How does the social structure of a neighborhood affect the linkage strategies chosen by politicians? Our argument links the social centrality of individuals and the social density of neighborhoods to the incidence of individual-level clientelism and the contingent provision of neighborhood benefits, respectively. We test the resulting hypotheses using a combination of original network, experimental and non-experimental survey data collected across 9,184 households in 217 slums across three cities in India. We find that socially central individuals are more likely to be targeted with private clientelism and that more socially dense neighborhoods are more likely to coordinate votes, attract benefits, and achieve the legal prerequisites for public services.
