Embed Size (px)
Citation preview
Ready to leave the ivory tower?: Academic scientists’appeal to work in the private sector
Michael Fritsch • Stefan Krabel
Published online: 5 June 2010� Springer Science+Business Media, LLC 2010
Abstract This study investigates the factors that shape the attitudes of scientists toward
starting their own business or working in a private sector firm. The analysis is based on
data collected from scientists working in the German Max Planck Society, a research
institution devoted to basic science. We find that the scientists’ attractiveness of working in
a private sector firm or of starting their own business differ considerably according to their
academic discipline and the self-reported commercial potential of their research. The
ability to take risks, prior work experience in private firms, and personal experience in
cooperating with industry lead to a positive attitude towards switching to private sector
employment or entrepreneurship. Strong willingness to freely distribute research findings is
related to a low appeal of private sector work.
Keywords Knowledge transfer � Science � Entrepreneurship � Innovation �Commercialization
JEL Classification O31 � O33 � L26 � L32
M. FritschSchool of Economics and Business Administration, Friedrich Schiller University, Carl-Zeiss-Strasse 3,07743 Jena, Germanye-mail: [email protected]
S. Krabel (&)Institute of Economics, Economic Policy Research, University of Kassel, Nora-Platiel-Str. 5,34109 Kassel, Germanye-mail: [email protected]
123
J Technol Transf (2012) 37:271–296DOI 10.1007/s10961-010-9174-7
‘‘… one of the strongest motives that lead men to art and science is escape fromeveryday life with its painful crudity and hopeless dreariness, from the fetters ofone’s own ever-shifting desires. A finely tempered nature longs to escape frompersonal life into the world of objective perception and thought.’’ (Albert Einstein)1
1 Introduction
Scientists have often been considered, in the notion of Albert Einstein, as people who try to
escape the ‘real world’ by hiding in the ivory tower of science. Recognizing that basic
science is an important source of industrial innovation politicians, however, have
increasingly contradicted to this view. Over the past decades universities have been
required to transform from ‘ivory towers’ to entrepreneurial enterprises (Etzkowitz 1998;
Powers et al. 2005). Scientists are supposed to make their research relevant to the private
sector and to effectively transfer their knowledge and research results (Mansfield 1995;
Dietz and Bozeman 2005). One of the challenges of linking results of academic research to
industry is the tacit nature of knowledge (Pavitt 1998). Scientific achievements are based
on competences and information ‘embedded’ in scientists, so that research results can only
to a limited extent be transferred via publications or patents (Agrawal and Henderson
2002). Therefore, the transition of academic researchers to private sector employment is
increasingly in the focus of a policy as well as of a scholarly debate (see e.g. Link and
Siegel 2007; OECD 2002; Noll 1998).
While large shares of scientists’ reveal that industry is interested in their research
(Libaers 2009), only few studies examine job migration between academia and industry.
Existing studies of scientists’ job movements into the private sector mainly focus on
academic entrepreneurship.2 This stream of literature emphasizes that scientific expertise
(Shrader and Siegel 2007), as well as contacts to industry (Shane and Khurana 2003) and
venture capitalists (Toole and Czarnitzki 2007) are conducive to the emergence and the
success of academic start-ups. Furthermore, the evidence suggests that individuals’ career
mobility from academia to private sector employment partly depends on their research field
(Mangematin 2000; Martinelli 2001). However, the factors that shape scientists’ appealand readiness to switch to the private sector—which can be regarded a first step of a switch
towards industry employment or entrepreneurship—are rarely discussed.
The present study aims to shed more light on the factors that shape scientists’ appeal to
work in private sector firms or to start their own venture. Our analysis is based on survey
interviews conducted with more than 2,000 scientists working at research institutes of the
German Max Planck Society (MPS). Using this survey data, we particularly examine
whether scientists’ self-reported attractiveness scores to work in the private sector and to
start their own firm are related to commercial research alignment, peer influence, and
commercialization experience. We deliberately focus on individuals’ attractiveness as this
measure reflects the allurement of private sector work and entrepreneurship. Our approach
enables us to identify factors that shape scientists readiness to work in the private sector or
to start a company—regardless of whether or not being followed by factual transition. The
1 A. Einstein, ‘Autobiographical notes’, in: A. Einstein: Philosopher-Scientist. 1949. Everton (Ill.), ed. P. A.Shilpp.2 Further studies on scientists’ job movements to the private sector deal with skills gained in academia thatare useful in the private sector (e.g. Zellner 2003) and the impact of academically trained personnel inprivate firms (Herrera et al. 2010).
272 M. Fritsch, S. Krabel
123
results are particularly relevant for policy and transfer institutions for targeting that group
of individuals which is relatively open towards job migration.
Our results suggest that scientists who work in fields with commercial orientation are
more attracted to working in the business sector or to start their own firm. Experience in
cooperative research projects with private partners enhances scientists’ appeal to work in
the private sector and entrepreneurship while patenting does not yield higher attractiveness
levels. Personal characteristics such as nationality, having a tenured position, or a positive
attitude towards risk are found to have significant influence on scientists’ attractiveness to
work in the private sector and to start an own venture. Comparing the determinants of
entrepreneurial attractiveness with the factors that influence early start-up activity (nascent
entrepreneurship) we find some considerable differences. While entrepreneurial attrac-
tiveness differs considerably according to scientists’ area of research, the respective effect
of research field on nascent entrepreneurship is rather weak.3
The remainder of the paper is structured as follows. Section 2 develops hypotheses
predicting why commercial research alignment, peer influence and commercialization
experience shape scientists’ appeal of working in the private sector or of starting their own
company is influenced by the nature of their research and their experience in the com-
mercialization of scientific results. We then provide some information on the Max Planck
Society which constitutes the framework of our empirical analysis (Section 3) and give
some basic information about sample characteristics and the measurement of variables
(Section 4). The results of our analysis are presented in Section 5. Section 6 discusses the
results and concludes.
2 Determinants of scientists’ appeal to entrepreneurship and work in private firms
Research on the importance of different knowledge transfer channels between science and
industry suggests that personal interaction of academic scientists and private sector firms is
a key element of successful transfer (e.g. Cohen et al. 2002; Cohen et al. 1998; Agrawal
and Henderson 2002). Particularly, employment of academic scientists is found to increase
firms’ ability to transform scientific inventions into new market products (Agrawal 2006;
Herrera et al. 2010). The latter studies indicate that innovating firms benefit from the
knowledge of academically-trained personnel. These results relate to the notion that sci-
entific knowledge often comprises important tacit components which can be most effi-
ciently conveyed by scientists’ personal involvement in applying and commercializing
scientific knowledge in the private sector (Pavitt 1998; Rosenberg 1990). Thus, despite of
potential costs due to scientists’ adaption to private sector work, firms may benefit from
employing academics—especially when firms try to innovate.
In view of the evidence that academic scientists accepting jobs in private firms repre-
sents a crucial source of knowledge and technology transfer, surprisingly few studies have
investigated scientists’ willingness to switch from academia to the private sector. A rare
exception is a study by Stern (2004) which revealed that the majority of scientists prefer to
stay in the science sector even when being offered higher paid jobs in private firms. This
finding suggests that non-pecuniary determinants also shape scientists’ interest to work in
the private sector.
3 This supports the supposition that attractiveness does not necessarily coincide with intention (see e.g.Ajzen 1991).
Ready to leave the ivory tower? 273
123
2.1 Commercial research focus, institutional peers and their relation to scientists’
appeal of entrepreneurship and work in private firms
Job migration of academic scientists is found to be most frequent in those areas of research
which promise to have some commercial potential (Martinelli 2001; Zellner 2003). After
completing their Ph.D. degree, relatively more researchers working in fields such as
engineering, information technology or biotechnology obtain a job in the private sector as
compared to researchers in humanities and social sciences (Martinelli 2001). This suggests
that scientists’ research orientation relates to their appeal to work in the private sector.
Scientists who perceive that their research has commercial potential may be aware of the
practical relevance of their work and regard application of their research as an attractive
opportunity. Therefore, they may consider it as an attractive idea to leave academia in
order to transform their ideas into marketable products. Moreover, scientists might be more
inclined to enter private sector employment if they expect that the subject matter of their
work and their routines would not drastically change. Hence, we hypothesize that a sci-
entist’s attractiveness to work in the private sector is related to scientist’s perception of the
commercial potential of their own or of related research. The larger the perceived potential
of their research for commercial application, the higher the appeal of working in the private
sector (H1a). We also predict that this perception is positively related to scientists’ pro-
pensity to start an own firm. Several studies report that scientists who become entrepre-
neurs often base their firm on scientific inventions (see e.g. Shane and Khurana 2003;
Buenstorf 2009). One might, therefore, expect that setting-up an own firm is a particularly
attractive option for scientists whose research is closely related to industrial needs (Gitt-
elman 1999). Therefore, similar to our argumentation regarding the appeal to scientists of
private sector work, we predict that the attractiveness of starting their own firm is posi-
tively related to the perception that their research has commercial potential (H1b).
H1a Scientists’ perception that their own or related research has commercial potential is
positively related to their attractiveness of working in the private sector.
H1b Scientists’ perception that their own or related research has commercial potential is
positively related to their attractiveness of starting their own company.
