Fredrik Andersson, Julia Lane, Erika McEntarfer
This study examines the effect of employer characteristics, types of coworkers, and residential location in promoting the advancement of Temporary Assistance for Needy Families (TANF) recipients in the labor market. It is the first to use new, large integrated employer-employee data with new measures of worker quality and firm pay premia to examine labor market outcomes and provide evidence that proactive welfare agencies may use in decision making.
The first part of the study examines changes in the low-wage labor market and finds that employment opportunities expanded in the low-wage labor market between 1997 and 2001, but that the growth in opportunities has not been associated with greater access to good wages.
In particular, we find that low-wage employment is concentrated in a few industries, and opportunities in these industries have expanded between 1997 and 2001. In addition, although many low-wage workers work with high-wage workers, the opportunity to do so is declining. We also find that firms in low-wage industries pay workers with the same set of skills less than firms other industries. While job quality has increased over time, this trend has been less strong in heavily low-wage firms. In terms of the importance of neighborhood characteristics, we find that low-wage jobs do exist in low-wage neighborhoods, but there remains a mismatch between the location of low-wage workers and jobs. In sum, most low-wage workers are in industries, firms, and neighborhoods that have low job quality indices, and the changes over time have not been positive.
The second part of the study examines the effect of employer characteristics, types of coworkers, and residential location on TANF recipients' transition out of low-wage work and finds that the industries and firms for which workers work, the characteristics of their coworkers, and the neighborhoods in which they live all affect the likelihood of exit from low-wage status.
In particular, we find the most important individual characteristic determining the likelihood is worker quality. In addition, the quality of coworkers and the quality of the employer matter. Finally, living in a low-wage neighborhood strongly reduces the likelihood of transitioning out of low-wage status.
In sum, we find that jobs were created for low-wage workers between 1997 and 2001. While opportunities exist, however, the challenge for TANF agencies is to identify the job-creating firms that provide opportunities for low-wage workers, as well as coworkers that are primarily non-low-wage. This will maximize TANF recipients' chances of transitioning out of low-wage status. In addition, given the importance of residential location, TANF agencies might want to examine alternative transportation options.
With the tremendous success achieved by welfare reform in moving large numbers of former recipients into jobs, attention is now turning to their ability to advance in the labor market. While anecdotal evidence suggests that the employer is critical to this advancement, little hard evidence exists despite the need for such information by Temporary Assistance for Needy Families (TANF) agencies examining ways to improve labor market outcomes for disadvantaged workers. This study expands on intriguing preliminary evidence (see Andersson, et al., forthcoming(1)) by directly investigating the impact of firm characteristics such as industry, firm quality, and firm location on worker earnings, earnings growth, and transitions out of low-wage status for TANF recipients. In addition, it highlights the substantial dynamism of the low-wage labor market, indicating opportunities for TANF agencies not just to emphasize the quality of program placement but also to identify successful retention strategies.
This analysis is possible because the Longitudinal Employer-Household Dynamics (LEHD) program at the Census Bureau has developed a new database that can be used to answer a number of these questions. Because the LEHD data integrate employer and employee files, they provide a unique opportunity to analyze how employment outcomes are influenced by both worker and firm characteristics and the interaction of the two. The database integrates several administrative data files, including internal records at the Census Bureau, Unemployment Insurance (UI) wage records, and state ES202 establishment (company-based) records. This analysis includes data for eight states, accounting for 55 percent of the U.S. labor market, and provides quarterly longitudinal information on firms and workers for approximately 10 years from the early 1990s to 2001. A smaller subset of observations is also integrated with various national survey datasets such as the 2000 Decennial Census, the Survey of Income and Program Participation, and the Current Population Survey.
The analysis will include a combination of basic descriptive statistics as well as more complex multivariate modeling. The analysis will include an examination of the following set of questions:
- Where are the low-wage jobs?
- How has the labor market for low-wage workers changed over time in terms of the availability and quality of employment opportunities in the key industries, firms, and neighborhoods where low-wage jobs are concentrated?
- How are TANF recipients doing in the labor market relative to all low-wage workers, and how are they progressing?
- What are the factors associated with transitions out of low-wage status for TANF recipients?
- What factors are most predictive of job retention and wage advancement when simultaneously controlling for a set of worker, firm, and location characteristics?
The new LEHD data enabled us to explore these questions in detail. Questions 1 and 2 characterize the labor market where disadvantaged workers are likely to seek employment and are addressed in sections 1 through 5. This labor market is characterized in terms of the industries, firms, and neighborhoods where low-wage workers are likely to be concentrated. Question 3 examines how TANF recipients and other low-wage workers are doing in the labor market, as well as how they are progressing, and is addressed in section 6. Questions 4 and 5 analyze the factors affecting TANF recipients' and low-wage workers' success in the labor market and are explored in sections 6 and 7.
A primary goal of this project is to inform efforts to improve job retention and wage advancement by providing a better understanding of the factors that influence employment. One aspect of the labor market highlighted by these data is the dynamism of the U.S. economy an illustrative example of which is provided in table 1 below.(2) This table demonstrates several new facts. First, while in this particular state, the employment picture was quite negative over the time period, this varied markedly across age-groups: 19-21 year olds lost about 0.7 percent of its jobs; 35-44 year olds lost 0.9 percent; and 55-64 year olds lost more than 2 percent. Second, even though jobs were lost on net, there were still jobs being created and this creation rate varied dramatically across age groups again. For jobs among 19-21 year olds, 18 percent were new, compared with about 5 percent for the older cohort. Finally, hiring continued even during a slowdown in economic activity more than 43 percent of the youngest cohorts were in new jobs in the next year, as were 12 percent of 35-44 year olds, and 7 percent of 55-64 year olds.
|Net Job Change||-1,988||-12,004||-11,183|
These preliminary results suggest the intriguing possibility that the economy still creates jobs for workers, even in relatively dark economic times. The challenge for TANF agencies is to identify the job-creating firms that provide opportunities for low-wage workers and focus on tactics to teach workers to retain jobs once hired.
Note: This report is available in its entirety in the Portable Document Format (PDF).
1. Andersson, F., H. Holzer, and J. Lane. Forthcoming. Moving Up or Moving On: Workers, Firms and Advancement in the Low-Wage Labor Market. Thousand Oaks, Ca, forthcoming Sage Press.
2. The LEHD data provide information on all age-groups; the table provides a subset for clarity of exposition.
3. Details on how these statistics are calculated is provided in Abowd, Lengermann, and Vilhuber. 2002. "The Creation of the Employment Dynamics Estimates." LEHD Technical Working Paper TP2002-13. http://lehd.dsd.census.gov.