I’m speaking at the Massachusetts Avenue Development Seminar at CGD on Tuesday.

The title: “Can employment programs reduce poverty and social instability? Experimental evidence from a Ugandan aid program”

The details are here.

Bloated abstract below the fold. I will have a draft paper posted in the next few weeks,. for the very curious.

Youth unemployment is widely considered a threat to development and to security. To reduce poverty and social instability, aid programs commonly provide youth with inputs to boost self-employment. Such programs are rooted in two assumptions. The first is that the poor possess high returns to investments but are constrained from reaching those returns unaided. The second is that increased income reduces youth alienation and aggression.

We look for evidence of these claims using an experimental study of Uganda’s largest employment program. In 2008, the government provided thousands of young people relatively unconditional cash transfers to pay for vocational training, tools, and start-up costs. Cash transfers were allotted randomly to small groups of youth, and we track roughly 2000 successful and unsuccessful applicants over two years.

First, in spite of the absence of central monitoring and accountability after the transfer, most youth invest the majority of the transfer in vocational skills and business assets. Group-based disbursement may have encouraged investment, but we see evidence that group quality and longevity impacted investment and performance.

Second, we find that the economic impact of the transfers is large: incomes increase by nearly 50% on average. Consistent with an inter-temporal model of investment under credit constraints, the program was most impactful when grantees were patient and faced high start-up costs. Surprisingly, however, formal schooling, health and cognitive abilities appear to have little effect on success—findings which suggest that entrepreneurial ability may be more informal or “non-cognitive” in nature.

Finally, these economic gains results in modest gains in social stability. Measures of social cohesion and community participation improve by 5 to 10%. We see a 50% fall in male aggression and disputes with authorities among males, but from a very low base.

A second stage of data collection in 2012 will collect longitudinal economic impacts, additional data on political violence and behavior, and explore alternative theoretical mechanisms.