Apart from the potential influence of the commercial orientation of their research,
scientists’ appeal of working outside academia may also be shaped by the respective
attitudes of their colleagues. If scientists’ institutional peers value working in the private
sector as relatively attractive, scientists may be likely to adopt a similar attitude. This may
be the case for two reasons. First, peers’ mean attractiveness of working in the private
sector or to start an own firm can be regarded to be an indicator for the commercial
orientation of research performed at the respective institute. Some research institutes orient
their agenda towards commercial requirements while others hardly focus on linking their
scientific findings to industrial needs (Shane 2004). In view of such differing research
agendas it is not surprising that high numbers of university spin-offs are observed in only
few universities (Wright et al. 2007). The latter study as well as Charles and Conway
(2001) argue that commercial research orientation at universities and research institutes
also shapes the institutional culture towards application of scientific findings. Frequent
commercialization activities may lead the majority of scientists at institutes to be open
towards entrepreneurial activity, commercialization of scientific results and working with
or for private sector firms. Mean attractiveness levels of institutional peers may serve as a
proxy for such an institutional culture. Hence, scientists’ own appeal to work in the private
274 M. Fritsch, S. Krabel
123
sector may be related to the perceived attractiveness level of colleagues working at the
same institute.
A second reason why peers’ attitudes may relate to an individual’s assessment of the
attractiveness of working in the private sector is that individuals tend to adopt attitudes of
colleagues. Social psychology studies emphasize that colleagues and peers are regularly
regarded as a reference group. Prevalent opinions in this reference group affect individual
decision-making due to a perceived pressure to conform to commonly stated opinions
within this group (Turner 1990; Sherif 1935). Several studies confirm this phenomenon in
the context of commercialization activities of scientists.4 Though not mutually exclusive,
we assume that both potential influences, research orientation of the respective institute
and peer effects, may lead to a positive relationship between institutional attractiveness
levels and scientists’ individual appeal toward working in the private sector (H2a).
Assuming that such a relationship also holds in the context of venture creation, hypothesis
H2b predicts that scientists’ appeal of entrepreneurial activity is positively related to the
respective attitude of institutional peers.
H2a The attractiveness of working in the private sector of institutional peers is positively
related to individual scientists’ attractiveness of working in the private sector.
H2b The attractiveness of becoming entrepreneurs of institutional peers is positively
related to individual scientists’ attractiveness of becoming entrepreneurs.
2.2 Commercialization experience and its relation to scientists’ appeal
to entrepreneurship and to work in private firms
We assume that scientists with commercialization experience value private sector work
and entrepreneurship as relatively attractive. Commercialization of research results reflects
scientists’ ability to link scientific findings to industrial needs (Perkmann and Walsh 2007).
Observing that their research output is suited for application in the private sector may
stimulate scientists’ appeal to progress their ideas while being employed in firms. More-
over, even if scientists do not intend to follow specific research ideas in the business sector
they may be more inclined to working in private firms when having patenting experience.
Toole and Czarnitzki (2009) as well as Azoulay et al. (2006) highlight that scientists’
patenting productivity may be regarded as an indicator of their commercial research ori-
entation. Similar to the aforementioned argument leading to hypothesis H1a we assume
that commercial research orientation is positively related to scientists’ appeal to work in
private firms. Thus, we predict that scientists with patenting experience value private sector
work as more attractive than scientists without patents (H3a). Furthermore, patents have
been shown to be a robust predictor of scientists’ entrepreneurial activity (Azoulay et al.
2006; Stuart and Ding 2006). This finding may indicate that the legal appropriation of
innovative ideas in form of patents tends to increase a scientist’s readiness to start a
company based on these ideas. Accordingly, we hypothesize that patenting activity is
positively related to scientists’ attraction to starting their own company (H3b).
H3a Patenting is positively related to scientists’ attraction to work in the private sector.
4 According to Stuart and Ding (2006), scientists are more likely to become entrepreneurs if peers at theirinstitute are involved in commercial science. Bercovitz and Feldman (2008) show that organizationalinfluence stimulates the entrepreneurial activity of scientists as scientists tend to learn to adopt the behaviorof peers.
Ready to leave the ivory tower? 275
123
H3b Patenting is positively related to scientists’ attraction to start their own company.
Another way for academic scientists to foster the commercial application of their research
results is through joint research projects with private firms. Scientists with experience in such
cooperation tend to be aware of the potential benefits and problems of trying to transform
research results into marketable products. This argumentation relates to a study by Lin and
Bozeman (2006) who argue that diverse sources of technical and human capital increase
scientists’ innovative capacity as they are more likely to (re-)combine different parts of
knowledge. Moreover, cooperative research with private firms may stimulate a scientist’s
orientation of his further research towards the demands of these firms (Segarra–Blasco and
Arauzo–Carod 2008). Collaboration with firms can particularly increase scientists’ aware-
ness of the usefulness of their individual skills and knowledge in the private sector (Thune
2007). We therefore predict that experience in research cooperation with private firms is
positively related to scientists’ attractiveness of working in the private sector (H4a).
Experience in research cooperation with private firms may also have a positive effect on
scientists’ appeal to start their own company. This may be explained with relatively high
prospects of success if firm founders have both scientific-technical and industry-specific
knowledge (Colombo and Grilli 2005). Agarwal et al. (2004) describe the ability to
evaluate the commercial potential of research results as market–pioneering know-how.
Combined with technological knowhow it is the key capability necessary to seize market
opportunities. Thus, we hypothesize that experience in research cooperation with private
firms is positively related to scientists’ attractiveness in starting their own company (H4b).
H4a Experience in research cooperation with private firms is positively related to sci-
entists’ attractiveness to work in the private sector.
H4b Experience in research cooperation with private firms is positively related to sci-
entists’ attractiveness to start their own company.
3 Max Planck Society scientists: framework of the study
3.1 The Max Planck Society
The Max Planck Society (MPS) is a German research association which was initially
founded in the year 1911 as the Kaiser Wilhelm Society. In 1948 the association adopted
its current name. The Society consists of 78 research institutes and three additional
research facilities in Germany that perform basic research in the natural sciences, life
sciences, social sciences, and the humanities. Approximately 10,200 scientists are
employed by Max Planck Plank Institutes, including professors, post-doctoral fellows,
doctoral students, and guest scientists (as of 31.12.2008). Researchers are supplemented by
roughly 3,000 non-scientific employees responsible for administration and research
assistance (Max Planck Society 2009). Around 82% of MPS expenditure is met by public
funding from the Federal Government and the German States. The remaining 18% stems
from donations, member contributions, and from funded projects. In 2008, the total budget
of the MPS accounted for 1.4 billion Euro.
MPS institutes focus on basic research. They are meant to take up new and innovative
research areas that German universities are not in a position to accommodate or deal with
adequately. Thus, research at Max Planck Institutes complements the work of universities
and other research facilities in relevant fields. The outcome of research conducted at MPS
276 M. Fritsch, S. Krabel
123
institutes is quite distinguished, as is documented by 32 Nobel Prizes awarded to the MPS
researchers since the society’s foundation.
There are a number of spin-off companies founded by MPS researchers. The success of
some of these companies reveals that basic research can have commercial applications with
high market potential. In order to support the transfer and commercialization of technol-
ogy, the MPS maintains a distinct technology transfer office, Max Planck Innovation,
which is responsible for all 78 institutes. This transfer office is responsible for the provision
of professional services and assistance for technology-based spin-off companies from the
MPS. Such services include the assessment of the commercial potential of a technology
and assistance in the creation of a business plan as well as assistance in searching for
potential financiers (venture capital companies, banks and business angels). However, the
MPS does not invest capital in its spin-offs. According to its records, the MPS transfer
office has coached 86 spin-off companies since 1990, when it began its professional
support of spin-offs. Max Planck Innovation lists 86 currently existing spin-off companies
which operate in high-tech industries such as biotechnology, biochemistry and physical
engineering. Altogether, these companies presently employ circa 2,260 people.
3.2 Data generation: the Max Planck scientist survey
Our data is based on a survey conducted in MPS institutes between mid-October and 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. We finished with 2,604
interviews, denoting a response rate of 33.35%.
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; scientists could refuse to respond
to any specific question or to skip the entire survey. Scientists who could not be contacted
after three calls were dropped from the study. The survey questions were particularly
designed to analyze whether or not scientists perceive entrepreneurship and research coop-
eration with firms as attractive and whether scientists actively engage in commercialization
of scientific research results. The feasibility and reliability of the survey questions were tested
and improved during a pilot study, conducted in August and September 2007. The questions
on the survey cover entrepreneurial attractiveness and business ownership experience as well
as experience in commercialization activities such as patenting or consulting. Additionally,
the survey contains questions regarding scientists’ individual attitudes toward commer-
cialization activities as well as questions on research experience, industrial experience,
education, socio-demographic and idiosyncratic characteristics as well as risk-attitude.
4 Data and estimation approach
4.1 Dependent variables: entrepreneurial attractiveness and attractiveness of work
in the private sector
The question ‘‘To what degree is working in the business sector an attractive idea to you?Would you say… not attractive at all; not attractive; neutral; attractive; highly
Ready to leave the ivory tower? 277
123
attractive?’’ tried to assess scientist’s attractiveness of working in the private sector.
Therefore, our dependent variable is of ordinal character and can assume five different
values from 1 to 5, such that higher values represent increasing attractiveness. Thus, a
value of 1 represents that work in the private sector is regarded not attractive at all while a
value of 5 denotes the highest level of attractiveness. In analogy, entrepreneurial attrac-
tiveness was raised by asking scientists: ‘‘To what degree is starting your own businessattractive to you? Would you say … not attractive at all; not attractive; neutral; attractive;highly attractive?’’ Again, this variable was coded with values from 1 to 5 with higher
values denoting higher attractiveness.
In order to compare our models of entrepreneurial attractiveness with models predicting
new business formation, we relate our analysis to a prior study by Krabel and Mueller
(2009) that analyzed nascent entrepreneurship within the same sample. Nascent entre-
preneurship is a concept introduced by the American Panel Study of Entrepreneurial
Dynamics (Reynolds et al. 2004). According to this concept, scientists were classified as
nascent entrepreneurs if they were engaged in any activity associated with starting a
business on the day of the interview. These activities may include applying for public or
private financing, seeking venture capital, writing a business plan, looking for office space,
or forming a founding team. Therefore, the dependent variable was dichotomous, indi-
cating whether the scientist is involved in start-up activities (coded as 1) or not (coded as 0)
(see Krabel and Mueller 2009, for details).
