Summary queued
This paper remains searchable while structured annotation is pending.
ECON CURRENT excludes papers without a reviewed, schema-valid annotation from editorial selection. Public metadata remains available below.
Public abstract
Physicians in the U.S. military health care system are quasi-randomly assigned to combat deployments, leading to gaps of varying durations in their clinical practice. We leverage this unique setting to provide novel causal evidence on skill depreciation (forgetting) and learning-by-doing among physicians. We find robust evidence that readmission rates increase by 48% in the first two months after a surgeon returns from a deployment. However, surgeons return to baseline performance within 6 months of resuming practice. In contrast, obstetricians maintain elevated cesarean section rates for up to 9 months, implying a deeper change in clinical judgment. Finally, we interpret these patterns through the lens of a learning model and estimate learning-by-doing rates for surgeons and obstetricians. A doubling of recent procedure volume reduces readmission and C-section rates by 17% and 11% for surgeons and obstetricians, respectively.