Working Paper: CEPR ID: DP2519
Authors: Pedro Pita Barros; Pau Olivella
Abstract: We develop a model of waiting lists for public hospitals when physicians deliver both private and public treatment. Public treatment is free but rationed, i.e., only cases meeting some medical criteria are admitted for treatment. Private treatment has no waiting time but entails payment of a fee. Both physicians and patients take into account that each patient treated in the private practice schedule reduces the waiting list for public treatment. We show that physicians do not necessarily select the mildest cases from the waiting list. We provide sufficient conditions on the rationing policy under which cream skimming is always partial. We show that, to a large extent, one can by-pass the analysis of doctors' behaviour in the characterization of patient selection.
Keywords: cream skimming; waiting lists
JEL Codes: H51; I11; I18
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Rationing policy (D45) | Cream skimming (D49) |
| Intermediate rationing policies (D45) | Full cream skimming (Y60) |
| Severity of cases treated in the private sector (I11) | Cream skimming (D49) |
| Doctors' willingness to treat varying severity levels (I11) | Selection process (C52) |
| Doctors' preferences and treatment costs (I11) | Selection process (C52) |