Other Authors: Erica Field, Leigh Linden, Shing-Yi Wang
Geographical Region: Asia
How do property rights institutions affect trust?
Preparer: Alicia Cooperman
A large body of literature in political science and economics links institutions, particularly property rights institutions, to economic growth, trade, and political participation. Underlying these theories is the role of generalized trust within society, where the expectations of trust developed in close social networks spill over to anonymous interactions in general society. Individuals develop their own sense of trust, which is influenced by the number of interactions they have and whether those interactions tend to be positive or negative.
Prior observational (non-experimental) studies on institutions and trust suffer from selection bias, which means that results from these studies cannot be generalized to the wider population because their sample is not representative of said population. It is possible that we only see property rights reforms in societies with high trust to begin with, and/or other unobserved factors in society predict both institutions and trust. Many parts of the world lack strong property rights institutions, but we know little about how institutions causally affect behavior.
The researchers theorize that when individuals gain secure property rights, they will be less likely to have negative interactions related to land, which will increase their generalized trust in society. This effect will be more pronounced for individuals that have a higher number of interpersonal interactions.
The researchers study a program that randomly assigned land property rights to small semi-nomadic herder groups in Mongolia: the Peri-Urban Rangeland Property Rights Project (PURP), financed by the Millennium Challenge Corporation (MCC). The Mongolian government provided long-term exclusive leases of rangeland plots, basic infrastructure, and training in herd and rangeland management to the treated herder groups during October/November 2011. The herder groups have four households on average.
The randomization occurred as follows: first, herder groups were allowed to apply to the program; second, among the 329 herder groups that applied and were eligible, 165 were randomly assigned to the property rights treatment through a public lottery. The research team collected baseline data in January-April 2012 from households in the treated groups, the control groups, and a randomly selected sub-sample of neighboring herder groups. In a follow-up survey, the team had a response rate of 93.1%. Researchers also conducted lab-in-the-field games to measure prosocial behavior such as altruism and fairness. A first follow up survey was conducted in 2014 in which mostly household economic activity and rangeland management was measured. A second follow up data collection is planned for winter 2017.
Survey on Trust: The property rights intervention had a significant effect on trust. Overall, the treated groups had 36% higher odds of indicating trust when asked the dichotomous question: “In general, in Mongolia, would you say that one can trust other people [=1] or that people cannot be trusted [=0]?” These results include controls for gender, age, and fixed effects for herder groups. Surveys of neighboring groups showed that control respondents are very similar to their neighbors, so there is no evidence that the effect is actually due to control members being resentful for losing the lottery. In the small group of respondents that engaged in high levels of interpersonal interaction, as measured by participation in joint business ventures, the treatment had an even stronger effect. However, there are also very few joint business participants (2.3% of respondents) so the effect is not statistically significant.
Lab Games: The lab games included the dictator and ultimatum games, which are common in behavioral economics. In the dictator game, individuals choose how much money to keep and how much to give to the receiver, an anonymous person. The ultimatum game is similar, but the receiver can decide if they are receiving an acceptable amount; if the receiver finds it acceptable, each player keeps their share, if the receiver finds it unacceptable, they both get 0. The participant also imagines being the receiver and chooses how much money she would have to receive to find the amount acceptable and keep it. The treatment group is more likely to engage in prosocial acts of altruism (giving non-zero amounts in the dictator game) and/or fairness (giving non-zero amounts in the ultimatum game) than the control group. However, the treatment effects are very small and many are not statistically different from 0. Still, these effects emerged after only a few months of treatment, as opposed to standard theories that prosocial behaviors only develop in the long run.
The study highlights the ability of property rights institutions to increase generalized trust, even within months of treatment. This is likely the first randomized allocation of land property rights to small groups, such that the increase in trust and prosocial behaviors can be causally attributed to the institutional treatment.
The treatment is also likely to improve the sustainability of rangeland in peri-urban Mongolia, and it is an example of a successful policy intervention to address changes in land-use and improve livelihoods for semi-nomadic groups.
Future data will enable researchers to study medium- and long-term effects of property rights on trust and other social, economic, and environmental outcomes. Since the increase in trust is due to random provision of property rights, the researchers can also study downstream effects of increased trust on other behaviors.