Abstract
Between 2000 and 2015, access to life-saving antiretroviral therapy (ART) in sub-Saharan Africa increased from virtually 0% to approximately 40% coverage. Despite the promise of ART to improve health, a variety of supply- and demand-side factors suggest at-scale ART may not improve employment outcomes. To measure the causal effect of ART on employment outcomes, I use triple-difference regression analysis, exploiting spatial, temporal and demographic variation in intensity of ART exposure during scale-up in Zambia. My results suggest that local ART introduction increased employment of likely HIV+ adults by eight percentage points, or approximately 15% relative to the sample mean. Cash employment and employment for no pay—the first and second largest employment categories in Zambia, respectively—each comprised roughly one-half of the employment gains. Temporal and spatial heterogeneity analyses support a causal interpretation of the results. These findings appear to be the first at-scale quasi-experimental evidence suggesting that ART, the single largest item in many countries’ foreign health aid budgets, can improve employment outcomes among the targeted.