Knowledge Spillovers and Absorptive Capacity—Institutional Evidence from the “German Mittelstand”

Abstract

Recent extensions to the knowledge spillover theory of entrepreneurship (KSTE) show that the successful commercialization of new knowledge by incumbents depends on their absorptive capacities. For policy-makers focusing on increasing incumbents’ innovative performance, the systematic reduction of knowledge filter through improved absorptive capacities thus becomes a crucial goal. While the general working mechanisms of knowledge filter have been analyzed within the KSTE framework, few institutional solutions to increase absorptive capacities have been put forth. This study provides an initial case study explaining a specific institutional framework fostering the systematic penetration of knowledge filters by incumbent firms in the case of German SMEs. Using a set of 177 in-depth interviews with firm representatives, the system of interrelated organizations, institutional arrangements, shared values, and economic incentives associated with the institutional structures for knowledge spillovers for German SMEs are described. I identify institutional characteristics connected to the dual system of vocational training, regulatory measures, and economic incentives mutually enforcing and fostering broad knowledge spillovers. This exploratory approach enables deriving hypotheses for the further study of knowledge filters as well as policy implications for the design of institutions increasing incumbents’ absorptive capacities.

Introduction

Building upon the endogenous growth theory (Romer 1986, 1994), the knowledge spillover theory of entrepreneurship (KSTE) addresses the issue of knowledge commercialization (Acs et al. 2009a; Audretsch and Keilbach 2008; Audretsch 1995), which had remained unconsidered in the original theory. This “missing link” (Braunerhjelm et al. 2010) between knowledge generation by universities, private research institutions, and business services as well as its commercial implementation in firms and (regional) economies has become the central issue studied within the KSTE (see Acs et al. 2013 and Ghio et al. 2015 for extensive reviews of the literature). Among the central results of this line of research is the existence and effectiveness of systematic barriers to knowledge commercialization despite its potential profitability to firms and welfare-enhancing effects for regional economies. These barriers have been denoted as “knowledge filters,” and a number of studies have been conducted to determine effective ways of penetrating the various filters hindering the efficient flow of new information (see, e.g., Acs et al. 2004; Acs and Plummer 2005; Mueller 2006; Carlsson et al. 2009).

The research on knowledge filter has primarily focused on technological start-ups close to universities and research facilities, as well as their effectiveness in commercializing knowledge. Overall, the perspective predominant in these studies is that entrepreneurs play the central role in commercializing new knowledge, while incumbent firms often tend to be structurally incapable of penetrating knowledge filter (Acs et al. 2009a, 2013; Audretsch and Keilbach 2007; Audretsch 1995). By contrast, building on Cohen and Levinthals’ (1990) and Qian and Acs’ (2013) consideration of firms’ absorptive capacities, Qian and Jung (2017) emphasize that incumbent firms can constitute a meaningful complement for the effective penetration of knowledge filters and should therefore be supported by policy-makers. This implies a shift of perspective: from improving conditions for entrepreneurs commercializing new knowledge in a specific innovative milieu to considering mechanisms supporting incumbents through designing institutions that aim at upholding and increasing absorptive capacities in firms and the overall workforce.

I argue that this shift of the basic perspective has substantial advantages for the design of policies aiming at the effective penetration of knowledge filters by extending the analysis to innovation systems characterized by low entrepreneurial dynamics yet substantial innovative capacities, an aspect scarcely covered in previous cross-national studies of national innovation systems (e.g., Audretsch and Caiazza 2016; Acs et al. 2016). In those innovation systems, incumbent firms can be considered the central driver of sectoral innovation. More specifically, when building upon the research on comparative institutional advantages—namely the Varieties of Capitalism approach (Soskice and Hall 2001)—it can be stated that the focus on entrepreneurs commercializing R&D-based innovations tends to disregard the working mechanisms of many innovative processes in “coordinated market economies,” i.e., most continental European countries. In these innovation systems, firms are comparably long-lived with a highly specialized workforce, whose fluctuation between firms is impeded, e.g., by high wages and strong employment protection, which fosters firm-specific investments in workers’ qualification. Overall, this system leads to firms, particularly SMEs, pursuing incremental, in-firm innovations, and continuously improving on niche products; properties captured by the term “German Mittelstand” (Massis et al. 2017). Consequently, in coordinated economies, the implementation of new knowledge in many sectors is conducted by incumbents focusing on incremental, low-key innovations, which are developed and implemented in close cooperation with customers and suppliers instead of formal R&D (Thomae 2017; EFI 2016; Jensen et al. 2007). To succeed in this specific innovation mode, firms rely on a high degree of absorptive capacity. Therefore, innovation policy in the respective countries and sectors does not primarily aim at increasing overall entrepreneurship to foster knowledge transfers. Rather, policy-makers are faced with the question of how to build, maintain, and improve institutions that succeed in increasing incumbents’ absorptive capacities.

This paper’s goal is to describe and analyze institutional mechanisms capable of systematically penetrating knowledge filters and fostering absorptive capacities in firms without direct links to the knowledge production. While based upon the example of a specific German sector, patterns are identified that are transferrable to other countries and economic contexts. This research goal resembles and—to some extent—complements Massis et al.’s approach of broadly describing the “German Mittelstand innovation model” and providing transferrable implications for other economic contexts, albeit with a more specific focus on the issue of knowledge transfer. This is achieved by conducting an in-depth investigation into a typical sector of German SMEs characterized by the traits described above and providing a broad description of its institutional setup in terms of knowledge spillovers. Given the holistic nature of the research question, a qualitative approach is chosen, namely the analysis of 177 in-depth interviews with firm representatives. The interview protocols are coded and a quantitative descriptive analysis and a qualitative analysis of the institutional mechanisms at play are conducted. Thus, the institutions, entrepreneurial values, regulations, organizations, and economic incentives are described that systematically support knowledge spillovers to incumbents in the German Mittelstand. Building upon these results, I derive policy implications applicable to other national innovation systems and sectors concerned with implementing incremental innovations and give propositions for the further study of knowledge filters in the context of the KSTE.

