Author`s personal copy

Transcription

LES SERVICES ÉCOSYSTÉMIQUES RENDUS PAR LES SOLS :
ANALYSE DES RÉGIMES INCITATIFS SOUS-JACENTS
ANALYZING INCENTIVE MECHANISMS BEHIND THE PROVISION
OF ECOSYSTEM SERVICES DELIVERED BY SOILS
Gilles Grolleau
Montpellier SupAgro – LAMETA Bat. 26
2, place Pierre Viala
34060 Montpellier Cedex 1
Tel : + 33 (0)4.99.61.28.81 – Fax : + 33 (0)4 67 63 54 09
[email protected]
Rapport final
15 mai 2013
Numéro de contrat Ministère/ADEME: CV 09-00006527
Les services écosystémiques rendus par les sols : Analyse
des régimes incitatifs sous-jacents
05/2013
REMERCIEMENTS
Nous tenons à remercier de nombreuses personnes qui ont contribué à ce projet. Ces
contributions ont pris de nombreuses formes tant dans un cadre formel qu’informel et nous
tenons à exprimer notre gratitude tout en sollicitant l’indulgence de ceux qui ne trouveront
pas mention de leurs noms dans ces remerciements. En effet, penser à tous et exprimer nos
sincères remerciements au bout de 38 mois est un véritable défi en soi et génère forcément
des oublis et des omissions.
Nos remerciements s’adressent d’abord au programme GESSOL pour son soutien financier,
l’accompagnement et les échanges fructueux permis tout au long du déroulement de ce
programme. Nous tenons tout particulièrement à remercier Dominique King qui bien qu’il
nous ait quitté a été pour nous une source d’encouragements. Dresser la liste de ceux qui
nous ont aidé de manière très diverse, mais souvent avec constance, compréhension et
gentillesse est un exercice périlleux. Qu’Anne-Véronique Auzet, Marion Bardy, Insaf Bekir,
Martial Bernoux, Philippe Billet, Antonio Bispo, Eric Blanchart, Douadia Bougherara, Claire
Chenu, Sophie Clot, Patrice Coll, Thomas Eglin, Nicolas Gallaï, Camille Guellier, Sana El
Harbi, Isabelle Feix, Lisette Ibanez, Robert Lifran, Yves Le Bissonais, David Masclet, Jean
François Ouvry, Véronique Souchère, Angéla Sutan, Luc Thiébaut, Alban Thomas, Sophie
Thoyer et Marc Willinger trouvent ici l’expression de notre gratitude pour leur commentaires
et leurs suggestions. Certains ont nourri nos réflexions dans des cadres formels, pour
d’autres autour d’un café et d’une discussion à bâtons rompus. Nous tenons également à
remercier l’équipe administrative du LAMETA et les collègues de l’unité qui ont fait preuve
d’un désir sincère de nous faciliter les procédures et nous leur en sommes très
reconnaissants.
L’ADEME EN BREF
L'Agence de l'Environnement et de la Maîtrise de l'Energie (ADEME) participe à la mise en œuvre des politiques publiques
dans les domaines de l'environnement, de l'énergie et du développement durable. Afin de leur permettre de progresser dans
leur démarche environnementale, l'agence met à disposition des entreprises, des collectivités locales, des pouvoirs publics et
du grand public, ses capacités d'expertise et de conseil. Elle aide en outre au financement de projets, de la recherche à la mise
en œuvre et ce, dans les domaines suivants : la gestion des déchets, la préservation des sols, l'efficacité énergétique et les
énergies renouvelables, la qualité de l'air et la lutte contre le bruit.
L’ADEME est un établissement public sous la tutelle du ministère de l'écologie, du développement durable et de l'énergie et
du ministère de l'enseignement supérieur et de la recherche. www.ademe.fr
Et un copyright
Toute représentation ou reproduction intégrale ou partielle faite sans le consentement de l’auteur ou de ses
ayants droit ou ayants cause est illicite selon le Code de la propriété intellectuelle (art. L 122-4) et constitue
une contrefaçon réprimée par le Code pénal. Seules sont autorisées (art. 122-5) les copies ou reproductions
strictement réservées à l’usage privé de copiste et non destinées à une utilisation collective, ainsi que les
analyses et courtes citations justifiées par la caractère critique, pédagogique ou d’information de l’œuvre à
laquelle elles sont incorporées, sous réserve, toutefois, du respect des dispositions des articles L 122-10 à L
122-12 du même Code, relatives à la reproduction par reprographie.
Programme GESSOL
2/24
05/2013
RESUME
Le sol rend de nombreux services aux communautés humaines, dont la valeur est considérable.
Néanmoins, les activités humaines perturbent et affectent souvent négativement ces services en
provoquant des détériorations sur les plans quantitatif et qualitatif. Le projet ECOSOL démarre du
constat que le comportement humain est souvent à l’origine des perturbations affectant les sols et
qu’en conséquence, il est nécessaire de mieux comprendre les déterminants de ce dernier afin de
pouvoir l’influencer. Le projet vise une meilleure compréhension de ces déterminants afin d’envisager
des modalités concrètes susceptibles de maintenir, voire de renforcer la fourniture de ces services, et
ce tout en se situant dans un contexte de fortes contraintes budgétaires.
De manière très concrète, nous mobilisons l’économie néo-institutionnelle pour comprendre les
facteurs permettant l’émergence de solutions contractuelles entre les agriculteurs et les municipalités
pour la fourniture du service d’épuration et de traitement des eaux. Grâce à l’étude détaillée de deux
cas (Munich et New York), nous tirons plusieurs leçons pratiques susceptibles de faciliter l’émergence
de ces accords décentralisés comme la structure de marché monopsonique, la présence de structures
institutionnelles crédibles et la capacité à transformer les bénéficiaires de ces services rendus par les
sols en supporters et payeurs.
De plus, une partie importante du projet est consacrée aux apports de l’économie comportementale
afin d’expliquer le décalage entre les promesses des instruments des politiques visant une gestion
plus durable des sols et les résultats réels. Cette analyse repose sur un modèle du comportement
humain alternatif qui utilise des raccourcis mentaux et des processus heuristiques dans la prise de
décision, en contraste avec le modèle purement rationnel de maximisation d’une fonction objective
sous contraintes. Ce modèle de comportement permet de générer des préconisations originales en
termes de politiques.
Plusieurs biais sont ainsi étudiés en rapport avec les politiques liées aux sols. Nous explorons le rôle
des considérations de statut et de position qui interfèrent avec le seul calcul économique et qui
peuvent pousser les agents à refuser certaines innovations pourtant profitables et
environnementalement préférables à cause de leur impact négatif sur leur statut dans un certain
groupe de référence. Nous montrons également que le recours systématique à des (des)incitations
monétaires pour encourager (ou décourager) des comportements préférables peut générer un effet
d’éviction des motivations préexistantes. Pour éviter ce risque, nous proposons un mécanisme original
permettant d’utiliser au mieux les motivations hétérogènes au sein d’une population donnée.
Nous montrons également que la qualification et l’origine perçue des éventuelles incitations financières
ne sont pas neutres et peuvent significativement affecter leur capacité à générer le comportement
recherché. Nous mettons également en évidence que le soutien à une politique donnée en faveur des
sols au-delà des bénéfices escomptés peut être lié à l’origine attribuée aux perturbations et aux points
de référence. Ainsi, nous démontrons que des éléments de contexte considérés a priori comme
mineurs peuvent influencer de manière significative les performances des politiques
environnementales et constituer des leviers d’action prometteurs.
Mots clés : biais, comportement, éviction, incitations monétaires, sols, statut.
Programme GESSOL
3/24
05/2013
ABSTRACT
Soils deliver several valuable ecosystem services to human communities. Nevertheless, human
activities frequently disturb these services both qualitatively and quantitatively. The ECOSOL project
starts from the premise that the human behaviour is at the root of the problem. Consequently, a better
understanding of drivers of human behavior can usefully inform policymakers seeking to influence
agents’ behavior. The ECOSOL project aims at investigating some non-classical drivers of human
behavior in order to design policies that can reinforce the provision of soils services, especially in times
of severe resource constraints.
We investigate contractual arrangements for water quality services between municipal water
organizations and farmers. Our study uses a transaction cost framework, in conjunction with detailed
case studies of two water quality payment schemes, to examine the factors that increase and decrease
transaction costs. Factors such as the lack of rural/urban antipathy, homogeneous land use, utilization
of well-developed organic standards, and strong demand for organic products decreased the
transaction costs in the Munich case relative to the New York case. Factors that decreased transaction
costs in both cases included: some highly sensitive land was purchased outright and there was one
large “buyer”, the municipality, which interacted with farmers. Time and flexibility of water quality
regulations allowed negotiation over and development of the watershed programs.
Moreover, we use insights from behavioral economics to explain why promises of soil related policies
are not fulfilled and why results differ so much from expectations. Our analysis considers an alternative
model of human behavior that includes the use of mental shortcuts and heuristics in decision making.
This model based on the influence of contextual factors offers original recommendations regarding soil
related policies. It contrasts with the rational model, which considers rational agents who analyze
incentives offered to them and behave in ways that reflect their best interests.
We study several biases related to soil policies. We emphasize that status concerns can interfere with
economic calculus and lead agents to refuse win-win innovations because of their detrimental impact
on their status in a reference group. We also show that the systematic use of monetary incentives to
influence some behaviours can generate a crowding-out effect on preexisting motivations. To avoid or
attenuate this risk, we propose an original mechanism, which is based on the heterogeneity of
motivations among agents.
Departing from standard applications of utility theory where only outcomes matter, we show that past
states and alternative causes of an environmental problem affect respondents’ interest and
willingness-to-pay to improve it. We also argue that the qualification and perceived origin of monetary
incentives are not neutral and can significantly affect their ability to generate the desired behavior. We
emphasize that ‘minor’ contextual elements can significantly shape the performances of soil related
policies and constitute good leverages.
Key-words: biases, behavior, crowding-out, monetary incentives, soils, status.
Programme GESSOL
4/24
05/2013
SOMMAIRE
Titre
Page
Remerciements
2
Résumé
3
1. Contexte général et objectifs généraux du projet
6
2. Quelques éléments de méthodologie et éventuelles difficultés rencontrées
7
3. Des arrangements contractuels afin de pérenniser les services d’épuration des sols
8
4. L’influence sociale : les considérations de statut et de position
9
5. L’effet d’éviction entre incitations et les stratégies mobilisables
10
6. Les effets de formulation
13
61. L’effet des causes de la dégradation des sols sur le consentement à payer
13
62. L’origine des incitations
15
7. Le timing des incitations
16
8. La dynamique des actions : l’effet de compensation morale
17
9. Partenariats mis en place, projetés, envisagés
17
10. Conclusion générale
18
11. Références mobilisés dans le rapport
18
12. Liste des opérations de valorisation et de transfert issues du contrat
20
Annexes
24
Programme GESSOL
5/24
05/2013
Les services écosystémiques rendus par les sols : Analyse des
régimes incitatifs sous-jacents
‘Nous connaissons mieux la mécanique des corps
célestes que le fonctionnement du sol sous nos pieds’
(Leonard de Vinci, 1452-1519)
1. Contexte général et objectifs généraux du projet
Les sols rendent de nombreux services écosystémiques dont la valeur est estimée de quelques
1
dizaines d’euros à plusieurs milliers d’euros par hectare , et ce souvent à l’insu des bénéficiaires et
même des fournisseurs. Certains de ces services semblent évidents (e.g., support de la production
agricole, stockage de déchets) alors que d’autres passent souvent inaperçus (e.g. réservoir de
biodiversité, épuration de l’eau, stockage de carbone). Par exemple, selon une étude commanditée
par le Ministère de l’environnement irlandais, les plus gros contributeurs naturels à la richesse
nationale seraient les vers de terre avec un apport estimé à 700 millions d'euros par an, notamment
grâce à leurs rôles dans le recyclage des déchets végétaux, l’aération des sols et la fertilisation de ces
derniers du fait de leurs déjections (Bullock et al., 2008). La perte, la diminution ou la détérioration de
la qualité de ces services et la nécessité de leur restauration voire de leur remplacement par des
solutions de substitution intensives en capital permettent parfois de se rendre compte de leur
importance pour les différentes parties concernées. Des rapports alarmants s’inquiètent de la
dégradation et des menaces pesant sur de nombreux sols du fait des activités humaines et ce dans
des proportions largement supérieures à ce qu’autorisent les cycles naturels de formation des sols ou
de réhabilitation de ces derniers. L’intuition sous-jacente à l’ensemble de ce projet est de considérer
que si l’être humain est à l’origine des menaces à travers son comportement, le changement (durable)
de son comportement pourrait aussi constituer une solution. Dès lors, une meilleure compréhension
des déterminants du comportement humain peut déboucher sur des politiques et stratégies originales
susceptibles de contribuer à modifier ce dernier dans un sens favorable aux intérêts de la société,
c’est à dire en faveur d’une gestion plus durable du capital sol.
Le recours à des solutions technologiques couplées à des incitations économiques appropriées a
souvent été préconisé afin de limiter les perturbations ou de compenser les déficiences des services
écosystémiques. Par exemple, les structures de traitement ont souvent été utilisées pour remplacer ou
compléter le service d’épuration rendue par les sols. L’économie traditionnelle considère que les
agents sont parfaitement rationnels (‘Econs’ de Sunstein et Thaler, 2008) et que suite à une analyse
exhaustive de l’information disponible et des incitations proposées, ils adopteront le comportement qui
correspond le mieux à leurs intérêts. Cette approche avec ses raffinements récents (e.g., économie
néo-institutionnelle, économie de l’information) a ainsi permis des avancées remarquables. Dans cette
direction, nous avons analysé certains problèmes liés à la dégradation des services écosystémiques
rendus par les sols comme des conflits entre différents groupes d’intérêt portant sur des usages
mutuellement incompatible de la même ressource. Nous avons ainsi étudié les conditions ayant permis
l’émergence de solutions contractuelles pour la fourniture du service d’épuration de l’eau par les sols
avoisinant les villes de Munich et de New-York et Munich afin d’identifier les déterminants des coûts de
transaction et les stratégies mises en place afin de rendre les coûts de transaction compatibles avec
un arrangement décentralisé.
L’homogénéité offerte par un modèle d’explication du comportement quasi-universel a eu pour
contrepartie une certaine ignorance d’autres paramètres notamment psychologiques, qui influencent
pourtant les décisions des agents. Tout en reconnaissant les apports considérables permis par ce
modèle rationnel, nous avons pris le parti de consacrer une partie importante des recherches réalisées
à un autre modèle de comportement qui reconnaît l’importance de l’interprétation d’une situation et du
contexte. L’économie comportementale étudie comment les agents économiques (‘Humans’ dans la
terminologie de Sunstein et Thaler, 2008) prennent leurs décisions dans la réalité, notamment en
utilisant les apports de la psychologie. Plutôt que de postuler une unique forme de rationalité des
agents, comme la maximisation d’une fonction objective sous contrainte, les économistes
1
Ces évaluations sont souvent controversées notamment du fait de l’incommensurabilité des objets concernés,
comme l’existence d’une espèce. Il y a une sorte de répugnance morale quant à la question d’attribuer un prix à
la nature (Bourgeois-Gironde et al., 2011).
Programme GESSOL
6/24
05/2013
comportementaux accordent une grande importance à l’observation des comportements réels et à
leurs déterminants. Cette logique inversée a permis l’identification de plusieurs biais cognitifs et
comportementaux, correspondant à des « erreurs régulières » ou à des « déviations systématiques »
dans les décisions des agents par rapport à ce qui est prédit par la théorie néoclassique
(Venkatachalam, 2008 ; Gowdy, 2008 ; Shogren et Taylor, 2008 Beretti et la., 2013). En résumé,
l’économie comportementale démontre la pluralité des normes qui guident les choix des individus. Elle
insiste sur le fait que le comportement est influencé par un ensemble de facteurs qui vont bien au-delà
de ce que suppose la rationalité conventionnelle mobilisée dans les modèles économiques.
Or, la conception des interventions est souvent basée sur le modèle de comportement rationnel, ce qui
est à la fois réducteur et susceptible de restreindre considérablement leur efficacité. En effet, sans
renier les apports de ce modèle du comportement individuel, des travaux récents en psychologie et en
neurosciences (Kahneman, 2003 ; Thaler et Sunsein, 2008) distinguent un autre modèle de
fonctionnement qui opère en parallèle dans le cerveau humain et qui accorde une importance
considérable au contexte en mobilisant des automatismes, des raccourcis mentaux et des processus
heuristiques dans la prise de décision. Alors que le système réflexif suppose une analyse consciente
et approfondie des éléments disponibles sur un domaine donné afin d’agir dans un sens qui maximise
les intérêts de l’individu, le système automatique opère de manière inconsciente, quasi-automatique et
superficielle sur de multiples domaines. Quelques éléments de comparaison sont évoqués dans le
tableau 1.
Tableau 1: Coexistence de 2 systèmes de prise de décision au niveau du cerveau humain
(Kahneman, 2003; Thaler et Sunstein, 2009)
Système automatique
Système réflexif
Non contrôlé
Contrôlé
Sans efforts
Nécessite des efforts
Logique d’association
Déduction
Rapide
Lent
Inconscient
Conscient
Poids des émotions
Se conforme à des règles
En ignorant ce deuxième système opératoire, les interventions publiques se sont vraisemblablement
sur-focalisées sur le système réflexif et se sont en quelque sorte privées d’un levier important de
compréhension et d’influence du comportement humain. Les éléments développés dans la suite de
notre contribution visent à encourager une prise en compte plus systématique du système
automatique afin de mieux comprendre les résultats parfois surprenants de certaines politiques, mais
aussi afin de favoriser des politiques plus efficaces, notamment en faveur d’une gestion plus durable
des sols.
2. Quelques éléments de méthodologie et éventuelles difficultés rencontrées
Le projet ECOSOL a permis d’explorer plusieurs situations et de mettre en évidence des résultats
originaux relatifs aux régimes incitatifs sous-jacents à la fourniture des services écosystémiques des
sols. Les méthodes d’investigation retenues reflètent un certain éclectisme et comprennent
notamment des études de cas détaillées, des (quasi)expérimentations en laboratoire
(décontextualisées) et sur le terrain et de la modélisation. De manière générale, les différentes
opérations du projet ECOSOL ont été menées en parallèle. Les opérations utilisant des questionnaires
quasi-expérimentaux ont généralement fait l’objet de discussions préalables avec des spécialistes de
sciences du sol, mais aussi avec des professionnels impliqués dans le conseil aux agriculteurs et de
pré-tests sur des populations néophytes, puis sur quelques professionnels et enfin sur l’échantillon
considérée. Les premiers résultats sont indiqués de manière succincte dans les sections suivantes,
sachant que les articles d’accompagnement contiennent l’ensemble des détails ainsi que les
références bibliographiques permettant de resituer l’apport des travaux réalisés. Nous faisons donc le
choix de surtout restituer (1) les intuitions/hypothèses sous-jacentes et une rapide justification
théorique/conceptuelle de ces dernières (2) les principaux résultats actuellement disponibles et (3)
quelques implications en termes de politiques publiques pour la gestion durable des sols. Nous avons
préféré mentionner les implications pratiques et les recommandations tout au long du texte pour
faciliter la compréhension.
De manière schématique, chaque ‘biais’ a été étudié en le définissant et en le caractérisant par
rapport aux prédictions dérivées de la théorie néoclassique, puis en montrant comment sa prise en
Programme GESSOL
7/24
05/2013
compte dans l’analyse est susceptible de modifier les résultats escomptés. Les recherches se sont
parfois appuyées sur une modélisation analytique permettant de comparer le comportement prédit
avec et sans biais. Une fois, ces bases théoriques posées, nous avons eu recours à l’économie
expérimentale en laboratoire et sur le terrain (List, 2011) pour apporter des éléments de validation
empirique ou de prolongement du débat engagé sur le plan théorique, avec des applications
spécifiques liées aux sols. Pour des facilités d’exposition et du fait des contraintes imposées en termes
d’espace disponible, nous présentons chaque biais de manière indépendante et nous insistons sur les
dimensions que nous souhaitons explorer, tout en étant conscient que ces raccourcis peuvent occulter
certains aspects, notamment l’importance des interaction entre biais (Beretti et al., 2013b).
Toutes les tâches prévues dans le projet ECOSOL ont été achevées, sauf celles relatives à
l’évaluation monétaire et dans une moindre mesure, celles relatives aux déterminants de l’épandage
2
des boues . Etant données la faible taille de l’équipe impliquée (5 membres sur 38 mois) et la
localisation de tous les participants et personnes ressources sur la région de Montpellier, les réunions
de coordination ont été régulières et suivies au rythme d’environ une réunion en moyenne tous les
deux mois en plus des réunions informelles quasi-hebdomadaires pour certains sous groupes. Des
résultats originaux ont été produits et la valorisation de ces derniers est déjà bien entamée. Ainsi, cinq
publications sont déjà parues ou acceptées dans des revues nationales et internationales à comité de
lecture, alors que d’autres sont soumises et en cours de rédaction.
Néanmoins, un certain nombre d’évènements ont perturbé le bon déroulement du projet ECOSOL.
Sans s’appesantir sur les aléas du quotidien auquel nul chercheur n’échappe, trois faits ont
sérieusement contraints l’achèvement de certains travaux. Premièrement, l’une des chevilles ouvrières
du projet, le doctorant Antoine Beretti, s’est désengagé à l’issue de la soutenance de son doctorat. Il
était en effet prévu qu’il continue à travailler sur plusieurs questions, tout en assurant la valorisation
des résultats déjà obtenus. L’expérience qu’il avait accumulée, les ressources qu’il avait su mobiliser
(par exemple l’accès à certaines données et collaborations), les travaux qui étaient en cours et qui
l’associaient se sont donc retrouvés fortement contraints. Deuxièmement, le porteur du projet, Gilles
Grolleau s’est mis en disponibilité en Septembre 2012 pour convenances personnelles, ce qui a
fortement diminué les interactions quotidiennes avec d’autres collègues impliqués dans le projet.
Troisièmement, certaines contraintes administratives liées à certains types de dépenses (notamment
lorsque certaines expériences sortaient des clous) ont été vécues comme extrêmement lourdes,
pénalisantes et consommatrices d’énergie. De fait, l’ensemble des fonds demandés n’a pas été
3
mobilisé et certaines tâches n’ont pas été menées à terme . Néanmoins, du point de vue du porteur, le
bilan reste globalement positif.
3. Des arrangements contractuels afin de pérenniser les services d’épuration
des sols
La qualité de l’eau est un enjeu crucial dans de nombreuses régions. A proximité de nombreuses
agglomérations, cette qualité est souvent menacée par des activités agricoles qui la rendent impropre
à la consommation humaine. Cette situation peut être analysée comme un conflit entre différents
groupes d’intérêt portant sur des usages mutuellement incompatible de la même ressource. Par
exemple, les agriculteurs souhaitent utiliser les eaux souterraines comme des réceptacles ultimes des
co-produits de fertilisation ou des traitements phytosanitaires, alors que les citoyens consommateurs
d’une ville avoisinante souhaitent bénéficier d’une eau pure, notamment grâce au service d’épuration
rendue par les sols. Ces deux usages mutuellement incompatibles sont à la source d’un conflit qui
pourrait potentiellement être résolu à travers l’échange de droits de propriété entre les parties (Coase,
1960 ; Anderson, 2004). Néanmoins, la réalisation de ces arrangements volontaires est conditionnée
par le niveau des coûts de transaction qui peuvent être prohibitifs et empêcher toute transaction. Ces
coûts se rapportent notamment aux coûts d’identification des parties impliquées et de leurs droits
respectifs, à la délimitation des droits à échanger, à la rédaction de l’arrangement et aux modalités afin
de le rendre exécutoire (McCann et al., 2005 ; Déprés et al., 2008).
2
Concernant le travail relatif aux déterminants de l’épandage des boues de station d’épuration dans les
exploitations agricoles, nous avons été pénalisés par le départ prématuré d’Antoine Beretti qui travaillait sur les
données du recensement agricole à Rennes en compagnie de Douadia Bougherara. En effet, n’ayant pas
l’accréditation de la commission du secret pour pouvoir travailler sur ces données librement, cette étude était
réalisée en collaboration avec D. Bougherara (UMR SMART Rennes).
3
L’ensemble des fonds n’a pas été utilisé (reliquat d’environ 10 000 EUR sur les 69 000 EUR accordés par le
programme GESSOL), ce qui explique que certaines expériences prévues n’ont pas été réalisées.
Programme GESSOL
8/24
05/2013
Plutôt que de se contenter de reconnaître que des coûts de transaction trop élevés ont empêché
l’émergence de solutions décentralisées, nous nous sommes intéressés à deux arrangements entre
des agriculteurs et les villes avoisinantes (New-York, Munich) afin d’identifier les déterminants des
coûts de transaction et les stratégies mises en place afin de rendre les coûts de transaction
compatibles avec un arrangement décentralisé. De manière plus anecdotique et souvent à des fins de
comparaison, nous avons également recouru à des arrangements mis en place par des
embouteilleurs d’eau minérale (e.g., Perrier, Vittel) en France (Déprés et al., 2008 ; Grolleau, 2013).
Les cas étudiés permettent de mettre en évidence de manière relativement fine les facteurs ayant
favorisé ou défavorisé l’émergence de solutions contractuelles et l’efficacité de ces dernières,
notamment en termes de performances à la fois environnementales et économiques (e.g., qualité de
l’eau, coût des solutions de substitution). Ces études de cas permettent d’attribuer une valeur au
service d’épuration rendue par les sols, valeur (plancher) qui est notamment indexée sur le coût des
solutions de remplacement dans des contextes spécifiques. Notre analyse montre que l’absence
d’antipathie entre urbains et ruraux (poids de l’histoire), l’utilisation homogène des sols sur la zone
considérée, l’existence de références technico-économiques fiables et reconnues ainsi que la
demande de produits biologiques ont contribué à diminuer le niveau des coûts de transaction dans le
cas de Munich en comparaison de la ville de New-York. Pour les deux villes, l’acquisition d’une partie
des terres, la situation de monopsone, la flexibilité (notamment temporelle et en termes de moyens
exigés) des réglementations relatives à la qualité de l’eau, la sensibilisation des citoyens (e.g.
écotourisme, écolabels signalant l’origine des produits et le rapport entre leur mode de production et la
qualité de l’eau) ont favorisé l’émergence et l’efficacité d’accords contractuels dans le cas des deux
villes (Grolleau et McCann, 2012).
4. L’influence sociale : les considérations de statut et de position
Les effets de position et de statut impliquent que l’individu maximise une fonction objective
interdépendante avec les fonctions d’utilité des autres. Il n’est pas intéressé par la seule utilité
obtenue, mais aussi par sa position par rapport aux autres de son groupe de référence (Salhi et al.,
2012). Une étude récente réalisée sur des agriculteurs en Chine montre que l’adoption de pratiques de
conservation favorables à la biodiversité est certes influencée par les incitations financières
proposées, mais aussi de manière substantielle par la décision des autres (Chen et al., 2009).
L’utilisation des normes sociales pourrait ainsi permettre d’augmenter considérablement l’efficience
des politiques incitatives. De nombreux exemples existent dans le domaine de l’environnement ou
certaines entités sont davantage préoccupées par leur position relative que par leur situation en
termes absolus (Griskevicius et al., 2010 ; Bougherara et al., En révision). L’importance de ces
interdépendances peut inciter l’individu à s’éloigner du comportement prédit par la théorie traditionnelle
et même l’amener à refuser certains arrangements pourtant gagnant-gagnant mais susceptible de
modifier sa position relative (voir d’inverser son statut, par exemple de membre du ‘club des 100
quintaux’ à l’agriculture biologique) dans un sens qu’il juge indésirable. Ainsi, un agriculteur encouragé
à adopter certaines pratiques favorables aux sols ne considère pas uniquement le gain net des
évolutions proposées, mais aussi comment ces évolutions modifient ou non sa position relative dans
son groupe de référence, en fonction de la (des) dimension(s) retenue(s) (Salhi et al., 2012).
Nous avons d’abord montré l’importance des considérations de statut à la fois grâce à des
questionnaires quasi-expérimentaux, puis grâce à une expérience en laboratoire. Les questionnaires
quasi-expérimentaux appliqués à des domaines variées montrent qu’un pourcentage non négligeable
des individus préfèrent la situation positionnelle (e.g., avoir plus que les autres quitte à avoir moins en
valeur absolue) à des scénarios alternatifs (e.g, comme avoir plus en valeur absolue, mais moins que
les autres). En effet, même lorsque des gains réels sont en jeu et que les individus doivent juste
choisir entre trois couplets (X, Y) qui déterminent leur paiement (X) et celui de leur partenaire (Y). Ces
trois couplets comprennent un couplet A prônant l’égalité et permettant d’évacuer l’aversion à
l’inégalité (Fehr et Schmitt, 1999) (3, 3) ; un couplet B permettant de maximiser le gain privé quitte à ce
que le second participant obtienne plus (5, 7) ; un couplet C correspondant à des préférences
positionnelles (4, 1). La prédiction néo-classique est évidente : tous les individus devraient opter pour
l’option B qui maximise le gain individuel. Néanmoins les résultats diffèrent significativement de cette
prédiction. Même si une proportion importante des individus choisit l’option égoïste, l’option égalitaire
et l’option positionnelle cumulées ensemble représentent 55%. Ainsi, les individus sont prêts à
renoncer à des gains privés afin de promouvoir soit une certaine forme d’égalité, soit la recherche
d’une meilleure position relative. L’expérience a été reproduite en doublant les montants en jeu pour
voir comment des montants plus élevés influençaient ce type de préférences (Tableau 2).
Programme GESSOL
9/24
05/2013
Tableau 2: Pourcentages d’individus ayant choisi différentes options
Option égalitaire
Option égoïste
Pourcentage
d’individus
Pourcentage
d’individus
(3, 3)
40 %
(6, 6)
46 %
Option positionnelle
(5, 7)
45 %
(10, 14)
34 %
(4, 1)
15 %
(8, 2)
20 %
Ces résultats soutiennent l’importance des considérations de conformité à une norme d’égalité et de
préférences positionnelles. A l’aide d’une analyse conceptuelle et d’exemples, nous montrons qu’il est
possible de mobiliser cette ‘influence sociale’ afin de favoriser l’adoption de comportements
socialement désirables. Ainsi et contrairement aux prédictions traditionnelles, la fourniture privée d’un
bien public pourrait être renforcée si un mécanisme de course à la position sur cette dimension est
enclenchée, ce qui peut se faire notamment en augmentant sa visibilité. En effet, alors qu’une analyse
coût bénéfice semble déboucher naturellement sur l’adoption de certains comportements jugés
préférables, les interactions entre profitabilité et statut peuvent sérieusement remettre en question
cette conclusion (Salhi et al., 2012). Très concrètement, certaines innovations permettant une gestion
plus durable des sols sont parfois rejetées, non sur la base d’un manque d’informations, ou d’une
profitabilité insuffisante en comparaison des alternatives, mais à cause de leur effet dépréciatif en
termes de statut. Plusieurs pistes et recommandations sont suggérées afin d’éviter ces interférences
négatives et visant même à instrumentaliser la recherche de statut comme motivation à l’adoption de
comportements préférables du point de vue de la gestion durable des sols (Salhi et al., 2012).
5. L’effet d’éviction entre incitations et les stratégies mobilisables
En retenant le cadre d’analyse néoclassique, on peut faire l’hypothèse que le choix des pratiques fait
par un utilisateur du sol est celui qui lui permet de maximiser son utilité, et que pour faire évoluer ces
choix, il faut lui fournir des incitations différentes qui permettent d’augmenter la désirabilité de la
pratique souhaitée par le régulateur public. Dans le cas de l’agriculture, les politiques agroenvironnementales se sont essentiellement cantonnées aux incitations économiques comme étant la
force principale susceptible de pousser les agriculteurs à changer de pratiques. Ainsi, les mesures
agro-environnementales de la PAC proposent des paiements aux agriculteurs volontaires qui
permettent de couvrir les coûts supplémentaires de mise en œuvre et le manque à gagner induit par
l’adoption d’un cahier des charges plus respectueux de l’environnement.
En contradiction avec les prédictions pessimistes de la théorie néoclassique, les agents n’adoptent
pas systématiquement un comportement de passagers clandestins et contribuent dans une certaine
mesure à la production de ces services écosystémiques, malgré leur nature publique, sans toutefois
atteindre le niveau pareto-optimal. Cette production volontaire témoigne en faveur de l’existence de
motivations intrinsèques favorables à la fourniture des services considérés. Les motivations
intrinsèques se réfèrent aux raisons d’agir qui proviennent de l’individu lui-même, comme le plaisir ou
la satisfaction personnelle. Une personne intrinsèquement motivée entreprend une activité même si
elle ne reçoit aucune compensation autre que celle dérivée de l’activité elle-même. Les motivations
extrinsèques sont « imposées » à l’individu de l’extérieur et peuvent prendre plusieurs formes comme
la reconnaissance sociale ou la compensation monétaire pour l’adoption d’un certain comportement ou
la peur d’une sanction dans le cas du non-respect du comportement souhaité (Mzoughi, 2011).
La question est donc de savoir comment encourager et soutenir cette production, à la fois en
renforçant les motivations pré-existantes et en réduisant le comportement de passager clandestin. La
réponse traditionnelle se base notamment sur l’utilisation d’incitations monétaires positives ou
négatives. Néanmoins, des travaux récents (Frey et Oberholzer-Gee, 1997 ; Gneezy et Rustichini,
2000 ; Vollan, 2008 ; Bolle et Otto, 2010) ont démontré que l’utilisation d’incitations monétaires pouvait
interférer avec les motivations pré-existantes et aboutir à des contre-performances, c’est à dire à une
diminution du comportement recherché plutôt qu’à son augmentation. De plus, cet effet serait
contagieux (il s’étend à des domaines initialement non concernés par les incitations monétaires) et
durable (c’est à dire qu’il perdure même après suppression des incitations monétaires) Par exemple,
réalisant une étude concernant la localisation d’un centre de stockage de déchets nucléaires dans
deux communes suisses, Frey et Oberholzer-Gee (1997) ont trouvé que le pourcentage de personnes
acceptant l’infrastructure polluante dans leur voisinage diminue de 50,4% sans incitations à 24,6%
Programme GESSOL
10/24
05/2013
lorsque des incitations monétaires leur sont proposées. L’explication selon laquelle les incitations
monétaires signalent la dangerosité est écartée, car l’augmentation des rémunérations n’entraîne pas
de diminutions plus fortes du taux d’acceptation. La focalisation sur les seules motivations
extrinsèques est susceptible, sous certaines conditions, d’évincer les motivations intrinsèques et de
générer un niveau de production du bien ou service environnemental inférieur à ce qu’il aurait été sans
intervention extérieure. En d’autres termes, une politique qui utilise comme seul levier les incitations
économiques est susceptible de générer chez l’individu un sentiment d’insatisfaction de voir ses
motivations intrinsèques ignorées, non reconnues, non valorisées et le conduisant ainsi à réduire son
effort (Frey et Oberholzer-Gee, 1997 ; Gneezy et Rustichini, 2001). Ainsi l’utilisation d’incitations
financières, comme dans les MAE peut durablement éroder ou au contraire renforcer de manière
significative la propension des acteurs à entreprendre des actions écologiquement préférables qu’ils
entreprenaient déjà en l’absence d’incitations financières (phénomène de renforcement ou d’éviction).
Notre apport a notamment consisté à vérifier l’existence de ce risque d’effet d’éviction sur des
problématiques environnementales et en particulier sur la gestion durable des sols et à explorer
quelques stratégies susceptibles de l’atténuer, voire de contribuer à un renforcement des motivations
pré-existantes. En effet, la littérature existante n’offrait guère de solutions à l’effet d’éviction, hormis la
non-utilisation des incitations monétaires dans certains domaines.
Nous sommes allés plus loin en explorant deux stratégies : la direction de l’argent vers la cause ou
vers l’individu avec une option où le répondant choisit entre lui-même et la cause et la qualification de
la somme sous forme de paiement, ou de dédommagement. Le protocole expérimental consistait en
une demande d’action ayant des vertus publiques, demande assortie d’une rémunération ou non.
Nous avons d’abord réalisé une étude auprès de 1800 ménages (sur des pratiques proenvironnementales génériques), puis sur un échantillon restreint de viticulteurs (environ 300
viticulteurs contactés et 47 réponses valides), puis sur un échantillon de 2000 viticulteurs de la région
montpelliéraine sur des actions en rapport avec la gestion durable des sols. Les actions choisies
devaient être réalistes (e.g., utilisation des parcelles à des fins expérimentales, invitation à une réunion
relative à la gestion durable des sols à SupAgro). Ces études ont permis de mettre en évidence
l’existence d’un effet d’éviction apparemment plus prononcé chez les particuliers en comparaison avec
les viticulteurs (Tableaux 3 et 4). Ce moindre effet peut être lié à la nature des relations préexistantes
chez les viticulteurs (et autres agriculteurs) où les approches en faveur de l’environnement en général
et des sols en particulier sont déjà largement promues à travers des incitations monétaires.
Tableau 3 : Taux de participation des ménages en fonction de l’incitation (n=1118)
Incitation
Montant de l’incitation EUR Taux de participation
Pas d’incitation
0
28.66 %
Pour l’individu
2
19.38 %
Pour l’individu
7
27.27 %
Pour une association environnementale
2
21.74 %
Pour une association environnementale
7
24.20 %
Choix : individu ou association environnementale
2
23.87 %
Choix : individu ou association environnementale
7
29.45 %
Les résultats concernant les stratégies susceptibles d’atténuer l’effet d’éviction montrent une efficacité
relative du fait de laisser l’individu choisir vers qui diriger l’incitation monétaire. Plus concrètement, ce
résultat peut correspondre à une stratégie institutionnelle permettant de solliciter les individus
monétairement motivés en évitant de décourager les individus intrinsèquement motivés (qui apprécient
de pouvoir diriger l’argent vers la cause) (Beretti et al., 2013a). Néanmoins, les résultats méritent des
investigations supplémentaires, surtout lorsque l’on prend en compte le critère de coût-efficacité de
cette stratégie. Une modélisation analytique de ce mécanisme d’évitement de l’effet d’éviction a
également été réalisée et est actuellement soumis en vue de publication (Beretti et al., Soumis).
L’investigation pilote menée sur 300 viticulteurs de la région montpelliéraine par téléphone en Juin et
Juillet 2010 montre que le risque d’effet d’éviction est bien réel (Tableau 4). Trois activités
hypothétiques ont été proposées aux répondants : une visite de leurs parcelles par des spécialistes de
sciences du sol à des fins d’analyse et de recherche, une réunion à Montpellier SupAgro sur la gestion
plus durable des sols et l’acceptation de boues de station d’épuration sur leurs parcelles. L’échantillon
a été divisé en deux sous échantillons, l’un où les activités sont proposées avec une contrepartie
Programme GESSOL
11/24
05/2013
financière faible (20 EUR), l’autre où aucune contrepartie financière n’est proposée. Cette étude nous
a permis d’affiner les activités retenues pour l’enquête postale adressée aux plus de 2000 viticulteurs
en 2011 et notamment d’écarter l’activité liée aux boues à cause de son caractère très controversée.
Tableau 4 : Taux d’acceptation de différentes activités par les viticulteurs selon le traitement
(n=47)
Visite
Réunion
Epandage
Sans rémunération
86 %
45 %
53 %
Rémunération
61 %
36 %
39 %
En outre, dans le sous-échantillon où la contrepartie financière est proposée, la moitié des répondants
a été soumis à une contrepartie financière de type contrôle (e.g., ‘Etant donné que nous sommes prêts
à vous payer 20 EUR pour nous assurer de votre participation, seriez-vous prêt à participer à une telle
rencontre ?’) alors que l’autre moitié a été soumis à une contrepartie de type soutien (e.g., ‘Etant
donné que nous sommes prêts à vous dédommager à hauteur de 20 EUR si vous venez, seriez-vous
prêt à participer à une telle rencontre?’). Les résultats préliminaires liés à la qualification de l’incitation
financière semblent montrer un effet non négligeable de l’étiquette attribuée à l’incitation au-delà de
son seul montant qui reste identique (Tableau 5). En effet, la situation sans contrepartie financière est
toujours celle qui obtient les taux de participation les plus élevés. Les résultats semblent indiquer au
moins partiellement que la formulation de l’incitation sous forme de dédommagement réduit l’intensité
du risque d’éviction. Néanmoins, le nombre de réponses réellement exploitables (n=47 soit un taux de
réponses utilisables de 15,6%) est trop faible pour permettre de tirer des conclusions ayant une
significativité statistique crédible.
Tableau 5 : Taux d’acceptation de différentes activités par les viticulteurs selon le traitement
Visite
Réunion
Epandage
Sans rémunération
86 %
45 %
53 %
Rémunération contrôle
72 %
27 %
36 %
Rémunération soutien
53 %
43 %
45 %
Grâce à ce pilote, nous avons pu soumettre un questionnaire affiné à environ 2000 viticulteurs par voie
postale, début 2011. Après de nombreux tests et entretiens préliminaires aves des experts et des
viticulteurs, tout a été fait pour avoir un questionnaire bref (1 page) et identique entre les traitements,
sauf sur un mot (payer versus dédommager). Dans ce cas, les résultats relatifs à la seconde stratégie
4
sur la qualification des sommes versées en tant que paiement (contrôle ) ou dédommagement
5
(soutien ) par rapport à l’absence de contrepartie financière sont surprenants (Tableau 6). En effet,
tous les tests statistiques visant à vérifier d’éventuelles différences entre les traitements sont non
significatifs. Même si ces résultats ne confirment pas les résultats de l’étude pilote, ils indiquent que la
proposition d’une contrepartie financière faible n’a pas eu d’effets sur l’acceptation de l’activité
proposée, en l’occurrence une ½ journée de rencontre à Montpellier SupAgro rassemblant viticulteurs,
chercheurs et autres acteurs du sol pour réfléchir ensemble à l’avenir des sols du Languedoc6
Roussillon et aux stratégies d’une gestion plus durable .
Tableau 6: Taux de réponse et d’acceptation des viticulteurs selon le traitement (n=1950)
Traitement Témoin Traitement Paiement
Traitement
(n=656)
(n=645)
Dédommagement (n=649)
Taux de réponses
9,45 %
8,22 %
8,01 %
Taux d’acceptation
4,72 %
4,03 %
3,7 %
La principale conclusion de cette section est que le risque d’effet d’éviction est réel, notamment dans
le cadre des politiques visant à encourager l’adoption de pratiques préférables à travers l’utilisation de
contreparties financières surtout si elles sont modestes. Il est également crucial de veiller à la direction
4
‘Nous nous engageons à vous payer 20 EUR pour votre participation.’
‘Nous nous engageons à vous dédommager de 20 EUR pour votre participation.’
6
Cette réunion a effectivement été organisée ultérieurement et a permis la restitution de certains résultats des
recherches menées au sein du projet ECOSOL. Les agriculteurs présents ont tous reçus la contrepartie
financière proposée, indépendamment du traitement auquel ils appartenaient lors de l’enquête postale. Cette
option a été retenue afin d’éviter tout sentiment de discrimination entre les participants.
5
Programme GESSOL
12/24
05/2013
et à la qualification des sommes utilisées qui peuvent renforcer ce risque ou au contraire le diminuer.
Dans cet ordre d’idée, les représentants des producteurs en agriculture biologique (FNAB) ont
récemment milité en faveur d’une ‘rémunération de reconnaissance’.
6. Les effets de formulation
Comme l’a abordé de manière indirecte la section précédente, les individus ne réagissent pas
uniquement aux changements nets, mais également en se basant sur une situation de référence.
Ainsi, une proposition aboutissant à un résultat net identique générera des réactions différentes en
fonction du point de référence utilisé par les individus. Contrairement à l’idée prédominante, Tversky et
Kahneman (1981) démontrent que les choix faits par un individu dépendent non seulement des
caractéristiques objectives des alternatives qui s’offrent à lui mais aussi de la manière dont elles lui
sont présentées. Dans le domaine environnemental, l’une des mesures les plus souvent avancées
pour lutter contre le réchauffement climatique est d’augmenter le prix des activités émettrices de gaz à
effet de serre de manière à refléter les dommages environnementaux causés. Ces activités devenant
plus coûteuses sont donc rendus moins attractives. Dans une étude empirique, Hardisty et al. (2010)
montrent que le terme utilisé pour désigner une augmentation de coût identique sur des vols aériens
(taxe versus compensation) peut avoir un impact considérable sur l’acceptabilité de la mesure. En
effet, le même supplément de prix qualifié de compensation (offset) est plus facilement accepté que si
ce dernier est qualifié de taxe. En d’autres termes, la même incitation/désincitation en termes de
montant peut entraîner, selon la formulation employée (e.g. aide, indemnisation, subvention,
dédommagement, rémunération, paiement, soutien, compensation), des comportements différenciés.
Appliquée au domaine agricole, ce résultat nous pousse à penser que la manière dont les agriculteurs
perçoivent une incitation donnée, comme une compensation ou comme une récompense, peut
influencer leurs décisions d’adopter ou non les mesures suggérées. Le Ministère de l’agriculture
français a ainsi préféré qualifier les aides financières des MAE du terme neutre de « paiements agroenvironnementaux ». Cependant, l’insistance de la Commission européenne pour que ces paiements
ne financent que des coûts supplémentaires ou des manques à gagner a le clair désavantage de
signaler aux agriculteurs qu’un changement vers des pratiques plus respectueuses de l’environnement
ne peut pas se faire sans subir une perte, ce qui ne correspond pas forcément à la réalité, la baisse du
coût des intrants pouvant souvent compenser une baisse de rendements. Par ailleurs, elle indique
aussi que la collectivité ne considère pas le changement de pratiques comme générateur d’un ‘service
environnemental’ dont la fourniture par l’agriculteur pourrait être ‘récompensée’. Même si le montant
payé est le même, l’effet psychologique peut être très différent, puisqu’il attire l’attention de l’agriculteur
sur les conséquences individuelles négatives qu’il aurait à subir au lieu de valoriser sa contribution à
un impact collectif positif.
61. L’effet des causes de la dégradation des sols sur le consentement à payer
Lors de la mise en place d’une politique visant à promouvoir une gestion durable des sols, vaut-il
mieux la présenter comme une amélioration nette ou comme la restauration d’un état passé ? Le
soutien des individus et leur consentement à payer est-il identique, quelle que soit la cause de la
dégradation ? De nombreux économistes et psychologues ont remis en question l’idée conventionnelle
que l’utilité des agents dépend exclusivement du résultat en suggérant que des points de référence et
des causes alternatives d’une situation donnée combinés au résultat peuvent modifier l’utilité perçue
(Bulte et al., 2005 voir également Benz et al., 2004). Selon la théorie des perspectives, le point de
référence détermine si le changement proposé est perçu comme un gain ou comme une perte, ce qui
va alors modifier la façon dont l’individu évalue les alternatives. En fait, l’asymétrie de la fonction
d’utilité implique qu’une perte par rapport à un point de référence va avoir un impact plus grand sur la
prise de décision qu’un gain de valeur équivalente à partir du même point de référence (Gregory et al.,
1993).
L’aversion aux pertes signifie que les individus sont plus motivés et sont disposés à payer davantage
pour éviter une perte que pour acquérir un gain de la même valeur. En parallèle, une littérature réduite
mais grandissante souligne que le consentement à payer (CAP) pour lutter contre des problèmes
environnementaux peut varier selon l’origine naturelle ou anthropique du problème en question. La
combinaison de ces deux paramètres, point de référence et causes alternatives, reste néanmoins rare
dans la littérature. L’objectif de cette opération était de combler cette lacune dans la littérature en
