1 Introduction

In universities and other public research institutions (e.g. federal labs), it is common to assemble heterogeneous teams, consisting of researchers with diverse cultural and educational backgrounds (OECD 2001; Boni et al. 2009). The rationale behind this strategy is that both research institutions and host countries can benefit from the inflow of human capital gained through the in-migration of foreign scientists. From the perspective of scientific institutions, researchers with different educational background and origin may complement each other while jointly working on research projects (Mahroum 2000; Phan et al. 2009). Such knowledge exchange is conjectured to stimulate diversity and creativity, which may lead to innovative and creative ideas (Fleming et al. 2007). Thus, employing scientists with different backgrounds is thought to increase the quality of research output and competitiveness of scientific institutions. In the long-run countries may also benefit from the inflow of foreign-born and foreign educated scientists since knowledge and human capital are assumed to be key drivers of sustainable economic growth and wealth creation (Lucas 1988; Romer 1990).

In this study, we examine the likelihood of foreign-born and foreign-educated scientists to become entrepreneurs. By doing so, we add an additional aspect to the literature on the potential benefits from ‘brain gain’ or ‘brain circulation’. Previous studies rather focus on how foreign-born and foreign-educated scientists contribute to research productivity via publications (Stephan and Levin 2001) and patenting (Hunt and Gauthier-Loiselle 2008; Kerr 2008). At the same time, there is increasing evidence emphasizing the economic importance of entrepreneurial activities of scientists. Zucker and Darby (2007) report that entrepreneurial scientists have a significant impact on the economy, even though the number of firms they create is relatively small. Typically, such firms are established in high-technology industries and have the potential to grow and to become leading firms in terms of innovation and employment (Darby and Zucker 2003). Thus, by spawning new technology-based firms public research institutions could have a substantial positive impact on regional economic development and play an important role in the development of high-technology districts (Castells and Hall 1994; Keeble and Lawson 1997). Unfortunately, the scholarly analysis of academic entrepreneurship has not focused on effects of in-migration and circulation of scientists. Given that foreign-born professionals have contributed to the development of high-technology industries in the US (Saxenian 2002; Saxenian and Hsu 2001), we argue that our study helps identifying whether similar patterns exist among academic entrepreneurs.

The argument that foreign-born and foreign-educated scientists are more entrepreneurial than their peers is based on the assumption that mobile scientists possess broader professional skills and social capital that are conducive for entrepreneurship. Due to such broader skills mobile scientists are more likely to identify entrepreneurial opportunities. On the other hand, foreign scientists may recognize entrepreneurial opportunities that were non-existent in their home countries and then establish businesses in foreign countries to exploit such opportunities.

We test this hypothesis using a sample of approximately 2,500 researchers from the Max Planck Society in Germany, which is regarded as one of the most distinguished research organizations in the world. Though focused on basic research, the Max Planck Society has an institutionalized technology transfer office that supports scientists to commercialize their research. The technology transfer office of the Max Planck Society—Max Planck Innovation—managed 3,000 invention disclosures and approximately 1,800 commercialization agreements. In sum, Max Planck Innovation has generated more than 260 million Euros in licensing revenue.Footnote 1

Our results suggest that both foreign-born scientists (non German citizens) and foreign-educated scientists (German scientists with Ph.D. degree from foreign universities) are more likely to become entrepreneurs than German scientists and German scientists with a Ph.D. degree from a German university. Thus, our findings reveal a positive relationship between career mobility and the propensity to engage in academic entrepreneurship. In the case of foreign scientists, one may argue that variation of entrepreneurial activity of scientists may also be influenced by the entrepreneurial culture of their respective home country, which may shape their entrepreneurial proclivity. We address this in our empirical strategy and use several controls for country-specific effects of entrepreneurial activity. Thus, our findings reveal that the relationship between mobility and entrepreneurial propensity is robust regarding entrepreneurial culture and a large number of other controls. Thus, evidence presented in this paper contributes to the important policy debate regarding immigration and the scientific workforce.

The remainder of the paper is organized as follows. Section 2 elaborates on factors influencing the entrepreneurial likelihood of scientists and the reasons why foreign-born and foreign-educated scientists should be more likely to become entrepreneurs. Section 3 provides some background information on the Max Planck Society. Section 4 introduces the research methodology and describes the data and variables examined while the empirical results are provided in Section 5. Section 6 provides conclusions and caveats.

