Frictional Labor Markets: Theory and Implications

该项目建立了一个基于信息不完全的摩擦性失业理论。人们认为，工人观察所有潜在雇主的特征是代价高昂的。相反，他们观察到一个随机的雇主子集，他们选择最优的样本量。该模型的均衡表现为摩擦型失业；一些工人失业，而一些符合条件的雇主有空缺职位。然而，与微观基础不那么稳固的现有摩擦性失业模型相比，该模型的含义更接近于基线的瓦尔拉斯模型。除了得出许多新的令人惊讶的结论外，这一理论还具有易操作性的额外优势。初步研究表明，可以使用该模型来检验风险厌恶和不完善的劳动收入保险市场对摩擦性失业的影响。随着失业、厌恶风险的工人消耗自己的资产，他们越来越愿意寻找低工资、低生产率的工作，这些工作的排队时间更短，就业机会更高。因此，最优求职行为意味着只失业一段时间的工人的期望工资远远高于极度厌恶风险的长期失业者的期望工资。该模型也可用于研究异质员工如何寻找异质工作。以前的研究侧重于匹配过程，而不是搜索过程，表明无摩擦匹配或分配模型的结论在没有额外假设的情况下不会延续到摩擦环境中。然而，通过仔细建模摩擦性失业背后的信息不完善，这一理论确立了无摩擦性结果推广到摩擦性环境。在任何时候，都有超过5%的美国劳动力遭受摩擦性失业之苦，仅出于这个原因，了解这一水平是否最优就很重要。但理解摩擦性失业还有其他重要原因：市场经济是否会为失业工人提供最佳的工作激励，或者人们是否会低效地寻找工作，接受不合适的工作，或者拒绝合适的工作？摩擦性失业如何影响创造新就业机会或投资于人力和物质资本的激励措施？这个项目解决了所有这些问题，初步研究得出了与现有文献所建议的完全不同的结论。

This project builds a theory of frictional unemployment based on information imperfections. It is assumed that it is costly for workers to observe the characteristics of all potential employers. Instead, they observe a random subset of employers choosing optimally their sample size. The equilibrium of this model exhibits frictional unemployment; some workers are unemployed, while some eligible employers have open vacancies. However, the implications of this model are much closer to the baseline Walrasian model than to existing modals of frictional unemployment with less solid microfoundations. Besides reaching a number of new and surprising conclusions, this theory has the added advantage of tractability. Preliminary research shows the possibility of using this model to examine the implications of risk-aversion and imperfect labor-income insurance markets for frictional unemployment. As unemployed risk-averse workers consume their assets, they become increasingly willing to look for low wage, low productivity jobs where queues are shorter and employment chances are higher. Thus, optimal job search behavior implies that the expected wage of a worker who has been unemployed for only a short period of time, is much higher than the expected wage of the long-term unemployed who grow extremely risk-averse. This model can also be used to study how heterogeneous workers search for heterogeneous jobs. Previous research, which focused on the matching process and not on the search process, showed that the conclusions of frictionless matching or assignment models do not carry over to a frictional environment without additional assumptions. However, by carefully modeling the information imperfections behind frictional unemployment, this theory establishes that frictionless results generalize to a frictional setting. At any point in time, over 5% of the U.S. labor force suffer from frictional unemployment, and for that reason alone it is important to understand whether the level is optimal. But understanding frictional unemployment is important for other reasons: does a market economy give optimal incentives for an unemployed worker to take a job, or do people search inefficiently, accept inappropriate jobs, or reject appropriate ones? How does frictional unemployment affect the incentives to create new jobs, or to invest in human and physical capita1? This project addresses all these issues and preliminary research leads to substantially different conclusions than the existing literature suggests