The remainder of this paper is structured as follows. In the second section, the theoretical context of this study is explained, while the third section describes the methodology and interview data. The fourth section presents the quantitative descriptive results (the “Descriptive results” section) as well as the qualitative analysis (the “Qualitative results” section) and an ensuing discussion with policy implications (the “Discussion and policy implications” section), before the final section concludes.

Theoretical Context

In pursuing the goal of providing exploratory evidence on the effectiveness of specific institutions for penetrating knowledge filter, this study is placed in the broader theoretical context of the KSTE. This strand of discussion is concerned with determining the relevance of knowledge for firms and working mechanisms for its transfer across firms within the economy (Acs et al. 2009a; Audretsch et al. 2006; Audretsch and Keilbach 2007).

Based upon the seminal contribution by Arrow (1962) regarding the allocation and role of knowledge for firms, the KSTE has identified several issues related to the entrepreneurial management of knowledge spillovers. Overall, knowledge has public good properties, making it non-rivalrous and non-excludable in most cases, which presents firms with the issue of preventing unwanted knowledge spillovers. This induces a risk aversion regarding knowledge sharing and cooperation for knowledge commercialization, which potentially weakens a firm’s market position. Therefore, firms often display a rational reluctance for conducting open sharing of knowledge, which, however, would be beneficial from a broader economy-wide perspective (cp., e.g., Driffield and Love 2005; Driffield et al. 2014; Cappelli et al. 2014).

This aspect is an important explanation for the “knowledge paradox” (Audretsch and Keilbach 2008), meaning that investment in knowledge production and dissemination does not automatically lead to increased regional or national competitiveness and growth. Thus, from the perspective of economic policy, measures for coping effectively with barriers to knowledge spillovers should rely on firm characteristics and build institutions capable of enabling and fostering knowledge spillovers. This should provide the missing link (Braunerhjelm et al. 2010) between knowledge production and commercialization, which is consequently the core subject of investigation in recent contributions to the KSTE.

Serving as an operationalization, the concept of knowledge filters has been introduced by Acs et al. (2004), which systematically and broadly captures the barriers from newly created to commercialized knowledge. Within this framework, a variety of different factors have been captured, such as the role of geographical characteristics (Acs et al. 2009a; Acs and Plummer 2005), different national institutions (Stenholm et al. 2013), the age of firms (Acs et al. 2009b; Carlsson et al. 2009), as well as social norms in the respective regions (Guerrero and Urbano 2014), all of which can enhance or hinder the effective flow of information from its production to the market. Within the research domain of knowledge filter, two topics are most relevant to this study, namely the investigation of knowledge networks and the discussions regarding firms’ absorptive capacities.

Knowledge networks have been investigated as a means of reducing the detrimental impact of knowledge filters and thus serve as effective transmission channels (see, e.g., Hayter 2013; Huggins et al. 2012; Shu et al. 2014; Huggins and Thompson 2015). It has been shown that, overall, knowledge networks between firms increase their overall performance (Shu et al. 2014) and that the existence and effectiveness of these networks account for regional differences in innovative capacities (Huggins and Thompson 2015). Therefore, regional policy strengthening those networks has a strongly positive impact in terms of fostering knowledge spillover (Huggins et al. 2012). This effect has similarly been shown for professional networks, which connect firms of a specific field which hold a common training and, therefore, knowledge base. These networks can similarly encourage innovative activities by the respective firms (Huggins and Thompson 2015). For the German example, Maennig and Ölschlaeger (2011) as well as Maennig et al. (2015) have shown that professional associations are associated with higher entrepreneurial and innovative activities. This line of research connects well to this study’s research goals, namely determining the effect of specific institutions in the German context, which are often based upon these knowledge networks. Therefore, determining in detail the institutional arrangements can further our understanding of the mechanisms prevalent in knowledge networks for reducing the effect of knowledge filters.

Secondly, the discussion of determinants and effects of absorptive capacities is discussed as a means of penetrating knowledge filter. This holds particular relevance when investigating incumbent firms and their ability to use and commercialize new knowledge. Defining absorptive capacity as “an ability to recognize the value of new information, assimilate it, and apply it to commercial end” (Cohen and Levinthal 1990, p 128), Qian and Jung (2017) show that high absorptive capacities of regional incumbents are associated with positive regional development. This contrasts previous results, e.g., by Acs and Plummer (2005), arguing that newly founded firms constitute a more effective mechanism of penetrating knowledge filter than incumbent firms. Based upon their contradictory result, Qian and Jung suggest that policy-makers should address the question of how to systematically increase incumbents’ absorptive capacities to effectively overcome knowledge filters and thus increase regional innovativeness. This study addresses this research goal by characterizing in detail for a specific market segment the institutional mechanisms associated with higher absorptive capacities, which connects the discussions on knowledge filter, knowledge networks and absorptive capacities for an in-depth analysis of a particular national market context. This particular example is then used to derive generalizable evidence regarding the further study of knowledge filters as well as more general policy implications for different national contexts.

Method and Data

Research Design

To achieve the research goals outlined above, this study describes a particular national structure for knowledge spillover. No systematic inquiry has so far been undertaken to describe the institutions connected to the German dual training scheme and the networks upheld by professional associations and chambers of craft from a knowledge spillover perspective. Consequently, this study applies an exploratory qualitative research framework to provide initial theoretical insight (Eisenhardt 1989; Eisenhardt and Graebner 2007). This approach is able to holistically cover the overarching institutions and the interconnected values and incentives. Developing hypotheses about the working mechanisms of the specific institutional framework opens up opportunities for subsequent empirical studies tackling more specific issues. This study thus builds on the qualitative research frameworks by various authors investigating knowledge spillovers (e.g., Schiller and Diez 2010; Yang and Steensma 2014; Schmidt 2015; Feser and Proeger 2017).

Due to the lack of previous theoretical and empirical work on the institutional structure in question, this study refrains from testing explicit hypotheses and instead centers on three guiding questions:

• 1. Which institutional mechanisms for the continual flow of information between producers of knowledge and SMEs exist?

