Semiparametric Modeling of Household Asset Welfare Dynamics.
This research develops a new semiparametric panel data estimator based on penalized splines to model potentially highly non-linear patterns of household asset accumulation and applies it to newly expanded household panel data from rural Andhra Pradesh and Maharashtra.
Quantitative analysis of household welfare dynamics and idiosyncratic shocks in rural Ghana (with ISSER University of Ghana).
This research will use newly expanded panel data from rural Ghana to assess how household-specific risk affects long-term asset accumulation patterns differentiating between ex ante and ex post risk exposure and between risk management mechanisms (self-insurance, formal insurance informal social insurance).
Idiosyncratic shocks and household welfare dynamics. A three country comparison (with IFPRI, Addis Ababa University, ISSER University of Ghana, DATA Bangladesh).
This USAID funded research aims to clarify what existing mechanisms help mitigate adverse idiosyncratic shocks, what gaps in coverage exist, and how different interventions affect insurance against idiosyncratic ex-ante and ex-post risk and its dynamic effects on rural households' asset holdings, productivity and well-being in Bangladesh, Ethiopia and Ghana.
The impact of government decentralization policies on child nutrition in rural Kenya (with International Livestock Research Institute, Nairobi, Kenya).
This research uses novel difference-in-difference stochastic dominance analysis and a new, large dataset on kids' Mid Upper Arm Circumference in arid and semi-arid rural Kenya to evaluate the effect on child nutrition of the Kenyan government's Arid Land Resource Management Program.
Measuring poverty over time - Accounting for the intertemporal distribution of poverty across households.
This research develops two new classes of poverty measures that capture the degree to which poverty is shared across households over time and, hence, the extent of chronic poverty. It compares these new indicators with two existing measures from the unemployment literature and applies them to panel data from rural Pakistan.