Main Sponsor: Fortis Bank Nederland
Academic Network in Entrepreneurship, Innovation and Finance
Entrepreneurship and Human Capital
Edited by
Prof. Dr. Mirjam van Praag
Amsterdam Center for Entrepreneurship
Faculty of Economics and Business
University of Amsterdam
The Netherlands
Assisted by Thomas Hemels and Martin Koudstaal (University of Amsterdam)
This book has been written with the financial support of
the Gate2Growth Academic Network in Entrepreneurship, Innovation and Finance
July 2006
Main Sponsor: Fortis Bank Nederland
©European Communities, 2006
Reproduction within the framework of the Gate2Growth Initiative is authorized provided the source is
acknowledged.
Neither the European Commission, nor any person acting on behalf of the Commission is responsible for the
use, which might be made of the following information. The views expressed in this study are those of the
authors and do not necessarily reflect the policies of the European Commission or any person acting on behalf
of the Commission
Table of Contents
Acknowledgments
Introduction
p1
p3
PART I
p4
Family Origins
Chapter 1* Families, human capital, and small business: Evidence
from the Characteristics of Business Owners Survey
*
Chapter 2 The entrepreneur’s entry mode: Business takeover or new
venture start
PART II
Education
p5
p 11
p 17
Chapter 3* Why are the returns to education higher for entrepreneurs
than for employees?
p 18
PART III
p 25
General Human Capital versus Specific Human Capital
Chapter 4* Wage policies and incentives to invest in firm-specific
human capital
Chapter 5* Returns to intelligence: Entrepreneurs versus employees
Chapter 6* Opportunity identification and pursuit: Does an
entrepreneurs’ human capital matter?
PART IV
Work Experience
p 26
p 32
p 39
p 46
Chapter 7* Inferring the unobserved human capital of entrepreneurs
Chapter 8* The role of initial and acquired human capital in the longterm survival and performance of immigrant entrepreneurs
p 47
p 52
PART V
p 58
Scientific Entrepreneurship
Chapter 9* Scientist entrepreneurship: Human capital or social capital? P 59
PART VI
Guidelines for Policy
p 67
Chapter 10* Implementing the Community Lisbon Programme: Fostering p 68
entrepreneurial mindsets through education and learning
Chapter 11 So what?
P 74
*
These chapters and their titles are based on the papers that were presented
by various authors at the ACE-G2G workshop “Entrepreneurship and Human
Capital” in Amsterdam, June 23-24, 2006. Please refer to the
acknowledgments and the chapters themselves for the authors’ names.
Acknowledgments
This book includes edited proceedings from the workshop “Entrepreneurship
and Human Capital” that was held in Amsterdam in June 2006. The workshop
was organized by The Amsterdam Center for Entrepreneurship (ACE) of the
Faculty of Economics and Business of the University of Amsterdam in
Cooperation with the EC Gate2growth Academic Network in Entrepreneurship,
Innovation and Finance (G2G). For ACE it was the first activity, resulting in
this first ACE-book. For ACE, a bridge between academia and practice, this
first activity was academically oriented. For G2G, it was the third research
workshop in their series.
The content that follows has resulted from the papers and presentations of
ten speakers at the workshop. Moreover, eight discussants have commented
on the papers, which added beneficial insights. Moreover, the session heads
acted such that valuable discussions resulted.
I first wish to thank all the speakers and their coauthors for presenting
interesting cutting-edge work. The presenters were David Audretsch from
the Max Planck Institute and Indiana University who coauthored with Taylor
Aldridge and Alex Oettl from the Max Planck Institute; George Baker from
Harvard Business School, coauthoring with Nancy Dean Bealieu and Cristian
Voicu from the same school; Simone Baldassari from the Commission of the
European Communities; Jean Bonnet from the University of Caen (France),
who coauthored with Nicolas Le Pape from the same university and Régis
Renault from Université de Cergy-Pontoise; Rob Fairlie from the University of
California with coauthor Alicia Robb from the same university; Miri Lerner
from the Tel-Aviv University with her coauthor Susanna Khavul from the
London Business School; Simon Parker from the University of Durham
(coauthored and copresented by Mirjam Van Praag from the Amsterdam
Center for Entrepreneurship of the University of Amsterdam); Justin Van Der
Sluis from the University of Amsterdam (coauthor Mirjam Van Praag); Arjen
Van Witteloostuijn (coauthors Justin Van Der Sluis and Mirjam Van Praag),
from the University of Groningen and Durham, and Mike Wright from the
Centre for Management Buy-Out Research, Nottingham University Business
School with his coauthors Deniz Ucbasaran (Nottingham University Business
School) and Paul Westhead (Warwick Business School).
Second, I express my gratitude to the discussants. The presentation by
David Audretsch was discussed by Jeremy Philpott from the European Patent
Office; the one by George Baker was discussed by Enrico Perotti from the
University of Amsterdam; Jean Bonnet’s presentation was discussed by Uzi
De Haan from the Technion University of Haifa; Miri Lerner’s presentation
was discussed by Mirjam Van Praag; the presentation by Simon Parker and
Mirjam Van Praag was discussed by Roy Thurik from the Erasmus University
of Rotterdam; Justin Van Der Sluis’ presentation was discussed by Gerard
Pfann from the University of Maastricht; the presentation by Arjen Van
Witteloostuijn was discussed by Magnus Henrekson from the Research
Institute of Industrial Economics from the University of Stockholm; and the
presentation by Mike Wright was discussed by Lorraine Uhlaner from the
Erasmus University of Rotterdam.
1
Third, I thank Bill Rees (Board of ACE) and Auke Ijsselstein, both from the
University of Amsterdam for being session heads. I am grateful to Kees Cools
(Board of ACE) from the University of Groningen and The Boston Consulting
Group for leading the final discussion about policy implications.
Finally I would like to thank Jean Bonnet for initiating the workshop, Tom
Schamp and Bart Clarysse (both from G2G) for their co-organization of the
workshop, Cecile Wentges for her invaluable assistance, and Thomas Hemels
and Martin Koudstaal for their assistance in writing this book.
