Academic journal article
By Altman, Morris
Review of Social Economy , Vol. 62, No. 4
Abstract Contrary to the conventional view that unemployment insurance serves to directly increase the rate of unemployment as well as reducing an economy's competitiveness by increasing the market wage of labor, the argument presented in this paper is that this worldview critically depends on unrealistic behavioral assumptions. A more realistic modeling suggests that unemployment rates need not rise and competitiveness need not deteriorate with the introduction of or improvements in unemployment insurance, which can also induce increases in economic efficiency. These analytical predictions are consistent with the empirics of unemployment insurance. Unemployment insurance can therefore protect the unemployed without damaging the economy.
Keywords: unemployment insurance, x-efficiency, competition, bargaining power, wellbeing
The conventional wisdom argues that unemployment insurance serves to increase the rate of unemployment by various means. Unemployment rates are said to increase by attracting into the labor market individuals who intend to quit their new jobs so as to collect this benefit, by increasing the voluntary job search time of those unemployed workers who are already in the labor force, by inducing increasing quite rates of the currently employed so that they search for better jobs, and by increasing the market wage thereby increasing the price of labor and reducing the overall competitiveness of the economy. Moreover, following upon the efficiency wage literature, unemployment insurance is expected to reduce the effort incentive effect of a given rate of unemployment, both forcing up real wages to compensate and, thereby, increasing the unemployment rate. The standard approach assumes that the increased duration of short term job search by the unemployment induced by unemployment insurance can have no positive effect on long term employment rates. It is also assumed that higher real wages necessarily or typically generate higher production costs and thereby higher rates of unemployment. It is further assumed that the marginal worker maximizes utility or economic wellbeing at low levels of real income thus allowing unemployment insurance to serve as a utility maximizing "wage of being unemployed". Thus, some workers maximize their utility by getting themselves laid off so as to take advantage of unemployment insurance.
The available empirical evidence provides no unambiguous support for the conventional proposition that unemployment insurance damages the economy. It is critically important from both an analytical and public policy perspective to develop a theoretical framework which can incorporate the empirics suggesting that unemployment insurance generates no long run negative economic effects. Situating itself in the context of the debate, this paper is most concerned with critically evaluating the assumptions underlying the theory-based arguments that various supply side attributes of unemployment insurance negatively affect the trend unemployment rate. To the extent that any of these assumptions are false, the negative analytical predictions that flow from standard analysis tend to be undermined, even prior to adjusting the basic model for institutional details and demand side effects. Such adjustments, by themselves, would cast some considerable doubt on the validity of the standard negative view of unemployment insurance (Atkinson and Micklewright 1991).
Focusing on the theoretical underpinnings of the conventional wisdom is critical since without a convincing theoretical alternative, scholars and policy makers tend to doubt empirical results which speak against the theory. Moreover, empirical studies of the impact of unemployment insurance on unemployment rates are built upon a theoretical infrastructure so that variables selected in the regression and the expected results are conditioned by the theoretical worldview of the analyst. Thus, theory either implicitly or explicitly plays a determining role in the empirical research program and in the results generated. …