Working Paper: CEPR ID: DP3290
Authors: David K. Miles; James Sefton
Abstract: This Paper considers the optimal design of public pension provision for the retired and income support for those of working age. We consider social security systems that differ in terms of the level of benefits, degree of means testing and the nature of contributions. We aim to find which system maximises the expected utility of agents behind an initial veil of ignorance about their future labour productivity. We also explore whether systems which differ from the (ex-ante) optimal one could be reformed. We ask whether agents who already know what their productivity is would vote for a move towards a system that they would have found optimal from behind a veil of ignorance about their own position on the wages ladder. We also consider which systems could be sustained in an economy where reform is decided by majority voting. We find a substantial role for means testing in optimal welfare systems. We also find the possibility of multiple equilbria in welfare systems.
Keywords: optimal taxes; pensions; social security
JEL Codes: H10; H20; H30
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| means testing (I38) | expected utility (D81) |
| means testing (I38) | labor supply (J20) |
| means testing (I38) | multiple equilibria in welfare systems (D50) |
| higher taper rates (E43) | increased distortions on labor supply (H31) |
| increased distortions on labor supply (H31) | expected utility (D81) |
| means testing (I38) | redistribution and insurance against low productivity outcomes (J68) |
| agents' awareness of productivity (O49) | favor less redistributive systems (P19) |