Working Paper: CEPR ID: DP6261
Authors: Michael Kuhn; Luigi Siciliani
Abstract: Performance indicators are increasingly used to regulate quality in health care and other areas of the public sector. We develop a model of contracting between a purchaser (principal) and a provider (agent) under the following scenarios: a) higher ability increases quality directly and indirectly (through a lower marginal cost of quality); b) the provider can game the system (or misreport); c) the purchaser is averse to inequality. The main results are: 1) if ability has a direct impact on quality, quality effort is lower and distortions from informational rents are higher; under some conditions more able providers exert lower quality effort; 2) gaming or misreporting reduces the optimal level of quality effort but generally increases the level of quality targets (and measured quality). Pooling may be optimal for high types (or all types) if the benefit function is sufficiently concave and the cost of manipulation is at intermediate levels; 3) If patients receive a low benefit from health care because quality is low then higher inequality aversion increases the optimal level of quality and the associated transfer, and reduces quality dispersion.
Keywords: adverse selection; falsification; inequality aversion; moral hazard; performance; quality
JEL Codes: D82; I11; I18; L51
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Provider ability (H42) | Quality of care (I11) |
| Quality of care (I11) | Quality effort (L15) |
| Provider ability (H42) | Distortions from informational rents (D89) |
| Gaming behavior (C70) | Quality effort (L15) |
| Gaming behavior (C70) | Quality targets (L15) |
| Inequality aversion (D63) | Optimal level of quality (L15) |
| Inequality aversion (D63) | Quality dispersion (D39) |
| Low quality (L15) | Patient benefits (I11) |