2 The likelihood of foreign-born and foreign-educated scientists to engage in entrepreneurship

In this section, we describe several studies indicating that the quality of scientific human capital increases with mobility. We also summarize key findings from the literature on academic entrepreneurship, which concludes that the probability of a scientist becoming an entrepreneur is positively associated with his/her human capital. By linking these two literatures, we conclude that scientists who work outside their home country and foreign-educated native scientists are relatively likely to engage in start-up activity due to their diverse knowledge.

Existing studies indicate that both foreign-born and researchers with intense experience abroad are likely to be among the most able scientists. Stephan and Levin (2001) analyze the productivity of researchers in science and engineering in the US and find that foreign-born and foreign-educated are disproportionately represented among the scientists making exceptional contributions in terms of high quality ‘hot’ papers, citations and elections to the National Academy of Engineering. Using the 2003 National Survey of College Graduates in the US, Hunt and Gauthier-Loiselle (2008) show that foreign-born hold on average higher degrees in science and engineering, and patent at double the native rate. Similarly, Wadhwa et al. (2007) report a dramatic trend in the inventive activities of foreign-born in the US. The authors estimate that the contribution of foreign-born scientists residing in the US to international patent applications filed from the US increased from 7.3% in 1998 to 24.2% in 2006.

Moreover, while the aforementioned studies mainly examine differences between foreign and native scientists, other studies assess the relationship between scientists’ career mobility and research productivity. Examining inventor mobility, Fleming et al. (2007) provide evidence that career mobility enhances individuals’ patenting activity. Using bibliographic databases of Spanish scientists, De Filippo et al. (2009) provide evidence that both scientists’ number of research stays in different institutions as well as the number of visits of external researchers are positively related to subsequent research output. Similarly, Jonkers and Tijssen (2008) suggest that work experience in foreign countries has a distinctive positive impact on the publication activities. Analyzing Chinese plant molecular life scientists in top level research organizations in China, the latter study suggests that scientists who returned to their home country after staying abroad are significantly more productive than Chinese scientists without such experience. Furthermore, by tracking biotechnology scientists from US universities over time, Stuart and Ding (2006) find that the number of job changes is positively related to the likelihood to become entrepreneur. Stuart and Ding (2006) reason this finding by claiming that individuals with many job changes have had frequent employment offers and are therefore assumed to possess valuable skills.

Observing the high research output of foreign-born and mobile scientists, several authors have asserted that career mobility is conducive to the human capital and innovative capacity of individuals (see e.g., Hoisl 2007; Barr et al. 2009). Individuals with work experience in many different environments are more likely to possess a greater diversity of ideas, perspectives, assumptions and creative techniques than individuals who only worked in few different environments (McEvily and Zaheer 1999). The argument that career mobility enables scientists to acquire broad and diversified knowledge is also supported by the social network literature. According to this stream of literature, individuals with access to many different sources of knowledge are more likely to be more creative and innovative than individuals with access to a fewer knowledge sources (Granovetter 1973). With broader knowledge and ideas individuals’ capacity to innovate increases since many ingenious ideas are born in the process of (re-)combining different parts of knowledge (Jacobs 1969). Thus, the ability to recognize innovative and entrepreneurial opportunities is thought to be especially pronounced among mobile scientists.

The literature often argues that scientists’ expertise and competences are conducive to academic spin-off formation, given that academic spin-offs are typically knowledge and technology-intensive (Siegel et al. 2003, 2007; Wright et al. 2007). Empirical support for this view was provided by Lowe and Gonzalez-Brambila (2007), who found that more productive scientists are more likely to become entrepreneurs. Further, the human capital and scientific expertise—measured in innovation output and years of research experience—of academic founders are positively related to scientists’ likelihood to become entrepreneurs and the success of the start-ups (Shrader and Siegel 2007; Stuart and Ding 2006; Toole and Czarnitzki 2009). Combining the two aforementioned separate argumentations, leads to our prediction that foreign-educated scientists (H1) and foreign-born scientists (H2) are more likely to become entrepreneurs.

