Unemployment and the Labor Market

‘Quit Turbulence and Unemployment,’ T. Sargent,  Lars Ljungqvist (with I. Baley), 2018.

According to physicist Steven Weinberg (2018): (1) new theories that target new observations should be constrained to agree with observations successfully represented by existing theories; and (2) the constraint to preserve successes of earlier theories is also a guide that can lead to unanticipated understandings of yet other phenomena. We illustrate how Weinberg’s advice helps answer the question: in more turbulent times, how do higher risks of skill losses coincident with involuntary layoffs (“layoff turbulence”) but also with voluntary quits (“quit turbulence”) affect equilibrium unemployment rates? Earlier theories that had included only layoff turbulence had established a positive relationship between turbulence and the unemployment rate within generous welfare states like those in Europe, but the absence of that relationship in countries without much of a welfare state. An influential earlier paper found that even small amounts of quit turbulence would lead to a negative relationship between turbulence and unemployment rates. We show that that finding was based on a peculiar calibration of returns to labor mobility that not only made the model miss the positive turbulence-unemployment rate relationship in the post 1970s data for European welfare states, but also miss observations about labor market churning. Repairing the faulty calibration not only brings models with quit turbulence into line with those observations but also guides us to shed light on labor market issues in the US raised by Alan Greenspan (1998)

‘Declining Search Frictions, Unemployment and Growth,’ G. Menzio, (with P. Martellini), 2018.

Over the last century, unemployment, vacancy, job-finding and job-loss rates as well as the Beveridge curve have no trend. Yet, the last century has seen the development and diffusion of many information technologies—such as telephones, fax machines, computers, the Internet—which presumably have increased the efficiency of search in the labor market. We explain this phenomenon using a textbook search-theoretic model of the labor market. We show that there exists an equilibrium in which unemployment, vacancies, job-finding and job-loss rates are constant while the search technology improves over time if and only if firm-worker matches are heterogeneous in quality, the distribution of match qualities is Pareto, and the quality of a match is observed before the start of the employment relationship. Under these conditions, improvements in search lead to an increase in the rate at which workers meet firms and to a proportional decline in the probability that the quality of a firm-worker match is acceptable leading to a constant job-finding rate, unemployment, etc… Interestingly, under the same conditions, unemployment, vacancies, job-finding and job-loss rates are independent of the size of the labor market even in the presence of increasing returns to scale in search. While declining search frictions do not lower unemployment, they contribute to growth. The magnitude of the contribution depends on the thickness of the tail of the Pareto distribution. We present a simple strategy to measure the decline in search frictions and its contribution to growth. A rudimentary implementation of this strategy suggests that the decline in search frictions has been substantial, it has been caused by both improvements in the search technology and increasing returns to scale in the search process, and it has had a non-negligible impact on growth..

‘Turbulence and Unemployment in Matching Models,’ Sargent, T., Ljungqvist L., (with I. Baley), 2018.

Ljungqvist and Sargent (1998, 2008) show that worse skill transition probabilities for workers who suffer involuntary layoffs (i.e., increases in turbulence) generate higher unemployment in a welfare state. den Haan, Haefke and Ramey (2005) challenge this finding by showing that if higher turbulence means that voluntary quits are also exposed to even a tiny risk of skill loss, then higher turbulence leads to lower unemployment within their matching model. We show (1) that there is no such brittleness of the positive turbulence unemployment relationship in the matching model of Ljungqvist and Sargent (2007) even if we add such “quit turbulence”, and (2) that if den Haan et al. had calibrated their productivity distribution to fit observed unemployment patterns that they miss, then they too would have found a positive turbulence-unemployment relationship in their model. Thus, we trace den Haan et al.’s finding to their assuming a narrower productivity distribution than Ljungqvist and Sargent had. Because den Haan et al. assume a distribution with such narrow support that it implies small returns to reallocating labor, even a small mobility cost shuts down voluntary separations. But that means that the imposition of a small layoff cost in tranquil times has counterfactually large unemployment suppression effects. When the parameterization is adjusted to fit historical observations on unemployment and layoff costs, a positive relationship between turbulence and unemployment reemerges.