• 2. How do these mechanisms work in practice, i.e., which incentives are at play? How does competition affect the mechanisms for knowledge spillover?

• 3. (How) can these institutional mechanisms be transferred to other economic/regional/national contexts?

To answer these questions, interviews with firm representatives from the German craft sector were conducted. Using the logic of theoretical sampling, the interview partners were selected to reflect different subsectors of this sector. This broad choice of firms assures theoretical saturation for the sample, and this particular subsection of the broad domain of German SMEs (see Glaser 1965; Glaser and Strauss 2008 for the basic methodological contributions). The interviews lasted between 5 and 30 min, depending on how extensively respondents answered. The interviewees were encouraged to explain at length their perspective on the respective matters. The interviewers followed a structured questionnaire but were encouraged to further inquire into the matter during the interviews. Anonymity was guaranteed to the interviewees and the interviews were recorded, transcribed and codified according to the research questions.

The questionnaire (documented in Appendix Table 6) comprised three parts: first, firm-specific details; second, details on the perception of the firm’s market situation; and third, details on the effects of the market specifics on the effectiveness of knowledge spillovers and their outlook on the effects of competition. In most interviews, the brief questionnaire was extended by individual inquiries depending on the interviewee’s respective answers. Thus, the interviews are not fully identical but rather aim at producing an extensive explanation of institutional working mechanisms, in which a formal questionnaire would have precluded. Thus, the questions did not aim at testing specific hypotheses from previous studies, but were chosen such that incumbents were encouraged to explain the structures of knowledge spillover faced in their markets.

The interviews were analyzed using the method of content analysis proposed by Mayring (2004). This method reduces the content to the statements relevant for the respective research questions. Thus, initially, open codings were developed and used based upon an initial reading of the transcripts. Once a set of broad categories had been developed, the open codings were further developed to more detailed codings and assigned to the categories and subcategories, which were subsequently used for the subsequent content analysis, upon which the results build.

While this research design is necessarily not representative of German SMEs in general, it can be assumed that theoretical saturation has been achieved as a requirement for the application of the grounded theory (Glaser and Strauss 2008). This methodological trait implies that no new information is gained regarding the respective questions in the respective segment of interviewees, even if an infinite number of additional interviews were conducted. The large number of interviewees and the recurring results within the interviews enables the assumption that theoretical saturation is fulfilled for the sample. This in turn enables more general statements about the underlying mechanisms of knowledge spillover in the respective sector and the derivation of policy implication.

Firm and Market Background

For this study, a sample of 177 interviews with German SMEs in the sector of craft products and services is used. This sector constitutes a core domain of the German economy. In 2016, it comprised about one million SMEs with about 5.45 million employees and 360,000 apprentices. This translates to about 12.5% of all German employees. The overall revenue of the German craft sector amounts to about 560 billion Euros for 2016. More specifically, interviews are conducted with entrepreneurs from the building sector, since it is the largest sector within the craft domain with about 60% of all craft SMEs.¹ The craft domain is also chosen since it is the classic example of a sector characterized by SMEs, almost exclusively relying on incremental innovation without formal ties to the systems of knowledge production (Thomae 2017; Jensen et al. 2007). The interviews are conducted on craft fairs with a focus on building, interior design and renovation in central Germany, at which firms present their products and services to potential customers.²

Choosing the German craft sector as an example of a traditional sector characterized by incremental innovativeness has the additional advantage that it basically includes two separate markets with specific properties (see Mueller 2016; Fredriksen and Runst 2016 for details):

• (1) A regulated sector, encompassing about half of all craft trades with a traditional entry regulation, in which the “Meister” certificate is required (see “Vocational Training” section). This certificate is a costly and time-intensive qualification encompassing technical and administrative training along with leadership skills and basic educational knowledge for training apprentices. In those sectors, the “Meisterpflicht” applies, meaning that only individuals with the respective qualification are allowed to establish and run craft firms. This regulation leads to a situation with higher average quality of products and services yet lower competition by entrepreneurial individuals, making it representative of the traits of the German Mittelstand.

• (2) In the second sector encompassing the other half of trades, market entry has been deregulated following a European Commission directive in 2004. Thus, the Meisterpflicht has ceased to apply and firms can be founded and run without a formal certificate of qualification. This has led to a substantial decrease of participants in formal training, a strong increase in new firms being established and a lower average durability of firms (Mueller 2016). Consequently, craft firms in these sectors have developed to more closely resemble more competitive, entrepreneurial economies with a substantially higher competitive pressure on firms.

Due to this two-part structure of an otherwise similar market, the institutions and incentives connected to knowledge spillovers can be discerned in the qualitative analysis. This extends the scope of the results from traditional firms in coordinated economies to firms working under market conditions more similar to those in liberal market economies. In the following, firms in domains with a limited market entry are denoted as being regulated, while markets without formal entry requirements are denoted as deregulated.

Table 1 provides descriptive statistics of the firms/individual entrepreneurs interviewed for this study. Summing up the basic information on the respondents, it can be stated that firms are generally older incumbents. They are mostly active on regional markets and have rather small numbers of employees, which also explains why mostly high-level staff and the owners themselves attend the craft fairs. About two thirds are active on the regulated markets and about one third on deregulated markets. About half are from eastern Germany, with the other half coming from western Germany. These basic descriptives point towards regionally active, smaller, older firms with a strong focus on their reputation for high-quality work. The interviews confirm that they are primarily active in market domains with quality competition and refrain from entering direct price competition with lower quality firms. Consequently, a strong emphasis is placed on high levels of technical training and service quality within those firms, which means stimulating and upholding knowledge spillover from various sources.

Table 1 Basic information on the firms interviewed for this study. Numbers smaller than the total number of interviews (177) imply a missing value, meaning that the interviewee either did not know a plausible answer or refused to answer

Firms with these traits are chosen since they are the most promising sources of information on the mechanisms of institutionalized knowledge spillovers for incumbent German SMEs. They are likely to be engaged in knowledge networks and have substantial economic incentives to adopt new knowledge, implement innovative practices, and monetize those knowledge spillovers, since their customers are willing to pay for quality improvements. Consequently, they are an interesting subgroup to learn about mechanisms of institutionalized knowledge spillover within the German institutional structure.

