Previous research on temporary employment found that it provided important on the job training that increased skill levels and allowed better employment outcomes. But a new study contradicts this work and finds that it is mainly better matching, not enhanced skills, that explains why temporary workers tend to do better in the labor market. This implies that there are informational or other problems preventing fully efficient labor market outcomes, and that there is room for government agencies or private sector firms to improve outcomes by acting as intermediaries between workers and firms:
Do temporary jobs improve workers long-term labour market performance?, by Fredrik Andersson, Harry J. Holzer, and Julia Lane, Vox EU: What are the long term labour market consequences of temporary help work? As the importance of temporary help work has increased across OECD countries, economists have started paying attention to this question.
If, on the one hand, temp agencies provide a productive stepping stone on the path to more stable employment, both by increasing access to “good” jobs and by imparting useful job skills, workforce development agencies should actively promote temp jobs. If, on the other hand, they are actually part of a “secondary” labour market in which low-wage workers churn from bad job to bad job, workforce development agencies should avoid placing workers in temp help agencies, and concentrate their efforts on more successful strategies. The importance of answering this question can be seen from a glance at the table. Across the OECD, about a quarter of young workers, and more than one in eight workers with low education are in temp help employment, and for some countries, like Spain, the proportions reach as high as two out of three and more than one in three, respectively.
One strand of the literature has suggested that temp help agencies can have an impact by increasing the level of workers job skills by providing on-the-job training. But a growing related literature suggests that temporary help agencies might help various groups of less-skilled workers get access to firms that offer “good” jobs with stable employment and higher-wage jobs by overcoming geographic and information gaps. This finding is based on important empirical research which shows that the characteristics of firms and jobs to which workers are matched, independently of worker skills, affect the labour market outcomes of workers in general. In other words, this research suggests that one way in which temp help could have an impact on low-wage workers is because it helps workers find better jobs, rather than adding training or skills.
However, as with most questions in labour economics, the most difficult challenge is identifying an appropriate counterfactual, so that workers working for temp agencies can be correctly compared to those who don’t. This requires developing approaches that address the important selection issues that are inherent in any such analysis. That is, researchers must be careful to rule out a confusion of correlation and causality. It might be that workers who do temp work have certain characteristics that are unobservable to the research but which make it (a) more likely both that they seek temp employment, and (b) subsequently do well in the labour market. In this case, the above average performance of temp workers is caused by the unobserved characteristics, not the temp work itself.
Initial empirical research based on both survey and administrative data supported the notion that temp agencies were providing pathways to more stable employment. Lane et al. (2003) applied ‘matched propensity score’ techniques (an econometric approach aimed at getting around the selection effect) to data from the Survey of Income and Program Participation. That paper concluded that spells in temp agency employment improved labour market outcomes relative to spells of unemployment. Heinrich, Mueser and Troske (2005) came to similar conclusions. In a more broad-ranging study, Andersson, Holzer and Lane (2005) used linked employer-employee data and found that low earners who were employed by temp agencies did better than those who were not. Further, they found that the positive impacts of earlier temp employment were largely accounted for by the characteristics of the firms in which they were subsequently employed: consistent with the notion that temp agencies provided better access to other higher-wage firms. More recent work continues to show positive effects. For instance, Benner et al. (2007) examined survey data on employers and workers in Milwaukee and Silicon Valley and found that workers who used temp agencies to find employment had higher earnings in subsequent jobs. A large number of European studies have similar positive findings (see Ichino et al., 2006, for a review).
But an important new study contradicts these findings. It gets around the selection problem much more directly, namely by looking at the long-term labour market performance of a group of workers, some of which were randomly assigned to temp work while others were not. The group were part of an innovative reform in US welfare programmes adopted in 1996 – the so-called Temporary Assistance to Needy Families program, or TANF for short. TANF requires that those who are receiving welfare assistance either work in the private economy or engage in constructive activities leading to employment. To ensure that this expectation is real, the current law specifies a minimum proportion of the overall welfare caseload that must be actively participating in constructive activities.