Amsterdam, July 2006,
Mirjam van Praag,
Professor of Entrepreneurship and Organization
Director of the Amsterdam Center for Entrepreneurship
2
Introduction
The Lisbon strategy dictates higher levels of innovation. Most politicians and
policy makers strongly believe that more innovation can only be realized by
more entrepreneurship. Therefore, entrepreneurship ranks high on the policy
agenda’s of most European countries. However, in what ways can the latent
entrepreneurial capacity and willingness be developed such that more
innovations will result? One of the most relevant and straightforward
instruments is human capital. Recently academic research into the effect of
human capital on entrepreneurship has regained interest, both from a micro
and a macro-economic perspective.
People have initial levels of human capital, dependent on family
background and intelligence. Besides, and perhaps more importantly from a
policy perspective, they also invest in acquiring more human capital by means
of education and experience. In this book, family background, intelligence,
(entrepreneurship) education and various sorts of experience are considered
as determinants of entrepreneurship at all engagement levels. What is the
effect of initial and acquired human capital on the likelihood that people
become entrepreneurs, either through start-up or through takeover? What is
the effect on the likelihood that people identify and pursue entrepreneurial
opportunities? How can human capital influence the performance of
entrepreneurs such as their survival, growth and incomes? Is the return on
general human capital investments higher or lower than the return on
(entrepreneurship) specific human capital? Do immigrants benefit less from
the human capital acquired in their original country than in their host country?
How can universities commercialize their human capital embedded in the
many researchers and what is the role of entrepreneurship? And how do
rigidities in the labor market affect the relationship between entrepreneurship
and education? These are the questions addressed in this book.
Based on the answers, implications and recommendations for researchers,
public policy makers and, last but not least, entrepreneurs are discussed.
3
Part I
Family Origins
4
CHAPTER 1
Based on
“Families, human capital, and small business:
Evidence from the Characteristics of Business Owners Survey”
by
Robert W. Fairlie
University of California
&
Alicia Robb
Foundation for Sustainable Development and University of California, Santa
Cruz
ABSTRACT
An important finding in the rapidly growing literature on entrepreneurship is
that the probability of business ownership is substantially higher among the
children of business owners than among the children of non-business owners.
Using data from the confidential and restricted-access Characteristics of
Business Owners (CBO) Survey, Fairlie and Robb provide evidence on the
causes of intergenerational links in business ownership and the related issue
of how having a family business background affects small business outcomes.
Estimates from the CBO indicate that more than half of all business owners
had a business-owning family member prior to starting their business.
Conditional on having a business-owning family member, less than 50 percent
of small business owners worked in that family member's business, suggesting
that it is unlikely that intergenerational links in business ownership are largely
due to the acquisition of business human capital. The estimates also indicate
that only 1.6 percent of all small businesses are inherited suggesting that the
role of business inheritances in determining intergenerational links in
entrepreneurship is limited at best. Instead, similarities across family
members in entrepreneurial preferences may best explain part of the
intergenerational relationship.
In contrast, estimates from regression models conditioning on business
ownership indicate that coming from a business-owning family plays only a
minor role in determining small business outcomes, whereas the business
human capital acquired from prior work experience in a family member's
business appears to be very important for business success.
5
1.1
Introduction
An important finding in the entrepreneurship literature is that the probability
of business ownership is two or three times higher among the children of
business owners than among the children of non-business owners. The
underlying causes, however, of intergenerational links in business ownership
have not been identified in the literature. They may be due to the acquisition
of specific types of experience in family-owned businesses, inheritances of
businesses, or a correlation among family members in preferences for
entrepreneurial activities.
Using the Characteristics of Business Owners (CBO), Fairlie and Robb
provide evidence on the importance of these factors and explore the related
question of whether having a business-owning parent or other family member
improves small business outcomes. Although strong intergenerational links in
entrepreneurship have been repeatedly documented in the literature, the
effects on small business outcomes conditioning on ownership are essentially
unknown.
1.2
Literature
A few patterns are beginning to emerge in the young and expanding literature
on entrepreneurship. One of them is that the likelihood of business ownership
is higher for individuals with a business-owning parent.
Several explanations for the intergenerational transmission of business
ownership have been offered. First, the informal learning or apprenticeshiptype training that occurs by growing up in the context of a family business
may provide an important opportunity for the acquisition of general or specific
business human capital.
Second, intergenerational links in business ownership may be caused by a
correlation among family members in preferences for entrepreneurial
activities and entrepreneurial ability. The correlation may simply be due to
similarities among family members in preferences for autonomy or selfemployment, or similarities in other personal characteristics that are
associated with entrepreneurship, such as entrepreneurial ability and attitudes
towards risk. Related to the issue of correlated preferences and ability,
intergenerational links may also be caused by business-owning parents
creating role models for their children to become business owners. Observing
a parent as a successful entrepreneur may improve a child's assessment of his
or her own entrepreneurial ability.
Third, intergenerational links may be due to the likelihood of becoming a
partner in the family business or inheriting it. Forming a partnership with a
child may represent a less expensive method for parents to help their
offspring becoming business owners. Also, partnerships and inheritances may
represent an efficient form of transmitting reputation capital or an established
clientele from one generation to the next.
Although strong intergenerational links in entrepreneurship have been
repeatedly documented in the literature, neither the exact underlying
mechanism, nor their effects on small business outcomes conditioning on
6
ownership are known. Fairlie and Robb fill these two essential gaps in the
literature.
1.3
Data
The 1992 Characteristics of Business Owners (CBO) survey was conducted
by the U.S. Census Bureau and provides economic, demographic and
sociological data on business owners and their business activities.1
The sample used below includes firms that meet a minimum weeks and
hours restriction, to rule out the very small-scale business activities.
1.4
Results
Family business background
More than half of all business owners had a business-owning family member
prior to starting their business. This suggests a high level of intergenerational
transmission of business ownership. Conditional on having a business-owning
family member, nearly half of the small business owners worked in that family
member's business. This finding suggests that intergenerational links in
entrepreneurship are not largely due to the acquisition of business human
capital. Overall, only 23 percent of small business owners worked in a family
business prior to starting or acquiring their business.