utilisant une enquête quasi-expérimentale réalisée auprès de résidents français en rapport avec la
bétonisation. Notre principale hypothèse est que les individus ne prennent pas en compte les points de
références de façon isolée, mais qu’ils considèrent également le contexte, tel que des causes
Programme GESSOL
13/24
05/2013
alternatives susceptibles d’expliquer l’état actuel. En d’autres termes, même avec des points de
référence identiques, les individus peuvent se comporter différemment selon les causes alternatives
ayant conduit à ces points. Notre étude porte sur les services écosystémiques fournis par les sols, qui
contrairement aux autres compartiments tels que l’air ou l’eau, n’a reçu que très peu d’attention de la
part des économistes. Plusieurs recommandations en termes de politiques publiques sont formulées,
suggérant des améliorations à faible coût susceptibles d’engendrer d’importants bénéfices. Par
exemple, un élément récurrent de nos investigations est la sensibilité différenciée des individus par
rapport à l’origine ‘perçue’ des fonds (et pas uniquement par rapport à l’arbitrage coûts-bénéfices des
arrangements proposés), un peu comme si tous les euros n’avaient pas la même valeur.
Nous testons ces hypothèses en interrogeant 1760 individus de Lunel (34400) et de Sommières
(30250), des communes touchées par de récentes inondations, à propos de l’importance qu’ils
accordent à un programme d’amélioration des sols, susceptible de contribuer à une diminution des
risques d’inondation. Selon le traitement, le changement proposé, identique dans chacun des
scénarios, est présenté comme une amélioration nette, comme la restauration d’une perte, comme la
restauration d’une perte due à l’activité humaine, ou encore comme la restauration d’une perte due à
des phénomènes naturels. Nous interrogeons également les sujets sur leur CAP pour ce programme.
A ces deux variables, nous ajoutons une troisième variable : le taux de réponse, comme indicateur
supplémentaire de l’intérêt porté au projet présenté.
Le détail des investigations empiriques est disponible dans Beretti et Grolleau (2013, Soumis pour
publication). Les résultats montrent que le taux de réponse est systématiquement et significativement
plus faible pour le traitement ‘gain’ comparé à chacun des traitements ‘perte’ (Tableau 7).
Etonnamment, l’intérêt déclaré pour le programme est plus important lorsqu’une cause humaine est
avancée par rapport à une cause naturelle, alors que pour le CAP, l’effet est inversé. La restauration
d’une perte ayant une cause expliquée, quelle qu’elle soit, entraîne un CAP plus important que la
restauration d’une perte non expliquée. Ce dernier résultat n’est toutefois pas significatif au niveau
statistique.
Tableau 7 : Résultats clés par rapport aux différentes causes
Traitement
Nombre de
Taux de
Intérêt pour le
réponses
réponse (%)
programme
Gain (G)
Perte non expliquée (UL)
Perte causée par la nature (NC)
Perte causée par l’homme (MC)
Total
15
33
32
28
108
3.52
7.99
7.71
6.65
6.45
5.36
5.61
5.19
6.04
5.56
CAP pour le
programme
(EUR)
14
13.44
17.10
15.36
15.09
A nouveau, nous avons réalisé une extension de cette expérience en interrogeant quelques viticulteurs
de la région (n=47) par voie téléphonique sur l’intérêt qu’ils portent, ainsi que sur leur CAP, concernant
sur un programme d’amélioration/restauration de la teneur en Matière Organique (MO) des sols. A la
lecture des résultats, il convient néanmoins de rappeler, ce ne sont pas tant les montants en valeur
absolue qui signifient quelque chose, mais la comparaison des CAP entre les différents scénarios. Les
résultats semblent partiellement confirmer nos hypothèses, notamment en termes de CAP.
Tableau 8 : CAP des viticulteurs pour un programme d’amélioration des sols
Programme d’amélioration environnementale présenté comme…
Consentement à payer
pour le programme (EUR)
… un gain net
16.0
… la restauration d’une perte non expliquée
12.5
… la restauration d’une perte due à des causes humaines
15.7
… la restauration d’une perte due à des causes naturelles
20.0
N
10
12
14
11
Ces résultats (Tableau 8) sont plus de nature indicative car ils ne permettent pas de générer des
résultats statistiquement robustes en raison du faible nombre d’observations. Ces recherches
soulignent notamment l’importance des points de référence et de la causalité (réelle ou perçue) dans
la réaction des agents par rapport aux politiques liées aux sols.
Programme GESSOL
14/24
05/2013
62. L’origine des incitations
L’analyse économique traditionnelle suppose que les alternatives sont évaluées en fonction de leur
contenu objectif et non en fonction de celui qui les propose. Néanmoins, plusieurs travaux, notamment
en psychologie sociale, démontrent que les individus, en particulier dans un processus de négociation,
pratiquent une sorte de ‘réaction dévaluatoire’. Par exemple, des individus ont jugé très différemment
le même plan de paix au Proche Orient en fonction de l’origine attribuée : la partie palestinienne, la
7
partie israélienne ou une tierce partie (Maoz, 2006) . Sans approfondir les mécanismes sous-jacents à
ce type d’obstacles à la négociation, l’intuition est relativement simple. La négociation est un
processus social interactif dans lequel chaque partie ne cesse de tirer des conclusions au sujet des
intentions, des motifs et de la bonne foi de l’autre (Mnookin, 2006). Ce qui est proposé par un
adversaire se fait vraisemblablement au détriment de l’autre partie, d’où une suspicion ou une
résistance liée à l’origine de la proposition. Ross et Stillinger (1991) arguent ainsi que la proposition
(l’incitation) faite par un individu considéré comme un adversaire subit généralement une
‘dévalorisation’, du fait de son origine. De manière concrète, ce biais pourrait expliquer des réactions
des agriculteurs très différentes à des propositions pro-environnementales pourtant identiques mais
dont la source serait différente (e.g. bureaucrates bruxellois versus syndicats agricoles). La principale
implication de ce biais est de porter une attention toute particulière à l’attribution que peuvent faire les
agents concernés à une proposition. Par exemple, certains auteurs (e.g., Mnookin, 2006) arguent que
l’intervention d’une tierce partie indépendante et crédible ou le fait de permettre aux agents visés de
s’attribuer la mesure envisagée peuvent considérablement influencer son niveau d’acceptation et de
soutien.
Afin de tester cet effet, nous avons réalisé une expérience pilote en Tunisie sur l’acceptation d’une
taxe visant à promouvoir une gestion plus durable des sols. Etant donné le contexte instable en
Tunisie et les enjeux qui pourraient sembler largement prioritaires par rapport aux sols, nous avons
voulu tester si le soutien des répondants à une mesure hypothétique parfaitement identique différerait
en fonction de l’origine de la mesure. Les résultats montrent que le soutien à la taxe (mesuré sur une
échelle de 1 [pas de soutien] à 5 [soutien total]) augmente significativement quand ce sont les
écologistes ou les scientifiques qui sont à l’origine de cette taxe par rapport à une situation neutre,
c’est-à-dire sans préciser l’origine. Par contre, il n’y a pas de différence significative entre l’origine
écologiste et l’origine scientifique. Cette configuration reste la même quel que soit le genre des
participants.
Tableau 9 : Réponse moyenne selon l’origine de la taxe (écologistes vs scientifiques vs neutre)
Ecolos
Scientifiques
Neutre
Test de Wilcoxon
(40)
(40)
(40)
E/S
E/N
S/N
Réponse moyenne
3,65
3,22
2,5
ns
***
**
Une expérience a également été menée à Madagascar en juin 2013 pour étudier le soutien des
individus à un programme de conservation des sols dans un design 2X2 croisant l’origine de l’aide
(organisation d’agriculteurs locaux versus organisme international) et la qualification de celle-ci
(rémunération versus indemnisation). 746 étudiants de l’université d’Antaranivo ont été sollicités pour
évaluer un programme de conservation des sols, notamment en lui attribuant sur une échelle de Likert
une note de 1 (très négatif) à 7 (très positif). Les résultats sont encore préliminaires, mais ils indiquent
que l’origine des fonds et leur qualification influencent le degré de soutien accordé au programme
considéré. En effet, les gens évaluent plus positivement un programme au contenu identique lorsque
ce dernier est une indemnisation d’un organisme international (5,3) que lorsque ce dernier est un
8
paiement accordé par une organisation d’agriculteurs locaux (4,8) .
Dans le cas des mesures de politiques liées aux sols, au moins deux suggestions peuvent être faites
pour mobiliser positivement ce biais. Le premier est de responsabiliser les agriculteurs dans le
montage du cahier des charges. S’ils se sentent parties prenantes dans la construction de la MAE, ils
seront plus prêts à s’engager par la suite. Cela peut se faire d’une part par des processus de
concertation et de co-décision avec les représentants de la profession agricoles, et d’autre part de
7
Bien que cet exemple soit éloigné des considérations liées aux sols, nous l’utilisons, car il est très parlant et
généralement considéré comme un exemple fondateur de ce courant.
8
La différence est statistiquement significative à 1%.
Programme GESSOL
15/24
05/2013
façon plus individuelle en invitant les agriculteurs à soumettre leur propre cahier des charges via des
systèmes d’appel d’offre qui existent depuis longtemps aux Etats-Unis ou en Australie (Saïd et Thoyer,
2007).
De manière plus globale, au-delà de la conception des programmes et des instruments des politiques,
il semble crucial de ne pas négliger les termes employés et le contexte évoqué dans l’esprit des
agents sollicités. En effet, ce contexte joue un rôle déterminant, qui jusque là a trop souvent été
négligé ou relégué à l’arrière-plan alors les options retenues (sans qu’il s’agisse forcément d’un choix
intentionnel) pourraient contribuer à renforcer ou à diminuer les objectifs visés.
7. Le timing des incitations : différer l’application d’une politique pour
renforcer son acceptation ?
Un résultat relativement robuste de l’économie comportementale se rapporte à l’incohérence
temporelle des préférences qui se traduit par une survalorisation du présent et une dévalorisation du
futur (Frederick et al., 2002). Or, certaines innovations environnementales en faveur d’une gestion plus
durable des sols se caractérisent par des coûts privés immédiats (e.g. augmentation des coûts de
production, perte de statut social dans une communauté donnée) et des bénéfices futurs ayant
souvent, en outre, une dimension collective. En conséquence, les pertes immédiates sont souvent
survalorisées alors que les gains futurs mêmes importants sont sous-valorisés, du fait d’une
actualisation hyperbolique des gains. Par exemple, l’adoption de l’agriculture biologique suppose
généralement des coûts immédiats et des bénéfices décalés dans le temps. Ainsi, si les agriculteurs
font une actualisation hyperbolique de leurs gains, la valeur accordée à ces bénéfices futurs serait
faible. Cette actualisation hyperbolique, par opposition à l’actualisation exponentielle de l’analyse néoclassique, génère également une tendance à la procrastination, qui permet aux exploitants, tout en
reconnaissant les bienfaits de certaines innovations agro-environnementales, de repousser
constamment leur adoption à plus tard.
Nous nous sommes intéressés à une proposition récente s’appuyant sur ce biais afin d’accroître
l’acceptation de politiques socialement désirables. Puisque les individus ont tendance à dévaloriser les
coûts et bénéfices futurs avec une forte préférence pour le présent, il peut être opportun de leur
proposer des choix pro-environnementaux certes coûteux, mais dont l’application et donc le coût est
différé dans le futur. En proposant aux agents une décision immédiate, mais dont les coûts et les
bénéfices sont différés, il serait possible d’obtenir un soutien plus important (Roger et Bazerman,
2008). Nous avons voulu répliquer dans un contexte différent les protocoles proposés par Roger et
Bazerman (2008). Nous avons ainsi réalisé une enquête auprès de deux échantillons : un échantillon
aléatoire (n=185) constitué de personnes abordées à proximité de la principale place de Montpellier et
l’autre de convenance (n=58) sur leur soutien (sur une échelle de 1 [aucun soutien] à 9 [soutien total])
vis à vis d’une taxe visant à promouvoir une gestion durable des sols. Chaque sous échantillon a été
soumis à un questionnaire proposant soit une taxe prenant effet immédiatement ou dans deux ans.
Contrairement aux hypothèses et aux résultats obtenus par Rogers et Bazerman (2008), nos résultats
ne mettent pas en évidence d’effet lié au décalage dans le temps de l’application de la mesure.
Plusieurs explications ont été avancées ainsi que la nécessité d’investigations supplémentaires
(Tableau 10) (Grolleau et Mzoughi, Soumis).
Tableau 10 : Soutien à une taxe relative à la gestion des sols selon l’entrée en vigueur
Echantillon aléatoire (N=185)
Echantillon de convenance (N=58)
Present
Future
Wilcoxon
Present
Future
Wilcoxon
(N=92)
(N=93)
test
(N=28)
(N=30)
test
Moyenne des
4,63
4,74
ns
4,85
4,73
ns
réponses
En résumé, alors que l’attention a surtout été dirigée sur le montant des incitations en jeu, plusieurs
études (e.g., Duflo et al., 2011) soutiennent que la répartition optimale de ces dernières dans le temps
en fonction des contraintes des agents peut s’avérer plus efficace et moins coûteuse, surtout dans un
contexte de fortes contraintes budgétaires. Une piste que nous aurions souhaité explorer aurait été
d’évaluer si les agriculteurs s’engageant dans des démarches privées (e.g., Lu’Harmony) ou publiques
(e.g., MAE) en faveur d’une gestion plus durable des sols ont des préférences temporelles différentes
des agriculteurs dits ‘conventionnels’ sur des dimensions privées et publiques.
Programme GESSOL
16/24
05/2013
8. La dynamique des actions : L’effet de compensation morale
Nous avons également consacré une partie de nos recherches à la dynamique des actions en faveur
de l’environnement et des sols en particulier. En effet, des recherches récentes montrent qu’une
‘bonne action’ pourrait contribuer à ‘désinhiber l’individu en lui permettant d’entreprendre des actions
‘moins souhaitables’ par la suite, comme si l’individu recherche une sorte d’équilibre entre bonnes et
mauvaises actions. Cet effet de compensation morale demeure relativement méconnu dans la
littérature académique en sciences économiques et juridiques, hormis quelques travaux récents qui lui
sont consacrés dans divers domaines (e.g., Mazar et Zhong, 2010 ; Khan et Dhar, 2006). Par exemple
Woodyard (2009) explique que les conducteurs de véhicules hybrides roulent plus, ce qui peut se
traduire par un effet global défavorable à la préservation de l’environnement. L’attention limitée
accordée à ce biais comportemental contraste avec son impact potentiel qui pourrait notamment
expliquer les effets contre-productifs de certaines politiques visant à encourager des comportements
pro-sociaux, comme la préservation de l’environnement. En effet, plusieurs exemples anecdotiques
montrent qu’une bonne action dans un domaine environnemental (e.g., recycler ou économiser l’eau)
peut inconsciemment ‘autoriser’ l’individu à commettre une action moins désirable dans un domaine
identique ou différent (Clot et al., A paraître). L’intuition sous-jacente serait d'appliquer ce concept aux
comportements favorables aux sols au sein d'une population pour vérifier s'il existe une sorte de
'comptabilité' où les actions déployées donneraient lieu à des substitutions, à des complémentarités
plus ou moins équivalentes pouvant aboutir à un effet global incertain et dans certains cas à une
dégradation des sols (ou d’autres éléments du capital naturel) plutôt qu’à leur amélioration.
Nos travaux nous ont permis de définir et de caractériser l’effet de compensation morale à différents
niveaux d’analyse intra-individuel et inter-individuel tout en soulignant les risques d’effets pervers sur
lesquels il peut déboucher. Ce biais comportemental est en effet susceptible de remettre en question
l’efficacité de certaines politiques. Nous utilisons à cette fin des études et des exemples tirés de
différents domaines, notamment en rapport avec les actions en faveur de l’environnement. Nous
avons identifié plusieurs pistes de recherche visant à mieux comprendre les conséquences directes et
indirectes de l’effet de compensation morale, tout en mentionnant quelques stratégies mobilisables
afin d’atténuer ce biais (Clot et al. A paraître). Par exemple, nous vérifions si le fait que la bonne action
soit rémunérée impacte le risque d’effet de licence. De même, que se passe t-il lorsque la bonne
action est imposée, par exemple par voie réglementaire ? Se pourrait-il que des exigences
réglementaires plus sévères sur certains aspects liés aux sols se traduisent par des comportements
moins désirables dans des domaines non couverts par la réglementation ? Les résultats des
expériences menées sur l’effet de licence sont en cours d’exploitation et ne seront disponibles sous
une forme diffusable que d’ici quelques mois.
9. Partenariats mis en place, projetés, envisagés
Le projet a permis aux chercheurs impliqués de solliciter régulièrement des spécialistes du sol afin de
discuter de la validité et de la crédibilité des enquêtes envisagées. Les travaux engagés dans le cadre
du projet ECOSOL ont permis de développer des partenariats plus ou moins forts avec différents
catégories d’acteurs que l’on peut grosso modo classer en quatre catégories :
-
des économistes et autres chercheurs en sciences sociales (sociologie, psychologie) en
France et à l’étranger ayant un intérêt pour les questions liées aux sols (e.g., G. Libecap, L.
9
McCann, M. Bell, S. El Harbi )
des chercheurs et experts en sciences du sol et/ou en services écosystémiques (e.g., Y. Le
Bissonais, N. Gallai, P. Coll, JM. Ouvry)
des praticiens en prise avec les politiques environnementales et agro-environnementales à
différents niveaux et dans des institutions diverses (e.g. Ministères en charge de l’agriculture,
de l’écologie, OCDE) (G. Lecat, A. Darpeix, T. Uetake),
des utilisateurs directs des sols, notamment des agriculteurs individuels qui constituent de
véritables personnes ressources et des associations diverses (e.g., Alterre Bourgogne,
AREAS) qui ont exprimé de l’intérêt pour les résultats du projet
9
Un partenariat privilégié est en train de se développer avec le laboratoire d’économie expérimentale de Sousse
(Tunisie).
Programme GESSOL
17/24
05/2013
Malheureusement, le départ anticipé d’Antoine Beretti (et dans une certaine mesure l’absence du
porteur du projet) a d’une certaine façon perturbé et menacé la pérennité de ces relations. Aujourd’hui,
certains de ces contacts peuvent être qualifiés latents, mais pourraient aisément être réactivés.
10. Conclusion générale
« The main reason behavioral science should be part of
the policy debate is that it provides in some cases a
perspective that is vastly different from economics. »
(Amir et al., 2005)
Sans remettre en question l’utilité des politiques visant à modifier le comportement des utilisateurs du
sol en recourant aux (dés)incitations monétaires et à l’information, nous avons montré que cette vision
est réductrice et que l’efficacité de ces dernières est également déterminée par d’autres paramètres
souvent ignorés ou relégués au second plan. Par exemple, au-delà du seul montant des incitations
financières, nous avons montré l’importance d’intégrer dans la réflexion des décideurs divers
paramètres comme les causes attribuées aux problèmes visés, la formulation, l’origine et le timing des
incitations. Ce projet, au-delà des exemples étudiés et des biais examinés est d’inviter les chercheurs
et les décideurs à prendre en compte de manière plus systématique la coexistence de deux systèmes
opératoires dans la prise de décision. Nos résultats suggèrent plusieurs pistes pratiques qui si elles
étaient confirmées pourraient générer des effets de premier ordre pour des coûts relativement faibles.
En outre, les résultats de quelques expériences sont en cours d’exploitation et devraient déboucher
sur des résultats dans des délais relativement brefs.
Bien que notre objectif soit d’encourager à repenser les modalités de conception et d’application des
politiques en faveur d’une gestion plus durable des sols, nous ne saurions passer sous silence un
certain nombre de critiques formulées à l’encontre de l’économie comportementale, comme l’absence
de cadre d’analyse homogène et fédérateur conduisant à une juxtaposition de petits modèles,
l’ignorance des interactions entre biais, le recours excessif à des expériences de laboratoire ne
reflétant pas forcément les conditions de prise de décision dans le monde réel ou le risque de dérive
vers une forme de paternalisme ou de manipulation (Beretti et al., 2013b). Ce débat en cours et loin
d’être stabilisé permet néanmoins des avancées remarquables de part et d’autres. Sans sombrer dans
le ‘biais de déformation professionnelle’, nous faisons le vœu que ce projet contribue à un dialogue
constructif aboutissant à des interventions plus performantes.
11. Références mobilisées dans le rapport
Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Koszegi, B., Lichtenstein, D., Mazar, N.,
Mullainathan, S., Prelec, D., Shafir, E., Silva, J. 2005, Psychology, behavioral economics, and public
policy. Marketing Letters,16 (3): 443-454.
Anderson, T.L., 2004, Donning Coase-coloured Glasses: A Property Rights View of Natural
Resources Economics, The Australian Journal of Agricultural and Resource Economics 48(3), 445462.
Benz M., Frey B.S., Stutzer A., 2004, Introducig Procedural Utility: Not Only What, but Also How
Matters, Journal of Institutional and Theoretical Economics, 160: 377-401.
Beretti, A., Grolleau, G., Do Past States and Causes of Soil Degradation Affect Stated Willingness-toPay Values? Evidence from a Quasi-Experimental Survey, Soumis pour publication à Journal of
Environmental Management.
Beretti, A., Figuières, C., Grolleau, G., An instrument that could turn crowding-out into crowding-in,
Soumis pour publication à Economica.
Beretti, A., Figuières, C., Grolleau, G., 2013a, Using money to motivate both "saints" and "sinners": a
field experiment on motivational crowding-Out, Kyklos, 66, 63-77.
Beretti, A., Figuières, C., Grolleau, G., 2013b, Behavioral innovations: The missing capital in
sustainable development?, Ecological Economics, 89: 187?195.
Bolle F., Otto P.E., 2010, A Price Is a Signal: on Intrinsic Motivation, Crowding-out, and Crowding-in,
Kyklos, 63: 9-22.
Programme GESSOL
18/24
05/2013
Bougherara, D., Costa, S., Grolleau, G., Ibanez, L., Do Positional Preferences Cause Welfare
Gains?, Under revision at Economics of Governance.
Bourgeois-Gironde, S., Doazan, Y., Figuières C., 2010, Le prix de la biodiversité et le coût de la
répugnance morale, Mimeo.
Bullock, C., Kretsch, C., Candon, E., 2008, The Economic and Social Aspects of Biodiversity,
Benefits
and
Costs
of
Biodiversity in
Ireland,
Report,
Government
of
Irland,
http://www.cbd.int/doc/case-studies/inc/cs-inc-ireland-en.pdf.
Bulte, E.; Gerking, S.; List, J.A.; de Zeeuw, A., 2005, The Effect of Varying the Causes of
Environmental Problems on Stated WTP Values: Evidence from a Field Study, Journal of
Environmental Economics and Management, 49: 330-342.
Chen, X.; Lupi, F.; He, G.; Liu, J., 2009, Linking social norms to efficient conservation investment in
payments for ecosystem services, Proceedings of the National Academy of Sciences of the United
States of America, 106(28): 11812-11817.
Clot, S., Grolleau, G., Ibanez, L., Ndodjang, P., A paraître, L'effet de compensation morale ou
comment les « bonnes actions » peuvent aboutir à une situation indésirable, Revue Economique.
Coase, R.H., 1960, The Problem of Social Cost, Journal of Law and Economics 3: 1-43.
Déprés C, Grolleau G, Mzoughi N., 2008, Contracting for Environmental Property Rights: The Case
of Vittel, Economica, 75(299): 412-434.
Fehr, E. and K.M. Schmidt, 1999. A theory of fairness, competition, and cooperation, Quarterly
Journal of Economics, 114, 817-868.
Frederick, S., G. Loewenstein, R. O’Donoghue, 2002, Time discounting and time preference: a
critical review, Journal of Economic Literature, XL, 351-401.
Frey B.S., Oberholzer-Gee F., 1997, The Cost of Price Incentives: An Empirical Analysis of
Motivation Crowding- Out, The American Economic Review. 87: 746-755.
Gneezy U., Rustichini A., 2000, A Fine Is a Price, The Journal of Legal Studies. 29 (2000) 1-17.
Gowdy, JM., 2008, Behavioral economics and climate change policy, Journal of Economic Behavior
and Organization, 68(3-4): 632-644.
Gregory, R.; Lichtenstein, S.; MacGregor, D., 1993, The Role of Past states in determining
Reference Points for Policy Decisions, Organizational Behavior and Human Decision Processes, 55:
155-206.
Griskevicius, V., Tybur, J., Van Den Bergh, B., 2010, Going Green to Be Seen: Status, Reputation
and Conspicuous Conservation, Journal of Personality and Social Psychology, 98(3): 392-404.
Grolleau, G., 2013, The French Case Study: The Vittel Case, Etude de cas, Providing Agrienvironmental
Public
Goods
through
Collective
Action,
OECD
Report,
COM/TAD/CA/ENV/EPOC(2012)11/FINAL.
Grolleau, G., McCann, L., 2012, Designing watershed programs to pay farmers for water quality
services: Case studies of Munich and New York City. Ecological Economics, 76, 87-94.
Grolleau, G., Mzoughi, N., Helping People to Do what They Should by Suggesting Future
Implementation: Does it Work?, Soumis pour publication à Journal of Environmental Management.
Hardisty DJ.; Johnson, EJ.; Weber, EU., 2010, A Dirty Word or a Dirty World? Attribute Framing,
Political Affiliation, and Query Theory, Psychological Science, 21(1): 86–92.
Kahneman, D., 2003, A Psychological Perspective on Economics, American Economic Review,
93(2): 162-168.
Kahneman, D.; Tversky, A., 1981, The Framing of Decisions and the Psychology of Choice, Science,
211, 453-458.
Khan, U.; Dhar, R., 2006, Licensing Effect in Consumer Choice, Journal of Marketing Research, 43
(2): 357-365.
Programme GESSOL
19/24
05/2013
List, JA., 2011, Why Economists Should Conduct Field Experiments and 14 Tips for Pulling One Off,
Journal of Economic Perspectives, 25(3) : 3-16.
Maoz, I., 2006, The effect of news coverage concerning the opponents’ reaction to a concession on
its evaluation in the Israeli-Palestinian conflict, The International Journal of Press/Politics, 11: 70-88.
Mazar, N., Zhong. C.,2010, Do green products make us better people? Psychological Science, 21(4):
494-498.
McCann, L., Colby, B., Easter, KW., Kasterine, A., Kuperan KV., 2005, Transaction Cost
Measurement for Evaluating Environmental Policies, Ecological Economics, 52(4): 527-542
Mnookin, R., 2006, Surmonter les obstacles dans la résolution des conflits, Revue Française de
Gestion, 154 : 237-254.
Rogers, T., Bazerman, M.H. 2008. Future lock-in: Future implementation increases selection of
‘should’ choices, Organizational Behavior and Human Decision Processes, 106 (1): 1-20.
Ross, L., Stillinger, C. 1991. Barriers to conflict resolution. Negotiation Journal, 8: 389-404.
Thoyer, S., Said, S., 2007, Mesures agri-environnementales : quels mécanismes d’allocation ?
Etudes et Synthèses, LAMETA, ES 2007-01, http://www.lameta.univ-montp1.fr/Documents/ES200701.pdf
Salhi, S., Grolleau, G., Mzoughi, N., & Sutan, A., 2012, How can positional concerns prevent the
adoption of socially desirable innovations? Journal of Economic Issues, 46, 799-810.
Shogren, JF.; Taylor, LO., 2008, On Behavioral-Environmental Economics, Review of Environmental
Economics and Policy, 2(1): 26–44.
Suntein, C., Thaler, R., 2008, Nudge: Improving Decisions about Health, Wealth, and Happiness,
Yale University Press.
Venkatachalam, L., 2008, Behavioral Economics for Environmental Policy, Ecological Economics, 67
(4): 640 - 645.
Vollan B., 2008, Socio-ecological explanations for crowding-out effects from economic field
experiments in southern Africa, Ecological Economics, 67: 560-573.
Woodyard, C., 2009, Hybrid car owners drive more and get more traffic tickets, Usa Today,
http://content.usatoday.com/communities/driveon/post/2009/07/6849 4710/1.
12. Liste des opérations de valorisation et de transfert issues du contrat
PUBLICATIONS SCIENTIFIQUES
Publications scientifiques parues
Beretti, A., Figuières, C., Grolleau, G., 2013, Using money to motivate both "saints" and "sinners": a
Beretti, A., Figuières, C., Grolleau, G., 2013, Behavioral innovations: The missing capital in
sustainable development?, Ecological Economics, 89: 187?195.
Publications scientifiques à paraître
Clot, S., Grolleau, G., Ibanez, L., Ndodjang, P., L'effet de compensation morale ou comment les
‘bonnes actions’ peuvent aboutir à une situation indésirable, Revue Economique.
Programme GESSOL
20/24
05/2013
Publications scientifiques prévues
Beretti, A., Grolleau, G., Do Past States and Causes of Soil Degradation Affect Stated Willingnessto-Pay Values? Evidence from a Quasi-Experimental Survey, Soumis pour publication à Journal of
Plusieurs autres publications sont envisagées ou en cours de rédaction notamment sur d’autres
biais mentionnés dans ce rapport (effet de compensation morale, réaction dévaluatoire, effet des
sanctions sur les comportements pro-environnementaux, etc.)
COLLOQUES
Participations passées à des colloques
Beretti, A., Grolleau, G., 2010, Do Past States and Causes of Soil Degradation Affect Stated
Willingness-to-Pay Values? Evidence from a Quasi-Experimental Survey, International Workshop ‘A
Primer in Economic Analysis of Soil-Related Ecosystem Services’, 18 Novembre 2010, Montpellier
(France).
Grolleau, G., McCann, L., 2010, Designing Watershed Programs to Pay Farmers for Water Quality
Services: Case Studies of Munich and New York City, International Workshop ‘A Primer in Economic
Analysis of Soil-Related Ecosystem Services’, 18 Novembre 2010, Montpellier (France).
Beretti, A., Bougherara, D., Grolleau, G., Mzoughi, N., 2010, The Determinants of On-Farm Sewage
Sludge Spreading, International Workshop ‘A Primer in Economic Analysis of Soil-Related
Ecosystem Services’, 18 Novembre 2010, Montpellier (France).
Beretti, A., Grolleau, G., 2011, Do Past States and Causes of Soil Degradation Affect Stated
Willingness-to-Pay Values? Evidence from a Quasi-Experimental Survey, Experimental Community
of Southern Europe Workshop (SEET), Février 2011, Agadir (Maroc).
Beretti, A., Figuières, C., Grolleau, G., 2011, Does Directing Monetary Incentives to the Cause
Reduce Crowding-out? A Simple Field Experiment, Experimental Community of Southern Europe
Workshop (SEET), Février 2011, Agadir (Maroc).
Beretti, A., Grolleau, G., 2011, Les services écosystémiques rendus par les sols: Analyse des
régimes incitatifs sous-jacents, Séminaire de suivi GESSOL 3, Mai 2011, Tours (France).
Reduce Crowding-out? A Simple Field Experiment, International Meeting on Experimental and
Behavioral Economics (IMEBE), Avril 2011, Barcelone (Espagne).
Reduce Crowding-out? A Simple Field Experiment, Association Française d’Economie
Expérimentale Meeting (ASFEE), Mai 2011, Fort-de-France (Martinique).
Beretti, A., Grolleau, G., 2012, L’effet des biais cognitifs et comportementaux sur les politiques
publiques liées aux sols, 11èmes Journées d’Etude des Sols (AFES), 19‐23 mars 2012, Versailles
(France).
Ibanez, L., Grolleau, G., Clot, S., 2012, Do good deeds make bad people?, International Meeting on
Experimental and Behavioral Economics (IMEBE 2012). Mars 2012, Castellon (Espagne).
Ibanez, L., Grolleau, G., Clot, S., 2012, Cheap Goodwill. The best deal?, European Association of
Programme GESSOL
21/24
05/2013
Agricultural Economists, EAAE Seminar, July 2012, Castelldefels (Espagne).
Bekir, I., El Harbi, S., Grolleau, G., Sutan, A., 2012, Do You Prefer Having More or More than
Others? Comparing Evidence from hypothetical surveys and laboratory experiments, Association
Française d’Economie Expérimentale Meeting (ASFEE), June 2012, Montpellier (France).
Participations futures à des colloques
Plusieurs présentations sont prévues en fonction du degré d’avancement des papiers en cours,
notamment lors de l’ European Association of Environmental and Resource Economists (EAERE) en
2013 à Toulouse.
THÈSES
Thèses passées
Beretti, A., 2012, Essais sur l’économie comportementale appliquée aux enjeux environnementaux,
Université de Montpellier 1 – Montpellier SupAgro, Janvier 2012.
Thèses en cours
Sophie, C., Application de quelques biais de l’économie comportementale aux politiques
environnementales, Soutenance prévue en 2014 (Certains essais prévus dans la thèse relève
d’opérations prévues dans le projet ECOSOL).
ARTICLES DE VALORISATION-VULGARISATION
Articles de valorisation parus
Asselineau, A., Galia, F., Sutan, A., Grolleau, G. (2013) "L'avenir de l'agriculture se joue dans un
océan bleu", La Tribune, http://www.latribune.fr, 26 février 2013.
Articles de valorisation à paraître
Néant
Articles de valorisation prévus
Plusieurs tribunes (e.g. Huffington Post, La Tribune, Le Monde) sont envisagées dans des
quotidiens papiers ou en ligne au fur et à mesure de la parution de certains articles. Des contacts
sont en cours avec une journaliste spécialisée (Anne Sophie Novel) sur les apports de l’économie
comportementale pour une conception renouvelée des politiques environnementales, notamment
par rapport aux sols. Des articles sont également envisagés dans des revues à lectorat plus large
(e.g., Responsabilité et Environnement, Aménagement et Nature).
AUTRES ACTIONS VERS LES MÉDIAS
Actions vers les médias effectuées et Actions vers les médias prévues
Voir le point précédent
ENSEIGNEMENT – FORMATION
Enseignements/formations dispensés
Etant enseignant à Montpellier SupAgro, j’ai pu proposer des séquences relatives à la gestion des
services écosystémiques rendus par les sols sous différentes formes (études de cas, cours
magistraux, mémoires, etc.), auprès des élèves ingénieurs agronomes (notamment dans la
spécialisation TERPPA (Territoires et Ressources: Politiques Publiques et Acteurs) et du master
recherche A2D2 (Agriculture Alimentation et Développement Durable). Certains des travaux menés
en compagnie de N. Mzoughi ont également été ré-investis dans certains des enseignements qu’il
assure au sein du master professionnel ‘Gestion Qualité Production Végétales’ à l’Université
d’Avignon.
Programme GESSOL
22/24
05/2013
Enseignements/formations prévus
Etant en disponibilité, rien n’est programmé jusqu’à mon retour à Montpellier SupAgro, hormis
quelques interventions ponctuelles.
EXPERTISES
Expertises menées
Grolleau, G., 2013, The French Case Study: The Vittel Case, Etude de cas dans le cadre du rapport
de l’OCDE, Providing Agri-environmental Public Goods through Collective Action,
COM/TAD/CA/ENV/EPOC(2012)11/FINAL.
Expertises en cours et expertises prévues
Néant pour l’instant
MÉTHODOLOGIES (GUIDES…)
Méthodologies produites
Plusieurs de nos recherches ont montré l’utilité des expériences pilotes permettant d’affiner très
significativement et à moindre coût les protocoles expérimentaux avant d’envisager leur utilisation à
plus grande échelle. De plus, les expériences peuvent s’avérer extrêmement utiles pour anticiper les
réactions des acteurs à certains designs de politique environnementale.
Méthodologies en cours d’élaboration et méthodologies prévues
Nous avons aussi exploré la confiance que l’on peut faire aux questionnaires quasi-hypothétiques
en comparaison des expériences monétairement motivés (incentive compatibility).
AUTRES
Le projet ECOSOL a notamment permis l’organisation de deux manifestations, l’une d’orientation
plutôt académique et l’autre plus à des fins de transfert et d’échange :
-
International Workshop ‘A Primer in Economic Analysis of Soil-Related Ecosystem
Services’, 18 novembre 2010, Montpellier (France)
La gestion (plus) durable des sols, Demi-journée d’échanges, 22 juin 2012, Montpellier
(France)
Un compte rendu a été diffusé aux acteurs présents lors de cette ½ journée d’échange.
Programme GESSOL
23/24
05/2013
Annexe : Textes des publications
Nous joignons les publications parues en lien avec le projet ECOSOL. Les restrictions concernant leur
utilisation sont celles définies par les supports de publication et auxquels les auteurs ont souscrits.
 Publications scientifiques parues
Beretti, A., Figuières, C., Grolleau, G., 2013, Using money to motivate both "saints" and "sinners": a
Beretti, A., Figuières, C., Grolleau, G., 2013, Behavioral innovations: The missing capital in sustainable
development?, Ecological Economics, 89: 187?195.
 Publications scientifiques à paraître
Clot, S., Grolleau, G., Ibanez, L., Ndodjang, P., L'effet de compensation morale ou comment les «
bonnes actions » peuvent aboutir à une situation indésirable, Revue Economique.
 Publications scientifiques prévues
Beretti, A., Grolleau, G., Do Past States and Causes of Soil Degradation Affect Stated Willingness-toPay Values? Evidence from a Quasi-Experimental Survey, Soumis pour publication à Journal of
Plusieurs autres publications sont envisagées ou en cours de rédaction notamment sur d’autres biais
mentionnés dans ce rapport (effet de compensation morale, réaction dévaluatoire, etc.)
Programme GESSOL
24/24
KYKLOS, Vol. 66 – February 2013 – No. 1, 63–77
Using Money to Motivate Both ‘Saints’ and ‘Sinners’:
a Field Experiment on Motivational Crowding-Out
Antoine Beretti, Charles Figuières and Gilles Grolleau*
I. INTRODUCTION
In a wide range of situations, individuals engage in activities for their own sake
without obvious external incentives, thereby showing intrinsic motivations (see
Frey & Jegen, 2001, for a survey). Governments and other organizations often
use additional monetary incentives to reinforce or increase these intrinsically
motivated activities. Nevertheless, several experiments showed that positive
(negative) monetary incentives can be counter-productive, leading to a decrease
(increase) rather than an increase (decrease) of the targeted activities (e.g.,
Titmuss, 1970; Frey and Oberholzer-Gee, 1997; Gneezy and Rustichini, 2000,
Ariely et al., 2009; Bolle and Otto, 2010).
This crowding-out of moral and social motivations by monetary incentives is
relatively well documented. But this literature, in its effort to attract attention on
crowding-out effects, does not pay enough attention to the likely heterogeneity
of intrinsic motivations among the population (a notable exception is Bénabou
and Tirole, 2006). In this, it takes the risk to replace the caricature of a society
populated by amoral “sinners” by that of a society populated by uniform
“saints”. More realistically, any society is made of people characterized by
various degrees of non-monetary concerns. By this we mean not only that
individuals may differ in their degree of compliance with some particular kind
*
Antoine Beretti: UMR LAMETA – SupAgro, place Pierre Viala, Bât. 26, 34060 Montpellier Cedex 1,
France. Email: [email protected]. Charles Figuières (corresponding author): UMR LAMETA,
INRA, place Pierre Viala, Bât. 26, 34060 Montpellier Cedex 1, France. Email: Charles.Figuieres@
supagro.inra.fr. Gilles Grolleau: UMR LAMETA – SupAgro, place Pierre Viala, Bât. 26, 34060 Montpellier Cedex 1, France and LESSAC, Burgundy School of Business, Email: [email protected]. The
authors greatly acknowledge the editors and anonymous referees of Kyklos for their feedback and
suggestions. We are also grateful to Naoufel Mzoughi and Marc Willinger, participants at the LEF seminar
(Nancy) and at various conferences and workshops for many helpful comments. And we thank the French
research program GESSOL sponsored by the Ministère de l’Ecologie, du Développement durable,des
Transports et du Logement, the Agency for the Environment and Energy Management (ADEME) for
financial support. The usual disclaimer applies.
© 2013 Blackwell Publishing Ltd., 9600 Garsington Road,
Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
63
ANTOINE BERETTI/CHARLES FIGUIÈRES/GILLES GROLLEAU
of intrinsic motivation, such as altruism, but that several logics of intrinsic
motivations – not only altruism but also shame, guilt, self-esteem, reputation
concern, etc. – may be at work to varying degrees in different individuals.
Neglecting this heterogeneity might distort our understanding of the crowdingout phenomenon at the aggregate level. For instance, this phenomenon might be
wrongly attributed to the interaction between extrinsic incentives and altruism
while it may be that, for the subject under study, what is observed is the net
effect of extrinsic incentives faced with several intrinsic motivations with only
a minor effect of altruism. And our conception of effective public policies is
also at stake because the same incentive design applied to all that would be
suitable for a particular kind of “saints”, for instance to activate agents’ altruism, would probably be suitable neither for, obviously, the “sinners” nor for all
the “saints”.
This paper postulates the existence of several intrinsic motivations among
heterogeneous contributors to the natural environment. More precisely, we consider only two different intrinsic motivations, which is sufficient for our
purpose. To altruism often cited as an intrinsic motivation in previous studies
(such as Bolle & Otto, 2010), we suggest to add a role for moral repugnance
inherently associated to monetary rewards. From this framework we conjecture
some important implications regarding the observability of aggregate crowdingout effects and regarding adequate policy designs in favour of the environment.
Our originality, at the level of individual decisions, is thus to rationalize the
crowding-out effect as the interplay of three logics: (i) extrinsic incentives, (ii)
distorted altruistic motivations and (iii) moral repugnance. A second originality,
as far as policy design is concerned, is that we go further than most papers
which take note of crowding-out but do not suggest alternatives other than
eliminating monetary incentives. Indeed, we scrutinize the strength of the
crowding-out effect under three modes of interventions: (A) a direct monetary
reward for participation, (B) monetary incentives directed to the alleged cause
underpinning the intrinsic motivation(s) rather than to the individual, (C) offering the choice to participants (rather than choosing for them) to direct the
money to themselves or to the cause. We implement an original field experiment to investigate the effect of our alternative incentive strategies on motivational crowding-out, in a context where environmental intrinsic motivations are
presumably activated. Randomly selected household are simply invited to
undertake an action related to environmental motivations under different
experimental treatments.
The remainder of the paper is organized as follows. Section 2 positions our
work in the literature. Section 3 gives more details about the theory of behaviours underlying the present paper, from which it deduces the theoretical properties of the three policy designs A, B & C. Section 4 reports the results of a
field experiment designed in order to test those properties. It shows that in
64
© 2013 Blackwell Publishing Ltd.
USING MONEY TO MOTIVATE BOTH ‘SAINTS’ AND ‘SINNERS’
most cases (i) monetary incentives (design A) has a significant negative impact
on participation, (ii) even when the monetary incentive is diverted to a proenvironmental association (design B), the negative impact is weakened but
participation remains lower (or similar) than without any incentive, and (iii)
giving the choice between A and B (design C) is the most efficient policy,
which can even turn crowding-out into crowding-in. The last section
concludes.
II. CONTRIBUTION TO THE LITERATURE
There is a sizeable and growing literature, especially in psychology and economics (e.g., Deci et al., 1999; Frey and Jegen, 2001; Bowles, 2008; for
reviews) where the crowding-out effect has been explored. Without purporting
to be exhaustive, most papers can be classified along the following categories:
i) defining, debating and refining the concepts of crowding-out and intrinsic
motivations and the circumstances under which this effect is likely to occur, ii)
testing empirically the existence and magnitude of crowding-out in different
contexts.
While the theoretical possibility of crowding effects is widely accepted
among economists, many of them have been critical about its empirical relevance, in spite of compelling evidence based on numerous circumstantial
insights, laboratory studies and field experiments (Frey and Oberholzer-Gee,
1997; Gneezy and Rustichini, 2000; Frey and Jegen, 2001; Reeson and Tisdell,
2008; Bolle and Otto, 2010; Gneezy et al., 2011). Simply put, external incentives can change the perceptions people have about a given task, notably by
turning goodwill into a market-like interaction leading to a decrease of the
overall contribution, contradicting the hypothesis of motivations additivity. In
these cases, rather than reinforcing the pre-existing motivation, the crowdingout effect dominates the traditional relative price effect making the end-result
counter-productive. The crowding-out effect is more likely to occur when (i) the
external intervention is perceived as controlling rather than supportive, (ii) there
is a high level of self-determination of individuals and (iii) there is a high level
of trust between people (Vollan, 2008)1. Moreover, it has been shown that this
effect lasts even after removing the monetary incentives and is contagious
to fields which were not initially targeted by the external intervention.
Nevertheless, recent empirical works showed evidence of the role played by the
1.
Interestingly, most empirical studies have examined the effects of positive monetary rewards (e.g., Frey
and Oberholzer-Gee, 1997; Mellström and Johannesson, 2008) while some scarce studies have considered the case of negative rewards (e.g., Gneezy and Rustichini, 2000). For example, Frey and
Oberholzer-Gee (1997) found that acceptance of “not in my backyard” projects is reduced if monetary
rewards are offered and Gneezy and Rustichini (2000) showed that a fine for picking children up late
from day care increased the number of late pick-ups.
65
nature of external interventions. For instance, the field study conducted by Falk
(2007) showed that including a postcard in a letter when asking for donation to
support children in need generates a positive effect on donations. This is indicative of the possibility to design more subtle interventions than just offering
money to encourage participation.
Conceptually, our work is mostly related to the recent contribution by Bolle
and Otto (2010). In their model of behavior, the representative agent is
endowed with preferences that account for both extrinsic and intrinsic motives.
Intrinsic motivation goes through altruism that renders a certain amount of
service to other agents desirable, absent any monetary rewards for doing so. In
this context, the occurrence of a motivational crowding-out is linked to the
perceived “benefit” of altruism, which is lowered down once a payment is
introduced.
But intrinsic motivations are many and varied. They are not mere synonyms
for altruism. Intrinsic motivations frequently refer to motivations that come from
inside an individual rather than from an external reward, such as money or social
recognition. The motivation comes from the sense of satisfaction in doing a task
per se. Others consider that intrinsic motivations refer to preexisting motivations
that are motivations that were present before the introduction of external
rewards. With such a definition, it is clear that “intrinsic motivation” covers a
variety of logics guiding behaviors, ranging from self-esteem to conformity to
social norms.
At the conceptual level, the distinction between Bolle & Otto (2010) and the
present work lies in the presumption that different modes of regulation may
interact with different intrinsic motivations, and therefore incur different risks of
motivational crowding-out2.
At the empirical level, another distinction is also to be made between our work
and that of Bolle & Otto (2010). They consider the correlation between a stated
contribution to the provision of a public good (blood donation to the Red Cross)
and the level of altruism (money donation to charity). But this differs from our
approach in that their subjects state their preferences about a hypothetical contribution under hypothetical incentive schemes, while in our experimental setting
subjects are revealing their preferences through a real contribution under real
incentive schemes.
2.
66
To complicate even further the matter, the nature of the contribution and the type of the prosocial activity
under consideration can have a critical influence on the distribution of intrinsic motivations within the
population. There might even be no intrinsic motivation attached to the contribution in some given
domains. Furthermore the persistence of crowding-out when the incentive is withdrawn shows that even
in identical settings this distribution is not stable over time for a given population (Gneezy and
Rustichini, 2000).
III. AN INFORMAL CONCEPTUAL FRAMEWORK OF CROWDING OUT
WITH HETEROGENEOUS AGENTS
This section briefly gives some precisions about the theory of behaviors we have
in mind when agents have several intrinsic motivations. Then, based on this
theory it sketches the logical implications of various external monetary incentives on contributions.
As in Bolle & Otto (2010), agents are endowed with preferences that render
participation to an environmental task attractive, to some extent, because it
produces advantages to other agents, even though there is an opportunity cost
attached to this participation. This is a sort of ecologically-mediated, or ‘green’,
altruistic concern. But our conceptual framework differs from Bolle & Otto
(2010) by (i) considering that monetary incentives can activate moral repugnance, as a driving force of behaviors in addition to altruism, (ii) assuming
heterogeneity (of altruism) among people.
Moral repugnance mentioned in point (i) is a notion that we borrow from other
disciplines like cognitive sciences, moral philosophy and psychology where it is
sometimes called the yuck factor (see for instance Kelly, 2011). It refers to a
particular kind of disgust because of its relation with moral injunctions. In
discussion of bioethics, Leon Kass argues that it is “an emotional expression of
deep wisdom, beyond reason’s power fully to articulate it.” (Kass, 2002). Critics
have attacked the possibility of a so-called “wisdom of repugnance”, in particular because of the risk to ground moral values on emotions rather than on reason.
Recall indeed that, in the 20th century, racism and antisemitism have been
supported by popular revulsions. However, for the purpose of our analysis, the
ability of this particular kind of emotion to indicate social goodness is not
