Leveraging the random assignment of over 50,000 Medicaid enrollees in New York, I present causal evidence that narrow networks are a blunt instrument for reducing health care spending. While narrower networks constrain spending, they do so by generating hassle costs that reduce quantity, with...