Role of malaria control in improving child health in mainland Tanzania: Evidence from a rapid policy scale-up Romeo J. Gansey University of Pennsylvania, PA, USA article info Article history: Accepted 24 September 2019 Keywords: Impact evaluation Malaria control Child mortality Nutritional health Tanzania abstract Malaria control is a major policy concern in Sub-Saharan Africa, where the disease imposes a high burden on child health. Between 2004 and 2010, in mainland Tanzania, malaria control interventions have expe- rienced a rapid scale-up evidenced by a sharp increase of the budget devoted to malaria control, rising from less than $10 million to nearly $140 million. However, the extent to which these interventions have improved child health remains an open policy question. I exploit the timing of the fast scale-up of the malaria control program, along with the variation in malaria endemicity across the country due to cli- mate, to evaluate the impact of malaria control interventions on child health with a difference-in- difference approach. My estimates suggest that, on average, malaria control interventions have helped avert approximately 17.9 deaths for every 1000 live births. In relative terms, they have contributed to 57.7 percent of the reduction in under-five mortality between 2004 and 2010. These interventions have also improved other measures of child health, such as anemia and stunting, reducing the odds of these conditions by 52 and 35 percent, respectively. Calculations suggest that it costs at most 4100 US dollars to save a child through the program. Finally, the malaria control program passes the test of a positive net present value with a value of a statistical life as low as 4600 US dollars for discount rates ranging from 10 to 20 percent. Ó 2019 Elsevier Ltd. All rights reserved. 1. Introduction The extent to which interventions geared toward malaria con- trol have improved child health is an important concern among policy makers and researchers, as this disease is thought to impose a high burden on children’s health (Rowe et al., 2007; Smithson et al., 2015; Steketee & Campbell, 2010; Tanzania Malaria Impact Evaluation Research Group, 2012; World Health Organization, 2014). Many studies have investigated the impact of national malaria control in African countries, where Plasmodium falci- parum, the dominant parasite strain, is responsible for high mor- bidity and mortality, especially among children, who typically do not have an immune system well developed against malaria (Rowe et al., 2007; Steketee & Campbell, 2010). Most of these stud- ies, relying on a before-and-after design combined with a ‘‘plausi- bility argument” (Habicht, Victora, and Vaughan 1999; Rowe et al., 2007), have suggested possible large health benefits attributable to malaria control. For example, many have documented a downward trend in indicators such as under-five mortality and the prevalence of anemia associated with scale up malaria control interventions (Rowe et al., 2007; Smithson et al., 2015; Snow, Trape, & Marsh, 2001; Steketee & Campbell, 2010). Yet, there are four shortcomings in past studies of the impact of national malaria control on child health. First, they usually fail to provide estimates of the effect size of health improvements that can unambiguously be related to malaria control interventions (Rowe et al., 2007). In other words, they often do not quantify the specific contribution of malaria control, as opposed to a pack- age of changes correlated with malaria control, to improvements in child health. Second, some studies have used data from 1995 to 1999 as the baseline, likely overestimating the impact of the interventions because of climatic changes (Thomson et al., 2017). Third, prior studies do not often control for confounders such as socioeconomic characteristics of children under study. Finally, they focus on mortality, but do not investigate the contribution of malaria control to improving child nutritional health. This paper addresses these shortcomings by employing a difference-in-difference strategy and by using the period from 1999 to 2004 as the baseline along with indicators of child nutri- tional status. It exploits the timing of the dramatic scale-up of interventions related to malaria control combined with the fact that malaria endemicity depends on climate to assess the impact of malaria control on child health in mainland Tanzania. https://doi.org/10.1016/j.worlddev.2019.104694 0305-750X/Ó 2019 Elsevier Ltd. All rights reserved. E-mail address: rgansey@sas.upenn.edu World Development 125 (2020) 104694 Contents lists available at ScienceDirect World Development journal homepage: www.elsevier.com/locate/worlddev