Exiting primary care providers.

This article studies the impact of primary care providers (PCPs) exit from the local health care system on patients' health care utilization. I compare patients with each other whose physicians have left the local health care system at different points in time due to retirement, relocation, or other reasons. Estimation results indicate that the imminent exit leads soon-leaving physicians to changing their treatment behavior, which has a significant impact on patients' health care spending. In addition, successors and new PCPs provide significantly more preventive services in the post-exit-period and refer patients more often to specialists for further examinations than the physicians who exit later. The increased inpatient expenditures in the post-exit period are caused by the new PCPs (through referrals). Self-initiated substitution behavior of patients (e.g., less PCP care, more specialist care) after the exit is observed but is low in magnitude. Although an overall increase in health service utilization is observed, mortality in the post-exit periods is significantly increased among affected patients. A possible explanation is the low frequency follow-up care of patients who were referred to hospitals by their former PCP in the notification-period.

The company you keep: health behavior among work peers.

There is widespread agreement that behavior crucially influences one's health. However, little is known about what actually determines health-related behavior. We explore the impact of the place where many people spend most of their time, at work, and analyze whether an individual's decision to participate in health screening is related to the observed behavior of peers at work. We use linked employer-employee data and exploit the transitions of workers to new jobs. We find that the health behavior of co-workers highly correlated. A comparison of individuals moving into new firms shows that participation in general health checks, mammography screening, and prostate-specific antigen tests increases with the share of work peers attending these screenings. To differentiate between peer effects and common influences at the workplace, we further separate the peer groups within firms and show that workers with similar characteristics tend to have a stronger effect on individual screening participation.