4.2 Covariates
According to our hypotheses as outlined in Section 2 we include the following information
about the nature of a scientist’s research, the assessment of peers working at the same
research institute and commercialization experience into the model.
• Nature of research: Three variables measure the nature of research, comprising
scientists’ own assessment of the commercial potential of their research as well as a
classification of their research field. The survey included the following statements on
commercialization activities within the scientists’ research community: ‘‘Commercial-ization activities are common in your field of research’’ and ‘‘Your research groupfocuses on basic research, which is not suitable for commercialization.’’ We again
provided the aforementioned 5-point Likert-type scale, ranging from ‘stronglydisagree’ to ‘strongly agree.’ These two measures identify scientists’ personal
perception of the commercial potential of their own or of related research and are used
to test our hypotheses H1a and H1b. In order to disentangle scientists’ perceptions from
research field effects, we also include three binary variables indicating if scientists
belong to the Life Science Section, the Natural Science Section or to the Humanities.
Variables take a value of 1 if scientists belong to the respective research section and are
coded 0 otherwise.
• Peer effects: Two variables are supposed to capture the potential influence of peers on
the attractiveness of working in the private sector and of entrepreneurship. One of these
variables is the mean entrepreneurial attractiveness stated by colleagues working at the
same institute. We excluded the respective scientists’ own assessment from this
computation to ensure that the variable measures only the assessment in the working
environment and not the respondent’s own evaluation. In the same manner, we
computed the mean attractiveness of working in the business sector at each institute,
again excluding scientists’ own evaluation.
278 M. Fritsch, S. Krabel
123
• Commercialization experience: Two binary variables indicate whether or not a scientist
has experience in research cooperation with private firms and whether or not scientists
have ever applied for a patent (yes = 1; no = 0). These two variables allow us to
investigate our predictions that experience with commercialization of scientific results
are positively related to his assessment of working outside academia.
A number of further variables were included that potentially influence the attractiveness
of working in the private sector and of starting one’s own firm. Our control variables
include information on scientific position, perception of open science, and work history as
well as personal and demographic variables.
• Doctoral degree: In order to control for differences between Ph.D. students and senior
researchers with doctorate degrees, we include a binary variable assuming a value of 1
if a scientist has already obtained a doctorate and a value of 0 to indicate that a scientist
has not yet completed a doctoral degree.
• Director position: Potential effects resulting from director positions are captured by
including a variable indicating whether or not a scientist holds a director position
(1 = yes, 0 = no). Directors of Max Planck Institutes are prestigious scientists who
can hardly advance in academia from their current position. This variable controls for
influences of both status and research productivity which have been found to be
conducive to entrepreneurship in earlier research (Lowe and Gonzalez-Brambila 2007;
Zucker et al. 1998).
• Perception of open science: In order to examine if and to what degree scientists’
attitudes towards the notion that science should be a public good that is freely available
to anyone plays a role with regard to the attractiveness of working in the private sector
or starting an own a firm, we include a measure of this open science attitude in our
analysis. This measure is based on the degree to which scientists agreed to the
following statement: ‘‘Your research results should be freely accessible to any otherresearchers and businesses.’’ Scientists were asked to agree or disagree with these
statements, based on a five-point Likert-type scale, ranging from ‘strongly disagree’ to
‘strongly agree’. We coded the variables with a value of 1 when scientists strongly
disagreed with the statement while a value of 5 indicates strong agreement. This control
variable for scientists’ attitudes towards free diffusion of scientific knowledge is
included as scientists who intend to pursue an academic career may place particular
importance on widespread research results.
• Work experience in the private sector: Two further control variables capture prior work
experience in the private sector as business owners or as employees in private firms.
Previous business ownership experience should predict further entrepreneurial activity
(Shane 2004; Stuart and Ding 2006). Our indicator of such experience is a binary
variable with the values 1 (=yes) and 0 (=no) which includes the possibility that the
respondent is currently engaged in an own firm. Work experience in the private sector
prior to occupation at Max Planck can be conducive for subsequent entrepreneurship
because it may indicate contacts to actors such as business owners, suppliers,
customers, and financiers (Shane and Stuart 2002). To control for such effects, a
variable is included denoting the number of years worked in the private sector.
• Personal and socio-demographic characteristics: Our empirical models also account
for gender, age, nationality and individual risk attitude. The measure of risk aversion is
adopted from the Socio-Economic Panel in Germany (Wagner et al. 2007; Dohmen
et al. 2005). Respondents were told that they have hypothetically won 100,000 Euro in
a lottery and are faced with the chance of a risky but lucrative investment. They could
Ready to leave the ivory tower? 279
123
either invest nothing, 20, 40, 60, and 80% or their entire lottery winnings. According to
the answers given, our risk variable takes six integer values from 0 to 5. A value of 0
denotes that the respondent would not invest anything and a value of 5 means that the
scientist would invest the entire winnings. Two further binary variables indicate
scientists’ genders (female = 1, male = 0) and scientists’ nationalities (1 = German,
0 = foreign). Scientists’ ages are taken into account by two variables denoting age in
years and the squared value of age in years. Finally, we include a binary control
variable regarding whether scientists have a tenured work contract (1 = tenured,
0 = not tenured). Tenured work contracts may also be an important factor shaping
scientists’ attraction to jobs outside academia as scientists in the Max Planck Society
only achieve a tenure position if they have accomplished an outstanding research
record. Thus, tenured scientists may have selected themselves (or were selected) to
pursue an academic career.
4.3 Sample characteristics and descriptive statistics
As interviewed scientists were allowed to skip any specific question, the following analysis
is restricted to the 2,331 scientists who have answered all questions pertaining to our
variables of interest. All sample characteristics reported here pertain to this group. Among
these scientists, 1,127 are doctoral students and 1,204 postdoctoral researchers, including
58 directors. Note that the majority of scientists in the MPS have temporary working
contracts. Only 25.91 per cent of the senior researchers with Ph.D. are tenured (312 out of
1,204). Among scientists without Ph.D. the share of tenured scientists is 9.23 per cent (104
out of 1,127).
A relatively high share of scientists in our sample values work in the private sector as an
attractive career opportunity. Around 40% of all scientists assess work in the private sector
as ‘attractive’ of ‘highly attractive’. Distinguishing between Ph.D. students and senior
researchers, we find considerable differences with regard to the appeal to work in the
private sector. Among Ph.D. students this share exceeds 50%, while less than 30% of
postdoctoral researchers find it attractive to work in the business sector (Fig. 1). Regarding
0
5
10
15
20
25
30
35
40
1 - not attractive atall
2 - not attractive 3 - neutral 4 - attractive 5 - highly attractive
Attractiveness measure
Per
cent
age
of a
nsw
ers
(%)
PhD students Senior Researchers
Fig. 1 Distribution of job attractiveness in private sector by doctorate
280 M. Fritsch, S. Krabel
123
scientists’ appeal towards entrepreneurial activity, we find that approximately 28% of the
scientists regard starting an own firm as ‘attractive’ or ‘highly attractive’. Mean attrac-
tiveness of entrepreneurial activity is only slightly higher among Ph.D. students as com-
pared to postdoctoral researchers (Fig. 2). Among doctoral students, 29.8% value
entrepreneurial activity as ‘attractive’ or ‘highly attractive,’ while 26.4% of postdoctoral
scientists report such high values.
It is noteworthy that more than half of the Ph.D. students in life sciences as well as in
physics, chemistry and technical sciences reported work in the business sector as either
‘attractive’ or ‘highly attractive.’ Comparing the share of Ph.D. students and postdoctoral
researchers with high attractiveness of working in the private sector it clearly stands out
that jobs in the private sector are more attractive for Ph.D. students. This relates to the
aforementioned finding that the completion of the doctorate represents a natural point in
academic careers for switching to a occupation in the private sector. A similar pattern can
be detected regarding entrepreneurial attractiveness. It can be seen that in all research
sections the share of Ph.D. students who reported high entrepreneurial attractiveness is
always higher than the respective share of scientists holding a doctorate degree (see
Table 1).
The measures of entrepreneurial attractiveness and attractiveness of working in the
private sector show a significant positive correlation coefficient of 0.33 (Table 2). It is
noteworthy that the correlation coefficient between our measure of entrepreneurial
attractiveness and the variable indicating nascent entrepreneurship is quite low (0.182).
Such a low correlation suggests that the two variables of entrepreneurial attractiveness and
nascent entrepreneurship reflect different issues which, therefore, should be explored
separately. Relatively high negative correlation (-0.423) can be found between the indi-
cators of basic research focus (‘research in group not suitable for commercialization’) and
the commonness of commercialization in a scientists’ research field (‘commercialization iscommon in my field of research’). Unsurprisingly, scientists working in basic research
place a stronger weight on the statement that ‘‘research results should be freely accessi-ble’’. Remaining correlations of individual perceptions of nature of research (C4–C6) are
weak or insignificant.
0
5
10
15
20
25
30
35
1 - not attractive atall
2 - not attractive 3 - neutral 4 - attractive 5 - highly attractive
Attractiveness measure
Per
cent
age
of a
nsw
ers
(%)
PhD students Senior Researchers
Fig. 2 Distribution of entrepreneurial attractiveness by doctorate
Ready to leave the ivory tower? 281
123
4.4 Model specification
We separately analyze the determinants of scientists’ attraction to work in the private
sector and their attraction to entrepreneurial activities. Since the two attractiveness mea-
sures used as dependent variables are of ordinal character (5 point scales), we apply an
ordered probit model. As our empirical analysis includes the effects of attractiveness of
work in the private sector as well as entrepreneurial attractiveness among peers, the
number of cases is reduced to 2,328 scientists (attractiveness of work in the private sector)
and 2,324 scientists (entrepreneurial attractiveness) due to missing values on the assess-
ment of institutional peers that did not allow us to compute the variable for peer effects in
seven (entrepreneurial attractiveness) and in three (attractiveness to work in the private
sector) institutes, respectively.