crucial. What is at stake is the relevance of this yuck factor as a driver for
individual behaviors.
Point (ii) is not merely a concern for realism; combined with Point (i), it is
necessary in order to understand the multi-facet nature of behaviors. The
crowding-out effect does not operate via identical behavioral channels under
different treatments, and this will prove useful to understand the advantage of
those kinds of regulation that mix up different kinds of incentives in presence of
intrinsic motivations.
Given this multiplicity of intrinsic motivations, let us now go through predictable behaviors in face of various instruments.
III.1. Neutral Treatment: pristine altruism alone
The neutral treatment is the benchmark case: there is no monetary incentive
scheme. Agents differ in their degree of green altruism and they make a
67
trade-off between the estimated “benefits” that their participation to an environmental goal produces to others and the self-centered opportunity cost of this
participation.
Under this treatment altruism is the only intrinsic motivation at work, and all
the agents who have a sufficient degree of altruism choose to contribute.
III.2. Direct Treatment (A): distorted altruism and moral repugnance
Under this treatment individuals are paid for their contribution to the environment. This payment has two potentially adverse effects on participation: i)
as in Bolle & Otto (2010) there is a signal effect because the monetary
payment reduces the individual estimation of the benefits of participation on
other individuals, ii) in addition, putting a price onto a territory previously
immune to the market forces is one of the list of events that generally spark
the yuck factor argument (see for instance Sandel, 2012, and also the
discussion about obnoxious markets in Kanbur, 2001). In other words, the
payment activates a moral repugnance component in agents’ preferences that
reduces the incentives to participate. Let us call this the moral repugnance
effect.
As far as the negative signal effect is concerned, other things equal the lower
the monetary payment, the lower the green altruistic incentives on participation
and the larger the crowding-out phenomenon, compared to the neutral treatment.
In the opposite, if the payment is increased crowding-out is reduced and, pushing
this to its logical limit, if the payment gets sufficiently large, extrinsic motivations can even dominate and result in crowding-in.
Regarding the moral repugnance effect, it is worth noting that it is somehow
connected to, though distinct from, altruism. Indeed we may expect that agents
with higher degrees of altruism are also those who have the sharper moral
repugnance. Therefore the direct treatment not only discourages those with an
insufficient degree of altruism – the negative signal effect mentioned above – but
it also repels a share of the population characterized by the larger degrees of
green altruism, because for them, the altruistic benefits of participation are more
than outweighed by their moral repugnance of being paid.
This two-end effect of direct monetary rewards has an interesting implication. When payment is increased it may well be that agents who start to participate are less altruistic than those who start to participate under the neutral
treatment, and yet there is crowding-out because this recruitment of low-end
altruistic agents is not enough to compensate the loss of participation from
those with the highest degree of altruism in the population. Let us keep in
mind this possibility that will turn out to be important for the properties of the
last instrument.
68
III.3. Indirect Treatment (B): distorted altruism alone
Under this design the payment is no longer given to contributors; rather it is
directed to a cause supporting the environment, for example a related association.
Individuals no longer bear the cost of moral repugnance, therefore there won’t be
crowding-out anymore among the most altruistic individuals in the population.
But of course the agents’ altruistic motivation is activated, for participation
still generates a benefit to the environment. Because the payment is directed to
the cause supported by the individual, we can reasonably consider that the
individual’s estimation of the environmental benefit of his participation is higher
than when the same amount of money is directed to the individual’s pocket; the
chosen destination, by its very nature, reinforces the belief of the agent on the
presence of high environmental values, or because the association is more
efficient than individuals in transforming a given amount of contributions in
environmental gains.
So, one can expect that the indirect treatment mitigates the negative signal
effect of the monetary reward, and crowding-out is less important compared to
the direct treatment. But participation under the indirect treatment is not any
larger than under the neutral treatment.
III.4. Choice Treatment (C): auto-selection of motivations
Under this treatment, individuals can choose whether the payment is directed to
themselves or to an environmental association. Giving the choice to individuals
(keeping the reward for them or giving it to the ‘environmental cause’) could
motivate a wider set of individuals, possibly leading to a higher overall contribution. It is even possible that crowding-in is obtained under the choice treatment, whereas there is crowding-out under each separate treatment. We expect
this possibility occurs when the first individuals recruited under the direct treatment are less altruistic than those who start to contribute under the neutral
treatment (and the indirect treatment), and because the most altruistic individuals
who do not participate under the direct treatment do contribute under the indirect
treatment. Crowding-in is obtained because the choice treatment combines the
positive recruitment effects that each separate treatment has on different subsets
of the population.
IV. EMPIRICAL ANALYSES
IV.1. Testable predictions
The previous simple theory of behaviours conveys general qualitative predictions
that empirical analyses can test and possibly refute:
69
• Conjecture 1: When an environmental-related intrinsic motivation preexists, a newly and ‘weak’ monetary reward crowds out this pre-existing
motivation, possibly leading to an overall reduction of the motivation to act
and hereafter to an overall decrease in contribution unless the monetary
reward is sufficiently important.
• Conjecture 2: A monetary reward that is not targeted directly at the individual but at the cause supported by the individual’s behaviour can mitigate
the crowding-out effect of a direct monetary reward.
• Conjecture 3: When a choice is given between a direct and an indirect
incentive, the overall contribution is higher than without choice at the same
level of monetary incentive, and can even result in crowding-in if the extrinsic incentives are strong enough.
IV.2. Experimental design
In order to investigate the above predictions, we designed a quasi-field experiment. The environmentally related contribution was a simple CO2 emission
compensation linked to a participation in a survey on pro-environmental behaviors. The principle of CO2 emission compensation tends to spread more and more
each day, among institutions, firms, households, and is today widely accessible.
In 2010 a total volume of 131 MtCO2e were traded on the market for voluntary
carbon offset for an overall value of US$ 424 million, a growth by 34% from
previous year. CO2 emissions are displayed on many consumption goods and
compensations are often suggested (plane travel for instance) if not tied to the
consumption act (many pro-environmental products). In January 2009, the
survey was administered to a randomly selected sample of 1498 French citizens
living in the Montpellier metropolitan area (South of France). This metropolitan
area, officially elected 2011 French Capital City of Biodiversity, was chosen
because of highly publicized events and initiatives promoting individual and
collective behavioral changes to help the environment, such as the both annual
‘Biodiversity Festival’ and ‘Week of the Environment’. The pro-environmental
behaviors surveyed are very common such as switching off the light when
leaving a room, turning off tap when brushing teeth, using a re-usable grocery
bag or purchasing eco-friendly products. These items were informed on a Likert
scale from ‘never’ to ‘always’.
Every participant received an envelope containing the one-page questionnaire,
an accompanying letter describing the overall study and presenting the incentive
offered (depending on the treatment), a reply coupon and a prepaid return
envelope. In the test phase, reading the accompanying letter and filling the
questionnaire took on average 6 mns. We wrote the letter of solicitation to make it
engaging, easy to read and promoting the benefits of results for the common good.
70
Table 1
Presentation of the Experimental Treatments
Name
N
D2
D7
I2
I7
C2
C7
Treatment
Amount given (€)
Target
N = 1118
No Reward
Direct Reward
Direct Reward
Indirect Reward
Indirect Reward
Chosen Reward
Chosen Reward
0
2
7
2
7
2
7
none
Respondent
Respondent
Association
Association
Respondent or Association
Respondent or Association
157
160
165
161
157
155
163
In order to activate intrinsic pro-environmental motives, we used a double strategy,
first by committing ourselves to offset for CO2 emissions – via the CO2Solidaire
program (see www.co2solidaire.org/) – corresponding to 300 kms by car for each
returned questionnaire and second, by stressing in the accompanying letter that a
better knowledge of pro-environmental behaviors can inform policy makers
and researchers. The letter was identical across treatments except for the variables (incentive level, money recipient) that changed from one treatment to
another.
Our experimental design allows us to compare response rates under different
treatments combining the incentive level (no reward, small reward, high reward)
for filling and returning the questionnaire and the money recipient (the individual, the ‘environmental cause’, choice between either the individual or the
‘environmental cause’). Exactly as in the analytical framework above, the treatments have been labeled either as ‘direct’ if the money is offered to the respondent himself or ‘indirect’ if the money is devoted to the environmental cause or
‘chosen’ if the respondent must choose between himself and the environmental
cause. The incentive levels were chosen to be as realistic as possible. For
example, the lower level is slightly more than twice the minimum wage rate in
France based on an estimation of the average time spent to fill (6 min) and send
back the questionnaire. We used three levels of incentive:
• a ‘no reward’ situation to control for the level of intrinsic motivation;
• a 2 euros (ª US $3) incentive corresponding to the small incentive
level;
• a 7 euros (ª US $10) incentive corresponding to the high incentive
level.
We sent 1498 questionnaires which were divided randomly into seven groups
of 214 contacts each group corresponding to one of the seven experimental
treatments presented in Table 1 below. Due to a poor update of the phone book
data and a high turnover in the region surveyed, 380 letters were returned by
71
postal services, stipulating that the person targeted does no longer live at the
indicated address at the delivery time.
After receiving all the responses, we purchased environmental offsets from an
environmental association for CO2 emissions (US $733) corresponding to the
questionnaires returned, sent back to participants their rewards (US $612) and
made a donation to an environmental association (US $806).
IV.3. Results and discussion
Between January and March 2009, we received 279 responses. The corrected
overall response rate reached 24.96%. No reminder has been sent, to avoid
introducing a bias. Table 2 provides key statistics for the seven incentives
schemes.
We compare the participation rates under different treatments and table 3
presents the one-tailed z-test for the comparison of two proportions from independent samples that we performed. Most results exposed below are consistent
with findings of previous studies.
First, consistently with Conjecture 1, the findings clearly show that there is a
crowding-out effect of a small monetary incentive on participation. When a
monetary reward is offered, participation falls by 32.4% when compared to the
benchmark case without reward. This result is significant at the 5% level
(p = 0.0264). Nevertheless, when the monetary incentive is sufficiently high the
price effect attenuates the crowding-out effect leading to a lower and even not
significant decrease of participation rate, when compared with the non-reward
case (-4.8% / p = 0.3906).
Second, with regard to Conjecture 2, offering the incentives to the cause
instead of directing it at the individual encourages participation, resulting in an
increase by 15.4% when compared to the case with a direct incentive. However,
this result is not significant at the 5% level (p = 0.2526). The most surprising
Table 2
Key-statistics regarding each treatment
Treatment
N
D2
D7
I2
I7
C2
C7
Total
72
Response rate
(%)
Women/men
ratio
Age
(years)
No high school
diploma (%)
University
education (years)
28.66
19.38
27.27
21.74
24.20
23.87
29.45
24.96
57.78
64.52
66.67
60.61
51.35
72.22
64.58
62.55
54.93
45.45
45.36
53.74
53.58
51.31
46.21
50.01
20.00
12.90
15.56
20.00
13.16
27.03
27.08
19.71
3.46
4.5
3.49
3.31
4.25
4.25
3.73
3.82
Table 3
One-tailed Z-test for the Comparison of Two Proportions from Independent Samples
Test number
1
2
3
4
5
6
7
8
9
Treatments
compared
X –Y
Value
X (%)
Value
Y (%)
Difference
X – Y (%)
z-value
Pr(Z > z)
Significant
difference?
N – D2
I2 – D2
C2 – D2
C2 – I2
N-D7
N-I7
N-I2
N-C2
N-C7
28.66
21.74
23.87
23.87
28.66
28.66
28.66
28.66
28.66
19.38
19.38
19.38
21.74
27.27
24.20
21.74
23.87
29.45
9.29
2.99
5.12
2.13
1.39
4.46
6.92
4.79
-0.79
1.9365
0.6663
1.1104
0.4517
0.2778
0.8958
1.4225
0.9614
-0.1547
0.0264
0.2526
0.1334
0.3258
0.3906
0.1852
0.0774
0.1682
0.5615
YES**
NO
NO
NO
NO
NO
YES*
NO
NO
*: significant at the 10% level; **: significant at the 5% level.
result is that directing the reward (regardless of its amount) to the cause engenders a smaller but still strong crowding-out effect, reducing participation by
24.1% when compared to the case without reward. This result is significant at
the 10% level (p = 0.0774). An increase of the size of the monetary reward
increases participation, but not significantly, by 11.3% (p = 0.3007) when compared to the small monetary reward directed to the cause. But it still corresponds to a decrease in participation by 15.6% (p = 0.1852) when compared to
the benchmark case.
As for Conjecture 3, when the choice is offered to individuals between a direct
and an indirect reward, participation increases when compared to both cases with
direct (+ 26.4%) or indirect (+ 9.8%) reward but differences are not significant
(respectively p = 0.1334 and p = 0.3258). Without offering the choice, we
induced a selection bias by attracting either money-motivated individuals or
intrinsically motivated individuals. By offering the choice, we can, to some
extent, reduce this bias by recruiting in the two types. Individuals may also value
the fact of having a choice per se which would be consistent with the theory of
procedural utility (Frey et al., 2004; Frey and Stutzer, 2005).
IV.3.1. Relative performance of the different incentive schemes
If we compare the different incentive schemes in terms of cost-effectiveness, the
findings are compelling. The only incentive scheme that allows a no-decrease
(slight increase, but not statistically significant) in participation corresponds to
the experimental treatment C7 where a higher reward is offered along with the
choice of its recipient. The marginal cost of this increase in participation rate
reaches the huge value of €383. Most of this marginal cost (€285) is directed by
respondents to the environmental cause, which can be considered as an addi© 2013 Blackwell Publishing Ltd.
73
Table 4
Tests of heterogeneity
Test number
1
2
3
Variable of interest
D2
N
t-value
p-value
Significant difference?
Age (years)
Sex (% women)
Study (years)
45.45
64.52
3.767
54.93
57.78
2.545
-2.2564
0.5870
1.8700
0.0137
0.2796
0.0330
YES**
NO
YES**
*: significant at the 10% level; **: significant at the 5% level; ***: significant at the 1% level.
tional component of effectiveness. Intuitively, rewards devoted to the environmental causes are more likely to be used for environmental improvement than
direct rewards offered to individuals. In sum, in all other cases, even if there are
(beneficial?) transfers to the environmental cause, these expenditures reduce the
response rate when compared to the no-reward case.
IV.3.2. Heterogeneity of green altruism and socio-demographic
characteristics
At this stage, we check whether individuals responding to various treatments
differ in terms of socio-demographic characteristics. Indeed, we assume that all
original sub-samples have similar socio-demographic characteristics because
they are randomly drawn from the same whole population. In theory, polar
treatments will recruit people who differ in their degree of green altruism. For
instance, neutral treatment N (i.e., no payment at all) is likely to attract a wide
array of intrinsically motivated contributors whereas treatment D2 (i.e., paying a
small amount to the individual) is likely to discourage contribution among the
most altruistic people. Altruism is not directly observable, but it might be
correlated to observable characteristic. More precisely, socio-demographic variables such as age and education could indirectly reflect the role played by the
degree of altruism. In order to investigate this issue, we performed a Student’s
independent two-sample t-test with unequal sample size and unequal variance on
three available socio-demographic variables: gender, age and education level
(Table 4).
According to Table 4, individuals who responded to treatment D2 are 45.45
years old on average, while individuals who responded to treatment N are 54.93
years old on average. This difference is significant at the 5% level (p = 0.0137).
Interestingly, D2 contributors are younger and more educated when compared
with N contributors and these differences are significant at 5% level3. This
3.
74
These results still hold if we group together some similar treatments. For instance, the results remain
consistent if we add people responding when a small amount is directed to the cause (I2) to the green
contributors group. If we add people responding under the choice treatments (C2 and C7) and directing
suggests that the most altruistic individuals are older and, perhaps because of a
generational effect, have less education. Nevertheless, these results are only
suggestive, given that we do not have robust information about non-respondents.
But they support the idea that studying heterogeneity among agents is relevant
when considering intrinsic motivations related to the environment.
V. CONCLUSION
When participation of some agents rests mainly on intrinsic motivations, the
usual monetary incentives can damage them and produce a result opposite to that
originally sought. There is a clear interest to preserve intrinsic motivations. But
the intrinsic motivations are many and varied, and coexist with extrinsic
motivations. Any society is made of heterogeneous people, ranging from “saints”
to “sinners”. A mode of regulation best designed for one category of them only,
such as directing monetary incentives to the cause, could fail to encourage the
participation of agents more responsive to other intrinsic motivations, not to
mention the agents motivated primarily by extrinsic incentives. This paper constructs a field experiment within which the properties of several incentive designs
can be empirically explored and compared.
Overall the main messages are as follows:
1. directing monetary incentives to the cause can quasi-cancel the decrease in
overall participation, but does not generate an improvement compared to a
situation without monetary incentives;
2. The negative impact is most reduced, and can even be turned into a positive
impact, when offering the choice to the respondent between directing a
large reward to the cause or to himself. The explanation might be that the
choice treatment avoids the usual single logic of regulatory instruments. A
wider set of motivations can then be recruited, thereby avoiding a counterproductive selection bias.
An interesting extension to this study will be to use gifts rather than money
per se which can avoid to stimulate a detrimental ‘market instinct’, under
which an individual expresses personal interests rather than social preferences
(Reeson and Tisdel, 2008)4, and directing them either to the individual or to
the cause.
4.
the money to the cause to the green group and those directing the money to themselves to the grey group,
the results still hold. The results of these tests are available upon request.
The individual does not consider the question “Am I willing to contribute freely to this survey for the
sake of the society and the environment?” but he switches to the question “Am I willing to earn x euros
to contribute to this survey?” which recruits a different scope of motivations.
75
REFERENCES
Ariely, Dan, Anat Bracha and Stephan Meier (2009). Doing Good or Doing Well? Image Motivation
and Monetary Incentives in Behaving Prosocially, The American Economic Review. 99: 544–555.
Bénabou, Roland, and Jean Tirole (2006). Incentives and Prosocial Behavior, The American Economic Review. 96: 1652–1678.
Bolle, Friedel, and Philipp E. Otto (2010). A Price Is a Signal: on Intrinsic Motivation, Crowding-out,
and Crowding-in, Kyklos. 63: 9–22.
Bowles, Samuel (2008). Policies Designed for Self-Interested Citizens May Undermine “The Moral
Sentiments”: Evidence from Economic Experiments, Science. 320: 1605.
Deci, Edward L., Richard Koestner and Richard M. Ryan (1999). A Meta-Analytic Review of
Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation, Psychological
Bulletin. 125: 627–668.
Falk, Armin (2007). Gift Exchange in the Field, Econometrica. 75: 1501–1511.
Frey, Bruno S., Matthias Benz and Alois Stutzer (2004). Introducing Procedural Utility: Not Only
What, but Also How Matters, Journal of Institutional and Theoretical Economics JITE, 160:
377–401.
Frey Bruno S. and Reto Jegen (2001). Motivation Crowding Theory, Journal of Economic Surveys.
15: 589–611.
Frey, Bruno S. and Felix Oberholzer-Gee (1997). The Cost of Price Incentives: An Empirical Analysis
of Motivation Crowding-Out, The American Economic Review. 87: 746–755.
Frey, Bruno S., and Alois Stutzer (2005). Beyond Outcomes: Measuring Procedural Utility, Oxford
Economic Papers. 57: 90–111.
Gneezy, Uri, Stephan Meier and Pedro Rey-Biel (2011). When and Why Incentives (Don’t) Work to
Modify Behavior, Journal of Economic Perspectives. 25(4): 191–210.
Gneezy, Uri, and Aldo Rustichini (2000). A Fine Is a Price, The Journal of Legal Studies. 29:
1–17.
Kanbur, Ravi (2001). On Obnoxious Markets, Revised version published in Stephen Cullenberg and
Prasanta Pattanaik (editors), Globalization, Culture and the Limits of the Market: Essays in
Economics and Philosophy. Oxford University Press, 2004.
Kass, Leon R. (2002). Life, Liberty, and the Defense of Dignity. Encounter Books.
Kelly, Daniel (2011). Yuck! The Nature and Moral Significance of Disgust, MIT Press.
Mellström, Carl, and Magnus Johannesson (2008). Crowding-out in Blood Donation: Was Titmuss
Right? Journal of the European Economic Association. 6: 845–863.
Reeson, Andrew F., and John G. Tisdell (2008). Institutions, Motivations and Public Goods: An
Experimental Test of Motivational Crowding, Journal Of Economic Behavior and Organization.
68: 273–281.
Sandel, Michale J. (2012). What Money Can’t Buy: the Moral Limits of Markets, Allen Lane, A/L
PENG PRESS.
Titmuss Richard M. (1970). The Gift Relationship: From Human Blood to Social Policy, Allen and
Unwin.
Vollan, Bjørn (2008). Socio-Ecological Explanations for Crowding-Out Effects from Economic Field
Experiments in Southern Africa, Ecological Economics. 67: 560–573.
76
SUMMARY
Economists recognize that monetary incentives can backfire through the crowding-out of moral and social
motivations leading to an overall decrease of the desired behaviour. Under the premise that agents are
heterogeneous and have various intrinsic motivations we suggest precise strategies to reduce counterproductive motivational crowding-out. In order to test our suggestions, we implement a field experiment where
participants are asked to fill a questionnaire on pro-environmental behaviours under different incentive
schemes, either with no monetary incentive (control) or with low or high monetary incentive directed either
to the respondents (design A) or to an environmental cause (design B), or with a choice offered between A
and B (design C). We investigate (i) whether there is a significant crowding-out effect, (ii) which design
performs better to promote participation. Except for a high monetary incentive where the respondent chooses
himself the end-recipient, we show that monetary rewards directed either at the individual or at the cause
actually harms intrinsic motivations, but not to the same extent.
77
Author's personal copy
Ecological Economics 89 (2013) 187–195
Contents lists available at SciVerse ScienceDirect
Ecological Economics
journal homepage: www.elsevier.com/locate/ecolecon
Behavioral innovations: The missing capital in sustainable development?
Antoine Beretti a, Charles Figuières a, Gilles Grolleau b, c,⁎
a
b
c
INRA, UMR 1135 LAMETA, F-34000 Montpellier, France
Montpellier SupAgro, UMR 1135 LAMETA 2, place Pierre Viala, F-34060 Montpellier, France
LESSAC, Burgundy School of Business, 29, rue Sambin BP 50608, 21006 Dijon, France
a r t i c l e
i n f o
Article history:
Received 12 April 2012
Received in revised form 28 February 2013
Accepted 3 March 2013
Available online 29 March 2013
JEL classification:
D03
Q55
Q57
Keywords:
Behavioral capital
Behavioral economics
Dual interest theory
Innovations
a b s t r a c t
Many scholars argue that environmental issues can be addressed through technological innovation, a proposal which echoes a lasting debate between environmental and ecological economics about the substitution
rate between natural and manufactured capital. In addition to these two established types of capital, this
paper introduces the idea of ‘behavioral capital’. We define behavioral capital as the latent potential of behavioral change to affect improvement in environmental quality. Our contribution argues that technological and
traditional regulatory innovations serve as insufficient tools for addressing modern environmental issues and
ensuring sustainable development. Without discarding these solutions, we contend that because human behavior is a significant contributor to environmental problems, it should be regarded as a key component of
continued solutions. We suggest that the dual interest theory can serve as an integrative framework for
behavioral innovations related to environmental issues. In suggesting this, we assume that behavioral innovations can both overcome some of the limitations of technological innovations and offer new solutions. Our
main insight is to suggest that some depletion of natural capital – but not all – can be offset by behavioral
changes without decreasing, or even increasing, subjective well-being.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
During a severe famine in the late 18th century France, locals
remained reluctant to eat potatoes despite having received encouraging
messages about their edibility and nutritional benefits. In response to
this crisis, a French pharmacist named Antoine-Augustin Parmentier
strategically planted 100 acres of fallow land with potatoes and had
royal soldiers stationed around it. However, he secretly ordered them
not to protect the planted area, but to accept bribes to leave it unguarded. With their curiosity piqued by the presence of the guards, local residents took the bait, paying off soldiers to leave their posts in order to
steal and eat the unguarded potatoes. This unconventional strategy effectively induced residents to eat the potatoes they had once refused.
Our current environmental predicament is strangely reminiscent of
the situation that Parmentier faced in 1785. The severity of modern environmental problems such as climate change, water shortages, and biodiversity losses contrasts sharply with the indifference, unresponsiveness,
and even active denial of those culpable for the damage. Some argue
that environmental issues can be addressed through technological innovation (Lomborg, 2009), a proposal which echoes a lasting debate between environmental and ecological economics about the substitution
rate between natural and manufactured capital. While standard
⁎ Corresponding author. Montpellier SupAgro, UMR 1135 LAMETA, F-34060 Montpellier, France. Tel.: +33 6 60 82 52 51.
E-mail addresses: [email protected] (A. Beretti), fi[email protected]
(C. Figuières), [email protected] (G. Grolleau).
0921-8009/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ecolecon.2013.03.004
environmental economics accepts manufactured capital as a pure substitute for natural capital, ecological economics maintains that the existing
stock of natural capital must not be depleted beyond a certain threshold.
Accordingly, this minimum stock level must be maintained and perhaps
even enhanced because the functions it performs cannot be replaced by
manufactured capital.1
In addition to these two established types of capital, this paper introduces the idea of ‘behavioral capital’. We define behavioral capital as the
latent potential of behavioral change to affect improvement in environmental quality. Our contribution, therefore, differs from that of most
studies to date by arguing that technological and traditional regulatory
innovations serve as insufficient tools for addressing modern environmental issues and ensuring sustainable development. Without discarding
these solutions, we contend that because human behavior is a significant
contributor to environmental problems, it should be regarded as a key
component of continued solutions. In suggesting this, we assume that behavioral innovations can both overcome some of the limitations of technological innovations and offer new solutions. Our main insight is to
suggest that some depletion of natural capital – though not all – can be
1
Ecological economics recognizes the importance of four primary types of capital
that contribute to human wellbeing: natural capital, human capital, built capital, and
social capital. Ecological economics further recognizes that they interact in a complex
and dynamic system (Vemuri and Costanza, 2006). Even, if some behavioral based initiatives could to some extent be classified into the existing four types of capital, we
contend that behavioral capital constitutes per se a fifth capital which interacts with
the other types of capital.
188
A. Beretti et al. / Ecological Economics 89 (2013) 187–195
offset by behavioral changes without decreasing, or even increasing,
subjective well-being. Additionally, behavioral changes often prove to
be less capital-intensive than technological innovations as a means of
precipitating change (Allcott and Mullainathan, 2010). By its very nature,
behavioral capital is centered on human behavior and can seem inconsistent with the non-anthropocentric ethics of the ecological economics paradigm. Nevertheless, considering human behavior as a powerful leverage
to reach sustainable development goals does necessarily imply that we
support nature as existing solely for the benefit of humans. Using behavioral capital is not inconsistent with considering the value of the environment per se, regardless of its human values. We contend that this
perspective can contribute to adequate management of the environment
for the environment's sake.
The remainder of this discussion is organized as follows. In Section 2,
we explain why technological change alone is not effective enough when
human behavior is at the root of an environmental issue. Section 3 proposes some solutions offered by behavioral innovations and illustrates
them using several micro-case studies and anecdotal evidence. We introduce the dual interest theory which can serve as an integrative framework for behavioral innovations related to environmental issues (Czap
et al., 2012). Section 4 makes justice to the on-going discussion between
neoclassical economists and behavioral economists and emphasizes
some challenges faced by behavioral interventions. The final section
(Section 5) extends this discussion by considering policy implications
and by developing general conclusions. Throughout the article, some
generalizable lessons are drawn from various studies.
2. Why is Technological Change Likely to be Insufficient to Address
Environmental Issues?
According to Ehrhardt-Martinez (2009), the first and most common
response from policy makers and the public alike is to look to technology
to provide answers to global climate change. Moreover, she argues that
when these technological solutions are not adopted, the next step is to
seek an economic explanation. Without negating the substantial contribution of this techno-economic model in addressing environmental issues,
we contend that it only treats one part of the problem. The recommendations drawn from this model are based on rational actors who perfectly
weight the costs and benefits of adopting certain behaviors. According
to this model, well-crafted economic incentives and disincentives can
guide people to adopt socially desirable technologies and behaviors.
While traditional economists assume that economic actors are rational
agents who are self-interested, fully-informed, outcome-oriented, and
time-consistent, most social scientists believe that human behavior is
often irrational, imperfect, uncontrolled, and time-inconsistent, yet predictable (Allcott and Mullainathan, 2010; see also Venkatachalam, 2008).
Clearly, the alleged performances that technological change can
deliver are contingent upon the reliability of the incentives devised
within the traditional (rational) model of economics. Consequently,
the goals of the techno-economic model may be unrealistic because
it neglects important psychological and emotional determinants of
human behavior (Gowdy, 2008; Venkatachalam, 2008). For instance,
the effectiveness of monetary incentives (disincentives) to encourage
(discourage) socially (un)desirable behaviors can be significantly
Table 1
Two cognitive systems (Kahneman, 2003; Thaler and Sunstein, 2009).
Automatic system
Reflective system
Uncontrolled
Effortless
Associative
Fast
Unconscious
Skilled
Controlled
Effortful
Deductive
Slow
Self-aware
Rule-following
reduced because of motivational crowding-out, inadequate timing
of payments or reactive devaluation (e.g., Duflo et al., 2011). In particular, recent research suggests that there are two distinct operational
systems in the brain (see Table 1 below). The reflective system offers
deep, rational analysis, despite its admittedly limited capabilities,
while the automatic system is able to process many concepts simultaneously but in a more ‘superficial’ way by employing heuristics and
biases (Thaler and Sunstein, 2009). This study indicates that some
of our decisions are more influenced by the automatic system than
was once thought, rendering the techno-economic model limited in
scope, as it assumes perfectly rational choice. In sum, a growing literature reveals that our intuitions about the drivers of behavior can be
flawed, notably because preferences, decision making, and resulting
behaviors are largely context-dependent (Ariely, 2008; Cialdini,
2005). Sanfey et al. (2003), for example, use functional magnetic resonance imaging to demonstrate that unfair offers in an ultimatum
game induce conflicting activity between the areas of the brain
which represent “twin demands […], the emotional goal of resisting
unfairness and the cognitive goal of accumulating money.”
Technological solutions which seem promising on paper can be
disappointing when implemented in the real world. Among the
counter-intuitive effects of technological improvements, let us mention rebound effects (Binswanger, 2001). In short, rebound effects
occur when the lower cost of energy services resulting from increased
efficiency ‘pushes’ individuals to consume more energy overall. While
there remains some disagreement about the magnitude of rebound
effects, the existence of these effects is not contested.
Additionally, technological change is frequently considered to be
highly capital intensive. A panel of Nobel laureate economists ranked
‘higher public funding and research and development of non-carbon
energy, in order of $100 billion a year’ as ‘one of the most effective responses’ to address climate change issues. However, in a context of
economic crisis, financial constraints can force agents to seek innovative solutions which are less capital intensive. Behavioral interventions, in particular, are frequently less costly than interventions
drawn from the techno-economic model, and they can deliver first
order environmental results (Allcott and Mullainathan, 2010).
3. Human Behavior: A Potential Solution to Environmental Problems?
Many behaviors, especially day-to-day ones, are frequently determined by the automatic system, making conventional interventions
(e.g. technological improvements, market incentives) less likely to
deliver desired results. We contend that by taking these mental shortcuts and biases into account, it is possible to encourage particular
types of behaviors. Well-crafted behavioral innovations can ‘nudge’
agents to make better choices and can therefore constitute powerful
solutions to sustainability issues. Two fundamental behavioral findings are essential to understand why decision makers (Humans)
diverge so systematically from those of standard economics (Econs).
The first is ‘construal’ which captures the notion that decision makers
need to construe a representation of the relevant situation in their
mind and the second is the ‘power of the situation’, implying that
such construal is heavily impacted by the context of the decision
(Shafir, 2008). Humans with their so-called behavioral anomalies
influence how the various “capitals” transform material and energy.
Most behavioral findings can be interpreted through the lens of the
new dual interest metaeconomics framework. Rather than just enumerating a list of biases by presenting them as exceptions to mainstream economics, the dual interest theory and metaeconomics approach offer an
integrative framework (Czap et al., 2012). This promising framework
notably recognizes the co-existence of two human tendencies, a tendency to egoistic–hedonistic based self-interest which needs to be tempered
by the human tendency to empathy–sympathy based other-interest.
Self-interest is considered as a force rather than presuming it as the
only motivating force and the dual interest theory acknowledges the
moral dimension reflected in norms, values, conscience and so forth, i.e.,
the we-interest or other-interest (Czap et al., 2012). This framework and
theory make explicit the “norm” being referred to in our paper. That is, it
is the other (internalized within self, but shared with others)-interest
that tempers and conditions the self-interest. Stirring and evolving the
other-interest better ensures not only environmental action now, but
also builds commitment to the shared cause over time, such that the behavior continues. In short, Humans (in contrast with Econs) have dual,
often conflicting egoistic–hedonistic (self-interest) and empathy–
sympathy (other/share with others-interest), seeking peace of mind,
and peace with nature, rather than the Econ presumption of a egoistic–
hedonistic, self-interest only “max U” person, few of which really exist
in this world and as a result, most can be nudged to act on empathy–
sympathy with the environment (Sheeder and Lynne, 2011). According
to Rifkin (2009), increased empathy would necessarily make people together to behave co-operatively and solve global problems, such as climate change.
Instead of defining a list of behavioral biases as a basis for behavioral
innovations, we have chosen to emphasize five behavioral-based interventions which can ‘nudge’ people in socially desirable directions:
social influence, causal attributions, framing effects (loss aversion,
default options), commitment and intentions manipulation, and visibility. Moreover, even if these behavioral-based interventions share the
behavioral background with those developed by previous scholars
(e.g., Gowdy, 2008; Shogren and Taylor, 2008; Venkatachalam, 2008),
we selected some interventions that were not mentioned in the preceding contributions.
3.1. Using Social Influence to Orient People in Socially Desirable Directions
As a frequently underestimated factor, social influence can be
remarkably powerful in directing people's decisions and actions
(Cialdini, 2005). However, both descriptive (what the majority does)
and injunctive social norms (what the majority approves of) are rarely
used in environmental policy design despite the fact that they are highly
effective and nearly costless. The potency of social norms was demonstrated through a simple experiment (Goldstein et al., 2008) in which
hotel guests received a card asking them to reuse their towels. Four different cards were presented according to each experimental treatment,
namely an environmental protection appeal, a social responsibility for
future generation appeal, an environmental cooperation appeal and a
descriptive norm.
Treatment
Message on the card
Environmental protection
appeal
“HELP SAVE THE ENVIRONMENT. You can show
your respect for nature and help save the
environment by reusing your towels during your
stay.”
“HELP SAVE RESOURCES FOR FUTURE
GENERATIONS. Future generations deserve our
concern. Please do your part to protect the
environment and conserve dwindling resources for
future generations to enjoy. You can help preserve
these precious resources for all of us by reusing
your towels during your stay.”
“PARTNER WITH US TO HELP SAVE THE
ENVIRONMENT. In exchange for your participation
in this program, we at the hotel will donate a
percentage of the energy savings to a nonprofit
environmental protection organization. The
environment deserves our combined efforts. You
can join us by reusing your towels during your
stay.”
“JOIN YOUR FELLOW GUESTS IN HELPING TO SAVE
THE ENVIRONMENT. Almost 75% of guests who are
asked to participate in our new resource savings
program do help by using their towels more than
once. You can join your fellow guests in this
program to help save the environment by reusing
your towels during your stay.”
Social responsibility for
future generations appeal
Environmental cooperation
appeal
Descriptive norm
189
The “descriptive norm” card increased towel reuse when compared to the other three cards. A simple and interestingly costless
presentation of what other people do – a descriptive social norm –
engendered a net increase of 34% in participation to this simple
conservation program.
In another study (Schultz et al., 2007), door hangers were placed
once a week for a month on the doors of a residential area of San
Diego. Each household received one of four messages, informing people
that (1) they could save money by conserving energy, (2) they could
save the earth's resources by conserving energy, (3) they could be
socially responsible citizens by conserving energy, or (4) the majority
of their neighbors tried regularly to conserve energy — information
learned from a prior experiment. The fourth message, carrying the descriptive social norm, was the only one that led to a significant decrease
in energy consumption: a reduction of almost 2 kWh per day (Nolan et
al., 2008). It is worth noting, however, that such use of a descriptive social norm is relevant only in cases where the desirable behavior also
stands for the norm, whereas, in any other context, disclosing this
kind of information might result in counterproductive effects.
For instance, Arizona's Petrified Forest National Park suffers from
the theft of more than one ton of wood each month by park visitors.
The local authority erected signs informing that many visitors were
vandalizing the forest every year by stealing small pieces of wood.
However, this message actually served to normalize the undesired
behavior and therefore achieved perverse results (Cialdini, 2003).
Framing, too, often plays a crucial role in determining environmental
behaviors. In response to the above perverse effect, Cialdini (2003)
conducted an initial experiment in which they placed two different
signs in high-theft areas of the park, along with a similar control
area without any sign. The first sign informed park visitors that
“Many past visitors have removed petrified wood from the Park, changing the natural state of the Petrified Forest” (a descriptive social
norm), while the second sign simply stated “Please do not remove petrified wood from the Park in order to preserve the natural state of the
Petrified Forest” (an injunctive social norm). The area with the first
type of sign exhibited the higher theft rate (8%), which proved significantly higher than either the control area (3%) or the area with the
second type of sign (1.7%). Cialdini (2003) then conducted a second
experiment, placing two new types of signs in high-theft areas. Both
signs urged visitors to not take wood. One depicted a scene of three
thieves in action, while the other showed a lone thief. The first type
of sign again induced a significantly higher theft rate than the second,
demonstrating the importance of the way in which a message is
framed. In this case, we find that revealing an undesirable social
norm, i.e., that many visitors steal wood, may in fact reinforce it as a
descriptive norm and consequently lead to an undesirable environmental outcome. This lesson is particularly relevant with respect to
alarmist messages about the detrimental environmental effects of
common behaviors. No matter how well-intentioned these messages
may be, it is clear that by confirming that an undesirable action is the
descriptive norm, these messages may only encourage such behavior,
thereby generating counter-productive effects.
It is well-known that groups and individuals are interested in their
performance not only in absolute terms, but also relative to the performances of others. Some people are willing to incur a loss (lower
absolute position) to get, in the dimension considered relevant, a
higher rank (higher relative position) in their reference group. For example, farmers can be reluctant to adopt environmentally friendly
technologies that deliver a net economic benefit for fear of losing
their position in another aspect (e.g., yields) in their reference
group (e.g. among farmers). These ‘status effects’ are frequently
ignored by neoclassical models, wherein agents maximize their profit or
their self-centered utility regardless of their position on the relevant social
ladder. Regarding farmer behavior, some recent papers (Chouinard et al.,
2008; Czap et al., 2012; Sheeder and Lynne, 2011) provide empirical evidence regarding the relevance of the dual interest theory. For instance,
190
Sheeder and Lynne (2011) show that farmers who temper their pursuit of
self-interest with shared other-interest reflecting empathy–sympathy are
more likely to adopt conservation tillage. They found support that farmers
pursue a joint and interdependent own-interest and not only self-interest
as presumed in traditional microeconomics.
These positional considerations, or peer performance comparisons,
have been used in other arenas to precipitate ‘races’ toward sustainable
development. For instance, The Yale Environmental Sustainability Index
ranks countries using a combination of 76 variables such as water quality,
overfishing, and greenhouse gas emissions. As a consequence, South
Korea, upset by its low rankings in the 2002 index, took specific actions
to improve its position by investing in reductions to air and water pollution. These efforts proved effective, moving it up 22 places within the
next three years. Yale environmental expert Daniel Esty (Ayres and
Nalebuff, 2005) describes another example: ‘After Norway came in second, the prime minister asked for a meeting to discuss the results. Rather
than crowing about Norway's superb showing, all he wanted to talk about
was what his country could do to overtake Finland and become number
one.' It is clear that ranking countries has proven to be a costless and powerful force to induce motivation for environmental improvements.
From this, we observe that the benefit of employing social influence
is the emergence of positive norms due to an increase in the visibility of
desirable behaviors. Highlighting top environmental performances may
induce virtuous competition, leading to greater environmental consciousness, and thus better behaviors. The experiment on towel reuse
in hotels (Goldstein et al., 2008) is a fitting example of this normative
effect, since the experimenter manipulated the provided descriptive
norm by communicating a 75% participation rate in the program. This
exaggerated information pushed guests to participate more, and thus
served as a normative signal.2
Lesson 1: Social influence can be used to promote sustainable development by describing the desired behavior as a social descriptive norm
when possible or as a moral prescriptive norm in other cases. Moreover,
messages describing widespread undesired behaviors can inadvertently
induce conformity and reinforce the undesirable behavior.
Lesson 2: Status seeking can be harnessed to promote sustainable
development by facilitating comparisons between entities based on environmentally friendly performances.
3.2. Using Causal Attributions: Superiority Bias and Varying Causes
Individuals make systematic and often biased inferences about the
causes of events and behaviors (Burger and Rodman, 1983; Forsyth et
al., 1981). A frequent attribution bias relies on the fact that individuals
attribute the responsibility of his successes to personal abilities, while
attributing the responsibility of his failures to external factors. Symmetrically, others' successes are attributed to external factors while others'
failures are attributed to their own responsibility. Therefore it seems
fairly intuitive that if others are understood to be responsible for an environmental degradation, a given individual will be more reluctant to
make a costly contribution to restore it or at least to avoid its further
degradation. Moreover, the superiority bias describes the tendency of
people to think that they are better than average in many dimensions.
