Abstract
Previous studies typically relate apprenticeship training or more generally ‘Vocational Education and Training’ (VET) to training that is highly specific and that uses well-established technologies. Accordingly, apprenticeship training is typically not expected to have positive effects on innovation. In contrast, we argue in this paper that the type of dual apprenticeship training seen in Switzerland (or Germany and Austria) does create positive innovation effects due to the VET system’s built-in and institutionalized curriculum development and updating processes. These processes ensure that firms participating in apprenticeship training gain access to knowledge that is close to the innovation frontier and that ultimately fosters innovation. We provide theoretical explanations of how this knowledge diffusion works and how it can help to generate innovation. We use the Swiss VET system as one example and derive hypotheses about the relationship between firms’ participation in apprenticeship training and their innovation outcomes. Empirical analyses support our hypotheses. In a VET system with a built-in curriculum-updating process like the one in Switzerland (or Germany), firms participating in apprenticeship training have higher innovation outcomes than do non-participating firms.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For the duration of the VET program, apprentices have a fixed-term contract with a firm. During their training, apprentices are fully integrated into the production process and perform productive tasks similar or identical to those of qualified workers.
Future orientation, and thus innovation, is a key component of the Swiss VET system (Bundesversammlung der Schweizerischen Eidgenossenschaft 2002; Wolter and Ryan 2011; Bauder and Osterwalder 2008; Der Schweizerische Bundesrat 2003). For an overview of institutions in the Swiss VET system, see Pedró et al. (2009) and Rauner (2008).
Although the original model uses the firm as the level of analysis, the model is also valid for higher aggregation levels such as industries or nations (Jensen et al. 2007).
An example of the revised curricula for commercial employees and knowledge aggregation appears in Pedró et al. (2009).
Bundesversammlung der Schweizerischen Eidgenossenschaft (2002).
One could argue that there may also be an innovation effect via apprenticeship graduates that change firms after apprenticeship training (mobility of apprenticeship graduates is comparatively high in Switzerland) and thus take the new knowledge with them. Although, in general non-training firms may also benefit from training via such a channel, they can only gain from that knowledge with a delay of 3–4 years (until the apprenticeship is finished). So training firms will always have an advantage over non-training firms; thus we still expect a measurable innovation effect for training firms. In any case, such mobility effects only make the test of our hypothesis stronger.
We use the process and product innovation indicators to construct a fourth innovation measure. This binary measure takes the value of one if either product or process innovation takes the value of one, and thus it is an indicator for general innovation activity. Thus, we give this constructed measure the title of general innovation.
To investigate the influence of vocational education on firms’ innovation outcomes, we exclude observations from waves 1990 to 1996. These waves do not provide the necessary information either for the innovation measures or for the main explanatory variables. Furthermore, we restrict our sample to German-speaking regions, as vocational education is more widespread in these regions and because firms in the French-speaking part often follow a more consecutive VET approach that is closer to the French than the German training tradition.
In deviation from the sample average of about 70% training participation, the average training participation across all companies in Switzerland is about 20%. This difference is due to the fact that our sample contains only companies that have at least 5 employees and that the data base purposefully overrepresented the number of larger firms. In addition we lose several small and medium-sized companies due to missing observations. And small and medium-sized companies train less because they often do not have the capacity to train apprentices or are too specialized (Schweri and Müller 2008). Thus, the comparably large percentage of training companies in our sample results from an overrepresentation of larger firms in our sample.
The KOF makes an initial assignment of the questionnaires to firms based on a regional linguistic categorization provided by the Swiss Post. However, if a firm is unable to fill in the designated questionnaire due to a difference in the regional and the company language, it can request a different language version of the questionnaire. Using this information, we can construct a binary variable that indicates whether a firm’s spoken language is German or not. As our sample contains firms that are located in German-speaking regions, we expect all firms that returned a German questionnaire to have a stronger training tradition than the remaining firms who have French or Italian as their spoken firm language and tradition.
Due to missing observations for firms’ R&D expenditures, we cannot include their lags in our model.