Another reason why foreign scientists might be especially inclined to become entrepreneurs is provided by Saxenian (see e.g., Saxenian 2002; Saxenian et al. 2002). These studies examine entrepreneurship in Silicon Valley and conclude that foreign scientists are especially likely to become entrepreneurs. Saxenian argues that foreign-born professionals may recognize opportunities that were not prevalent in their home countries. Thus, the change of environments induces the likelihood of identifying novel business ideas. Foreign scientists, thereby, do not necessarily move with the idea to found a business. It is rather the appropriate environment for high-tech entrepreneurship—as the availability of potential investors—which positively stimulates their entrepreneurial activity. In a similar vein, Krueger and Brazeal (1994) argue that individuals may become more likely to start an own venture when their environment becomes more conducive to entrepreneurship.

H1

Within the group of native scientists, foreign-educated scientists are more likely to become entrepreneurs.

H2

Foreign-born scientists are more likely to become entrepreneurs than their native peers.

3 Framework and methodology of the study

3.1 The Max Planck Society: research setting and employment policy

Our sample of entrepreneurial and non-entrepreneurial scientists is based on a survey we conducted at the Max Planck Society (henceforth, MPS) in Germany. The MPS consists of 78 independent research institutes comprising more than 9,000 scientists who perform basic research in natural sciences, life sciences, and humanities. Within the public science sector in Germany, the MPS is complementing German universities by performing research in new and innovative fields that universities are not able to embed in their research agenda. The MPS focuses on basic research and seeks scientific excellence in cutting-edge research fields. Therefore, particularly new and innovative fields are pursued, even if such research requires costly equipment or long-term funding. Research excellence of the MPS is documented by 32 Nobel Prizes awarded to the MPS since 1911. According to Times Higher Education, the MPS is ranked as the number one of non-university institutes in science and number 3 in technology after AT&T and Argonne National Laboratory in 2006.Footnote 2

The Max Planck Society is seen as one of Europe’s science powerhouses and researchers with different backgrounds and origins are attracted by its research excellence and work environment. Employment policy within the Max Planck Society seeks to hire those scientists who can contribute to the knowledge frontier. Thus, analyzing entrepreneurial activity within the Max Planck Society enables us to analyze to what degree the circulation of scientists’ effects entrepreneurial activity.

3.2 Academic entrepreneurship in the Max Planck Society

Max Planck Innovation, the technology transfer office of the Max Planck Society, closes on average 80 license agreements a year, about half of them with companies abroad. The nature and extent of the licenses are as varied as the payment modalities. They range from upfront payments and sales-based royalties to additional milestone payments (e.g., in the case of drug development). Furthermore, according to their self-reported records Max Planck Innovation currently oversees more than 1,113 inventions and has shareholdings in 15 companies. Since 1979, Max Planck Innovation managed about 3,000 inventions, have closed more than 1,700 license agreements and, since 1990, coached 89 start-ups. The total revenues from Max Planck inventions currently amount to about 260 million Euros.

In 1990, Max Planck Innovation started professional support for MPS scientists who intended to create spin-off companies. These services are equally available to all institutes. Services for founders include help in assessing the feasibility of scientific research results for a business start-up, assistance in writing a business plan as well as finance planning and search for potential financiers. Thereby, the technology transfer office relies on a network of venture capital companies, banks and business angels. However, neither the MPS nor Max Planck Innovation invests capital in spin-offs.

Since 1990, Max Planck Innovation has spawned 89 companies, all in high-technology industries as life sciences, biotechnology, biochemistry and physical engineering. The technology transfer office coached 53 out of these 89 companies in the founding process. Out of these 89 companies 58 firms are licensed-based spin-offs. Such a firm is subject to a license, know-how or option agreement with Max Planck Innovation at the time of founding. Further, 26 companies are scientist start-ups meaning that a former researcher of Max Planck Innovation is founder or employee of a company that is not subject to a technology license agreement with Max Planck Innovation. Moreover, 45 companies were backed by venture capital and 13 start-ups were later subject of M&A deals with existing companies in the field. Currently (as of December 31st, 2009) these 89 companies created about 2,330 jobs with nearly 150 patents licensed to these spin-offs by the MPS.

