Jeroen Content, Nicola Cortinovis, Koen Frenken (p), Jacob Jordaan

Since the seminal article by Hidalgo and colleagues (Science, 2007), we have witnessed a surge of studies looking at the emergence of new regional specializations. It has become clear that the chance of developing a new industrial specialization is highly dependent on pre-existing activities in related industries. Hence, economic development is highly path dependent.

A follow-up question holds how regions can escape the forces of path dependence, that is, how regions can diversify into industries that are – to some extent – unrelated to their current industrial base. Here, we look at two conditions that may enable regions to diversify into more unrelated directions: the presence of Knowledge-Intensive Business Services (KIBS) and the presence of Global-Value-Chain Services (GCVS). KIBS and GVCS provide the generic competences of running business and managing global chains, respectively, which can be used in many different – and unrelated – industry contexts.

We use a new dataset that combines detailed information of sectoral employment at the 4-digit level. With this dataset, we can establish in what industries regions are specialised during the period of 2005 until 2014 for each year. Looking at the co-occurrence of industry specialisations we can measure inter-industry relatedness in terms of the number of regions that is co-specialised in two industries. We then estimate both the direct effect and relatedness-moderating effect of KIBS and GVCS on a region’s ability to specialize in new industries. Preliminary results suggest that KIBS and GVCS indeed support the emergence of new specializations as well as moderate the effect of relatedness. The latter result suggests that regions hosting many KIBS and GVCS allow them to diversify into more unrelated directions compared to regions with few KIBS and GVCS.

Michael Wyrwich (p), Michael Fritsch, Martin Obschonka

Policymakers around the world strive for creating the next Silicon Valley in their region. The Valley is a synonym for a vibrant local entrepreneurship culture and innovativeness. Research has shown that extraordinary entrepreneurial hotspots like the Silicon Valley developed due to “historical accidents” or have a very specific history that makes it difficult to replicate the success of these places elsewhere. We ask in this paper whether there is also a systematic historical component explaining regional differences in entrepreneurship culture today.
We measure entrepreneurship culture by aggregate regional personality traits. To be more precise, research has identified a combination of certain personality traits that can be regarded as making people entrepreneurship-prone. We regard regional differences in the average entrepreneurial personality profile as indicator for regionally different entrepreneurship culture. Comparing aggregate psychological traits across space is a well-established concept in psychological research that was recently also applied to the study of entrepreneurship (e.g., Obschonka et al, 2013). Mapping the average entrepreneurship-personality profiles across German regions reveals huge differences. In our econometric analysis, we are interested in understanding the role of historical conditions for explaining current spatial variation for the entrepreneurial personality profile of the regional population or entrepreneurship culture. In a nutshell, we investigate whether historical self-employment rates explain regional differences in entrepreneurship culture today. Earlier research demonstrates that historical self-employment rates are strongly correlated with current levels of new business formation and self-employment (Fritsch & Wyrwich 2014, 2017). This persistence indicates the presence of a local entrepreneurship culture that is self-perpetuating over longer periods of time. To analyze this pattern we are interested in whether there is indeed a relationship between historical levels of entrepreneurship and aggregate personality traits of today’s local population. We also consider migration of people with an above-average entrepreneurial personality score and different measures for historical self-employment rates in our empirical assessment.
Our findings reveal a positive relationship between historical entrepreneurship and current entrepreneurship culture. These results have two important implications. First, regional differences in entrepreneurship culture are to a significant degree determined by historically. Second, for this reason, it may be rather difficult to promote an entrepreneurship culture in the short run.
Our research calls for understanding the spatial differences of historical entrepreneurship, for example, over the course of industrialization. Why have some regions become more entrepreneurial than others? It also calls for sustainable long-term oriented policy designs to make places entrepreneurial in the future.

