Between Physician Remuneration and Performance: Evidence from a Natural Experiment of Incentive Scheme Shift in China

Despite the growing interest in the compensation packages of hospital-employed physicians, little evidence exists regarding how payment from the hospital affects the medical services provided and, ultimately, the costs and quality of healthcare. Using a proprietary dataset obtained from a public tertiary hospital in China, this study investigates the impact of a shift in its physician’s compensation scheme from the department-based profit-sharing model to the Resource-based Relative Value Scale (RBRVS) on physician behavior and patient welfare.

To identify the effect, we exploit the rollout of the new compensation scheme within the hospital and adopt the difference-in-difference (DID) research design. Our treatment group consists of three pioneer departments that were randomly chosen. The rest departments serve as our control group and did not reform their compensation scheme until the end of a trial period. Two major changes of the new scheme are that (1) it imposes a tighter tie between individual physician’s performance and his/her compensation, and (2) that it introduces a large-pool profit sharing rather than within a single department.

Our results show that compared to those in the control group, the physicians in the treatment group perform 11% more surgeries following the incentive scheme shift, their patients occur 6.3% higher daily expenditures yet become 1.3% less likely to recover. These findings suggest that the new compensation scheme provides physicians greater incentives to get more work done but the workload is increased at the expenses of lower treatment quality and patient well-being. A further examination shows that the patients in the treatment group have shorter length of stay, i.e., a higher turnover rate, compared to the control group, and spend more on the examinations and treatment, which are subject to physicians’ discretion. Another interesting finding is that physicians’ personal characteristics, including sex, education, and their position/ranking, play important roles in their reactions to the compensation scheme reform.

This study makes several contributions. Although much of the theory in personnel economics relates to effects of monetary incentives on output, the theory was largely untested due to the lack of appropriate data (Lazear, 2018). We mitigate this gap in the literature by adopting a unique dataset that provides information on patient hospitalization and physician compensation. More specifically, it consists of approximately 300 physicians and 100,000 patients between 2012-2016. For each physician, we trace their monthly income (basic salary, benefits, and bonuses) and their patients’ medical treatment, expenditures, and health outcomes. Moreover, prior literature often exploits variations in insurance payment at the state level, which ignores interesting interactions between physicians within the hospital. We add to this line of research by demonstrating how the profit-sharing, in addition to the type and amount of payment, would influence physician performance. Finally, we bring attention to the heterogeneous response to financial incentives along different physician characteristics.