The effect of temporary workers and works councils on process innovation

Abstract

PurposeThis study aims to investigate the effects of temporary workers and works councils on process innovations at manufacturing sites. The impact of temporary workers, commonly viewed as a means of operational flexibility and cost savings, on firms’ ability to innovate is underexplored. Works councils represent and help integrate temporary workers, but are often equated with unions, which have been criticized as barriers to innovation, especially in the US.Design/methodology/approachThe authors use secondary data collected by the Institute for Employment Research (IAB) of the German Federal Employment Agency. Specifically, the authors conduct a series of regression analyses using 11-year panel data covering the period 2009–2019 with 11,641 manufacturing site-year observations.FindingsThe results suggest that the use of temporary workers initially promotes process innovation, but at too high a level, it impairs firms’ ability to innovate. Furthermore, the results suggest that works councils have a positive impact on innovation and dampen the curvilinear effect found with respect to temporary workers.Originality/valueResearch has largely focused on the cost and flexibility benefits of temporary workers. The authors analyze the effectiveness of temporary workers in terms of innovativeness. By including works councils, the study also consider the contextual environment in which temporary workers are employed. Finally, the results reject the assumption that works councils have a similar negative impact as unions on innovation; in fact, the authors find the opposite.

Does global supply chain integration payoff? The case of maritime shipping firms

“…To give further confidence that our results are not an artifact of our model specification, we performed a robustness check in which we ran our Difference GMM estimation with a two‐step Windmeijer (2005) finite sample correction following Roodman's (2009) guidance on two‐step GMM estimation. The Windmeijer (2005) correction addresses any downward biased standard errors in the two‐step estimation (Durach et al, 2023; Roodman, 2009). Our results are largely consistent between the two‐step and the one‐step estimates (see Table S3 (Models 7 and 8) in Appendix S1).…”