An alternative explanation is that the children of business owners become
partners with their parents or directly inherit businesses. In contrast to the
high likelihood of having a business-owning family member and working for
that family member for current entrepreneurs, very few small businesses are
inherited. The authors’ estimates indicate that only 1.6 percent of all small
businesses are inherited. This finding suggests that the role of business
inheritances in determining intergenerational links in entrepreneurship is
limited at best.
The CBO also includes information on whether the owner acquired the
business through a "transfer of ownership/gift". Even with the concern that it
may contain many other forms of business transfers besides transfers by
family members, only 6.6 percent of owners received their business through a
transfer of ownership or gift, suggesting that direct parent-to-child transfers
of businesses cannot represent a large percentage of all small businesses.
Related to business inheritances, financial transfers from parents to children
are shown to be an uncommon source of start-up capital among small
business owners. Only 6.4 percent of owners borrowed capital from their
family.
Fairlie and Robb thus conclude that similarities across family members in
entrepreneurial preferences must explain part of the positive effect of having
1
It is the only nationally representative dataset containing information on prior work
experience in businesses owned by family members and prior work experience in businesses
providing similar goods and services, which is critical for the analysis by Fairlie and Robb.
The main disadvantage is that the CBO does not contain information on a comparison group of
wage/salary workers.
7
a business-owning family member on the likelihood of becoming an
entrepreneur.
Human capital and background determinants of small business outcomes
Next, Fairlie and Robb examine whether these factors play a role in
determining small business outcomes conditioning on ownership. Small
business outcomes are measured using a number of performance measures:
the probability of a business closure from 1992-1996, the probability that a
firm has profits of at least $10,000 per year, sales, and the probability of
having employees.2
Similar to previous studies, business outcomes are strongly and positively
associated with the education level of the business owner. Having a family
business background persé is unimportant for small business outcomes. In
contrast, working in this family member's business has a positive and
statistically significant effect on all performance measures. Hence, the main
effect appears to be through the informal learning or apprenticeship type
training that occurs in working in a family business and not from simply
having a business-owning family member.
The strong effect of previous work experience in a family member's
business on small business outcomes suggests that family businesses provide
an important opportunity for family members to acquire human capital related
to operating a business. Correlations across family members in
entrepreneurial preferences are less important in contributing to the
intergenerational link in business success conditioning on business ownership
than in contributing to the intergenerational link in business ownership.
Management experience prior to starting or acquiring a business generally
improves business outcomes, but has a less consistent effect than experience
working for a close relative. Previous work experience in a business whose
goods and services were similar to those provided by his/her business, a
more general case of acquiring specific business human capital appears to be
very important.
The fact that the large and positive effect of previous work experience in a
family business is obtained while controlling for prior managerial experience
and similar business experience suggests that the measured effect is not
simply capturing the effects of management experience or specific business
human capital on small business outcomes. Instead, prior work experience in a
family member's business has an independent effect on small business
outcomes, which may in part be due to the acquisition of less specific, general
business human capital.
2
Nearly one quarter of small businesses existing in 1992 were not operating by 1996, and
slightly more than 30 percent of businesses report a net profit of at least $10,000. Small firms
hire 1.77 employees on average with only 21.3 percent hiring any employees. Finally, small
businesses had mean sales of $213,000 in 1992.
8
1.5
Conclusion
From a policy perspective the findings are important. Most disadvantaged
business development policies currently in place, such as set-asides and loan
assistance programs, are targeted towards alleviating financial constraints not
towards providing opportunities for work experience in small businesses.
Even programs providing mentoring, such as the Small Business
Administration's 8(a) Business Development Mentor-Protégé Program,
generally focus on technical, management and financial assistance,
subcontract support, and assistance in performing prime contracts through
joint venture arrangements. These programs do not explicitly provide
opportunities for would-be entrepreneurs to acquire general and specific
business human capital by working for other small business owners. The
findings from this research suggest that governmental programs providing
mentoring, internships or apprenticeship-type training may help to reduce
historical inequalities in business ownership patterns. More research,
however, is needed on the potential effectiveness of these types of programs,
especially from evaluations of experimental programs.
9
CHAPTER 2
Based on
“The entrepreneur’s entry mode:
Business takeover or new venture start?”
by
Simon C. Parker
University of Durham
&
C. Mirjam Van Praag
University of Amsterdam
ABSTRACT
Parker and Van Praag analyze the mode of entry of entrepreneurs. Do
individuals become entrepreneurs by taking over an established business or
by starting up a new venture from scratch? Their theoretical model predicts
that several individual- and firm-specific characteristics influence the optimal
mode of entry. The new venture creation mode is associated with higher
levels of schooling and wealth, whereas managerial experience, start-up
capital requirements and risk promote the takeover mode. Entrepreneurs
whose parents run a family firm are predicted to invest the least in schooling,
since schooling reduces search costs and these individuals have the lowest
probability of needing to search for a business opportunity outside their
family. A sample of data on entrepreneurs from the Netherlands is used to
test the theory; implications for policy-makers concerned about the survival
of family firms that lack within-family successors are also discussed.
10
2.1
Introduction
Recent literature focuses on entrepreneurship as a transition into independent
business ownership, and usually frames entrepreneurship in terms of new
venture. Individuals, however, can also take over an existing firm. One can
therefore separate the mode of entry from the entry decision itself.
There are at least two reasons why policy makers may be concerned with
the mode of entry. First, a key, but sometimes overlooked consideration in the
entrepreneurship debate is that policy makers need to remember the
importance of preserving economic value of existing firms as well as the
creation of value via new starts. In the Netherlands, for instance, 20,000 firms
per year are expected to seek takeover candidates in the next five years.
Moreover, the percentage of firms that is taken over by family members is
decreasing sharply. It can be costly and time-consuming to find suitable
successors from outside the family, which suggests that aggregate transaction
and operation costs are likely to increase as the number of family firms taken
over by ‘outsiders’ rises.
