Working Paper: NBER ID: w2030
Authors: Richard O. Frank; Judith R. Lave; Carl A. Taube; Agnes Rupp; Howard H. Goldman
Abstract: Medicare's Prospective Payment System (PPS) for hospitals was phased-in during the 1884 Federal Fiscal Year. While many providers of psychiatric inpatient care were exempted from PPS patients treated in general hospital beds outside of psychiatric units (scatterbeds) were not. This allows for an initial assessment of the impact of PPS on psychiatric patients. We use a single equation model of hospital length of stay to estimate the impact of PPS. We allow for the possibility of both anticipating behavior and slow adjustment to the new payment scheme. The results indicate a substantial response to PPS over the first year of implementation. The estimated response includes sizable anticipatory and slow adjustment components. The findings suggest that policy discussions may be weighted too heavily in the direction of concern over hospital financial status given the ability of hospitals to change their behavior.
Keywords: Medicare; Prospective Payment System; Psychiatric Care; Length of Stay
JEL Codes: I11; I18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Medicare prospective payment system (PPS) (I13) | reduction in length of stay for psychiatric patients (C41) |
| timeprior variable (C29) | increase in length of stay (C41) |
| timepost variable (L87) | significant negative impact on length of stay (I15) |
| Medicare prospective payment system (PPS) (I13) | hospitals' sensitivity to financial incentives (I11) |
| hospital ownership type (for-profit vs. nonprofit) (L39) | no significant alteration in effects of PPS on length of stay (C41) |