‘Discretizing Unobserved Heterogeneity,’ Manresa, E. (with S. Bonhomme, and T. Lamadon), 2017.

We study panel data estimators based on a discretization of unobserved heterogeneitywhen individual heterogeneity is not necessarily discrete in the population. We focus on two-step grouped-fixed effects estimators, where individuals are classified into groups in a first step using kmeans clustering, and the model is estimated in a second step allowing for group-specific heterogeneity. We analyze the asymptotic properties of these discrete estimators as the number of groups grows with the sample size, and we show that bias reduction techniques can improve their performance. In addition to reducing the number of parameters, grouped fixed-effects methods provide effective regularization.For instance, when allowing for the presence of time-varying unobserved heterogeneity we show they enjoy fast rates of convergence depending on the underlying dimension of heterogeneity. Finally, we document the finite sample properties of two-step grouped fixed-effects estimators in two applications: a structural dynamic discrete choice model of migration, and a model of wages with worker and firm heterogeneity.

‘A Note on the Estimation of Job Amenities and Labor Productivity,’ Galichon, A. (with A. Dupuy), 2017.

This note introduces a maximum likelihood estimator of the value of job amenities and labor productivity in a single matching market based on the observation of equilibrium matches and wages. The estimation procedure simultaneously fits both the matching patterns and the wage curve. Our estimator is suited for applications to a wide range of assignment problems.

‘High Wage Workers Work for High Wage Firms,’ Borovičková, K. (with R. Shimer), 2018.

We develop a new approach to measuring the correlation between the types of matched workers and firms. Our approach accurately measures the correlation in data sets with many workers and firms, but a small number of independent observations for each. Using administrative data from Austria, we find that the correlation between worker and firm types lies between 0.4 and 0.6. We use artificial data sets with correlated worker and firm types to show that our estimator is accurate. In contrast, the Abowd, Kramarz, Margolis (1999) fixed effects estimator suggests no correlation between types in our data set. We show both theoretically and empirically that this reflects an incidental parameter problem.

‘Risk Premia and Unemployment Fluctuations,’ Borovičková, K. and Borovičková, J., 2018.

We study the role of fluctuations in discount rates for the joint dynamics of expected returns in the stock market and employment dynamics. We construct a non-parametric bound on the predictability and time-variation in conditional volatility of the firm’s profit flow that must be met to rationalize the observed business-cycle fluctuations in vacancy-filling rates. A stochastic discount factor consistent with conditional moments of the risk-free rate and expected returns on risky assets alleviates the need for an excessively volatile model of the expected profit flow.

‘The Proportional Hazard Model: Estimation and Testing using Price Change and Labor Market Data,’ Borovičková, K. (with F. Alvarez and R. Shimer), 2016.

We use labor market data and data on price changes to examine the role of structural duration dependence and heterogeneity in shaping the aggregate hazard rates. In line with an extensive literature we examine this question through the lens of a mixed proportional hazard model. While we think that this model is a convenient representation of the data, we recognize that its structure can be too restrictive. We focus on environments where we observe two observations per individual as this not only allows us to estimate the model non-parametrically, but also test whether the true data-generating process is likely to have a structure imposed by a mixed proportional hazard model. We reject that this is the case both for the price change data and labor market data. We then turn to data simulated from reasonable structural models, none of which can be represented as a mixed proportional hazard model, to examine implications of estimating a misspecified mixed proportional hazard model. We use a “CalvoPlus” model for price changes, while for the labor market data, we assume that individual durations follow an inverse Gaussian distribution. We find that, in fact, the mixed proportional hazard model is a good approximation of the CalvoPlus model and therefore the estimated baseline hazard rate is very similar to the true structural hazard rate. This is not the case for the inverse Gaussian model for the labor market, where the mixed proportional hazard model cannot be viewed as a good approximation. As a consequence, fitting a mixed proportional hazard model to these data vastly understate the importance of heterogeneity in the economy.

