Working papers:
Intergenerational Effects of Early Childhood Shocks on Human Capital: Evidence from Ethiopia .

This paper studies the intergenerational effects of maternal early childhood shocks on the human capital outcomes of children. I exploit the 1983-1985 Ethiopian famine as an exogenous source of variation to study the effects of exposure to severe shocks in utero and/or during the first three years after birth on the cognitive, non-cognitive and health capabilities of children of mothers who were exposed to the shocks in early childhood. Using data that track children from early childhood through adolescence, I estimate the effects of maternal early childhood shock over their children's life cycle. I find that the famine has had a lasting intergenerational effect. Mothers' early childhood famine exposure reduces their children's height-for-age z-score, schooling, locus of control and self-esteem. These effects are persistent from age one through early adolescence. The main inter-generational transmission channel of the shock is children's maternal human capital endowment. Mothers who suffered the famine in early childhood are shorter and have less schooling. I also find a critical maternal shock duration threshold of three months. These findings point to ineffectiveness of remediation once the damage is done to mothers as young girls. The policy implication is that girls under the age of three with high risk of crossing the critical famine duration threshold should be targeted for health and nutritional interventions.

Insuring Wellbeing? Buyers Remorse and Peace of Mind Effects from Insurance . with Chris Barrett and Erin Lentz, 

In this paper we estimate the causal effects of index insurance coverage on the subjective well-being
(SWB) of a poor population in rural southern Ethiopia. Insurance coverage may be welfare enhancing ex ante by reducing exposure to risk. Yet, if the insurance policy lapses without paying out, the purchaser may experience buyer's remorse ex post of the resolution of uncertainty about stochastic losses. The prospective (ex ante) and retrospective (ex post) well-being effects of insurance may therefore differ, especially in the absence of indemnity payments. We separately identify (1) the ex ante SWB effects of current insurance coverage that arise from reducing ex ante risk exposure to potential shocks, and (2) the ex post buyer's remorse effects of lapsed insurance policies that did not pay out. By exploiting the randomization of incentives to purchase a newly introduced index-based livestock insurance product and three rounds of household panel data, we establish that current coverage generates statistically significant gains in buyer's SWB. The ex ante gains more than offset the statistically significant buyer's remorse effect of having lost money on insurance that did not pay out. These findings suggest that insurance can have significant impact on SWB and illustrate that failure to control for potential buyer's remorse effects can bias downwards estimates of the welfare gains from insurance coverage.​ 

Diversification and Productivity in African Agriculture: Evidence from Uganda

This paper explores the determinants of crop diversification in Ugandan agriculture. I use three rounds of the Uganda National Panel Survey (UNPS) data, which collects detailed information on land holding and characteristics, crop production, agricultural inputs and farm management practices to examine the prime drivers of observed crop diversification practices. The findings suggest that crop diversification is determined by a combination of yield and variance considerations, and that these considerations vary by crop type. Among the main crops in Uganda, the beans and sweet potatoes mix appears to be primarily driven by average yield considerations while variance (risk) appears to prominently factor in maize planting decisions on diversified farms. Maize and beans are best suited for joint production, whereas sorghum and matoke yield better results when planted as mono crops. The maize-beans combination is the best crop mix. I also find that crop yields are lower and yield variance higher on larger plots, suggesting the inverse productivity-size phenomenon is present in Ugandan agriculture. 

           Aspirations and Wealth Dynamics in Rural Ethiopia

This paper investigates the effects of aspirations on wealth dynamics in poor communities. It uses a unique dataset from rural Ethiopia, which elicits responses on four dimensions of aspirations: income, assets, social status, and children's education. Data on specific aspirations is used to construct an aspiration index. Using a reference dependent utility and a simple model of aspiration adaptation, I study the dynamic equilibrium effects of aspirations on wealth dynamics. The total effect of aspirations is then disentangled into risk preference, time preference and motivational effects. I find that aspirations are conductive to upward mobility, especially in relatively poor households. The findings point to motivation as the primary mechanism through which the effects of aspirations are channeled. In terms of policy, the implication is that interventions aimed at improving the aspirations of the poor can play complementary role to programs aimed at relieving external constraints and produce greater impact.

​Papers in progress:

  • Healthcare and educational achievement in Ethiopia. with Kevin Croke, Andualem Mengistu and Stephen O'Connell
  • The gender gap in the disciplining of politicians: Evidence from Brazil. with Stephen O'Connell
  • Education and weather shocks in East Africa. with Anthony Mveyange.
  • Formal and informal insurance: experimental evidence from Ethiopia. with Ruth Vargas Hill, Guush Berhane, Stefan Dercon and Alemayehu Seyoum Taffesse