The World Health Organization (WHO) estimates that an additional four million health workers will be needed to address the global shortfall of health workers. This shortage is especially hard felt in Africa, which has approximately 11% of the world's population and 24% of the world's disease burden, but only 1% of its doctors, nurses and midwives. Task shifting, which involves the redistribution of tasks from highly skilled health workers to less highly skilled health workers, is emphasized in policy circles as a solution to the problem of shortage of skilled medical professionals. This policy has gained considerable momentum and an increasing number of African countries have expanded production of mid-level health professionals. The appeal of mid- level providers is easy to understand: if a less skilled worker can do the same work (with comparable results) as a more highly skilled provider at a lower cost, then it makes sense to expand the production of such workers. There is however limited data and evidence for whether the substitution of more highly skilled health providers with less highly skilled workers influences quality, safety, and patient outcomes. In thi R01, we randomly assign providers with different levels of training to primary health care clinics and study the effect on a range of maternal and child health outcomes. This will to our knowledge be the only existing randomized trial in which providers with varying levels of training/skill are assigned to a health facility to provide primary health care services. We combine randomized provider assignment with a randomized offer to participate in a conditional cash transfer. This unique design allows us to address the issue of non-random patient sorting and to provide some evidence about the extent to which supply-side barriers (such as provider skill) may help explain why increased use of services (as a result of demand-side programs) has not translated into the anticipated improvements in health outcomes.
