State governments contract with health maintenance organizations (HMOs) to coordinate medical care for nearly 20 million Medicaid recipients. Identifying the causal effect of HMO enrollment on government spending and health care quality is difficult if, as is often the case, recipients have the option to enroll in a plan. To estimate the average effect of HMO enrollment, this paper exploits county-level mandates introduced during the last several years in the state of California that required most Medicaid recipients to enroll in a managed care plan. The empirical results demonstrate that the resulting switch from fee-forservice to managed care was associated with a substantial increase in government spending but no corresponding improvement in infant health outcomes. The findings cast doubt on the hypothesis that HMO contracting has reduced the strain on government budgets.