by Pranab Bardhan
As a development economist I am celebrating, along with my co-professionals, the award of the Nobel Prize this year to three of our best development economists, Abhijit Banerjee, Esther Duflo and Michael Kremer. Even though the brilliance of these three economists has illuminated a whole range of subjects in our discipline, invariably, the write-ups in the media have referred to their great service to the cause of tackling global poverty, with their experimental approach, particularly the use of Randomized Control Trial (RCT).
Of course, the Prize as such is not for great policy achievements in poverty reduction (if it were, the Chinese policy-makers enabling the lifting of nearly half a billion people above the poverty line in their country would have got prior attention), but for methodological breakthroughs, which the pioneering effort in extensive application of RCT in field experiments in several poor countries clearly is.
I should also proudly point out that these three economists did some of their major work while they were active members of a MacArthur Foundation-funded international research group on Inequality that I co-directed for more than 10 years starting in the mid-1990’s. Duflo was the youngest member of our group; incidentally, the French speakers (including, apart from Duflo, Thomas Piketty, Philippe Aghion, Roland Benabou, and Jean-Marie Baland) and the Bengalis (apart from myself, Abhijit Banerjee and Dilip Mookherjee) were together nearly half in strength in this group of about 18 members from different countries and social science disciplines.
The media write-ups on the Nobel Prize (including in leading magazines like The Economist), however, give a somewhat misleading impression about the evolution of thinking in development economics, as if after decades of pontification on structural transformation and prudential macro-economic policy and associated cross-country statistical exercises to understand the mainsprings of growth and development, the practitioners of RCT finally came along focusing our attention to the micro level, and providing us with a magic key, the so-called ‘gold standard’ in assessing poverty alleviation policies, telling us what ‘works’ at the ground level of policy intervention and what does not.
First of all, this kind of write-up shows some ignorance of a large, several-decades long (starting at least in the 1970’s), literature in development microeconomics, both theoretical and empirical, the latter often applying non-experimental methods to scrupulously collected field-level data (measuring at least up to a ‘silver’ standard from the point of view of avoiding various kinds of statistical bias) to a diverse range of important policy and institutional questions related to the roots of endemic poverty.
Secondly, there have been critics of the application of RCT (including some Nobel laureates) and doubts about how ‘golden’ the standard really is and how generalizable and reliable the micro results are. To be fair, the current laureates have responded to these critics in their recent work, trying, for example, to face the issues in scaling up their results from the individual micro studies.
Thirdly, there are diverse ranges of pressing development policy questions where the method of RCT, by its very nature, cannot be applied at all (for example, in many types of monetary-fiscal policy issues or questions like where to locate a power plant or a port or the feasibility of other infrastructural projects), even though here too the experimental approach has now been diversified beyond the earlier preoccupation with health and education interventions in localized contexts to, for example, larger issues of governance or information networking.
The profession of development economists should be grateful to the new laureates (as well as to the four laureates earlier—Lewis, Schultz, Sen, and Deaton) for bringing the limelight to their particular field, which for too long has been relegated to the periphery of the subject of Economics, even though the forefathers in the discipline—the classical economists—were all development economists as they were writing about a developing economy (more often than not, Britain) that was going through a major industrial transformation in the late 18th and the early 19th century. With the glitter of the Prize and the associated public attention more of the younger generations of bright people are now likely to be attracted to this field.
But there is also a risk that the glamor of RCT may divert attention from some of the big questions of development economics that remain unresolved. I think the new laureates will agree with me that we do not still have a clue why some developing countries succeed (some of them in East Asia), and why others do not, and why some initially successful cases cannot escape from stagnating at the middle stage. Why do some dysfunctional institutions persist? Why is it so difficult to create good jobs for the restive young population in today’s poor countries aching to get out of the low-productivity informal sector? What are the ingredients of state capacity involved in making industrial policy more of a success in East Asia than elsewhere? Are these mere historical contingencies or are there meaningful patterns in the diverse case studies? Our discipline is as yet in its infancy in understanding the forces and motivations behind formations of political coalitions and different kinds of group bargains that work in different historical contexts, which underpin the process of institutional change or atrophy. Taking the major coordination mechanisms of any economy—the market, the state, and the community—we are now as familiar with their ‘failures’, as with their successes; but how to strike a balance in their operation so that we can sustain their advantages without getting crippled by their failures? It may be premature to try to answer these questions with the same methodological fastidiousness with which the experimental approach to global poverty alleviation policies is now handled. But I think in its search for theoretical and empirical rigor the profession undervalues the need for in-depth country or regional studies of political and economic processes, which sometimes provide deeper insights into the origin and persistence of poverty than those gleaned from either cross-country standardized data or micro experiments. The latter may be more journal-publishable or headline-grabbing, but not necessarily more valuable in the larger scheme of things.
In a world of crushing inequality and aggregate demand stagnation and the associated political-economy impasse and the prevailing macro-pessimism, the new laureates bring some cheer in showing us how changes at the margin and small policy improvements are feasible using statistically ‘clean’ methods, and that incremental changes steadfastly pursued can add up to a lot. While we should be thankful for such mercies (which are often the outcome of a great deal of research effort and money—for RCT’s are usually frightfully expensive), the way experimental methods have now swept most other types of research work in development economics off the radar (and the top journals), one can only hope that the Prize, instead of exacerbating this trend, will reinvigorate the whole field and encourage diverse attempts to understand a whole range of issues, big and small.