Equity and Efficiency in Unemployment Insurance

This paper is concerned with equity and efficiency in social security. If social security, ormore specifically unemployment insurance, leads to losses in employment and output, menwhy is it such a widely observed phenomenon? The first rationale for social security isutility interdependence or a preference for equality, combined with the public goodcharacter of income redistribution. Secondly, the combination of risk, aversion to risk andprivate insurance market failures is an important raison d'être for social security. Addi-tional reasons for the existence of social security are the administrative efficiency of thesystem and patemalism. However, social security remains a second-best solution, since inpractice taxes that do not distort economie decisions and benefits without mond hazard areimpossible. The second-best character of unemployment insurance raises the question ofthe optimal balance between efficiency and equity (equality and risk reduction). Somemodels of optimal unemployment insurance leave out the equality aspect and indicate theelasticity of unemployment with respect to the replacement ratio, the degree of riskaversion, and capital market imperfections as the determinants of the optimum benefitlevel. However, one's taste for equality plays a major part in determining the optimumlevel of income redistribution. The paper determines a relation between the feasible levelsof equity and efficiency, the output possibilities curve. Three typical views on equity(capturing equality and income protection) are confronted with this output possibilitiescurve: the extreme liberal view, the moderate liberal and the egalitarian view. With regardto unemployment insurance in the Netherlands, we constructed two output possibilitiescurve's: one, assuming the Dutch labour market is characterized by excess-supply and one,assuming the Dutch labour market is instantaneously cleared by flexible wages. Themodels differ significantly in their evaluation of the efficiency cost of Dutch unemploymentinsurance: for example, according to the equilibrium model, the 1982-replacement ratio of65.6 percent causes output to be 13.5 percent smaller than it would be in a situationwithout unemployment insurance, while the output loss is only 1.4 percent, according tothe disequilibrium model. Finally, we calculate the output gains of a mini-system ofunemployment insurance. Again, the results heavily depend on the assumption of aclearing versus a non-clearing labour market in the Netherlands.