Second, public policy in many countries is increasingly being used to
promote entrepreneurship. If targeted policies are to provide the correct
incentives, it is necessary to take into account the mode of entry into
entrepreneurship as well as the gross entry flow. A set of policies designed to
promote new starts will not necessarily be suitable for individuals who are
contemplating entry by taking over an existing firm seeking a successor.
However, the entry mode of entrepreneurs has been little studied to date.
In particular, we still know little about which types of individual will match
with which types of firm as the owner of that firm.
Parker and Van Praag present a theoretical and empirical analysis of the
entrepreneurial entry mode decision to shed light on the following question:
When will individuals become entrepreneurs by starting up a business from
scratch and when will they take over instead an established firm looking for a
successor? And, in the case of a takeover, when will individuals take over a
family business, and when will they taker over one acquired from a third
party? Please note that the authors only analyze the entry strategies of
individuals who have chosen to be entrepreneurs; the decision to enter
entrepreneurship in the first place will not be analyzed below.
The issues discussed in this paper relate to two other strands of literature.
First, the authors shed additional light on the issue of family firm succession.
However, most studies in the succession literature to date have focused on
the problem from the founder’s perspective rather than the potential
successor’s as Parker and Van Praag do. A second strand of literature
analyzes business transfers and acquisitions. The focus of the current paper
is on entry into entrepreneurship: Business acquisition and transfer strategies
used by entrepreneurs wishing to grow their incumbent business are beyond
its scope.
2.2
Model
The theoretical model that Parker and Van Praag develop is based on a multistage decision process in which individuals choose formal human capital
11
(years of schooling) and an entry mode to maximize expected utility. The
model is used to explore the characteristics of entrepreneurs choosing
between entry modes, and the types of firm they will match with. In particular,
the authors trace the effects of schooling and family background on the mode
of entry, and analyze the roles played by several other variables including
business risk, wealth, required start-up capital, and previous managerial
experience. They also include the option of family takeover in their model,
and by comparing this with the other occupational choices of children of
family firms, the authors shed additional light on the important issue of family
firm succession.
Two types of entrepreneurs and three types of firms are distinguished in
the model. The two entrepreneur types are f and g: f types are born into a
family with parents who are (or used to be) entrepreneurs, and g types are
born into non-entrepreneurial families. The three firm types are family firms
looking for successors within the family, F; family firms seeking a successor
from outside the family, T; and new start-ups, N. The last two types of firms
can be operated by anyone, but by definition only f-type entrepreneurs can
operate the first type. The model is based on several assumptions.
Assumptions
The first assumption made is that for any entrepreneur there is only one firm
within each given firm type with a potentially successful match. All
entrepreneurs know the value of the chance of a successful match in advance.
The second assumption is that entrepreneurs first select a given firm type
to explore, and then search to identify the most favorable firm within the set
of firms of a given type to try out a match. In the first round, no search is
needed for f types who have available and try the F firm type first. For the
other firm types (i.e., N and T) search is always performed. Individuals
searching for the first time incur search costs. But they learn how to search
from this experience, so any subsequent searching for the most favorable firm
within another firm type occasioned by a bad match previously is cheaper.
General human capital (namely schooling) reduces initial search costs,
because analytical skills increase an individuals processing power.
The third assumption is that entrepreneurs cannot know whether they will
make a good match with the most favorable firm within a firm type until they
try it. If they do not make a good match, they must try another firm type
altogether. If they still fail to make a good match after exhausting all available
firm types they receive an outside wage in paid employment. This assumption
recognizes that not everyone is cut out to be an entrepreneur.
The fourth assumption is that new firms offer riskier payoffs than firms
that were acquired by a takeover.3
The fifth assumption is that in the event of a bad match, individuals make a
zero return, but they are able to wholly reclaim the assets invested in the
firm. All entrepreneurs can borrow if they cannot self-finance the initial
investment.
3
This assumption is consistent with evidence that new firms have more variable growth and
profit rates and lower survival rates than established firms do on average (Parker, 2004,
Chap. 9; Van Praag, 2005, Chap. 6).
12
Timing
An entrepreneur’s decision process can be divided into the following stages:
1. The entrepreneur chooses his optimal years of schooling.
2. Type f entrepreneurs now learn whether their family firm, F, is available:
for the part for whom it is they try F first. Those for whom it is a good match
remain in F, paying off the bank any loan; those for whom it is not a good
match proceed to stage 3.
3. Type g entrepreneurs and the part of unmatched f types from the previous
stage identify their preferred firm type, search and locate the most favorable
firm within that type. If a match between an entrepreneur and their chosen
firm type is good, the entrepreneur remains in that firm, obtains his payoff,
and repays the bank any loan. If it is a poor match, the entrepreneur invests in
the other firm type.
4. Remaining f and g type entrepreneurs who are unmatched with any firm are
forced to exit entrepreneurship and receive the net return in paid
employment.
Schooling choices
At stage 1, each individual chooses his optimal years of schooling and incurs
the cost to maximize expected future returns, taking into account the
possibility of unsuccessful matches in various entry modes in
entrepreneurship. f type entrepreneurs optimally acquire less formal human
capital than g types. The reason is that f’s face a smaller expected benefit
from their costly investment in human capital given the possibility that they
take the unique intra-family opportunity, whose net payoff does not depend
on costly human capital-enhanced search.
Entry strategies and firm purchase prices
Having chosen their human capital as above, both the g type entrepreneurs
and the unmatched f type entrepreneurs from stage 2, have to make the entry
choice at the start of stage 3 between T and N. As noted above, the benefit of
entering via T rather than N is that T is less risky. On the other hand, a T
firm has to be purchased for a higher price than the cost of starting N.
Furthermore, unmatched f types are willing to pay more to take over a
business than g types are because they have lower human capital and hence
income – and so are more risk averse. For similar reasons, the price of the T
firm will be higher the greater the risk in N and the lower the exogenous
capital endowment, as less wealthy individuals are more risk averse and less
willing to start a risky N. This is summarized in the following testable
proposition:
Proposition 1. Entrepreneurs who start new ventures are more likely to be
born into non-business-owning families than are entrepreneurs who take over
existing firms. As a consequence, f-types have pursued lower levels of
education than g-types.