Due to the panel structure of our data set, we risk getting biased standard errors if we do not correct for clustering at the firm level (Moulton 1990). Therefore, we use cluster-corrected standard errors for the basic equation and the instrumental variable equation.
Endogenous training decisions might be a potential source of bias in Eq. (1). This bias occurs if unobservable decisions influence both the training decision and the innovation output. Strategic management decisions, for example, might aim to foster innovation and simultaneously introduce the training of apprentices. To take this endogenous training decision into account, we test the robustness of the OLS estimates against endogenous training decisions with an instrumental variable strategy (Angrist and Krueger 2001). As instrumental variables, we use the firm’s age and firm’s language to measure its training tradition.
The decision to train apprentices might result in innovation outcomes that occur only a few years after that decision. To analyze such effects, we included the lag of the training variable in our model. Results in Table 6 in the “Appendix” still support our hypothesis of a positive effect of training on innovation. However the effects from the lagged variable are smaller than the ones of the original variable reported in Table 2. Thus, the innovation effect seems to occur rather immediately.
We present the results of a GMM estimation in this section, as GMM is an efficient estimator if heteroscedasticity and clustering occur. We also run the IV estimations with TSLS and limited information maximum likelihood (LIML). LIML is more robust to weak instruments than the procedures mentioned above (Stock et al. 2002). The two alterative IV estimations generate results that are in line with those in Table 5.
We also perform over-identification tests in every specification possible. Results are reported in Table 5 and basically support our IV approach. However, the test relies on the validity of the respective other instrument and is therefore limited in its informativeness. We therefore discuss potential threats to the validity of both instruments in the discussion and conclusion section of the paper.
By excluding young firms we also exclude potentially highly innovative start-ups. But the innovation process in start-ups may anyways strongly differ from the innovation process we describe in this paper and start-ups may not offer apprenticeship training directly from the beginning.
We also estimate specifications with a squared and cubic age term. All results are available from the authors upon request. Generally, results of the second stage are in line with the results we describe here.
Vocational education systems can also be an important part of a national innovation system and contribute in combination with other national institutions to the innovation outcomes of firms (Meuer et al. 2015).
References
Aghion P (2008) Higher education and innovation. Perspektiven der Wirtschaftspolitik 9:28–45
Aghion P, Howitt P (2006) Appropriate growth policy: a unifying framework. J Eur Econ Assoc 4(2–3):269–314
Aghion P, Bloom N, Blundell R, Griffith R, Howitt P (2005) Competition and innovation: an inverted-u relationship. Q J Econ 120(2):701–728
Angrist JD, Krueger AB (2001) Instrumental variables and the search for identification: from supply and demand to natural experiments. J Econ Perspect 15(4):69–85
Backes-Gellner U (1996) Betriebliche Bildungs- und Wettbewerbsstrategien im deutsch-britischen Vergleich, 1st ed. Rainer Hampp Verlag, Munich and Mering
Backes-Gellner U, Rupietta C, Tuor Sartore SN (2017) Reversed educational spillovers at the firm level. Evidence-Based HRM: Glob Forum Empirical Scholarship 5(1):80–106
Bauder T, Osterwalder F (2008) 75 Jahre eidgenössisches Berufsbildungsgesetz, 1st edn. h.e.p. verlag ag, Bern
Bolli T, Renold U, Wörter M (2018) Vertical educational diversity and innovation performance. Econ Innovat New Technol 27(2):107–131
Bundesversammlung der Schweizerischen Eidgenossenschaft (2002) Berufsbildungsgesetz: BBG
Carlsson B, Stankiewicz R (1991) On the nature, function and composition of technological systems. J Evolut Econ 1(2):93–118
Dosi G (1982) Technological paradigms and technological trajectories. Res Policy 11(3):147–162
