Why do Americans get their Impalas from Canada? Classic

4/3/12!
Classic American car.!
Why do Americans get their Impalas from
Canada?
2008 Impala.!
Photo by Brett Weinstein, published under a Creative
Commons license.!
! 
All Impalas are now made at Oshawa, Ontario.
! 
Roughly 200,000 imported into the US per year.
! 
All Cobalts are made in Lordstown, Ohio
(similar but smaller Chevrolet).
For good measure.....!
1!
4/3/12!
! 
! 
! 
! 
! 
! 
! 
! 
! 
! 
! 
! 
Comparative advantage?
Requires Canada to have comparative
advantage in Impalas, US to have one in
Cobalts.
! 
Result: Big 3 automakers mostly produced in
Canada for Canadian market.
Little trade; exports from Canada virtually
zero.
Duplicated assembly lines on both sides of
the border.
Alternative: Two factors:
!  I. Canada-US Auto Pact of 1965
!  II. Increasing returns to scale
Let’s list some reasons a country might have a
comparative advantage in something, and see
if that is reasonable.
Before 1965, both countries had high tariffs
on imported cars and auto parts.
Comparative advantage doesn’t seem to
explain US imports of Impalas.
! 
Pact provided for free trade between US and
Canada.
! 
(Some exceptions: e.g., fire engines.)
! 
Some restrictions applied:
!  US side: Minimum Canadian content required for
duty-free status.
!  Canadian side: Employment at plants in Canada
needed to keep on increasing.
Production on smaller scale in Canada: High
cost per unit.
Free trade in autos and auto parts between
US and Canada.
But Detroit had to maintain the same level of
production in Canada.
Agreement was grandfathered into CUFTA
(1988) and NAFTA (1994).
2!
4/3/12!
! 
! 
! 
Setting up and maintaining an assembly line
for one model requires huge fixed costs.
E.g., even just to maintain the machines in
line to be able to produce 1 car per month.
Induces increasing returns to scale (IRS).
! 
! 
! 
! 
! 
! 
Need to maintain same level of production in
Canada
! 
But can save on costs by reducing number of
models produced in each country.
! 
! 
! 
! 
! 
! 
Then cost of producing 2Q Impalas at
Oshawa is equal to F + 2waQ.
LESS THAN TWICE.
Essence of IRS: Double output, less than
double cost.
Suppose that GM needs to produce 11
models.
Each model:
!  200,000 units for the US market,
!  20,000 units for the Canadian market.
! 
! 
Suppose cost of producing Q Impalas at
Oshawa is equal to C(Q) = F + waQ.
Assume that in either country, production
costs are given by F + waQ for each model.
Pre-Auto Pact.
11 models produced in Canada; 20,000 units
each.
! 
! 
Cost for each model = F + wa20,000.
Post Auto Pact.
Now, GM can concentrate production of each
model in one location.
11 models produce in US; 200,000 units
each.
! 
Produce 1 model in Canada, 10 in the US.
! 
Produce 220,000 units at each plant.
Cost for each model = F + wa200,000.
! 
Total cost for GM: 22F + wa11x220,000
= 22F + wa2,420,000.
Now costs are equal to 11x(F + wa220,000)
= 11F + wa2,420,000.
3!
4/3/12!
! 
! 
Thus, GM has saved 11F.
Same number of each type of car produced,
but costs are lower.
! 
Note: Before, there was no trade in cars.
! 
Now, every car is traded.
! 
! 
200,000 cars exported from the US to
Canada.
Intra-industry trade.
! 
Internal IRS.
Double the firm’s inputs and more than
double that firm’s output.
! 
Can result from fixed cost.
! 
Can also result from learning by doing.
! 
Implies a downward-sloping AC curve.
! 
Note: Incompatible with perfect competition.
! 
! 
IRS provides a reason for trade, by creating
an incentive to concentrate production of
each product in one location.
200,000 cars exported from Canada to the
US.
! 
! 
! 
! 
! 
! 
! 
External, international IRS.
Case when external, national IRS are not
present, but if you double the inputs of all
producers in the industry worldwide it will
more than double worldwide output.
External, national IRS.
Double all inputs employed by all firms in an
industry in one country, and more than
double total industry output.
Can result from learning-by-doing
spillovers.
Can also result from shared infrastructure
improvements.
! 
I. How to tackle a foreign market.
! 
II. Role of monopolistic competition in trade.
! 
III. Intra-industry trade.
4!
4/3/12!
! 
You are the CEO of GM.
! 
You want to break into the European market.
! 
Two options: Produce here and export, or
produce over there.
! 
Set up a plant in Spain.
! 
Fixed cost: F.
! 
After that, each unit requires a* units of
labor.
! 
Each unit of labor costs w*.
! 
Choose P to maximize (P - a*w*)Q(P) - F.
! 
This yields maximum profit from FDI option.
! 
FDI option is more likely to be attractive if:
! 