Extending the model
The model can be extended naturally in several ways. One extension might
allow for structural differences in the firm types. Entrepreneurs who face the
highest start-up capital requirements and start-up risk will outbid
13
entrepreneurs of the same type for the right to take over an existing firm. It
therefore follows that:
Proposition 2. Entrepreneurs facing higher start-up capital requirements and
risk are more likely to take over an existing firm than to start up a new one.
A second extension might recognize that established firms are larger on
average than new starts, being more likely to employ large workforces.
Therefore, it is likely that entrepreneurs who take over these firms would
benefit management experience.
Proposition 3. Entrepreneurs with greater managerial experience will bid more
to take over an existing firm and so are more likely to take over a firm than to
start one up, compared to otherwise identical entrepreneurs with less
managerial experience.
A third extension recognizes that entrepreneurs seeking to identify new
opportunities have to search in a broader domain than those seeking to take
over firms whose existence is known. Entrepreneurs with higher levels of
education of both types are likely to be more efficient at processing large
amounts of information.
Proposition 4. Entrepreneurs with higher levels of education are more likely to
start a new firm than to take over an existing one, even after controlling for
the entrepreneur’s family (business) background.
2.4
Data
To test the practical validity of the four propositions resulting from the
theoretical model, Parker and Van Praag use a 1994 sample of 700
entrepreneurs that is broadly representative of the Dutch population of
entrepreneurs. Entrepreneurs are defined as individuals who started their own
business from scratch or who took over an existing firm.
The dataset contains a wide range of economic and demographic variables
including ones related to family background, entry mode, and human and
financial capital. With respect to entry mode, 9.5% took over a family firm
(actually by acquisition or inheritance), 7.4% took over another firm and
83.1% started a firm from scratch. Hence in total 17% of the firms were
started through a takeover of some kind.4 In the sample, 46 percent of the
entrepreneurs are born into entrepreneurial families.5 Apparently, this
percentage is much higher than the rate of business ownership in general,
pointing at strong intergenerational links in business ownership. It is
noteworthy that f types are significantly less likely to start a new firm from
scratch (70%) than g types are (94%).
4
The 9.5 percent of family firms that are either inherited or acquired from the family is
consistent with the numbers pertaining to the United States as provided in Chapter 1.
5
This high percentage is again very consistent with the percentages shown in the previous
chapter that pertain to the US.
14
Considering human capital variables, f types have significantly less
education on average than g types (14.1 versus 15.2 years). This is in
accordance with the predictions of the model
When it comes to comparing both average start-up capital requirements
and risk between firm types, as is required to test Proposition 2, individual
data are clearly unsuitable. Start-up requirements in terms of finance and
start-up risk can actually only be measured at the individual level for
entrepreneurs who really started up a business from scratch. However,
measurement of these variables is required for all entrepreneurs in the
sample. For this reason, the authors define both variables at the industry
level; 9 industries are distinguished.
2.4
Empirical results
Schooling choices
Looking at schooling choices, the second stage of the model generated a key
result, summarized in Proposition 1: that f types optimally choose lower levels
of education than g types. Consistent with the raw data, the regression results
(where the determinants of the number of years of formal education are
estimated while controlling for various relevant individual characteristics) also
support the notion that f-types acquire significantly lower levels of formal
education than g-types who do not have the opportunity to take over a family
business.
Entry modes
The authors test the four propositions using a simple probit model in which
the dependent variable equals one if a start-up (N) is chosen, and takes the
value zero if a takeover (T) is chosen.
F type entrepreneurs turn out to be significantly more likely to take over a
non-family business than to start up a business from scratch, compared with g
type entrepreneurs. Second, highly educated entrepreneurs are more likely to
start up a new firm instead of entering entrepreneurship through takeover.
These results are supportive of Propositions 1 and 4, the latter being
consistent with the notion that education has a productive role in new venture
starts by reducing search costs and enhancing success in managing high risk
and high return projects. When adding to the specification other forms of
human capital, in the form of various kinds of general and specific labor
market experience, the only variable in this category that becomes marginally
significant is a dummy for previous experience of managing people, which is
associated with a higher probability of becoming an entrepreneur through
takeover instead of start-up. This is consistent with Proposition 3. The
authors further include controls for (industry-specific) entry capital
requirements and risk, as defined in the previous section in order to test
Proposition 2. The results show that takeover becomes indeed relatively more
attractive when industry entry is more risky and/or more expensive. These
findings are consistent with the proposition, though the entry risk variable is
only marginally significant.
15
2.5
Conclusion
Parker and Van Praag have produced an economic theory about the entry
mode decisions and related human capital decisions of entrepreneurs. The
theory rendered four testable propositions. Parker and Van Praag have tested
these propositions empirically based on a representative dataset of Dutch
entrepreneurs. They find empirical support for all four propositions.
The subject that Parker and Van Praag discuss is new and requires followup research. Questions to be answered are, for instance, to what extent is
entry mode affected by these determinants in other countries than the
Netherlands? Are there any important differences across the EU and the US?
Are the returns to education different for start-ups than for takeovers?
If the results can be generalized, this study shows that the likelihood of
finding takeover candidates is highest among the offspring of entrepreneurs,
people with lower levels of education and more management experience.
Given the current scarcity of takeover candidates in Europe this is relevant
information for the organization of takeover markets where supply and
demand can meet, for the organization of capital markets for the takeover of
family firms as well as for the owner-managers of family firms needing a
successor themselves.
References
Parker, S. C. (2004), “The Economics of Self-employment
Entrepreneurship”. Cambridge: Cambridge University Press.
and
Van Praag, C.M. (2005), “Successful Entrepreneurship: Confronting Economic
Theory with Empirical Practice”, Cheltenham: Edward Elgar Publishers.