Table 1 Descriptive statistics: Attractiveness of work in business sector
Ph.D. students Postdoctoral researchers t-test
Mean Standarddeviation
Numberof cases
Mean Standarddeviation
Numberof cases
Total sample 3.485 1.103 1,127 2.862 1.108 1,201 ***
Life science section 3.557 1.042 535 2.958 1.150 505 ***
Chemistry, physics & technology section 3.521 1.135 505 2.868 1.071 592 ***
Humanities section 2.839 1.077 87 2.356 0.975 104 ***
Ph.D. students Postdoctoral researchers t-test
Share of scientists with high attractiveness of working in the business sector
Total sample 601 (1.127) 53.33% 340 (1.201) 28.31% ***
Life science section 303 (535) 56.64% 165 (505) 32.67% ***
Chemistry, physics & technology section 276 (505) 54.65% 163 (592) 27.53% ***
Humanities section 22 (87) 25.29% 12 (104) 11.54% **
Ph.D. students Postdoctoral researchers t-test
Mean SD N Mean SD N
Entrepreneurial attractiveness: distribution
Total sample 2.804 1.187 1,125 2.652 1.193 1,199 ***
Life science section 2.925 1.198 535 2.804 1.221 506 *
Chemistry, physics & technology section 2.744 1.178 504 2.594 1.175 589 **
Humanities section 2.507 1.067 86 2.24 1.029 104 n.s.
Ph.D. students Postdoctoral researchers t-test
Share of scientists with high entrepreneurial attractiveness
Total sample 336 (1.125) 29.87% 318 (1.199) 26.52% *
Life science section 178 (535) 33.27% 156 (506) 30.83% n.s.
Chemistry, physics & technology section 144 (504) 28.57% 150 (589) 25.47% n.s.
Humanities section 14 (86) 16.28% 12 (104) 11.54% n.s.
The asterisks *, ** and *** report significance at the 10, 5 and 1% level, respectively
282 M. Fritsch, S. Krabel
123
The analysis is performed for the entire sample as well as for scientists who have not
obtained their Ph.D. degree yet and senior researchers with doctorate, separately. This
allows us to disentangle determinants of attractiveness values in different career stages.
Such a distinction may be important as senior researchers have already made a decision to
stay in academia after completing their doctorate. They may, therefore, have higher bar-
riers to and different incentives for considering a job in the private sector. In contrast,
doctoral students may face the pressure to search for job opportunities outside academia
after finishing their dissertation, as positions in science are scarce (Schomburg and Teichler
2006). Thus, the distinction between researchers without a Ph.D. and senior scientists may
be very helpful for analyzing the attractiveness of entrepreneurship or of working in a
private firm.
In order to assess to what extent determinants of entrepreneurial attractiveness are
similar to determinants of entrepreneurial action, we provide additional estimations ana-
lyzing scientists’ likelihood of becoming nascent entrepreneurs. We thereby build upon
results of a prior study by Krabel and Mueller (2009) on nascent entrepreneurship using the
same dataset. As in this study, we apply rare event logistic estimations to analyze scien-
tists’ likelihood of becoming nascent entrepreneurs. The models include the same set of
explanatory variables so that we are able to compare the results for entrepreneurial
attractiveness and nascent entrepreneurship. Thus, we are able to provide a comprehensive
picture of entrepreneurial attractiveness to scientists and their choices to start up compa-
nies. Finally, we perform several tests in order to analyze to what degree our results may be
Table 2 Correlation Matrix
C1 C2 C3 C4 C5 C6 C7 C8 C9
Correlations of individual perceptions and commercialization experience
C1 Attractiveness towork in businesssector
1
C2 Entrepreneurialattractiveness
0.330* 1
C3 Nascententrepreneur
0.065* 0.182* 1
C4 Commerciali-zation is commonin field of res.
0.136* 0.204* 0.111* 1
C5 Researchnot suitable forcommercialization
-0.114* -0.212* -0.068* -0.423* 1
C6 Research resultsshould be freelyaccessible
-0.138* -0.126* -0.098* -0.158* 0.232* 1
C7 Patent applied orfiled for
-0.076* 0.070* 0.151* 0.174* -0.191* -0.102* 1
C8 Experience inR&D cooperationwith industry’’
0.010 0.075* 0.100* 0.207* -0.221* -0.115* 0.305* 1
C9 Doctorate -0.272* -0.065* -0.003 -0.004 -0.033 0.084* 0.196* 0.164* 1
Notes: Pairwise correlations are reported. An asterisk (*) reports significance at the 1% level
Ready to leave the ivory tower? 283
123
effected by a selection bias. Thereby, we empirically analyze scientists’ willingness to
partake in further surveys. This analysis allows an identification of factors that potentially
lead to a selection bias.
5 Results: determinants of scientists’ appeal to work outside academia
5.1 Attractiveness of working in the business sector
The results of the ordered probit models suggest that scientists’ perception of commer-
cialization as common in their field of research is positively related to their assessment of
the attractiveness of work in the private sector (Table 3) while a focus of the research
group on basic research is negatively related to this attractiveness. These relationships are
both statistically significant at the 5% level in the entire sample (columns 5 and 6) and
significant at the 1% level in the subsample of senior researchers (columns 3 and 4) who
have already completed their doctorate. This evidence suggests support for hypothesis H1a
which predicted a positive relationship between scientists’ perceptions that their research
has commercial potential and their attraction to work in the private sector. Scientists’
attitude towards open science (‘‘Research results should be freely accessible’’) is nega-
tively related to attractiveness values, which indicates that research orientation and the
general attitude towards commercialization affects scientists’ appeal of working outside
academia. Again, these findings are mainly driven by the senior researchers and are not
statistically significant in the subsample of researchers without Ph.D. degree (see columns
1 and 2). Scientists in the humanities value working in the private sector as significantly
less attractive than their peers in life sciences and natural sciences.
The mean assessment of the attractiveness of work in the private sector by institutional
peers is significantly and positively related to the individual assessment supporting our
hypothesis H2a. However, this result is driven by the subgroup of Ph.D. students while no
significant relationship can be found among senior researchers. With regard to commer-
cialization experience, we find that holding a patent has no significant influence on a
scientist’s assessment of attractiveness to work in the private sector so that hypothesis H3a,
which predicted a positive relationship between patenting and scientists’ appeal to work in
the private sector, is clearly rejected. Prior experience in R&D cooperation with private
firms is positively related to attractiveness of working in the business sector in the entire
sample as well as within both subgroups. Thereby, our evidence is in support of hypothesis
H4a, which predicted such a positive relationship.
As is already apparent from the descriptive statistics, working in the business sector is
significantly more attractive to doctoral students than for senior researchers. Including a
binary variable indicating a doctoral degree reveals that the difference in attractiveness
levels is significant at the 1% level. Age has a positive influence on Ph.D. students’
attraction to work in the private sector while we find a negative effect of age for senior
researchers. This corresponds to results of a study by Mangematin (2000), which finds that
the completion of a doctoral thesis is a natural point in time for a change of career tracks.
Directors as well as scientists with tenured working contracts find it significantly less
attractive to work in the business sector. While gender is not significantly related to the
attractiveness of working in the private sector, German doctoral students are found to
perceive work in the private sector significantly more appealing than non-German Ph.D.
