Behavioral Development Economics

This article is part of our special issue “Nudge Turns 10,” which explores the intersection of behavioral science and public policy. View the complete issue here.

Behavioral economics is often studied in the contexts of consumer behavior—overeating, going to the gym, doing your homework, and other relatively “first world” problems. However, for many people around the world, behavioral biases and irrationalities can have more pernicious effects on well-being.

Consider, for example, farmers in Kenya, deciding whether or not to purchase fertilizer for the upcoming season. Purchasing it is a hassle—they have to go to the store, pick a fertilizer, get it home, set up a place to store it, and so on. In other words, this is the exact kind of behavior we humans are hardwired to put off until the last possible moment (think about your taxes—when do you usually do those? Early April? Yep, us too!). As a result, farmers often put off fertilizer purchase day after day until it is “too late,” and they end up not fertilizing their crops properly. Months later, they suffer when their farm yields are much smaller than they could be. Their families eat less, and perhaps most importantly, they lose disposable income, which is critical for people living on the financial edge. Less money can mean less schooling for children, less medicine, and precarious living conditions. What to do?

Behavioral science has come a long way in the past 50 years. While many of the early, pioneering studies took place in sanitized “lab” environments, with subjects from Western countries, the past decade has seen an explosion of behavioral science research in the messier environment of the developing world. This work has given us greater insight into how and why the world’s poorest populations make the decisions they do. But perhaps more importantly, this work has allowed behavioral scientists to directly improve the well-being of the world’s poorest and most vulnerable populations.

Behavioral economics is often studied in the contexts of consumer behavior…and other relatively “first world” problems.

Back to our Kenyan farmers. A group of researchers noticed the systematic underuse of fertilizer by farmers and suggested a solution based on behavioral research: offering these farmers free delivery on fertilizer at a better moment in time, namely when farmers are flush with cash right after the harvest. The researchers tested this solution and found that free delivery of fertilizer had minimal impact when it was offered at the “wrong” time (growing season, when money was scarce) but had huge impacts when it was offered at the “right” time (just after the harvest, when farmers had some cash). One little, logical, well-timed “nudge” was all that was needed to help farmers make a better decision.

Research in this relatively new and growing field of “behavioral development economics” can be broadly classified into two categories.

The first category of research involves work that gives us insights into how people from poorer countries behave, and why they behave the way they do. For example, Coren Apicella and her colleagues studied the endowment effect (our tendency to raise our valuation of goods once we own them) with Hadza bushmen, one of the last hunter-gatherer populations of the world. They found that the Hadza living in areas with greater exposure to modern society display the endowment effect, whereas the ones who live in isolated regions do not. This suggests that behavioral biases may have roots in cultural factors.

Recent work has given us an understanding of another way in which poverty in the developing world impedes the ability to escape it.

New research suggests a critical role for environmental factors as well. For example, Josh Dean and a team of researchers used a lab-in-the-field experiment in Kenya to study the effects of ambient noise (ubiquitous in the developing world) on productivity. He found that an increase of 10 decibels in background noise reduced worker productivity by about 5 percent, by impeding cognitive functions such as attention and working memory. Similarly, Matthieu Chemin and his collaborators found evidence for a causal link between poverty and stress. Specifically, the authors studied the cortisol levels (a biological measure of stress) of farmers in Kenya and found that low levels of rain in the preceding year caused farmers to become a lot more stressed. Why? Farmers need rain for their crops! So less rain means more stress, which means a worse quality of life. Indeed, Anandi Mani and her colleagues found that the cognitive capabilities of farmers are dependent on the harvest cycle—farmers show diminished cognitive capabilities before harvest (when poor) as compared to after harvest (when rich).

Taken together, this recent work has given us an understanding of another way in which poverty in the developing world impedes the ability to escape it. It seems that psychological traps of poverty are formed from years of living in extreme stress and suffering from mental taxation, and the issues that come with both, including poor nutrition, vulnerability to addictions, and myopic decision-making. A new generation of researchers affiliated with institutions like the Busara Center for Behavioral Economics, The Abdul Latif Jameel Poverty Action Lab, Precision Agriculture for Development, and ideas42, among others, continues to explore the mechanisms that go into these psychological poverty traps. These efforts can help policymakers better understand how to best assist those living in poverty in the poorest places on earth.

The second category of research involves the actual design and testing of behavioral interventions meant to improve the lives of the very poor. In an early, pathbreaking study, Nava Ashraf and her colleagues developed and offered to account holders in a rural bank in the Philippines a “commitment device” for savings. Specifically, they created a bank account that restricted access to deposits (without interest), and found that such accounts are surprisingly popular (despite the fact that there is no rational reason to want them). Similarly, Frank Schilbach conducted a study with poor cycle-rickshaw drivers in Chennai, India, finding that they struggled to control their consumption of alcohol. He devised a commitment device that provided a high-return savings opportunity and incentives for sobriety.

In the domain of insurance, Lorenzo Casaburi and Jack Willis studied the timing of payment for a crop insurance premium in Kenya. They found that by deferring the time of payment of the premium to the time of the harvest (when money is less scarce for farmers), they were able to increase the take-up rate of crop insurance from 5 percent to 72 percent, an astounding increase.

Researcher-led interventions have also had sizable health impacts; Erez Yoeli and his colleagues (including Syon Bhanot, a co-author of this piece), for example, found that a simple behaviorally-informed text message platform designed to encourage tuberculosis (TB) medication adherence in Kenya was able to reduce failure rates for TB treatment from 13.1 percent to 4.2 percent.

If the last decade has taught us anything, there is a growing vanguard of researchers willing to break the academic mold and apply themselves to help the world’s poorest populations.

That said, there are important topics in the developing world in need of further research. For any policy to work effectively, the basic infrastructure of government must work in tandem—which is often not the case in developing countries. In many parts of the developing world, corruption, delays, and sloppy governance are commonplace. Behavioral science has the potential to help us understand the causes of these problems. Interventions that encourage better processes are urgently needed.

Furthermore, as automation gradually reaches the developing world, social welfare programs need to be adapted to the changing times, as more and more people may face reduced opportunities for work. Research on the potential of new approaches for supporting growing populations during the rise of automation, such as universal basic income or public works programs, is vital for our understanding of globalization and how we can build social systems that are resilient to its effects. But these challenges are not only critical in the developing world. They are also vital in the so-called “developed” world, where these issues are often just as acute. This provides the developed world with an opportunity to learn from the policy interventions of the developing world, thus facilitating a two-way street of exchange of ideas.

To address these big challenges, it is critical for researchers to partner more with governments and key stakeholders on the ground in the developing world and to develop interventions that use the insights from behavioral science to do good. This will not be an easy process. Needless to say, academics often have different incentives than policymakers in the developing world. But if the last decade has taught us anything, there is a growing vanguard of researchers willing to break the academic mold and apply themselves to help the world’s poorest populations.