This is an inter-disciplinary project, joint with researchers in Earth Sciences and Sociology, which allows us to link high-resolution, high-frequency, and fine-scale environmental data (remote sensing and in situ) with socio-economic data on migration and adaptation. Early findings suggest that environmental factors have a significant effect on migration, though the magnitude is relatively modest, in part because many in the vulnerable population lack the necessary resources to move. More recent work, published in Nature Climate Change, reveals that, though flooding per se has only modest effects on migration, saltwater flooding has large effects via saline contamination of soils. We are working towards the integrated development of a natural and social science framework, employing robust predictive modeling towards adaptation to and mitigation of sea-level rise and other hazards in coastal Bangladesh. An innovation of this project is the collection of high-frequency data on migration via mobile phones. This will allow us to identify even the most vulnerable migrants, particularly those who may migrate over short distances for short periods of time and perhaps lack the resources to move permanently to a more environmentally and climatically stable location. Support from the Belmont Forum and the Institute for Population Research and the Sustaiable and Resilient Economy Program at The Ohio State University is gratefully acknowledged.

This research program examines the motives and potential for cooperative behavior, focusing on the roles of information,incentives and social norms. I find evidence of significant information asymmetries between spouses, resulting in varying degrees of inefficiency in the intra-household allocation of resources. In Ghana, where spouses independently manage separate agricultural plots, husbands and wives exhibit large discrepancies with respect to their spouses' earnings, expenditure, and Engel curves. Moreover, these discrepancies significantly reduce household profit, with the average effect being roughly equivalent to 15% of the variation between households. In China, split-household migration exacerbates information problems but effects only allocations that are difficult to monitor, such as time and nutrient allocation. In contrast, observable goods such as child BMI and school enrollment remain unchanged, controlling for income. Therefore, subsidies for agricultural inputs – where the value is transparent – may have very different effects than subsidies for agricultural credit – where the value could be more easily obscured – even if the two programs have the same overall effect on household income. Finally, as suggested by the Samaritan's Dilemma, altruism is found to reduce incentives for cooperation, all else equal. This suggests that, in scenarios requiring voluntary provision of public goods, it may be optimal to maintain greater social distance between participants.





Data on migration largely fail to distinguish permanent (establish a new residence and household) and temporary (intend to rejoin the origin household at a later date) migrants, leading to ambiguity in how both migrants (e.g., by duration or household membership) and income (e.g., as a remittance or as earnings of a current household member) are characterized and obscuring the interaction between migrants and origin households. The primary objectives of this research program are to document the characteristics of these two disparate types of migrants and analyze the determinants of migration for each, as well as to estimate the causal effect of migration, by type on well-being and economic activity. Research in progress utilizes data from Pakistan and Mexico, as well as an ongoing survey in Bangladesh (supported by the International Growth Centre, and the Institute for Population Research and the Sustaiable and Resilient Economy Program at The Ohio State University) in which we use mobile phones to collect high-frequency migration data.

This line of research examines processes of household bargaining and the determinants of intra-household resource allocation. The impact of economic incentives will also be limited by parental preferences; understanding the import of preferences highlights where economic policy should end and where social policy may begin. Depending on the context, the role of parental preferences, particularly related to gender, may be quantitatively quite large. Moreover, this suggests that the impact of skill-based immigration restrictions on future generations will be dependent not only on the skill-composition of current immigrations but also on patterns of selection with respect to preferences for human capital investment.