Thread: The Labor Market Effects of Employer Recruitment Choice

Employers and job seekers are brought together for potential matches through their recruitment and job search activities. Employers may post help wanted signs, run newspaper advertisements, or seek referrals from private employment agencies. Job seekers may speak with friends and relatives, solicit the aid of the state employment agency, or simply walk in and apply. These recruitment and search activities help both parties acquire information about each other, and the more information they obtain prior to entering an employment agreement the higher the likelihood of a good employment match. The crucial role of information in the labor market has been recognized since Stigler (1962), but despite a voluminous literature on job matching and organizational behavior in labor markets, we know far less about employers’ recruitment strategies than about job seekers’ search strategies. As Granovetter (1995) notes, “while people are finding jobs, employers are finding people to fill them, and their behaviors, strategies, and purposes play a central but often neglected role in the process of matching people to jobs.”
This imbalance in research effort is explained more by a dearth of adequate data describing employer recruitment behavior than by lack of scholarly interest. Indeed, labor economists, sociologists, psychologists, and human resource management specialists have spent the last half-century exploiting the meager existing data sets in efforts to learn something about employer recruitment behavior. The typical approach is to regress some labor market outcome, such as wages or vacancy duration, on a set of explanatory variables including “recruitment dummies”, which are assumed exogenous.
This paper is the first to estimate a dynamic structural model treating the employer’s recruitment choice as endogenous. In the model, employers optimally select recruitment strategies, wage offers, and desired worker skill levels to maximize profits, with the knowledge that some recruitment strategies are more likely than others to generate a worker who is well matched to the firm. This gives rise to a discrete-choice dynamic optimization problem under uncertainty (the stochastic match quality being the source of uncertainty) that I solve numerically. I estimate the parameters of the model using a newly available cross-sectional employer telephone survey, the Multi-City Study of Urban Inequality (MCSUI), that contains detailed information about recruitment behavior.
The advantage of a structural approach, as opposed to the standard approach of treating the recruitment choice as an exogenous regressor, is that it can be used to analyze the effects of government policies that have labor market effects by influencing employers to alter their recruitment decisions, thereby changing hiring outcomes. The model allows us to predict how employers alter their recruitment behavior in response to changes in the policy environment. Given a set of parameter estimates, I conduct simulations to analyze and compare the labor market effects of two types of government policies, “information policies” and “hiring incentive policies”, both of which are designed to improve job placement rates for targeted groups of low-skilled workers.
Information policies, such as the Workforce Investment Act of 1998, attempt to improve the information available to potential employers about these workers, thereby improving match quality. Hiring incentive policies, such as the Work Opportunity and Welfare-to-Work tax credits, give employers a direct incentive, in the form of a tax credit, to hire workers from targeted groups. By changing the relative attractiveness of worker-types in the employer’s eyes, both types of policies induce employers to alter recruitment strategies so as to attract the targeted workers. The main conclusion that emerges from the policy simulations is that hiring incentives are superior to information policies, in that they induce greater substitution from skilled to unskilled labor, and they increase unskilled starting wages by a larger amount. This finding is of particular interest given that current federal expenditure on information policies far exceeds that on hiring incentives.
A main contribution of the paper is that it provides a framework for analyzing how an important employer decision, the choice of recruitment methods for hiring new workers, affects labor market outcomes. This is useful for analyzing a wide array of government policies that affect employer recruitment choice, beyond the two “information” and “hiring incentive” policies analyzed in this paper. For example, if workers are differentiated by race or gender, the model could be used to analyze how affirmative action policies affect the hiring process. Another contribution of the structural model is that it sheds light on how the distribution of starting wages is affected by recruitment choices, a mechanism I call the “recruitment-wage effect.” Finally, as there has been no prior structural work in the area of employer recruitment, I hope that this paper will contribute to a deeper understanding of the effect of information and employer behavior on the job matching process.
The discussion begins with a survey of the previous literature on employer recruitment and wage-posting models of job search. I then present the economic model of employer recruitment choice, a description of the data, and some empirical support for the assumptions of the economic model. I then formulate an empirical model and discuss its estimation by the method of simulated moments. After estimating the parameters of the model, I conduct policy simulations to analyze the labor market effects of “information” and “hiring incentive” policies that are designed to improve job placement rates for targeted groups of low-skilled workers. Finally, I discuss the “recruitment-wage” effect, which illustrates how wages are affected by employer recruitment choice. The paper concludes with some suggestions for future research.