An example of a spin-off company is Lambda Physik, which was founded in 1971 as a commercial off shoot of the renowned Max Planck Institute for Biophysical Chemistry in Goettingen, Germany. This enterprise was created by two scientists of the Max Planck Society, Dirk Basting and Bernd Steyer. Soon after foundation, Lambda Physik has developed into a worldwide technology leader in pulsed lasers for industrial, medical and scientific applications including pholitography, flat screen production and laser ablation for the cleaning of art works. In 1981, Steyer left the enterprise, and the US American company Coherent, Inc. occurred as a majority partner. In 1994 the enterprise received the innovation price of the German economy. In September 2000, Lamda Physik became a stock company, as the enterprise was listed on the German exchange. In 2004, however, the US company Coherent took over Lamdy Physik by gathering more than 95% of the stocks and compensating other equity holders financially. Thus, today Lamda Physik is a subsidiary of Coherent, which currently employs 1,712 workers in the US and approximately 500 employees in Germany.

4 Data and research methodology

4.1 Survey analysis: identifying entrepreneurial scientists

In order to analyze entrepreneurial activity and their determinants we performed a survey analysis with MPS scientists in the period of mid-October to mid-December of 2007. Before performing the survey we contacted the executive directors of each institute to obtain permission to interview the scientists. The majority of directors (67 out of 78) permitted us to conduct the interviews and provided us with the necessary contact information of scientists if it was not publicly available. The basic population for the survey consisted of 7,808 scientists working in 67 institutes.

The survey was conducted by TNS Emnid GmbH, a professional opinion research institute. Trained interviewers from TNS Emnid GmbH contacted every scientist in the sample by phone. Participation in the survey was voluntary; that is, scientists could refuse to respond to any specific question or skip the entire survey. Scientists who could not be contacted after three calls were dropped from the study. From the targeted population of 7,808 scientists 2,604 responded to the survey (response rate of 33.35%). There has been no information collected about scientists in institute whose directors refused interviews as well as about scientists that rejected participation.

The survey questions were particularly designed to analyze the entrepreneurial activities of scientists. The feasibility and reliability of the survey questions were tested and improved during a pilot study, conducted in August and September 2007. Apart from entrepreneurial activities, the questions on the survey cover a broad range of individual attitudes toward commercialization, work experience, scientific competences, and social-demographic factors.

4.2 Measurement of variables

4.2.1 Dependent variable

In order to identify entrepreneurs among the Max Planck scientists, the definition of nascent entrepreneurship provided by the Global Entrepreneurship Monitor and US-American Panel Study of Entrepreneurial Dynamics was adopted (Reynolds et al. 2004a, b). Accordingly, scientists were classified as nascent entrepreneurs if they were engaged in any activity associated with starting a business. Such activities may include applying for public or private financing, seeking for venture capital, writing a business plan, looking for office space or forming the founding team. The dependent variable, nascent entrepreneurship, is binary, indicating whether the scientist is involved in start-up activities (coded as 1) or not (coded as 0).

4.2.2 Variables of main interest

  • Foreign-born is a binary variable, which is equal to 1 if a scientist is a non-German citizen; 0 otherwise.

  • Foreign-educated is a binary variable, which is equal to 1 if a scientist has received his/her Ph.D. degree from a non-German university; 0 otherwise.

4.2.3 Control variables

  • Entrepreneurial culture: Previous studies suggest that there are significant differences in entrepreneurial attitudes across countries (cf. Reynolds et al. 2005). Hence, the entrepreneurial attitude of a scientist might differ according to environment they have been socialized in. One plausible explanation for this phenomenon is entrepreneurial culture, which is defined as a national system of shared values in a particular society that embraces and supports entrepreneurship (cf. Thomas and Mueller 2000). The latter studies argue that appreciation of firm formation is shaped within the economic framework in which individuals grow up. Further, individuals are more likely to know someone who has founded a business when entrepreneurial activity in the environment is high. Thus, business environment and entrepreneurial attitude of the home country may have a sustainable effect on individual’s entrepreneurial decision and the disposition to become an entrepreneur. The stronger the impact of a society on individuals’ behavior is, the higher should be the likelihood for such differences. The same might apply to scientists who stay abroad for a while, if their behavior is influenced by entrepreneurial culture of the host country. In order to identify the pure effect of scientific mobility these effects have to be controlled for. The general entrepreneurial attitude in a scientist’s country of origin is proxied by several variables. First, we apply the Total Entrepreneurial Activity index (TEA) of the respective country of origin (TEA_country_origin) as described by Reynolds et al. (2005).Footnote 3 Second, we use three distinct indicators from the World Bank Group Entrepreneurship Database: the number of total registered corporations divided by total working age population (Business_Density_Home_Country), the number of newly registered corporations divided by total working age population (New_Density_Home_Country), and the number of newly registered corporations divided by the number of total registered corporations (Entry_Rate_Home_Country).