students. This result may well be explained by a self-selection process: Foreign doctoral
284 M. Fritsch, S. Krabel
123
Ta
ble
3R
esult
so
fo
rder
edp
rob
ites
tim
ates
of
bu
sin
ess
sect
or
wo
rkat
trac
tiv
enes
s
Ph.D
.st
uden
tsS
enio
rre
sear
cher
sE
nti
resa
mple
(1)
(2)
(3)
(4)
(5)
(6)
Do
ctora
ld
egre
e–
––
–-
0.3
28
**
*(0
.07
4)
-0
.32
8*
**
(0.0
74)
Lif
esc
ien
ce-
0.0
73
(0.0
8)
–-
0.0
25
(0.0
69)
–-
0.0
27
(0.0
47)
Ch
emis
try
,p
hy
sics
and
tech
no
log
y–
0.0
73
(0.0
8)
–0
.02
5(0
.06
9)
0.0
27
(0.0
47)
Hu
man
itie
s-
0.3
60
**
(0.1
61)
-0
.28
7*
(0.1
62)
-0
.34
8*
*(0
.13
5)
-0
.32
3*
*(0
.14
3)
-0
.35
8*
**
(0.0
94)
-0
.33
1*
**
(0.0
96)
Res
earc
hco
nte
xt
Com
mer
cial
izat
ion
isco
mm
on
infi
eld
of
rese
arch
(5-p
oin
tra
tin
gsc
ale,
1–
5)
0.0
21
(0.0
36)
0.0
21
(0.0
36)
0.1
08
**
*(0
.03
2)
0.1
08
**
*(0
.03
2)
0.0
64
**
*(0
.02
4)
0.0
64
**
*(0
.02
4)
Res
earc
hin
gro
up
isn
ot
suit
able
for
com
mer
cial
izat
ion
(5-
po
int
rati
ng
scal
e,1
–5
)
-0
.03
0(0
.03
3)
-0
.03
0(0
.03
3)
-0
.07
4*
**
(0.0
27)
-0
.07
4*
**
(0.0
27)
-0
.04
3*
*(0
.02
1)
-0
.04
3*
*(0
.02
1)
Res
earc
hre
sult
ssh
ould
be
free
lyac
cess
ible
(5-
po
int
rati
ng
scal
e,1
–5
)
-0
.06
1(0
.03
8)
-0
.06
1(0
.03
8)
-0
.10
0*
*(0
.04
1)
-0
.10
0*
*(0
.04
1)
-0
.08
2*
(0.0
33)
-0
.08
2*
(0.0
33)
Mea
nat
trac
tiven
ess
of
wo
rkin
gin
ind
ust
ryat
inst
itu
te
0.5
18
**
*(0
.16
6)
0.5
18
**
*(0
.16
6)
0.0
42
(0.1
11)
0.0
42
(0.1
11)
0.2
42
**
*(0
.09
0)
0.2
42
**
*(0
.09
0)
Wo
rkex
peri
ence
Yea
rso
fw
ork
inin
du
stry
0.0
35
**
(0.0
14)
0.0
35
**
(0.0
14)
0.0
37
**
(0.0
17)
0.0
37
**
(0.0
17)
0.0
33
**
*(0
.01
0)
0.0
33
**
*(0
.01
0)
Com
mer
ciali
zati
on
exper
ience
Pat
ent
file
do
rap
pli
edfo
r-
0.1
09
(0.1
38)
-0
.10
9(0
.13
8)
-0
.04
0(0
.10
3)
-0
.04
0(0
.10
3)
-0
.06
4(0
.08
2)
-0
.06
4(0
.08
2)
Co
op
erat
ion
wit
hin
du
stry
0.1
93
**
(0.0
89)
0.1
93
**
(0.0
89)
0.1
54
*(0
.08
5)
0.1
54
*(0
.08
5)
0.1
65
**
*(0
.04
8)
0.1
65
**
*(0
.04
8)
Ready to leave the ivory tower? 285
123
Ta
ble
3co
nti
nued
Ph.D
.st
uden
tsS
enio
rre
sear
cher
sE
nti
resa
mple
(1)
(2)
(3)
(4)
(5)
(6)
Per
sonal
chara
cter
isti
cs
Ris
kat
titu
de
0.0
24
(0.0
26)
0.0
24
(0.0
26)
0.0
46
*(0
.02
4)
0.0
46
*(0
.02
4)
0.0
35
*(0
.01
8)
0.0
35
*(0
.01
8)
Gen
der
(1=
fem
ale)
0.0
35
(0.0
63)
0.0
35
(0.0
63)
0.0
43
(0.0
82)
0.0
43
(0.0
82)
0.0
23
(0.0
51)
0.0
23
(0.0
51)
Ag
e(y
ears
)0
.08
3*
*(0
.03
6)
0.0
83
*(0
.03
6)
-0
.07
9*
*(0
.03
6)
-0
.07
9*
*(0
.03
6)
-0
.02
(0.0
23
)-
0.0
2(0
.02
3)
Ag
e2(y
ears
squ
ared
)-
0.0
01
**
(0.0
00)
-0
.00
1*
*(0
.00
0)
0.0
01
(0.0
00)
0.0
01
(0.0
00)
0.0
00
(0.0
01)
0.0
00
(0.0
01)
Ger
man
citi
zen
ship
0.1
29
**
(0.0
65)
0.1
29
**
(0.0
65)
-0
.01
4(0
.07
6)
-0
.01
4(0
.07
6)
0.0
55
(0.0
44)
0.0
55
(0.0
44)
Dir
ecto
rp
osi
tio
n–
–-
0.6
95
**
*(0
.18
2)
-0
.69
5*
**
(0.1
82)
-0
.57
5*
**
(0.1
83)
-0
.57
5*
**
(0.1
83)
Ten
ure
dw
ork
ing
con
trac
t-
0.5
33
**
(0.2
12)
-0
.53
3*
*(0
.21
2)
-0
.31
2*
**
(0.0
93)
-0
.31
2*
**
(0.0
93)
-0
.42
5*
**
(0.1
00)
-0
.42
5*
**
(0.1
00)
Nu
mb
ero
fo
bse
rvat
ion
s1
,127
1,1
27
1,2
01
1,2
01
2,3
28
2,3
28
McF
add
en’s
R2
0.0
40
00
.040
00
.070
80
.070
80
.073
50
.073
5
Lo
gli
kel
ihoo
d(f
ull
mo
del
)-
15
83
.32
-1
58
3.3
2-
16
64
.27
-1
66
4.2
7-
32
73
.99
-3
27
3.9
9
Wal
dte
st1
69
.10
16
9.1
03
19
.84
31
9.8
46
57
.84
65
7.8
4
Ro
bu
stst
and
ard
erro
rs,
wh
ich
are
adju
sted
for
inst
itu
tes,
are
rep
ort
edin
par
enth
eses
.T
he
aste
risk
s*
,*
*an
d*
**
repo
rtsi
gn
ifica
nce
atth
e1
0,
5an
d1
%le
vel
,re
spec
tiv
ely
286 M. Fritsch, S. Krabel
123
students decide to move to Germany and work for the Max Planck Society with the aim of
pushing their academic career by starting in a ‘science powerhouse.’ Therefore, these
students are less likely than their German counterparts to consider work in the private
sector to be attractive. Older Ph.D. students assess working in the private sector as being
relatively attractive while this attractiveness decreases with the age of senior scientists. The
willingness to take risks has a slightly positive and statistically significant relation to
scientists’ assessment of the attractiveness of work in the private sector.
5.2 Entrepreneurial attractiveness
According to our results, scientists’ assessment of entrepreneurial attractiveness is rela-
tively high when working in research fields where commercialization is common
(Table 4). In contrast, the perception of a strong focus on basic research of the scientists’
research group is negatively related to the appeal of working in the private sector. This
negative relation is statistically significant in each of the subsamples. Among doctoral
students, the perception of commercialization as common is positively related to the
attractiveness of work in the private sector. This evidence is in support of hypothesis H1b,
which stated that scientists’ perception of commercial potential of their own or related
research is positively associated with their assessment of entrepreneurial attractiveness.
Scientists in the life sciences value starting their own firm as significantly more attractive
than their peers in natural sciences and in the humanities.
Quite surprisingly, there is no significant relationship between the mean entrepreneurial
attractiveness at the institutional level and scientists’ individual assessment of the attrac-
tiveness to start a firm. Therefore, our results suggest rejecting hypothesis H2b, which
predicted a positive relationship between mean institutional and scientists’ individual
assessments of entrepreneurial attractiveness. We also do not find any significant positive
relationship between a scientist’s patenting experience and their proclivity to start a
business. Thus, hypothesis H3b, which predicted that patenting is positively related to
entrepreneurial attractiveness, is to be rejected. Research cooperation with industry has a
positive effect on the assessed attractiveness of starting one’s own firm only in the entire
sample, supporting hypothesis H4b. Several of the control variables show a statistically
significant relationship with entrepreneurial attractiveness. While risk-proclivity has the
expected positive sign and significance at the 1% level, we find that females and scientists
with German nationality find starting their own firm significantly less appealing than their
male and foreign-born counterparts. Scientists with tenured working contracts place the
attractiveness of entrepreneurship at a relatively low level.
Our data reveals a tremendous difference between scientist’s assessment of entrepre-
neurial attractiveness and current engagement in setting-up a firm. Roughly 28 per cent of
the surveyed scientists regard it as ‘attractive’ or ‘highly attractive’ to start their own firm
while only 3.2 per cent of scientists in the same sample actually engage in start-up activity.
An obvious reason for the higher shares of respondents that report high entrepreneurial
attractiveness as compared to the much lower share of scientists that can be regarded
nascent entrepreneurs is that a high appeal of being self-employed does not necessarily
imply to immediately undertake activities to set-up an own firm. Moreover, individuals
may attach high attractiveness values to entrepreneurship but at the same time consider
themselves as not well suited for starting an own business.