Freeman C (1988) Diffusion: the spread of new technology to firms, sectors and nations. In: Heertje A (ed) Innovation, technology and finance. Basil Blackwell, Oxford
Gonon P, Maurer M (2012) Educational policy actors as stakeholders in the development of the collective skill system: the case of Switzerland. In: Busemeyer MR, Trampusch C (eds) The political economy of collective skill formation. Oxford University Press, New York
Jensen MB, Johnson B, Lorenz E, Lundvall BA (2007) Forms of knowledge and modes of innovation. Res Policy 36(5):680–693
Justman M, Teubal M (1995) Technological infrastructure policy (TIP): creating capabilities and building markets. Res Policy 24(2):259–281
Knutti P (2007) Dokumentation Berufsbildung. http://www.doku.berufsbildung.ch/download/dokubb/index.htm
Krueger D, Kumar KB (2004a) Skill-specific rather than general education: a reason for US-Europe growth differences? J Econ Growth 9(2):167–207
Krueger D, Kumar KB (2004b) US–Europe differences in technology-driven growth: quantifying the role of education. J Monetary Econ 51(1):161–190
Lynn LH, Mohan Reddy N, Aram JD (1996) Linking technology and institutions: the innovation community framework. Res Policy 25(1):91–106
Lynn LH, Aram JD, Reddy NM (1997) Technology communities and innovation communities. J Eng Tech Manage 14(2):129–145
Meuer J, Rupietta C, Backes-Gellner U (2015) Layers of co-existing innovation systems. Res Policy 44(4):888–910
Moulton BR (1990) An illustration of a pitfall in estimating the effects of aggregate variables on micro units. Rev Econ Stat 72(2):334–338
OPET (2010) Vocational and professional education and training in Switzerland 2010: facts and figures. Federal Office for Professional Education and Technology OPET, Bern
Pakes A, Griliches Z (1984) Patents and R&D at the firm level: a first look. In: Griliches Z (ed) R&D, patents and productivity. University of Chicago Press, Chicago
Pedró F, Burns T, Ananiadou K, de Navacelle H (2009) Systemic Innovations in VET: OECD Studie zur Berufsbildung Schweiz. OECD
Rauner F (2008) Steuerung der beruflichen Bildung im internationalen Vergleich. Kurzfassung. Gütersloh: Bertelsmann Stiftung
Sahal D (1985) Technological guideposts and innovation avenues. Res Policy 14(2):61–82
Schweri J, Müller B (2008) Die Ausbildungsbereitschaft der Betriebe geht nicht zurück. Panorama 5:25–26
Staiger D, Stock JH (1997) Instrumental variables regression with weak instruments. Econometrica 65(3):557–586
Watkins A, Papaioannou T, Mugwagwa J, Kale D (2015) National innovation systems and the intermediary role of industry associations in building institutional capabilities for innovation in developing countries: a critical review of the literature. Res Policy 44(8):1407–1418
Wolter SC, Ryan P (2011) Apprenticeship. In: Hanushek EA, Machin S, Woessmann L (eds) Handbook of the economics of education 3. Elsevier, Amsterdam
Acknowledgements
We thank John Addison, Simone Balestra, Bernd Fitzenberger, Simon Janssen, Edward Lazear, Johannes Meuer, Kira Rupietta, Paul Ryan, Rainer Winkelmann and the participants of the annual meetings of the Canadian Economics Association, Swiss Society for Economics and Statistics, the Bildungsökonomischen Ausschusses im Verein für Socialpolitik, the Spring Meeting of Young Economists, the Colloquium on Personnel Economics, the DRUID Society Conference and the research seminars at the University of Zurich for helpful comments and suggestions. We acknowledge and thank Natalia Abrosimova and Jan Hagen for research assistance and Natalie Reid for editorial support. Furthermore we thank the Swiss Economic Institute (KOF) for data provision.
Funding
This study is partly funded by the Swiss State Secretariat for Education, Research and Innovation through its Leading House on the Economics of Education, Firm Behaviour and Training Policies.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Rights and permissions
About this article
Cite this article
Rupietta, C., Backes-Gellner, U. How firms’ participation in apprenticeship training fosters knowledge diffusion and innovation. J Bus Econ 89, 569–597 (2019). https://doi.org/10.1007/s11573-018-0924-6
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11573-018-0924-6