Three Essays on Environment and Development: A Behavioral Perspective

This dissertation comprises three essays with a behavioral perspective on environmental topics in developing countries. The first two essays are based on experimental methods and datasets that have been collected during field research in Zambia. Zambia is one of the most densely forested countries in Sub-Saharan Africa, but is also characterized by rapid land-use changes, in particular deforestation. The first paper evaluates the scope of Payments for Ecosystem Services (PES) among smallholder farmers in central Zambia. Agriculture is considered to be one of the major drivers of deforestation worldwide. In developing countries in particular this process is driven by small-scale agriculture. At the same time, many African governments aim to increase agricultural productivity. Empirical evidence suggests, however, that win-win relationships between agricultural intensification and forest conservation are the exception. Payments for Ecosystem Services (PES) could be linked to agriculture support programs to simultaneously achieve both goals. Due to potentially higher profits from intensified agriculture than from pure cash transfers, potential payment recipients may prefer inkind over conventional cash payments. Nevertheless, little scientific evidence exists regarding the preferences of potential PES recipients for such instruments. We report from a discrete choice experiment in Zambia that elicited preferences of smallholder farmers for PES contracts. Our results suggest that potential PES recipients in Zambia value in-kind agricultural inputs more highly than cash payments (even when the monetary value of the inputs is lower than the cash payment), highlighting that PES could potentially succeed in conserving forests and intensifying smallholder agriculture. Respondents who intended to clear forest within the next three years were found to require higher payments, but could be motivated to enroll in appropriately designed PES. The second paper investigates the effects of internal migration on the cooperation in host communities in rural Zambia, where in-migration is not confounded with increased ethnic or religious diversity. Potentially, in-migration could trigger discrimination, decrease overall levels of trust and hence negatively impact the propensity for collective action at the village level. We measure cooperative behavior through both self-stated survey information on public good contributions and incentivized decisions in a lab-in-the-field experiment. Different group compositions with respect to migrants and locals were introduced in a linear public good experiment. Our findings provide a nuanced perspective on how in-migration can affect cooperation in host communities that have been exposed to migration for relatively long periods. First, we find no evidence in the survey and experimental data that in-migration negatively affects cooperation across villages. Second, the particular effect of in-migration depends on the characteristics of the migrants relative to the villagers in the host communities. In our research area, migrants are generally wealthier and have higher incomes. We find evidence that in villages where these inequalities are more pronounced, migrants contribute more to public goods if exposed as the minority in the experiment. In these villages, migrants also contribute more to real public goods the more recently they have settled in the village and the higher their household income. The cooperation of better-off migrants is likely considered a signal of pro-sociality and the intention to integrate into the host community. Our findings indicate that the effects of migration on social dynamics in host communities are highly context specific and contingent on characteristics of the migrants in relation to the autochthonous population. More importantly, we provide evidence that communities that have been exposed to migration in the past can successfully accommodate migrants without negative consequences for the social fabric in these communities. The third paper is a methodological contribution to economic experiments that are increasingly applied in field settings to study decision-making with environmental consequences. In this paper, I investigate whether different degrees of subject-experimenter anonymity influence pro- and anti-social behavior in lab-in-the-field experiments. To do this I conducted the Dictator Game (DG) and Joy-of-Destruction Mini-Game (JoD) with 480 subjects in rural Namibia. In addition to a strict double-anonymous treatment two single-anonymous treatments are implemented. The results carry relevant implication for a methodologically sound implementation of lab-in-the-field experiments. Both in the DG and JoD, strict double-anonymous procedures are not necessarily required to minimize experimenter demand effects. However, if subjects are required to reveal their decision personally to experimenters, observed behavior is significantly more pro-social in the DG and significantly less anti-social in the JoD. Minimizing behavioral artifacts in lab-in-the-field experiments consequently requires sufficient privacy for subjects from experimenters during decision-making, however not necessarily a strict double-anonymous procedure.