Behavioral Development Economics
* Ravi Kanbur
The “Stiglitzian” paradigm is alive and well in modern development economics. Indeed, in many ways it might be said to dominate analytical development economics. Taking account of the imperfections of information and markets to understand and explain phenomena in developing countries, following on from Joe’s seminal papers on sharecropping, urban unemployment, credit constraints, and so on, is now well established. Thus key aspects of the “textbook model” are no longer dominant in analytical development economics, although there is a fight to bring these perspectives into the policy discourse on development.
But there is one aspect of the textbook model that has been central to the Stiglitzian paradigm, and is therefore also central to modern development economics. This is the assumption that individuals behave according to the textbook axioms of rational choice. In all of his Nobel Prize winning work modeling the consequences of imperfect information, or in his hugely influential work on imperfect competition, or on imperfect markets more generally, rational choice according textbook axioms is the unexamined assumption.
It is a mark of the influence of Joe’s work on modern development economics that
it too is characterized by a thoroughgoing acceptance of the textbook rational choice framework. Neither in Joe’s models, nor in the models of modern development economics, do individuals procrastinate; overweight low probability outcomes; take the current wealth position as a reference point from which departures are assessed; change choices when alternatives are presented with a different default option; leave dollar bills lying on the sidewalk because they disapprove of the process by which they got there; and so on.
But these and other “imperfections” of individual choice behavior (i.e. departures
from the textbook model) are increasingly accepted by the profession as at least viable empirical phenomena to be explained and incorporated. In the last two years, Sendhil Mullainathan won the MacArthur “genius” award, Matt Rabin the American Economic Association’s Clark Medal, and Daniel Kahneman and Vernon Smith the Nobel Prize, all
for their work in Behavioral Economics.
* T.H. Lee Professor of World Affairs and Professor of Economics, Cornell University. Notes for brief comments at the conference on Economics for An Imperfect World, in honor of Joseph Stiglitz, Columbia University, October 24-25, 2003. The notes draw on my 2002 paper, “Conceptual Challenges in Poverty
and Inequality: One Development Economist’s Perspective.” The paper can be downloaded from
Three decades ago, when Joe was writing his seminal papers on imperfect information, the leading journals were full of articles based on the textbook model, including perfect markets, perfect information, and rational choice. At that time, the development literature talked much about various imperfections, including that of information, markets and individual choice, but that literature was not in a form that could have an impact on the profession. As we all know, thanks to the work of Joe’s
generation, and thanks to the work of his students’ generation, the leading general
journals started to publish articles on imperfect information and imperfect markets, development economics entered the mainstream, and the development journals followed suit.
Thirty years later, however, there is one striking difference between the leading general journals in economics and the leading development journals. The general journals are increasingly publishing articles in behavioral economics, articles that integrate insights from psychology to enrich models of individual choice and entertain departures from the textbook axioms of rational choice. But the development journals are by and large stuck in the rational choice paradigm. Thus the leading general journals have articles which address the question of why New York city cab drivers stop work early on rainy days; why Ulysses would have himself tied to the mast; why workers choice between pension plans changes when the choice is presented in a different order; why deadlines matter; why individuals reject highly unequal division of a pool of money even if it means losing out themselves; why individuals seem to keep separate mental accounts for particular flows of income and expenditure and do not seem to operate with a unified budget constraint; and so on. However, to caricature somewhat, the articles in the leading development journals seem to start with the obligatory, “We assume the peasant
maximizes a Utility function U(….) subject to a single budget constraint.”
But, of course, the landscape of development economics is changing fast. There is now a mini-Industry of graduate students doing trust experiments and choice under risk experiments in developing countries. Standard micro data, which in the last twenty years have been analyzed within a rational choice framework, are now being analyzed in a broader framework, revealing that individuals in developing countries, no less than their counterparts in developed countries, seem to display similar psychological dimensions to behavior. Such frameworks may better explain, for example, why sharecropping contracts usually specify a share of a half across a wide variety of contexts; why certain types of public works schemes fail to attract the obviously poor and starving to participate even though it seems to be in their simple rational interest to do so; why changing relative prices may not change resource allocation as much as might be predicted by standard analysis; why peasant behavior may appear to be more myopic than predicted by standard analysis; why peer monitoring in credit may work, over and above the usual informational arguments; why reduction in poverty as measured by the usual consumption based measures may not correspond to people’s ground level sense of improvement in wellbeing; and so on.
None of the above is to say that the Stiglitzian paradigm in development economics is to be jettisoned. Imperfect information and imperfect markets will continue
to play a central role. But an Economics for an Imperfect World must surely also take on board “imperfections” in individual choice, in the sense of entertaining departures from the textbook axioms of rational choice. The mainstream economics profession has already begun to do so. Development economics should follow suit.