This paper presents an analysis of workers who persistently have low earnings in the labor market over a period of three or more years. Some of these workers manage to escape from this low-earning status over subsequent years, while many do not. Using data from the Longitudinal Employer Household Dynamics (LEHD) program at the U.S. Census Bureau, we analyze the characteristics of persons and especially of their firms and jobs that enable some to improve their earnings status over time.
Overall, the main results of this analysis are as follows:
- A significant fraction (about 12%) of prime-age adults in the United States with regular labor force attachment have very low earnings (i.e., $12,000 per year or less) that persist over a period of at least three years;
- These low earnings are associated both with their own demographic characteristics (i.e., race/gender and where they were born) and many characteristics of the firms for which they work (i.e., industry, size, turnover and net employment growth rates, and firm wage premia);
- Of those with persistently low earnings, nearly half manage to escape this status in subsequent years, though earnings improve only partially for most of them (i.e., they continue to earn less than $15,000 in at least some years);
- Of those with persistently low earnings, white males enjoy the highest subsequent earnings gains and highest rates of "escape" from this status of any race/gender group, while blacks endure the lowest improvements;
- Job and industry changes are associated with large percentages of the observed improvements in earnings, though a significant fraction (i.e., roughly a fourth to a third) of all escapes from low-earning status also occur among those who stay on initial jobs;
- Most earnings improvements for low-earning women occur within the service sector in areas such as financial services, health care and education while a larger fraction of those for males occur in the "traditional industries" like construction, manufacturing, transportation and wholesale trade. Their escape rates, which are similar to most groups of men, illustrate the key point that service sector work can also provide a springboard to better earnings and employment outcomes;
- Significant parts of the lower subsequent earnings of black and other (mostly Hispanic) males among initial lower earners are accounted for by their lesser access than white men to high-quality jobs;
- Improvements in earnings for workers with successful job changes are due not only to better characteristics of the new firms for which they work, but also higher rates of return to the workers' tenure and experience accumulated on the new job. This leads to both improvements in current earnings as well as increased wage growth over time; and
- Low-wage workers who start out in a temporary help agency certainly earn lower pay during the time they are employed by the agency, but job change for these workers is associated with higher subsequent wages and better job characteristics than is job change for other, similar, workers.
These findings have some important implications for the low-wage labor market. For one thing, some degree of upward mobility for persistently low earners is certainly possible, and in fact is being achieved even if these improvements remain fairly modest in most cases. Also, there is no single path for achieving earnings growth. Job changes are important to many who achieve earnings improvements, though staying on the job also works in a significant percentage of cases.
A range of characteristics also seems to be associated with these good jobs including not only firm wage premia (the mark-up that some firms pay workers, particularly in unionized and highly concentrated industries) but also industry, firm size, rates of turnover and employment growth (which are observable). The findings suggest trying to place low earners into high-wage sectors, firms with low turnover, and larger firms that provide job ladders and possibilities of upward mobility.
The positive results found for temp agencies suggest that these or other types of labor market intermediaries assist low earners in making the transition to better job opportunities. The overall results also suggest a strong need to improve access to good jobs for many low earners especially those who are not white males.
The paper's analysis is subject to a variety of limitations, such as selection issues and unobservable characteristics of workers and jobs. But we manage to mitigate some of these concerns with controls for person as well as firm fixed effects. Were data available on educational outcomes, hourly wages, and family/household structure (such as spouse's earnings and presence of young children), we could distinguish between the persistently low earners who might choose such a status voluntarily as opposed to those who face very constrained labor market opportunities.