In the standard account in economics, people are self-interested. In other words, individuals “free ride,” taking advantage of others, whenever they can. But is that how people actually behave?
A group of researchers wanted to find out. So they went around the world and got people to play an experimental public goods game. What they found was quite remarkable. Across the eight societies they studied, not one location supported the free riding model. Rather, the majority of individuals behaved as conditional cooperators. As can be seen in the figure below, although the proportion of cooperators varied substantially by country, in not one of them did free riders make up a majority share of the population. In other words, the canonical model of human behavior in economics, which is widely used in policy circles, does not accurately describe how most people act.
In experimental situations, most people behave as conditional cooperators rather than free riders.
In this spirit, the World Development Report 2015 emerged from the view that taking a more realistic account of human behavior can help make development policy more effective. Development economics and policy, we suggest, are due for a redesign.
If the assumptions of Development Policy 1.0 where that humans make decisions deliberatively, independently, and on the basis of consistent and self-interested preferences. Development policy 2.0 takes a different approach. It has three central pillars:
- People think automatically: when deciding, we usually draw on what comes to mind effortlessly.
- People think socially: social norms guide much of behavior, and we are conditional cooperators.
- People think with mental models: what we perceive and how we interpret it depends on the concepts and categories we have available to us.
A new perspective on development
These three principles have major implications for how we diagnose the constraints that poor people face in their day-to-day lives.
When seen under a psychological and social lens, living in a context of poverty is more than a deprivation in material resources. The stresses and strains of poverty impose “taxes” on cognitive resources. This means that the mental resources that fuel the deliberative thinking often required to make future oriented decisions are often in low supply. Policy makers should, therefore, try to move crucial decisions out of time periods when mental resources are especially scarce. They can, for example, shift school enrollment decisions to periods when poor farmers’ seasonal income is higher. They can target assistance to important decisions that require a lot of cognitive resources, such as applying to a higher education program. These ideas apply to any initiative in which program take up is a challenge.
Poverty early in life also affects psychological resources. The gaps between the cognitive and noncognitive development of children in rich and poor households are substantial and emerge well before school age. In addition to this, high stress and insufficient socioemotional and cognitive stimulation in the earliest years can impair cognitive development. This means that investing in polices that help disadvantaged families provide better psycho-social support for their young children can have high rates of return.
Expanding the policymakers’ toolkit
Just as importantly, adopting a psychological and social perspective enlarges policy makers’ toolkits. Consider some policy insights that come about from the fact that we think socially.[i]
Integrating opportunities for people to observe others’ behaviors—for example, by making behaviors more public—may be a useful way to of bolstering cooperation in the design of policies. Consider what happened when researchers created the illusion of “being watched” at an honor beverage bar in a university department in England. Researchers alternated pictures of watchful eyes with pictures of flowers above the price list for drinks each week and measured the contributions to the cash box. The results were striking. Every time the picture was changed to a pair of eyes, contributions for the week soared.
Consider also the utility of activating norms in an effort to reduce traffic deaths. Every year, about 1.25 million people die from traffic accidents—more than twice the number of victims from war and violence combined. Ninety percent of the road deaths occur in low- and middle-income countries. In Kenya, many of the people killed are passengers in minibuses, and people are aware of the danger. Researchers decided to try an inexpensive behavioral intervention to reduce accidents. Buses were randomly divided into two groups. In one group, nothing was done. In the other group, passengers were reminded of their right to a safe ride on public transportation. Stickers posted in the buses encouraged passengers to “heckle and chide” reckless drivers. The intervention was a remarkable success. Insurance claims involving injury or death fell by half, from 10 percent to 5 percent of claims. Results of a driver survey during the intervention suggested that passenger heckling played a role in improving safety. The cost per year for a life saved was about $5.80, making the program even more cost-effective than childhood vaccination, one of the most cost-effective health interventions available.
Consider also, the scope for Community-Led Total Sanitation (CLTS), a methodology for engaging communities, to help eliminate open defecation. It tries to trigger collective shame and disgust for the implications of open defecation. Local and national governments in rural India and Indonesia, with technical support from an international sanitation program, began implementing the first large-scale CLTS programs to be experimentally evaluated. Some communities were randomly selected to receive the treatment, while others were randomly selected to serve as controls and not to receive the treatment within the period of the evaluation. The CLTS programs were found to decrease open defecation by 7 percent and 11 percent from very high levels in Indonesia and India, respectively, compared to the control villages.
Striking lessons from writing the Report
Our opening goal was to explore how taking into account the specific psychological and social influences that guide decision making and behavior in a particular setting can improve policy. This journey brought us to unexpected places.
For example, recognizing that cultural and political outlooks affect how we interpret data led us to explore whether development professionals unconsciously interpret the data they use in a biased way. Identical sets of data were presented to World Bank staff, but in different frames. In one frame, staff were asked which of two skin creams was more effective in reducing a rash. In the other, they were asked whether or not minimum wage laws reduce poverty. Even though the data were identical, World Bank respondents were significantly less accurate when considering the data for minimum wage laws than for skin cream and views on whether minimum wage laws lower poverty tended to be related to cultural and political outlooks.
In addition to this, we came to realize that without the unified framework found in standard economics, we had to re-examine the challenge of “operationalizing” social and cognitive insights. Behavioral interventions work best when they are an iterative process of discovery, learning, and adaptation. This led us to conceive of an intervention cycle that looks quite different to the standard approach (see below). It places greater resources in defining and diagnosing the problem and the implementation period involves numerous phases of testing using multi-armed interventions – to learn what works in the context and what does not. This approach also makes policymakers think slow where they, too, have a tendency to think fast. What matters, we conclude in the Report, is not only which tool to use, but also how it is used.
The opinions expressed on this blog are those of the authors and do not necessarily reflect the official position of their institutions or of AFD.