We employ a unique data set on members of an elite service club in Germany to investigate how social connections in elite networks affect the allocation of resources. Specifically, we investigate credit allocation decisions of banks to firms inside the network. Using a quasi-experimental research design, we document misallocation of bank credit inside the network, with bankers with weakly aligned incentives engaging most actively in crony lending. Our findings, thus, resonate with existing theories of elite networks as rent extractive coalitions that stifle economic prosperity.

We study the wealth accumulation of Indian state politicians using public disclosures required of all candidates. The annual asset growth of winners is 3–5 percent higher than that of runners-up, a difference that holds also in a set of close elections. The relative asset growth of winners is greater in more corrupt states and for those holding ministerial positions. These results are consistent with a rent-seeking explanation for the relatively high rate of growth in winners’ assets.

We provide microeconomic evidence on ethnic frictions and market efficiency, using dyadic data on managers and borrowers from a large Indian bank. We conjecture that, if exposure to religion-based communal violence intensifies intergroup animosity, riot exposure will lead to lending decisions that are more sensitive to a borrower’s religion. We find that riot-exposed Hindu branch managers lend relatively less to Muslim borrowers and that these loans are less likely to default, consistent with riot exposure exacerbating taste-based discrimination. This bias is persistent across a bank officer’s tenure, suggesting that the economic costs of ethnic conflict are long-lasting, potentially spanning across generations.