How to Model “Churn”

Print/Save PDF

Theorists are always alert for interesting phenomena that lack a name. Lazear and Spletzer (L&S) recently called attention an unlabeled labor-related fact that they named “churn”: the hiring of replacements for employees who have left their jobs. Much of the hiring and job-separation that actually occurs reflects churn, not employment expansion or contraction. From L&S (2012, p.575): “Hires occur for two reasons—to grow a business and to replace those who have left. Hiring can be for expansion or it can be associated with churn. Analogously, separations reflect a decrease in the size of the business or the departure of a current employee who is replaced by a new employee…. Churn is procyclical. Churn declines during recessions because separations, which during good times would have been associated with a replacement hire, are allowed to go unfilled during recessions. As a result, employment declines. Churn also declines during recessions because workers become reluctant to quit their jobs, and in response businesses reduce their hiring. Hiring declines during recessions.”

Here is my claim. GEM Project always provides more robust explanations for rational labor-market behavior than does mainstream theory. Churn is no exception.

Problematic L&S Analysis

L&S work wholly within the mainstream market-centric DSGE framework, a choice that causes their explication of churn to be – putting it kindly – incomplete. Two omissions are especially damaging. First, involuntary job loss (IJL), shown by the evidence to be the dominate reason that employment contracts during recessions, is excluded. L&S mention layoffs but never effectively integrate the concept into their analysis. Instead, they loosely assign cyclical-employment causality to firms “allowing separations to go unfilled”. Forced layoffs are a crucial labor adjustment that cannot be understood as allowing separations to go unfilled.

Second, persistent wage rents are ignored. A substantial incidence of labor pricing chronically in excess of market opportunity costs, as indicated by the evidence, powerfully informs the useful analysis of churn. Harris and Todaro (1970) taught us that, if one economic venue pays rents and another does not, a queue of workers seeking high-wage employment rationally organizes itself. The size of the queue is motivated by the magnitude of the wage premium and perceptions of the likelihood of obtaining a rent-paying job. In specialized economies, rational Harris-Todaro labor transfer must be a critical part of any adequate model of churn. (There will be more on H-T next week.)

GEM Project Analysis

The generalization of rational, price-mediated exchange from the marketplace to the large-establishment workplace is central to adequately modeling churn. The GEM Project enables the derivation of meaningful wage rigidity (MWR) needed to suppress wage recontracting and thereby introduce involuntary job loss and chronic labor rents into coherent general decision-rule equilibrium. In the simplest version of the generalized-exchange model class, all price-mediated exchange other than between LEV employers and employees occurs in the marketplace and is governed by familiar textbook analysis.

In the small-establishment venue (SEV), worker hours supplied and demanded are equilibrated in the labor market with short lags rooted in price-discovery frictions – the familiar textbook labor-search story. Cost-effective supervision of on-the-job behavior (OJB) restricts worker discretionary action to the marketplace, inducing investment in information acquisition related to stay-quit decisions as well as the on-going quest for rent-paying LEV jobs. Rational behavior that equates the market wage and the marginal disutility of work generates labor churning, into and out of SEV jobs, into and out of unemployment, and into and out of the labor force. Market information costs limits workers’ grasp of the true state of the economy as well as their own opportunity costs, requiring stochastic decision-making. That market-centric framework is used by L&S to analyze churn. In highly specialized economies, the story is badly incomplete.

In the large-establishment venue (LEV), cooperative labor input is restricted by costly, asymmetric workplace information as well as routinized jobs and, as a result, cannot be measured by hours alone. Optimal exchange relocates from the marketplace to the workplace, a distinct venue constructed by the firm to facilitate effective labor management. Profit-seeking firms pay wages that minimize unit costs while workers use their latitude on the job to pursue axiomatic preferences that govern their satisfaction with management policies. (Chapter 2)

The rational LEV wage, embodying cyclical downward rigidity as well as chronic rent, is the lynchpin of dynamic workplace decision-rule equilibrium. Worker stay-quit decisions, for those who remain in the labor force, generally collapse to rationally choosing to stay. Employers and employees both invest in acquiring information relevant to managing workplace reference standards that influence the transformation of labor hours into cooperative input. LEV practitioners understand their central labor problem to be nonmarket in nature: How to induce, given costly, asymmetric workplace information and routinized jobs, employees’ voluntary acceptance of management’s goals?

Given imperfect information, management choices concerning employee reference standards and production capability are stochastic in nature. Firms play the averages with respect to OJB and unit labor costs, while their employment decisions are motivated by stationary and nonstationary profit expectations. LEV compensation practices, rationally suppressing wage recontracting in lieu of forced job loss, powerfully constrain marketplace decision-rule optimization and generate crucial intermarket spillovers from temporary and chronic job/hours rationing. Those constraints on household and small-firm decisionmaking permit the coexistence of dynamic general (decision-rule) equilibrium and the failure of markets to clear. The economy instead features involuntary job loss and joblessness at both business-cycle and nonstationary frequencies. Putting the pieces together, interacting workplace and marketplace venues, especially as the economy adjusts to their inconsistent labor pricing, uniquely microfounds recognizable churn.

Parting Comment

I want to emphasize the bold assertion made above. The GEM Project always provides more robust explanations for rational labor-market behavior than does mainstream theory. The next two weeks’ posts will provide further support for that claim, looking closely at Harris-Todaro labor transfer and the natural rate of unemployment. Like churn, they are not exceptions to the unrivaled power of GEM modeling.

Blog Type: Wonkish San Miguel de Allende, Mexico

Write a Comment

Your email address will not be published.