We argue that a common practice of evaluating portfolio managers relative to a benchmark has real effects. Benchmarking generates additional, inelastic demand for assets inside the benchmark. This leads to a "benchmark inclusion subsidy:" a firm inside the benchmark values an investment project...