Eric Werker (guest author)
I enjoyed the chance to present a work in progress that attempts to measure the impact of AIDS on the economies and populations in Africa at the Applied Statistics Workshop on Wednesday, November 9. Given the possibility for some omitted variable to influence both the national AIDS rate and economic performance or some other outcome variable, I chose to pursue an instrumental variable strategy using variations in the male circumcision rate (which the bulk of the medical literature on this subject believes to have a causal impact on the spread of HIV/AIDS). Comments from the audience were useful and illuminating, and the debate was most interesting around potential violations of the exclusion restriction as well as the use of 2SLS in a small sample setting.
Posted by James Greiner at November 23, 2005 5:47 AM