16
Part II
Education
17
CHAPTER 3
Based on
“Why are the returns to education higher for entrepreneurs than for
employees?”
by
Justin Van Der Sluis
University of Amsterdam
Mirjam Van Praag
University of Amsterdam
&
Arjen Van Witteloostuijn
University of Groningen & University of Durham
ABSTRACT
Van Der Sluis et al. compare the returns to education (RTE) between
entrepreneurs and employees in the United States. What percentage annual
income gain is derived from investing in a marginal year of education? Is this
percentage higher for entrepreneurs than for employees? By using improved
econometric techniques (IV), the authors find that the RTE are significantly
higher for entrepreneurs than for employees (14.8 percent and 10.8 percent,
respectively). Van Der Sluis et al. perform various analyses in an attempt to
explain the difference. They find (indirect) support for the argument that the
higher RTE for entrepreneurs is due to fewer (organizational) constraints
faced by entrepreneurs when optimizing the profitable employment of their
education in comparison to employees.
18
3.1
Introduction
There is one factor that both academic scholars and policy makers see as an
important determinant of entrepreneurial performance, namely human capital.
This study focuses on the measurement of the returns to human capital, in
particular formal education, for entrepreneurs relative to employees.
3.2
Empirical literature
Van Der Sluis, Van Praag and Vijverberg (2003) -in an attempt to synthesize
the empirical literature about the relationship between education and
entrepreneurship by means of a meta-analysis- conclude that there are
several consistent findings with respect to the relationship between education
on the entry and performance in(to) entrepreneurship. Four outcomes are
relevant to the current study.
First, the relationship between education and selection into an
entrepreneurial position is mostly insignificant, i.e., in 75 percent of the cases.
Hence, education plays no significant role in the decision whether people
become entrepreneurs (instead of wage employees). However, the
relationship between schooling and performance is unambiguously positive
and significant in 67 percent of the observed studies. Irrespective of how the
performance of entrepreneurs is measured, i.e. in terms of profit, income,
growth or survival, entrepreneurs with higher levels of education are
associated with better entrepreneurial outcomes.
Second, the return to a marginal year of schooling for entrepreneurs in
terms of income is 6.1 percent, on average. Third, the returns to education
turn out to be of similar levels for entrepreneurs and employees.
The fourth conclusion from the analysis of the literature is that previous
studies have not yet employed estimation strategies that account for the
endogenous nature of schooling and unobserved individual characteristics.
There are at least two possible sources of inconsistency when education is
taken to be exogenous in the income or performance equation. First, there
may be unobserved individual characteristics, such as ability and motivation,
that affect both the schooling level attained and subsequent entrepreneurial
performance. Second, the schooling decision is probably endogenous in a
performance equation because individuals are likely to base their schooling
investment decision, at least in part, on their perceptions of the expected
payoffs to their investment. Clearly, the entrepreneurship literature can much
better use econometric methods that take account of these difficulties. Such
methods are common practice in labor economics where researchers measure
the returns to education of wage employees. One of these methods is used in
the article by Van Der Sluis, Van Praag and Van Witteloostuijn as an
estimation strategy: the instrumental variables (IV) approach.6
6
Since the meta-analysis of the literature, two studies have been performed that use the IVmethodology to measure the returns to education for entrepreneurs [Parker and Van Praag
(2006) and Van Der Sluis et al (2003)], whereas the current study is the third. In the current
study, the returns to education for entrepreneurs relative to employees are re-evaluated,
repairing some of the drawbacks that characterized the earlier attempts.
19
3.3
Data
Sample
The effect of education on incomes for both entrepreneurs and employees is
estimated on a sample drawn from the National Longitudinal Survey of Youth
(NLSY) in the US. The nationally representative part of the NLSY consists of
6,111 individuals aged between 14 and 22 years in 1979. They have been
interviewed annually up to 1994, and since then on a bi-annual basis. Within
each observed year, the authors include in their sample all persons who are
entrepreneurs or employees. The resulting sample size per year includes, on
average, 2,646 entrepreneurs/employees.7
Features
An important feature of the sample is that it includes both entrepreneurs and
employees, and it records individuals' switches between these states over
time. All entrepreneurship spells, also short ones, are recorded. Therefore,
the sub-sample of entrepreneurs does not suffer from survival bias, i.e., the
returns to education will not pertain to surviving entrepreneurs only.
Moreover, the incomes and all other relevant variables are measured in a
comparable way for both groups such that the returns to education for
employees and entrepreneurs can be compared.
Another appealing feature of the NLSY is that it contains the Armed
Services Vocational Aptitude Battery (ASVAB), an IQ-like test score that is
used as a proxy for general ability (also used in the study discussed in
Chapter 5).8 An ability measure like this is beneficial for the correct
measurement of the returns to education: Individuals with higher ability levels
are more likely to opt for higher levels of education and they are more likely
to have higher levels of income, ceteris paribus.
Another quality of the NLSY is the presence of detailed family and
individual background variables. Some of these qualify as identifying
instruments as they are possibly good predictors of the educational level of
the respondent or the respondent's choice for entrepreneurship, while
otherwise independent of their future earnings.
Empirical methodology
An income equation is estimated by means of a random effects (RE) model.
Using the IV-estimation strategy in this case implies that two first-stage
equations are estimated, i.e. of education and the decision to become an
occupational entrepreneur. The predicted values resulting from these
equations can be used as the instrumented values of education and
entrepreneurship in the RE model.9
7
The econometric tools used to analyze these data, i.e. panel data analyses, take into account
that the observations over time for each individual are not independent from each other.
8
The authors recalculate the values of the test score by using a common approach to correct
for age and education effects.
9
A set of four identifying instruments for education is extracted from the NLSY data: (1)
“Magazines present in the household at age 14”, (2) “Library card present in the household at
age 14”, (3) “The presence of a stepparent in the household”, (4) “Number of siblings in the
household”. This set of instrument turns out to be of sufficient quality and validity.
20
Besides instruments for schooling, an instrument is required for the
entrepreneurship selection equation. Correcting for this kind of selectivity has
proven to be difficult. Van Der Sluis et al investigate two different routes.
