Student Access to Apprenticeships: Evidence from a Vignette Experiment

• Author: Ilse Tobback,Dieter Verhaest,Stijn Baert
• Published date: 01 July 2020
• DOI: http://doi.org/10.1111/irel.12258
• Date: 01 July 2020

Student Access to Apprenticeships: Evidence

from a Vignette Experiment

ILSE TOBBACK , DIETER VERHAEST and STIJN BAERT*

We identify the causal effects of student characteristics on the likelihood of being

hired for an apprenticeship and explore the mechanisms underlying the employ-

er’s decision. To this end, we perform a vignette experiment among human

resources professionals in Belgium, focusing on less-qualified youth. Our results

indicate that students with favorable educational records and students revealing

being motivated are more likely to obtain an apprenticeship. Furthermore, we find

that these characteristics are used by human resources professionals as signals of

trainability, employability, and quit intentions.

Introduction

The great recession has fostered a renewed interest in dual-learning pro-

grams that combine classroom-based learning with an apprenticeship in a firm

or organization (Cahuc et al. 2013; Eichhorst et al. 2015). While many coun-

tries saw their already high youth unemployment rates soar in the aftermath of

the financial crisis, the relatively low youth unemployment rates of traditional

dual-apprenticeship countries such as Germany,

1

Austria, and Denmark were

barely hit (Bell and Blanchflower 2011). This anecdotal evidence on the role

of dual-learning programs in tackling youth unemployment is complemented

with several micro-econometric studies that indicate that young workers with

apprenticeship experience face higher employment chances at the start of their

careers in comparison to those without apprenticeship experience (Fersterer,

JEL codes: I24, J24.

*The authors’affiliations are, respectively, KU Leuven, Brussels, Belgium. E-mail: Ilse.Tobback@kuleu-

ven.be; KU Leuven, Ghent University, and GLO, Brussels, Belgium. E-mail: Dieter.Verhaest@kuleuven.be;

Ghent University, University of Antwerp, Universit

e catholique de Louvain, IZA, GLO, and IMISCOE,

Ghent, Belgium. E-mail: Stijn.Baert@UGent.be. The authors thank Eva Van Belle for her valuable sugges-

tions and Sander De Boeck for his research assistance.

1

For more information as to why we might observe the apparent resilience of the German training sys-

tem, we refer to Busemeyer et al. (2012).

INDUSTRIAL RELATIONS, DOI: 10.1111/irel.12258. Vol. 59, No. 3 (July 2020). ©2020 Regents of the

Universit y of Calif ornia. Published by Wiley Periodicals, LLC., 350 Main Street, Malden, MA 02148,

USA, and 9600 Garsington Road, Oxford, OX4 2DQ, UK.

435

Pischke, and Winter-Ebmer 2008; Parey 2016; Riphahn and Zibrowius 2016;

Sollogoub and Ulrich 1999).

2

While this success in easing school-to-work transitions is well established,

dual educational systems face several challenges. In particular, because the

system crucially depends on the willingness of employers to provide training

places, not all youth may gain access to an apprenticeship position (Cahuc

et al. 2013). A small number of descriptive studies have indicated that individ-

uals encountering difficulties in gaining access to regular employment, such as

individuals with inferior educational records in terms of grade point average

(GPA) and grade retention or ethnic minorities, also fail to begin apprentice-

ships (Helland and Støren 2006; Hupka-Brunner, Sacchi, and Stalder 2010).

Put differently, many of the individuals for whom the problem of youth unem-

ployment is most acute, seem precluded from using apprenticeships as a step-

ping stone toward regular employment. However, given the non-experimental

nature of these studies on access to apprenticeships, this conclusion is poten-

tially biased. Not only may employers select on characteristics that are unob-

servable to the researcher but correlated with observables, such as one’s

minority status or past educational record, but these observable factors may

also be correlated with supply-side factors such as job choice and job search

intensity.

In this article, we report the results of a study that accounts for these prob-

lems by running a survey experiment among 276 human-resources (HR) pro-

fessionals in Belgium. These professionals are each asked to evaluate six

randomized table vignettes that describe apprenticeship applicants that differ

with respect to their past educational records (educational track, GPA, and

grade retention), their personal characteristics (gender and ethnic ancestry), and

their knowledge of the company or organization, allowing us to assess whether

these characteristics affect the likelihood of being hired for an apprenticeship

and subsequent regular employment in a causal way.

To the best of our knowledge, only two other studies have already investi-

gated access to apprenticeships in a causal way.

3

Relying on similar survey

2

While this evidence pertains to employment chances at the start of the career, the evidence on starting

wages (Parey 2016; Sollogoub and Ulrich 1999) or on employment chances at the end of the career (Hampf

and Woessman 2017; Hanushek et al. 2017) are more mixed. In general, similar conclusions on the impor-

tance of vocational skills and workplace-based learning for initial labor-market outcomes emerge when voca-

tionally educated individuals (with or without apprenticeship) are compared to more generally educated

individuals (Verhaest, Lavrijsen, et al. 2018; Vogtenhuber 2014). For an overview of the literature on the

labor-market effects of apprenticeships, we refer to Ryan (2001). Some of the more recent studies are

reviewed in Lerman (2017).

3

Many of the considered factors have been investigated more extensively in the context of the regular

labor market. However, the context of the apprenticeship market is largely different, with strict regulations

on aspects like the provision of training, apprentice pay, or the duration of the apprenticeship contract.

436 / ILSE TOBBACK,DIETER VERHAEST,AND STIJN BAERT