Results

Descriptive Results

In the following, the coded results to the main questions are presented and briefly interpreted. Overall, the dataset encompasses 31 variables (i.e., codings), which can be used to analyze the results of the 177 interviews quantitatively. Of these variables, the descriptive results of the most important ones are presented below to give an initial impression of overall answering patterns in the interviews. The dataset of respondents is categorized according to their market situation, meaning whether they worked in regulated or deregulated domains of the market (Table 2). This section is followed by an in-depth qualitative analysis in the section “Qualitative Results” and a discussion with policy implications in the respective section.

In the initial descriptive section, three domains of results are reported:

• (1) Vocational training. Do firms conduct knowledge spillover by participating in the vocational training scheme and what are the effects of competition?

• (2) Monetization of quality. Can knowledge spillovers be monetized on the markets?

• (3) Development of knowledge intensity. How has knowledge intensity developed on the markets in past years?

Participation in Collective Qualification Schemes

Looking at the descriptive results in Table 3, it can be assessed that eight out of ten firms in regulated markets participate in the qualification schemes, while two thirds of firms in deregulated markets refrain from doing so. Further, about half of respondents in regulated markets expect that increasing competitive pressure from deregulation would result in fewer participants of the qualification scheme (Table 4). Almost no firms expect additional trainees, while some expect no substantial effects. This basic result resonates with previous studies on the effects of deregulation on vocational training (Mueller 2016).

Table 2 Share of respondents participating in the collective vocational qualification scheme according to the market structure

Table 3 Responses to the anticipated change of qualification intensity in case of a deregulation, leading to higher competitive intensity, for firms in regulated markets

Table 4 Descriptive results on firms’ ability to monetize knowledge intensity according to the regulated/deregulated structure of the respective markets

Monetizing Knowledge Spillover and Knowledge Intensity

Regarding the aspect of monetization of a higher knowledge intensity in firms, several initial clues can be drawn from the descriptive results.

In regulated markets, firms assume that the Meister certification implies a higher level of quality to customers. Firms believe that customers tend to look for and act on different certifications of higher quality when deciding for a firm. Further, most firms assume that customers are able to anticipate the product and service quality, i.e., the firm’s qualification prior to the transaction.

In deregulated markets, the results are mixed. For about half of all respondents, the highest level of certification is accepted as a good indicator of quality by customers. Similarly, about half of all firms assume that customers base their decisions on other quality signals. Consequently, of those respondents who hold the highest degree of certification, a substantial share of respondents does not advertise this certification. Moreover, they acknowledge that this quality signal itself would not lead to a higher willingness to pay for products or services by their customers. Finally, the assessment regarding customers’ ability to anticipate product/service quality beforehand is comparable to firms in regulated markets, meaning that a little more than half of all firms assume that customers can determine the quality ex ante.

Regarding the economic incentives for increasing a firm’s knowledge intensity, these initial results hint at continuously high monetary incentives in regulated markets. For deregulated markets, this might apply to a lesser degree: while the highest certification serves as a quality signal, most firms do not assume that customers act strongly on quality signals and therefore refrain from advertising their higher qualification levels. This result yields the (preliminary) interpretation that price competition might not be sufficient to induce the costly acquisition of a high level of qualification when compared to markets with mandatory levels of qualification.

Development of Knowledge Intensity

The third dimension of descriptive results is concerned with the development of knowledge intensity on the two markets. For the analysis, it is assumed that firms’ assessment of the quality development on their markets in recent years is a broad indicator for the degree of knowledge intensity. Thus, evidence can be derived concerning whether the institutional framework at play has fostered knowledge spillovers in the respective markets. Overall, firms in regulated markets are somewhat more optimistic about the average quality in their markets, whereas about the same number of firms in deregulated markets assumes that quality has improved as compared to those thinking that it has declined (Table 5). Overall, it can be stated that SMEs in regulated markets tend to view the development of average quality on their markets, interpreted as knowledge intensity, more positively than SMEs on deregulated markets.

Table 5 Assessment of firms regarding the change in average product/service quality on their market in recent years according to the market structure. Further, firms on deregulated markets are asked whether the deregulation had an influence on the average quality in their opinion

Qualitative Results

For the qualitative analysis, the full interview protocols are used. There are 18 overarching categories with 5–10 subcategories each. This provides a structure of responding patterns by interviewees, which enables deriving evidence for the research questions. Using this methodological approach, three distinct channels fostering knowledge spillovers and increasing entrepreneurs’ absorptive capacities can be identified that are described in their respective sections: vocational training , regulation, and market mechanisms. In the following, those channels are described; in the section “Discussion and Policy Implications”, the results are summarized and policy implications are presented. For illustrative purposes, representative quotations from the interviews are provided in the footnotes to the respective chapters.

Vocational Training

Institutional Background

To provide an understanding of the basic channel for knowledge spillovers for German SMEs, the dual training scheme for apprentices is briefly explained. Its structure and working mechanisms apply to the craft sector but also to most other industrial or service sectors in Germany. It can therefore be considered basic knowledge for all interviewees. This system combines theoretical knowledge taught in publicly financed vocational schools with practical training at the chambers of craft and in firms and is therefore described as a mixed institutional approach to vocational training (for overviews, see, e.g., Busemeyer and Schlicht-Schmälzle 2014; Lassnigg 2016; Hernandez 2016).

From a knowledge spillover perspective, this system is quite effective in disseminating the current state of technological development in the respective field by combining the spillover of established practical experience-based knowledge from practitioners with newly created knowledge by industrial and university-based research institutions via the public vocational school system. Publicly financed vocational teachers are either required to have Meister-training and practical experience or are required to study their respective field at technical colleges or universities, whereby many teachers are experienced craftsmen or engineers who start teaching after or during their work in the field. Public schools work closely with the semi-private chambers of crafts to ensure the applicability and relevance of the classes. New knowledge is implemented through new teachers graduating from technical colleges or universities who are taught the current state of technology in their respective fields and within the school system by chambers of crafts who regularly update teaching requirements based upon the current technological developments.

