Horizontal mismatch and vocational education

ABSTRACT

Recent literature suggests that vocational education and training (VET) provides individuals with smoother transitions into the labour market but lower wages over the lifecycle, compared to general education. A possible mechanism explaining lower wages is horizontal mismatch, defined as a mismatch between the type of qualifications acquired by individuals and those required for their current job. Some studies have found higher mismatch wage penalties when individuals’ education is more specific, suggesting higher penalties for workers with VET. Therefore, we analyse horizontal mismatch in Switzerland, the country with the highest proportion of firm-based VET in the OECD. We use two measures from the Swiss Household Panel that cover different aspects of horizontal mismatch. While we find sizable mismatch wage penalties in OLS estimations, effects are small or insignificant in fixed-effects regressions. This holds for workers with vocational and general education background alike. We conclude that VET is more transferable than often assumed. We finish with recommendations on concept and methods for future analyses of horizontal mismatch.

KEYWORDS:

• Horizontal mismatch
• qualification
• training
• vocational education
• wages

JEL CLASSIFICATION:

• I21
• J24
• J31
• J62

Acknowledgments

The authors thank two anonymous referees, Michael Gerfin, Joop Hartog, Andreas Kuhn, Cain Polidano and congress participants at BIBB Bonn, IWAEE Cantanzaro, IAB/ZEW Nuremberg and SFIVET Bern for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data of this article can be accessed here.

Notes

¹ Germany has concluded memoranda of understanding with several EU member states with the aim of introducing dual apprenticeships, among them Greece, Italy, and Portugal.

² They mention the coefficient of −.036 as insignificant at the 1% level, but it is significant at the 5% level in Model 2 of Table 4.

³ Most students who attended Gymnasium (3A) go on to university and join group 5A.

⁴ This study has been realized using data collected by the Swiss Household Panel, which is based at the Swiss Center of Expertise in the Social Sciences (FORS). The SHP project is financed by the Swiss National Science Foundation.

⁵ Excluding self-employed may underestimate the labour-market opportunities for individuals with a 5A degree, but even more so those with a 5B degree, since the Swiss PET system (5B) offers many programs which prepare for self-employment, typically in the crafts sector.

⁶ See Bütikofer (2013) for an analysis of female labour supply estimated with SHP data.

⁷ This concerns 309 men in the full sample and 115 in the bio subsample. We exclude them as this group is likely to be highly selective. Including them in the estimations does not change the results substantially.

⁸ In Switzerland, wages below 2,000 CHF a month are not credible for a full-time job. We consider wages below 24,000 and above 300,000 as outliers. Trimming eliminates 388 person-year observations due to the lower bound and 277 person-year observations due to the upper bound requirement.

⁹ Pecoraro (2016) uses this variable in an ORU-type analysis of over-education.

¹⁰ In the pooled sample, ‘81.1%’ is a shorthand formulation for ‘81.1% of all person-year observations’. The text is explicit if we consider individuals instead of person-year observations, which include multiple observations for most individuals.

¹¹ Note that we use differences in 2-digit ISCO code. Using 3-digit ISCO differences, the proportion of the mismatched increases to 56.0%. Using 1-digit ISCO differences, the proportion of mismatched decreases to 42.7%.

¹² Results of the first model are statistically equal when performed in the full sample and the bio subsample.

¹³ Extended regression output was available as table A.4 in the online appendix.:

¹⁴ Estimation results are available in Table A.7 in the online appendix.