In the only study to date that has used random assignment of TANF recipients to temporary help agencies, Autor and Houseman (2005, 2007) found that temp agencies increased the short-term earnings for workers; but their longer-term employment was characterised by lower earnings, less frequent employment and higher welfare recidivism.
The critical difference between this study and earlier studies is the random trial, since all the other studies relied on econometric techniques to identify the appropriate comparison groups, and concerns remain about selection on variables (including time-varying characteristics in studies that control for person-fixed effects) that are unobservable to the econometrician.
Of course, questions remain about the external validity of the Autor and Houseman results. They are based only on TANF recipients – a group that is clearly not a random sample of workers, not even of low-wage workers. Moreover Autor and Houseman only use data from only one agency, the “Work First” agency, in only one city (Detroit). Furthermore, they also found that other intermediary agencies that generated longer-term placements had more lasting positive effects on employment outcomes.
A recent NBER conference on Labour Market Intermediation was organised by David Autor and held in Cambridge May 17-18, 2007 with the aim of bringing together many of the researchers interested in the topic and probing further on the selection and comparison group issues that need to be addressed to answer the policy challenge.
New work presented at that conference, by Andersson, Holzer and Lane, was more nuanced than their earlier results. It showed that temp earners clearly have lower earnings than others while working at these agencies. Even their subsequent earnings are somewhat mixed. However, it confirmed their earlier findings that these earnings are generally higher if they manage to gain stable employment with other employers. They found that there is some self-selection among low earners into temp employment. Even controlling for individual characteristics and job tenure, however, the positive wage effects from temp employment seem mostly to occur because those working for temp agencies subsequently work for higher-wage firms than do comparable low earners who do not work for temps. And the positive effects we estimate seem to persist over time, for as much as six years beyond the base period during which the temp employment was observed.
These results are consistent with the notion that low earners, in addition to any deficiencies in skills that they bring to the labour market, sometimes have difficulty “matching” themselves to higher-wage employers in the labour market.
For policy purposes, the findings confirm the merits of using “labour market intermediaries” to improve the quality of job placements for less-skilled workers. But they also imply that incentives should be provided to these intermediaries – whether temp agencies or others – so that they would be rewarded for creating longer-term opportunities for these workers.
Andersson, F., Harry Holzer, and Julia Lane (2005). Moving Up or Moving On: Workers, Firms and Advancement in the Low-Wage Labour Market. New York, Russell Sage.
Autor, D., and Susan Houseman (2007). Temporary Agency Employment: A Way Out of Poverty? Working and Poor: How Economic and Policy Changes are Affecting Low-Wage Workers. S. D. a. R. S. R. Blank. New York, Russell Sage.
Autor, D. a. S. H. (2005). Do Temporary Help Jobs Improve Labour Market Outcomes for Low-Skilled Workers? Evidence from Random Assignments. NBER Working Paper, NBER.
Benner, C., Laura Leete, and Manuel Pastor (2007). Staircases or Treadmills? Labour Market Intermediaries and Economic Opportunity in a Changing Economy. New York, Russell Sage.
Heinrich, C. J., Peter R. Mueser, and Kenneth R. Troske (2005). "Welfare to Temporary Work: Implications for Labour Market Outcomes." Review of Economics and Statistics 87(1): 154–173.
Ichino, A., Fabrizio Mealli, and Tommasso Nannicini (2006). From Temporary Help Jobs to Permanent Employment: What Can We Learn from Matching Estimators and their Sensitivity? IZA Discussion Paper. Bonn, IZA.
Lane, J., Kelly Mikelson, Patrick Sharkey, and Douglas Wissoker (2003). "Pathways to Work for Low Income Workers: The Effect of Work in the Temporary Help Industry." Journal of Policy Analysis and Management 22(4): 581-598.