Thus, the German dual system combines extensive in-firm training of apprentices becoming qualified workers (“Gesellen”) and prospective entrepreneurs (Meister) with a publicly financed school system and—in recent years—specialized fields of study at universities. Apart from the scale of this system of vocational training, the main difference to other national vocational systems lies in the specific role of chambers of craft, which essentially serve as a knowledge network. They connect the different organizations and firms, whereby membership is mandatory. Further, the chambers have a privileged judicial status, allowing them—in part—to enforce training requirements for apprentices and Meisters alike.

The curricula followed in schools and training programs are updated based upon recommendations by members of the chambers and other professional associations as well as vocational teachers educated in universities. Both the chambers and other professional associations as well as (technical) universities therefore play a role in establishing training programs to include basic vocational knowledge but also implement innovations in the field. Consequently, knowledge spillovers can occur either from institutions of higher education down to firms via the training of apprentices, through additional qualifications offered by the chambers to incumbents or through new entrepreneurs implementing their newly acquired knowledge from Meister schools or university studies in their own firms. If field-specific technological innovations have been realized by a firm and tend to constitute a new standard in the field, chambers of craft can include these innovations in their various training programs and might also set the respective innovation as a mandatory standard enforced across firms. Thus, there are both top-down and bottom-up spillover effects.

This leads to a situation in which the diffusion of new knowledge happens broadly, as craft firms are often required to conform to the most recent standards and all firms connected to the dual system of training regularly receive new knowledge through the public school system. Thus, not only very innovative firms receive knowledge spillover, but also smaller firms are effectively connected to the domain of knowledge production via the chamber networks. This effect is particularly positive, since knowledge filters are obviously stronger for smaller firms with little or no R&D capabilities, and the institutional framework tends to compensate these disadvantages in terms of knowledge spillovers. While this still does not guarantee an automatic absorption of new knowledge by the firm, it reduces a central barrier to knowledge diffusion and increases firms’ absorptive capacities via supportive external institutional structures.

In the following, the institutional mechanisms that additionally foster participation in this system of vocational training are described, based upon the interview results.

Craft Identity and Vocational Training

The most traditional channel of knowledge spillover is the entrepreneurs’ ethical connection to the vocational training system described above. Established by early German guilds in previous centuries, the ideal of a master-to-apprentice training has prevailed in many of the interviewed modern craft firms. Consequently, a strong emphasis is placed on a high and thorough level of training, i.e., knowledge and skill acquisition by all craftsmen within the firm. As a part of their professional identity, traditional firms are willing to invest in training new apprentices, sending them to vocational schools and receiving training classes from different knowledge providers for their more senior staff.³ Interviewees even argue that the strong identity effect on the willingness to initiate knowledge spillovers is independent of the economic risk of training potential competitors and that the underlying ethos is stronger.⁴

Interestingly, the traditional ethos passed on through the craft organizations—while implementing quality-based barriers to market entry—is all but hostile to knowledge spillovers; rather, the transmission of existing knowledge as well as the acquisition of new knowledge is supported by the traditional codes of conduct.

When asked about the potential effects of the requirement of higher qualification for creating and running a firm on the degree and quality of apprenticeship training, firms in regulated domains are fairly skeptical. Overall, they assume that professional training will suffer and that the traditional motivation for training apprentices will erode.⁵ Asking firms in the deregulated domains, they tend to state highlight that the traditional mentality can often no longer be maintained due to competition.⁶ Consequently, increasing competitive pressure might lead to a crowding out of the traditional mentality, since upholding it becomes more difficult from an economic perspective.⁷ This qualitative finding resonates well with the aggregate findings, showing that two thirds of firms in regulated markets train apprentices, compared to only one third of those in deregulated markets.

Continuous Qualification of Incumbents

Connected to the identity-based incentive to train new apprentices in the crafts, incumbents show a consensus that continuous qualification is a basic requirement in their field. Consequently, firms show an intrinsic willingness to adopt new technologies and invest in the continuous training of their employees. Consequently, the mindset described above is not limited to the training of new employees. Rather, essentially, all firms acknowledge that continuous qualification is a basic requirement of succeeding on their markets. This mindset is reinforced by traditional values, various regulatory and industrial certification requirements, and customers’ pressure to implement up-to-date technology. The combination of these factors leads to a mentality that actively seeks knowledge spillovers by working to increase absorptive capacities within firms.⁸ This requirement is not only limited to classes organized by the chambers, but is also fostered by private knowledge networks⁹ or large firms offering in-house training for their workers.¹⁰ Overall, the conviction that continuous vocational training for incumbents is a prerequisite for success on the market is ubiquitous across the interviews and independent of firm size and the regulation/deregulation of the respective domains. Firms thus use various channels to receive knowledge spillovers and continue to build up absorptive capacities.

• Proposition 1.

  Maintaining a state-wide dual training scheme fosters knowledge spillover.

• Proposition 2.

  The effectiveness of the dual training scheme hinges on substantial government expenditures and a participatory identity among firms, which might erode in market situations characterized by strong competition.

• Proposition 3.

  Upholding a mentality of continuous qualification among incumbents, chambers and industrial suppliers stabilizes knowledge spillover.

Regulation

A second separate institutional framework fostering knowledge spillover in the case of German SMEs is extensive public and chamber-based regulations regarding the quality of products and services and the level of training required to continue practicing the crafts. Therefore, regulators systematically determine quality standards, with the option of revoking licenses for market entry if the criteria are not met. While this is obviously a strict measure and strong intervention in the market, it also ensures that innovations can be declared a standard in the field and are therefore implemented by each firm within the respective markets, which again leads to a very broad dissemination of technologies and administrative practices. Three different layers of regulation and different institutions enforcing standards can be identified for the case of German craft SMEs.

