Job Separation Behavior of Welfare Recipients: Results from a Unique Case Study

Abstract

Working Paper 2004-12 May 2004 Abstract: This paper uses a unique personnel data set to explore job separation behavior among welfare hires. Our results indicate that welfare hires are no less stable than similar nonwelfare hires; however, time until separation does differ across welfare status by reason for separation. We also found that the presence of a mentoring program will increase time until separation for both welfare and nonwelfare hires. JEL classification: I38, J24, J64 Key words: TANF, welfare, job search, job separation I. Introduction and Background In 1996 the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) became law, eliminating the Aid to Families with Dependent Children (AFDC) program. (1) In place of AFDC, states receive a block grant to establish a Temporary Assistance for Needy Families (TANF) program. New rules under the TANF program impose stricter time limits and work requirements. Once a TANF recipient is determined to be 'job ready' or she has received assistance for 24 months, she is required to work. (2) Under AFDC, relatively few recipients were required to work. For example, the Job Opportunity and Basic Skills (JOBS) program, which TANF also replaced, required only about 10% of the AFDC caseload to participate. But under TANF, there is pressure on the states to increase their work participation rates in order to receive their full block grant amount. By 1999, states must have had at least 35% of their mandatory TANF caseload participating in work activities for at least 25 hours (for all families) and 90% of their two-parent TANF families working 35 hours or more. Due to the changes brought about by the PRWORA, an increasing number of individuals are required to find a job, some entering the labor force for the first time. Many of these job seekers have a low skill level, little work experience, and other barriers to employment, such as transportation and child-care needs. (3) Most of the literature examining questions related to the transition from welfare to work focuses on employment barriers of former welfare recipients and the availability of jobs for former welfare recipients. Few studies have examined the turnover behavior of former welfare recipients. (4) The fear of many employers is that welfare recipients are less stable (higher turnover) workers than non-welfare recipients. And, high turnover is, of course, costly because of personnel processing, orientation, training, etc. This paper utilizes unique, firm-level post-PRWORA data to examine the separation behavior of welfare and non-welfare hires and the effectiveness of a firm's work programs on reducing turnover. The following questions are addressed: (1) Is the separation behavior of former welfare recipients, in fact, different from other workers? If so, how and why does it differ? (2) What characteristics of a firm's work program contribute to higher retention rates? (3) Does the time to separation vary by welfare status for different reasons for separating? In other words, do welfare hires separate from their employer sooner than non-welfare hires for personal versus professional reasons? 5 The answers to these questions have implications for firms' hiring policies and the potential effectiveness of work programs. By identifying factors contributing to higher retention rates, especially factors related to the type of work programs in place, more effective work programs can be designed. After all, the determination of a successful welfare-to-work transition does not conclude merely when the individual finds employment. Most likely due to the relatively new public policy interest in hiring welfare recipients, there are few studies which empirically examine the separation behavior of welfare hires. Whenever examined, turnover among welfare hires has uniformly been found to be high. For example, Berg et al. (1992) find that 46 percent of recent welfare recipients lost or quit their job in the first three months of employment and 73 percent have lost or quit their job within one year. …