Monday, Nov 3
Melvin Stephens, Estimating Program Benefits
Nicola Branson, David Lam, Murray Leibbrandt, Vimal Ranchhod, Leticia Marteleto (PSC, U of M)
04/27/2009, at noon in room 6050 ISR-Thompson.
PSC researchers have been collaborating with researchers from the University of Cape Town to analyze the causes and consequences of teen sexual activity and pregnancy in South Africa. This research takes advantage of longitudinal data in the Cape Area Panel Study, a collaborative UM-UCT project that began in 2002. The brownbag will include several South African researchers who will be in town for the PAA meetings. Results will be presented from several current papers coming out of the project. The first looks at the impact on sexual debut of exposure to older classmates as a result of the high variance in age-for-grade in black schools. The results suggest that students who are ahead in school are more likely to become sexually active because of these peer effects. The second paper looks at alternative methodologies for identifying the causal impact of teen pregnancy on schooling outcomes. Estimates based on propensity score matching show much smaller effects of teen pregnancy than OLS estimates, but still show statistically significant negative effects. The third paper looks at the outcomes for children born to teen mothers. The apparent negative effect of being born to a teen mother is substantially reduced, though not entirely eliminated, by controlling for socio-economic status and by using sibling fixed effects.