Chamber Regulation

The initial and most basic regulatory instances are the chambers, which offer periodic classes for further qualification. Many interviewees refer to their mandatory classes for the acquisition of their Meister certificate as well as further training for all employees.¹¹ Consequently, craft organizations can help impose certain technological and administrative standards to a certain degree by offering classes to their members. While attendees are not always required to implement the respective standards, they are at least informed of important developments in their field. Consequently, a broad knowledge dissemination is fostered by these offers and—in part—requirements, thus penetrating the knowledge filter on a large scale for SMEs. Another regulatory factor is the element of firm-specific control by the chambers of crafts. This institution provides firms with mandatory certifications allowing them to perform specific tasks. The task of renewing certifications allows the chambers to fully access firms’ protocols of previous works at any given time. The chamber then tests whether the work has been completed according to the quality standards currently required. If certain standards have repeatedly not been met, firms can lose their licenses. This obviously provides a very strong incentive for firms to conform to current standards in the field and provides the chambers with an instrument to foster the dissemination of certain quality standards. According to the interviewees, this instrument is used particularly in the domain of ecological regulations, forcing firms to continually screen and adopt the most recent standards in the field. Note again that this is not limited to the most innovative firms or specific subsectors of the craft sector, but rather that it applies to all firms in the regulated domains and—less strictly—to firms in some deregulated sectors.

• Proposition 4.

  While reducing firms’ entrepreneurial scope, vocational chamber requirements are an effective means of continually enforcing knowledge spillovers even to small firms.

National and Supranational Regulations

A second layer of regulatory requirements is national or regulations by the European Union implementing specific technological standards. According to firms, this process has been especially strict in the field of ecological requirements in recent years. In this domain, the use of eco-friendly building materials, processes, and lower energy consumption is fostered. Due to the strict regulatory requirements, there are substantial R&D investments by industrial suppliers and firms are often required to implement the resulting technologies. Otherwise, at best, no federal subsidies for home-building or renovation are granted and, at worst, homeowners and firms have to pay fines. Due to these ever-increasing regulatory requirements, chambers of craft have to ensure that new regulations and the corresponding correct use of new technologies are implemented by their members. This constitutes a substantial incentive to penetrate the knowledge filter in the respective technological fields and uphold absorptive capacities in firms. Interestingly, these strict regulations in certain fields are considered by some to serve as a full compensation for the expected loss of quality due to the deregulation.¹² Overall, it becomes obvious from the interviews that the ever-increasing regulatory standards force firms to continue training and adapting new technologies by upholding or increasing their absorptive capacities.

• Proposition 5.

  State-wide regulation can trigger knowledge spillovers. Continually enforcing higher standards (e.g., in environmental domains) and connecting market entry to compliance triggers knowledge spillover.

Private Certification Systems

The third layer of regulation is privately organized certification systems, which can be acquired by firms to show their quality commitment. There are various different labels and associations screening firms for certain criteria and offering labels for higher quality in return. These instruments apply in combination with customer demand for high quality and can counter information asymmetries on the markets. The certifications enable external networks to control firms and ensure that a certain current state of the art in terms of service provision and technological standard is met. Often, the certification organizations similarly serve as knowledge networks for member firms, organizing the exchange of ideas and experiences with new technologies.

Finally, apart from national regulations, chamber-based or voluntary certifications, there is another form of supervision by an external institution that provides mandatory quality audits: as a semi-public institution, the German “TÜV” is required to monitor the quality of certain aspects of house building and renovations. It is financially independent on firm membership fees or homeowners and (among other things) file reports on the correct provision of craft services. If certain standards are not met, firms are required to re-do the inadequate aspects of their work. According to the interviewees, this external monitoring institution with far-reaching rights to intervene leads to a strong pressure to maintain and increase quality standards that go along and consequently implement new technologies and processes.¹³ Again, an external institution is able to impose certain standards—particularly in the field of eco-friendly materials—upon firms and has the right to sanction firms that are unable or unwilling to meet those standards. This provides another channel for the systematic transfer of new knowledge to a very broad base of incumbents. Therefore, being enforced by different layers of regulation, knowledge transfer might be more effective than through a purely market-based diffusion of technologies by individual entrepreneurs interested in first-mover advantages and the subsequent imitation by incumbents.

• Proposition 6.

  Private or semi-private certification institutions foster knowledge spillover with the side-effect of increasing market transparency in situations of asymmetric information.

Market Mechanisms

Market mechanisms are all effects related to market and customer demand on the development of firms’ absorptive capacities and—ultimately—knowledge spillover. Four main channels can be identified that determine whether firms aim to increase spillovers: their ability to monetize a higher level of knowledge, customer preference for higher quality goods and services, long-term regional reputation-building by incumbents through high-quality services, as well as market pressure to adopt new technologies by industrial suppliers.

Effects of Prices

The first and most straightforward motive for increasing knowledge spillovers through training and the implementation of innovations is a firm’s ability to increase revenues and charge higher prices for higher quality goods and services resulting from continuous knowledge spillovers. Interviewees stress that their focus on customers with a high-quality demand in most cases enables them to realize higher prices for most complex and innovative product solutions, which in turn provides an incentive to pursue the current state of technology in their respective fields. Most respondents agree that their customers value a high level of training and an effort to implement the current standard of technology, enabling them to charge higher prices due to their higher knowledge intensity.¹⁴ Thus, the continuity and incentive to uphold knowledge spillovers depends on the ability to charge higher prices for higher quality services. In regulated markets, this is less of an issue, since all firms are required to uphold a fairly common level of training. This common standard supports their willingness to absorb new knowledge, since costly knowledge transfer does not automatically lead to a competitive disadvantage through higher costs.