‘Job Flows, Worker Flows and Labor Market Policies,’ Borovičková, K., 2016.

I study an equilibrium model of the labor market with firm- and worker-level shocks and evaluate the impact of labor market policies in this framework. Firms hire and shed workers in response to firm-specific productivity shocks. Workers and firms learn about the quality of their employment match and separate when they realize they are mismatched. Match quality and productivity shocks must interact in order to explain the hazard rates of separation in the cross section of firm growth rates and workers’ tenures. The model is estimated using a large panel dataset of individual labor market histories in Austria. I find that accounting for worker flows generated by learning and direct job to job transitions, and job flows driven by firm-specific productivity shocks plays a crucial role for the evaluation of the impact of labor market policies on the unemployment rate, duration and average productivity.

‘Decomposing Duration Dependence in a Stopping Time Model,’ Borovičková, K. (with F. Alvarez and R. Shimer), 2016.

We develop a simple dynamic model of a worker’s transitions between employment and non-employment. Our model implies that a worker finds a job at an optimal stopping time, when a Brownian motion with drift hits a barrier. The model has structural duration dependence in the job finding rate, in the sense that the hazard rate of finding a job changes during a non-employment spell for a given worker. In addition, we allow for arbitrary parameter heterogeneity across workers, so dynamic selection also affects the average job finding rate at different durations. We show that our model has testable implications if we observe at least two completed non-employment spells for each worker. Moreover, we can nonparametrically identify the distribution of a subset of our model’s parameters using data on the duration of repeated non-employment spells. We use a large panel of social security data for Austrian workers to test and estimate the model. Our model is not rejected by the data, while a mixed proportional hazard model with arbitrary heterogeneity and an arbitrary baseline hazard rate is rejected using the same data set. Our parameter estimates indicate that dynamic selection is important for the decline in the job finding rate at short durations, while structural duration dependence drives most of the decline in the job finding rate at long durations.

‘A Nonparametric Variance Decomposition Using Panel Data,’ Borovičková, K. (with F. Alvarez and R. Shimer), 2014.

We consider a population of individuals who draw a random variable from an individual-specific distribution that is fixed over time. We propose an unbiased within-between variance decomposition using a short panel of two observations for each individual. We illustrate the usefulness of our decomposition with two applications: decomposing heterogeneity versus structural duration dependence in unemployment, nonemployment, and employment durations; and calculating the importance of frictional wage dispersion for labor market outcomes.

‘Unemployment fluctuations, match quality, and the wage cyclicality of new hires,’ Gertler, M. (with C. Huckfeldt  and A. Trigari), 2018.

We revisit the issue of the high cyclicality of wages of new hires. We show that after controlling for composition effects likely involving procyclical upgrading of job match quality, the wages of new hires are no more cyclical than those of existing workers. The key implication is that the sluggish behavior of wages for existing workers is a better guide to the cyclicality of the marginal cost of labor than is the high measured cyclicality of new hires wages unadjusted for composition effects. Key to our identification is distinguishing between new hires from unemployment versus those who are job changers. We argue that to a reasonable approximation, the wages of the former provide a composition free estimate of the wage flexibility, while the same is not true for the latter. We then develop a quantitative general equilibrium model with sticky wages via staggered contracting, on-the-job search, and variable match quality, and show that it can account for both the panel data evidence and aggregate evidence on labor market volatility.

‘Discount Rates, Learning by Doing, and Employment Fluctuations,’ Midrigan, V. (with P. Kehoe and E. Pastorino), 2015.

We revisit the Shimer (2005) puzzle in a search and matching model with on-the-job human capital accumulation in which households exhibit preference for consumption smoothing. We parameterize the model so that it accords with the micro-evidence on returns to tenure and experience as well as individual life-cycle earning profiles. We find that employment fluctuations in response to productivity shocks are greatly amplified in this environment.