A familiar illustration of this tendency is that most people think that
their chances of having a car accident are significantly lower than the
average person's chances of experiencing this event (DeJoy, 1989).
2
The results obtained in this study are not based on the use of deception, given the
‘75% participation rate’ was a number provided by the company that supplies such
cards to hoteliers. Deceiving is common in psychology experiments but raises several
ethical issues in economics, making its use very scarce. Moreover, the consequences
of deception overtime have to be considered. Indeed, one can imagine that the positive
effect resulting from the use of a manipulated social norm can vanish and even backfire
if people understand what the reality is.
An interesting example supporting this insight arises from an energy consumption experiment conducted in California (Schultz et al.,
2007). 3 In this study, the authors showed that when the monthly
electric bill listed the average neighborhood level of consumption,
people in above-average households significantly decreased their
consumption. At the same time, those with the below-average bills
reacted by significantly increasing their consumption. To avoid this
unanticipated increase in consumption, experimenters included an
image along with the consumption information: a smiling face (☺)
on a below-average bill or a frowning face (☹) on an above-average
bill. They appropriately labeled this effect the ‘injunction effect’. As
a result, heavy users exhibited larger reductions in electricity consumption, while light users remained frugal. In sum, helping people
to measure their performance relative to others by exposing the
true ‘average’ with an adequate framing can help people acting
more pro-environmentally. Using a descriptive social norm, an injunctive social norm, or even a combination of both can help to
guide individual consumers in a socially desirable direction.
Since environmental degradation is often the result of multiple
factors, pinpointing a single cause can be challenging. A recent
study (Beretti and Grolleau, 2011) tested the effect of the presence
and the nature of the cause of an environmental degradation on participants' interest in and willingness-to-pay (WTP) for the implementation of a restoration program. They found that the presence of a
given cause has an effect on interest, as well as on WTP. When considering man-caused degradations, individuals were more interested but
less willing to pay for the same net restoration than when considering
nature-caused degradations. They suggest that people are more
concerned (outrage effect 4) but feel less responsible when the root
of the degradation is man-made as opposed to when it is natural.
Lesson 3: Given that people misattribute causes of events and
overestimate their desirable qualities relative to others, stipulating a particular cause for an environmental issue and disclosing the ‘true’ average and
the individual's position can nudge people in more desirable directions.
3.3. Using Framing Effects: Default Option (Opt-out, Opt-in) and Loss
Aversion
A major finding of behavioral economics is the loss aversion bias
which maintains that people value a certain loss more than its equivalent net gain. Many laboratory and field experiments have supported
the fact that people are generally more motivated (around two times
more) to avoid a particular loss than to acquire the same net gain.
These findings have been used to explain the well-known WTP-WTA
disparity. Applied in the environmental realm, policies can be more or
less effective (e.g., in terms of support, of willingness to contribute)
according to whether the suggested changes are perceived as avoiding
losses or as acquiring net gains. Unfortunately, many environmental
messages are framed as sacrifices or losses for individuals in terms of
well-being (e.g. reducing car use, heating less) rather than in terms of
gains. Consequently, reframing environmental messages accordingly
constitutes a natural candidate to help increase participation in environmental preservation. Martin (2007) also recognizes the importance
of framing: ‘for messages to have the best chance of being effective it
would seem that not only should they be framed in terms of what we
stand to lose [by not adopting the desired behavioral change]. That
loss should be something we currently possess and should also be accompanied with a clear and specific action we can personally take to avoid
such as loss.’ For instance, researchers from the University of California
“told one group of homeowners that they could save 50 cents a day by
carrying out energy efficiency improvements in their home. They told
3
4
This example can also be used to illustrate the power of social influence.
See Bulte et al. (2005) for more details about the ‘outrage effect’.
a second group of homeowners that they would continue to lose 50
cents a day if they did nothing. The second group was 300% more likely
to carry out the improvements”. In sum, framing the benefits in terms of
loss due to the non-adoption can significantly increase the adoption of a
desirable behavior (Goldstein et al., 2007).
Another interesting and under-exploited finding of behavioral economics relates to the importance of default options. Even in situations
characterized by uncertain preferences and heavy consequences, it can
be shown that the option described as the default strongly influences individuals' decisions. Notably, the default option often inadvertently indicates the social norm. This phenomenon is reinforced by the omission
bias, which describes the tendency to judge harmful actions as less
moral than equally harmful omissions. In this way, the omission bias explains inaction because it is easier for an individual to restrain himself
from making a ‘good deed’ (e.g., signing a form to donate his/her organs)
than it is to perpetrate a ‘bad deed’ (e.g., signing a form not to donate his/
her organs). Consequently, appropriate choice architecture can be used
to encourage individuals to select the socially desirable option without
infringing on individual freedom. In 2009, organizers of the Behavior,
Energy and Climate Change Conference (BECC) proved the effectiveness
of such an architecture. At a prior BECC conference, meat was the default
option for lunch, while vegetarians were able to request a vegetarian
dish. Seventeen percent of the attendees chose the vegetarian meal. In
2009, conference organizers instead made the vegetarian lunch the default option. After this change, 80% of the 700 attendees chose a vegetarian lunch. Karen Ehrhardt-Martinez, the conference chair, exposes the
stakes behind such an experiment: “Omnivores contribute seven times
the greenhouse gas emissions, when compared to vegans.” The default
option effect is currently used in tax systems, organ donations, insurance, and membership renewals. To our knowledge, however, it has
not been used in many environmental policies, despite compelling
evidence of its power, as demonstrated in field experiments on organ
donation (Johnson and Goldstein, 2003) and more recently on green
electricity (Pichert and Katsikopoulos, 2008).
Lesson 4: People are more sensitive to losses than to gains of a
similar amount and tend to stick to default option. So, framing the benefits in terms of loss and make the desired behavior the default option can
promote sustainable development.
3.4. Commitments and Intention Manipulation
When an individual detects an inconsistency between his attitude
and his behavior, he experiences an uncomfortable psychological
state called cognitive dissonance. Interestingly, human dual interest
theory reflects the reality that the cognitive dissonance is natural, inherent in the brain. The dissonance is reflected in the human tendency to egoistic–hedonistic based self-interest (probably more primal,
more biological) which needs to be tempered by the human tendency
to empathy–sympathy based other-interest (this one also internalized within self, when it is effective, and, also biological, but perhaps
not as primal, more secondary in the brain). Both are jointly pursued,
non-separable, quite interdependent, with the dissonance resolved
once the two interests are squared, and the individual achieves a kind
of peace-of-mind. Also, both interests reflect a bit of self-sacrifice, to
make all this work. In this case, a bit of self-sacrifice in the self-interest
to achieve an environmental payoff in the other (shared with others)interest (but realizing there is also sacrifice in the environmental payoff
when one pursues “max self-interest U” only). Humans achieve a kind of
altruism in both domains of interest, as the dissonance declines, and a
“peaceful” state of mind is achieved, in this case, in peace with one's
environment (Czap et al., 2012; Sheeder and Lynne, 2011).
Beretti et al. (2009) suggest that in order to relieve this cognitive
dissonance, the individual should in theory (i) either change his behavior such that it becomes consistent with his attitudes, (ii) change
his attitudes, (iii) expose himself to selective information that ‘makes’
his behavior consistent, or (iv) remain ignorant (bliss ignorance).
191
Cornforth (2009) explains that when experiencing cognitive dissonance “people form ‘socio-psychological denial mechanisms’, meaning they overestimate costs and underestimate benefits of shifting
to less energy-intensive behavior while blaming other people's and
government's inaction.” Several economists have commented on the
potential of formally integrating cognitive dissonance into economic
theory (e.g., Akerlof and Dickens, 1982; Rabin, 1994). By asking the
individual to publicly commit in advance to a given behavior,
policy-makers can nudge individuals to reveal their attitudes toward
pro-environmental behaviors. This insight is consistent with dual interest theory. Indeed, when oneself commits to the other-interest, in
public, and others-so commit too, and, as a result, the self-interest is
tempered, conditioned, and each sacrifices a bit in that domain
(Sheeder and Lynne, 2011). This can help people to avoid cognitive
dissonance by, for example, overcoming procrastination and other
time inconsistency issues. Such commitments have been successfully
used by economists and psychologists in experiments related to
speeding (Elliot and Armitage, 2006), using public transport, and
buying organic food (Bamberg, 2002). In order to increase bus use,
Bamberg (2002) asked people for specific details about a day and
time for taking the bus instead of simply asking to commit to take it
on an unspecified day. This technique, known as implementation intention, has been studied by Gollwitzer and Sheeran (2006) in a
meta-analysis of 94 independent experiments. Their results confirm
the power of this tool to increase goal achievement. In the same
vein, Rogers and Bazerman (2008) showed that people are more likely
to choose what they believe they should choose when the (binding)
choice will be implemented in the future rather than in the present.
They argue that policymakers could leverage the benefits of future
lock-in by advocating for reforms that would be decided upon in the
present, but go into effect in the future.
Interestingly, the addition of a financial incentive to commitments
does not necessarily increase the strength of the ‘nudge.’ The introduction of monetary incentives can evoke a business frame and transform the nature of the considered decision from a moral one to an
economic one. Shifting the individual to think in terms of monetary
incentives also shifts the individual over to the self-interest domain,
which requires even more monetary payment. This shift has been
found to be descriptive of interventions to bring about more recycling
behavior, where “too high” financial incentives can actually result in
less recycling (Kalinowski et al., 2006; see also Sheeder and Lynne,
2011). In the same vein, Messick and Tenbrunsel (1999) showed
experimentally that the introduction of a weak monetary sanction
increased fraudulent behavior when compared to a situation without
monetary sanctions. The authors suggest that this introduction led
people to base their behavioral decision on a cost–benefit analysis
rather than on what is considered the morally right thing to do.
Dual interest theory can easily explain these paradoxes, ironies, as
individuals are “nudged” from one domain of interest to the other.
According to Self-Determination Theory (Ryan and Deci, 2000),
people are indeed less motivated when they sense that a behavior
arises due to external pressure, and more motivated when they feel
their motivation is self-generated. From these results, it appears
that internalizing commitments is a key factor in determining behavioral intention. Based on this research, senior economists at Yale University created an online commitment device (http://www.stickk.
com/) which encourages the individual to make a very specific type
of contribution, that is a monetary contribution that will be given to
an ‘anti-charity’ or an organization the individual dislikes if he does
not reach his stated goal, thereby triggering the loss aversion
discussed above. The approach also recommends the involvement of
a referee who is either a relative or another member of the online
community. Additional effects of social exposure on individual
decision-making are discussed below.
Lesson 5: People make less suboptimal decisions when they make
choices further in advance of their consequences, so asking people to
192
commit in advance to a given behavior, preferably precisely and publicly,
can increase its adoption.
3.5. Visibility and Exposure Effect
Environmental outcomes caused by individual behaviors are
frequently intangible and delayed in time. For instance, driving too
quickly contributes to increases in CO2 emissions and private costs
that are not immediately perceptible at the individual level. Even if
a cost increase is perceptible later, it is easily attributed to other
causes (causal ambiguity). Despite good intentions, people often fail
to attain their goals because they get derailed along the way due to
a lack of progress markers. Behavioral sciences show that people are
more likely to adapt their behavior when they receive feedback in
real time and in an easily accessible and understandable way. For example, the Prius mile-per-gallon display is an innovation that makes
motor performance immediately visible. Sarah Darby argues that
“once you start making fuel consumption more visible, you have
something that comes to the forefront of people's minds instead of
lurking in the background. This gives you feedback that alters actions,
and encourages you to try and improve things” (Rosenwald, 2008).
Equivalent devices (e.g., Kill a Watt plug, Wattson console) are
currently implemented in various countries to inform people on
their electric consumption at home, in real time, and sometimes in
comparison with peers.
The disclosure of a particular environmental measure could trigger some of the above discussed effects. For instance, Derek Foster,
a post-graduate at University of York, developed a simple application
which uses a personal energy meter to record live and historical
household energy consumption and then transmits this data to
Facebook (Foster et al., 2010). He concludes that ‘by using Facebook
as the delivery platform, it introduces social psychology elements
such as peer-pressure and normative behavior between friends
which can introduce competitiveness to reduce energy usage’
(University of York, 2009). In our estimation, experimental tests of
the exposure effect need to be undertaken to adequately measure
the extent to which the desired behavior can be induced using
these forces at a low cost (relative to the cost of alternatives to
reach the same result). Even if the ranking effect and social norms
have already proven to be effective, we contend that the acceptability
of the public disclosure of private information about energy
consumption remains contentious, especially among heavy users.
Nevertheless, all environmental dimensions (electricity or water
consumption versus biodiversity losses) are not equally amenable to
measurement and feedback because of measurement easiness or
temporal dimension.
Lesson 6: Giving people an immediate feedback on the consequences
of their behavior and allowing social comparison by increasing visibility
can increase the adoption of socially desirable behaviors.
4. Some Critiques of Behavioral Economics
Despite its recent success, behavioral economics have been criticized
on several grounds, notably by eminent economists (e.g., Aumann in
Hart (2005), Rubinstein (2006) Glaeser (2006) and Becker in Clement
(2002) and Herfeld (2012)) and even in trade press (e.g., Ferguson
(2010), Lehrer (2011), Farrell and Shalizi (2011)). Even if these criticisms are not directly related to behavioral environmental economics,
they can easily be extended to policies drawn from behavioral studies
in the environmental realm. So, to make justice to this current debate
among economists and practitioners, we will present and discuss
some major points of debate. The presentation will be synthetic, bearing
in mind that reviewing the whole range of criticisms and responses
related to behavioral economics is beyond the scope of our article.
First, several economists emphasize that many insights of behavioral economics have been tested in laboratory experiments or in
settings that do not reflect real-world situations. Aumann (2005,
interviewed in Hart, 2005) argues that lab experiments are artificial,
notably because they use small monetary payoffs, proposes unusual
decisions upon an act-by-act basis. Moreover, these experiments frequently use limited samples of students from Western universities to
support their theories, leading some critics to describe behavioral
economics not as ‘the study of humans in markets’ but as ‘the study
of kids in psychology labs’ (Ferguson, 2010). For instance, Gary Becker
(interviewed in Clement, 2002) stated that ‘economics theory is not
about how people act in experiments, but how they act in markets’.
Rubinstein (2006) goes further and considers that some experiments
are inherently flawed and over-generalized. Moreover, he found that
behavioral economists use a double standard: severe scrutiny for neoclassical theories and overconfidence and lack of critical assessment
for behavioral contributions. In some cases, the behavioral anomalies
crop up in one-shot experiments while they disappear in repeated experiments. Similarly, individual or household behavior can conform
to the non-rational behavior in the short run, but this behavior can return to the rational behavior (in the neoclassical sense) in the long
run. Given that environmental policies aim at achieving long term sustainable development, behavioral findings from one-shot experiments
and in the short run may be less relevant. While over-generalization
from one-shot experiments on small samples of very specific individuals5 can be dangerous, several economists agree that experiments,
given that some conditions are respected, can be a relevant source of
data to reach a better understanding of human behavior (Levitt and
List, 2007). For instance, Falk and Heckman (2009) argue that empirical
methods and data sources are complements, not substitutes. Indeed,
causal knowledge requires controlled variation and experiments, both
lab and field, play a significant role to improve the state of knowledge
in the social sciences. They reject the idea of a hierarchy among these
methods and conclude that the issue of generalizability of results is
universal to all of them.
Second, behavioral economics is frequently perceived as a list of
effects, not as a unified theory of human behavior that has the explanatory and predictive power of standard theories. There are multiple
behavioral explanations for many types of behaviors. ‘At its core it is
a collection of psychological phenomena—norms, biases, and heuristics—that are connected only in the sense that each runs counter to
some fundamental tenet of traditional economics’ (Ellis and Hayden,
2005). Rubinstein (2006) describes the typical behavioral contribution as messy, long with poorly formulated models which lack the
elegance and generality of neo-classical models. Rosch (2010)
disagrees with this reasoning and argues that ‘rejecting behavioral
economics whole cloth for lack of an organizing principle is arguably
just another way of saying that an answer is better than no answer,
however wrong that answer may be’. Moreover, there are some
attempts to provide an integrative framework to behavioral findings.
Metaeconomics or dual interest theory is a promising candidate.
Nevertheless, the process is on-going and it is preliminary to predict
whether it will be successful.
Third, several concerns are related to policy insights drawn from behavioral studies. One of the strongest criticisms against policies based
on behavioral economics is the risk of paternalism, fear of slippery
slopes and manipulation (Glaeser, 2006; Rizzo and Whitman, 2009).
Accepting new paternalist policies based on behavioral economics findings creates a risk of accepting, in the long run, greater restrictions on
individual freedom than have been heretofore acknowledged. Thaler
and Sunstein (2009) recognized the risk of slippery-slope and stressed
that this criticism does not address the issue of whether their ‘proposals
have merit in and of themselves’. Moreover, they emphasize that their
suggested policies limit the steepness of the slope and are designed to
retain the freedom of choice. They also advance that in many situations,
5
WEIRD: Western Educated Industrial Rich and Democratic.
some kind of nudge is inevitable. Indeed, there are always default rules
and contexts that frame choices in certain directions. Moreover, behavioral economics models predict irrationality among regulated entities,
but seem to ignore that regulators are subject to the same kind of biases,
heuristics and so on. This asymmetric application of behavioral theories
raised concerns (Farrell and Shalizi, 2011; Glaeser, 2006). For instance,
Posner (2009) argues as follows: “behavioral economists are right to
point to the limitations of human cognition. But if they have the same
cognitive limitations as consumers, should they be designing systems
of consumer protection?” Nevertheless, this criticism ignores the fact
that, unlike human beings who make decisions in a vacuum, government regulators have the ability to study over time how many individuals behave in certain settings (Rosch, 2010).
As stated before, the aim of this part was to mention explicitly the
current debate regarding the use of behavioral economics. We do not
pretend to give a definitive answer, but indicating some elements of
the debate is a means to give a more critical and balanced view of
so-called ‘behavioral innovations’, especially among the academic
community. In sum, while most traditional economists tend to think
of environmental matters as driven primarily by prices (even by
their absence) we advocate for a more systematic consideration of
other important drivers of human behavior.
5. Policy Considerations and Conclusion
‘The main reason behavioral science should be part of the policy
debate is that it provides in some cases a perspective that is vastly
different from economics.’ (Amir et al., 2005)
Behavior is often at the root of environmental issues, even in the
presence of technological solutions. In this paper we contend that in
some circumstances, behavioral innovations represent a cost-effective
and under-exploited method by which to address these issues.
Influencing behaviors in this way has the potential to result in better
environmental outcomes at a lower cost compared to expensive
research on technology, as we've seen in reducing energy consumption.
However, a lack of unifying theory, coupled with the generally uncertain nature of psychological factors, has limited the usefulness of
research to merely drafting strategies and raising questions regarding
scalability of the discussed effects. The dual interest theory and
metaeconomics approach constitute a promising candidate to synthesize behavioral anomalies related to environmental issues within a consistent framework. This approach offers a refreshing way to analyze
individuals' decisions regarding environmental issues without negating
the self-interest component but by complementing it. Despite recent
contributions in this domain, a lot of work remains to be done.
Behavioral findings can understand how individuals make their
decisions and inform public policy in the environmental domain in
several ways. In a recent project, the OECD environment directorate
states that behavioral economics can improve benefit–cost analysis
through methodological adjustments to nonmarket valuation techniques and the development of policy mechanisms to influence environmental behavior. For instance, recent advances in risk perceptions,
loss aversion and hyperbolic discounting have important implications
for benefit–cost analysis and can strongly influence related decisions
(Brown and Hagen, 2010; Shogren and Taylor, 2008).
Of course, a prerequisite for invoking the behavioral change solution is
either negligible transaction costs or exorbitant transaction costs of an alternative outcome (Coase, 1960). Coase (1960) stated that “all solutions
have costs and there is no reason to suppose that government regulation
is called for simply because the problem is not well handled by the market
or the firm.” Behavioral-based interventions are not immune to transaction costs and fulfillment of promises depends on the ability to keep
193
transaction costs sufficiently low in comparison to alternatives. For instance, Shogren (2012) suggests several directions where behavioral insights can help to decrease transaction costs and ease the use of Coasean
bargaining. Moreover, behavioral changes extend ‘the richness of social
alternatives among which we can choose’ (Coase, 1974). A coasean
analysis frequently assumes that agents' preferences are given while
behavioral sciences show that preferences are context-dependent.
Interventions can change preferences and subsequently changing preferences can become in some situations the least-cost way of reducing
harm. For instance, in order to fight global warming, several alternatives
can be considered such as government regulation, Coasean bargaining,
laissez faire but also encouraging people to adopting a vegetarian diet.
There is a clear need to examine the implications of behavioral changes
on the transaction costs in comparison to other kinds of interventions
but addressing adequately this issue is beyond the scope of our
contribution.
In a different vein, neoclassical economics traditionally suggests policy
instruments in order to reduce undesirable behavior such as quotas, taxes
and tradable permits. While neoclassical economics typically assumes
that the selection of the relevant instrument lies in economic and administrative efficiency, psychologists contend that policy measures impact
the way in which people consider the environmental good to be conserved (OECD, 2012). Recent models of moral decision-making suggest
that because of the complex interferences with moral motivations, performances of policy instruments depend on how those instruments are perceived by regulated entities (Frey, 1999; Nyborg, 1999). Rather than
suggesting a substitution between traditional interventions and behavioral ones (Loewenstein and Ubel, 2010), we argue that insights from behavioral economics can help to increase the effectiveness of traditional
interventions (Pollitt and Shaorshadze, forthcoming).
Experimental methods in the laboratory and in the field can contribute to policy design. Indeed, experimental methods constitute a
useful testbed for alternative institutions in order to provide evidence
of success before actual implementation, especially in the context of
developing countries. In a recent paper devoted to the role that the
experimental method and mindset can play in environment and
development economics, Ehmke and Shogren (2008) state that ‘the
relative benefits of using experiments for resource policy in developing
countries may even be greater than in the developed world. Due to the
low capital resource constraints, the rewards may be significant in
using cost-effective experimental research relative to pilot projects’
(see also Shogren, 2012).
Without negating technological innovations, we highly recommend that policy-makers and research funders divert some resources
to design, test, and promote behavioral interventions. As magnified
by some described initiatives, solutions frequently require a mix of
different types of capitals, namely built capital, human capital, social
capital, natural capital and… behavioral capital. We believe that
behavioral innovations can change the traditional frontiers between
‘weak’ and ‘strong’ sustainability. While monetary incentives and
technological improvement are expensive, behavioral changes possess the capacity to deliver good results not only at a lower cost, but
also over a shorter time horizon. Importantly, even researchers suffer
from biases in the way they perceive the environment in the context
of their field of expertise. We suspect the existence of a self-serving
bias, more precisely, a déformation professionelle bias, which causes
traditional scientists to devote too much attention to technological
solutions (as we, admittedly, do with behavioral solutions).
Presumably, behavioral solutions are not immune to deficiencies,
and raise several ethical issues. For instance, little is known on how
behavioral and cognitive biases interact. For instance how loss aversion interacts with hyperbolic discounting? Understanding these
interactions is crucial to design the optimal package of behavior
change‐inducing measures to reach sustainable development goals.
We also need to better understand the relationship between short‐
run and long‐run behavioral changes or conversion (Pollitt and
194
Shaorshadze, forthcoming). Moreover, behavioral policy measures
may discriminate against less-educated populations and could conceivably result in an overly paternalistic society — one in which the
tendency towards behavioral manipulation surpasses the preservation of individual freedoms. Consumer, citizen, or researcher, we are
all subject to attribution biases, inclined to externalize the responsibility of environmental problems by looking to external solutions
(i.e., technological capital) instead of the internal potential that lies
with changing our own behavior — by tapping into our ‘behavioral
capital’. At the other extreme, however, it is evident that purely behavioral solutions to our current environmental predicament would
also fail in important ways. Indeed, behavioral interventions and
other nudges can be used (and are already used) by various agents
who diverge in some respects (e.g., financial resources, leverage)
and ultimately pursue competitive goals. Some environmental problems can be so crucial that behavioral soft interventions or so-called
nudges seem insufficient and call for shoves. For this reason, it is
incumbent upon policy makers to combine behavioral innovations
with more traditional tools. Indeed, change is most likely to be reached
through a range of complementary instruments and approaches. In sum,
there is a need to improve policy design by taking into account recurrent
and predictable cognitive and behavioral biases in order to address
enduring environmental issues, while simultaneously maintaining and
even improving subjective well-being.
Acknowledgments
We thank the editors and anonymous referees of Ecological Economics for their feedback and suggestions, which helped us to improve
significantly the paper. We are also very grateful to Isabelle Feix, David
Masclet, Naoufel Mzoughi, Alban Thomas and Marc Willinger for many
helpful comments. We thank the French research program GESSOL
sponsored by the Ministère de l'Ecologie, du Développement durable,
des Transports et du Logement, the Agency for the Environment and Energy Management (ADEME) for financial support. The usual disclaimer
applies.
References
Akerlof, G., Dickens, W., 1982. The economic consequences of cognitive dissonance.
American Economic Review 72, 307–319.
Allcott, Mullainathan, 2010. Behavior and energy policy. Science 324, 1204–1205.
Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Koszegi, B., Lichtenstein, D., Mazar,
N., Mullainathan, S., Prelec, D., Shafir, E., Silva, J., 2005. Psychology, behavioral
economics, and public policy. Marketing Letters 16 (3), 443–454.
Ariely, D., 2008. Predictably Irrational. HarperCollins, New York.
Ayres, I., Nalebuff, B., 2005. Peer Pressure, Forbes Magazine, April 11th, 2005. http://
www.law.yale.edu/news/2302.htm.
Bamberg, S., 2002. Effects of implementation intentions on the actual performance of
new environmentally friendly behaviours — results of two field experiments. Journal
of Environmental Psychology 22, 399–411.
Beretti, A., Grolleau, G., 2011. Do past states and causes of soil degradation affect stated
willingness-to-pay values? Evidence from a quasi-experimental survey. 2nd
Southern Europe Experimentalists Meeting, February 24–27, Agadir, Morocco.
Beretti, A., Grolleau, G., Mzoughi, N., 2009. How cognitive biases can affect the performance of eco-labeling schemes. Journal of Agricultural & Food Industrial Organization, Berkeley Electronic Press 7 (2), 1–11.
Binswanger, M., 2001. Technological progress and sustainable development: what
about the rebound effect? Ecological Economics 36, 119–132.
Brown, G., Hagen, D.A., 2010. Behavioural economics and the environment. Environment
and Resource Economics 46, 139–146.
Bulte, E., Gerking, S., List, J.A., de Zeeuw, A., 2005. The effect of varying the causes of environmental problems on stated WTP values: evidence from a field study. Journal
of Environmental Economics and Management 49, 330–342.
Burger, J.M., Rodman, J.L., 1983. Attributions of responsibility for group tasks: the egocentric bias and the actor—observer difference. Journal of Personality and Social
Psychology 45, 1232–1242.
Chouinard, H.H., Paterson, T., Wandschneider, P.R., Ohler, A.M., 2008. Will farmers
trade profits for stewardship? Heterogeneous motivations for farm practice selection. Land Economics 84 (1), 66–82.
Cialdini, R.B., 2003. Crafting normative messages to protect the environment. Current
Directions in Psychological Science 12, 105–109.
Cialdini, R.B., 2005. Basic social influence is underestimated. Psychological Inquiry 16,
158–161.
Clement, D., 2002. Interview with Gary Becker. The Region. (June, http://m.minneapolisfed.
org/article.cfm?id=3407).
Coase, R.H., 1960. The problem of social cost. Journal of Law and Economics 3, 1–44.
Coase, R.H., 1974. The lighthouse in economics. Journal of Law and Economics 17 (2),
357–376.
Cornforth, A., 2009. Behaviour change: insights for environmental policy making from
social psychology and behavioural economics. Policy Quarterly 5 (4), 21–28.
Czap, N.V., Czap, H.J., Khatchaturyan, M., Lynne, G.D., Burbach, M.E., 2012. Walking in
the shoes of others: experimental testing of dual-interest and empathy in environmental choice. Journal of Socio-Economics 41 (5), 642–653.
Dejoy, D.M., 1989. The optimism bias and traffic accident risk perception. Accident
Analysis and Prevention 21, 333–340.
Duflo, E., Kremer, M., Robinson, J., 2011. Nudging farmers to use fertilizer: theory and
experimental evidence from Kenya. American Economic Review 101 (6), 2350–2390.
Ehmke, M.D., Shogren, J.F., 2008. Experimental methods for environment and development economics. Environment and Development Economics 14 (4), 419–456.
Ehrhardt-Martinez, Karen, 2009. Pursuing energy-efficient behavior in a regulatory
environment: motivating contexts for policymakers, program administrators, and
program implementers. Paper prepared for the California Public Utilities Commission. (Washington, DC).
Elliot, M.A., Armitage, C.J., 2006. Effects of implementation intentions on the selfreported frequency of drivers' compliance with speed limits. Journal of Experimental Psychology. Applied 12, 108–117.
Ellis, S.E., Hayden, G.M., 2005. Beyond tinkering: economics after behavioral economics
Hofstra Univ. Legal Studies Research Paper No. 05–20 (Available at SSRN: http://
ssrn.com/abstract=824405 or http://dx.doi.org/10.2139/ssrn.824405).
Falk, A., Heckman, J.J., 2009. Lab experiments are a major source of knowledge in the
social sciences. Science 326, 535–538 (23 October).
Farrell, H, Shalizi, C., 2011. Nudge No More, Slate Magazine, November, 12, http://www.
slate.com/articles/health_and_science/new_scientist/2011/11/does_nudge_policy_
work_a_critique_of_sunstein_and_thaler_.html.
Ferguson, A., 2010. Nudge nudge, wink wink. The Weekly Standard 15 (29) (April, 19,
http://www.weeklystandard.com/articles/nudge-nudge-wink-wink).
Forsyth, D.R., Berger, R.E., Mitchell, T., 1981. The effects of self-serving vs. other-serving
claims of responsibility on attraction and attribution in groups. Social Psychology
Quarterly 44, 56–64.
Foster, D., Lawson, S., Blythe, M., Cairns, P., 2010. Wattsup? Motivating reductions in
domestic energy consumption using social networks. NordiCHI '10: Proceedings
of the 6th Nordic Conference on Human–Computer Interaction: Extending Boundaries, October 2010.
Frey, B., 1999. Morality and rationality in environmental policy. Journal of Consumer
Policy 22 (4), 395–417.
Glaeser, E.L., 2006. Paternalism and psychology. The University of Chicago Law Review
73 (1), 133–156.
Goldstein, N.J., Martin, S., Cialdini, R.B., 2007. Yes!: 50 Secrets from the Science of
Persuasion. Profile Books Ltd., London.
Goldstein, N.J., Cialdini, R.B., Griskevicius, V., 2008. A room with a viewpoint: using social
norms to motivate environmental conservation in hotels. Journal of Consumer
Research 35 (3), 472–482.
Gollwitzer, P.M., Sheeran, P., 2006. Implementation intentions and goal achievement: a
meta-analysis of effects and processes. Advances in Experimental Social Psychology
38, 69–119.
Gowdy, J.M., 2008. Behavioral economics and climate change policy. Journal of
Economic Behavior and Organization 68 (3–4), 632–644.
Hart, S., 2005. An interview with Robert Aumann. Macroeconomic Dynamics 9, 683–740.
Herfeld, C., 2012. The potentials and limitations of rational choice theory: an interview
with Gary Becker. Erasmus Journal for Philosophy and Economics 5 (1), 73–86.
Johnson, E.J., Goldstein, D., 2003. Do defaults save lives? Science 302, 1338–1339.
Kahneman, D., 2003. Maps of bounded rationality: psychology for behavioral economics.
The American Economic Review 93 (5), 1449–1475.
Kalinowski, C.M., Lynne, G.D., Johnson, B., 2006. Recycling as a reflection of balanced
self-interest: a test of the metaeconomics approach. Environment and Behavior
38 (3), 333–355.
Lehrer, J., 2011. Is ‘nudging’ really enough? Wall Street Journal (April, 16. Retrieved from:
http://online.wsj.com/article/SB10001424052748703385404576259030599041442.
html).
Levitt, S.D., List, J.A., 2007. What do laboratory experiments measuring social preferences
tell us about the real world. Journal of Economic Perspectives 21 (2), 153–174.
Loewenstein, G., Ubel, P., 2010. Economics behaving badly. The New York Times A31 (July
14. Retrieved from: http://www.nytimes.com/2010/07/15/opinion/15loewenstein.
html?_r=0).
Lomborg, B., 2009. Technology, not talks, will save the planet. Finance & Development
46 (4), 13–14.
Martin, S., 2007. No penguins in my garden. The Guardian. (December, 16, http://www.
guardian.co.uk/commentisfree/2007/dec/16/nopenguinsinmygarden).
Messick, D.M., Tenbrunsel, A.E., 1999. Sanctioning systems, decision frames, and cooperation. Administrative Science Quarterly 44 (4), 684–707.
Nolan, J., Schultz, P.W., Cialdini, R.B., Goldstein, N.J., Griskevicius, V., 2008. Normative social
influence is underdetected. Personality and Social Psychology Bulletin 34, 913–923.
Nyborg, K., 1999. Informational aspect of environment policy deserves more attention:
comment on the paper by Frey. Journal of Consumer Policy 22 (4), 419–427.
OECD, 2012. Behavioural economics and environmental policy design, OECD Environment
Directorate. http://www.oecd.org/env/consumptioninnovationandtheenvironment/
Behavioural%20Economics%20and%20Environmental%20Policy%20Design.pdf.
Pichert, D., Katsikopoulos, K.V., 2008. Green defaults: information presentation and
pro-environmental behavior. Journal of Environmental Psychology 28, 63–73.
Pollitt, M.G., Shaorshadze, I., forthcoming. The role of behavioural economics in energy
and climate policy. In: Fouquet, R. (Edward Elgar) (Ed.), Handbook on Energy and
Climate Change. Chapter #24. Available at http://www.eprg.group.cam.ac.uk/wpcontent/uploads/2011/12/EPRG1130_Main.pdf.
Posner, R.A., 2009. Treating financial consumers as consenting adults. Wall Street Journal
(July 23, A15, http://online.wsj.com/article/SB100014240529702039469045743022
13213148166.html).
Rabin, M., 1994. Cognitive dissonance and social change. Journal of Economic Behavior
and Organization 23 (2), 177–194.
Rifkin, J., 2009. The Empathic Civilization: The Race to Global Consciousness in a World
in Crisis. Jeremy P. Tarcher/Penguin.
Rizzo, M.J., Whitman, D.G., 2009. Little brother is watching you: new paternalism on
the slippery slopes. Arizona Law Review 51 (3), 685–739.
Rogers, T., Bazerman, M.H., 2008. Future lock-in: future implementation increases
selection of “should” choices. Organizational Behavior and Human Decision Processes 106 (1), 1–20.
Rosch, T., 2010. Behavioral economics: observations regarding issues that lie ahead.
Federal Trade Commission, Vienna Competition Conference, June, 9 (http://www.
ftc.gov/speeches/rosch/100609viennaremarks.pdf).
Rosenwald, M.S., 2008. For hybrid drivers, every trip is a race for fuel efficiency. Washington
Post. (May 26th, 2008. http://www.washingtonpost.com/wp-dyn/content/article/2008/
05/25/AR2008052502764.html).
Rubinstein, A., 2006. Comments on behavioral economics. In: Blundell, R., Newey, W.K.,
Persson, T. (Eds.), Advances in Economic Theory (2005 World Congress of the
Econometric Society), vol. II. Cambridge University Press, pp. 246–254.
Ryan, R., Deci, E., 2000. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. The American Psychologist 55, 68–78.
195
Sanfey, et al., 2003. The neural basis of economic decision-making in the ultimatum
game. Science 300, 1755.
Schultz, P.W., Nolan, J., Cialdini, R.B., Goldstein, N.J., Griskevicius, V., 2007. The
constructive, destructive, and reconstructive power of social norms. Psychological
Science 18, 429–434.
Shafir, E., 2008. A behavioural perspective on consumer protection. Competition and
Consumer Law Journal 15, 302–317.
Sheeder, R.J., Lynne, G.D., 2011. Empathy conditioned conservation: ‘walking-in-theshoes-of-others’ as a conservation farmer. Land Economics 87 (3), 433–452.
Shogren, J.F., 2012. Behavioural economics and environmental incentives. OECD Environment Working Papers, #49.OECD Publishing. http://dx.doi.org/10.1787/
5k8zwbhqs1xn-en.
Shogren, J.F., Taylor, L.O., 2008. On behavioral–environmental economics. Review of
Environmental Economics and Policy 2 (1), 26–44.
Thaler, R.H., Sunstein, C.R., 2009. Nudge: Improving Decisions About Health, Wealth
and Happiness. Yale University Press.
University of York, 2009. Student's ‘Green’ Use for Online Social Networking. Posted on
August, 17, 2009 (Retrieved from: http://www.york.ac.uk/news-and-events/news/
2009/green-social-network/).
Vemuri, A.W., Costanza, R., 2006. The role of human, social, built, and natural capital in
explaining life satisfaction at the country level: toward a National Well-Being
Index (NWI). Ecological Economics 58 (1), 119–133.
Venkatachalam, L., 2008. Behavioral economics for environmental policy. Ecological
Economics 67 (4), 640–645.
Ecological Economics 76 (2012) 87–94
Contents lists available at SciVerse ScienceDirect
Ecological Economics
journal homepage: www.elsevier.com/locate/ecolecon
Analysis
Designing watershed programs to pay farmers for water quality services: Case studies
of Munich and New York City☆
Gilles Grolleau a, Laura M.J. McCann b,⁎
a
b
ENSAM–LAMETA Bat. 26, 2, place Pierre Viala, 34060 Montpellier Cedex 1, France
Dept. of Agricultural and Applied Economics, 212 Mumford Hall, University of Missouri, Columbia, MO 65203, USA
a r t i c l e
i n f o
Article history:
Received 18 September 2011
Received in revised form 12 January 2012
Accepted 7 February 2012
Available online 3 March 2012
Keywords:
Environmental transactions
Transaction costs
Water quality
a b s t r a c t
While preserving water quality by contracting with farmers has been examined previously, we analyze these
arrangements from a different perspective. This study uses a transaction cost framework, in conjunction with
detailed case studies of two water quality payment schemes, to examine factors that increase and decrease
transaction costs in order to improve policy choice as well as policy design and implementation. In both
the Munich and New York City cases, agreements with farmers to change land management practices resolved the water quality problems. In Munich, factors including lack of rural/urban antipathy, homogeneous
land use, utilization of well-developed organic standards, and strong demand for organic products decreased
transaction costs. Using existing organic institutions addressed a range of environmental issues simultaneously. Factors that decreased transaction costs in both cases included: highly sensitive land was purchased
outright and the existence of one large “buyer”. Adequate lead time and flexibility of water quality regulations allowed negotiation and development of the watershed programs. Tourism and eco-labels allow
urban residents to become aware of the agricultural production practices that affect their water supply.
We conclude with recommendations based on the experiences of these cities, both of which have been proposed as models for other schemes.
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
Preserving and restoring water quality is a major concern in numerous countries. The success of regulations in reducing pollution from
point sources 1 has led to an increased focus on nonpoint, unregulated,
sources of pollution such as agriculture, which may have lower abatement costs. Point–nonpoint source trading has been enabled by
☆ Acknowledgements: We are particularly grateful to Naoufel Mzoughi, Christophe
Déprés and participants at the workshop ‘A Primer in Economic Analysis of SoilRelated Ecosystem Services (Montpellier, November, 18, 2010) for many helpful comments and suggestions. We also benefited from comments at the World Congress of
Social Economics meetings (Montreal, June 28–July 1, 2010). We are particularly indebted to two anonymous referees and the editor of Ecological Economics for suggestions that greatly improved the paper. We are also grateful to the French research
program GESSOL sponsored by the Ministère de l'Ecologie, du Développement durable,
des Transports et du Logement, the Agency for the Environment and Energy Management (ADEME), the EU Fellows Program at the University of Missouri, and the Missouri
Agricultural Experiment Station for financial support. The usual disclaimer applies.
⁎ Corresponding author. Tel.: + 1 573 882 1304.
E-mail addresses: [email protected] (G. Grolleau), [email protected]
(L.M.J. McCann).
1
According to an OECD report (2003, 43), point-source pollution reduction has been
successful in most OECD countries, given that ‘industrial discharges of heavy metals
and persistent chemicals have been reduced (…) by 70–90% (or more) in most cases’.
Nevertheless, the same report indicates these successes are not enough to meet water
quality standards.
0921-8009/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.ecolecon.2012.02.006
legislation in several places in the U.S. as a way to reduce abatement
costs but there has been less trading than expected due to issues
such as transaction costs (Fang et al., 2005). Contractual arrangements
and payments for water quality services from municipal water organizations to nonpoint sources represent a similar policy instrument. The
current study uses a transaction cost framework in conjunction with
detailed case studies of two water quality payment schemes (i.e.,
Munich and New York City) to examine the factors that increase and
decrease transaction costs in order to improve policy choice as well
as policy design and implementation. Rather than just admitting that
high transaction costs can prevent contractual arrangements as emphasized by Coase (1960), we devote attention to factors explaining
the level of transaction costs and strategies employed to shape them
(Anderson and Libecap, 2005; Déprés et al., 2008; Libecap, 1989).
The water supply of Munich, Germany, with a population of
1.2 million, is mainly from the Mangfall Valley. The water supply for
New York City, with a population of over 9 million people, comes
from the Catskill Mountains and the headwaters of the Delaware
River. In both cities, decreasing water quality in the 1980s meant
that expensive water filtration systems would need to be installed
($6 billion in the case of New York City), or that land management
changes would need to occur in the watersheds. In both cases, agreements with farmers to change land management practices resolved
the water quality problems, however the two cases differed in a number of ways. Using these two detailed cases, we examine the factors
88
G. Grolleau, L.M.J. McCann / Ecological Economics 76 (2012) 87–94
that influenced the level of transaction costs, the nature of the transaction costs that are likely to affect the potential for exchange, and the
opportunities for greater reliance on voluntary contractual solutions. 2
Transaction costs arise throughout the process of an exchange transaction, even if it ultimately does not occur. From an operational viewpoint, these costs notably include the costs of defining, enforcing and
exchanging property rights (Dahlman, 1979; McCann et al., 2005).
Under some circumstances, they may be high enough to prevent voluntary exchanges (Coase, 1960).
The contributions of this paper are at least threefold. First, it adds
empirical content to basic transaction cost concepts by analyzing the
design and implementation of real contractual arrangements for
nonpoint source pollution in a multi-player setting. Second, our comparative analysis of Munich and New York indicates how specific
transaction costs were modified to enable efficiency gains and successful arrangements. Third, it uses the case studies to develop recommendations regarding the design of similar contractual solutions
to water quality issues. While our analysis is focused on water quality,
it has applicability to other complex environmental and natural resource issues where both physical location and the specific institutional environment are important.
