A frequent argument for managed care is that oversight of plan members’ utilization can reduce unnecessary care while improving coordination and quality of that care. Our study estimates the effects of Medicare Advantage (MA) on enrollees’ utilization patterns relative to enrollees in traditional Medicare (TM). Simple comparisons of utilization between MA and TM enrollees, even those that control for observable demographic and health characteristics, are susceptible to the confounding influence of unobservable individual factors that drive enrollment-type selection. Our two-way fixed effects approach attempted to reduce that confounding by using enrollees’ own previous experience to control for preferences and other time-invariant characteristics. We analyzed data on persons over age 65 enrolled in Medicare during both 2018 and 2019 and compared changes in several key utilization outcomes between 2018 and 2019 for those who switched plan type relative to those who did not. We found that 2019 outcomes for 2018 TM enrollees appear to confirm prior findings that MA significantly reduces costly and low-value care. However, among 2018 MA enrollees, those who switch to TM do not show marked reductions in costly care, but rather show several improvements in cost and quality after switching. These apparent contradictions call into question the hypothesized improvements of managed care over fee-for-service Medicare.