Publications
Publications
- 1998
- HBS Working Paper Series
Some Evidence on the Optimal Welfare State Based on Subjective Data
By: Rafael Di Tella and Robert MacCulloch
Abstract
It is often difficult to evaluate all the costs and benefits of the welfare state. This paper suggests an alternative approach based on surveys of citizen satisfaction with welfare programs. In the first part of the paper we estimate the level of unemployment benefits desired by the employed and the unemployed in 9 countries. Perhaps surprisingly, the evidence suggests that sometimes both the employed and the unemployed wish to see higher benefits than the level which the government actually provides. In the second part we study the structure (as opposed to the level) of the welfare state. We find that preferences for higher benefits depend on individual characteristics such as age, income, skill and area of residence. For the United Kingdom, we find that preferences for high unemployment benefits are correlated with the regional unemployment rate, even after controlling for fixed effects. Using regional data helps overcome identification problems. Our results imply that some countries could enjoy a free lunch in terms of welfare reform by making the generosity of unemployment benefits depend on the regional unemployment rate. They also imply that previous work finding that unemployment benefits increase the unemployment rate could have a simultaneity bias.
Keywords
Personal Characteristics; Employment; Surveys; Programs; Government and Politics; Age; Income; Residency; Welfare; Governing Rules, Regulations, and Reforms; Cost vs Benefits; Satisfaction; United Kingdom
Citation
Di Tella, Rafael, and Robert MacCulloch. "Some Evidence on the Optimal Welfare State Based on Subjective Data." Harvard Business School Working Paper, No. 98-092, March 1998.