The context of this evaluation is the rural Kosirai Division in western Kenya, which has an area of 76 square miles and a population of 35,383.2 The primary healthcare provider in Kosirai is the Mosoriot Rural Health Training Center, a government-run heath center that offers medical services to the region. The health center features an HIV clinic, founded in 2001 by the Academic Model for the Prevention and Treatment of HIV/AIDS (AMPATH), which provides free ARV therapy to HIV patients. By the end of the study period, the clinic was serving approximately 2,150 individuals from both inside and outside Kosirai.
Individuals become eligible for ARV treatment when their HIV infection progresses to AIDS. In developing countries, death usually occurs within one year of progression to AIDS without ARV treatment.3 ARVs are able to prolong the lives of HIV-infected individuals and improve functional capacity after several months of consistent treatment.
The primary research goals were (1) to estimate the effect of ARV treatment on patients’ labor supply, measured through labor force participation and number of hours worked, and (2) to estimate the effect of ARV therapy on the labor supply of other household members in a patient’s household. The authors implemented two rounds of a household survey in the Kosirai Division between March 2004 and March 2005, measuring demographic characteristics, health status, and information on income and employment.
The survey sample included two groups of households: 266 HIV-positive households with at least one AMPATH patient, and 503 households chosen randomly from the population of non-AMPATH patients. The authors used multiple time periods to estimate the impact of ARV treatment on labor supply with fixed effects. Fixed effects can reduce potential biases from time-invariant factors in the labor supply, such as family land and human capital endowments. Labor supply data from the random sample of households were used to control for time varying trends in the survey area, such as seasonality in agriculture and health.
Results and Policy Implications
ARV treatment had a significant impact not only on health, but also on labor supply. After three to six months on ARV treatment, patients experienced an average 12.3 percentage point increase in labor force participation (a 20 percent increase from the baseline level of 61.6 percent). Furthermore, the treatment allowed for an average 6.9-hour increase in weekly hours worked (a 35 percent increase from the baseline level of 19.7 hours worked). Even after accounting for patients who did not complete the treatment period because of death or other reasons, labor market effects remained significant.
The authors also found that ARV treatment significantly reduced the labor supply of younger boys in patient households, suggesting that younger boys reduced their school attendance and entered the labor market after an adult fell ill. The results for older boys and girls were not significant. Older boys are more likely to already be engaged in the labor market, while girls are more likely to engage in housework, and thus are not counted in the labor force. Finally, the labor supply for other adults in patient households did not significantly change after ARV treatment, suggesting that increased labor supply of HIV-positive individuals after ARV treatment is a complement to — rather than a substitute for — other household adult income, and may therefore be welfare-enhancing.
As the first evidence of the effect of ARV treatment on labor supply, this paper highlights the benefits of ARV treatment on the ability of symptomatic HIV-positive individuals to work in the labor market. It also illustrates the potential for positive spillovers, as ARV treatment can reduce the need for younger children in the same household to leave school in order to work.