The remainder of the paper is organized as follows. Sections 2
and 3 provide historical background and context, analyze the transaction cost issues in each city, and show how various issues were overcome leading to satisfactory arrangements. 3 The final section
provides an overall assessment of arrangements, draws some
generalizable lessons and policy implications, and then concludes.
2. Transaction Cost Issues in Munich
Munich is the third largest city in Germany with about 1.2 million
inhabitants. For more than 125 years, Munich drinking water was
extracted mainly from springs in the foothills of the Bavarian Alps,
namely the Mangfall Valley. This valley, located 40 km from the city
of Munich, supplies around 80% of the tap water consumed in Munich, that is, about 90 million m 3 per year (SVM, 2008). The land in
the catchment areas is mainly used for farming or forestry. Interestingly, since the end of the nineteenth century, the city purchased
lands surrounding the Mangfall springs and devoted them to forestry
activities. Despite this, in the 1980s, the Munich water organization
noticed a slow but significant increase in nitrates (15 mg/l) and pesticides (0.065 μg/l) in groundwater supplying the city. Between
1974 and 1992, the nitrate level increased 250%, from 6 mg/l to
15 mg/l, and there was also deterioration in the taste of the water. Although this increased nitrate level was lower than regulatory requirements for tap water, the city decided in 1991 to behave proactively
and encouraged farmers in the catchment areas to adopt organic
farming (Table 1).
First, the Stadtwerke München (SVM or Munich City Utilities)
delimited the target area (6000 ha) by using hydro-geological data
and defined the desirable changes in farming practices that were
needed to ensure water quality. While half of the area is covered by
forest, around 2250 ha are used for agriculture and 120 farmers
(mainly dairy farmers) were in the target area. Then, SVM organized
a public information campaign for local farmers in order to encourage
2
In our analysis we do not use the externality as the basic unit of analysis but the
transaction. Like other authors (e.g., Cheung, 1970; Coase, 1992; Bougherara et al.,
2009), we suggest shifting the basic unit of analysis from the traditional externality
to the transaction. The transaction framework emphasizes conflicting uses of natural
resources by humans and the potential of various institutional arrangements as ways
to resolve this conflict.
3
Multiple sources of case evidence were gathered and analyzed, e.g., reports by various institutions, academic papers, websites of key organizations, internal documents,
interview transcripts, and the popular and technical press. These documents describe
the history of each case. Of course, any error or misinterpretation is the sole responsibility of the authors.
them to switch to more environmentally friendly practices. Farmers
were initially very reluctant because the suggested changes would
have profoundly modified their production practices. To overcome
the initial reluctance of most farmers, SVM organized meetings to discuss the issue and provided in-depth guidance to individual farmers
by ecological/sustainable farming experts (SVM, 2005). In addition
to persuading farmers at meetings, SVM considered the possibility
of offering financial incentives tailored to each practice change such
as limitations in nitrogen use or manure discharge, or transition to
pasture.
The practice by practice approach was abandoned because to be
effective, it required extensive monitoring and verification, which
would have been very costly and time-consuming. Rather than following its initial plan, SVM switched to a comprehensive approach
by encouraging farmers to switch to organic farming, as defined by
existing regulations (SVM, 2005). Moreover, the SVM selected 3 reputable associations of organic producers already operating in the targeted areas, namely Bioland, Naturland and Demeter. The SVM paid
for the first evaluation by producers' unions of potential candidates
for organic conversion, reducing their fear regarding the extent of
practice changes and their consequences. SVM recognized that the
extensive involvement of the unions helped to overcome farmers'
fears (SVM, 2005).
Contractual arrangements were proposed to farmers who had
land located in the catchment area. In the Munich case, it is clear
that by basing the contractual arrangements on organic farming requirements, SVM economized on several types of transaction costs.
Indeed, referring to organic standards in Germany facilitated several
phases of the transaction. For example, rather than defining and
explaining each practice change, negotiating the value of a specific
mix of changes according to a particular situation (e.g., the portion
of the farm located in the targeted area) or acquiring skills to effectively monitor and enforce the contractual requirements, SVM
benefited from synergies offered by existing organic standards. In
Germany, disputes over appropriate payments were seemingly less
intense compared to the Vittel case 4 thanks to baselines regarding
technical aspects of organic farming and potential losses due to conversion. Rather than starting from scratch, SVM used the experience
accumulated by producers' unions to convince farmers. Relationships
with farmers were also facilitated to some extent by the unions. In
some areas, producers’ unions requested stricter requirements than
official organic requirements such as only 1.5 cows per ha and banning the use of manure from another farm (Naturland). The monitoring is conducted by conventional certification bodies and violations
range from admonition to immediate contract termination. This enforcement strategy is very cost-effective since it uses existing
procedures.
Using available knowledge, the payments offered took into account expected lost income and the investments required to switch
to organic farming. As a financial incentive, the city offered farmers
an annual payment of 280€/ha/year for agricultural land during the
first 6 years, regardless of whether the land was owned or leased.
For the next 12 years, farmers received compensation of 230 €/ha/
year. Interestingly, farmers also benefited from European and national subsidies to switch to (and remain in) organic farming. According
to Simonet (2005), farmers received European subsidies of 250€/ha/
year for a 5-year period, corresponding to the Bavarian agrienvironmental program, the so-called “Kulturlandschaftsprogramm’
(KULAP). At the end of the 5-year period, these contracts can be
renewed. Given the average farm size of 24 ha in the Mangfall watershed, overall compensation per farmer amounted to more than
10,000€ per year (SVM, 2005). In sum, the financial investment
4
Vittel is a French bottler of mineral water which developed contractual arrangements with farmers located in their firm's catchment area to change some agricultural
practices and preserve its water quality. The case is detailed in Déprés et al. (2008).
required from the city to make the switch profitable for farmers was
less than expected because of financial synergies. SVM was also able
to leverage consumer market payments for organic produce. Although organic products command a significant price premium that
can exceed 30% compared to conventional products, organic farming
is less profitable than conventional farming during the initial conversion phase and becomes more profitable than conventional farming
about seven or eight years after the switch. 5
Time is usually required to develop the skills to master organic
farming (learning by doing). Moreover, the relatively high homogeneity of farms (in terms of products and methods of production) in
the targeted area reduced transaction costs to identify and assist
farmers. It also facilitated the process of convincing farmers since results from the early adopters were available. Despite the willingness
to prescribe organic farming in the whole targeted area, the arrangement was somewhat flexible since it allowed farms that could not
comply with all requirements of organic farming to adopt practice
changes favorable to water quality and receive compensation of 200
€/ha/year and have the status of ‘supporting members’ (SVM, 2005).
Given that these supporting members benefit from municipal and
state (CAP) financial incentives they are monitored by independent
examining teams and spot checks are made by the Department of Agriculture (SVM, 2005). In addition to financial incentives and efficient
organizational design, personal characteristics of the farmers involved also played a strong role as admitted by SVM. SVM states
that it “has some very conscientious farmers (…) who are practicing
ecological farming not only for the sake of financial remuneration
but with enormous personal conviction and enthusiasm” (SVM,
2005).
Another noteworthy point is the involvement of Munich and the
producers' unions in the processing and retailing of the organic products. For instance, the city is an important purchaser of products from
the targeted area to supply its schools and municipal restaurants. The
city also funded advertising campaigns to promote the purchase of
organic products from the catchment area by Munich inhabitants.
These ads frequently emphasized the relationship between purchases
of locally produced organic products and water quality. A special promotion involving children's milk indicated “One liter of Bio milk
contributes to the protection of 4000 liters of Munich's drinking
water” (Höllein, 1996). Interestingly, some broader marketing initiatives aim at creating a link between regional origin and Munich inhabitants. For instance, the ‘UnserLand’ (Our Land) umbrella
organization promotes the marketing of regional products (including
organic products) from the counties surrounding Munich (Schafer,
2006). The city involvement was also perceived by farmers as a real
and sustainable commitment reinforcing the strategy's credibility.
It is very likely that the initiative benefited from the institutional
environment (Dietz et al., 2003; Libecap, 1989; North, 1990;
Ostrom, 1990; Williamson, 2000) due to a strong and historically
rooted trend in favor of organic farming. For instance, Germany is
considered as “one of the founders of organic farming” and as the
“largest market for organic food in Europe” and Munich as the “largest market for natural food and organic products in Germany”
(Schafer, 2006, p. 3). The city of Munich itself owns 11 farms covering
2800 ha, not necessarily in the catchment area, and offers tours of
these farms to promote its citizens' interest in organic farming
(Schafer, 2006). Moreover, the city of Munich organizes bicycling
tours from Munich to the Mangfall Valley, at the heart of the
5
Burkhard Schaer, Expert in economic analysis of organic farming in France and
Germany Ecozept, Personal communication, January 2010. Organic farming is frequently perceived as less profitable than organic farming, but several studies (e.g.,
Delate et al., 2003; Badgley et al., 2007) comparing organic and conventional farming
find that organic farms are often more profitable per acre, even if the time by which
organic farming becomes more profitable varies across systems (see Nieberg and
Offermann, 2003).
89
catchment area, and some tours have included picnics and explanations of the city's initiatives (Simonet, 2005).
In sum, the switch to organic farming was self-enforcing thanks to
at least two mechanisms, namely, a mix of financial incentives until
organic farming became more profitable than conventional farming,
and strong involvement of the city in purchasing and promoting organic products from the targeted area as well as acquisition of specific
skills and reputation. If a farmer is largely known as having proenvironmental values by adopting organic farming, this reputational
asset (and desire for consistency) can make a return to conventional
farming less likely. It is expected that, for established organic farms,
the opportunity costs of returning to conventional farming will be
so high that it will not be profitable to do so. Another factor in Munich's favor was that the strong existing demand for organic agricultural products in Germany, as well as German farmers' acceptance of
sustainable and organic production mentioned earlier (Schafer,
2006), meant that existing institutions, such as well-developed organic production standards, extension organizations, and marketing
channels for organic produce could be used, thus decreasing transaction costs. Use of the existing organic institutions also meant that a
range of environmental issues related to conventional farming were
addressed simultaneously. The farming systems near Munich were
also quite suitable to organic production (Ministère de l'Ecologie et
du Développement Durable, 2005).
The results are inspiring. Most of the targeted area (110 farmers
and 80% of agricultural area) is now under contract and is considered
to be the largest contiguous area of organic farming in Germany
(Schafer, 2006). There has been a significant increase in water quality
with levels of nitrates down to 7 mg/l and some pesticides containing
terbuthylazin down to 0.02 μg/l (Simonet, 2005, see also Höllein,
1996). The price increase for water consumers due to the whole
water protection scheme is estimated to be 0.005€/m 3 (SVM, 2005)
whereas the avoided cost of water treatment equipment was estimated at 0.23/m 3 (Simonet, 2005). Good advance planning and sufficient
time were clearly necessary to reach this mutually beneficial arrangement. If the city had not acted early, before they were required to,
time considerations may have prevented the use of contractual approaches and forced the water utility to consider engineering solutions (e.g., building a filtration plant). The Munich case has also
served as a model and inspired similar arrangements in other German
cities (e.g., Stuttgart, Leipzig). Unfortunately, since the New York City
Table 1
Main events regarding the protection of water quality in Munich.
Main dates
Main events
1873–1878 First use of the Mangfall valley (Mangfalltal) to supply drinking
water to Munich
Late 1800s First outright purchase of forest and agricultural fields in the
catchment area
1930
Purchase of 30 properties (500 ha) and plantation with spruces
and larches
1950s
Seeking of new catchments
1950–1970 Plantation of 100 ha with conifers
1980s
Slow but regular increase of NO3 and pesticides in groundwater
1991
SVM's decision to encourage farmers to switch to organic farming
1992
Public information campaign for farmers located in the center of the
targeted area and initial reluctance of farmers
From 1992 Several meetings, individual and collective discussions between the
city, farmers and unions
1993
First 23 farms (800 ha) under contract
1994
50 farms under contract
1998
Updating of contract length after the pilot phase: from 6 years
(conversion period) to 18 years (maintenance period)
1999
92 farms (2200 ha)
2005
110 farms (2500 ha) i.e., over 80% of agricultural areas in the
conversion area were under contract; only about 10 farms refused
to contract
Various sources as indicated in the text.
90
case was unfolding concurrently with the Munich one, they were not
able to benefit from the Munich experience.
Table 2
Main events regarding the protection of water quality in New York City.
Main dates
Main events
3. Transaction Cost Issues in New York City
1905
New York City is the largest city in the United States, with about
9 million inhabitants, and consumes almost 5 million m 3 of water
per day. 6 The New York City watershed consists of 518,000 ha in
the Catskill Mountains and Hudson Valley regions. The portion west
of the Hudson River is located 120–200 km north of New York City
(Appleton, 2002). It is the largest unfiltered water supply in the U.S.
and provides drinking water for the residents of New York City and
some other locations, or almost half of the population of the state of
New York (Catskill Center for Conservation and Development). Land
acquisition and construction of the 18 reservoirs and associated infrastructure began in 1905 and continued through the 1960s and now
supplies 90% of the City's water (Table 2). Land was taken by eminent
domain and entire towns were relocated or disappeared which contributed to resentments that continue to the present (Gold, 1990;
Hoffman, 2008; Porter, 2006). The Croton Watershed, east of the
Hudson River, supplies 10% of the City's water but due to the severity
of the problems, water from this reservoir will be filtered (Appleton,
2002), therefore this watershed is not part of the discussion below.
Relative to New York City, the Catskill Mountains area is economically depressed, relying on tourism, farming (especially dairy), and
construction (Gold, 1990). Three quarters of the watershed is forested, and 85% is in the hands of individuals (Nickens, 1998); half of
that is owned by people from the New York City area who have purchased land for vacation or retirement homes. Changes in these local
industries eventually jeopardized the natural filtering abilities of the
ecosystem, thus compromising drinking water quality in the 1980s
(Appleton, 2002; Porter, 2006). More intensive agricultural practices,
due to economic pressures, led to more pollutants in the runoff from
farms. Forestry practices became less sustainable since forests were
not actively managed. There were also increased discharges from
wastewater treatment plants and faulty septic systems. Increased development contributed to runoff from impervious surfaces and there
was increased building on hills and near attractive streams, which increased erosion (Appleton, 2002).
In the early 1990s, because of the deteriorating water quality,
coupled with changes in federal regulation and potential compliance
costs, New York City reexamined its water supply strategy. Under the
Safe Drinking Water Act Amendments of 1986, the City would normally have been required to filter its surface water supplies unless
it could demonstrate that it had taken other actions to protect its customers from harmful water contamination. For instance, a filtration
avoidance determination or waiver can be granted to water suppliers
if they have a comprehensive watershed protection program ensuring
natural filtration and compliance with water quality standards (Murphy
et al., 1995). The alternative was to spend $4–6 billion to install a filtration plant (plus ¼ billion per year operating costs) (Appleton, 2002).
The filtration plant would have doubled the cost of water to residents
and worsened water taste as well. Presented with a choice between provision of clean water through a major investment in a new treatment facility or managing the watershed, New York City decided that using
natural filtration through watershed management was more cost effective. A state law dating from the beginning of the reservoir projects gives
the City the right to take what measures are necessary to preserve the
quality of its drinking water (Gold, 1990). In 1990, the NYC Department
of Environmental Protection Agency, headed by Commissioner Albert
Appleton, developed a set of regulations, the first since 1954, which restricted development and farming activities in the watershed. This was
1905
Construction on reservoirs and other infrastructure in the Catskill
Mountain and Hudson Valley watersheds.
City obtains legal authorization to take land in the watersheds by
eminent domain and begins to do so.
Regulations to preserve drinking water quality enacted.
U.S. Environmental Protection Agency promulgates new rules to
reduce microbial contamination under the Safe Drinking Water Act.
Water quality deteriorates
Regulations to preserve drinking water quality to comply with the
Safe Drinking Water Act are drafted by the City's Department of
Environmental Protection. Conflict ensues.
The New York State Department of Agriculture works with farmers
and the City to facilitate dialogue and problem-solving.
The City and about 30 watershed communities sign the New York
City Watershed Agreement.
500 farmers were signed up for the Whole Farm Program.
93% of farmers had chosen to participate in the program, well
above the 85% that was stipulated in the Agreement.
1800 acres are in conservation easements and 1375 acres are in
agricultural easements.
Over 200 business listings are in the Pure Catskills Guide to Farm
Fresh Products
6
The New York Metropolitan Statistical Area has almost 19 million inhabitants, including some in New Jersey and Connecticut, but we focus on those on New York City
water.
1954
1986
1980s
1990
Early 1990s
1997
1998
2002
2003
2009
Various sources as indicated in text.
seen as unfair by farmers and other property owners who would be affected; they thought that New York City residents should pay for what
the farmers perceived as changes in property rights (Gold, 1990).
Years of what was described as “open warfare” ensued (Nickens, 1998).
Appleton's office worked to diffuse the confrontation. The City
approached the New York State Department of Agriculture to ask
them to help (Appleton, 2002). High level administrators of the Department made helpful suggestions for creating a dialogue with
farmers such as mutual education on the problems faced by the City
and by farmers in the Catskills. Ultimately, Appleton agreed to let
the farmers themselves design and run a program to reduce pollution
from farms.
In 1997, the City and about 30 watershed communities signed the
New York City Watershed Agreement which paid farmers and others
to implement changes to preserve water quality. The agreement was
estimated to cost $1.4 billion, a significant saving over the new water
treatment facility (Nickens, 1998). Activities included upgrading sewage treatment plants and septic systems, acquiring conservation easements, buying land from willing sellers, and developing innovative
agriculture and forestry programs (Nickens, 1998). The latter included having farmers sign up for a Whole Farm Planning Program to reduce pollution by implementing best management practices and
upgrading manure handling facilities. Technical teams worked to
find combinations that improved water quality and also addressed
the business and labor needs of farmers. The City also paid for technical assistance, research, and monitoring support that the new Watershed Agricultural Council (consisting of both farmers and
government staff) contracted for. Appleton (2002) indicated that it
was an ecosystem approach, not a pollutant by pollutant approach, although less comprehensive than the Munich case. In exchange for a
voluntary program, the farmers committed to obtaining participation
by 85% of the relevant farmers within five years. Farmers were also
involved with recruiting and training farmers in the Munich case,
but in a less formal way. If the program didn't work, the City had
the right to impose restrictive regulations. This represented a credible
threat to underpin the search for mutually beneficial programs and
policies.
By 1998, almost 500 dairy and livestock farms had signed up for
the program, including one who had threatened physical violence
(Nickens, 1998). Within five years after the Whole Farm Program
was created, 93% of farmers had chosen to participate and there are
reports of farmers wondering how they can become part of the watershed (Appleton, 2002). Similar to the Munich case, there is an
eco-label for agricultural products from the region, “Catskill Family
Farms”, now “Catskills Pure” that is managed by the Watershed Agricultural Program (Lydon, 1999). It represents farmers who are participating in the program to improve water quality but the products are
not organic. Increasingly, the program seems to be benefiting from
the “buy local” movement in the U.S.; they have a Farm to Market
Program that helps to connect communities to farmers (Watershed
Agricultural Program, 2010).
Land acquisition, as in the Munich case, was another important
strategy to improve water quality. This corresponds to the wellknown strategy for dealing with pollution, internalizing the externality by the agent being negatively affected purchasing the polluting
firm, i.e. integrating the two activities in a single decision unit
(Castle, 1965). Demsetz (1967) indicates that both contracting and
purchase are options to internalize externalities. However, “…if
there are several externalities, so that several such contracts will
need to be negotiated, or if the contractual agreements should be difficult to police, then outright purchase will be the preferred course of
action” (p. 357–358). This was certainly the case for the Munich and
NYC watersheds. Land acquisition was voluntary (willing seller, full
market value) but the City also needed to prioritize its acquisitions
to efficiently meet water quality objectives (Pires, 2004). The property owned by New York City in 2001 was located closest to the reservoirs. High priority areas for further land acquisition are located in the
lower portions of the watersheds, near the reservoirs (Catskill Center
for Conservation and Development, 2001). However, the Watershed
Agreement stipulated that the City solicit purchase of (not necessarily
actually purchase) 143,745 ha of land in the watershed between 1997
and 2007 and this has gone more slowly than expected (Pires, 2004).
Another strategy was conservation easements. By 2003, the agency
had acquired over 1800 ha in conservation easements (Pires, 2004).
There were also 1375 ha under agricultural easements. The parties
were able to take advantage of the existing Conservation Reserve
Enhancement Program from the United States Department of Agriculture. Munich was also able to benefit from existing programs, as previously mentioned.
The long-standing animosity between people living in the watershed and the governmental agencies from New York City represented
an obstacle to this agreement. Pires (2004) and other authors put
these relationships in a more general category of critical water resource conflict narratives where a powerful place dominates and extracts resources from a subordinate place. In addition, there was a
lack of understanding of the situation of each party. The complexity
of the changes required was also an issue. There were two other
major parties in addition to the farmers, residents of towns and
owners of forested areas. Diversity of the farms as well as the diversity of the stakeholders would tend to increase transaction costs compared to a more homogeneous landscape and population as was the
case in Munich. While various organic standards have existed in the
United States, national legislation was only passed in 1990 and the
regulations were implemented in 2001. Also, until the 2008 farm
bill, there was not specific funding available in the United States to facilitate the switch to organic farming. The institutional environment
in the U.S. thus did not enable the City to use the same institutional
infrastructure that existed in Germany. 7 In addition, organic production alone would not have addressed one of the main water quality issues in the New York case, pathogens from human and animal waste.
7
As pointed by one of the referees, even if the U.S. has had federal standards for organic certification since 2002, there is a history of conflict regarding issues such as
GMO use and sewage sludge. It is not as well-accepted as the organic standard in Germany. There is evidence showing how consumer preferences have shifted in response
to that (e.g., Ward et al., 2004; see also Rosenthal, 2011).
91
A factor that facilitated this agreement was that the Commissioner
of the New York City Department of Environmental Protection, Albert
Appleton, was open to new ways of thinking about the problem and
new solutions (Appleton, 2002). He had a background in management and environmental policy, not engineering, so was more willing
to consider non-engineering solutions to the problem. At the level of
institutional environment, the 1989 USEPA rule that precipitated the
need for filtration to address pathogens allowed the agency to issue a
filtration avoidance determination for facilities that had not previously filtered their water. This regulatory flexibility allowed them to consider non-engineering solutions. Appleton found people in the
Department of Agriculture that were willing to help address the
rural/urban impasse. He thus made use of existing institutions and
the trust that the farmers had in that Department. The fact that the
farmers in the Catskills had a high level of social capital also helped
them to use collective action to design and run the program themselves and recruit the farmers. The City was also willing to purchase
land outright and obtain conservation easements rather than use its
eminent domain rights. This was a major concession (Pires, 2004).
It is also the case that there were large potential gains available
since the cost of the filtration plant was so high. The City could be fairly generous with farmers and still save billions of dollars over time.
The City paid the staff costs of the Whole Farm Program, and also
the capital costs for some pollution abatement investments
(Appleton, 2002). Hoffman (2008; see also Sabatier et al., 2005;
Lubell et al., 2002) indicates that sufficient resources make watershed
collaborations more likely to succeed. 8 As indicated by the high participation rates and desire of those not in the watershed to participate, the program did more than make the farmers whole, it
enabled them to improve their farms and their economic situation.
Hoffman (2008) using federal, state, regional and rural controls
found that the collaboration in the New York City case generally
had positive or neutral effects on population, incomes, unemployment, wages and agricultural employment but negative effects on
the construction industry. From the point of view of the environment,
EPA issued filtration avoidance determinations in 1993, 1997, 2002
and 2007, indicating that water quality was satisfactory. Thus, while
likely having higher transaction costs than the Munich case, the
agreement solved the problem satisfactorily.
4. Some Lessons from the Munich and New York City Strategies to
Cope with Non-point Source Pollution
The previous two sections on the historical background of water
supplies in Munich and New York City show that the two cities
were benefiting for free from natural filtration services for many
years. They eventually noticed a deterioration in their water quality
and made the connection to watershed activities, notably intensive
farming, and were forced to act. Previously, these two groups of
agents, farmers and water suppliers, were using the same environmental assets for different purposes (farming, filtration services) but
in a compatible way. Then, because of increased use of chemical inputs in farming, more intensive livestock production, and stricter regulatory requirements applied to water supplies, the previous balance
was disturbed. These two competing groups were now using the
same scarce resources in an incompatible way (Anderson, 2004). In
the case of NYC, the formal or legal allocation of rights to the provision of water quality (dating from the early 1900s) was now in conflict with the perceived rights of individuals in the watershed to use
the land as they wished for farming or development. The differences
in wealth between the Catskills and the City, and the knowledge
that in the U.S. environmental practices for farmers are usually
8
Using a comprehensive survey of empirical literature devoted to watershed partnership, Leach and Pelkey (2001) proposes a comprehensive list of factors contributing
to their success including several quoted in our study.
92
voluntary, may also have contributed to the strong initial reaction
against regulation of farming practices in the New York case.
This history also showed that several conditions are likely to influence the level of transaction costs, affecting whether a voluntary contractual agreement is feasible and at what cost it can be implemented.
In cases where large gains from trade exist (e.g. NYC case), transaction cost issues may become secondary to win–win opportunities as
far as whether an agreement is reached, but are still important as
far as the overall cost of the program. Without purporting to be exhaustive, we elucidate several factors that played a significant role
in the Munich and New York City arrangements. We then draw
some lessons and recommendations based on the experiences of
these cities, both of which have been proposed as models for other
schemes in their respective countries. Interestingly, the lessons
drawn from our case studies are consistent with factors identified in
other streams of literature about watershed collaborations (Leach
and Pelkey, 2001; Lubell et al., 2002; Sabatier et al., 2005).
regulations) can determine, at least partly, reactions towards a potential win–win contractual arrangement (see also Lubell et al., 2002). A
clear implication is that decision-makers have to consider both short
and longer term time horizons since short term benefits can be more
than offset by additional costs generated later. 9 Once precedents are
set, either formally or informally, transaction costs are incurred to reverse them (Challen, 2000). We do not contend that voluntary approaches are always better but suggest asking under what
circumstances would contracting/negotiating for improved water
quality be better than regulations or a tax? For example, in places
that already have a well-developed organic system, requiring organic
methods could be an efficient way to address multiple environmental
issues simultaneously. However, in a country without such standards
and markets, more conventional input restrictions might be better.
These considerations of time and path dependency are in addition
to the problems of sustainability, intergenerational equity, choice of
discount rates, etc. that are addressed by several authors in Bromley
(1995).
4.1. The Importance of the Institutional Environment at Different Levels
The two cases show that the institutional environment (Dietz et
al., 2003; Libecap, 1989; Lubell et al., 2002; North, 1990; Ostrom,
1990; Williamson, 2000), including norms, legal frameworks and policy, matters. These deeper levels of institutional environment also
shape, to some extent, the level of transaction costs in reaching a
quasi-Coasean solution. For instance, it is clear that contractual arrangements were considered, at least partly, because of looming regulatory pressures regarding water quality and the high cost of
engineering solutions (i.e. filtration plants). Voluntary strategies
may require the existence of conventional regulatory threats with
significant economic consequences to be seriously considered. Several contributions have already stressed the role of regulatory threat
(and associated costs and loss of flexibility) as a significant driver of
voluntary approaches to environmental management (e.g., Alberini
and Segerson, 2002; Khanna and Anton, 2002).
As a second example, economically efficient rearrangements of
property rights can seem unfair, regardless of their formal validity,
if they contradict the perceived existing allocation of rights. Also,
loss aversion can make beneficiaries very reluctant to become payers
- that is, to pay for environmental services which were previously
free. Interestingly, in the New York case, both parties (i.e., farmers
and the municipality) thought they had, to some extent, the property
rights regarding water quality, which created conflict and increased
transaction costs. Property rights affect the direction of compensation. While ‘polluter pays’ is an official OECD policy, in practice, it
has been the case in the U.S. and E.U. that public authorities do subsidize farmers to improve environmental outcomes (Baylis et al., 2008).
Both NYC and Munich water authorities wanted to avoid paying for
high-cost engineering solutions, so a compromise with farmers had
to be reached.
The historical and strong trend towards organic farming in Germany benefited Munich in comparison to New York City. European and
German (formal and informal) incentives in favor of organic farming
allowed the city to take advantage of pre-existing policies, while this
dimension was less important in the New York City case. Moreover,
the historically rooted urban–rural antipathy in New York City obviously increased transaction costs and generated the need for intermediaries (e.g. representatives of the Department of Agriculture) and
costly and time-consuming efforts to develop less contentious relationships while Munich benefited from better pre-existing relationships with the farming community.
It is obvious that history matters and policy instruments rarely
start from scratch. In short, transaction costs of reaching a contractual
agreement are impacted by path-dependence (Arthur, 1989; see
Rammel and van den Bergh, 2002 for an application to environmental
policies). Choices made previously (e.g. imposing statutory
Lesson 1: Stricter regulations about water quality (with flexibility
about how to reach the regulatory targets) may be needed as an impetus to seriously consider voluntary arrangements. Nevertheless,
policies rarely start from scratch and overall transaction costs can
be increased or decreased according to the weight of history and
the institutional environment (Challen, 2000; Libecap, 1989).
4.2. Homogeneity and Number Issues
In water quality issues, there is a fairly strong location specificity 10
(watershed services are not fungible like carbon markets), making a
case-by-case approach necessary and limiting the scale and scope of
the market structure. Watershed services must be exchanged at the
watershed level which makes the local context, physical and institutional, very important. The two case studies show that homogeneity
(heterogeneity) among farmers and other stakeholders can lead to
lower (higher) transaction costs of reaching and enforcing a contractual arrangement (Ostrom, 1990). For instance, the homogeneity of
farms in the Munich catchment area contributed to standardized definitions of requirements and allowed mutual exchange of information
among farmers. The fact that requirements were standardized in accordance with organic farming rules facilitated enforcement by
using pre-existing institutions that are expert in organic farming
monitoring. Adopting a whole farm approach rather than a practiceby-practice approach in the two cases increased some costs (by acquiring rights on actions that are not directly tied to water quality issues), but decreased other costs, notably monitoring costs, and
probably increased efficiency since synergies among practices could
be exploited. A recent USDA report from long-term research programs shows that suites of practices work better than individual practices (USDA, 2010). Contracts may sidestep the measurement
problem by specifying the actions in terms of means to an end rather
than the end itself, sometimes at the price of sacrificing some allocative efficiency (Bougherara et al., 2009). Interestingly, in the two
cases, the arrangement was an interesting mix of one-size-fits-all or
shared requirements, but with some flexibility. For instance, Munich
enabled some farmers who were not able to meet all organic requirements to participate in the program at a less stringent level with reduced financial incentives. In the New York City case, whole farm
plans were developed that incorporated both environmental and
9
These time horizon considerations can be inconsistent with other time horizons
such as political elections which can create additional complications.
10
Cities need selected farmers located on the catchment areas and these farmers can
propose and get significant payments for their ecosystem services mainly/only from
these cities.
business criteria. Since these were developed on an individual basis,
this probably increased transaction costs compared to the Munich
case.
In addition, the ‘monopsony’ market structure with a municipality
as buyer and a limited number of farmers (in comparison with other
situations such as the Mississippi River Basin) has seemingly reduced
some transaction costs. 11 In a recent paper, Kemkes et al. (2009) suggested that creating a ‘monopsony’ can provide an effective way of
delivering ecosystem services, because it is relatively easy to calculate
the willingness of the purchasing organization to pay by measuring
the benefit of the ecosystem service to the organization's objective
function. Indeed, city size and location (developed versus developing
countries) can very roughly determine the money available to address water quality issues. Part of the solution can be found in transforming ‘large number cases into smaller numbers when the large
numbers form associations, clubs or firms, such as river basin associations’ (Yandle, 1998, p.148; see also Libecap, 1989; Mailath and
Postlewaite, 1990; Ostrom, 1990; Paavola and Adger, 2005). Similarly,
the efforts of some organizations (e.g. SVM), as a stand-in for individual citizens to contract with farmers, may sometimes constitute a
transaction cost economizing strategy to artificially create a quasimonopsony. In the two cases, the arrangement was also facilitated
by producers' unions but in different ways. For instance, in Munich
the unions of organic producers were effective at promoting organic
farming in targeted farms, while in New York City, representatives
of farmers were responsible for helping to design the program and
getting a minimum percentage of farmers to participate in the program. 12
Lesson 2: Voluntary contractual arrangements are more likely when
contractors are homogeneous on each side, embedded in a
monopsony-like market structure and not too numerous on the supply side. To some extent, the market structure can be modified in
order to reduce overall transaction costs.
4.3. Time and Good Science
It is obvious from the two cases that adequate lead time and good
science were necessary to reach mutually beneficial agreements.
Reaching agreements between two very different groups (i.e., urban
water users and farmers) required time to create mutual knowledge
and understanding, especially in the New York City case. There are
also lag times with management approaches since there is a stock of
pollutants in the environment, while changing practices affects the
flow. However, timing of regulatory requirements may not allow
time for considering contractual approaches as a possible solution
and may inadvertently favor engineering solutions, which can clean
up drinking water more quickly. Interestingly good science, from
neutral and reputable sources, can increase scientific consensus on
causes and strategies to fix the problems, which lead to lower transaction costs. Better knowledge and understanding of multidimensional interactions between surface activities and watershed
11
At the same time, a monopsony structure can make the transaction very specific
(just one buyer). This effect has been avoided to some extent in the Munich strategy,
where the city reduced the asset specificity of the transaction (by prescribing organic),
making farmers less dependent on the city of Munich (European incentives under the
CAP, market development of organic products in Germany, etc.). Another issue is that
creating a new monopsony buyer would incur transaction costs.
12
In some cases, too high a level of concentration on each side of the market can lead
to bilateral monopoly issues with uncertain outcomes. Indeed, while negotiating with a
unique ‘seller’, e.g., a farmer pool in the Vittel case, was likely to reduce transaction
costs on the one hand, it can also serve to increase the monopoly power problem on
the other (Déprés et al., 2008).
93
services, especially at the local level, is likely to decrease transaction
costs of designing mutually acceptable contractual arrangements.
Conversely, lack of such a shared understanding, based on good science, can make the bargaining very contentious where each party
suspects the other one of exploiting the situation to his advantage
(e.g., Déprés et al., 2008). Scientific findings may not be automatically
accepted by all involved groups (Michaels, 2008).
Lesson 3: Advance planning and good science can decrease transaction costs of reaching mutually beneficial agreements. Without ignoring other parameters, regulators have to take into account how time
flexibility can restrain or enlarge the opportunity set of parties. Given
the public nature of good science, this situation can legitimize investments in acquiring better knowledge of relationships between land
use and watershed services.
4.4. Visibility Issues
Increasing visibility (direct and indirect) of ecosystem services can
increase support of citizens for the proposed arrangement. 13 Indeed,
being close and visible allows the development of a mutual understanding that reduces transaction costs. This effect was stronger in
Munich because of physical proximity and the city's initiatives in
terms of sourcing food products from local areas, preserving water
quality, and organizing bicycling tours in the catchment. New York
City is also much more limited in using natural visibility than is the
case in Quito, Ecuador, where residents can see the snowcapped
peaks that provide their water, which facilitates transforming beneficiaries into supporters and payers (Hahn, 2006.). As described by
Hawn, “walking out of Penn Station onto 34th Street, any New Yorker
who looks up will see concrete and neon, taxi-lines and donut shops.
The forested slopes of the Catskills, some of the city's main watersheds, lie roughly eighty miles to the north; they are very much out
of sight and, usually, out of mind”. Tourism in the Catskills allows
some urban residents to observe the agricultural production practices
that affect their water supply, and natural visibility has been
substituted to some extent with education campaigns. On the marketing dimension, New York City also supported an ecolabel initiative
Pure Catskills which was adapted in 2004 to “emphasize the connection between our region's good food and the famously pure water
that flows to New York City taps” (Watershed Agricultural Council,
2010). So, increasing visibility can also provide suppliers with additional (and not necessarily monetary) benefits (e.g. social image, access to new markets).
Lesson 4: Visibility matters. It can decrease overall transaction costs
by facilitating mutual knowledge and appreciation among potential
transactors or their representatives. Increasing visibility can be a
strategy to decrease the costs of reaching and maintaining an agreement over time.
In sum, the design of watershed protection programs needs to
begin with an understanding of the unique institutional and physical
environment which facilitates some options and constrains others.
For instance, building on existing policies and institutions (as in the
Munich case) can decrease transaction costs. Nevertheless, lessons
can be learned from the successes (and failures) in other situations.
13
This recommendation is consistent with recent research in behavioral economics
(Sunstein and Thaler, 2008). Moreover, Pedersen (2000) provided empirical support
to show that environmental concerns are likely to be accentuated in areas where the
environmental impact is visible and tangible.
94
References
Alberini, A., Segerson, K., 2002. Assessing voluntary programs to improve environmental quality. Environmental and Resource Economics 22, 157–184.
Anderson, T.L., 2004. Donning coase-coloured glasses: a property rights view of natural
resources economics. The Australian Journal of Agricultural and Resource Economics 48 (3), 445–462.
Anderson, T.L., Libecap, G.D., 2005. Forging a new environmental and resource economics paradigm the contractual bases for exchange. Workshop on Environmental
Issues and New Institutional Economics, CESAER, December 8, Dijon (France).
Appleton, Albert A., 2002. How New York City used an ecosystem services strategy carried out through an urban–rural partnership to preserve the pristine quality of its
drinking water and save billions of dollars and what lessons it teaches about using
ecosystem services. Katoomba Conference, Tokyo.
Arthur, W.B., 1989. Competing technologies, increasing returns and lock-in by historical events. The Economic Journal 99, 116–131.
Badgley, C., Moghtadera, J., Quintero, E., Zakem, E., Chappell, M.J., Aviles-Vazquez, K.,
Samulon, A., Perfecto, I., 2007. Organic agriculture and the global food supply. Renewable Agriculture and Food Systems 22, 86–108.
Baylis, K., Peplow, S., Rausser, G., Simon, L., 2008. Agri-environmental policies in the EU
and United States: a comparison. Ecological Economics 65 (4), 753–764.
Bougherara, D., Grolleau, G., Mzoughi, N., 2009. The ‘make or buy’ decision in private environmental transactions. European Journal of Law and Economics 27 (1), 79–99.
Bromley, D.W., 1995. The Handbook of Environmental Economics. Blackwell Publishers, Cambridge, MA.
Castle, E., 1965. The market mechanism, externalities, and land economics. Journal of
Farm Economics 47 (3), 542–556 (Aug., 1965).
Catskill Center for Conservation and Development, 2001. New York City Watershed Protection Priority Areas. www.catskillcenter.org/atlas/nycwatershed/nycw_priorityareas_
1.htm.
Challen, R., 2000. Institutions, Transaction Costs, and Environmental Policy: Institutional Reform for Water Resources. Edward Elgar, Cheltenham, England.
Cheung, S., 1970. The structure of a contract and the theory of a non-exclusive resource. Journal of Law and Economics 13, 49–70.
Coase, R.H., 1960. The problem of social cost. Journal of Law and Economics 3, 1–43.
Coase, R.H., 1992. The institutional structure of production. American Economic Review
82 (4), 713–719.
Dahlman, C.J., 1979. The problem of externality. Journal of Law and Economics 22,
141–162.
Delate, K., Duffy, M., Chase, C., Holste, A., Friedrich, H., Wantate, N., 2003. An economic
comparison of organic and conventional grain crops in a long-term agroecological
research (LTAR) site in Iowa. American Journal of Alternative Agriculture 18,
59–69.
Demsetz, H., 1967. Toward a theory of property rights. American Economic Review 57,
347–359 (May).
Déprés, C., Grolleau, G., Mzoughi, N., 2008. Contracting for environmental property
rights: the case of Vittel. Economica 75 (299), 412–434.
Dietz, T., Ostrom, E., Stern, P., 2003. The struggle to govern the commons. Science 302,
1907–1912.
Fang, F., Easter, K.W., Brezonik, P.L., 2005. Point–nonpoint source water quality trading:
a case study in the Minnesota River Basin. Journal of the American Water Resources Assoc. 645–658 (June).
Gold, Allan R., 1990. New York's Water Rules Worry Catskills. New York Times 17, 1990
(September).
Hahn, 2006. Watershed Services: The New Carbon?, http://www.ecosystemmarketplace.com/
pages/dynamic/article.page.php?page_id=461&section=newsletters&eod=1#close.
