still the case, however, that no labor statistic in developing countries captures the
efficiency of hours worked, i.e., Q/H.
Yet ultimately the quality or efficiency of employment matters. Either people
must attain higher productivity in their current employment, or they must transfer
to higher-productivity sectors. This means, for example, increasing productivity
in agriculture through higher private and public investments in that sector; the
infusion of new entrepreneurship and the linking of small farm operators into
higher value-added chains; extension, training, and education for small farmers
and their families; and the gradual movement away from agriculture and fisheries
into manufacturing and better service-sector jobs. (Always remembering of
course that there are also low-productivity jobs in manufacturing and services.)
There is no direct reason the effect of such measures will be reflected in the
unemployment rate; but they are more substantial and more relevant to welfare
nonetheless.
This paper is offered to Hal C. Hill, development economist, steadfast friend, and patronus
vini to the U.P. School of Economics. Support from the Philippine Center for Economic
Development is gratefully acknowledged.
References
Hoynes, H., M. Page, and A. Stevens [2006] “Poverty in America: trends and
explanations”, Journal of Economic Perspectives 20(1): 47-68.
Krugman, P. [1994] “The fall and rise of development economics”, Available
from: http://web.mit.edu/krugman/www/dishpan.html. Accessed September
2014.
Mapa, D., S. Dineros, K. Flores, and C. Japlit [2013] “The link between job
misery index and the net satisfaction rating of the President: evidence from
household surveys” (Paper presented at the National Conference on Statistics)
245
Myrdal, G. [1968] Asian drama: an inquiry into the poverty of nations (3 vols.).
Harmondsworth: Penguin Press.
Nichols, A. [2013] “Poverty in America”, Urban Institute Fact Sheet. Available
from: http://www.urban.org/publications/412898.html. Accessed September
2013.
Nichols, A. and T. Callan [2013] “Unemployment and poverty”, Urban Institute
Fact Sheet. Available from: http://www.urban.org/publications/412400.html.
Accessed September 2013.
World Bank [2013] Philippine development report: creating more and better jobs.
Philippine Office. Asia Pacific Region.

December 2015 Volume LII No. 2

Transcription

Similar documents

Anna Marie Periquet Profile JL

The Philippines: A New Awakening

Memorandum

Sure-Win Dip

STEM Human Resources - British Council | Philippines

Converging Texts: Teaching Culture throughTranslaVon and

Higher profitability and increased customer satisfaction!

immediate press release 11th aug`2014

Newsletter (December 2015) - Philippine Charity Sweepstakes Office

ASEAN`s Dispute Resolution Mechanisms