According to a recent report by the International Labour Organization, the number of temporary agency workers has been increasing in many countries (Countouris et al. 2016). While temporary employment is often associated with lower pay and lower income security, labour economists argue that temporary work agencies have the potential to improve job-matching efficiency and reduce unemployment (Katz et al. 1999, Houseman et al. 2003).
In particular, it is said that temporary work can serve as a stepping stone to regular employment. However, there is a debate as to what extent this stepping-stone effect is present. A study by Autor and Houseman (2010) finds that jobs mediated through a temporary work programme in the US do not improve, and may even be detrimental to, long-run labour market outcomes.
Temporary work agencies are frequently used nowadays by unemployment benefits administrations to help unemployed workers return to the labour market. Direct matching of unemployed job seekers to vacancies is often found to be one of the most effective ways to reduce benefits dependency (Bolhaar et al. 2018). Temporary work agencies can complement the role of public employment services because they often have access to a different pool of vacancies and employers.
To intensify cooperation with temporary work agencies, the Dutch unemployment insurance administration has introduced ‘labour market speed dates’ – events where unemployed job seekers meet different temporary work agencies. Participants spend around two hours at these events to have short talks with, on average, four agencies. The size of the speed dates varies from small, sector-specific events where only about ten job seekers are present to more general speed dates with 300 to 400 attending job seekers.
Organising a speed date event entails very low costs to the public employment service. Compared to, for example, intensive counselling of job seekers by caseworkers, labour market speed dates may therefore be a more cost-effective instrument to reduce benefits dependency than, for example, counselling.
Evaluating the labour market speed dates
To evaluate the effectiveness of labour market speed dates, we conducted a randomised experiment in which only a random subsample of more than 12,000 unemployed benefit recipients are invited to attend such an event (Van der Klaauw and Ziegler 2019). In total, our experiment covers 18 speed date events. Job seekers who do not get invited never participate and serve as a control group, while 24% of the invited job seekers attend the speed date event.1
Our empirical analysis reveals that participants are 6 to 7 percentage points more likely to work in the month after a speed date. Figure 1 illustrates the corresponding increase in the number of monthly working days for treated job seekers. A large part of this increase is driven by work at temporary employment agencies, but participants are also more likely to find regular employment.
Figure 1 Speed-date participation’s effect on days of employment
Source: Van der Klaauw and Ziegler (2019).
Because the speed dates are able to reduce the number of benefit recipients, the programme is highly cost-effective for the unemployment insurance administration. However, the positive employment effects disappear again in the following weeks. Participants also tend to have slightly lower wages in the longer run. This implies that unemployed job seekers, on average, experience an earnings loss when they participate in a speed date event and that jobs mediated by participating agencies are no stepping stone towards regular employment.
Using additional survey evidence, we find that labour market speed dates also affect the search behaviour of job seekers. Participants claim to be more motivated and report lower wages for which they are willing to accept work. The latter finding suggests that talking to temporary work agencies causes unemployed job seekers to lower their expectations about future jobs. The agencies can point them to vacancies which they were not aware of, but they also convey a more realistic idea of potential earnings at a new job. This is consistent with evidence from Spinnewijn (2015), who shows that recently unemployed workers are too optimistic about their job-finding probabilities.
Active labour market programmes are considered an important element of unemployment insurance. If these programmes reduce moral hazard and speed up job finding, unemployment benefits can be relatively generous without stimulating inactivity. However, active labour market programmes are often expensive and empirical evaluations show mixed results on their effectiveness (Card et al. 2018). In many countries, the recent economic crisis increased the number of unemployment benefits recipients while government budgets were cut. This forced public employment services to adopt less expensive programmes that can serve larger groups of benefits recipients. Labour market speed dates have these properties since the participating temporary work agencies are not remunerated and the costs of organisation are very modest.
For unemployed job seekers, temporary work is an important channel to return to the labour market. While it is difficult for many job seekers to find regular employment, flexible work arrangements are often more readily available. In the Netherlands, about one-third of exits from unemployment insurance are to jobs provided by temporary work agencies. Speed date events with such agencies might thus have a considerable impact on the labour market.
Compared to unemployed job seekers, workers in a temporary job are less stigmatised, obtain additional work experience, and are less likely to lose skills. In theory, this should make it easier to find subsequent employment. However, our empirical results show that for jobs provided by temporary work agencies, such stepping-stone effects are absent. In the short run, the agencies are able to mediate jobs to unemployed workers but this increase in job finding does not have long-run consequences.
Labour market speed dates with temporary work agencies should thus be considered as an effective programme to reduce the costs of unemployment insurance, but not as a programme to stimulate labour market prospects of participants in the long run.
Autor, D H, and S N Houseman (2010), “Do temporary-help jobs improve labour market outcomes for low-skilled workers? Evidence from ’work first’”, American Economic Journal: Applied Economics 2(3):96–128.
Bolhaar, J, N Ketel and B van der Klaauw (2018), “Caseworker’s discretion and the effectiveness of welfare-to-work programs”, CEPR Discussion Paper 13047.
Card, D, J Kluve and A Weber (2018), “What works? A meta analysis of recent active labour market program evaluations”, Journal of the European Economic Association 16(3), 894–931.
Countouris, N, M Freedland, S Deakin, A Koukiadaki and J Prassl (2016), Report on temporary employment agencies and temporary agency work – A comparative analysis of the law on temporary work agencies and the social and economic implications of temporary work in 13 European countries, International Labour Organization.
Houseman, S N, A L Kalleberg and G A Erickcek (2003), “The role of temporary agency employment in tight labour markets”, Industrial and Labor Relations Review 57(1):105–127.
Katz, L F, A B Krueger, G Burtless and W T Dickens (1999), “The high-pressure US labour market of the 1990s”, Brookings Papers on Economic Activity 1999(1):1–87.
Spinnewijn, J (2015), “Unemployed but optimistic: Optimal insurance designs with biased beliefs”, Journal of the European Economic Association 13(1): 130–167.
Van der Klaauw, B, and L Ziegler (2019), “A field experiment on labour market speeddates for unemployed workers”, CEPR Discussion Paper 13516.
 Instrumental variable estimation is used to estimate the average treatment effect on the treated (ATET).