In our post earlier this month, Louise Fox and I explored the question of the treatment effects of supply-side interventions on employment outcomes when the setting is one of seemingly weak labor demand. Do programs create new jobs or rather crowd-out others in the labor market? One approach to get at this question is to run a double-sided experiment involving both workers and firms with random assignment of both to treatment and control groups.
In their forthcoming ReStud paper (earlier working paper version), Crépon and Premand indeed do this in their randomized experiment in Côte d’Ivoire and aim to tackle our question. They study the impact of subsidized dual-apprenticeships (formal center-based training with on-the-job training), to examine whether & which market failures are at play in this labor market. Is it that youth under-invest in apprenticeship due to financial constraints (and limiting supply), or firms under-invest in training for arguably standard economic reasons, or firms under-hire apprentices due to information asymmetries or imperfect intermediation? As they describe in their paper “Our double-sided randomized design is tailored to estimate the direct impacts of subsidized dual apprenticeships on youth participation and the indirect crowding-out effect in firms, which allows us to estimate the net effect on the number of filled apprenticeship positions.” Yay!
Spoiler alert on findings: They find low youth demand hindering apprenticeship expansion and, thus, firms facing high recruitment costs and hiring frictions. Treated youth are more likely to enter into apprenticeships and, after completion, they have higher earning (but not through wage employment, suggesting an increase in productivity in informal self-employment). Lastly, to the question Louise and I had posed, they find limited crowding out, with the intervention resulting in a net increase in apprenticeships in firms (that is, more jobs being created).
The intervention: The program they study is run by the government of Côte d’Ivoire. The main features on the formal apprentice offer to treated youth are:
- Subsidies: about US$54/month for 12 or 24 months (varying by occupation) to cover meals and transportation costs, insurance coverage, and work equipment.
- Technical, theoretical courses at vocational centers to complement the on-the-job training (about 180 hours per year).
- Contracts signed by the apprentice.
- Certification.
On the firm side, treated firms are matched to a formal apprentice, and:
- get a small toolkit of materials to facilitate practical learning.
- commit to not request payment of tuition fees at the start of the apprenticeship, in contrast to the traditional apprenticeship model in West Africa (half of the firms in this setting report charging apprentices a fee)
The study was implemented in seven urban localities, which were stratified by ‘micromarkets’ (trades in a given locality). Firms interested in hosting program apprentices were listed by trade^ (resulting in 731 firms). These are micro/small, informal businesses (84% with no formal legal status; 68% lack book-keeping). Including owner, firms have 6.4 permanent employees, just over half are traditional apprentices.
Interested and eligible youth were also listed by trade (resulting in 1,842 youth; more than the number off firms since most firms offered more than one apprenticeship position). (Standard blog disclaimer: see the paper for a lot more details). The criteria for matching youth to firms included the distance between the firm and residence of the youth.
Spill-over effects? One concern discussed at length in the paper is the potential for spill-over effects driven by changes to the tightness of the apprenticeship market (i.e., the chance that firms and potential apprentices match). Specifically, this intervention could change the matching opportunities between youth and firms in the control group – raising questions about the validity of the control group. The paper takes this issue seriously, presenting a conceptual framework in Section 3 and estimates in Section 5. Bottom line (and a gross oversimplification): this is a relatively small experiment in these urban labor markets and the scope for spillovers seems limited (they show this in their estimates). The net increase in hiring of apprentices in treated firms does not reflect a decline in hiring among control firms. But as the scale of such a program grows, this will be a more relevant concern. Where they do find some limited crowding-out effect, it is in localities with a higher share of treated youth or firms.
Crowding-in jobs? At midline (20-months into the program), treated firms have 0.6 more apprentices than control firms. The program is not crowding-out jobs. The greater hiring is mostly driven by more formal apprentices (0.8) and slightly fewer traditional apprentices (-0.2) compared to control firms. Measuring the aggregate numbers since the start of the intervention shows that treated firms have had 1.1 more apprentices compared to control firms (1.4 formal and -0.3 traditional). But as noted above, not all workers in these firms are apprentices. They do not find a difference in the number of full-time non-apprentice workers (0.030 and not significant) between treated and control firms. These are ITT estimates.
…So what is keeping firms from hiring more apprentices? Crépon and Premand find that traditional and formal apprentices make similar total contributions to firm activities (the latter get more training but have higher absenteeism and work fewer days). They conclude that firms are constrained in hiring apprentices by the limited supply of youth rather than the firms’ ability to train or employ them in relatively profitable tasks.
…So it is difficult for firms to find apprentices. But why? Well, as the paper notes and is well documented, these positions pay very little (or sometimes youth have to pay to work for almost nothing). But why? Seemingly because these firms have very low productivity (among other reasons, due to the low skill level of youth). So in some sense, the root causes of the dilemma of creating jobs is the low productivity firms (and, in this setting, at least partly offset by providing firms a better-trained workforce).
Development Impact is now on summer break, and will resume after Labor Day. We wish you all a wonderful August!
^ I have complained about studies that don’t give the reader details about the trades of businesses/jobs (see this old blog). I am happy to report that this paper tells us (in a footnote) the most popular trades are: car or motor mechanic (21% of positions), metalworker, boilermaker, welder (14%), bricklayer, painter, plumber (11%), carpenter (9%), car electrician (9%), electrician (8%), coach-builder (8%), repairman of fridges and freezers (7%).
Join the Conversation