We also find considerable differences with regard to the factors that affect scientist’s
assessment of entrepreneurial attractiveness and the determinants of their propensity to
Ready to leave the ivory tower? 287
123
Tab
le4
Res
ult
so
ford
ered
pro
bit
esti
mat
esof
entr
epre
neu
rial
attr
acti
ven
ess
Ph.D
.st
uden
tsS
enio
rre
sear
cher
sE
nti
resa
mple
(1)
(2)
(3)
(4)
(5)
(6)
Do
ctora
ld
egre
e–
––
–-
0.1
35
**
(0.0
64)
-0
.135
**
(0.0
64)
Lif
esc
ien
ce0
.306
**
*(0
.07
4)
–0
.116
(0.0
75)
–0
.20
9*
**
(0.0
45)
–
Ch
emis
try
,p
hy
sics
and
tech
no
log
y–
-0
.30
6*
**
(0.0
74)
–-
0.1
16
(0.0
75)
–-
0.2
09
**
*(0
.04
5)
Hu
man
itie
s-
0.0
52
(0.1
19)
-0
.35
7*
**
(0.1
31)
-0
.079
(0.0
93)
-0
.195
*(0
.11
6)
-0
.04
6(0
.06
5)
-0
.255
**
*(0
.08
0)
Res
earc
hco
nte
xt
Com
mer
cial
izat
ion
isco
mm
on
infi
eld
of
rese
arch
(5-p
oin
tra
tin
gsc
ale,
1–
5)
0.0
93
**
(0.0
41
)0
.09
3*
*(0
.04
1)
0.0
47
(0.0
42)
0.0
47
(0.0
42)
0.0
65
**
(0.0
29)
0.0
65
**
(0.0
29)
Res
earc
hin
gro
up
isn
ot
suit
able
for
com
mer
cial
izat
ion
(5-
po
int
rati
ng
scal
e,1
–5
)
-0
.092
**
*(0
.03
1)
-0
.09
2*
**
(0.0
31)
-0
.152
**
*(0
.03
7)
-0
.152
**
*(0
.03
7)
-0
.12
1*
**
(0.0
23)
-0
.121
**
*(0
.02
3)
Res
earc
hre
sult
ssh
ould
be
free
lyac
cess
ible
(5-
po
int
rati
ng
scal
e,1
–5
)
-0
.061
*(0
.03
6)
-0
.06
1*
(0.0
36
)-
0.1
21
**
*(0
.03
2)
-0
.121
**
*(0
.03
2)
-0
.09
4*
**
(0.0
27)
-0
.094
**
*(0
.02
7)
Mea
nat
trac
tiven
ess
of
wo
rkin
gin
ind
ust
ryat
inst
itu
te
-0
.03
(0.1
42)
-0
.03
(0.1
42
)0
.068
(0.1
24)
0.0
68
(0.1
24)
0.0
36
(0.0
93
)0
.036
(0.0
93)
Wo
rkex
peri
ence
Yea
rso
fw
ork
inin
du
stry
0.0
25
**
(0.0
11
)0
.02
5*
*(0
.01
1)
0.0
12
(0.0
15)
0.0
12
(0.0
15)
0.0
18
*(0
.01
)0
.018
*(0
.01
)
Pri
or
bu
sin
ess
ow
ner
ship
0.7
31
**
*(0
.19
6)
0.7
31
**
*(0
.19
6)
0.6
64
**
*(0
.12
3)
0.6
64
**
*(0
.12
3)
0.6
82
**
*(0
.10
8)
0.6
82
**
*(0
.10
8)
Com
mer
ciali
zati
on
exper
ience
Pat
ent
file
do
rap
pli
edfo
r0
.048
(0.1
99)
0.0
48
(0.1
99
)0
.130
(0.1
09)
0.1
30
(0.1
09)
0.1
07
(0.1
)0
.107
(0.1
)
288 M. Fritsch, S. Krabel
123
Tab
le4
con
tin
ued
Ph.D
.st
uden
tsS
enio
rre
sear
cher
sE
nti
resa
mple
(1)
(2)
(3)
(4)
(5)
(6)
Co
op
erat
ion
wit
hin
du
stry
0.1
00
(0.0
62)
0.1
00
(0.0
62
)0
.088
(0.0
64)
0.0
88
(0.0
64)
0.0
94
**
(0.0
46)
0.0
94
**
(0.0
46)
Per
sonal
chara
cter
isti
cs
Ris
kat
titu
de
0.1
02
**
*(0
.02
0)
0.1
02
**
*(0
.02
0)
0.0
71
**
*(0
.02
1)
0.0
71
**
*(0
.02
1)
0.0
86
**
*(0
.01
4)
0.0
86
**
*(0
.01
4)
Gen
der
(1=
fem
ale)
-0
.393
**
*(0
.08
1)
-0
.39
3*
**
(0.0
81)
-0
.324
**
*(0
.07
3)
-0
.324
**
*(0
.07
3)
-0
.37
1*
**
(0.0
51)
-0
.371
**
*(0
.05
1)
Ag
e(y
ears
)0
.042
(0.0
44)
0.0
42
(0.0
44
)-
0.0
67
**
*(0
.02
5)
-0
.067
**
*(0
.02
5)
-0
.02
8(0
.02
)-
0.0
28
(0.0
2)
Ag
e2(y
ears
squ
ared
)-
0.0
01
(0.0
01)
-0
.00
1(0
.00
1)
0.0
01
**
(0.0
00)
0.0
01
**
(0.0
00
)0
.00
0(0
.00
1)
0.0
00
(0.0
01)
Ger
man
citi
zen
ship
-0
.437
**
*(0
.08
7)
-0
.43
7*
**
(0.0
87)
-0
.298
**
*(0
.06
2)
-0
.298
**
*(0
.06
2)
-0
.36
3*
**
(0.0
58)
-0
.363
**
*(0
.05
8)
Dir
ecto
rp
osi
tio
n–
–0
.174
(0.1
66)
0.1
74
(0.1
66)
0.1
89
(0.1
63
)0
.189
(0.1
63)
Ten
ure
dw
ork
ing
con
trac
t-
0.2
27
*(0
.13
5)
-0
.22
7*
(0.1
35
)-
0.2
17
*(0
.10
6)
-0
.217
*(0
.10
6)
-0
.23
3*
**
(0.0
90)
-0
.233
**
*(0
.09
0)
Nu
mb
ero
fo
bse
rvat
ion
s1
,12
51
,125
1,1
99
1,1
99
2,3
24
2,3
24
McF
add
en’s
R2
0.0
63
50
.063
50
.067
70
.06
77
0.0
64
20
.064
2
Lo
gli
kel
ihoo
d(f
ull
mod
el)
-1
62
0.7
5-
16
20
.75
-1
70
6.9
2-
17
06.9
2-
33
39
.68
-3
33
9.6
8
Wal
dte
st5
90
.21
59
0.2
13
68
.56
36
8.5
65
08
.67
50
8.6
7
Rob
ust
stan
dar
der
rors
,w
hic
har
ead
just
edfo
rin
stit
ute
s,ar
ere
po
rted
inp
aran
thes
es.
Th
eas
teri
sks
*,
**
and
**
*re
po
rtsi
gn
ifica
nce
atth
e1
0,
5an
d1%
level
,re
spec
tivel
y
Ready to leave the ivory tower? 289
123
take entrepreneurial action and become nascent entrepreneurs (Table 5).5 It is quite
remarkable that nascent entrepreneurship seems to be largely independent of the respective
research field and the assessed suitability of research for commercialization while both
issues, the nature of scientists’ research and its perceived commercial potential, have a
robust positive and significant effect on entrepreneurial attractiveness. Furthermore, patent
activity seems to be a strong predictor of nascent entrepreneurship but is not related to
entrepreneurial attractiveness. Having a director position has a positive effect on nascent
entrepreneurship but does not seem to affect entrepreneurial attractiveness. Individual risk
attitude and gender, which have a significant relation to entrepreneurial attractiveness, do
not contribute to the explanation of nascent entrepreneurship.
These differences between our results for entrepreneurial attractiveness and nascent
entrepreneurship clearly show that a high appeal of entrepreneurship and taking action of
setting up a business are two different issues. Existing literature in social psychology often
denotes ‘attractiveness’ as a value individuals perceive (see e.g. Fishbein and Ajzen 1975;
Ajzen 1991) while following action is described as a result of value activation. However,
value activation does not occur automatically but can be invoked by feasible conditions for
the action or enhanced self-focus (Verplanken and Holland 2002). In view of our results,
the gap between attractiveness and action might indicate that a potential for more academic
spin-offs exists if the institutional framework and other conditions would alleviate entre-
preneurial activity.
With respect to the robustness of results we compared whether our sample of analysis is
representative for the Max Planck Society with respect to research discipline, gender and
the share of non-German scientists. Thereby, we find no significant difference between the
entire population of and our sample of analysis.6 Further, we analyzed scientists’ readiness
to participate in a further study including the same set of regressors as in the analysis of
scientists’ readiness to work in the private sector. Thereby, we find that scientists who
cooperated with private firms and who find commercialization common in their research
field are more willing to answer survey interviews. As this tendency may influence our
empirical results, we performed additional regressions of scientists’ attractiveness to work
in the private sector excluding cooperators and scientists who highly agree that their
research has commercial value. Further, we separately analyzed our models among the
subgroups of female and male scientists, respectively.
6 Discussion and conclusions
Based on a survey of scientists working in the German Max Planck Society we have
analyzed the level as well as the determinants of the attractiveness that respondents attach
to working in the private sector and to running an own business. Furthermore, we have
compared entrepreneurial attractiveness to factual involvement into founding a firm. We
found that about two out of five scientists regard working in the business sector as
‘attractive’ or ‘highly attractive’. About 28% of the scientists attach high levels of
attractiveness to starting an own firm. These appear to be remarkably high shares, given
that most of the MPS researchers in our sample are devoted to basic science with no
5 Note that among doctoral students in the humanities, no nascent entrepreneur is found, so that wedisregard this variable in the analysis.6 Statistics are available upon request.
290 M. Fritsch, S. Krabel
123
Tab
le5
Rar
eev
ent
logis
tic
esti
mat
esof
scie
nti
st’s
likel
ihood
tobe
anas
cent
entr
epre
neu
r
Ph
.D.
