هزینه چرخه های کسب و کار و ارزش تثبیت بیمه بیکاری

This paper offers a new perspective on why labor market policies aimed at reducing the cost of business cycles may be warranted and how such policies can be designed in order to improve welfare. To this end, we develop a quantitative dynamic equilibrium model to illustrate how the contractual structure of the labor market may hide significant undiversified wage risk induced by aggregate fluctuations. The environment analyzed is such that the only imperfectly diversified risk workers bear is the risk of losing their job when the market for new contracts is depressed. When we fit the model to replicate the amount of wage variation estimated from micro-data, we obtain estimates of the potential value of stabilization policies that are substantially larger than those found in the literature. We use this framework to examine several policy issues and, in particular, to show why state-contingent unemployment insurance may dominate non-contingent unemployment insurance schemes.

When analyzing policies aimed at reducing the cost of business cycles, the ultimate goal is to evaluate these policies in terms of their effectiveness in generating welfare gains. However, ever since Lucas (1987) placed in doubt the intrinsic value of eliminating aggregate fluctuations, it has become contentious to claim that any stabilization policy is worth pursuing. In his influential study, Lucas presents a simple estimation of the cost of aggregate instability. His estimate computes the uniform percentage increase in consumption that is needed to leave a consumer indifferent between a consumption stream with the U.S. consumption variability, and a smooth consumption path. This calculation indicates that the business cycle generates an almost negligible welfare cost – less than one-tenth of a percentage point of consumption. Lucas thus concludes that the potential gains attainable from stabilizing the economy may be negligible. Lucas's computation is based on the assumption that markets are complete and hence all but aggregate risk is diversifiable. Although this is a questionable assumption, Lucas's work clearly shows that stabilization policies generate substantial welfare gains only if they help to reduce some undiversifiable components of business cycle risk.1 Following this observation, Imrohoroglu (1989) and Atkeson and Phelan (1994) have examined the potential gains associated with reducing economic instability in economies where markets are incomplete.2 Both papers argue that employment fluctuations are not shared evenly among different individuals. Instead, the burden of recessions falls disproportionately among those few who lose their jobs in recessions. Based on this premise, both papers focus on unemployment risk as the principal undiversified risk associated with the business cycle. For example, Imrohoroglu measures the welfare cost of the business cycle in an economy where individuals have access to a storage technology but limited access to credit and insurance. She compares a cyclical and a stabilized economy that differ only in the transition probabilities between employment and unemployment. In the cyclical economy, the probability of becoming unemployed and the probability of remaining unemployed is more likely during a recession than it is during a boom. Imrohoroglu's estimates of the cost of aggregate fluctuations in this environment are also small, on the order of 0.3% of consumption.3 The reason for this result is the relatively small time variation in the average duration of U.S. unemployment. Thus, the risk of being unemployed longer in a recession is relatively small and smoothing out the cycle in order to have a constant expected duration of unemployment does not lead to big welfare gains.4 We also premise our analysis on the view that the burden of the business cycle is borne mainly by those who are laid off in recessions. However, we claim that focusing only on the time variability of unemployment duration may underestimate the welfare gains of stabilization policies and bias policy analysis. Indeed, the main purpose of this paper is to emphasize that the mild variability of the aggregate wage may hide important business cycle fluctuations in individual wages and that this source of risk induces substantial welfare costs. We argue that there may be significant welfare gains from stabilization policies even when abstracting from unemployment risk (or risk associated with variation in asset returns, as in Atkeson and Phelan (1994)). The second aim of this paper is to examine different economic policies designed to reduce wage risk. We find that simple unemployment insurance schemes may be inefficient at diversifying the risk associated with economic fluctuations, and we indicate how and why alternative programs may be preferable. Our model builds on evidence suggesting that the labor market may be better characterized as a market for contracts than as a spot market. In particular, we rely on the work of Bils (1985), Beaudry and DiNardo 1991 and Beaudry and DiNardo 1995 and Jacobson et al. (1993), which present empirical evidence that labor market outcomes exhibit a pattern of history-dependence suggestive of dynamic enforcement-constrained implicit contract theory. In our model, capital market imperfections and different commitment possibilities for workers and firms, give rise to risk-sharing contracts between workers and firms that are similar in nature to those examined by Harris and Holmstrom (1982). The environment considered assumes two types of shocks: aggregate and allocative. In this environment, it is optimal for an employer to insure workers against deteriorating labor market conditions induced by aggregate shocks. However, if a job is subject to a reallocation shock, the worker is laid off and must endure the burden associated with labor market conditions. Hence, the risk in our model is the result of the interaction between allocative and aggregate shocks and the undiversifiable risk is that of being laid off when the market for new contracts is depressed. We fit our model to replicate the amount of wage risk estimated from micro-data and use it to asses the cost of the business cycle in an environment where workers receive unemployment subsidies (UI) when unemployed. This calibration exercise relies heavily on the estimates of cyclical wage movements estimated by Beaudry and DiNardo (1991). The first finding in this paper is that accounting for contractual wage risk leads to estimates of the costs of business cycles that are considerably larger than those found in the literature. Intuitively, the reason why contractual wage risk leads to estimates of the value of stabilization policies that are substantially larger than estimates based on unemployment-duration risk is that wage risk has a very significant time-varying dimension. Moreover, this risk induces changes in workers’ income that are very persistent and thus difficult to smooth away with personal savings. The second finding is that recognizing the importance of contractual wage risk has implications for the design of policy. In particular, we find that in an environment characterized by substantial variations in contract wages, unconditional UI may be an inefficient way of reducing the cost of economic fluctuations.5 In contrast, we find that state-contingent UI can generate substantial welfare gains. In fact, by increasing wages during recessions and mildly decreasing wages during booms, such a policy improves risk-sharing and reduces the cost of aggregate fluctuations. The paper is structured as follows. Section 2 presents the model. Section 3 discusses how the model is calibrated and uses it to assess the potential welfare gains associated with stabilization policy. Section 4 evaluates the extent to which different policies may achieve these welfare gains. Finally, Section 5 draws some conclusions.

In this paper we present a quantitative dynamic general equilibrium model aimed at replicating several observed micro and macroeconomic features of wage and employment behavior. The objective has been to highlight how the contractual structure of the labor market can hide a significant degree of undiversified wage risk associated with aggregate fluctuations. Our estimates show that this source of risk may be substantial. It leads to an evaluation of the welfare cost of economic fluctuations that is much larger than previous estimates which have been based on aggregate consumption or unemployment duration risk. The intuition behind our result is simple: whereas unemployment risk induces temporary shocks in current income, contractual wage risk induces persistent changes in workers’ income and therefore creates risk that cannot easily be smoothed away by personal savings. We also find that the identification of what might be an important component of undiversified business cycle risk has clear policy implications. The first obvious one is rehabilitating the potential role for stabilization policies as policies that can bring about important welfare gains. The second is that by identifying one reason business cycles may be socially costly, we can recommend stabilization policies that are more directly focused on reducing this risk borne at the individual level. To this end, we have examined the risk-diversification role of UI and find that a simple non-contingent UI system is not a well-targeted policy for addressing this risk. Furthermore, we show why a state-contingent UI program can be a more efficient way of reducing the cost of the business cycle. We have also investigated the role of wage subsidies in improving allocative and risk-sharing efficiency. Overall, we have attempted to explain why labor market policies aimed at mitigating the cost of business cycles may be warranted and how such policies can be designed in order to improve welfare.