Temporary help service firms' use of employer tax credits: implications for disadvantaged workers' labor market outcomes.

1. Introduction

In recent years, important labor market changes have influenced the employment experiences of low-skilled and disadvantaged workers. One strong trend has been an increase in these workers' participation in the temporary help services (THS) employment sector. As THS employment has increased, (1) temporary help service firms have become increasingly prominent labor market intermediaries for disadvantaged groups (Blank 1998; Lane et al. 2003). Furthermore, as the industry more than doubled in size between 1992 and 2000 (Pressler 2002), workers with recent participation in public welfare programs came to make up a disproportionate share of THS employees, a share that has continued to increase over time (Barker and Christensen 1998; Autor and Houseman 2005; Heinrich, Mueser, and Troske 2005).

Another important trend has been an increase in employers' participation in hiring subsidy programs. The Work Opportunity Tax Credit (WOTC) and Welfare-to-Work Tax Credit (WtW) reimburse employers (in the form of federal income tax credits) for up to 50% of the wages of workers with certain employment barriers, with the objective of helping low-income and disadvantaged individuals gain labor market experience, improve their work skills, and keep their jobs longer. (2) In fact, a U.S. Department of Labor publication used to market the program attempts to attract employers with the bold headline "Hiring Welfare Recipients Can Reduce Employer Taxes by $8500 Per New Hire" (http://www.doleta.gov/wtw/resources/fact-taxcredits.pdf). Unsurprisingly, certifications of newly hired employees have increased since their inception in 1996 (WOTC) and 1997 (WtW), totaling more than 630,000 in fiscal year 2005.

Little attention has been paid to the fact that an increasing fraction of workers being certified for the WOTC or WtW are actually being certified through work at THS firms. In contrast to the legislative intent to subsidize long-term positions, most THS firms are qualified to claim the WOTC and WtW subsidies despite their role as labor market intermediaries rather than end-user employers. We investigate how worker outcomes in subsidy-certified THS jobs are affected by both the subsidies and the temporary nature of employment. An empirical investigation is particularly important in this case since economic theory does not offer any clear prescriptions about whether THS employment should help these workers; THS firms drive a wedge between the worker and his or her employer and have limited legal obligations to their workers (such as protection from unjust dismissal) relative to traditional employers. More specifically, we focus on two key questions. First, do THS firms' claims of these subsidies affect the labor market outcomes of their workers? Second, among disadvantaged workers in subsidized firms, how do the labor market outcomes of THS workers compare to those of workers in traditional jobs? The answers to these questions may help policymakers determine whether continuing to subsidize temporary jobs is an effective strategy for improving employment outcomes of the disadvantaged.

There are several interesting parallels between the WOTC/WtW programs and the emergence of THS firms as a growing source of employment for public assistance recipients. For example, proponents of both would argue that they make employers more willing to hire workers who they might not otherwise consider by reducing costs. In this way, both subsidies and temporary work may allow workers to get a foothold in the labor market, an effect that may be magnified when THS firms themselves are subsidized. However, critics have suggested that both employer subsidy programs and THS firms could contribute to worker "churning," that is, the dismissal of workers whose subsidies run out or the perpetuation of unstable, short-term work assignments. Thus far, there is no evidence of churning as a result of the WOTC and WtW programs (U.S. General Accounting Office 2001). (3) And although a number of studies of temporary workers have shown that they are more likely than other workers to change employers, to work fewer and more variable hours, and to have less attachment to the labor force, findings on the longer-term implications of temporary employment for disadvantaged workers' labor market outcomes are more mixed. (4) In light of the disparate findings, concerns persist that these workers may not be obtaining jobs that promote labor market attachment and retention, while THS firms are claiming federal tax credits designed for that purpose.

This is the first study, to our knowledge, to examine the distinct effects of THS employment and WOTC/WtW participation on the labor market outcomes of WOTC/WtW-certified THS workers. (5) We use uniquely available administrative data on workers affected by the WOTC/WtW and THS employment to investigate these issues along with a new survey of THS firms' WOTC/WtW awareness and utilization. Because national statistics do not distinguish WOTC/WtW approvals (or "certifications") by job type, we compiled administrative microdata available to us from Wisconsin for our empirical analysis. (6) We found that among employers in Wisconsin, three of the six firms with the largest number of applications were THS employers. WOTC/WtW applications (and certifications) from THS firms have increased steadily over time, expanding from fewer than 16% of all applications in 1999 to 26% in 2002. (7)

We approach the empirical estimation by comparing labor market outcomes for a sample of WOTC/WtW-certified THS workers (the "treatment group") to two comparison groups: workers in the THS industry who are not WOTC/WtW certified (but are WOTC eligible) and workers who are WOTC/WtW certified but do not work in the THS industry. These groups are assembled using three sources of administrative data on the population of public assistance recipients in Wisconsin over several years. Our analysis of treatment effects uses econometric matching estimation to isolate the effects of WOTC/WtW and THS employment on labor market outcomes by controlling for observable factors and also addressing individuals' selection into THS or subsidized employment. We first examine the effects of the subsidy programs on workers in the THS sector. We hypothesize that if, in fact, the WOTC/WtW programs are working as intended, earnings and labor market attachment will be greater among those THS workers for whom tax credits are claimed than for those whose employers are not participating, all else equal. Second, we examine the effects of THS employment versus traditional employment among WOTC/WtW-certified workers. We hypothesize that earnings will be lower and/or that tenure will be shorter for workers in THS jobs. However, we describe how this relationship may be different in the context of subsidized employment, and, thus, this investigation represents an important new contribution to both the temporary employment and the tax credit literature.

We find large differences in labor earnings and job tenure across the three groups, some of which reflect significant (and previously undocumented) differences in worker and employer characteristics. Among THS workers, those who are WOTC/WtW certified have quite similar job tenure to eligible unsubsidized workers, but the WOTC/WtW-certified workers have much higher earnings per quarter employed; this holds even when we control for observable characteristics. This suggests that some of the WOTC/WtW funds pass through to the THS worker in the form of increased earnings per quarter, just as they do (though to a far lesser extent) for workers in other industries (Hamersma 2008). Using panel estimation to examine longer-term outcomes, however, the earnings difference between these workers is no longer statistically significant. Including all WOTC/WtW-certified workers, we find that THS workers have much lower total earnings than non-THS workers because of much shorter average job tenure, and again, this relationship holds when we control for observable characteristics. Fortunately for these workers, panel estimates show no evidence of continued lower earnings over time.

In light of the limited information available to us to explain firms' selection into the WOTC/WtW, we added a second component to our study--a telephone survey of Wisconsin THS firms--to further investigate their awareness and use of the WOTC and WtW tax credits. Among the 101 firms that completed surveys, more than 70% were aware of the tax credits, and of these, 60% reported that they currently claim the tax credits. Contrary to legislative intent, however, only one firm reported that prospective employees' eligibility for the tax credits might affect their hiring decisions. Thus, our findings suggest that concerns about the use of WOTC/ WtW funds to subsidize THS employment are likely warranted since subsidies do not appear to improve the job outcomes of disadvantaged workers in these jobs or encourage the hiring of additional disadvantaged workers as intended.

2. The Characteristics of the WOTC/WtW and THS Employment