Abstract

The central question that we ask in this paper is: do people cheat or behave dishonestly when they are unsupervised? This question is motivated by our work on environmental management programs in developing countries where local conservation effort can be either too expensive or difficult to monitor by outsiders. We combine responses from a household survey with results from a set of field experiments in rural Vietnam to examine how people behave when they are unsupervised, and how well are they able to predict the behavior of others in their community. Both the survey and the field experiments are structured along the lines of an environmental project in which local people receive an incentive for providing their time and labor to undertake project activities. Our total sample size is 400 with experimental treatments varying by group size, whether people receive cash or in-kind incentives, and whether they are monitored. In contrast to existing studies that predict compulsory cheating, we find very little cheating behavior among our subjects. Moreover, people can accurately predict the behavior of others except in a scenario when people collect cash incentives for others. In that case, people become extra cautious than what they were predicted to be. We explain these results by invocating the theory of self-concept maintenance and conclude with a discussion on potential applications of these results.