(37
nas
cen
ts)
Sen
ior
rese
arch
ers
(36
nas
cen
ts)
En
tire
sam
ple
(73
nas
cen
ts)
(1)
(2)
(3)
(4)
(5)
Do
ctora
ld
egre
e–
––
-0
.494
(0.3
59)
-0
.49
4(0
.35
9)
Lif
esc
ien
ce0
.273
(0.5
04)
0.0
40
(0.4
06)
–0
.064
(0.3
23)
–
Chem
istr
y,
physi
csan
dte
chnolo
gy
––
-0
.398
(0.4
06)
–-
0.0
64
(0.3
23
)
Hu
man
itie
s–
1.2
88
*(0
.71
2)
1.2
48
(0.7
72)
0.0
48
(0.5
39)
-0
.01
6(0
.64
5)
Res
earc
hco
nte
xt
Com
mer
cial
izat
ion
isco
mm
on
infi
eld
of
rese
arch
(5-p
oin
tra
tin
gsc
ale,
1–
5)
0.3
56
*(0
.19
1)
0.1
98
(0.1
96)
0.1
98
(0.1
96)
0.2
79
*(0
.15
6)
0.2
79
*(0
.15
6)
Res
earc
hin
gro
up
isn
ot
suit
able
for
com
mer
cial
izat
ion
(5-p
oin
tra
ting
scal
e,1–5)
0.0
83
(0.1
76)
0.0
78
(0.2
68)
0.0
78
(0.2
68)
0.0
42
(0.1
55)
0.0
42
(0.1
55
)
Res
earc
hre
sult
ssh
ould
be
free
lyac
cess
ible
(5-p
oin
tra
ting
scal
e,1
–5
)-
0.1
73
(0.1
77)
-0
.506
**
(0.2
04)
-0
.506
**
(0.2
04
)-
0.3
49
**
*(0
.12
8)
-0
.34
9*
**
(0.1
28)
Mea
nat
trac
tiv
enes
so
fw
ork
ing
inin
du
stry
atin
stit
ute
-0
.938
(1.0
37)
1.0
44
(0.9
94)
1.0
44
(0.9
94)
0.0
56
(0.0
44)
0.0
56
(0.0
44
)
Wo
rkex
per
ien
ce
Yea
rso
fw
ork
inin
du
stry
0.0
40
(0.0
61)
0.0
90
(0.0
62)
0.0
90
(0.0
62)
0.0
56
(0.0
44)
0.0
56
(0.0
44
)
Pri
or
bu
sin
ess
ow
ner
ship
1.4
54
**
*(0
.50
5)
1.2
76
**
*(0
.41
6))
1.2
76
**
*(0
.41
6))
1.3
48
**
*(0
.29
5)
1.3
48
**
*(0
.29
5)
Com
mer
ciali
zati
on
exper
ience
Pat
ent
file
do
rap
pli
edfo
r1
.349
**
*(0
.39
3)
1.2
42
**
*(0
.46
2)
1.2
42
**
*(0
.46
2)
1.2
98
**
*(0
.29
9)
1.2
98
**
*(0
.29
9)
Co
op
erat
ion
wit
hin
du
stry
0.6
62
(0.4
08)
0.7
59
(0.4
75)
0.7
59
(0.4
75)
0.7
02
**
(0.2
89)
0.7
02
**
(0.2
89)
Per
sonal
chara
cter
isti
cs
Ris
kat
titu
de
-0
.051
(0.1
24)
0.1
50
(0.1
00)
0.1
50
(0.1
00)
0.0
68
(0.0
78)
0.0
68
(0.0
78
)
Gen
der
(1=
fem
ale)
-0
.299
(0.3
47)
0.2
68
(0.4
63)
0.2
68
(0.4
63)
-0
.039
(0.2
55)
-0
.03
9(0
.25
5)
Ag
e(y
ears
)-
0.2
01
(0.1
40)
-0
.142
(0.1
65)
-0
.142
(0.1
65)
-0
.112
(0.1
14)
-0
.11
2(0
.11
4)
Ag
e2(y
ears
squ
ared
)0
.002
(0.0
01)
0.0
01
(0.0
01)
0.0
01
(0.0
01)
0.0
01
(0.0
01)
0.0
01
(0.0
01
)
Ready to leave the ivory tower? 291
123
Tab
le5
con
tin
ued
Ph
.D.
(37
nas
cen
ts)
Sen
ior
rese
arch
ers
(36
nas
cen
ts)
En
tire
sam
ple
(73
nas
cen
ts)
(1)
(2)
(3)
(4)
(5)
Ger
man
citi
zen
ship
-1
.362
**
*(0
.42
9)
-0
.554
(0.3
86)
-0
.554
(0.3
86)
-0
.935
**
*(0
.34
4)
-0
.93
5*
**
(0.3
44)
Dir
ecto
rp
osi
tio
n–
–1
.858
*(1
.01
6)
1.6
03
**
(0.7
89)
1.6
03
**
(0.7
89)
Ten
ure
dw
ork
ing
con
trac
t-
0.4
22
(0.6
02)
-0
.594
(0.8
24)
-0
.594
(0.8
24)
-0
.449
(0.4
75)
-0
.44
9(0
.47
5)
Co
nst
ant
2.7
33
(4.1
48)
-3
.129
(5.3
44)
-3
.129
(5.3
44)
-0
.259
(3.6
38)
-0
.25
9(3
.63
8)
Nu
mb
ero
fo
bse
rvat
ion
s1
,124
1,1
97
1,1
97
2,3
21
2,3
21
Pse
ud
oR
20
.145
50
.267
30
.267
30
.267
30
.182
9
Wal
dte
st2
07
.17
15
6.8
51
56
.85
23
6.0
72
36
.07
Robust
stan
dar
der
rors
,w
hic
har
ead
just
edfo
rin
stit
ute
s,ar
ere
port
edin
par
anth
eses
.T
he
aste
risk
s*,
**
and
***
report
signifi
cance
atth
e10,
5an
d1
%le
vel
,re
spec
tiv
ely
292 M. Fritsch, S. Krabel
123
obligation to cooperate with private firms.7 We—cautiously—interpret our numerical
results as an indication that there exists a considerable potential for career mobility
between basic academic science and the private sector.
Our empirical analysis of the determinants of scientists’ appeal to work in the business
sector or to start their own firm yields (at least) four notable results. First, researchers who
do not hold a Ph.D. degree find working in the private sector and starting an own firm
especially interesting while senior researchers report significantly lower attractiveness
levels. Our results confirm the view expressed by Mangematin (2000) that the completion
of a doctoral thesis is a natural point in time for a change of career tracks. In contrast to
Ph.D. students senior researchers have already opted for an academic career when staying
in academia after having obtained their doctorate. This reasoning may well explain why
senior researchers find switching to the private sector less appealing.
Second, scientists regard it particularly appealing to work outside academia if they
regard their research as suited for commercialization. Experience in joint research with
private firms is related to high attractiveness of working in the business sector. Yet,
patenting experience is unrelated to scientists’ appeal to work in the private sector. This
result is similar to the finding by Link and Ruhm (2010) that academic entrepreneurs
(compared to business entrepreneurs) tend to rather publish their results than to patent
research outcome. Thus, these findings suggest that patenting activity is hardly related to
scientists’ intention to become entrepreneurs. In contrast, patenting is positively related to
start-up activity (Stuart and Ding 2006; Krabel and Mueller 2009). This is a puzzle that
demands further research.
Both indicators, working in a research field with a potential of commercial application
as well as experience of cooperation with private firms are also positively related to high
levels of entrepreneurial attractiveness. Moreover, we find some considerable differences
of attractiveness levels across scientists’ fields of research. The appeal to work in the
business sector or to start an own company is especially high in life sciences and is
relatively low among researchers in the humanities. These results may be of interest for
policy makers and leaders of scientific institutions. When research is supposed to meet
practical requirements and involves cooperative research with private partners, scientists
tend to be more attracted towards accepting jobs in the private sector or starting an own
company. Hence, the ‘‘supply’’ of scientists who are ready to leave academia can be
influenced by the design of research agendas. Further, scientists’ attitude towards open
science is negatively related to scientists’ attractiveness values to start an own company or
to work in the private sector. This result contrasts prior findings by Ponomariov and
Boardman (2008), which suggest that scientists worry that interaction with the private
sector could distract them from doing good research is not a crucial barrier for university-
industry interaction.8
Third, we found that scientists’ proclivity to work in the business sector is considerably
shaped by prior experience with private sector employment. Similarly, previous entre-
preneurial experience is an important driver of scientists’ attractiveness to start an own
firm. One plausible explanation for this result is that more ties to the private sector lead to a
higher likelihood that scientists switch (back) to private sector work.
Fourth, we interestingly do not find evidence that the mean entrepreneurial attractive-
ness of institutional peers relates to the individual appeal to start a firm or the likelihood of
7 We are not aware of any comparable study that could serve as a reference to the shares detected by us.8 However, the study by Ponomariov and Boardman relates to informal interaction of academic scientistswith private firms rather than job movements between sectors.
Ready to leave the ivory tower? 293
123
being a nascent entrepreneur. This is in contrast to related studies by Stuart and Ding
(2006) and Bercovitz and Feldman (2008) which do find peer effects with respect to
entrepreneurial activity. However, this divergence might be due to different measures
used.9 There are considerable differences between the determinants of entrepreneurial
attractiveness and the factors driving nascent entrepreneurship among the scientists of our
sample. While patenting of research results is a robust indicator of nascent entrepreneur-
ship, it is not related to entrepreneurial attractiveness. Conversely, scientists’ perceived
commercial suitability of their research and the perceived commonness of commerciali-
zation in their research field clearly impact their interest in starting a firm but have no
significant effect on the likelihood to be involved in setting up a firm. Hence, our findings
suggest that scientists’ appeal towards starting an own business (entrepreneurial attrac-
tiveness) and factual entrepreneurial activity tend to be rather different issues.
To sum up, the present study contributes to the broader debate on benefits and costs of
job mobility from academics to the private sector. On the one hand, prior literature reveals
that firms benefit from employing academic scientists by improving their innovative
capacity (Herrera et al. 2010; Agrawal 2006). Thus, especially for innovating firms it is
beneficial to integrate former academics in their research teams—even if academics need
some time to adapt to non-academic research. On the other hand the costs and benefits of
universities associated with job mobility of academic personnel has hardly been analyzed.
With our study we contribute to this lack of knowledge as we regard scientists’ individual
motivations and obstacles to private sector work. Scientists’ appeal to work in the private
sector is positively related to their commercial research orientation while being negatively
related to scientists’ wish that research results should be freely available. Thus, scientists
benefit from working in the private sector as they are able to advance their ideas while the
wish to publish research results freely rather impedes on scientists’ demand to work in the
private sector. Thus, the demand for high publication output may inversely affect scien-
tists’ wish to move to the private sector. Yet, as our study focuses on the individual level
the question remains open to what degree job mobility affects universities on the orga-
nizational level. Therefore, we encourage further research aiming to analyze universities’
costs and benefits when academics switch to the private sector.