‘The Fundamental Surplus,’ Sargent, T., and L. Ljungqvist, 2017.

To generate big responses of unemployment to productivity changes, researchers have reconfigured matching models in various ways: by elevating the utility of leisure, by making wages sticky, by assuming alternating-offer wage bargaining, by introducing costly acquisition of credit, by assuming fixed matching costs, or by positing government mandated unemployment compensation and layoff costs. All of these redesigned matching models increase responses of unemployment to movements in productivity by diminishing the fundamental surplus fraction, an upper bound on the fraction of a job’s output that the invisible hand can allocate to vacancy creation. Business cycles and welfare state dynamics of an entire class of reconfigured matching models all operate through this common channel.

‘What Nonconvexities Really Say about Labor Supply Elasticities,’ Sargent, T., and L. Ljungqvist, 2014.

Rogerson and Wallenius (2013) draw an incorrect inference about a labor supply elasticity at an intensive margin from premises about an option to work part time that retiring workers decline. We explain how their false inference rests on overgeneralizing outcomes from a particular example and how Rogerson and Wallenius haven’t identified an economic force beyond the two — indivisible labor and time separable preferences — that drive a high labor supply elasticity at an interior solution at an extensive margin.

‘Firms’ Choices of Wage-Setting Protocols in the Presence of Minimum Wages,’ Flinn, C. (with J. Mabli and J. Mullins), 2017.

We study the formation of wages in a frictional search market where firms can choose either to bargain with workers or post non-negotiable wage offers. Workers can secure wage increases for themselves by engaging in on-the-job search and either moving to firms that offer higher wages or, when possible, leveraging an outside offer into a higher wage at the current firm. We characterize the optimal wage posting strategy of non-negotiating firms and how this decision is influenced by the presence of renegotiating firms. We quantitatively examine the model’s unique implications for efficiency, wage dispersion, and worker welfare by estimating it using data on the wages and employment spells of low-skill workers in the United States. In the estimated steady state of the model, we find that more than 10% of job acceptance decisions made while on the job are socially sub-optimal. We also find that, relative to a benchmark case without renegotiation, the presence of even a small number of these firms increases the wage dispersion attributable to search frictions, deflates wages, and reduces worker welfare. Moving to a general equilibrium setting, we use the estimated model to study the impact of a minimum wage increase on firm bargaining strategies and worker outcomes. Our key finding is that binding minimum wages lead to an increase in the equilibrium fraction of renegotiating firms which, relative to a counterfactual in which this fraction is fixed, significantly dampens the reduction in wage dispersion and gains in worker welfare that can typically be achieved with moderate minimum wage increases. Indeed, the presence of endogenous bargaining strategies reverses the sign of the average welfare effect of a $15 minimum wage from positive to negative.

‘Simultaneous Search in the Labor and Marriage Markets with Endogenous Schooling Decisions,’ Flinn, C. (with L. Flabbi), 2015.

Labor market decisions are not taken in isolation when individuals are engaged in stable relationships. There now exist a number of estimated models of household search able to address and estimate the impact of these decision processes. However, in these cases a number of simplifying assumptions have been made that limited the usefulness of the models for policy evaluation purposes, notably the lack of intrahousehold behavior and of the process that led to the formation of the household. Our analysis, instead, develops and estimates a model designed to determine the joint equilibrium distribution of schooling levels, labor market outcomes, and marriage market statuses. The model is estimated using the Method of Simulated Moments (MSM) using labor market information from the Current Population Survey (CPS) and marriage market information from the American Community Survey (ACS). We plan to use the estimates of the model to perform several comparative statics exercises in order to: (i) separate the impact of the labor market from the impact of the marriage market in determining lifetime returns to schooling; (ii) explore the impact of eliminating gender differences in the wage offers distribution on marriage rates, assortative mating patterns, and schooling investments; (iii) assess the importance of marital status in determining labor market outcomes. Finally, we plan to use the parameter estimates to perform a series of policy experiments comparing a labor income tax system based on individual taxation with a system based on joint taxation.