Hoffman, J., 2008. Census Peek: Collaboration in the New York City Catskill/Delaware
Watershed: Case Study 1990–2000. Environment, Development and Sustainability
10 (129–156).
Höllein, K., 1996. Clean water through organic farming in Germany. Pesticides News 32
(June), 18.
Kemkes, R.J., Farley, J., Koliba, C.J., 2009. Determining when payments are an effective
policy approach to ecosystem service provision. Ecological Economics.
doi:10.1016/j.ecolecon.2009.11.032.
Khanna, M., Anton, W.R.Q., 2002. Corporate environmental management: regulatory
and market-based incentives. Land Economics 78, 539–558.
Leach, W., Pelkey, N.D., 2001. Making watershed partnerships work: a review of the
empirical literature. Journal of Water Resources Planning and Management 127
(6), 378–385.
Libecap, G.D., 1989. Contracting for Property Rights. Cambridge University Press, New
York.
Lubell, M., Schneider, M., Scholz, J.T., Mete, M., 2002. Watershed partnerships and the
emergence of collective action institutions. American Journal of Political Science
46 (1), 148–163.
Lydon, Betsy, 1999. Eco-labels: what's in a name? Green Guide 62–63, January 1999.
www.thegreenguide.com/doc/62-63/ecolabels.
Mailath, G., Postlewaite, A., 1990. Asymmetric-information bargaining problems with
many agents. Review of Economic Studies 57, 351–368.
McCann, L., Colby, B., Easter, K.W., Kasterine, A., Kuperan, K.V., 2005. Transaction cost
measurement for evaluating environmental policies. Ecological Economics 52 (4),
527–542 (March).
Michaels, D., 2008. Doubt is Their Product: How Industry's Assault on Science
Threatens Your Health. Oxford University Press, New York, New York, USA. 384 pp.
Ministère de l'Ecologie et du Développement Durable, 2005. Utilisation de l'agriculture
biologique pour la qualité de l'eau: Exemple de la ville de Munich. D4E: Expertise
Economique et Evaluation, juillet, pp. 1–4.
Murphy, S., Tone, J.W., Schwartzberg, P., 1995. Land acquisition for water quality protection: New York City and the Catskills watershed system. http://www.ucowr.
siu.edu/updates/pdf/v100-A9.pdf.%20Visited%20on%20June%2011www.ucowr.siu.
edu/updates/pdf/v100-A9.pdf. Visited on June 11, 2010.
Nickens, Eddie, 1998. A watershed paradox — New York City's water quality protection
efforts. American Forests (Winter) 103 (4), 21–24.
Nieberg, H., Offermann, F., 2003. The profitability of organic farming in Europe. OECD,
Organic Agriculture: Sustainability, Markets and Policies. CABI Publishing, Wallingford, U.K., pp. 141–152.
North, D.C., 1990. Institutions, Institutional Change and Economic Performance. Cambridge University Press.
OECD, 2003. Improving Water Management: Recent OECD Experience. OECD
Publishing-IWA Publishing.
Ostrom, E., 1990. Governing the Commons: The Evolution of Institutions for Collective
Action. Cambridge University Press.
Paavola, J., Adger, W.N., 2005. Institutional ecological economics. Ecological Economics
53, 353–368.
Pedersen, L.H., 2000. The dynamics of green consumption: a matter of visibility? Journal of Environmental Policy & Planning 2, 193–210.
Pires, Mark, 2004. Watershed protection for a world city: the case of New York. Land
Use Policy 21 (2004), 161–175.
Porter, K.S., 2006. Fixing our drinking water: from field and forest to faucet. 23 Pace Environmental Law Review, 389. Available at: http://digitalcommons.pace.edu/pelr/
vol23/iss2/4.
Rammel, C., van den Bergh, J., 2002. Evolutionary policies for sustainable development:
adaptive flexibility and risk minimizing. Ecological Economics 47, 121–133.
Rosenthal, E., 2011. Organic Agriculture May Be Outgrowing Its Ideals. The New York
Times 30, 2011 (December).
Sabatier, P.A., Focht, W., Lubell, M., Trachtenberg, Z., Vedlitz, A., Matlock, M. (Eds.),
2005. Swimming Upstream, Collaborative Approaches to Watershed Management.
MIT Press.
Schafer, M., 2006. The sustainable food and agricultural movement in Munich, Germany. Paper presented at the Joint Organic Congress, Odense, Denmark, May 30–31,
2006.
Simonet, N., 2005. L'agriculture biologique au service de l'eau. Report. .
Sunstein, C., Thaler, R., 2008. Nudge: Improving Decisions about Health, Wealth, and
Happiness. Yale University Press.
SVM, 2005. La protection de l'eau potable grâce à l'agriculture écologique: travailler la
terre pour un meilleur environnement. Report. . 9 pp.
SVM (Stadtwerke München), 2008. Stadtwerke München: Committed to the Future. Annual report. http://www.swm.de/dokumente/swm/pdf/dok/annualreport-2008.pdf.
USDA, 2010. Assessment of the effects of conservation practices on cultivated cropland
in the Upper Mississippi River Basin. http://www.nrcs.usda.gov/Technical/nri/
ceap/.
Ward, R., Hunnicutt, L., Keith, J., 2004. If you can't trust the farmer, who can you trust?
The effect of certification types on purchases of organic produce. International
Food and Agribusiness Management Review 7 (1), 60–77.
Watershed Agricultural Council, 2010. Pure Catskills. http://www.nycwatershed.org/
ei_food.html. Visited on June, 11th, 2010.
Watershed Agricultural Program, 2010. 2009 Annual Report and 2010 Workload for
the New York City Catskill/Delaware and Croton Watersheds, February 2010.
Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead.
Journal of Economic Literature 38 (3), 595–613 (September).
Yandle, B., 1998. Coase, Pigou and environmental rights. In: Hill, P., Meiners, R. (Eds.),
Who Owns the Environment? Rowman & Littlefield Publishers, Inc., Lanham,
Maryland, pp. 119–152. chapter 5.
JOURNAL OF ECONOMIC ISSUES
Vol. XLV I No. 3 September 2012
ISSN 0021−3624/2012 $9.50 + 0.00
DOI 10.2753/JEI0021-36244603XX
Notes and Communications
How Can Positional Concerns Prevent the Adoption
of Socially Desirable Innovations?
Abstract: We study the role of positional concerns in explaining the (non-)
diffusion of profitable and environmentally friendly innovations in situations
where there is no conventional prisoner dilemma. Based on anecdotal evidence
from farming activities, we introduce the concept of negative positional goods, that
is, goods that destroy status in a given reference group, and show how interactions
between status concerns and profitability can lead to unexpected outcomes. Several
policy implications are drawn.
Keywords: behavior, innovation, positional concerns, status, sustainable
development
JEL Classification Codes: Q0, Q50
Apart from economic payoffs, social status seems
to be the most important incentive and motivating
force of social behavior.
— John Harsanyi 1976, Nobel laureate
Social influence constitutes a convincing but frequently underestimated candidate to
explain numerous a priori irrational human behaviors (Cialdini 2005). Among ways
through which individuals are influenced by others in society, status and positional
concerns are of a paramount importance. Indeed, in many cases, individuals might
engage in activities or purchase a given good not in response to a basic need, but to
gain social status. The relevance of positional concerns is now well-established. Since
Veblen (1899), an increasing number of scholars have explored the issue of whether
people prefer having more of a good in absolute terms or more than others in society and
Salima Salhi is a Ph.D. student at the Centre de recherche en économie appliquée pour le développement
(CREAD). Gilles Grolleau is a professor at Montpellier Supagro and member of the Laboratoire
Montpellierain d'Economie Théorique et Appliquée (LAMETA). Naoufel Mzoughi is a research fellow at
INRA Ecodéveloppement. And, Angela Sutan is an associate professor at Burgundy School of Business and
member of the Laboratoire d'Expérimentation en Sciences Sociales et Analyse des Comportements
(LESSAC).
799
©2012, Journal of Economic Issues / Association for Evolutionary Economics
800
showed that relative standings matter (e.g., Alpizar, Carlsson and Johansson-Stenman
2005; Bekir, El Harbi and Grolleau 2011; Brekke, Howarth and Nyborg 2003;
Carlsson, Johansson-Stenman and Martinsson 2007; Carlsson and Qin 2010; Chao
and Schor 1998; Duesenberry 1949; Easterlin 1995; Frank 1985a, 1985b; Greene and
Nelson 2007; Grolleau and Saïd 2008; Hirsch 1976; Johannsson-Stenman, Carlsson
and Daruvala 2002; Neumark and Postlewaite 1993; Solnick and Hemenway 1998,
2005, 2009; Solnick, Hong and Hemenway 2007). Without purporting to review
consistently this sizeable and increasing literature, we would like to emphasize four
outstanding points.1 First, relative position matters and matters differently across
domains and reference groups. Status markers are not fixed but evolve over time.
Second, many people seek status as an end in itself and are willing to give up money
just to be better than others. Third, positional issues affect several important aspects
of life such as happiness and longevity. In a similar vein, Brekke, Howarth and
Nyborg (2003) investigate the Hirsch hypothesis (i.e., an increasing portion of
expenditure is allocated to status seeking as average income rises) and show that status
seeking can lead to detrimental effects on environmental quality.2
The purpose of this note is to examine the role of positional concerns as drivers
or barriers to the adoption of socially desirable behaviors in situations where there is
no conventional prisoner dilemma (i.e., where social and private interests converge).
The originality of our manuscript is at least fourfold. First, unlike previous literature
that has mainly investigated positional concerns with regards to consumers’ decision
making, that is, considered the demand side, we focus on the supply side, especially
farmers. We present and discuss some applications related to farming activities in
developed and less developed countries because they are major actors in the
environmental realm. Second, we introduce the concept of negative positional goods,
that is, goods destroying status in a given reference group. In spite of possible benefits,
displaying these goods generates stigma in some reference groups and can be counterproductive in terms of status seeking. Third, we discuss some recent results
concluding to weak positional concerns in poor societies.3 We contend that not only
thresholds matter (reaching a minimum standard before exhibiting positional
preferences) but especially researchers may have missed taking into account the
adequate status dimensions. Fourth, we suggest ways to overcome positional bias in
order to encourage agents to adopt socially desirable behaviors.
The remainder of this note is organized as follows. The second section
characterizes positional concerns, especially in the context of farming activities, and
shows how they can prevent the adoption of socially desirable behaviors such as
protecting the environment. Examples and micro-case studies emphasize how
positional concerns can make people deviate from neoclassical predictions. The third
section is devoted to strategies that can be used to overcome the positional bias. The
fourth section concludes.
801
How Positional Concerns Can Prevent the Adoption
of a priori Win-Win Strategies?
Positional concerns rely on the postulate that individuals have interdependent utility
functions. They frequently compare themselves to others from a reference group on
some dimensions. Comparisons can be upward and downward. In short, people
express higher (lower) happiness levels when they are better (lower) ranked than peers.
According to Solnick and Hemenway (2009), “the literature on positional goods
sometimes suggests that people typically want to be superior to others.” Agents not
only take into account their absolute level on a given dimension (e.g., income, office
size) but also how they perform in comparison to people from the same reference
group (e.g., colleagues, peers). These comparisons can push individuals to adopt
behaviors4 that will improve their relative position, either by using constructive or
destructive strategies (Grolleau, Mzoughi and Sutan 2009).
A crucial point is to identify the dimensions on which the social comparison is
performed. For example, according to societies, farmers can compare on the revenue,
crop yields, farm size, ethnic origin or number of spouses. Some papers empirically
supported that positional concerns are weak in poor societies but they may have
missed the dimension on which these “poor” people are positional (e.g., yields or farm
size rather than income). Using a tailored survey experiment in Northern Ethiopia,
Akay, Martinsson and Medhin (2011) found “positional concerns neither in income
per se nor in income from aid projects among the farmers” and concluded that
“positional concerns are positively correlated with absolute level of income of a
country.” Without questioning their result regarding income, we contend that various
reference groups can use various status markers. Unlike western societies, local
reference groups do not necessarily use income as a major status marker.
Consequently, extrapolating positional concerns by considering only income can
eclipse other status markers that matter much more in a given reference group. For
example, according to Bevan and Pankhurst (2006), in rural Ethiopia, “most power
relations between social groups and categories have historically involved wellestablished hierarchies of status supported by norms and rules favoring superiors
which are often violently enforced. Historic status discriminators include age, gender
and class and various manifestations of ethnicity, clan, religion, occupation and race.”
In the same vein, Carlsson and Qin (2010) indicated that, compared to results in
developed countries, “the concern for relative standing seems to be equally strong
among rural households in China.” In other words, dimensions on which agents from
a given group compare and seek status are culturally determined and evolve over time
(Bekir, El Harbi and Grolleau 2011). Figure 1 presents anecdotal evidence regarding
the wide use and misuse of status symbols in farming activities in various cultural
settings.
Lesson 1: Before aiming at modifying agents’ behavior, it is crucial to identify what dimensions
are used as status markers.
802
Figure 1. Some Status Symbols in Farming Communities
In developed and developing countries, status seeking is a major driver explaining the adoption or nonadoption of some innovations (e.g., Mzoughi 2011; Rogers 2003; Veblen 1899). In some cases, innovations
that can increase income are purchased such as ploughs, but not used for their functional properties but
just displayed at the front of observers (Duran 1967; Lanneau 1967). Even if suggested innovations are
income-enhancing and environmentally-friendly, farmers can be reluctant to adopt them if these
innovations decrease their status on other dimensions in their reference groups (Mzoughi 2011). We
provide below some documented examples.
The Yield Clubs in Farming Communities
In the 1970’s, reaching high yields became a status marker in several farming communities. For instance,
in United Kingdom and France, several clubs (such as the 10-tons wheat club) were created on the
initiative of chemical inputs producers. To become a member of these clubs, farmers have to reach high
yields. Concretely, farmers were encouraged to adopt very intensive farming methods to reach these yields,
resulting in environmental and health problems. These clubs became very prestigious and farmers were
willing to devote considerable resources to participate (Groussard and Colomer 2001). Interestingly, while
these clubs have lost their prestige in developed countries, they won their spurs in developing countries.
For instance, a 5-ton wheat club has been recently created in Algeria with an increasing number of farmers
seeking to become members (AïtSaïd 2011).
Fantasist Statistics and Propaganda in Communist Romania
During the communist era, among countries of COMECOM (Council for Mutual Economic Assistance),
Romania had a bad reputation in reporting agricultural statistics (Labaronne, 1992). Before 1984 (the year
with the highest export value for Romania), the production of cereals at harvesting time was announced
region by region on the public television, before publication in the statistics books, leading to a kind of
race between regional leaders. Winning that race was important at the time, since the regional leader
would win extra salary and consideration. Therefore, evening after evening, announced quantities were
higher and higher, even from unlikely regions. The reporting of quantities was four to ten times larger than
the real production (Le Cacheux 1990). The executive committee of COMECOM concluded that
Romania had an excess supply of cereals and turned to large exports. Noteworthy, even the International
Monetary Fund (IMF) was convinced that Romania had an excess supply. Merely nothing remained for the
national market and the country was confronted to the biggest alimentary crisis in its history.
The Haverstore Silos and Other Status Conveying Goods
In his seminal book, Rogers (2003) emphasized how the status dimension can explain the diffusion of
some innovations. “A spectacular example of the status-providing capacity of certain farm innovations is
provided by the diffusion of ‘Harvestore’ silos in the rural United States. These silos are constructed of
steel and glass, painted navy blue, and prominently display the maker’s name. Their height dominates a
farm’s skyline, so they are easily visible from public roads. Because Harvestores are extremely expensive . . .
most agricultural experts recommend that U.S. farmers buy a much cheaper type of silo for storing their
corn and hay silage. But the status-conferring quality of the Harvestores appeals to many farmers. In fact,
some American farmers own, and prominently display, two Harvestores, perhaps the rural equivalent of
the two-car garage in a suburban home.”1 Constructing these silos was harming both profits and landscape,
but they were strong status symbols in the farming communities.
____________
“It was sort of a good status symbol,” Lucile Cole says of silos. “If you didn’t have a silo, you weren’t quite
as good a farmer” (http://starbeacon.com/currents/x343649743/SILOS/print).
1
803
Of course, various reference groups are not isolated and influence each other in
different ways. For instance, Bruni and Stanca (2006) argue that television
“contributes to shifting up the benchmark for people’s positional concerns: income
and consumption levels are compared not only to those of their actual social reference
group, but also to those of their virtual reference group, defined and constructed by
television programs. As a consequence, television viewing makes people less satisfied
with their income and wealth levels. In this perspective, TV can be seen as a powerful
factor in speeding up the positional treadmill, through comparison with higher
benchmark groups.” Knight, Song and Gunatilaka(2009) found that “most rural
people [in China] confine their reference groups to the village: their orbits of
comparison are narrow.” Using survey data on several European countries, Clark and
Senik (2009) found that colleagues are the most frequently cited reference group.
Moreover, there is evidence that low status people are more likely to engage in
compensatory behaviors and “over-consume” goods that confer status (Caplowitz
1967; Van Kempen 2007; Veblen 1899). Applied to environmental issues, this
overconsumption of status signalling goods can ultimately reinforce the
environmental impacts generated by low-status population through status-related
strategies (Brekke, Howarth and Nyborg 2003).
Let us now consider agents (e.g., farmers) which face the decision to adopt an
environmentally friendly innovation. Rational agents will weigh the benefits against
the costs of adoption and adopt if the expected utility derived from adopting is greater
than the status quo (reservation) utility, that is, if the net expected benefits are
positive. For the sake of exposition, we assume that agents have a utility function
which includes only two components, profit and status. While the former is wellknown in the literature, the latter is less considered. Interestingly, one of the leading
researchers in innovation diffusion, Everett Rogers (2003), devoted several pages of
his seminal book to emphasize that in addition to profitability issues, the relative
advantage of an innovation includes the ability to deliver status. In many
circumstances, status dimensions interact with sustainable development issues, either
in a neutral, positive or negative way. Interestingly, observable features of
environmentally friendly behaviors may constitute (positive) negative positional goods,
goods that convey social information that stigmatizes (magnifies) the adopter and
lowers (increases) his/her social rank. The stigma that arises from negative positional
goods may also create negative social attitudes and discrimination against adopters
(Orbach 2006).
Table 1 describes the expected behavior of agents regarding the introduction of
an innovation. When the innovation impacts status and profits differently, the
aggregate effect is determined by the relative weights of these two components at the
individual level. Indeed, individuals can differ in the weights they give to these components leading to different outcomes. Moreover, loss aversion can reinforce a nochange option in indeterminate cases because a gain will be less valued than a loss of a
similar amount.
804
Table 1. Expected Reactions to the Introduction of an Innovation
Profit enhancing
No profit variation
Profit reducing
Status enhancing
Adoption
Adoption
Indeterminate
No status variation
Indeterminate
Status quo
Non adoption
Status reducing
Indeterminate
Non-adoption
Non adoption
Lesson 2: Rather than just verifying whether the proposed changes are economically profitable
and environmentally friendly, it is crucial to understand how the proposed modifications interact
with relevant status dimensions.
Moreover, an adequate analysis of relationship between profitability and status
can require taking into account different time horizons. Let us consider the example
of an environmentally friendly innovation that delivers immediate but negative status
output and delayed economic benefits. Later, the same innovation can also deliver
status benefits. Innovations addressing climate change are frequently perceived as
immediately and individually costly in economic and status terms while promising
future collective benefits. For instance, it is well-known that in spite of its
environmental benefits organic farming is frequently badly perceived in some farmers
communities while its profitability becomes better than traditional farming after some
years. Because of hyperbolic discounting bias, farmers can express a real interest in
adopting the innovation but perpetually procrastinate (Frederick, Loewenstein and
O’Donoghue 2002), leading to a lower diffusion compared to “rational” expectations.
Lesson 3: Taking into account the timing of costs and gains regarding status and money is
crucial to predict whether agents will adopt a given innovation.
Strategies to Channel Positional Concerns
Given the rooted desire of human beings for status differences, Frey (2008) argues
that blocking a dimension that can confer status (e.g., by using taxation) will simply
push individuals to seek other ways to differentiate themselves from others.
Consequently, rather than preventing people from seeking status, a better solution is
to channel relative issues and status seeking in a socially desirable direction.
McAdams (1992) argues that relative preferences can be channeled to promote private
provision of public goods. For instance, California has passed a bill that allows owners
of hybrid cars (e.g., Prius) with a specific sticker to drive in the high-occupancy vehicle
lanes with just a single occupant, reinforcing both status benefits and reducing
commuting time. This strategy publicizes and increases the visibility of the socially
805
desirable behavior, which results in a higher diffusion of the considered behavior
(Griskevicius, Tybur and Van den Bergh 2010). Even at a country level, positional
preferences over public goods can be initiated by increasing the visibility of the
concerned dimension. Ayres and Nalebuff (2005) report an anecdote regarding how
Norway expressed positional preferences based on the publication of the
Environmental Sustainability Index: “After Norway came in second, the prime
minister asked for a meeting to discuss the results. Rather than crowing about
Norway’s superb showing, all he wanted to talk about was what his country could do
to overtake Finland and become number one.” In short, increasing the visibility
constitutes an excellent candidate to guide people in socially desirable directions.
Lesson 4: Increasing the visibility of socially desirable innovations can increase their status
conveying properties, resulting in higher diffusion, ceteris paribus.
To align status seeking with socially desirable objectives, innovation promoters
must understand what confers status in a given community to tailor the innovations
for this community. A clear strategy can be to eliminate the negative positional
dimensions of suggested innovations which will mechanically increase the likelihood
of adoption, ceteris paribus. Related to this point, innovations are sometimes framed in
a way that emphasizes the loss on a given status dimension (e.g., by adopting organic
farming, the yields are lower), even if there are gains on other dimensions, such as an
improved profitability. If some pre-existing status dimensions are harmed by the
suggested innovation, a strategy frequently used is to redefine the dimensions on
which status can be delivered or the reference group (Clark and Senik 2009;
Wilkinson 2006). Indeed, getting status in other reference groups can counterbalance
the negative status effects in the initial reference group. For instance, several contests
have been created recently to compare entities on socially desirable dimensions, such
as the Newsweek environmental ranking of the biggest companies in developed and
emerging world markets (www.thedailybeast.com/) or at a lower level, the French
national contest of the most bloomed grassland.
Lesson 5: Eliminating or at least attenuating negative positional dimensions of socially desirable
innovations can increase their diffusion, ceteris paribus.
Conclusion
Positional concerns matter a lot and form an important part of the motivation system
in societies, especially in rural communities. Positional issues can explain behaviors
that would otherwise be labeled as “irrational,” either in developed countries or in
developing countries. Consequently, ignoring the interactions between suggested
innovations and status considerations in various reference groups can lead to flawed
recommendations. Even if some more conventional conditions are fulfilled (from
which profitability is the most natural candidate), status effects can prevent the
adoption of innovations that are proven to be economically profitable and
806
environmentally friendly. Indeed, some win-win innovations are perceived as negative
positional goods, which result in lower rates of adoption. Rather than just asserting
this disappointing outcome, we also suggested several strategies by which innovation
promoters can reinforce the diffusion of innovations by explicitly taking into account
relevant status dimensions. Nevertheless, the propositions advanced are only a first
step toward a coherent analysis of how positional issues influence the adoption of
innovations. There is ample room for further study of interactions between statusseeking behavior and pro-environmental changes.
Notes
1.
2.
3.
4.
An exhaustive review of these works is beyond the scope of our contribution. The interested reader
can refer to Grolleau, Mzoughi and Saïd (2012) for a recent review of the literature.
More formally, the authors demonstrate that Hirsch’s hypothesis holds true but under particular
assumptions regarding the structure of individual preferences.
Brekke, Howarth and Nyborg (2003) report anecdotal evidence supporting that “status-seeking
behavior is common in societies with comparatively low consumption levels.”
Individuals can also have psychological defense mechanisms to maintain their subjective well-being
even if they experience a disadvantageous relative position (Frey 2008; Wilkinson 2006). For instance,
they can increase the number of dimensions on which individuals compare or the number of
reference groups for a given dimension.
Salima Salhi
CREAD, Alger
Gilles Grolleau
LAMETA – Montpellier Supagro
Naoufel Mzoughi
INRA – Ecodéveloppement
Angela Sutan
LESSAC – Burgundy School of Business
References
Aït Saïd, Walid. “L'irrigation qui vient à point nommé... ” L’Expression, October17, 2011. Available at
www.lexpressiondz.com/actualite/141188-l-irrigation-qui-vient-a-point-nomme.html. Accessed May
2, 2012.
Akay, Alpaslan, Peter Martinsson and Haileselassie Medhin.“Does Positional Concern Matter in Poor
Societies? Evidence From a Survey Experiment in Rural Ethiopia.” World Development, 2011.
doi:10.1016/j.worlddev.2011.06.004.
Alpizar, Francisco, Fredrik Carlsson and Olof Johansson-Stenman.“How Much Do We Care about
Absolute versus Relative Income and Consumption?” Journal of Economic Behavior and Organization 56
(2005): 405-421.
Ayres, Ian and Barry Nalebuff. “Peer Pressure.” Forbes, April 11, 2005. Available at www.law.yale.edu/
news/2302.htm. Accessed May 2, 2012.
807
Bekir, Insaf, Sana El Harbi and Gilles Grolleau.“(Deceptive) Status Seeking Strategies in France and
Tunisia.” Journal of Economic Issues 45, 3 (2011): 717-732.
Bevan, Philippa and Alula Pankhurst. “Power, Poverty and Wealth in Rural Ethiopia: Lessons from Four
Case Studies.” Fourth International Conference of the Ethiopian Economics Association, Addis
Ababa, June, 2006.
Brekke, Kjell, Richard Howarth and Karine Nyborg. “Status-Seeking and Material Affluence: Evaluating the
Hirsch Hypothesis.” Ecological Economics 45 (2003): 29-39.
Bruni, Luigino and Luca Stanca. “Income Aspirations, Television and Happiness: Evidence from the
World Values Survey.” Kyklos 59 (2006): 209-225.
Caplovitz, David. The Poor Pay More: Consumer Practices of Low-Income Families. London: Macmillan, 1967.
Carlsson, Fredrik, Olof Johansson-Stenman and Peter Martinsson. “Do You Enjoy Having More Than
Others? Survey Evidence of Positional Goods.” Economica 74, 296 (2007): 586-598.
Carlsson Fredrik and Ping Qin. “It is Better to Be the Head of a Chicken than the Tail of a Phoenix: A
Study of Concern for Relative Standing in Rural China.” Journal of Socio-Economics 39 (2010): 180186.
Chao, Angela and Juliet Schor. “Empirical Tests of Status Consumption: Evidence from Women’s
Cosmetics.” Journal of Economic Psychology 19 (1998): 107-131.
Cialdini, Robert. “Basic Social Influence is Underestimated.” Psychological Inquiry 16 (2005): 158-161.
Clark, Andrew and Claudia Senik. “Who Compares to Whom? The Anatomy of Income Comparisons in
Europe.” Economic Journal 120, 544 (2009): 573-594.
Duesenberry, James. Income, Saving, and the Theory of Consumer Behaviour. Cambridge: Cambridge University
Press, 1949.
Duran, Pierre. “La consommation ostentatoire en milieu rural à Madagascar.” L'Homme 7, 2 (1967): 30-47.
Easterlin, Richard. “Will Raising the Incomes of All Increase the Happiness of All?” Journal of Economic
Behavior and Organization 27 (1995): 35-47.
Frank, Robert. “The Demand for Unobservable and other Nonpositional Goods.” American Economic
Review 75, 1 (1985a): 101-116.
———. Choosing the Right Pond. Human Behavior and the Quest for Status. Oxford: Oxford University Press,
1985b.
Frederick, Shane, George Loewenstein and Ted O’Donoghue. “Time Discounting and Time Preference: A
Critical Review.” Journal of Economic Literature 40, 2 (2002): 351-401.
Frey, Bruno. Happiness: a Revolution in Economics. Munich Lectures in Economics. Cambridge: MIT Press,
2008.
Greene, Kenneth and Philip Nelson. “Is Relative Income of Overriding Importance for Individuals?”
International Journal of Social Economics 34, 11 (2007): 883-898.
Griskevicius, Vladas, Joshua Tybur and Bram Van den Bergh. “Going Green to Be Seen: Status,
Reputation and Conspicuous Conservation.” Journal of Personality and Social Psychology 98, 3 (2010):
392-404.
Grolleau, Gilles, Naoufel Mzoughi and Sandra Saïd. “Do You Believe that Others Are More Positional
than You? Results From an Empirical Survey on Positional Concerns in France.” Journal of SocioEconomics 41, 1 (2012): 48-54.
Grolleau, Gilles, Naoufel Mzoughi and Angela Sutan. “The Impact of Envy-Related Behaviours on
Development.” Journal of Economic Issues 43, 3 (2009): 795-803.
Grolleau, Gilles and Sandra Saïd. 2008. “Do You Prefer Having More or More than Others? Survey
Evidence on Positional Concerns in France.” Journal of Economic Issues 42, 4 (2008): 1145-1158.
Groussard René and Jean-François Colomer. Vive l’agriculture. Le politique, les hommes, le vivant. Paris :
France Agricole Editions, 2001.
Harsanyi, John. Essays on Ethics, Social Behavior, and Scientific Explanation. Dordrecht: D. Reidel, 1976.
Hirsch, Fred. Social Limits to Growth. Cambridge: Harvard University Press, 1976.
Johansson-Stenman, Olof, Fredrik Carlsson and Dinky Daruvala. 2002. “Measuring Future Grandparents’
Preferences for Equality and Relative Standing.” Economic Journal 112 (2002): 362-383.
Knight, John, Lina Song and Ramani Gunatilaka. “Subjective Well-being and its Determinants in Rural
China.” China Economic Review 20, 4 (2009): 635-649.
808
Lanneau, Gaston. “L’adoption du tracteur dans une zone de polyculture.” Revue Française de Sociologie 8, 3
(1967): 325-347.
Labaronne, Daniel. “La transition de l'économie roumaine vers l'économie de marché : réformes
économiques et institutionnelles.” Revue d’études comparatives Est-Ouest 23 (1992): 105-136.
Le Cacheux, Jacques. “Roumanie.” Revue de l’OFCE 34 (1990): 117-128.
McAdams, Richard. “Relative Preferences.” The Yale Law Journal Company 102, 1 (1992): 1-104.
Mzoughi, Naoufel. “Farmers Adoption of Integrated Crop Protection and Organic Farming: Do Moral and
Social Concerns Matter?” Ecological Economics 70 (2011): 1536-1545.
Neumark, David and Andrew Postlewaite. “Relative Income Concerns and the Rise in Married Women’s
Employment.” Journal of Public Economics 70 (1993): 157-183.
Orbach, Barak. “Unwelcome Benefits: Why Welfare Beneficiaries Reject Government Aid.” Law and
Inequality: A Journal of Theory and Practice 24 (2006): 107-155.
Rogers, Everett. Diffusion of Innovations. New York: Free Press, 2003.
Solnick, Sara and David Hemenway. “Is More Always Better? A Survey about Positional Concerns.” Journal
of Economic Behavior and Organization 37, 3 (1998): 373-383.
———. “Are Positional Concerns Stronger in Some Domains Than in Others?” American Economic Review 95,
2 (2005): 147-151.
———. “Do Spending Comparisons Affect Spending and Satisfaction?” Journal of Socio-Economics 38 (2009):
568-573.
Solnick, Sara, Li Hong and David Hemenway. “Positional Goods in the United States and China.” Journal
of Socio-Economics 36 (2007): 537-545.
Van Kempen, Luuk. “Status Consumption and Ethnicity in Bolivia: Evidence from Durables Ownership.”
International Journal of Consumer Studies 31, 1 (2007): 76-89.
Veblen, Thorstein. The Theory of the Leisure Class: An Economic Study of Institutions. New York: Macmillan,
1899.
Wilkinson, Will. “Out Of Position: Against the Politics of Relative Standing.” Policy 22, 3 (2006): 3-9.
Do Past States and Causes of Soil Degradation Affect Stated Willingness-to-Pay Values?
Antoine Beretti
2, place Viala, Bâtiment 26
Montpellier SupAgro, UMR1135 LAMETA, 34060 Montpellier, FRANCE
[email protected]
Gilles Grolleau (Corresponding author)
2, place Viala, Bâtiment 26
Montpellier SupAgro, UMR1135 LAMETA, 34060 Montpellier, FRANCE
[email protected]
Abstract : Departing from standard applications of utility theory where only outcomes matter,
we investigate whether past states and alternative causes of an environmental problem affect
respondents’ interest and willingness-to-pay to improve it. We use a quasi-experimental
survey over a random sample of citizens in a small French region. Respondents stated their
perceived value of solving an environmental problem related to soil degradation. We find that
respondents’ interest (measured by the response rate) is strongly affected by the combination
of past states and alternative causes whereas results on stated willingness-to-pay are
inconclusive. Our results suggest that loss aversion can be cancelled depending on the
presence of a cause and on its nature. Several policy implications are drawn.
Keywords: Behavioral economics; Field experiment; Loss aversion; Reference-dependence;
Soil quality.
JEL codes: C93, Q28.
Acknowledgements: We are particularly grateful to Isabelle Feix, Gary Libecap, David
Masclet, Naoufel Mzoughi, Alban Thomas, and participants at the workshop ‘A Primer in
Economic Analysis of Soil-Related Ecosystem Services (Montpellier, November, 18, 2010)
for many helpful comments and suggestions. We are also grateful to the French research
program GESSOL sponsored by the Ministère de l'Ecologie, du Développement durable, des
Transports et du Logement, the Agency for the Environment and Energy
Management(ADEME). The usual disclaimer applies.
1
Do Past States and Causes of Soil Degradation Affect Stated Willingness-to-Pay Values?
Evidence from a Quasi-Experimental Survey
1. Introductive remarks
Numerous economists and psychologists have challenged the conventional perspective where
utility depends exclusively on outcomes and consequences and have suggested that reference
points and alternative causes of a given situation combined with outcomes can shape utility
(Gregory et al., 1993; Brown et al., 2005; Bulte et al., 2005; Masiero and Hensher, 2011;
Hjorth and Fosgerau, 2011). According to Kahneman and Tversky (1979), reference points
will determine whether a proposed change is framed as a loss or a gain, which will in turn
affect how an individual evaluates alternatives. Indeed, the asymmetry of the value function
implies that losses from a reference point have a greater impact upon choices than the
equivalent sized gain from the same reference point. Loss aversion means that people are
more motivated and more willing to pay to avoid losses than to acquire gains of the same
amplitude. For instance, researchers from the University of California “told one group of
homeowners that they could save 50 cents a day by carrying out energy efficiency
improvements in their home. They told a second group of homeowners that they would
continue to lose 50 cents a day if they did nothing. The second group was 300% more likely
to carry out the improvements” (Goldstein et al., 2007). At the same time, there is a small but
growing literature stressing that willingness-to-pay (WTP)1 for addressing an environmental
problem differs if the problem is man-caused or nature-caused (Walker et al., 2009; Bulte at
al., 2005). The roots of this idea lie in the psychological literature related to the attribution
bias and will be discussed later (Brown et al., 2005).
Nevertheless, investigating the combined effects of these two parameters, namely reference
points and alternative causes, on an environmental problem is scarce. The purpose of our
contribution is to fill this gap by using quasi-experimental survey to test whether reference
points and alternative causes affect respondents’ interest2 and WTP values regarding an
environmental problem. The originality of this contribution is at least threefold. First, we
combine reference points and alternative causes in the same quasi-experimental survey on a
1
There is a recent controversy regarding whether the endowment effect is more a by-product of the experimental
procedure than a real phenomenon (see the contributions of Isoni et al., 2011 and Plott and Zeiler, 2011).
2
We define respondents’ interest as the importance they devote to the survey and we measure it by the response
rate.
2
random sample of French inhabitants. Indeed, we contend that people do not take into account
reference points per se in an isolated way, but also consider the context such as alternative
causes. In other terms, even with identical reference points, people are likely to react
differently according to the alternative causes leading to these points (Bulte et al., 2005).
Second, unlike water or air, soil degradation in France and European Union has not received
the same level of attention in policymaking. At the same time, environmental policies are
under consideration, such as well as the European Commission proposal for a Soil Framework
Directive (COM(2006) 232) aiming at ensuring ‘an adequate level of protection for all soils in
Europe’3. Third, investigating this issue can usefully inform policymakers on how to improve
policy framing by considering small, low-cost interventions that could have first-order effects.
The remainder of this exploratory paper is organized as follows. Section 2 overviews the
related literature and formulates testable hypotheses. Section 3 is devoted to the empirical
strategy by describing why and how data have been collected and treated. Section 4 provides
the results and discusses them. Section 5 emphasizes policy implications and concludes.
2. Related literature and hypotheses
Given that the quasi-experimental design combines reference points and alternative causes,
we successively overview the literature related to each of these two dimensions by
emphasizing applications to the environmental realm. As stressed before, reference points and
underpinning alternative causes are likely to interact and for sake of exposition we first
discuss them independently, then in combination, and finally we draw some testable
hypotheses.
21. The loss aversion bias
Casual observations and numerous experiments show that losses from a reference point have a
greater impact upon choices than the equivalent sized gain from the same reference point.
Loss aversion means that people are more motivated to avoid losses than to acquire gains. In
other words, individuals are more reluctant to give up something they already own rather than
acquiring the same thing if they do not own it previously (endowment effect). This bias is
frequently used to explain the observed disparity between WTA and WTP (Tversky and
Kahneman, 1991). The two common explanations of this WTA-WTP disparity are either a
3
Unlike France, soil conservation has been a major concern in United States since several decades.
3
disutility from parting with one’s endowment and/or an extra utility from ownership which is
not anticipated by individuals who are not endowed with the good (Bischoff, 2008). Horowitz
and McConnell (2002) reviewed more than 40 studies devoted to the WTA-WTP gap and
found that the ratio WTA/WTP is ‘highest for non-market goods, next highest for ordinary
private goods, and lowest for experiments involving forms of money’. The fact that the WTAWTP discrepancy tends to be larger when valuing public goods compared to private goods is
surprising. Indeed, the non-excludability property of public goods implies that loss aversion is
likely to be weaker for public goods (Bischoff, 2008; Biel et al., 2011). Biel et al. (2011)
argue that the WTA-WTP discrepancy when valuing public goods is largely a result of
asymmetric emotional experiences and moral perceptions in the two cases.
A major implication of this literature is that all environmental improvements are not identical
in contributors’ mind. If the environmental improvements delivered by a given policy are
perceived by respondents as a way not to lose something they already have, we contend that
they will be more willing to pay. A direct implication is related to policy framing where
messages are likely to be more effective if they are framed in terms of what people stand to
lose (Cornforth, 2009). ‘That loss should be something [they] currently possess and should
also be accompanied with a clear and specific action [they] can personally take to avoid such
a loss.” (Martin, 2007). Applying this suggestion to draft agro-environmental policies for
instance could have first order effects at low cost. Most experiments testing loss aversion have
logically concerned private goods and to a lesser extent public goods (Bischoff, 2008; Biel et
al., 2011). Despite the formal definition of public goods in economics (non-rivalry and nonexclusion), we contend that loss aversion makes sense even if the good considered has public
good properties (Gregory et al., 1993; Horowitz and McConnel, 2002; Beretti et al., 2009;
Bischoff, 2008). Indeed, people can feel a special sense of ownership even on some
environmental goods characterized by public good properties (Biel et al., 2011). Soil quality
is an interesting candidate in this respect as it combines private and public good properties.
Indeed, lands in France are mostly privately owned, and landowners suffer first from soil
degradation. As a production factor either for agriculture or development, non-users can be
excluded from using the land. At the same time, soil has additional functions, which deliver
public goods such as carbon sequestration, biodiversity sink or landscape. We contend that
people will support more efforts to restore a lost state in terms of soil quality than to acquire a
net gain. Our first hypothesis can be formulated as follows:
4
H1: People will express more interest and a higher WTP if environmental initiatives are
framed as restoring a lost state rather than as creating a net gain.
22. The alternative causes bias
People make systematic and often biased inferences about the causes of a given situation and
these attributions strongly influence their behaviors (Heider, 1958; Kelley, 1973; Weiner,
1986; Tetlock, 1995). The psychological literature on attribution is abundant and has mainly
analyzed biases regarding whether people infer that a situation is due to personal factors
(internal attribution) or to situational factors (external attribution). Attributions are frequently
classified according to three causal dimensions: locus of control (internal versus external),
stability (whether the cause changes over time), and controllability (whether oneself can
control the cause). For instance, natural catastrophes can be classified as external, unstable
and uncontrollable while oneself’s efforts can be classified as internal, unstable and
controllable. In judging policymakers, people may give too much credit to those who were
simply lucky and too much blame to those who were unlucky (Mankiw, 2006).
In comparison, economic applications regarding the effect of attributed causes on behaviors
remain scarce (e.g. Babcock et al., 1996; 1997; Bazerman et al., 1997; Ettinger and Jehiel,
2010) and relate mainly to the self-serving bias where people tend to arrive at judgments of
what is fair or right that are biased in the direction of their own self-interests. In the
environmental realm, some contributions (e.g., Walker et al., 1999; Brown et al., 2002; Bulte
et al., 2005) investigated whether people are willing to pay more for man-caused problems
than nature-caused problems. Their results diverge, but the most recent study (Bulte et al.,
2005) found evidence in favor of the hypothesis that people are willing to pay significantly
more to correct man-caused problems compared to nature-caused problems. They found no
support for the hypothesis stating that WTP will be driven by the degree of responsibility
people feel for the damages (Walker et al., 1999). Recently, Beretti et al., (2009) commenting
on the findings of Bulte et al. (2005) called for further research regarding the relevance of the
external versus internal duality usually addressed in the literature. Indeed, if the attribution
can be external from the viewpoint of the individual, it can at the same time remain quite
internal from the point of view of the group or society the individual feels attached to. Even if
the individual does not feel himself responsible for the degradation of the environment, he
could exhibit some sort of solidarity with others and thereby could feel responsible for the
5
society he belongs to. Based on the previous contributions, we formulate two additional
hypotheses:
H2: People will express more interest and a higher WTP when a cause is stipulated compared
to a situation where no cause is indicated.
H3: People will express more interest and a higher WTP when a human cause is stipulated
compared to a situation where a non-human or natural cause is involved.
As far as we know, hypothesis H2 has not been formally investigated in the existing
literature4. Actually we consider that when a cause is explained, the inferences about the
cause of the degradation are easier, and the attribution of the responsibility is suggested. In
opposition, not explaining the cause of the degradation consists in offering the choice to the
individual to infer by himself, or even not infer at all. We thus support the idea that the
presence of a cause relieves the individual from the cost of his inferences. Furthermore it also
reduces the possibility of mixed inferences by focusing him on a particular attribution (human
or natural).
23. Combining the two biases
This exploratory paper has no prior claim regarding the combined effect of these two biases.
Nevertheless our hypotheses clearly imply that all losses of the same extent but due to
different causes are not equal in an individual’s mind. Deriving from the standard graphical
representation of the loss aversion bias, we constructed the figure 1 in order to illustrate the
combination of alternative causes and loss aversion. As presented above, the stated interest
and the stated WTP are hypothetically correlated. In the following, we use the concept of
perceived value to describe any of them. The hypothesis of a correlation between the stated
interest in a environmental problem and the WTP for it, according to the existence or to the
nature of the cause of the related degradation, will be discussed in the following, along with
its implication in terms of environmental valuation methodology.
[ INSERT FIGURE 1 HERE]
4
One could compare studies that estimate WTA or WTP for environmental goods that are either caused by man
(e.g., red tide events due to agricultural runoff) or nature (e.g., beach renourishment after a hurricane).
Nevertheless, such a comparison cannot mimic adequately an experimental design which allows to disentangle
the effects of various factors that can explain the results.
6
The resulting effect of varying the causes of the loss is an expected mitigation of the loss
aversion bias. The gap between the valuation of a gain and the valuation of a particular loss
could be widening or shrinking depending on the presence of a cause, as well as of the nature
of this cause.
Using the following notations, our experimental hypotheses can be formalized:
|VG|: individual valuation of an environmental improvement
|VG| < |VUL|
H1 
|VG| < |VNC|
|VG| < |VMC|
H2 
H3