By contrast, in deregulated markets, knowledge spillovers depend upon firms’ ability to convert them into higher prices through various channels. The effectiveness of these channels in turn depends on customers’ ability and willingness to understand and value qualitative differences and pay higher prices accordingly. Since this criterion is often not met, knowledge spillovers suffer in markets with strong price competition.¹⁵ More generally, this issue is at play in markets characterized by asymmetric information, in which customers are unable to see through quality differences prior to consumption. In the worst case, this leads to a market-for-lemons effect (Akerlof 1970), in which lower quality products crowd out higher quality products with lower prices from the respective markets. The interviews provide initial evidence that this process might in fact occur due to the partial abolition of mandatory certification and the market’s inability to develop effective private certifications that successfully signal quality to customers. Consequently, customers mostly do not understand product quality ex ante,¹⁶ an effect that is aggravated by stronger price competition.¹⁷ This finding is in line with previous case study evidence suggesting that asymmetric information inhibits knowledge spillover on the market for energy efficiency consultants (Feser and Proeger 2017).

• Proposition 7.

  Market transparency is a vital prerequisite for strong monetary incentives of increasing knowledge intensity in firms. If quality signaling fails, continuous qualification can become a competitive disadvantage.

Customer Preferences for Quality

The second market-endogenous mechanism is customer preferences for quality and high levels of qualification in SMEs. Overall, despite the issues associated with signaling quality differences described above, interviewees emphasize that German customers strongly demand high levels of quality and generally have a willingness to pay for such quality products. They ascribe this to specific, culturally reinforced requirements for quality products in the domain of craft services. Even though they are eager to pay lower prices, customers are described as very uncompromising when it comes to quality and the implementation of the current state of the art in the respective field. This cultural aspect of demand is a strong push factor for firms to initiate knowledge spillovers, as a failure to do so would result in a loss of customer satisfaction and ultimately orders.¹⁸

Thus, as customers are accustomed to firms implementing modern technologies with a high level of quality, firms can draw back on this cultural preference to justify higher prices due to higher qualification. This culturally determined option thus serves as a corrective force against the issues related to asymmetric information and the crowding-out of quality products. If customers were less uncompromising in terms of accepting lower quality, the incentive to uphold knowledge spillovers would certainly be lower for German SMEs. However, since firms can always explain higher prices with their quality commitment and customers overall accept this connection, the overall level of knowledge spillovers is supported, which might have led to quality increases in the regulated domains.¹⁹

• Proposition 8.

  Customer preferences for firm qualification constitute a major incentive for conducting knowledge spillovers. Customers expecting high-end solutions thus enforce knowledge spillover on a broad scale.

Reputation Seeking by Regionally Oriented Firms

As a third market-endogenous mechanism, the interviews show that firms invest in positive regional reputation. Knowing that most of their customers have a recurring, long-term demand for high-quality products and services, firms aim to establish a strong reputation, which enables them to charge higher prices compared to larger, non-regional firms and build customer loyalty.²⁰ Given that this quality reputation can only be achieved by consistently investing in qualification at all levels of a firm’s activities to provide satisfactory products and services to quality-interested customers, it provides a long-term incentive for upholding knowledge spillovers.

Reputation is also a classic example as a means of reducing the issues of asymmetrical information by making product specifics visible through transparent long-term quality standards. Consequently, it is a potential remedy to market failures in the domain of craft products and services precluding market-for-lemons developments. Connected with high-quality standards and the effectiveness of word-of-mouth information dissemination, it upholds high-quality standards. Consequently, particularly, regionally active firms are forced to uphold the acquisition and implementation of new knowledge.

• Proposition 9.

  Reputation-building by regionally active, long-lived firms can provide an additional stabilization of qualification incentives.

Technological Development and Craft-Industry Linkages

A further market mechanism increasing knowledge spillovers is the intense linkage between industrial, R&D-intensive industries providing intermediate products and craft services responsible for adapting and implementing those products at the final customers. While the craft sector includes very few firms conducting formal R&D, the large industrial suppliers strongly invest in new technologies and have strong linkages with universities and research institutions. While individual crafts and even chambers cannot sustain innovative cooperation with these institutions due to size constraints in most cases, the industrial suppliers are able to foster technological developments.

As industrial intermediate products become more complex, versatile and—in recent years—heavily influenced by information technology, the requirements for firms have substantially increased. In order to implement the technological state of the art produced by the suppliers, substantial investment in qualification is necessary, since a failure to do so would preclude firms from offering the latest technologies to customers.²¹ Thus, there is a specific linkage between the market interests of suppliers, who either need to implement increasing regulatory standards or want to gain market advantages over the competitors and the craft firms required to implement the technological standards.

Industrial suppliers thus have a strong interest in craft firms being able to implement their products and therefore try to foster the necessary knowledge spillover, e.g., by financing classes on their products and requiring firms to acquire private quality certifications allowing them to use the respective products. Therefore, market pressure induces penetration of the knowledge filter by larger firms connected to technological innovation in favor of smaller firms. This specific industry-craft linkage has a certain momentum—fueled by technological innovations, market incentives and regulatory standards—which forces incumbent SMEs to pursue knowledge spillovers.²²

• Proposition 10.

  Strong linkages to sectors conducting R&D and market pressure from suppliers and customers to implement new technologies helps to penetrate knowledge filters.

Discussion and Policy Implications

This study has so far addressed the institutional mechanisms for the continuous flow of information between producers of knowledge and SMEs and how they work in practice. Figure 1 provides a summarization of these results.

Fig. 1

Summary of the institutional framework for knowledge spillover for German craft SMEs. Mechanisms with a dotted structure are weakened or drop out altogether in more competitive markets without entry barriers according to qualification

Building upon these basic results, the third research goal can be addressed by providing generalizable results from the German case study.

First, market conditions should be influenced such that the continuous implementation of new technologies leads to a price premium for firms choosing to invest in knowledge acquisition. This can be achieved by providing a broad set of certifications and quality labels, whereby effective signaling of quality is crucial in markets characterized by asymmetric information, in which consumers cannot easily determine product quality ex ante. If mandatory certifications cannot be implemented, efforts should be made to increase market transparency to foster knowledge acquisition by incumbents. The deregulation in the German case shows that the abolishment of mandatory standards can lead to erosions of quality standards under conditions of asymmetric information. Markets forces might be unable to develop private certifications that credibly signal quality standards compensating the effects of mandatory standards.

• Policy implication 1.