  • Income and Reputation: Individual norms and values might influence the decision to become an entrepreneur. In the case of scientists, Etzkowitz (2003) suggests that entrepreneurship might be a feasible option to use their scientific expertise to generate personal income. Further, Owen-Smith and Powell (2003) point out that successful commercialization may be associated with an increase of scientists’ academic reputation. In order to control for such effects the scientists in our sample were provided with the statements: “There is little, if any, money to be made from commercialization” and “Commercialization activities increase the reputation of a scientist in your scientific community”. Scientists were asked to what degree they agree to these statements, given a 5-point Likert-type scale, ranging from 1 ‘Strongly disagree’ to 5 ‘Strongly agree’. As strong agreement to the statement on financial benefits reflects that there are little pecuniary incentives, we recoded the variable income such that higher values indicate lower agreement. Reputation is coded in analogy to the answers. Thus, our two measures Income and Reputation can assume five integer values from 1 to 5.

  • Risk: The measure of risk utilized in our study is adopted from the German Social Economic Panel (GSOEP) and relates to the financial risk attitude of the scientists (Dohmen et al. 2005). Respondents hypothetically won 100,000 Euro in a lottery and were confronted with a financially risky, but yet lucrative investment. They could either invest nothing, 20, 40, 60, 80% or the entire lottery winnings. According to the answers given, the variable Risk can assume six integer values, ranging from 1 to 6. While a value of 0 indicates complete risk-aversion with no investment, a value of 5 indicates that scientists would invest the entire amount of 100,000 Euro.

  • Career stages: In order to capture variation in human capital associated with different career stages, we include measures of scientists’ research position. Thereby, four binary variables—Ph.D. student (PhDstudent), postdoctoral researcher (PostDoc), group leader or professor (Groupleader/Professor) and director (Director)—indicate the research position of the scientists. These are coded as 1 if a scientists holds a particular position and 0 otherwise.

  • Inventions: Inventive activities, in particular patents, indicate the creation of novel technological knowledge with some commercial value and are, therefore, supposed to increase the likelihood to start an own business (Stuart and Ding 2006). In order to control for this effect, a binary variable Patent is included. It takes the value of 1 if scientist has ever granted a patent and 0 otherwise.

  • Private sector work experience and entrepreneurial experience: Private sector work experience and entrepreneurial experience are typically assumed to be strong predictors of a scientist’s entrepreneurial intentions. That is because such experience, skills, and contacts to other business owners, suppliers and customers, and external finance sources are important for establishing a successful business (Shane and Khurana 2003). Two binary variables are included to control for such effects. The first variable Work_Private_Sector indicates whether a scientists have prior work experience in the private sector (coded as 1) or not (coded as 0). The second variable Entrepreneurial_Experience indicates whether a scientist has ever owned a firm (coded as 1) or not (coded as 0).

  • Gender and age: Previous literature suggests that the entrepreneurial likelihood differs with gender and age. Thus, we include a binary variable Female (coded as 1), which indicates the gender status of a scientist, and a variable Age measuring scientists’ age in years.

  • Tenure: Employment contracts of scientists at the Max Planck Society are typically restricted to 12 years. However, this restriction does not apply for directors and a relatively small group of designated postdoctoral researchers who have lifetime contracts. These scientists may well be less inclined to entrepreneurship due to their secure employment prospects. Hence, we include a binary variable Tenure to control for (potential) lower entrepreneurial propensity of tenured scientists (coded as 1) in comparison to scientists with temporary work contracts (coded as 0).

  • Research discipline: Finally, we control for variation in entrepreneurial intentions across scientific disciplines. We rely on the categorization provided by the Max Planck Society, which classifies research institutes into the three different research sections: Chemistry, Physics and Technology (Natural Sciences), Biology and Medicine (Life Sciences), and humanities (Humanities). Thus, three binary variables indicate respectively the affiliation of the scientists to a particular research section.