framed as a net Gain (G);
|VUL|: individual valuation of an environmental improvement
framed as a restoration of an Unexplained Loss (UL);
|VUL| < |VNC|
|VNC|: individual valuation of an environmental improvement
|VUL| < |VMC|
framed as a restoration of a Nature-Caused loss (NC);
|VNC| < |VMC|
|VMC|: individual valuation of an environmental improvement
framed as a restoration of a Man-Caused loss (MC).
In the next section, we design a quasi-experimental survey which allows us to test these
hypotheses.
3. Experimental design
In order to investigate the above mentioned hypotheses, we designed a short questionnaire
regarding a soil quality improvement policy. The interviewees were asked about their interest
in soil quality improvement along with their WTP for a hypothetical (but plausible) soil
quality improvement program. In order to test for the hypotheses, the improvement has been
described either as a net gain, or as the restoration of a previous loss, or as a restoration of a
previous loss which is man-caused, or as a restoration of a previous loss which is naturecaused. Our experimental methodology relies on the two following principles: (1) exactly the
same environmental improvement is described across treatments, only the framing of this
improvement varies and (2) we managed to change as few words as possible between the
different treatments, in order to get the smallest changes. As a consequence, if the smallest
7
differences in wording have an impact in interest or WTP, we can infer that deeper changes in
the framing could have even greater impacts on behavior.
We first performed a pre-test of the questionnaire on a convenience sample (e.g., friends,
colleagues, family) and gathered elements regarding the understanding of soil quality
dynamics, the relevance of proposed scenarios and so on. After, we still improved the
understanding and attractiveness of the questionnaire by administrating the questionnaire to a
random sample of subjects in the streets of Montpellier, allowing us to improve both. In April
2010, the survey was administered to a randomly selected sample of 1760 French citizens
living in Lunel (34400) and Sommières (30250) (Montpellier metropolitan area, South of
France). These two cities were chosen because of their high frequency of flooding events
during the last few years, partly due to a relatively poor soil quality and a high level of
concreting. These events received wide news coverage, especially in local media. Every
participant received an envelope containing the one-page questionnaire, an accompanying
letter describing our study, its purpose and a prepaid return envelope. In an ultimate test
phase, reading the accompanying letter and filling the questionnaire took on average 3
minutes. We wrote the letter of solicitation to make it engaging, easy to read and promoting
the public benefits of results for environmental policymaking in this particular domain. The
letters were identical across treatments. Our experimental design allows us to compare three
seemingly relevant variables: response rates under the different treatments, degree of interest
in program’s implementation and stated WTP5. We contend that the response rate reflects the
interest of interviewees in the tackled issue, which is broader than the interest in the
program’s implementation per se.
We sent 1760 questionnaires which were randomly divided into four groups of 440 contacts,
each group corresponding to one of the four experimental treatments presented hereafter. Here
is a presentation of the main differences (in bold) between treatments corresponding to small
changes in the framing of the soil quality improvement program, the full survey material
being provided in annexes 1a, 1b, 1c, 1d and 1e. In an effort to remain realistic in the
5
We selected the contingent valuation approach. This approach is not immune to criticisms which can lead to
consider alternative approaches such as best-worst, conjoint choice or contingent ranking. Interestingly,
antecedents of contingent valuations have been used to value soil conservation (Carlson and Hanemann, 2005).
Nevertheless, it is important to stress that we are not interested in the elicited value in itself, but the ranking of
the values elicited under different treatments.
8
scenarios, we designed these options jointly with researchers of a unit 6 specialized in soil
related sciences.
Treatment “Gain”:
Situation: Consider that soil quality is measurable on a scale ranging from 1 (bad quality) to
10 (excellent quality). In a given French area, soil quality is estimated at 4/10.
Project: A sustainable investment program is under consideration and its purpose is to reach
a quality of 7/10.
Treatment “Unexplained Loss”:
10 (excellent quality). In a given French area, soil quality is estimated at 4/10.
a quality of 7/10, corresponding to the soil quality 10 years ago.
Treatment “Man Caused Loss”:
10 (excellent quality). In a given French area, soil quality is estimated at 4/10. One of the
major reasons for this mediocre quality is the water proofing of soils due to concreting
which increases the flooding risks.
Treatment “Nature Caused Loss”
10 (excellent quality). In a given French area, soil quality is estimated at 4/10. One of the
major reasons for this mediocre quality is natural erosion which is the reduction of the
superficial layer of soil caused by wind and rain.
6
LISAH – ECO&SOL
9
The only other variation not reported above is the replacement of the words ‘improvement’
and ‘improve’ by the words respectively ‘restoration’ and ‘restore’ (three occurrences) from
the treatment ‘Gain’ to the other experimental treatments.
In line with the strategy suggested by Bulte et al. (2005), each respondent was presented with
a randomly drawn bid from the set (€ 10, 20, 30, 40, where € 1 ≈ US$ 1.3) in order to evaluate
his/her WTP for the considered program. These amounts have been previously discussed with
experts in order to be meaningful for respondents and reflect plausible real world choices 7.
We also followed the experimental procedure in Bulte et al. (2005) related to a hypothetical
scenario presented with a cheap talk (treatment called ‘Hypothetical/Cheap Talk’ in their
paper). We finally collected some usual socio-demographic variables such as gender, age,
number of children, income, education level and ownership of undeveloped land. Key sociodemographic data on respondents are provided in annex 2. No reminder has been sent, to
avoid introducing a bias. We admit that sending reminders is a standard practice. Our unusual
procedure may have lowered the number of respondents compared to other studies, but we
were strongly interested by the response rate drawn from early respondents without any
reminder. Indeed, late respondents can shape results and have characteristics that differ from
early respondents8.
4. Results and discussion
In order to test our hypotheses, we consider three main variables in the data for each
treatment: the response rate, the stated interest in the soil quality improvement program (on a
Likert scale from 1=no interest, to 7=high interest), and the stated WTP for this program. The
former is supposed to reveal the aggregated interest of respondents in the questions asked.
Between April and May 2010, we received 108 responses. Due to a poor update of the file and
a relatively high turnover in the region surveyed, 85 letters were returned by postal services,
stipulating that the person targeted does not live at the stipulated address at the delivery time.
The corrected overall response rate reaches 6.45%. This low response rate can raise concern
regarding the representativeness of the sample. Nevertheless, given the exploratory nature of
7
For instance, we used the costs of CAP per citizen and per year as a reference to select realistic amounts.
8
The interested reader can find an extensive discussion regarding this point in Dalecki et al. (1993) in the
context of contingent valuation applied to environmental resources. Given that our objective is not the
aggregation of WTP for the whole population, we decided not to use reminders.
10
the survey and the current practice in experimental economics and psychology (e.g., nonrandom samples, small samples), we argue that these potential biases do not jeopardize the
validity of the obtained insights. Indeed, the issues addressed do not depend on
representativeness, but on whether small variations about past states and causes affect
respondents’ reactions in comparable samples. Table 1 provides key statistics for the four
Table 1. Key-statistics Regarding each Treatment
Treatment
Number of
responses
Response rate
(%)
Gain (G)
Unexplained Loss (UL)
Nature Caused Loss (NC)
Man Caused Loss (MC)
Total
15
33
32
28
108
3.52
7.99
7.71
6.65
6.45
Interest in
program’s
implementation
5.36
5.61
5.19
6.04
5.56
WTP for the
program
(€)
14
13.44
17.10
15.36
15.09
We first compare the participation rates under different treatments. Table 2 presents the onetailed z-test for the comparison of two proportions from independent samples. For sake of
simplicity, we grouped some experimental treatments under the same label as follows:
-
Loss = Unexplained Loss + Man Caused Loss + Nature Caused Loss
-
Explained Loss = Man-caused Loss + Nature-caused Loss
For sake of exposition, we labeled ‘loss’ what is actually a ‘restoration of a previous loss’. We
invariably presented an environmental improvement of the exact same extent across
Table 2. One-tailed Z-test for the Comparison of Two Proportions from Independent Samples:
response rate
Hypothesis
tested
Group X
Group Y
H1
Loss
Gain
H2
Explained
Loss
Unexplained
Loss
H3
Man-caused
Loss
Value Value Difference z-value Pr(Z>z) Significant
X
Y
X – Y (%)
difference?
(%)
(%)
7.44 3.52
3.92
2.8482 0.0022 YES***
7.18
7.99
-0.81
-0.5151 0.6968
NO
Nature-caused 6.65
Loss
7.71
-1.06
-0.5937 0.7236
NO
*: significant at the 10% level - ** : significant at the 5% level - ***: significant at the 1% level
11
We also performed the classical Wilcoxon-Mann-Whitney nonparametric test for the
comparison of means regarding stated interest for the proposed program. The results are
presented in Table 3.
Table 3. Wilcoxon-Mann-Whitney nonparametric test for the comparison of two means: stated
interest in program’s implementation
Hypothesis
tested
Group X
Group Y
M1: mean
M2: mean
p-value
stated interest stated interest
for X
for Y
5.36
5.59
0.7075
Significant
difference?
H1
Gain
Loss
H2
Unexplained
Loss
Explained Loss
5.61
5.58
0.5235
NO
H3
Man-caused
Loss
Nature-caused
Loss
6.04
5.19
0.0795
YES*
NO
In terms of response rates, there is a significant difference between questionnaires presenting
the improvement as a net gain and questionnaires presenting the improvement as a
compensation of a previous loss. More precisely, people responded significantly more to lossrelated questionnaires than to gain-related questionnaires (H1). In terms of stated interest for
suggested programs, the stated interest for compensating a loss is higher than the stated
interest for acquiring a gain of the same extent, but this difference is not statistically
significant.
Regarding H2, there is no significant difference in terms of response rate and stated interest
when a cause is stipulated compared to situations where no cause is indicated. Moreover,
response rates do not significantly differ among loss-related questionnaires when causes vary.
Interestingly, respondents expressed a higher interest for suggested program when the loss is
man-caused compared to a scenario where the loss is nature-caused (significant at the 10%
level) (H3). Regarding soil protection, our results can indicate that people are more sensitive
to the nature of the cause (man-caused versus nature-caused) than to the presence of a cause
per se (compared to a situation without a stipulated cause). Nevertheless, our findings do not
allow to reach a clear cut conclusion.
We finally performed the Wilcoxon-Mann-Whitney nonparametric test for the comparison of
mean WTP. The results are presented in Table 4.
12
Table 4. Wilcoxon-Mann-Whitney nonparametric test for the comparison of two means: stated
WTP for the program
Hypothesis
tested
Group X
Group Y
M1: mean
M2: mean
stated WTP stated WTP
for X
for Y
14.00
15.27
H1
Gain
Loss
H2
Unexplained Loss
Explained Loss
13.44
H3
Man-caused Loss
Nature-caused
Loss
15.36
p-value
Significant
difference ?
0.6627
NO
16.27
0.2700
NO
17.10
0.5697
NO
The mean stated WTP for an improvement framed as a restoration of a previous loss is higher
than the mean stated WTP for the same improvement framed as a gain. The WTP for the
compensation of a loss is higher when a cause is explicated rather than without any cause.
Lastly, the WTP for compensating a man-caused loss is lower when compared with the WTP
for compensating a nature-caused loss. Nevertheless, these three results are not statistically
significant.
Given the absence of an unified theory and the exploratory nature of our results, the following
interpretations are tentative and must be considered with caution. We contend that the
response to our study can be considered, to some extent, as a contribution to an
environmentally related public good – namely research in social sciences related to soil
protection. Indeed, the public benefits of contributing by filling the questionnaire were
emphasized in the accompanying letter as we stipulated some essential soil-related ecosystem
services. Considering this, our first major result is that individuals seem9 to be more
motivated to respond when the study is concerning the restoration of a previous
environmental loss than an environmental improvement of the same net extent.
This exploratory study of the combining effect of two cognitive biases suggests a second
major and counter-intuitive finding: an individual’s revealed or stated interest in the
program’s implementation is not necessarily a good predictor of its own WTP for it.
Interestingly, respondents expressed a higher interest but a lower WTP to compensate mancaused losses when compared to nature-caused losses. In short, interest could reveal to what
extent respondents support the need of doing something while WTP reveals to what extent
9
Though statistically significant, we consider it as a tentative result because of the relatively small sample.
13
they attribute to themselves responsibility of the situation and are willing to pay to solve it.
This finding might be related to the attribution theory stipulating that people generally
attribute the responsibility of problem or failure to external factors, thus placing the
responsibility of doing something into others’ hands. This could explain both the fact that
their WTP is lower in case of man-caused losses as respondents do not feel themselves
responsible for, whereas interest in implementing the plan is higher due to the precise
identification of other human beings as the external cause. This could also be partly explained
by a concern for equity. In that regard, the cause attribution is very important10.
The result that respondents are less willing to pay for man-caused degradation rather than for
nature-caused degradation is not consistent with Bulte et al. (2005). The higher stated interest
in program’s implementation in the ‘man-caused’ treatment is however consistent with their
findings. A plausible explanation relies on the fact that their study is based on a real public
good, whereas our study is based on a particular good, with several private dimensions, but
encompassing some public good characteristics (ecosystem services provided by soils such as
carbon and biodiversity sink or water filtration). Another difference is a supposed ‘time
effect’. We clearly stipulated in the survey a previous good state of the environment that was
explicitly ten years ago11. In addition, the restoration program was intended to reach
accomplishment by year 2015, almost six years ahead from the survey. Bulte et al. (2005)
stipulated neither a date for the previous state nor a date for the restoration. According to
Ostrom (1990), the ability to observe the results of (your) efforts can be a factor leading to
collective action. This may as well have impacted the credibility of the program and hence
respondents’ statements of their WTP for it along with their stated interest in it. Moreover,
soil degradation frequently occurs through small and cumulative losses over a long horizon
time.
Considering the mitigating effect of varying the causes of the degradation on the loss aversion
bias, this experiment suggests that the absence of a cause can cancel loss aversion, even
reverse it when we look at the WTP. The nature of the cause has a strong effect as well, as we
observe a higher interest in program’s implementation when restoring man-caused losses and
10
An alternative explanation is free riding of respondents by admitting the need of doing something but leaving
the burden on others.
11
Indeed, it seems intuitively convincing that loss aversion is likely to decline over time.
14
a lower interest when the program is aimed at restoring nature-caused losses when both are
compared to the interest in a program framed as providing a gain. This finding reinforces the
recommendation of behavioral economists to pay attention to loss aversion. We suggest
giving almost the same level of importance to the existence and the nature of the cause of the
loss when considering such issues.
Let us finally consider our results from a fundraiser point of view. The strategy is to maximize
the overall WTP for an environmental improvement, which can here be consider as an overall
willingness-to-donate (WTD), while minimizing the cost of fundraising, which can be
considered at first glance as proportional to the number of people interviewed. Below, we
provide a table summarizing these two variables.
Table 5. Fund raised trough each treatment
Experimental
Overall WTD
Number of people
Average donation
Treatment
(euros)
interviewed
(euros/person)
Gain (G)
210
426
0.49
Unexplained Loss (UL)
430
413
1.04
Nature Caused Loss
(NC)
Man Caused Loss (MC)
530
415
1.28
430
421
1.02
Total
1580
1675
0.94
When the environmental improvement is framed as the restoration of a loss, the overall WTD
is doubled when compared to the ‘net gain’ framing. Introducing a human cause for this
previous loss has virtually an effect neither on the overall WTD nor on the efficiency of
fundraising. Additionally, fundraising is almost 30% more efficient when the improvement is
framed as the restoration of nature-caused loss, when compared to an unexplained loss or a
man-caused loss. However we do not have any information on non-respondent and therefore
we cannot claim that their WTD is zero, thus making these average WTD per interviewee
inconclusive. For this reason, any test of significance of differences between these values
would be irrelevant.
15
5. Conclusion
Using a quasi-experimental survey over a random sample of a small French region on efforts
to improve soil quality, we found support for the hypothesis that past states affect
respondents’ interest whereas results of specifying alternative causes on WTP are
inconclusive. More generally, loss aversion is mitigated by the presence and the nature of the
causes of the loss. As mentioned earlier these mixed results can be due to the status of soil
compared to other environmental media such as endangered species or water. They are also
certainly partly due to the small size of our experimental sample. Verifying whether and to
which extent loss aversion effects are limited by public good properties and time constitutes a
challenging question. Interestingly, our results show that framing policies adequately can
significantly increase support among citizens and potential contributors. For instance, policies
aiming at correcting environmental problems framed as losses with well-identified causes are
more likely to be supported, ceteris paribus. These small and low-cost interventions,
especially at stages where just preliminary policy drafts are circulating can help to find the
expected support, or at least decrease reluctance. These improvements can be used in various
situations, ranging from environmental public policy to fundraising by a NGO.
Finally, potential discrepancies between one’s stated interest in an environmental program’s
implementation and one’s WTP for it is of a major interest for environmental valuation
studies. These researches often provide direct information on project feasibility and
sustainability. The implementation of environmental projects depends on the quality of the
WTP data. Consequently, given the consequences, great attention must be devoted to the
gathering of such information. This paper suggests paying attention to the presence and to the
nature of the cause of the environmental degradation in valuation studies.
References
Babcock, L.; Wang, X.; Loewenstein, GF., 1996, Choosing the Wrong Pond: Social
Comparisons in Negotiations that Reflect a Self-Serving Bias, Quarterly Journal of
Economics, 111: 1-19.
Babcock, L.; Loewenstein, G., 1997, Explaining Bargaining Impasse: The Role of SelfServing Biases, Journal of Economic Perspectives, 11: 109 - 126.
Bazerman, MH.; Morgan, KP.; Loewenstein GF., 1997, The Impossibility of Auditor
Independence, MIT Sloan Management Review, 38: 89-94.
16
Beretti, A.; Grolleau, G.; Mzoughi, N., 2009, How Cognitive Biases Can Affect the
Performance of Eco-Labeling Schemes, Journal of Agricultural & Food Industrial
Organization: Vol. 7: Iss. 2, Article 10.
Biel, A.; Johansson-Stenman, O.; Nilsson, A., 2011, The willingness to pay-willingness to
accept gap revisited: the role of emotions and moral satisfaction, Journal of Economic
Psychology, 32(6): 908-917.
Bischoff, I., 2008, Endowment effect theory, prediction bias and publicly provided goods: an
experimental study, Environmental and Resource Economics, 39(3): 283-296.
Brown T.C.; Nannini, D.; Gorter, RB.; Bell, PA.; Peterson, GL., 2002, Judged Seriousness of
Environmental Losses: Reliability and Cause of Loss, Ecological Economics, 42: 479-491.
Brown TC.; Peterson, GL.; Brodersen, RM.; Ford ,V.; Bell, PA., 2005, The judged
seriousness of an environmental loss is a matter of what caused it, Journal of Environmental
Psychology, 25(1): 13–21.
Bulte, E.; Gerking, S.; List, J.A.; de Zeeuw, A., 2005, The Effect of Varying the Causes of
Environmental Problems on Stated WTP Values: Evidence from a Field Study, Journal of
Environmental Economics and Management, 49: 330-342.
Carson, R.T.; Hanemann, WM., 2005, Contingent valuation, in Mäler, KG., Vincent, JR. (eds)
Handbook of environmental economics, Amsterdam, 2: 821-920.
Cornforth, A., 2009, Behaviour Change: Insights for Environmental Policy Making from
Social Psychology and Behavioural Economics, Policy Quarterly, 5(4): 21-28.
Dalecki, MG.; Whitehead, JC.; Blomquist, GC., 1993, Sample nonresponse bias and
aggregate benefits in contingent valuation:
An examination of early, late and non-
respondents. Journal of Environmental Management, 38, 133-143.
Ettinger, D.; Jehiel, P., 2010, A Theory of Deception, American Economic Journal:
Microeconomics, 2(1): 1–20.
Goldstein, NJ.; Martin, S.; Cialdini, RB., 2007, Yes!: 50 Secrets from the Science of
Persuasion, Profile Books Ltd, London.
Gregory, R.; Lichtenstein, S.; MacGregor, D., 1993, The Role of Past states in determining
Reference Points for Policy Decisions, Organizational Behavior and Human Decision
Processes, 55: 155-206.
Heider, F., 1958, The Psychology of Interpersonal Relations. New York: Wiley.
Hjorth, K.; Fosgerau, M., 2011, Loss Aversion and Individual Characteristics, Environmental
and Resource Economics, 49: 573-596.
17
Horowitz, JK.; McConnell, KE., 2002, A Review of WTA/WTP Studies, Journal of
Environmental Economics and Management, 44(3): 426-447.
Isoni, A.; Loomes, G.; Sugden, R., 2011, The Willingness to Pay—Willingness to Accept
Gap, the "Endowment Effect," Subject Misconceptions, and Experimental Procedures for
Eliciting Valuations: Commen, American Economic Review, 101(2): 991–1011.
Kahneman, D.; Tversky, A., 1979, Prospect Theory: An Analysis of Decision under Risk
Econometrica, 47(2), pp. 263-292.
Kelley, H. H., 1973, Attribution in social interaction. In E. E. Jones et al (Eds.), Attribution:
Perceiving the causes of behavior. Morristown, NJ: General Learning Press.
Mankiw,
G.,
2006,
Fundamental
Attribution
Error,
April,
28,
http://gregmankiw.blogspot.fr/2006/04/fundamental-attribution-error.html
Martin, S., 2007, No Penguins in My Garden, The Guardian, December, 16,
http://www.guardian.co.uk/commentisfree/2007/dec/16/nopenguinsinmygarden
Masiero, L.; Hensher, D., 2011, Shift of reference point and implications on behavioral
reaction to gains and losses, Transportation, 38(2): 249-271,
Ostrom, E., 1990, Governing the Commons, Cambridge University Press, Cambridge, UK.
Plott, R.; Zeiler, K., 2011, The Willingness to Pay-Willingness to Accept Gap, the
Endowment Effect, Subject Misconceptions, and Experimental Procedures for Eliciting
Valuations: Reply, American Economic Review, 101(2): 1012-1028.
Tetlock, PE., 1985, Accountability: A social check on the fundamental attribution error,
Social Psychology Quarterly, 48: 227-236.
Tversky, A.; Kahneman, D., 1991, Loss Aversion in Riskless Choice: A Reference-Dependent
Model, The Quarterly Journal of Economics, 106(4): 1039-1061.
Weiner, B., 1986, An attributional theory of motivation and emotion. New York: SpringerVerlag.
Walker, ME.; Morera, O.F.; Vining, J.; Orland, B., 1999, Disparate WTA–WTP Disparities:
the Influence of Human versus Natural Causes, Journal of Behavioral Decision Making, 12:
219–232.
18
Figure 1 : Effect of varying causes on the perceived values of an environmental improvement
VALUE
G
VG
LOSS
GAIN
VUL
VNC
UL
VMC
NC
MC
same improvement
same improvement
G: Improvement framed as a net Gain
UL: Improvement framed as the restoration of a previous Unexplained Loss
NC: Improvement framed as the restoration of a previous Nature-Caused loss
MC: Improvement framed as the restoration of a previous Man-Caused loss
Reference point when the improvement is framed as a net gain
Reference point when the improvement is framed as the restoration of a loss
19
Annex 1a: ‘Gain treatment’ translated questionnaire (G)
QUESTIONNAIRE
This survey analyses the acceptability of environmental initiatives targeted at improving
the state of soils. The following scenario describes a hypothetical realistic situation.
There are no right or wrong answers : we are only interested in your opinion!
SITUATION: Consider that soil quality is measurable on a scale ranging from 1 (bad quality)
to 10 (excellent quality). In a given French area, soil quality is estimated at 4/10.
PROJECT : A sustainable investment program is under consideration and its purpose is to
reach a quality of 7/10.
QUESTION 1 : The overall cost being reasonable, which importance to you attach to the
implementation of this plan? (circle the corresponding number)
no
importance
to the
1
implementation
of the plan
2
3
4
undecided
5
6
7
high
importance
to the
implementation
of the plan
IMPLEMENTATION : A bureau for soil sustainable management wishes to implement this plan in
order to improve soil quality by 2015. However such measures have a cost. In the following question,
we ask whether you accept to pay a certain amount to support this initiative. Usually interviewees
declare more than what they were actually ready to pay. While answering the next question, think
about your actual willingness to pay this amount for soil improvement, knowing that this money won’t
be available for other uses anymore.
QUESTION 2 : Are you willing to pay a unique amount (once for all) of X12 € in order to support soil
improvement? Yes 
No 
Please indicate the following information by filling the blank or ticking the corresponding box:
1. Interviewee’s age : ___ years old
4. Number of people in your household (you included) : ____
2. Interviewee’s sex : Male  or Female 
5. Are there children under 16? Yes  No 
3. Member of an environmental NGO 
6. Study level : before Bac13  or Bac + ___years 
4. Are you a farmer? Yes  No 
8. Do you own undeveloped land? Yes  No 
Please indicate the monthly level of aggregated resources for all the individuals in your household:
1.Less
than 2.Between 1000 and 3.Between 1500 and 4.Between 2000 and 5.More
than
1000€/month  1500 €/month 
2000 €/month 
3000 €/month 
3000€/month 
We would like to remind you that this survey is anonymous and that each question matters.
Thanks for your contribution !
12
X is a randomly drawn bid from the set (10, 20, 30, 40)
13
‘Bac’ is the standard abbreviation for ‘Baccalaureate’, a French nation-wide exam occurring at the end of high
school. A regular student is 18 years old when passing the exam.
20
Annex 1b: ‘Unexplained loss treatment’ translated questionnaire (UL)
QUESTIONNAIRE
SITUATION: Consider that soil quality is measurable on a scale ranging from 1 (bad quality)
to 10 (excellent quality). In a given French area, soil quality is estimated at 4/10.
reach a quality of 7/10, corresponding to the soil quality 10 years ago.
no
importance
to the
1
implementation
of the plan
2
3
4
undecided
5
6
7
high
importance
to the
implementation
of the plan
order to restore soil quality by 2015. However such measures have a cost. In the following question,
about your actual willingness to pay this amount for soil restoration, knowing that this money won’t
QUESTION 2 : Are you willing to pay a unique amount (once for all) of X € in order to support soil
restoration ? Yes 
No 
6. Study level : before Bac  or Bac + ___years 
Please indicate the monthly level of the aggregated resources of all the individuals in your household:
1. Less than 2. Between 1000 and 3. Between 1500 and 4. Between 2000 5. More than
1000€/month  1500 €/month 
2000 €/mois 
and 3000 €/mois  3000€/mois 
21
Annex 1c: ‘Nature-Caused loss treatment’ translated questionnaire (NC)
QUESTIONNAIRE
SITUATION: Consider that soil quality is measurable on a scale ranging from 1 (bad
quality) to 10 (excellent quality). In a given French area, soil quality is estimated at 4/10. One
of the major reasons for this mediocre quality is natural erosion which is the reduction of the
superficial layer of soil caused by wind and rain.
no
importance
to the
1
implementation
of the plan
2
3
4
undecided
5
6
7
high
importance
to the
implementation
of the plan
No 
1.Less
than
1000€/month  1500 €/month 
2000 €/month 
3000 €/month 
3000€/month 
22
Annex 1d: ‘Man-Caused loss treatment’ translated questionnaire (MC)
QUESTIONNAIRE
SITUATION: Consider that soil quality is measurable on a scale ranging from 1 (bad
quality) to 10 (excellent quality). In a given French area, soil quality is estimated at 4/10. One
of the major reasons for this mediocre quality is the water proofing of soils due to concreting
which increases the flooding risks.
no
importance
to the
1
implementation
of the plan
2
3
4
undecided
5
6
7
high
importance
to the
implementation
of the plan
No 
1.Less
than
1000€/month  1500 €/month 
2000 €/month 
3000 €/month 
3000€/month 
23
Annex 1e: accompanying letter (identical across treatments)
Montpellier, March 29th, 2010
Object : Soil quality related survey
Dear Sir or Madam,
In relation to a research project carried by the Laboratoire Montpelliérain d’Economie
Theorique et Appliquée (LAMETA), we submit you a short survey (around 3 minutes)
concerning soil quality. Soil, like air or water, is an essential environmental component. It
plays a major role in maintaining biodiversity and it provides numerous services such as
water filtration and agricultural production medium.
We can assure you of the strict confidentiality of responses. Furthermore, if you wish
we can provide you with the aggregated results of this survey14. In order to respect our
deadline, we would be grateful to you if you return the questionnaire by April 9th, 2010, with
the prepaid and pre-labelled return envelope. Your participation is very important to us and
we thank you in advance for it. For any question or remark, please contact us.
Yours faithfully,
In charge of the study
Gilles Grolleau / Antoine Beretti
LAMETA, UMR 1135, SupAgro
2, Place Pierre Viala, Bât. 26, 34060 Montpellier Cedex 1, France
E-mails : [email protected] / [email protected]
Phone : 04.99.61.28.81 – Fax : 04.67.63.54.09
14
In that regard, please indicate your email adress at the back of the questionnaire if you want to receive the
results of the study.
24
Annex 2. Key socio-demographic data on respondents according to the treatment
Name
Treatment
G
Gain (no cause)
Unexplained
Loss
Man-Caused
loss
Nature-Caused
loss
UL
MC
NC
Women/men Age
Ratio (%) (years)
Education
level
(years)
Child
under
16
Ownership of
Number
Household
Undeveloped
of
income
land
responses
(category)
(%)
(N=108)
38.5
3.85
15
Total
(N=
1675)
23.08
57.36
2.23
0.57
426
40.63
56.72
1.66
0.29
27.3
2.93
33
413
46.43
52.64
1.46
0.29
14.3
3.39
28
421
34.38
54.84
1.35
0.28
46.9
3.59
32
415
25
Immediate versus future implementation of a soil conservation tax: does it matter?
Gilles Grolleau
Supagro, UMR 1135 LAMETA, F-34060 Montpellier
Burgundy School of Business – LESSAC, F-21000 Dijon
[email protected]
Naoufel Mzoughi (Corresponding author)
INRA, UR 767 Ecodéveloppement, F-84914 Avignon
[email protected]
Abstract: A recent psychological literature argues that people are more likely to make “should”
choices rather than “want” choices when the decision’s implementation is scheduled in the future. We
investigate this psychological concern in the case of a soil conservation tax. Surveying both a
convenience and randomly-selected samples of individuals in Montpellier (Southern France), we test
whether the suggestion of a future implementation of a soil conservation tax increases public support,
compared to an “as-soon-as-possible” implementation. Our results show that future implementation
does not matter, since the support of the considered tax is similar in both samples, regardless of the
time perspective of its implementation.
Key-words: future implementation, soil conservation, taxes, want/should conflict.
JEL codes: C90, Q28.
Immediate versus future implementation of a soil conservation tax: does it matter?
1. Introductive remarks
Protecting the environment frequently involves tensions and tradeoffs between immediate (private)
benefits and delayed (collective) ones. Individuals frequently agree they should adopt an
environmentally friendly behavior but do not want to do it immediately because of monetary,
behavioral, or some other kind of (present) costs. For instance, using public transportations rather than
personal cars is frequently perceived as a “should” choice, but it also means giving up convenience,
making many people unwilling to do it immediately. This intra-subjective tension between the
immediate and future selves has been referred to as a “want/should” conflict (Bazerman et al., 1998;
Milkman et al., 2008, 2009). A relatively recent and limited literature (e.g., Khan et al., 2005; Rogers
and Bazerman, 2006 and references therein) provides theoretical arguments and empirical evidence
supporting that individuals tend to think that their future selves will behave more virtuously (by
making should choices) than their present selves (by making want choices). For instance, using four
survey experiments about a variety of domains (e.g., donations, physical exercise, money saving),
Rogers and Bazerman (2006) found empirical evidence to the future lock-in effect, since individuals
across the considered domains were significantly more likely to choose and support a should-choice
when its implementation is scheduled in the distant future, compared to near future.
The literature on intrapersonal conflict in decision making devotes a lot of attention to commitment
devices, frequently illustrated by the metaphor of Odysseus tying himself to the mast of his ships in
order not to succumb to the Sirens (Ashraf et al., 2006). In short, commitment devices constrain
individuals from submitting themselves to temptations of the future want self. Several scholars showed
that situational factors can influence should choices, independently of commitment devices (e.g.,
Ariely and Wertenbroch, 2002; Ashraf et al., 2006; Bazerman and Rogers, 2008). Among these
contextual factors, temporal distance from an event is a convincing candidate. Given that perceptions
of the near and distant future are different, modifying the temporal distance between a decision and the
time its consequences are felt can increase should choices. The Construal Level Theory (Liberman and
Trope, 1998) argues that the same entity can be construed at different levels of abstraction. It also
suggests that varying the temporal distance from an event can activate these different construal levels.
Events in the distant future are more likely to be construed at a high level (abstract, schematic, purpose
focused qualities) while events in the near future are more likely to be construed at a low level
(detailed, concrete and specific qualities) (Trope and Liberman, 2000; Kivetz and Tyler, 2007;
Bazerman and Rogers, 2008).
2
This note extends the above-mentioned literature to the environmental realm. Given that
environmentally friendly decisions frequently reflect should choices, which are rather abstract,
purpose-focused and superordinate, they are more attractive when construed at a higher level. In short,
delayed or future implementation of a “should” decision can nudge people and increase their support
for the considered decision compared to a situation where the implementation is immediate. Hence,
using survey data, we explore whether the support of an environmental policy, namely a soil
conservation tax, is influenced or not by the temporal focus of its implementation. In other words, we
examine whether public support of such a tax increases or not when a future implementation is
suggested, compared to an “as-soon-as-possible” implementation. Moreover, while the few available
empirical studies have been conducted in the United States, we survey individuals in a French
metropolitan area, namely Montpellier (Southern France). Furthermore, we consider two samples of
individuals: a convenience one and a random one. The remainder of this note is organized as follows.
Section 2 presents the empirical strategy and main results. Section 3 provides some policy
implications and concludes.
2. Immediate versus future implementation of a soil conservation tax: a quasi-experimental
survey
2.1. Data and methods
In order to empirically investigate whether time implementation perspectives matter, we designed a
short questionnaire regarding willingness to support a soil conservation tax which would be
implemented to ensure a soil sustainable management program. The survey was conducted in
Montpellier (Southern France) in June and September 2012. We chose this region because of its
commitment to environmental issues. For instance, Montpellier received in 2011 the awards of being
both the French and European capital of biodiversity. Moreover, this region suffered from dryness and
flooding events during the last few years, partly due to relatively poor soil quality and a high level of
concreting. These events received wide news coverage, especially in local media, likely to raise
awareness among inhabitants.1 We surveyed two samples of individuals: the first sample is a
convenience one, composed of 58 students reaching an economics course.2 The second one is
composed of 185 randomly selected people met at a popular square of the city. Some descriptive
statistics about our samples are provided in Table 1. Except for gender distribution, the two samples
1
See, for example, http://www.herault-tribune.com/articles/6647/languedoc-roussillon-la-vitrine-de-vos-talents/
and http://www.lemonde.fr/planete/article/2009/03/28/inventories-les-sols-francais-sont-dans-un-etat-mediocremais-pas-catastrophique_1173728_3244.html, accessed on March 8, 2013.
2
An increasing number of studies in the behavioral economics literature use convenience samples and argue that
such a choice does not skew results. For instance, investigating the importance of relative standings, Grolleau et
al (2012) show that convenience and random samples yield very similar findings.
3
are significantly different. Respondents in the convenience sample are younger and earn less than
randomly-selected individuals. Before conducting the ultimate survey, we performed a pre-test of the
questionnaire among a group of friends, colleagues, and family members. We gathered information
regarding the understanding of soil quality dynamics, relevance of the proposed scenarios and so on.
This step allowed us to improve the understanding and attractiveness of the questionnaire.
Table 1: Some characteristics of our samples
Whole population (N=243)
Random (N=185)
Convenience (N=58)
Wilcoxon test
Mean age
33 years
37 years
22 years
***
Men
45%
44%
48%
ns
Gender
Women
55%
56%
52%
ns
<500€
