اصلاحات بازار کار، بی ثباتی شغلی، و انعطاف پذیری روابط اشتغالی

We endogenize separation in a search model of the labor market and allow for bargaining over the continuation of employment relationships following productivity shocks to take place under asymmetric information. In such a setting separation may occur even if continuation of the employment relationship is privately efficient for workers and firms. We show that reductions in the cost of separation, owing for example to a reduction in firing taxes, lead to an increase in job instability and, when separation costs are initially high, may be welfare decreasing for workers and firms. We furthermore show that, in response to an exogenous reduction in firing taxes, workers and firms may switch from rigid to flexible employment contracts, which further amplifies the increase in job instability caused by policy reform.

Bargaining between firms and workers often takes place in the presence of private information. As an example, consider a firm that has the opportunity to adopt a new production technology that is effective only if the firm's workers invest in learning how to use it. The firm is likely to be only incompletely informed about the workers’ costs of making the investment and the workers are likely to be only incompletely informed about the firm's benefits of adopting the new technology. The presence of this private information will hamper the negotiations over the continuation of the employment relationship and may possibly lead to the firm and the worker separating. In this paper we analyze the functioning of decentralized labor markets under circumstances like these. To this end we endogenize separation in a search model of the labor market and allow for negotiations over separations to take place under private information. In most existing search models, bargaining between firms and workers takes place under complete information (see e.g. Pissarides, 2000). An important implication of this assumption is that bargaining is efficient, in that firms and workers instantaneously agree to form, or continue, employment relationships if there are gains from trade and they instantaneously agree to separate if there are none. In contrast, as a well-established literature has shown, if bargaining takes place under asymmetric information the parties may fail to realize all gains from trade (see Myerson and Satterthwaite, 1983, Hall and Lazear, 1984 and Hall, 1995).1 Our paper combines the insights of these two literatures. We present a dynamic random matching model in which firms post vacancies and workers search for jobs. We assume that job creation takes place under complete information (like in Pissarides, 2000) but job destruction may be caused by private information. In particular, we assume that during the course of their relationship a firm and a worker may experience productivity shocks and that the cost of the adjustments necessary to cope with these shocks is private information. As a consequence, wage renegotiation following productivity shocks may be inefficient and the firm and worker may separate failing to realize gains from trade (like in Myerson and Satterthwaite, 1983). Our analysis of this model provides three main insights. First, we show that in a labor market in which workers and firms bargain under asymmetric information, a reduction in the cost of separation (e.g. a firing tax) leads to an increase in job instability, i.e. in the probability that workers and firms separate following productivity shocks. The reason for this is that a reduction in separation costs increases the joint value of the workers and the firms’ outside options and thus induces them to bargain more aggressively. This, in turn, makes it more likely that firms and workers will separate following productivity shocks and thus makes employment relationships more unstable. In this regard it is important to note that, contrary to what would happen in a model in which workers and firms bargain under complete information, in our model workers and firms may sometimes separate even when it would be efficient for them not to do so, i.e. even when their joint expected utility from continuing the employment relationship is greater than their joint expected utility from separating and returning to the labor market. Second, we show that, because separations can be privately inefficient, a reduction in the cost of separation can make some groups in society—and possibly society as a whole—worse off by increasing the probability of separation. To see this, consider the effect of a reduction in the firing tax that has to be paid whenever a firm and a worker separate. On the one hand, such a reduction has a direct positive effect on the worker and the firm's expected welfare, because it lowers separation costs when the productivity shock is so large that separation is actually privately efficient. On the other hand, however, precisely because it becomes less costly to separate, the reduction in the firing tax induces the workers and firms to bargain more aggressively and increases the probability of privately inefficient separation. This negative effect can dominate the direct benefit of a reduction in firing taxes, with the possible consequence that firms and workers can be actually made worse off by a reduction in these taxes. Our analysis allows us to derive precise, and potentially testable, predictions about the conditions under which different groups in society are made worse off by a labor market reform. In this respect, our analysis contributes to the literature that seeks to explain why reforms that reduce labor market frictions have often faced substantial political opposition in spite of their strong support by institutions such as the OECD and the IMF (see, for instance, Saint-Paul, 2000 and Pissarides, 2001). Third, our analysis contributes to our understanding of contractual arrangements between firms and workers and of the effects of changes in the labor market environment on these arrangements. In Section 5 we endogenize contractual arrangements between firms and workers and study the effects of changes in the labor market environment on these arrangements. Specifically, firms and workers often use employment contracts that make it more costly for them to separate in the future. For instance, since the mid-1980s firms in Europe have been able to choose between hiring workers on ‘rigid’ permanent contracts for which separation costs are significant or on ‘flexible’ fixed-term contracts for which they are negligible.2 We provide a novel explanation for why firms and workers may agree to adopt rigid employment contracts that increase future separation costs. In particular, we show that firms and workers that anticipate future bargaining inefficiencies can use these contracts as a means of committing to less aggressive bargaining behavior and thus ‘stabilize’ their employment relationship. Our model shows that whether or not firms and workers want to use such contracts for this purpose depends crucially on the labor market environment and the workers’ productivity. In particular, it shows that a sufficiently large reduction in exogenously imposed separation costs, such as a reduction in firing taxes, leads to a one-off switch from rigid to flexible employment contracts. This endogenous contractual response to exogenous labor policy changes induces firms and workers to bargain even more aggressively and thus further destabilizes employment relationships. The rest of the paper is organized as follows. In the next section we discuss the related literature. In Section 3 we present the structure of the model. In Section 4 we solve the model, taking as given the contractually specified separation costs and discuss the effects of a reduction in firing taxes. We then endogenize the contractually specified separation costs in Section 5, check for the robustness of our results in Section 6 and finally conclude in Section 7. All proofs are in Appendix.

When firms and workers bargain over compensation they are often asymmetrically informed about relevant payoffs, such as profits and the workers’ opportunity costs. It is well-known that in such situations bargaining is likely to be inefficient, in the sense that firms and workers fail to realize all gains from trade. In this paper we investigate the implications of this type of bargaining inefficiency for the functioning of labor markets. Our analysis provides three main insights. First, it shows that when bargaining between firms and workers takes place under asymmetric information, a reduction in separation costs leads to an increase in job instability. Essentially, such a reduction makes it less costly for firms and workers to separate which, in turn, induces them to bargain more aggressively. In equilibrium this leads to more job separations. Second, a reduction in separation costs can make firms and workers worse off. In other words, the increase in job instability can outweigh the direct benefit of a reduction in separation costs. Finally, our analysis shows that when wage renegotiations take place under asymmetric information, firms and workers may find it optimal to adopt rigid employment contracts that make it more costly for them to separate in the future. Moreover, a reduction in firing taxes can lead to a one-off switch from rigid to flexible employment contracts. This contractual response induces firms and workers to bargain even more aggressively and thus further destabilizes employment relationships. As regards the robustness of our conclusions to alternative modelling choices, we address possible concerns that our results may depend on the assumed bargaining game by allowing firms and workers to play the most efficient bargaining game and show that for a number of common distributions our results continue to hold. In contrast, throughout we maintain the simplifying assumption that productivity shocks are match-specific. We believe that it would be interesting to analyze a model in which shocks are correlated—as this would shed some light on the, possibly different, effects of changes in separation costs on high and low productivity workers and firms. We leave this analysis for future work.