4.3 Sample characteristics and descriptive statistics

As noted above, the total sample population comprises 2,604 researchers. However, guest researchers (59 out of 2,604) and scientists who gave inconsistent answers are excluded.Footnote 4 Due to this exclusion, our final sample was reduced to 2,432 scientists. Based on our data, 3.3% of the Max Planck scientists (81 out of 2,432) scientists were currently engaged in activities associated with starting their own business. This share is comparable to the share of nascent entrepreneurs on the national population in Germany (Brixy et al. 2009). Table 1 presents descriptive statistics providing the mean value, standard deviation, minimum and maximum value of all variables. Moreover, a correlation matrix of all variables is provided in the “Appendix”.

Table 1 Descriptive statistics
Full size table

The descriptive statistics imply that the Max Planck Society is an ideal setting for analyzing the international career mobility of scientists. The share of foreign-born scientists in the total sample is 39.7%. 22.2% of the scientists hail from European countries (2.9% from Italy, 1.9% from France, 2.2% from Russia, 1.7% from The Netherlands, 1.6% from Austria, 1% from the UK). Further 8% of the scientists come from Asia and Oceania (with Chinese 3.5% and Indians 2.7%) and 3.8% stem from North America (with 2.3% US Americans). South Americans account for 2.4% and scientists from Africa and Middle East for 2.3%. Within the group of nascent entrepreneurs, 64.2% are foreign-born.

German scientists who received their Ph.D. degrees from foreign universities account for approximately 9.6% of German senior scientists (researchers with Ph.D. degree) and for approximately 4.9% of all German scientists. A substantial share of German scientists who graduated outside Germany received their Ph.D. degree in the USA (29.2%) or in Switzerland (20%). Further, the UK (16.9%) and Austria (9.2%), account for a relatively large share of foreign-Ph.D. degrees among German scientists. The Netherlands and Japan each hosted 4.6% of German scientists who graduated abroad.

Among all respondents, approximately 18% have worked in the private sector, on average 0.64 years. Focusing on the group of nascent entrepreneurs, 37% have private sector work experience of 1.3 years. While 5.5% of all scientists were business owners or founders in the past, 25% of the nascent entrepreneurs have entrepreneurial experience in the past. 11.6% of all surveyed scientists have ever filed for a patent. Among the scientists identified as nascent entrepreneurs the share of patentees rises to 33.8%.

While Ph.D. students are nearly equally represented in both the group of nascent entrepreneurs and the whole sample, the proportion of directors in the group of nascent entrepreneurs is relatively high. Directors account for about 2.5% in the entire sample and for approximately 10% of nascent entrepreneurs. Further, group leaders are slightly overrepresented in the group on nascent entrepreneurs. Moreover, in the group of nascent entrepreneurs there are more scientists that acknowledge that entrepreneurship may increase reputation than in the entire population. The group of nascent entrepreneurs is also—on average—less risk averse than the entire population. Finally, nascent entrepreneurs are more likely to come from research fields such as biology and medicine, and physics, chemistry and engineering.

5 Results

In this section, we report our empirical results for the likelihood of foreign-born and foreign educated scientists becoming entrepreneurs. Given the binary nature of the dependent variable indicating the nascent entrepreneurial status of a scientist, we employ logistic regression.

5.1 Foreign-educated and entrepreneurship

Table 2 presents our econometric results relating to the relationship between foreign-education and the probability of being a nascent entrepreneur. Our analysis is restricted to German scientists who already obtained their Ph.D. degree. Thus, non-Germans and doctoral students are excluded since these are not a proper reference group. We present two model specifications. In column (1) the baseline model is presented in which foreign-education is used as the sole explanatory variable. In column (2) the fully-loaded model with all control variables is presented.