Although both two routes are far from perfect, they both lead to the same
result, leading to more confidence about the findings. Since the results
indicate that correcting for selectivity is not too relevant, I do not discuss the
methods to correct for selectivity.
3.4
Estimation results
As a benchmark, the authors first show the result from estimating the income
equation without using an adequate identification strategy and without
controlling for ability. The result from this approach is comparable to the
(possibly inconsistent) estimates that had been obtained previously. The
estimated returns to education are 7.3 percent for entrepreneurs and 6
percent for employees. The returns are thus somewhat higher for
entrepreneurs than for employees, and this difference is significant.
The next step is to include ability controls and instrument education with
the discussed set of family background variables (IV). The results show
significantly higher estimates of the returns to education. For employees, the
returns jump from 6 percent to 10.8 percent. A novel finding is the greater
jump in the returns to education for entrepreneurs from 7.3 percent to 14.8
percent. This leads to the remarkable result that the returns to education for
entrepreneurs are a significant 37 percent higher than the comparable returns
for employees.
3.5
Why are entrepreneurs’ returns to education higher?
This section is devoted to finding an explanation for the result that the
estimated returns to education are higher for entrepreneurs. The authors test
the validity of various possible explanations.
The first test relates to the question as to whether the difference in
returns to education between entrepreneurs and employees can be attributed
to a risk premium. More highly educated individuals would perhaps require a
higher risk premium for being an entrepreneur if higher educated individuals
experience more additional income risk as an entrepreneur compared to an
employee vis-à-vis lower educated individuals. However, the authors find
that the higher returns to education for entrepreneurs are not a kind of risk
premium.
The second check concerns underreporting of incomes. Recent evidence
shows that entrepreneurs underreport more than employees and that bluecollar entrepreneurs underreport more than white-collar entrepreneurs
(Lyssiotou et al., 2004). Since blue-collar entrepreneurs have a lower
education on average than white-collar entrepreneurs, the returns to
education estimate for the total population of entrepreneurs could be upward
biased. This in turn might explain the gap in returns to education between
employees and entrepreneurs. However, a test leads to the conclusion that
underreporting is not the explanation.
21
A third check relates to the possibility that some entrepreneurs have
erroneously included the returns to (business) capital in their reported
income. This could explain the result if more highly educated entrepreneurs
have higher returns to capital than lower educated entrepreneurs. To test
this, the income for entrepreneurs who receive a business income from
unincorporated businesses is adjusted and the previously estimated income
equation is re-estimated. The adjustment for capital returns hardly reduces
the difference in returns between entrepreneurs and employees.
Fourth, could the difference in returns to education (in terms of hourly
earnings) between entrepreneurs and employees be explained by the inclusion
of part-time entrepreneurs and employees in the sample? For instance, this
could explain the result if working part-time is punished more heavily in
terms of hourly earnings for entrepreneurs than for employees and if parttime workers have lower levels of education. A test reveals that the
difference becomes smaller, but remains significant.
The fifth check is based on the idea that professional workers like lawyers
and medical doctors have high earnings, are highly educated and are often
self-employed. However, excluding professional workers from the sample
does not decrease the estimated difference between entrepreneurs and
employees substantially.
So why is education more valuable for entrepreneurs than for employees?
An organization-oriented explanation could be that entrepreneurs have more
freedom to optimize their employment of education than employees.
Entrepreneurs are not constrained by rules from superiors and can decide on
how to employ their education most productively. If entrepreneurs are in a
position to better control the profitable employment of their education, this
might be an explanation for the higher returns to education for entrepreneurs
vis-à-vis employees.
If it is true that a better control of the environment influences the
possibility to optimize the returns to education, it might also be true that
individuals' perceived control of the environment affects their returns to
education. Entrepreneurs and employees who have the perception that they
are in control of their environment, should then experience higher average
returns to education than others. This would support the control-related
explanation indirectly. Individuals’ perceived control of the environment is
measured by the personality trait ‘locus-of-control’. Such a measure has been
included in the dataset at hand. The control-related explanation turns out
valid. It holds true for both employees and entrepreneurs, but for
entrepreneurs to a higher degree, that the returns to education are higher the
more an individual perceives to be in control. The authors conclude that
control matters and that it is likely to be an explanation for the higher returns
to education obtained by entrepreneurs.
3.6
Conclusion
Van Der Sluis et al. have compared the returns to education for entrepreneurs
and employees. The methodological rigor applied in studies of the returns to
education for employees has been the benchmark. A novel finding then shows
22
up: The returns to education are significantly higher for entrepreneurs (14.8
percent) than for employees (10.8 percent).
The explanation supported by the test outcomes that utilize the locus-ofcontrol concept is that entrepreneurship gives better opportunities to optimize
one's education and subsequent returns. Altogether, these findings bear
implications for researchers and policy makers alike.
The observation that conventional estimates are biased and that the extent
of bias differs per labor market group is an interesting starting point for
researchers investigating returns to education for entrepreneurs. Further
research should first produce more evidence of the relative returns to
education for entrepreneurs by using modern estimation strategies, and then
aim at understanding the differences in terms of returns to education between
entrepreneurs and employees.
Before discussing policy implications, the authors elaborate on the
remaining untested assumptions required to translate the estimation results
into policy implications. First, they assume that the development of more
entrepreneurship is economically valuable. Second, they assume that the
difference between the social and private benefits of entrepreneurial activity
is at least as large as this difference is for employees. A successful
entrepreneur is, for example, more likely to influence competition and
innovation in a market positively than is an employee. Third, it is assumed
that individuals invest in schooling at a stage in their lives at which they do
not know yet, in general, whether they will become entrepreneurs or
employees, or a (sequential) combination of both. The fourth assumption is
that individuals and policy makers share the common opinion that the returns
to education are similar or slightly different, at most, for entrepreneurs and
for employees.
