Economic field experiments in developing countries – Topics, methods and implications

One of the most important tools in development economics are field labs – economic laboratory experiments conducted in the field. In order to explain different aspects of poverty, growth and human development, assumptions about the preferences and strategic behavior of individual economic agents are essential. However, the continuously growing literature on economic experiments in developing countries shows that economic agents tend towards not acting according to classic economic beliefs regarding preferences. Cardenas and Carpenter (2008) deliver an account on the most important topics and lessons learned from field experiments conducted in the developing world. This blogpost gives an introduction to the most important topics, concepts and implications that can be drawn from experimental economics literature in developing countries and illustrate the groundwork for my essay, in which I would like to work out the current discussion on some of these aspects in a more detailed fashion.

In social dilemma experiments, individuals face decisions regarding contributions to public goods in groups. Typically, people can either contribute to the public good, which, if every group member does so, would lead to the optimal group result, or free-ride by taking advantage of the contributions of the other group members. Typical experiments testing the cooperation in social dilemma situations are the famous prisoner’s dilemma, the voluntary contribution mechanism and the common pool resource game. The common property of these experiments is that the strategy leading to the social optimum is never the individual best response. Mirroring the settings in these sort of experiments, developing communities often rely on locally defined norms and rules of conduct regarding public good provision in absence of governmental regulation, so that people face social dilemma situations frequently. Cardenas and Carpenter (2008) find inconclusive results with varying levels of cooperation in social dilemma experiments in developing countries. The main findings regarding such experiments are that only a minority of individuals free-ride on first impulse and cooperation increases both with the introduction of sanctioning mechanisms and discussions between rounds into the experiment structure. Other factors influencing cooperation are the homogeneity and age structure of the group. Generally, individual information is regarded as an important determinant of individual cooperativeness. Information directly influences the norm of social cooperation, which allows social dilemma games to be solved in a cooperative manner.

Communities that value norms of trust and reciprocity highly are assumed to be more productive as people are more likely to engage in profitable interpersonal relationships. A typical experiment to test for individual trust and reciprocity (or trustworthiness) is the trust game (Berg et al., 1995). In these experiments, two players are endowed with the same amount of money. The first mover then has the possibility to send a portion of his endowment to the other player. This amount is multiplied by some factor to render contributions among players socially efficient. Finally, the second mover (the other player) has the chance to send back a monetary amount of his total endowment. As the best response of the second mover would be not to send back any money, the first mover would not have an incentive to send any money to the other player in the first place. Trust game results throughout different developing countries are varying significantly, while a positive relationship between trust (amount sent by the first mover) and trustworthiness (the amount of money sent back) is detected. While there is consensus that the behavior of the second mover in such experiments can be interpreted as reciprocity, one key issue with the trust game is what the behavior of the first mover really measures. As the first mover cannot rationally expect the second mover to send back any money, the amount sent by the first mover might as well be interpreted as a purely altruistic transfer. Following Cox (2004), experimenters combine trust games with so-called dictator games, where money is transferred to another anonymous entity, without any payback mechanism. Thus, the difference between trust game contributions of the first mover and his contributions in a dictator game can be interpreted as pure trust. Comparing experimental data on trust with GDP data, Cardenas and Carpenter (2008) find a significant negative relationship between trust levels and inequality, unemployment and poverty, and a significant positive relationship between trust and GDP growth.

The third group of experiments tries to measure the norms of altruism and fairness. These norms are interesting to evaluate in a developing country context, as such societies often solve their conflicts locally within their immediate social networks. Altruism may be measured by the previously introduced dictator game, while fairness is often evaluated by ultimatum games in which the first mover offers a share of a total amount to the second mover, who then has to either accept or reject the offer. Accepted offers lead to the corresponding allocation, while rejections let both players end up with nothing. Thus, the second mover only has an incentive to reject zero offers as rejecting a small amount would make receive nothing, and the first mover should optimally offer very small amounts. Cardenas and Carpenter (2008) find mean allocations to the second mover to be substantially larger than zero, and low offers to be rejected frequently, indicating existing norms of fairness. Additionally, the authors find significant variation in the types and stability of fairness norms throughout developing countries.

The final group of experiments tries to measure individual risk and time preferences. An important proposition in development economics is that poor people might remain poor because of their impatience and high degree of risk aversion. Risk preferences are typically measured with experiments in which participants have to choose between two or more lotteries, while time preferences are measured in various forms, differing in the type of assumed discounting method and payment mechanisms. Cardenas and Carpenter (2008) find that people in poorer countries do not have different risk preferences from those in wealthier regions of the world, while their findings regarding time preference (impatience) are inconclusive.

This post gives an overview about the most important concepts of experimental economics in the context of human development. Please feel free to use the comment section.

References

Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and economic behavior, 10(1), 122-142.

Cardenas, J. C., & Carpenter, J. (2008). Behavioural development economics: Lessons from field labs in the developing world. The Journal of Development Studies, 44(3), 311-338.

Cox, J. C. (2004). How to identify trust and reciprocity. Games and economic behavior, 46(2), 260-281.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s