The American Economic Association (AEA) wrapped up its annual three-day meeting last weekend with the 2013 John Bates Clark Medal recipient Raj Chetty addressing a packed ballroom in Boston’s Sheraton Hotel. Known for his keen eye for the importance of details often ignored by other researchers, Chetty used the spotlight to make the case for a pragmatic approach to incorporating behavioral insights into economics. He outlined why behavioral factors should be used when they fit the data, and how these behavioral insights can lead to better public policy outcomes in three concrete ways: providing new tools, offering better predictions, and through new welfare implications.
One well-known tool in the behavioral economist’s toolkit is the default option. Chetty noted that making enrollment in a retirement plan the default option increases enrollment from 20% to 80%. Perhaps of more importance however is a bigger question: do these defaults actually increase overall levels of savings compared to the standard approach of subsidizing employee contributions? In short, the answer is yes: not only do default options increase total savings, but these savings increases persist, lasting for more than a decade and resulting in experienced gains in retirees’ wealth.
Chetty also showed how behavioral principles can be used to help design policies that might not even seem, at first, to be linked to behavioral economics. He discusses how the Earned Income Tax Credit (EITC) has affected the way families earn income. If low-income families were optimizing their income in response to the EITC, we’d expect to see that many of them would earn an income close to $12,600 (at which point the tax credit decreases). In some areas, we see exactly this behavior. However, there are many areas of the country where families do not seem to be responding to this incentive, and there is no change near that threshold. Chetty theorizes that this is simply because, in some areas, people do not know about the EITC. After all, you are not going to respond to an incentive you do not know about!
In a final example, Chetty outlined how utilizing a behavioral approach offers new welfare implications. The well-known “Moving to Opportunity” experiment showed that moving from a high-poverty area to a low-poverty area can increase future earnings potential for children under 13 by 30%. However, despite only slightly greater costs, many parents don’t move their children to these neighborhoods. Why? A neoclassical economist would argue that perhaps families are maximizing in their current neighborhood and placing a low weight on their child’s future earnings. A behavioral approach, however, recognizes that parents don’t make the move because of present bias—in which the immediate costs of moving have a tendency to outweigh the far off, long-term benefits for children. He also pointed out that this bias is exacerbated by poverty-induced scarcity making it difficult for parents to conduct long-term planning when faced with short-term stresses.
On these more divisive issues, Chetty emphasized the need to use the model that best fits the specific context. He downplayed what he views as an arbitrary and improper distinction between behavioral and neoclassical economics. In many respects, it doesn’t matter if a given model is “correct”, but whether that model gives us useful insights. Using Chetty’s framework, “nudges” are “Optimal Policy with Model Uncertainty”. In other words, if you are not sure why people are behaving strangely, a nudge can be an effective and cautious tool. If it doesn’t work, it might be because people are optimizing correctly.
The behavioral successes cited in this lecture reflect many of the key insights informing the work ideas42 is doing to impact real lives for social good. Chetty’s embrace of employing behavioral models alongside economic methods in a setting as prestigious as the AEA speaks to the massive ground behavioral science is gaining within public policy.