Behavioural nudges may be a part of good economic policymaking, but much more will be involved. Tweets, even if it is a Thaler that sends them, will add nothing.
Making good economic policy anywhere requires a deep understanding of human behaviour (including societal norms), empirical knowledge of responses to changes in prices at the individual and systemic levels, and careful design of policies to take this knowledge into account. In my last column, I discussed some challenges of improving sanitation and household finance outcomes in India, in the context of these issues of policymaking. Since then, Richard Thaler’s Nobel Prize in economics has focused attention on his advocacy of behavioural nudges as a policy tool, among his other insights into the quirks of human behaviour in the sphere of economic decision-making. The Tarun Ramadorai report has mentioned using nudges to improve financial decision-making in savings and insurance, something Thaler has written about in detail. Previously, India’s NITI Aayog had supposedly set up a nudge unit to aid policy design and implementation. However, it is not clear that they have made any specific policy design recommendations that have been true “nudge” type interventions, and there is no evidence that anything they have done has successfully leveraged Thaler’s insights.
Most controversially, it has emerged that Thaler tweeted support for India’s demonetisation of 500- and 1,000-rupee notes, though not for the introduction of 2,000-rupee notes. It may be that Thaler saw this as a “nudge” toward using digital technology for transactions, which is what the Indian government claimed. However, the policy was more like a push off a cliff rather than a nudge, and, whatever Thaler’s other intellectual accomplishments, his tweet on Indian demonetisation must be one of his most ill-informed and misguided judgements. It is unfortunate that the Nobel Prize and the superficial simplicity of the “nudge” concept are distracting attention from what really needs to be done for good economic policymaking in India. One deficiency of demonetisation, and Thaler’s approval of it, was the lack of empirical analysis or evidence to buttress the policy decision. In previous columns, I have highlighted how Ideas for India has become a portal for non-technical presentations of empirical evidence that can guide policy, including in areas such as sanitation, education and health. But are these enough? Recently, five researchers, in a piece on the same site, plainly titled Three barriers that make it hard for policymakers to use the evidence that development researchers produce, try to shine a light on remaining gaps between research and policymaking. They posed various questions to over 1,600 civil servants (about 1,500 in Pakistan, the rest in India). These were designed to understand these bureaucrats’ knowledge and incentives in interpreting and using evidence generated by economic research.
The three barriers are: (1) policymakers can misinterpret quantitative evidence, such as confusing relative comparisons of absolute numbers and proportions or rates; (2) organisational and institutional barriers prevent the effective use of evidence, as when politicians impose their preferences without regard for broader economic or social impacts; (3) policymakers use qualitative and quantitative evidence to update their beliefs in ways that are not always objectively “rational”. Note that the first and third barriers are related, and have roots in the broader behavioural factors that Thaler’s research (and that of Amos Tversky and Daniel Kahneman before him) has illuminated. This simply means that researchers have to work harder to convey the meaning and implications of their results. A complementary improvement could be better training of policymakers, although one suspects that all of these efforts might be overwhelmed by the second barrier, which seems to have been at work in the demonetisation debacle. One can also note that policymakers may not place enough trust in certain kinds of research: a localised experiment under artificially controlled conditions may not give them the confidence to extend or scale up an innovation. Again, researchers have to address this credibility gap.
Making democracy work better, which is what overcoming the second barrier requires, is a long-run desideratum, but there can still be smaller improvements. Recently, Ajay Shah has critiqued the current state of public policy making in India, using the symbol of the deficiencies of the “three page note” to argue against lack of depth of empirical knowledge and analysis. He uses the lack of understanding and progress in dealing with India’s banking crisis as an example. Another example is monetary policy, where there is basic conflict among different economists on interpreting the data, as well as on the mechanisms of impact of changing policy variables such as the interest rate.
Shah argues for “teams that combine intellectual capacity, empowerment and long time horizons.” The question this raises is why does the monetary policy apparatus, or that for dealing with the banking crisis, or the sanitation crisis, or anything that NITI Aayog seems to be charged with, not have these characteristics? Is there a lack of talent (probably not) or is it a lack of political or administrative will? To return to the Tarun Ramadorai report on household finance, how will its many recommendations be refined, tested and implemented? It will certainly require many specialised teams with intellectual capacity as well as domain expertise. Behavioural nudges may be a part of this policymaking, but much more will be involved. Tweets, whoever sends them, will add nothing.
Professor of economics, University of California, Santa Cruz