By Jean-Baptiste Michau
This paper investigates the provision of insurance to workers against search-induced wage fluctuations. I rely on numerical simulations of a model of on-the-job search and precautionary savings. The model is calibrated to low skilled workers in the U.S.. The extent of insurance is determined by the degree of progressivity of a non-linear transfer schedule. The fundamental trade-off is that a more generous provision of insurance reduces incentives to search for better paying jobs, which is detrimental to the production efficiency of the economy. I show that progressivity raises the search intensity of unemployed worker, which reduces the equilibrium rate of unemployment, but lowers the search intensity of employed job seekers, which results in a lower output level. I also solve numerically for the optimal non-linear transfer schedule. The optimal policy is to provide almost no insurance up to a monthly income level of $1450, such as to preserve incentives to move up the wage ladder, and full insurance above $1650. This policy halves the standard deviation of labor incomes, increases output by 2.4% and generates a consumption-equivalent welfare gain of 1.3%. Forbidding private savings does not fundamentally change the shape of the optimal transfer function, but tilts the optimal policy towards more insurance at the expense of production efficiency.
There is no doubt this paper will generate controversy, but it makes sense. Suppose that workers do not like fluctuations in their wages as they move from job to job. Clearly, they would like to obtain insurance against such fluctuations. But if they get it, a moral hazard problem arises whereby they would not search hard enough for a new job if their current one has low pay before insurance. Such an economy would have a poor allocation of resources, as output could be higher with better job matches. The solution appears to be that low-wage jobs should not be insured at all, to preserve incentives for search.
People will object that this provides no insurance to the most vulnerable. We need to define vulnerable here. The common definition would be low-skilled workers who can only obtain low wage jobs. This is not what this paper is about. People who lost out in the life-lottery because they were born with fewer skills or in an environment that is less conducive to accumulate skills should obtain a different type of insurance, likely through social welfare. What this paper is about is how typically young workers bounce around from job to job until they find the right match. You want to provide them some insurance while giving the right incentives to search. And sometimes this involves not giving insurance.