Working Paper: NBER ID: w29613
Authors: Joseph J. Doyle Jr.; Becky Staiger
Abstract: Treatment intensity varies remarkably across physicians, yet key drivers of this variation are not well understood. Meanwhile, physicians are increasingly working in groups that may influence how they practice. This paper tests whether physicians' group affiliation matters for practice styles and patient health. Using Medicare inpatient claims data, we compare these outcomes before and after physicians switch between groups of varying treatment intensity while remaining in the same hospital to control for practice setting. Event studies show that internists who join groups with higher Medicare payments per patient immediately increase their own treatment intensity, with an elasticity of approximately 0.27; the opposite is found for internists who switch to groups that are less intensive. This change in Medicare spending largely stems from higher-priced services, although an increase in the quantity of care is also affected. We do not detect a change in health outcomes, suggesting that treatment intensity induced by group affiliation may not be productive.
Keywords: No keywords provided
JEL Codes: I10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| physician group affiliation (I11) | treatment intensity (C32) |
| treatment intensity (C32) | health outcomes (I14) |
| group-specific factors (C92) | treatment intensity (C32) |
| physician-side components (I11) | treatment intensity (C32) |
| physician group affiliation (I11) | healthcare spending (H51) |