One hopes that decision-makers in the Indian government will make administrative data readily available to researchers from all nations for its own development. There is much to be learned from administrative data despite the challenges in identifying clear causality between policy and outcome
This year’s Nobel prize in Economics was particularly noteworthy for several reasons. For the world as a whole, the fact that one of the winners, Esther Duflo, was only the second woman—and, at 46, by far the youngest ever—Economics winner stood out. Duflo has been vocal in pointing out the deficiencies of the profession in terms of making women feel welcome and valued. Her prize will help accelerate a corrective process already underway.
The prize, of course, was based not on gender, but the work done, and it was noteworthy also for recognising a methodological approach that has sought to understand the causes of poverty by doing field experiments, or randomised controlled trials. Duflo’s co-winners, Michael Kremer and Abhijit Banerjee, while a decade or so older, are still relatively young prize winners, which adds to the statement being made about the value of this kind of research. The benefit, of course, is precision with respect to causes—if one is interested in whether A causes X (and how much is the impact), controlling for other factors, applying the ‘treatment’ A, and having a benchmark, or ‘control’ (‘not A’), to further isolate the impact of A on X is much more reliable than using data that was collected for other purposes. In the latter case, one might pick up correlations and confuse them with causality, or be unable to control for other factors, and end up with misleading estimates of impacts. I will return to the methodology shortly.
Kremer was cited for pioneering experimental work in East Africa (interesting fact: my colleague, Jonathan Robinson, once co-authored an important paper with Duflo and Kremer, explaining seeming puzzles in the use of fertiliser by Kenyan farmers, and he has gone on to become an important practitioner in this academic area), but Banerjee and Duflo have mainly worked in India. Indeed, Banerjee and Duflo are husband and wife, and the supreme academic power couple. Their story, and the fact that they have also been producing non-technical books that explain the relevance of their work (Poor Economics, published in 2011, and Good Economics for Hard Times, being published next month), has been dominating the headlines, overshadowing their co-winner to some extent.
Of course, the fact that Banerjee is of Indian origin, and has been commenting and advising on Indian economic policy, has been extremely important for a nation that has been somewhat starved of good news in recent months. There are certain ironies in his academic pedigrees from India—Presidency College and Jawaharlal Nehru University—in terms of the state of higher education in India, as well as national and state politics. For the moment, everyone is celebrating his prize, and he is “fully Indian,” something that Raghuram Rajan was once accused of not being. Coincidentally, Banerjee and Rajan wrote and spoke together exactly on the day the prize was announced; Banerjee’s recommendations included prescriptions for fiscal and monetary policy in the short and medium terms, as well as the suggestions that might most appeal to the government—’pray’ in the short run, and ‘pray more’ in the longer run. Banerjee and Rajan were two of over a dozen co-authors of a report on India’s economic strategy, which I commented on in these columns at the beginning of 2019.
Banerjee and Duflo, and many others, have made working on the Indian economy more mainstream, and respected in the economic profession in the US—a welcome outcome of their passion for doing relevant economic research. I have previously highlighted the portal, Ideas for India, where non-technical accounts of much of this work can be found. A look at the range of that research reminds us that there is much to be learned from so-called administrative data (such as household surveys) as well, despite the challenges in identifying clear causal relationships between policy ‘A’ and outcome ‘X.’ Diane Coffey and Dean Spears used different kinds of data—and not just field experiments—to make the case for paying more attention to sanitation and access to toilets, with visible impacts on policy and politics. In other cases, such as the status of household savings in India, the only useful data may be from large-scale surveys, and not smallish field experiments.
One hopes, therefore, that after the immediate glow of the prize has passed, decision-makers in the Indian government—both politicians and bureaucrats—at the state and national levels will be more and more open to making administrative data readily available to researchers from all nations, and giving freedom to researchers to conduct field experiments. It must be noted, university researchers are already subject to strict standards for how they deal with human ‘subjects’ by their home institutions. A corollary of this openness is being willing to allow NGOs or non-profits the flexibility to fund, and collaborate on these studies since they can often bring in local, ground-level expertise that makes the research more reliable, or make it possible to do the ‘trials’ at scales that are more informative, and policy-relevant. One should also remember that bureaucrats can be an important source of information for asking the right questions, and designing the right policies to test—strengthening that avenue of interaction could also reap rewards for India’s economy.
Professor of Economics, UC Santa Cruz. Views are personal.