Table 2 Foreign-education and the propensity to engage in academic entrepreneurship
Full size table

The estimated coefficient for foreign-educated is positive and statistically significant at the 10%-level in both, the baseline and the extended model. Thus, our results provide weak evidence in favor of our hypothesis H1, which predicted that foreign-educated scientists are more likely to become entrepreneur than their German counterparts having graduated within Germany. In line with prior studies, we find evidence that patenting and entrepreneurial experience stimulates entrepreneurial experience. Furthermore, the positive and statistically significant coefficient for our director variable indicates that such scientists have slightly higher likelihood to be nascent entrepreneur than post-doctoral researchers and groupleaders/professors. This suggests that the general level of scientific capabilities is positively associated with the propensity to become an entrepreneur. The remaining control variables have the expected sign are, however, insignificant.

5.2 Foreign-born and the propensity to engage in academic entrepreneurship

Econometric results relating to the entrepreneurial intentions of foreign-born scientists are presented in Table 3. The appropriate reference group comprises of German scientists. However, in order to test the robustness of the results, we present model specifications in which foreign-educated German scientist and Ph.D. students are excluded. Furthermore, since the choice of the most appropriate control for cultural and other effects due to scientists’ origin is ex ante not clear, the impact of our cultural control variables TEA_country_origin, Business_Density_Home_Country, New_Density_Home_Country, and Entry_Rate_Home_Country is tested in different model specifications.

Table 3 Foreign-born and the propoensity to engage in academic entrepreneurship
Full size table

Across model specifications with different reference groups and different number of controls, the results provide robust support for our hypothesis. The estimated coefficient for Foreign-Born is positive and statistically significant—mostly at the 1 and 5% level—indicating that foreign-born scientists have higher likelihood to become entrepreneur than native scientists. An important result is that the coefficient for Foreign-Born is statistically significant even after controlling for the scientific abilities of the foreign-born researchers, since we can not exclude that these are selected according to their scientific abilities. Moreover, the results for Foreign-Born seem to be robust with respect to cultural and other effects from scientists’ home country. None of the measures we apply to control for cultural and country-specific effects (TEA_Country_Origin, Business_Density_Home_Country, New_Density_Home_Country, and Entry_Rate_Home_Country) is significant. Thus, our results provide support for the notion that foreign-born scientists are more likely to become entrepreneurs.

Similar to the results obtained in the analysis of the entrepreneurial likelihood of foreign-educated, we find that scientists with prior work experience in the private sector, scientists with entrepreneurial experience, and scientists that have applied for a patent are more likely to be nascent entrepreneurs. Furthermore, our results suggest (see model specifications 3–10) that scientists reporting that commercialization of research may contribute to reputation in the scientific community and scientists with temporary contracts are more likely to be nascent entrepreneurs. However, the estimated coefficients for Reputation and Tenure are statistically insignificant when excluding Ph.D. students suggesting that these effects are likely to driven by the particular characteristics of this group.

6 Conclusions

In this paper, we investigated whether foreign-born and foreign-educated scientists are more entrepreneurial than their “domestic” peers. We find evidence that the mobility of scientists might stimulate academic entrepreneurship. Therefore, our study contributes to the policy debate regarding the benefits and costs associated with scientists’ career mobility. Our findings may be useful to policymakers and university administrators who wish to stimulate technology commercialization.

Our results integrate two existing streams of literature. First, as noted in a previous section of the paper, several authors have found that career mobility of scientists stimulates commercialization and patenting. Our study extends these findings by showing that entrepreneurial activities of academic scientists might be also stimulated by career mobility. Second, our findings relate to prior studies by Saxenian on entrepreneurial activity of highly-skilled foreigners in Silicon Valley. Saxenian (2002) finds a positive relationship between in-migration and economic development since immigrant entrepreneurs are of key importance to job and wealth creation in Silicon Valley. Our study relates to these findings by elaborating on the relation between career mobility and entrepreneurship in academic settings.

Our results also have important implications for additional research on this topic. We conjecture that the higher propensity of foreign-educated and foreign-born scientists to engage in entrepreneurship is due to their experience with different research methods and cultural environments, which broadens their scientific and social capabilities. However, given our data and analytical tools, we cannot rule out the possibility that both career mobility and entrepreneurial inclination are due to the same individual unobserved characteristics. Thus, additional research is required to tease out cause and effect more precisely. It is especially important to learn more about the nature of scientific mobility and the role of monetary and intrinsic factors in stimulating such activity. Moreover, further research is required to improve our understanding of how mobility influences human and social capital and the recognition of entrepreneurial opportunities.