37%
29%
65%
***
Between 500 and 1000€
28%
28%
28%
ns
Income
(/month)
Between 1000 and 1500€
17%
19%
7%
**
More than 1500€
18%
24%
0%
***
The Wilcoxon test compares the random and convenience samples. (**) and (***) stand for parameter significance at the 5 and 1 percent
level, respectively. (ns) stands for not significant.
We considered two treatments for each sample: one where the tax would be implemented as soon as
possible (denoted Present) and another one where the tax would be implemented after two years
(denoted Future). Our quasi-experimental methodology relies on the two following principles: (1) we
described exactly the same environmental improvement and tax across treatments, and (2) changed as
few words as possible between treatments. As a consequence, if the differences in time
implementation have an impact on support, we can infer that deeper changes in the framing could have
even greater impacts on behavior.
After describing briefly the importance of safeguarding soil quality by emphasizing the importance of
soil services to human beings (a full version of the questionnaire is available in Appendix 1), the
respondents were asked to answer the following question: “In order to promote sustainable
management of soils, suppose that public authorities created a tax of 40€/person/year. This tax would
be voted the next month and go into effect as soon as possible [in 2 years]. In other words, you would
start paying this tax as soon as possible after passing [after two years in the second treatment].
Choose the level which corresponds to your support of this measure, on a scale from 1 (no support at
all) to 9 (strongly support).” The level of the tax policy has been discussed with experts of the
European Common Agricultural Policy to reflect a realistic amount. We also followed the
experimental procedure in Bulte et al. (2005) related to a hypothetical scenario presented with a cheap
talk to avoid an overstatement regarding the support of the tax policy.
4
2.2. Results
The main results are reported in Table 2. We report the mean level of responses to the abovementioned question by type of sample and treatment. We also present results by gender.
Table 2: Mean response by type of sample and treatment
Random sample (N=185)
Convenience Sample (N=58)
Present
Future
Wilcoxon
Present
Future
Wilcoxon
(N=92)
(N=93)
test
(N=28)
(N=30)
test
Mean response
4.63
4.74
ns
4.85
4.73
ns
Mean response (men only)
4.66
4.34
ns
5.23
4.73
ns
Mean response (women only)
4.60
5.12
ns
4.53
4.73
ns
The Wilcoxon test compares responses across the two treatments. (ns) stands for not significant.
Surprisingly, the results suggest that the hypothetical choice by public authorities of an immediate
versus future implementation of a soil conservation tax does not matter. In the two samples, there is no
significant difference between the as-soon-as-possible and future implementation of the tax. Moreover,
no gender effect has been detected between the two treatments.3 These results are only tentative and
should not be over-interpreted. However, these findings raise questions about the applicability of
future lock-in in various contexts. Several factors can explain why our results differ from those
obtained by other scholars (e.g., Rogers and Bazerman, 2008). For instance, the environmental issue
selected (soil versus fossil-fuel consumption, for example) or the time frame can play a role in
explaining individuals’ responses. People may feel more concerned at the individual level by fossil
fuel consumption than by soil issues, especially at a short term horizon. In addition, our sample
includes French inhabitants and may suggest cross-cultural differences regarding the effect of time
framing. Moreover, the environmental policy considered in this paper, taxation, may have limited the
relevance of future implementation, compared to an immediate one. In other words, individuals may
have not reported significantly different answers not only because they are insensitive to time
perspectives but also because they are reluctant to pay additional taxes. The context of our study,
characterized by a huge economic crisis since the year 2008 may have also reinforced such findings.
3. Policy implications and conclusion
Using a quasi-experimental survey among a sample of French individuals in the Southern city
Montpellier, we did not find empirical evidence for an effect of time framing (present versus future)
3
Our data do not allow us to perform a more rigorous econometric analysis, mainly due to the sample size and
potential omitted variables bias. Nevertheless, we tested a logistic model of the effect of the treatment on the
level of tax support, controlling for age, gender, income, and sample type. The results (not reported) show that
the level of responses is not influenced by the suggested treatment, neither control variables. In other words, men
and women surveyed in both samples behave the same with regards to their support of a soil conservation tax,
regardless of their age and monthly income.
5
on the support of the implementation of a soil conservation tax. Our findings are different from the
results of previous studies (e.g., Rogers and Bazerman, 2008). Hence, based on our survey results, one
may conclude that policy-makers do not need to (strategically) use time perspectives in order to
increase individuals’ willingness to support an environmental tax. The unwillingness of individuals to
pay taxes may be related to other factors. It is also possible that the nature of the instrument (i.e., an
additional tax) itself in a context of economic recession has cancelled the effect of time framing. So, a
natural extension has to consider the possible effect of time framing on the support of various
instruments of environmental policy, such as command-and-control instruments. It would be also
interesting to investigate the effect of time framing when the considered measure does not apply to the
surveyed individuals (e.g., voters) but other entities (e.g., companies). A more thorough analysis is
needed by comparing different time frames in various contexts and considering other environmental
policies. The relevance of future implementation remains an under-researched area and requires more
examination to usefully inform decision-makers. In other words, we believe that time framing can be a
crucial dimension to nudge people in socially desirable directions and must not be neglected.
Similarly, considering different environmental issues, other than soil conservation, may reveal
different behavioral patterns. Indeed, some environmental issues are closer to people than others and
this feature can affect their reactions to time framing. In short, rather than providing clear cut
conclusions, our contribution must be considered as an additional stone calling for more academic
attention. Robust findings in this area can lead to practical and low-cost tips that can help
policymakers to gain support for the policy or at least decrease reluctance.
References
Ariely, D., Wertenbroch, K. 2002. Procrastination, deadlines, and Performance: Self-control by
Precommitment. Psychological Science 13 (3): 219-224.
Ashraf, N., Karlan, D., Yin, W. 2006. Tying Odysseus to the mast: Evidence from a commitment
savings product in the Philippines. Quarterly Journal of Economics 121 (2): 635-672.
Bazerman, M.H., Tenbrunsel, A.E., Wade-Benzoni, K. 1998. Negotiating with yourself and losing:
Making decisions with competing internal preferences. Academy of Management Review 23 (2):
225-241.
Bulte, E., Gerking, S., List, J.A., de Zeeuw, A. 2005. The effect of varying the causes of
environmental problems on stated WTP values: Evidence from a field study. Journal of
Environmental Economics and Management 49 (2): 330-342.
Grolleau, G., Mzoughi, N., Saïd, S. 2012. Do you believe that others are more positional than you?
Results from an empirical survey on positional concerns in France. Journal of Socio-Economics 41
(1): 48-54.
6
Khan, U., Dhar, R., Wertenbroch, K. 2005. A behavioral decision theory perspective on hedonic and
utilitarian choice. In Ratneshwar, S., Mick, D.G. (Eds.), Inside consumption: Frontiers of research
on consumer motives, goals, and desires. London: Routledge: 144-165.
Kivetz, Y., Tyler, TR. 2007. Tomorrow I’ll be me: The effect of time perspective on the activation of
idealistic versus pragmatic selves. Organizational Behavior and Human Decision Processes 102
(2): 193-211.
Liberman, N., Trope, Y. 1998. The role of feasibility and desirability considerations in near and
distant future decisions: A test of temporal construal theory. Journal of Personality and Social
Psychology 75 (1): 5-18.
Milkman, K.L., Rogers, T., Bazerman, M.H. 2008. Harnessing our inner angels and demons: What we
have learned about want/should conflicts and how that knowledge can help us reduce short-sighted
decision making. Perspectives on Psychological Science 3 (4): 324-338.
Milkman, K.L., Rogers, T., Bazerman, M.H. 2009. I’ll have the ice cream soon and the vegetables
later: A study of online grocery purchases and order lead time. Marketing Letters 21 (1): 17-35.
Rogers, T., Bazerman, M.H. 2008. Future lock-in: Future implementation increases selection of
‘should’ choices. Organizational Behavior and Human Decision Processes 106 (1): 1-20.
Trope, Y., Liberman, N. 2000. Temporal construal and time-dependent changes in preference. Journal
of Personality and Social Psychology 79 (6): 876-889.
7
Appendix 1: Survey instrument (translated from French)
Changes among treatments (Present versus Future) are mentioned in bold.
Anonymous survey
(Please, answer the following questions as much precisely as possible. There is no right or wrong
answer. Only your opinion matters.)
Soils, just like air and water, are an important environmental component. They play an important role
in maintaining biodiversity and provide numerous useful services to society, such as water refining or
carbon storage. In order to encourage a sustainable management of soils, notably through erosion and
concreting control, and increase soil organic material, public authorities are considering the
implementation of several measures. Nevertheless, such measures are costly.
Generally, surveyed people indicate more than what they are effectively willing to pay. When
answering the question below, try to think if you would effectively pay the suggested amount,
knowing that this money will not be available for other uses.
A. In order to promote sustainable management of soils, suppose that public authorities created a tax
of 40€/person/year. This tax would be voted the next month and go into effect as soon as possible [in 2
years]. In other words, you would start paying this tax as soon as possible after passing [after two
years]. Choose the level which corresponds to your support of this measure, on a scale from 1 (no
support at all) to 9 (strongly support).
1
2
3
4
5
6
7
8
No support at all
9
Strongly support
B. Give the answers corresponding to your situation:
1. Age : _____ years
2. Gender : Man
Woman
3. Monthly net income:
a) Less than 500€
b) Between 500 € and 1000 €
c) Between 1000 € and 1500 €
d) More than 1500 €
C. Comments:
__________________________________________________________________________________
__________________________________________________________________________________
__________________________________________________________________________________
Thanks for your participation
8
An instrument that could turn crowding-out into
crowding-in
Antoine Beretti, Charles Figuièresyand Gilles Grolleauz
February 14, 2013
5
Abstract
Using a simple decision-theoretic approach, we formalize how agents with different
kinds of intrinsic motivations react to the introduction of monetary incentives. We contend
that empirical results supporting the existence of a crowding-out effect in various contexts
hide a more complex reality. We also propose a new policy instrument which taps into
agents' heterogeneity regarding intrinsic motivations in order to turn a situation subject
to crowding-out into a crowding-in outcome. This instrument uses a self-selection mechanism to match adequate monetary incentives with individuals' types regarding intrinsic
motivations.
10
Keywords: crowding-out, heterogeneity, moral motivation, environmental regulation.
JEL Classi cation: D03, D64, H23, Q58.
15
LAMETA - SupAgro.
UMR LAMETA, INRA, place Pierre Viala, Bât.
[email protected]
z
LAMETA - SupAgro.
y
1
26, 34060 Montpellier Cedex 1, France.
Email:
1
20
25
30
35
40
45
50
55
Introduction
Monetary incentives matter and are a powerful instrument to change behaviour. They are a part
of the story but not all the story. Indeed, the crowding-out effect, which has recently received
considerable attention in economic literature, stresses that the preferred leverage of economists,
monetary incentives and disincentives can back re and lead to inferior outcomes. This counterproductive effect of introducing monetary (dis)incentives has been theoretically (Bénabou and
Tirole, 2003) and empirically investigated (see Bowles, 2008 for a recent review). It has been
proven to be relevant in a wide variety of contexts such as blood donation (Titmuss, 1970; Mellstrom and Johannesson, 2008), acceptance of a polluting infrastructure (Frey and OberholzerGee, 1997) or to address late coming parents in day-care centers (Gneezy and Rustichini, 2000)
or bed-blocking in hospitals (Holmås et al., 2010).
Most papers consider that all agents behave similarly when facing monetary incentives.
Nevertheless, it is more realistic to assume that people are heterogeneous and have various intrinsic motivations according to the considered domain (Bénabou and Tirole, 2006; Beretti et
al., 2013). Consequently, the overall result in a whole population supporting or not the existence
of a crowding-out effect can overlook the fact that various subgroups of this whole population
react differently when facing the introduction of monetary incentives. Indeed, moderate monetary incentives can both motivate people who were originally not intrinsically motivated and
harm the intrinsic motivations of people who were originally intrinsically motivated. Moreover,
most papers to date explore the likelihood of crowding-out effects in various contexts but do
not propose alternatives to address this counterproductive result except the elimination of monetary incentives. Our paper addresses these two issues in a theoretical model that (i) allows to
take into account the heterogeneity of individuals when faced with the introduction of monetary
incentives (ii) suggests a mechanism to reduce the likelihood of getting a crowding-out result
by tapping into the heterogeneity of individuals. We assume that monetary incentives matter
and change behaviours in predictable directions according to the matching between to whom
they are directed (i.e. paying the individual versus paying the cause) and the preexisting level
of intrinsic motivation of the individual (i.e., low versus high level). Our model can explain a
large variety of outcomes stressed in recent empirical studies and has policy relevance by suggesting a new instrument which eventually turns crowding-out into crowding-in (e.g., Beretti et
al., 2013).
The remainder of the paper is organized as follows. The second section describes the hypotheses of a simple decision-theoretic approach and considers the benchmarking case where
only agents exhibiting a pristine altruism contribute to public-related cause, such as environmental conservation while non-intrinsically motivated people do not contribute. Section 3 analyzes what happens when this pristine altruism is distorted by adding monetary incentives
which generates moral repugnance. At the same time, these monetary incentives can push
non-intrinsically motivated people to contribute. Section 4 describes what happens when this
pristine altruism of intrinsically motivated individuals is distorted by adding monetary incentives but without moral repugnance because the money is directed at the supported cause and
2
not at the individual. Section 5 introduces a treatment where people can choose whether the
payment is directed to themselves or to an environmental association. Section 6 concludes.
2
60
65
70
Crowding-out with heterogenous agents: a simple model
This section constructs a simple behavioral model, in the environmental realm, in which it is
possible to explore the logical implications of various external monetary incentives on contributions when agents have heterogeneous intrinsic motivations. In the kind of situation we analyze
in this paper, people know that they affect each other by their decisions, but their interactions
are largely anonymous. They clearly don't know the set of strategies of the other individuals,
nor do they know their utility functions. Actually they even ignore how many "others" there
are. Therefore we prefer to analyze the issue using a "decision-theoretic approach", in connection with the work of Bolle & Otto (2010), rather than with a "game-theoretic approach" as in
Bénabou & Tirole (2006).
There is a continuum of agents of unit mass. Each agent i is endowed with an exogenous
income yi . The decision xi of agent i is to contribute (xi = 1) or not (xi = 0) to some environmental cause and the opportunity cost of contributing, in monetary terms, is c(xi ). The
standard assumption is that this cost is an increasing function: c (0) < c(1). Units are chosen
in such a way that c (0) = 0; c(1) = c > 0. The remainder of his income is then affected to
some alternative use ci = yi c(xi ). His decision-relevant preferences, de ned over bundles
(yi c(xi ); xi ), are captured by the following decision utility function:
U i (xi ) = yi
75
80
(1)
c(xi ) + ai t xi :
We use here the popular distinction between decision utility which prompts actions, and experienced utility, or hedonic satisfaction1 , which results from actions (see for instance Kahneman,
Wakker & Sarin, 1997).
In expression (1):
t is agent i's estimation of the marginal "monetarized" bene ts produced by his contribution xi on the other agents,
and ai 2 [0; 1] is a parameter that captures an attitude towards the other individuals via
the environment, a sort of ecologically-mediated, or `green', altruistic concern. Those
parameters are uniformly distributed on [0; 1] and each agent can be identi ed with a
particular point in this interval. At one extreme, Agent 0 with a0 = 0 does not feature
1
Experienced utility functions would rather incorporate altruism, and account for the public good created by
aggregated contributions, within a formulation like:
Z
Z
U i (xi ; x i ) = yi c(xi ) + ai
U h (xh ; x h )dh + bi
xh dh:
h6=i
3
h2[0;1]
any environmental concern; at the other polar case, Agent 1 with a1 = 1 has a strong
ecological conscience.
85
We now study the effects of different incentive schemes on contribution.
2.1
Neutral Treatment: pristine altruism alone
In the neutral treatment there is no incentive scheme and altruism is the only intrinsic motivation at work. Assume that, for the most altruistic agent, with a1 = 1; the marginal bene t
of contributing covers its marginal cost, that is c (1) c (0) = c < t . Individuals choose to
contribute or not with a view to maximize (1). Hence, agents who settle for zero contributions
are those such that:
yi c(0) > yi + ai t
c(1) ;
and the others contribute. Put differently, those in the interval
0; aN
where aN = [c (1)
interval
c (0)] =t = c=t , are non-contributors, with a mass aN ; and those in the
CN
aN ; 1
are contributors. The total number of contributors is
1
2.2
aN :
(2)
Direct Treatment (A): distorted altruism and moral repugnance
When individuals are paid for their contribution to the environment their decision utility function becomes:
V i (xi ) = yi c(xi ) + wxi + ai txi m (ai ; w; xi ) ;
(3)
90
95
where w is the monetary payment for participation. The introduction of the payment has two
effects as far as intrinsic motivations are concerned.
First, the altruistic motivation is distorted. This incentive leads the agents to revise their
estimation of the bene ts of their contribution on the environment, which is now t instead of
t ; with t < t : The idea is that the presence of a monetary transfer acts as a signal of the value
of participation (Bolle and Otto, 2010), upon which individual estimations can be anchored.
Moreover we assume that the estimation is at least equal to the payment offered (t w).
Second, agents who have a concern for the environment now suffer from a moral repugnance associated with the fact of being paid for contributing. This aspect is captured by function m (:; :; :) in expression (3). Putting a price onto a territory previously immune to the market
forces is one of the list of events that generally spark the yuck factor argument (see for instance
4
Sandel, 2012, and also the discussion about obnoxious markets in Kanbur, 2001). The monetarized value of this psychological "cost" is m (ai ; w; xi ) 0 . It is natural to assume that the
larger the green altruism ai , or the larger the payment w, and the larger the moral repugnance.
To put it formally, m (ai ; w; xi ) is non decreasing in the two rst arguments:
@
m (ai ; w; xi )
@ai
@
m (ai ; w; xi )
@w
ma (ai ; w; xi )
0;
mw (ai ; w; xi )
0:
We also assume that the marginal moral repugnance increases when altruism gets larger:
@2
m (ai ; w; xi )
(@ai )2
maa (ai ; w; xi )
0:
On the other hand, the mere fact to contribute could mitigate moral repugnance, so m (ai ; w; xi )
is non increasing in the last argument:
m (ai ; w; 0)
m (ai ; w; 1) :
Finally, without altruism, or without payment (when w = 0 or/and xi = 0), there is no moral
repugnance, therefore
m (0; w; xi ) = m (ai ; 0; xi ) = m (ai ; w; 0) = 0 :
Agents decide to contribute or not by comparing the levels of utility attached to each possibility.
De ne the utility change of contributing:
4V i (ai ; w)
100
V i (1) V i (0) ;
= w + ai t + c(0) c(1) + m (ai ; w; 0)
= w + ai t c 4m (ai ; w) ;
m (ai ; w; 1) ;
(4)
where 4m (ai ; w) m (ai ; w; 1) m (ai ; w; 0) = m (ai ; w; 1) : Function 4V i (ai ; w) is supposed to be of class C 1 . (meaning that r (:; :; 1) as a function of ai and w is itself C 1 .) Notice
that increasing the degree of green altruism can have two opposite effects on the utility change.
The rst effect is positive; it goes through the marginal bene t on others that is more valued by
a more altruistic agent. The second effect is negative, because more altruism goes along with a
more stringent moral repugnance under the direct treatment.
Contributors are those agents with 4V i (ai ; w) 0 and non contributors are agents i with
4V i (ai ) < 0:
5
Assumption 1. Let us further assume that:
lim 4V i (ai ; w) = w
ai !0
lim 4V i (ai ; w) = w + t
ai !1
d
4V i (ai ; w) = t
ai !0 dai
d
lim
4V i (ai ; w) = t
ai !1 dai
lim
(5)
c<0;
c
4m (1; w) < 0 ;
(6)
ma (0; w; 1) > 0 ;
(7)
ma (1; w; 1) < 0:
(8)
Item (5) of Assumption 1 focuses the analysis to payments w that are not high enough
to encourage participation of the less altruistic agents (the payment alone is not enough to
compensate the opportunity cost of contributing). This is the most interesting case, because if
extrinsic incentives are too strong, no crowding-out effect can occur. Item (6) of Assumption
1 means that for the most altruistic agents, contributing is not optimal because their feeling of
altruism towards others, though important, is overwhelmed by their moral repugnance of being
paid. From items (7) and (8) of Assumption 1, given that:
d2
i
2 4V (ai ; w) =
(dai )
maa (ai ; w; 1)
0;
and by the intermediate value theorem, 9ai such that t @a@ i m (ai ; w; 1) = 0: Therefore function
4V i (ai ) has an inverted U shape: it is rst increasing, until ai , then decreasing. Assume that
4V i (ai ) > 0: Then there exists a neighborhood
C D (ai )
105
110
[ai
"; ai + "] @ [0; 1]
of contributing agents around ai ; i.e. 4V i (ai ; w)
0; 8ai 2 C D (ai ) : Note that C D (ai ) ;
which is also the mass of contributors, is a proper subset of [0; 1] ; for all the elements of [0; 1]
do not belong to C D (ai ); in particular, because of parts (5) and (6) of Assumption 1, agents
a0 and a1 are not contributors. It is interesting to emphasize the peculiarity of this treatment.
The choice not to contribute can be explained by two intrinsic motivations of different natures:
a degree of altruism too low or a moral repugnance too strong (precisely in the most altruistic
agents).
For future reference, let us denote:
aD = ai " ;
aD = ai + " :
the agents who, among those who contribute, have the lowest and largest altruism respectively.
By de nition, those two values solve the equation:
4V i (ai ; w) = w + ai t
6
c
4m (ai ; w) = 0:
(9)
Does crowding-out necessarily occur? To answer this question, one has to compare the
mass of 1 aN of contributors under the neutral treatment with the mass aD aD . Unless more
structure is given to the moral repugnance function 4m (ai ; w), equation (9) cannot be solved
explicitly for aD and aD . Still, important qualitative pieces of information can be obtained. The
possibility of crowding-out depends on the relative position of aN with respect to aD and aD .
The answer is ambiguous when aN 2 aD ; aD and when aN > aD ; but there is crowding-out
for sure when aN
aD . All those situations can be associated with particular conditions on
parameters. Since 4V i (ai ) has an inverted U shape and takes on positive values around ai , by
construction aN 2 aD ; aD if and only if:
4V i aN ; w
0:
Using (4) and the fact that aN = c=t :
4V i aN ; w
0
,
w+
c
t
t
c
m
c
; w; 1
t
0:
(10)
c
; w; 1 < 0 :
t
(11)
By the same logic, if aN < aD or if aN > aD , then necessarily:
4V i aN ; w < 0
115
,
w+
c
t
t
c
m
The last inequality means that, for agent aN , the payment alone does not provide suf cient
incentives to compensate the cost of moral repugnance and the decrease in altruistic motivation
following the change of bene ts on others from t to t. This is consistent both with a too low
value for w and with a too high value (recall that moral repugnance increases with w). This
assessment of the weakness of monetary incentives is not absolute, but relative to parameter t .
So, rewriting equality (11):
De nition 1 (t -weak extrinsic incentive). Extrinsic incentives are t -weak when:
w<c
t
t
t
+m
c
; w; 1 :
t
The above reasoning has therefore established:
Proposition 1. If aN < aD ; incentives are t -weak and there is crowding out.
A last question is about the interactions between internal and external incentives. It is generally considered that the phenomenon of crowding-out gets weaker as external (monetary)
incentives gets stronger. Let us analyze the effect of increasing the external incentive w on the
upper and lower bounds of C D (ai ), i.e. on aD and aD : By equation (9) and the implicit function
theorem, using the properties that t ma > 0 until ai and t ma < 0 after ai , and under the
7
assumption that moral repugnance increases less than proportionally with the external incentive,
mw < 1; we can conclude:
da
dw
da
dw
=
1
t
mw aD ; w; 1
<0;
ma (aD ; w; 1)
=
1
t
mw aD ; w; 1
>0:
ma aD ; w; 1
a=aD
a=aD
Therefore, as w increases the mass of contributors C D (ai ) = aD ; aD gets wider. But it is
important to notice that a crowding-out is not a necessary consequence of the direct treatment.
A necessary condition for crowding in is aD < aN . This may well happen if the monetary
payment w is high enough. But this condition is not suf cient, because there is a mass of highly
altruistic agents, 1 aD ; who do not participate. By continuity, when w increases so that aD
decreases and falls below aN , and if:
da
dw
120
125
a=aD
da
dw
Proposition 2. Under Assumptions (7), (8), and when moral repugnance increases less than
proportionally with the external incentive, mw < 1; the stronger the external incentive w, the
weaker the crowding-out effect, if any, under the direct treatment.
Indirect Treatment (B): distorted altruism alone
Under this design the payment is no longer given to contributors; rather it is directed to a cause
supporting the environment, for example a related association or NGO. Hence individual i no
longer bears the cost of moral repugnance, but of course his altruistic motivation is activated,
for participation still generates a bene t to the environment.
The decision utility functions are now:
W i (xi ) = yi + ai b
txi
135
;
a=aD
that is if at the margin the increase of lower-end contributors is less than the decrease of upperend non contributors, there is a continuum of values for w consistent with crowding-in, even
though the less altruistic agent who participates under the direct treatment is less altruistic than
the less altruistic agent who participates in the neutral treatment.
To summarize,
2.3
130
<
c(xi ) ;
where b
t is the corrected individual estimation of bene ts on the environment.
Regarding the fact that a payment is directed to the cause supported by the individual, we
can reasonably consider that the individual estimation of the bene t on the environment of his
participation is higher than when the same amount of money is directed to the individual's
8
140
pocket, because the chosen destination, by its very nature, reinforces the belief of the agent on
the presence of high environmental values, or because the association is more ef cient than individuals in transforming a given amount of contributions in environmental gains. An assumption
on parameters consistent with that view is t b
t.
Hence, agents who settle for zero contributions are those such that:
c(0) > yi + ai b
t
yi
c(1);
and the others contribute. Put differently, there is an interval
0; aI ;
where
c
aI = ;
b
t
of non-contributors, and an interval
CI
(12)
aI ; 1
of contributors. The total number of contributors is
1
145
(13)
aI :
Compared to the neutral treatment there is crowding-out when b
t
t ; because the mass of
contributors has shrunk (compare expression (13) with expression (2)). This non ambiguous
result is rather intuitive: there is no direct own bene t and the estimation of the bene ts on
others has been cut down (b
t t ), so the incentives to participate are weaker compared to the
neutral treatment.
However, the comparison with the direct treatment is more subtle. It is worth noting that
there is a mass of agents near a1 who contribute under the indirect treatment and who do not
contribute under the direct treatment. In two cases, when aI 2 aD ; aD and when aI > aD ;
we cannot state which policy better encourages participation. But the indirect treatment outperforms the direct treatment for sure when aI < aD . Again all those situations can be associated with particular conditions on parameters. For aD
aI
aD , a necessary and suf cient
condition on the fundamentals of the model derives from the observation that 4V i aI ; w =
4V i c=b
t; w
0 in such a situation. Or equivalently, using (4):
w
c
b
t
t
b
t
+m
c
; w; 1
b
t
:
When this condition is not met, a necessary condition is obtained for aI < aD :
De nition 2 (b
t-weak extrinsic incentives). Extrinsic incentives are b
t-weak when:
w<c
b
t
t
b
t
+m
9
c
; w; 1
b
t
:
(14)
150
155
t-weak and crowding-out is unambiguously less important
When aI < aD ; incentives are b
under the indirect treatment. Intuitively, even if there are no monetary rewards for the agents
under policy B, altruistic motives are less corroded than under the direct treatment and, in
addition, the extrinsic motivation is not strong enough under policy A to compensate the weaker
altruistic motivation and moral repugnance.
In a nutshell:
b
Proposition 3. When the estimations of the bene ts on others are such that t
t
t , if
I
D
a < a the extrinsic incentives are b
t-weak and the indirect treatment unambiguously mitigates
the crowding-out effect compared to the direct treatment. When extrinsic incentives are not b
tD
I
weak, or when a > a , the ability of the indirect treatment to improve participation compared
to the direct treatment is ambiguous. However, in any case, participation under the indirect
treatment is never larger than under the neutral treatment, aN aI .
2.4
160
165
Choice Treatment (C): auto-selection of motivations
Under this treatment, individuals can choose whether the payment is directed to themselves
or to an environmental association. Giving the choice to individuals (keeping the reward for
themselves or giving it to the ènvironmental cause') could motivate a wider set of individuals,
possibly leading to a higher overall contribution2 .
In a sense, by choosing the target of the payment individual i chooses which decision utility
function to activate. Then, agent i's decision utility functions exists in two expressions:
it is:
V i (xi ) = yi
c(xi ) + wxi + ai txi
m (ai ; w; xi ) ;
when the payment is direct.
and it is:
W i (xi ) = yi + ai b
txi
c(xi ) ;
when the payment is indirect.
Does the choice treatment minimize the countervailing effect of external incentives?
Notice rst that the utility attached to non participation is the same, whatever the chosen
target:
U i (0) = V i (0) = W i (0) = yi :
170
But the utility derived from participation differs according to the target of the payment. Agents
who increase their utility by contributing are those who belong to at least one of the sets of
2
Moreover, in addition to obvious intuitive reasons based on empirical evidence, we argue that people enjoy
the possibility of choosing by themselves, even at a cost (Benz et al., 2004; Frey and Stutzer, 2005). This is
conceptualized through the idea of 'procedural utility'.
10
175
contributors previously identi ed. Clearly, the set C C of contributors under the choice treatment
is the union of the two sets of contributors of each separate treatment, i.e. C C = C I [ C D . The
set C C encompasses agents ai such that 4V i (ai ) 0; and/or 4W i (ai ) 0 and, therefore, the
choice treatment promotes participation as least as much as the two policies A & B separately
do. But more precision can be added.
We will keep on assuming that estimations of the bene ts of altruism are such that t b
t
t : Even under this assumption, several con gurations for the different sets of contributors are
possible:
Case 1 A rst case is when aN aI < aD ; so the extrinsic motive is both t -weak and b
t-weak
(de nitions 1 and 2). Then the different sets of contributors are such that:
CD @ CI = CC v CN :
180
Contributors under the choice treatment are exactly those who contribute under the indirect treatment and they are not more numerous than those who contribute under the
neutral treatment.
Case 2 A more interesting case is when the monetary payment is suf ciently important to produce the following ranking aN
aD < aI
aD ; that is the extrinsic motive is t -weak
but it is not b
t-weak. The sets of contributors are then in the following con guration:
CD; CI @ CN ;
CD; CI @ CC ; CC v CN :
Contributors under the choice treatment are more numerous than those who contribute
under any of the two separate treatments. But the choice treatment does not perform any
better than the neutral treatment.
Case 3 The most interesting case is when the monetary incentives are pushed slightly further
so that aD < aN < aI aD : The sets of contributors are such that:
CD; CI @ CN ; CD; CI @ CC ;
185
190
CN @ CC:
This is a case featuring crowding-out under each separate treatment, but there is crowdingin under the choice treatment. This possibility occurs because several intrinsic motivations exists and because agents are heterogenous. As a results those who contribute are not
necessarily identical across treatments and, even more, C D is neither a proper subset of
C I nor a proper subset of C N . The corresponding necessary and suf cient conditions on
parameters have been identi ed in (10) and (14). They must be imposed simultaneously,
as a new assumption:
Assumption 2 (Conditions for crowding-in).
w
c
t
t
t
+m
c
; w; 1
t
and w
11
c
b
t
t
b
t
+m
c
; w; 1 :
b
t
195
aI : It is not possible to conclude - without further
Case 4 Finally when aD < aN < aD
information on the moral repugnance function - about the extent of the crowding-out phenomenon, if any, because there is a mass of agents characterized by intermediate degrees
of altruism in the interval aD ; aI who are not contributors. However, this situation is
discarded when the extrinsic motivation is not b
t-weak.
To summarize, the choice treatment combines the incentive effects of both the direct and
indirect treatments:
200
Proposition 4. Let Assumption 1 holds and assume also t < b
t
t . Participation under
policy C (choice treatment) is at least as large as under policies A & B. The choice treatment
even results in crowding-in, although there is crowding-out under policies A & B, if and only if
Assumption 2 is satis ed.
3 Conclusion
205
210
215
We made a strong case for how different intrinsic motivations among agents can play an instrumental role in explaining the effectiveness of introducing monetary incentives. We have
formalized how the heterogeneity of intrinsic motivations among agents impacts their reactions
to the introduction of monetary incentives. We showed that overall results supporting (or not) an
undesired crowding-out effect can occult a more complex reality where some individuals contribute thanks to these additional monetary incentives while others reduce their contributions.
Moreover, we proposed a new instrument which taps into agents' heterogeneity in order to suppress, or at least to reduce, the risk of crowding-out effect. This instrument avoids a òne-sizets-all' policy and allows agents to self-select the most relevant arrangement. A considerable
advantage of our mechanism is that it does not require that policy makers have an extensive
knowledge about the various levels of intrinsic motivations of agents. In sum, we believe that a
ne analysis of agents' heterogeneity can usefully complement studies of crowding-out effects
and lead to innovative and more effective policy interventions. Last, but not least, our paper
also acts as a call for further empirical research about the heterogeneity of intrinsic motivations.
References and Notes
220
[1] Bénabou R., Tirole J., 2006, Incentives and Prosocial Behavior, American Economic Review. 96, 1652-1678.
[2] Benz M., Frey B.S., Stutzer A., 2004, Introducing Procedural Utility: Not Only What, but
Also How Matters, Journal of Institutional and Theoretical Economics JITE. 160, 377-401.
12
225
[3] Beretti A., Figuières C., Grolleau G., 2013, Using Money to Motivate Both Saints and
Sinners: a Field Experiment on Motivational Crowding-Out, Kyklos, 66/1, 63-77.
[4] Bolle F., Otto P.E., 2010, A Price Is a Signal: on Intrinsic Motivation, Crowding-out, and
Crowding-in, Kyklos. 63, 9-22.
[5] Bowles S., 2008, Policies Designed for Self-Interested Citizens May Undermine "The
Moral Sentiments": Evidence from Economic Experiments, Science. 320, 1605.
230
[6] Frey B.S. , Oberholzer-Gee F., 1997, The Cost of Price Incentives: An Empirical Analysis
of Motivation Crowding- Out, The American Economic Review. 87, 746-755.
[7] Frey B.S., Stutzer A., 2005, Beyond Outcomes: Measuring Procedural Utility, Oxford
Economic Papers. 57, 90-111.
235
[8] Kanbur R., "On Obnoxious Markets", 2001. Revised version published in Stephen Cullenberg and Prasanta Pattanaik (editors), Globalization, Culture and the Limits of the Market:
Essays in Economics and Philosophy. Oxford University Press, 2004.
[9] Kelly D., 2011, Yuck! The Nature and Moral Signi cance of Disgust, MIT Press.
[10] Gneezy U., Meier S., Rey-Biel P., 2011. When and Why Incentives (Don't) Work to Modify
Behavior, Journal of Economic Perspectives, 25(4), 191-210.
240
[11] Kahneman D., Wakker P., Sarin R., 1997. Back to Bentham? Explorations of Experienced
Utility, The Quarterly Journal of Economics, 112(2), 375-405.
[12] Mellström C., Johannesson M., 2008, crowding-out in Blood Donation: Was Titmuss
Right?, Journal of the European Economic Association. 6, 845-863.
245
[13] Reeson A.F., Tisdell J.G., 2008, Institutions, motivations and public goods: An experimental test of motivational crowding, Journal Of Economic Behavior and Organization.
68, 273-281.
[14] Sandel M., 2012, What money can't buy: the moral limits of markets, Allen Lane, A/L
PENG PRESS.
250
[15] Titmuss R.M., 1970, The gift relationship: From human blood to social policy, Allen and
Unwin.
[16] Tor Helge Holmås, Egil Kjerstad and Hilde Lurås, Odd Rune Straume, 2010, Does monetary punishment crowd out pro-social motivation? A natural experiment on hospital length
of stay”, Journal of Economic Behavior & Organization, 75, 261-267.
13

Author`s personal copy

Transcription

Similar documents

Spring Programming - Cornwall Public Library

News Letters - Collections

To view this issue please click this link

journal - PHE Canada

Untitled - System 24

Le Canard des Tropiques - Alliance Française de Townsville

contents - Geneva Association

here - South Asian Women`s Community Centre

Le journal Underground

Mont-Tremblant / Tremblant (Station)