In political economy, government provision of services has been treated as both a core driver of political engagement and as a key factor by which citizens judge government performance. However, non-state actors frequently step-in to supplement or expand government provision, and little is known about how these efforts influence the beliefs and behavior of citizens. Despite a long history, the impact of non-state service delivery in Africa has new relevance due to the rise of development NGOs. I argue that service delivery by NGOs has an important impact on domestic politics, but not in ways anticipated by extant theories. I argue that service delivery by NGOs has two unanticipated consequences. First, government actors receive credit for NGO projects because citizens believe that incumbents play a role in their allocation and citizens update their beliefs about the quality of incumbents due to changes in access to NGO services. Second, direct experience with an NGO program increases the preference for their broader role in providing services to the population. To provide support for these arguments, I use a variety of data and methods, including original temporal and spatial data on service delivery projects and infrastructure, protest event data, official election returns from three successive elections, and an original survey of villages that were randomly assigned to receive a highly effective NGO health intervention. I find that access to NGO services increases citizen preferences for NGO service delivery over government delivery. However, rather than seeing NGOs as a substitute for government, citizens see government actors as controlling access to superior NGO services and reward government actors for increases in access. These findings suggest that understanding the political consequences of non-state service delivery is critical for assessing the impact of specific NGO interventions and the rise of NGOs generally.