Do works councils improve the quality of apprenticeship training? Evidence from German workplace data

Abstract

Purpose Works councils have the legal right to participate in a firm’s training process and, where necessary, call for a replacement of training instructors. The purpose of this paper is to empirically test whether works councils are associated with a higher quality of apprenticeship training – or its inputs or outputs – in Germany. Design/methodology/approach The authors use two representative cross-sectional surveys of German workplaces in 2007 and 2012/2013 that were conducted by the Federal Institute for Vocational Education and Training in Germany. To account for selection on observables, the authors apply nearest neighbor matching models to estimate the extent to which works councils are associated with training quality. Findings The results shed light on the influence of works councils on the quality of apprenticeship training in Germany. Based on a quality model, the authors show that works councils are associated with a (moderately) higher output quality of apprenticeships, particularly with respect to the share of retained apprentices. However, the authors do not find empirical evidence for a positive association between works councils and input- and process-quality indicators. Research limitations/implications Although the identification of causal effects due to the existence of works councils is difficult and cannot be fully addressed in the analysis, the authors can use a number of important control variables at the workplace level. The results suggest that a works council only plays a moderate role in enhancing the quality of the German apprenticeship system. Originality/value The authors provide the first direct empirical evidence of how the existence of a works council is associated with the input-, process- and outcome-quality measures of the German apprenticeship system.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 4 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.