No additional fixed cost. (Important.)
! 
Each unit requires a units of labor.
! 
Labor costs w per unit.
!  t is high;
Transport cost of k(d) per car, where d is
distance to market.
!  F is small.
! 
! 
Tariff of t per car.
! 
Choose P to maximize (P-wa-k(d)-t)Q(P).
! 
! 
! 
! 
Brainard (1997) study examined US exports
and foreign sales of foreign affiliates of US
multinationals.
Key variable: Export share of total foreign
sales of each industry.
Found pretty good evidence for the last three
of these points.
(Didn’t really look at the first one; that’s
harder.)
!  w*a* is small;
!  d is high;
! 
How does GM actually serve Europe?
! 
Mostly through OPEL subsidiary (FDI option).
! 
Cars are made in Europe.
! 
! 
BUT: Each model is made in only one location
in Europe (e.g., Zaragosa, Spain).
Thus, intercontinentally -- FDI option; within
Europe, export option.
5!
4/3/12!
! 
! 
! 
! 
IRS matters not only for giant firms like GM.
Pervasive in manufacturing, including smallscale manufacturing.
E.g., furniture.
A pretty good description of this type of
industry: Monopolistic competition.
Baronet Java dining set.!
! 
! 
! 
! 
! 
Two medium-sized firms.
Each has a tiny share of the total furniture
market.
! 
Baronet is Canadian; Moser is American.
! 
Fixed cost from production and design.
! 
Distinctive styles.
Thos. Moser Hawthorne dining set.!
Large number of firms; each small compared
to the whole market.
Each produces a unique product; hence,
monopoly power.
Free entry, hence zero profits in equilibrium.
6!
4/3/12!
! 
! 
! 
! 
! 
! 
! 
Opening up trade flattens Baronet’s demand
curve.
Raising its price will send some of its
customers to a US competitor;
Lowering its price will grab some customers
from US competitors.
Thus, each firm’s demand is more elastic.
Thus, Baronet now has an incentive to lower
its price and sell more dining sets.
But at the same time all other firms have the
same incentive.
All other furniture makers therefore cut their
prices, shifting Baronet’s demand curve
down.
7!
4/3/12!
! 
! 
! 
! 
! 
! 
Plenty of trade even between identical
countries.
Once again, trade results from IRS.
Trade is intra-industry.
Trade reduces number of products produced
in each country (“shake-out”).
But increases variety available to each
consumer.
In addition, trade tends to increase elasticity
of demand for each product: P closer to MC.
Exports of k from country i to country j:!
! 
Important feature of trade between similar
countries where monopolistic competition is
important.
Total trade between i and j:!
Net trade, or inter-industry trade, !
in industry k:!
Net trade as a fraction of total trade:!
Intra-industry trade as a fraction of
total trade:!
! 
! 
! 
It’s actually a stretch to assume that all
furniture makers are equally productive.
Suppose there are high-cost and low-cost
producers in the same industry -heterogeneous firms?
Question was explored in an influential
paper by Marc Melitz (2003).
Suppose that to produce q units of output, a firm
must hire f + q/! units of labor.
! 
! 
The constant f is a fixed labor requirement,
and is the same for all firms.
Therefore, the fixed cost is equal to wf,
where w is the wage.
8!
4/3/12!
Suppose that to produce q units of output, a firm
must hire f + q/! units of labor.
! 
! 
! 
! 
! 
! 
The parameter ! is a constant for each firm,
but varies from firm to firm. More
productive firms have higher values of !.
! 
The marginal cost for each firm is equal to
w/!, where w is the wage.
! 
If a firm wants to export, it must pay an
additional fixed cost (e.g., setting up a
distribution network).
As a result, only the most productive firms
choose to export.
Less productive firms are hit by imports but
don’t benefit from exports.
Therefore, less productive firms produce less
and have lower profits than before trade;
some drop out.
! 
Firms enter until the profits for the marginal
firm are equal to zero.
Only the most efficient firms enter.
More productive firms (higher !) produce
more and make higher profits than less
productive firms.
More productive firms benefit from less
productive firms dropping out; their output
and profits go up.
!   Market share of less productive firms falls;
market share of more productive firms rises.
!   Average productivity of industry therefore
rises.
!   Call this the ‘Melitz effect.’
! 
! 
! 
! 
! 
! 
IRS is a reason for trade: Motive to concentrate
production of each product in one spot.
Three types of IRS: Internal, external national, and
external international.
In serving a foreign market, stronger IRS argues for
exporting, but higher tariffs or transport costs argue for
local production via FDI.
Many industries have small-scale IRS, many producers,
differentiated products: monopolistic competition.
Explains intra-industry trade.
9!
4/3/12!
! 
Finally, heterogeneous firms plus IRS yields
the Melitz effect: Trade raises industry
productivity by increasing market share of
more productive firms at the expense of less
productive firms.
10!