Jeroen Content (p), Nicola Cortinovis, Koen Frenken, Jacob Jordaan

One of the key findings of the growing literature on drivers of industrial specialisation is that countries and regions are more likely to diversify into new industries that are related to the economic activity already existing within that country or region (Hausmann & Klinger, 2007; Neffke et al., 2011; Boschma & Capone, 2015; Essletzbichler, 2015). In essence, this notion rests on the idea that it is more likely that related products are recombined to developed new goods or services than that unrelated products are recombined. Countries and regions endowed with a more developed economy therefore experience more opportunities for recombination, as they usually have more related industries. However, this reasoning does not clarify why and when regions diversify into unrelated industries. Unrelated diversification can be defined as the development of an activity that requires capabilities different from those supporting existing local activities (Neffke et al., 2014). Entrepreneurship and/or intrapreneurship may be able to introduce such capabilities and facilitate regions to diversify into unrelated industries. To assess whether this is the case, we use a newly constructed dataset for EU regions for the period 2005-2014 that combines detailed information of regional sectoral employment at the 4-digit level with information on total regional entrepreneurial activity. Using this dataset, we calculate annual levels of regional sectoral specialisations. Furthermore, looking at the timing of the co-occurrence of industry specialisations, we identify sectors that are related. To do so, we assume that industries are related to each other when regions are more often specialised in them together. With this information, we can determine the number of related industries that a region is specialised in, which we call industry density. Low density implies that a region has few related industries in which it is specialised, whereas high density implies that a region is specialised in numerous of the related industries. In our analysis, we estimate both the direct effect and moderating effect of entrepreneurship on a region’s ability to specialize in new industries, i.e., whether regions with high entrepreneurial activity have a higher propensity to specialise in new industries and whether it helps regions to diversify into unrelated industries. Preliminary results suggest that entrepreneurship in some cases supports regions to develop new specializations. Moreover, a higher regional activity of certain types of entrepreneurship seems to dampen the positive effect of density on regional relatedness, suggesting that it drives unrelated diversification.

Roberta Capello (p), Camilla Lenzi

Research on regional innovation processes has achieved important advances in the understanding of the conditions under which innovation occurs in a region and the spatial variants of regional innovation processes. Yet, several authors are increasingly emphasizing the need of a deeper and richer comprehension of how these innovation patterns can evolve over time and the economic implications of such evolutionary processes (Asheim et al., 2016). Consequently, an unexplored research avenue is the explanation of the determinants and implications of the dynamics of regional innovation patterns, through the analysis of the system of relationships (internal and external to the region) supporting them (Hassink and Klaerding, 2012).
This paper aims to contribute to this debate by focusing on the long term implications of the evolution of regional innovation patterns. This work starts from the concept of regional innovation patterns; they are conceived as alternative spatial variants/combinations of context conditions and of specific modes of performing and linking the different phases of the innovation process. This framework has been now conceptually accepted and empirically proved (Capello and Lenzi, 2013) and presents the advantage of considering all types of innovations, from radical to imitative approaches.
In the literature, studies on variations of regional innovation processes, possibly leading to progresses towards more complex ones exist. They are interpreted as the aggregate outcome of strategic individual behaviors of local actors pushing towards an endogenous switch and opening the way to alternative innovation modes, in accordance with recent debates in evolutionary economic geography (Asheim et al., 2015). Changes in regional innovation patterns have been recently identified for 262 NUTS2 regions of the EU in the period 2002-2006. The existence of these evolutionary changes in regional innovation patterns has called for analysis of territorial conditions enabling these structural changes to occur (Capello and Lenzi, 2017). What is missing in the literature is the analysis of the influence that these evolutionary, structural changes generate in the regional economic performance and dynamics in the years following such changes (2006-2014). This is the subject matter of this study.
The importance of this empirical analysis is also at the normative level. The analysis helps in highlighting under which conditions a normative effort in making regions change their mode of innovation is reasonable, and helps in suggesting advances in the current scientific and policy debate on the future of the smart specialization strategy.