Blog posts

image

November 26, 2012

New Research: Three papers from Sendhil Mullainathan

By Thea Garon

We do our best (not always successfully) to keep up with new research relevant to finance, poverty and development. Today, I’ll be sharing highlights from some new papers by FAI affiliate Sendhil Mullainathan.

In “Behavioral Design: A New Approach to Development Policy,” Mullainathan and Saugato Datta advocate for employing a behaviorally-informed economic perspective to design development policies and programs. Since behavioral economics helps us understand why people behave as they do, analyzing development policies through a behavioral lens allows us to make better policy diagnoses, which in turn lead to better-designed policies.

Mullainathan and Datta outline three ways in which behavioral economics can improve program design. First, it can change how we diagnose problems. For example, it might be tempting to think that parents don’t vaccinate their children because they don’t understand the importance of vaccines. However, behavioral economics encourages us to consider whether parents are failing to vaccinate not because they don’t recognize the value of vaccines, but because they simply don’t get around to doing it. Second, a behavioral economics perspective can change how we design solutions to problems. In some cases, this could be as simple as sending a reminder or reducing obstacles to action. And third, behavioral economics can change how we define the scope of a problem. We're often tempted to focus on increasing access to a product (such as a drug), but behavioral economics encourages us to think beyond the access component to additional factors such as whether people are actually taking the drugs they receive.

Mullainathan puts this behavioral economics approach to practice in a recent experiment conducted with seaweed farmers in Indonesia. In “Learning through Noticing; Theory and Experimental Evidence in Farming,” Mullainathan and his co-authors Rema Hanna and Joshua Schwartzstein explore whether traditional learning models accurately represent people’s learning processes. Traditional learning models suggest that one of the primary reasons why people fail to learn is because they lack sufficient information. That view suggests information interventions—but anyone who has been involved in community education initiatives can tell you that just making information available doesn’t necessarily  change people’s behavior. After observing a program designed to increase the productivity of seaweed farmers, Hanna, Mullainathan and Schwarzstein conclude that a failure to notice available information might actually better explain why people fail to change their behavior even when presented with good information.

In their experiment, the researchers introduce a new process that allows the seaweed farmers to increase their productivity. The farmers do not, however, adopt this process because they fail to notice important features of it. Only when the researchers presented the farmers with summary data highlighting the important aspects of the process did the farmers correctly replicate the process. We’ll be blogging more about this paper and its implications for financial access interventions soon.

In “Some Consequences of Having Too Little,” Mullainathan further explorers how a behavioral perspective can change the way we understand people’s decision-making behavior. Mullainathan and coauthors Eldar Shafir and Anuj Shah examine how scarcity itself can influence people’s behaviors using a series of experiments that create scarcity. In their first experiment, they measure whether scarcity leads people to borrow excessively. They set up a game in which participants are allocated time budgets of different sizes for multiple rounds of a game. Some participants can borrow time across rounds cheaply (ie. without interest), some can borrow time expensively (ie. with interest) and some cannot borrow time at all. The authors found that “time poor” participants borrowed a much higher proportion of their time budget than “time rich” clients. In a second experiment, the researchers revised the game so that excessive borrowing would essentially lead participants into a state of debt. Even with this disincentive for over-borrowing, “time poor” participants accumulated significantly more debt than “time rich” participants.

In two additional experiments, the researchers slightly modified the game to test how other dimensions of scarcity might affect borrowing habits. They found that “time poor” participants exhibited constraint focus – the tendency to shift one’s attention toward areas of perceived scarcity and neglect areas of perceived abundance.

The authors acknowledge that these experiments are “far-removed from the complexities of poverty,” yet therein lies their power. Because we rarely encounter scarcity in an abstract form, it can be challenging to measure the effects of scarcity itself rather than those of the larger context in which it is found. Mullainathan and his colleagues’ experiments reveal how the very act of having less can lead to detrimental decision-making that has very real and profound consequences for those living in a state of poverty. Again, we’ll be writing more about this paper in the future. 

Blog Tabs