Finally, we encourage further research that may complement our study. Due to the pure
cross-section character of our data we are currently not able to follow scientists over time
in order to analyze their further career steps. A longitudinal study, which is highly
desirable, would allow an examination as to what extent perceived attractiveness predicts
later career progression outside academia, be it as dependent employee or starting an own
firm. Such a study could complement our work and further contribute to our understanding
of career choices of researchers and of the potentials for an improved transfer of science
into commercial application. Moreover, further research may examine the causality
whether cooperative projects with the private sector lead to higher appeal of working in the
private sector or vice versa. From a policy perspective, the gap between entrepreneurial
attractiveness and nascent entrepreneurship demands attention. Our results indicate that a
potential for more academic spin-offs exists. However, the question how to exploit such
potential remains open for further research.
9 Our measures of peer influence relate to the mean entrepreneurial attractiveness and the share of nascententrepreneurs—regardless of scientists’ own observation. Bercovitz and Feldman (2008) observe peereffects by regarding cases when scientists switch from one institution to another while Stuart and Ding(2006) regard the entrepreneurial activity of past co-workers. Therefore, the three measures can hardly becompared.
294 M. Fritsch, S. Krabel
123
References
Agrawal, A. (2006). Engaging the inventor: Exploring licensing strategies for university inventions and therole of latent knowledge. Strategic Management Journal, 27(1), 63–79.
Agarwal, R., Echambadi, R., Franco, A. M., & Sarkar, M. B. (2004). Knowledge transfer throughinheritance: Spin-out generation, development and survival. Academy of Management Journal, 47(4),501–522.
Agrawal, A., & Henderson, R. (2002). Putting patents in context: Exploring knowledge transfer from MIT.Management Science, 48(1), 44–60.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes,50(2), 179–211.
Azoulay, P., Ding, W. W., & Stuart, T. (2006). The impact of academic patenting on (public) researchoutput. NBER Working Paper #11917.
Bercovitz, J., & Feldman, M. (2008). Academic entrepreneurs: Organizational change at the individuallevel. Organization Science, 19(1), 69–89.
Buenstorf, G. (2009). Is commercialization good or bad for science? Individual-level evidence from the MaxPlanck Society. Research Policy, 38(2), 281–292.
Charles, D., & Conway, C. (2001). Higher education–business interaction survey: A report to the UK HEfunding bodies and the Office of Science and Technology. Newcastle: University of Newcastle uponTyne.
Cohen, W. M., Florida, R., Randazzese, L., & Walsh, J. (1998). Industry and the academy: Uneasy partnersin the cause of technological advance. In R. Noll (Ed.), Challenges to the research university(pp. 171–198). Washington, DC: Brookings Institution.
Cohen, W. M., Nelson, R. R., & Walsh, J. (2002). Links and impacts: The influence of public research onindustrial R&D. Management Science, 48(1), 1–23.
Colombo, M. G., & Grilli, L. (2005). Founders’ human capital and the growth of new technology–basedfirms: A competence–based view. Research Policy, 34(6), 795–816.
Dietz, J. S., & Bozeman, B. (2005). Academic careers, patents, and productivity: industry experience asscientific and technical human capital. Research Policy, 34(3), 349–367.
Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schrupp, J., & Wagner, G.G. (2005). Individual riskattitudes: New evidence from a large, representative, experimentally-validated survey. DIW Discus-sion Paper 511, available at http://193.174.141.131/documents/publikationen/73/43553/dp511.pdf.Accessed October, 15th, 2009.
Etzkowitz, H. (1998). The norms of entrepreneurial science: cognitive effects of the new university-industrylinkages. Research Policy, 27(8), 823–833.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory andresearch. Reading, MA: Addison-Wesley.
Gittelman, M. (1999). Knowledge as property: Innovation in biotechnology in the United States and inFrance. In: Academy of Management Proceedings, IM: L1–L6.
Herrera, L., Munoz-Doyague, M. F., & Nieto, M. (2010). Mobility of public researchers, scientificknowledge transfer, and the firm’s innovation process. Journal of Business Research, 63(5), 510–518.
Krabel, S., & Mueller, P. (2009). What drives scientists to start their own company?: An empiricalinvestigation of Max Planck Society scientists. Research Policy, 38(6), 947–956.
Libaers, D. (2009). Industry relationships of DoD-funded academics and institutional changes in the USUniversity system. Journal of Technology Transfer, 34(5), 474–489.
Lin, M.-W., & Bozeman, B. (2006). Researchers’ industry experience and productivity in university-industry research centers: A ‘scientific and technical human capital’ explanation. Journal of Tech-nology Transfer, 31(2), 269–290.
Link, A. N., & Ruhm, C. J. (2010). Public knowledge, private knowledge: The intellectual capital ofentrepreneurs. Small Business Economics, (forthcoming).
Link, A. N., & Siegel, D. S. (2007). Innovation, entrepreneurship and technological change. Oxford: OxfordUniversity Press.
Lowe, R. A., & Gonzalez-Brambila, C. (2007). Faculty entrepreneurs and research productivity. Journal ofTechnology Transfer, 32(3), 173–194.
Mangematin, V. (2000). PhD job market: Professional trajectories and incentives during the PhD. ResearchPolicy, 29(6), 741–756.
Mansfield, E. (1995). Academic research underlying industrial innovation. Review of Economics and Sta-tistics, 77(1), 55–65.
Ready to leave the ivory tower? 295
123
Martinelli, D. (2001). Labour market entry and mobility of young French PhDs. In: Innovative people—Mobility of skilled personnel in national innovation systems. OECD Proceedings, 2001, 159–173.
Max Planck Society (2009). Max Plank Society: Annual Report 2008, available at http://www.mpg.de/pdf/jahresbericht2008/jahresbericht2008.pdf, Accessed February, 9th, 2009.
Noll, R. (1998). Challenges to research universities. Washington, DC: Brookings Institution.OECD (2002). International mobility of the highly skilled. OECD Report, Paris, available at: http://pesona.
mmu.edu.my/*chtan/Seminar/9202011E.pdf?bcsi_scan_2A2102422E3B7B6B=0&bcsi_scan_filename=9202011E.pdf, Accessed August, 15th, 2009.
Pavitt, K. (1998). The social shaping of the national science base. Research Policy, 27(8), 793–805.Perkmann, M., & Walsh, K. (2007). University-industry relationships and open innovation: Towards a
research agenda. International Journal of Management Reviews, 9(4), 259–280.Ponomariov, B., & Boardman, P. C. (2008). The effect of informal industry contacts on the time university
scientists allocate to collaborative research with industry. Journal of Technology Transfer, 33(3),301–313.
Powers, J. B., & McDougall, P. P. (2005). University start-up formation and technology licensing with firmsthat go public: A resource-based view of academic entrepreneurship. Journal of Business Venturing,20(3), 291–311.
Reynolds, P. D., Carter, N. M., Gartner, W. B., & Greene, P. G. (2004). The prevalence of nascententrepreneurs in the United States: Evidence from the panel study of entrepreneurial dynamics. SmallBusiness Economics, 23(4), 263–284.
Rosenberg, N. (1990). Why do firms do basic research (with their own money)? Research Policy, 19(2),165–174.
Schomburg, H., & Teichler, U. (2006). Higher education and graduate employment in Europe: Results fromgraduates surveys from twelve countries. Dordrecht: Springer.
Segarra–Blasco, A., & Arauzo–Carod, J. (2008). Sources of innovation and industry–university interaction:Evidence from Spanish firms. Research Policy, 37(8), 1283–1295.
Shane, S. (2004). Academic entrepreneurship—University spinoffs and wealth creation. Cheltenham:Edward Elgar.
Shane, S., & Khurana, R. (2003). Bringing individuals back in: The effects of career experience on new firmfounding. Industrial and Corporate Change, 12(3), 519–543.
Shane, S., & Stuart, T. (2002). Organizational endowments and the performance of university start-ups.Management Science, 48(1), 154–170.
Sherif, M. (1935). A study of some social factors in perception. New York City: Columbia University Press.Shrader, R., & Siegel, D. S. (2007). Assessing the relationship between human capital and firm performance:
Evidence from technology-based new ventures. Entrepreneurship Theory & Practice, 31(6), 893–908.Stern, S. (2004). Do scientists pay to be scientists? Management Science, 50(6), 835–853.Stuart, T., & Ding, W. W. (2006). The social structural determinants of academic entrepreneurship: An
analysis of university scientists’ participation in commercial ventures. American Journal of Sociology,112(1), 97–144.
Thune, T. (2007). University–industry collaboration: The network embeddedness approach. Science andPublic Policy, 34(3), 158–168.
Toole, A., & Czarnitzki, D. (2007). Biomedical academic entrepreneurship through the SBIR program.Journal of Economic Behavior & Organization, 63(4), 716–738.
Toole, A. A., & Czarnitzki, D. (2009). Exploring the relationship between scientist human capital and firmperformance: The case of biomedical academic entrepreneurs in the SBIR program. ManagementScience, 55(1), 101–114.
Turner, R. (1990). Some contributions of Muzafer Sherif to Sociology. Social Psychology Quarterly, 53(4),283–291.
Verplanken, B., & Holland, R. W. (2002). Motivated decision making: Effects of activation and self-centrality of values on choices and behavior. Journal of Personality and Social Psychology, 82(3),434–447.
Wagner, G. G., Frick, J., & Schrupp, J. (2007). German Socio.Economic Panel. http://www.diw.de/documents/dokumentenarchiv/17/56579/personen_en_2007.pdf. Accessed October, 28th, 2009.
Wright, M., Clarysee, B., Mustar, P., & Lockett, A. (2007). Academic entrepreneurship in Europe.Cheltenham, UK.l: Edward Elgar.
Zellner, C. (2003). The economic effects of basic research: Evidence for embodied knowledge transfer viascientists’ migration. Research Policy, 32(10), 1881–1895.
Zucker, L. G., Darby, M. R., & Brewer, M. B. (1998). Intellectual human capital and the birth of US.biotechnology enterprises. American Economic Review, 88(1), 290–306.
296 M. Fritsch, S. Krabel
123