  Increasing market transparency through voluntary or mandatory quality signals increases the economic incentives for continuous knowledge spillover by incumbents.

Second, the linkages between highly innovative firms in the supply chains and incumbents should be strengthened to ensure that the implementation of innovations on a broad scale is pursued by innovating firms and the final customers. If successful, these two factors exert a strong pressure on incumbents to increase absorptive capacities and cooperate closely with innovative suppliers to satisfy customers demand. Firms’ inability to uphold their absorptive capacities thus leads to a substantial disadvantage on their markets, which optimally leads to a self-increasing process of knowledge spillovers from highly innovative sectors to incumbents. These linkages can also be fostered through the continuous increase in product and quality standards by regulators, which essentially force incumbents to uphold their absorptive capacities to avoid fines by the state or judicial claims by consumers.

• Policy implication 2.

  Regulation implementing the technological state of the art issued by industrial producers as mandatory or publicly visible voluntary standards can exert pressure on incumbents to increase their absorptive capacities and conduct knowledge spillovers with industrial firms.

Third, substantial public investments in vocation training institutions offering free vocational training can help to ensure the continuous spillover of knowledge transmitted from university-trained teachers to workers who—at the same time—implement the new knowledge in firms. A public training scheme covering the majority of incumbents provides an opportunity for the broad-scale diffusion of new knowledge among the workforce and incumbent firms, which otherwise would not have received the current state of technology. This direct link between knowledge production and incumbents is probably the single most effective means of knowledge spillover, which directly bridges the knowledge filter. In this regard, it might be as effective as entrepreneurs directly implementing innovations, but functions on a much larger scale, as innovations can easily be set as a training standard, thus reaching thousands of apprentices each year and thereby influencing a similar number of firms.

• Policy implication 3.

  The creation of strong linkages between incumbents, public vocation schools, and professional associations through the cooperative training of apprentice can be initiated through maintaining a broad public vocational school system, which continually updates curricula according to the state of technological development.

Overall, this study’s policy implication is that a stronger coordination of knowledge producers, networks, professional associations, regulators, and incumbents can be an effective tool to foster knowledge spillover. This coordination can be achieved through monetary investments in public schools, market transparency measures, and the enforcement of regulatory standards. This result generally resonates well with the research on knowledge networks and their positive effect on knowledge spillover (Hayter 2013; Huggins et al. 2012; Huggins and Thompson 2015; Shu et al. 2014 and (for the German case) Maennig et al. 2015; Maennig and Ölschlaeger 2011). While many aspects pointed out in this study might primarily apply to coordinated market economies and incremental, non-R&D innovators, many of their traits can be transferred to other national and regional contexts, particularly to regions or nations characterized by a low entrepreneurial dynamic, little knowledge production yet a large number of incumbent firms capable of implementing incremental innovation if provided supportive institutions.

Conclusion

Extending Qian and Jung (2017) as well as Massis et al. (2017), this study has investigated the German Mittelstand innovation model, and more specifically, its means of systematically increasing absorptive capacities in incumbents using the example of a typical German SME sector characterized by low entrepreneurial dynamics, few formal connections to the knowledge production and an incremental, non-R&D-based innovative process. Using interviews with SMEs, I have identified the main institutions, mechanisms, values, and economic incentives driving knowledge spillovers in this sector despite its considerable disadvantages when compared to knowledge-intensive entrepreneurial sectors.

This investigation has led to a number of propositions that can be used to further investigate institutional mechanisms capable of penetrating knowledge filters. Among those are the maintenance and investment in a national dual training scheme with technical training in public schools and occupational learning within firms. Further, the establishment and support of a mentality of continuous learning among the incumbents is organized and supported by vocational chambers and enforced by industrial suppliers. Similarly, requirements by chambers, state-wide regulation and certification systems are an effective means of enforcing knowledge spillover across firms. Considering market conditions, transparency, and thus, monetary incentives for increased knowledge intensity are vital, as well as a strong customer preference for quality. Those incentives can be further stabilized by reputation-building by regional firms. Finally, connections to firms conducting F&E or knowledge-intensive industrial suppliers support and—in part—enforce knowledge spillover and lead firms to increase their absorptive capacities. Overall, it can be shown that industry-SME linkages, differentiated regulation and—most prominently—the dual training system are effective mechanisms for broad knowledge spillovers to a substantial share of firms in the respective sector. This coordinated system effectively and continuously penetrates knowledge filters, leading to a situation in which the dissemination of new knowledge does not exclusively hinge on the existence of entrepreneurial individuals disseminating innovations through market processes. Instead, incremental innovations are continuously disseminated across incumbents in a combination of voluntary and mandatory technology adaption, mainly through the institutions connected to the chambers of craft, which effectively support firms’ absorptive capacities.

From a policy perspective, this firstly leads to the implication that market transparency regarding quality signals should be supported to stabilize incentives to uphold absorptive capacities. Secondly, regulation prescribing the state of the art through suppliers, chambers, as well as national regulators successfully stabilizes knowledge spillovers, even to small firms, and should therefore be upheld. And thirdly, strong linkages in the field of vocational training between the different organizations and firms associated with occupational training should be fostered. Overall, this leads to the conclusion that the stronger coordination of market participants, non-market organizations, and regulators can be an effective way of increasing absorptive capacities in incumbent firms.

More generally, it can be suggested that the further study of knowledge filters can profit from a practical institutional perspective, describing in detail the mechanisms and structures supporting incumbents’ absorptive capacities. Investigating institutions and their working mechanisms thus supports the derivation of applied policy implications for an increase in absorptive capacities fostering broader knowledge spillover.

Appendix 1

Table 6 Guiding questions for the interviews. Please note that the overall number of questions asked to interviewees was higher. Only the questions relevant for the analysis in this paper are documented in this appendix. Overall, 21 questions were discussed with respondents (apart from the initial 8 questions on firm specifics), of which only the 14 questions documented below are related to knowledge spillovers. Note, further, that the guiding questions were often discussed in further detail with respondents. Thus, the questions documented below in many cases only constitute the starting points of a brief discussion of the respective content