Clearly, the finding that the entrepreneurial returns to education are high,
and that education is therefore a key success factor for a starting enterprise,
is informative for individual labor market decisions, the development of
educational policies, as well as for bankers' and capital suppliers' strategies
with respect to (selecting) starters. The finding could motivate governments
to stimulate higher education for (prospective) entrepreneurs. Alternatively,
policy makers could stimulate higher educated individuals to opt for an
entrepreneurial career. The first route would increase the likelihood that
entrepreneurs will perform better, and that they will generate more benefits
that will not only accrue to the entrepreneurs themselves, but also to society
as a whole. The second route appeals to the fact that entrepreneurship seems
not to be the favored option among highly educated individuals. Van Der Sluis
et al. strongly believe in the benefits of governmental programs to stimulate
the awareness among college and university students of the attractiveness of
entrepreneurship. Future research into the entrepreneurial returns to
education in general and of specific types of education may further increase
the effectiveness of such policies.
References
Lyssiotou, P., P. Pashardes and T. Stengos (2004), “Estimates of the black
economy based on consumer demand approaches”. Economic Journal,
114(497), pp. 622-640.
23
Parker, S.C. and C.M. Van Praag (2006), “Schooling, capital constraints and
entrepreneurial performance: The endogenous triangle”. Forthcoming in
Journal of Business and Economic Statistics .
Van der Sluis, J., C.M. Van Praag and W. Vijverberg (2003),
“Entrepreneurship, selection and performance: A meta-analysis of the role of
education”. Working Paper. University of Amsterdam.
24
Part III
General Human Capital
and
Specific Human Capital
25
CHAPTER 4
Based on:
“Wage policies and incentives to invest in firm-specific human capital”
by
George Baker
Harvard Business School
Nancy Dean Bealieu
Harvard Business School
&
Cristian Voicu
Harvard Business School
26
4.1
Introduction
Firm-specific human capital creates a gap between what an employee is
worth to her current employer, and what she is worth in her next-best
alternative job. This gap is a source of rents for both the employee and the
firm, allowing employees to make more than they would make in another job,
and allowing a firm to pay less than the worker is actually worth to the firm.
How this rent is divided is an important part of a firm’s wage policy. Becker
argued that employees should pay for all investments in general human
capital, while investments in firm-specific human capital should be split
between the firm and the worker.
An implicit assumption of this argument is that investments in human
capital are contractible, and that firms and workers can agree on the levels of
these investments. However, often investment in human capital involves
subtle actions by employees that are difficult to contract on. In addition, they
often involve private costs to the employee that are hard to measure and
compensate for. Thus, investment in human capital has many of the
characteristics of a moral hazard problem. In this theoretical paper, the
authors examine how a firm’s wage policy should split the rents generated by
the existence of firm-specific human capital, under the assumption that the
firm needs to induce employees to make such costly non-contractible
investments.
The tradeoff is simple: the more of the rents the firm’s wage policy gives
to the employee, the higher are the employee’s incentives to invest in firmspecific human capital. This is not the end of the story, however. Investment
in firm-specific human capital, if it crowds out investment in general human
capital, is risky to the employee since it lowers her outside wage and raises
the possibility that the firm will exploit the worker by lowering wages in the
future. Baker et al. examine this possibility theoretically, and show that firms
would like (ex ante) to commit to not exploit workers in this way.
The authors develop several models that allow them to explore the
implications of the fact that employees choose their levels of investment in
firm-specific human capital, and that firms design wage policies to affect this
choice. The models probe the trade-offs discussed above. In addition, the
models allow a better understanding of, what the authors believe to be a
vicious/virtuous circle of fragility that affects human capital-intensive firms: it
often seems that some professional firms organized as a partnership are
perpetually on the brink of dissolution, with their most valuable partners
constantly on the lookout for better opportunities, threatening to take their
skills elsewhere, and devoting energy to increasing their value on the market
(that is, investing in general human capital) rather than maximizing their value
to their current firm.
Baker et al. derive four results by means of a one-period model. They
then extend their analysis in the framework of a multi-period model in which
they distinguish two cases: full commitment and no commitment.
In what follows, the model assumptions are addressed first, followed by a
discussion of the results and implications of the one-period model and a very
brief and incomplete discussion of the multi-period model, respectively.
Finally, I [editor] add a conclusion of the empirical implications of these
models for entrepreneurship and small firms.
27
4.2.
Model
Assumptions
A number of assumptions are made. First, investments in human capital
require the worker's personal effort in order to become productive.
Furthermore, there is always the possibility of seeking alternative
employment. The outside offer is a function of the amount of general human
capital, but not of the amount of firm-specific human capital. The employee's
best outside offer might be higher than their general capital due to uncertainty
about the worker’s current productivity, which is partly resulting from luck.
Baker et al. further assume that the stocks of general and firm-specific
human capital are not contractible. The assumption that value to the firm is
contractible while general capital is not seems reasonable. First, firms invest
in many performance measurement systems to allow them to determine an
employee’s total contribution to the firm. They invest in few such systems to
determine an employee’s value to other firms. In addition, using general
capital in the contract would induce a different kind of moral hazard problem,
since the worker would have an incentive to generate outside offers.
Moreover, the employee and the firm are assumed to be risk neutral. In order
for the firm to make profits, it must satisfy the worker’s participation
constraint, i.e. the worker must not obtain higher levels of satisfaction at her
next best alternative job.
Timing of the model within each period
In the first stage, the firm offers a wage contract to the employee. Second,
the employee chooses her levels of effort with which she invests in a certain
combination of general and firm-specific human capital. The employee may
choose not to invest in firm-specific human capital at all. Baker et al. refer to
this as a participation choice by the worker. In order for the firm to make
profits, it must satisfy the worker’s participation constraint and induce the
employee to invest in firm-specific human capital. After making her
investment decisions, the employee then checks in the third stage to see if
the realized outside alternative is better than the expected wage inside the
firm. If so, the employee produces output and the firm pays the wage in the
fourth stage. Otherwise, the employee leaves with her stocks of human
capital, earns her next best alternative wage and the firm gets no profit in the
fourth stage.
One period model
To solve the moral hazard problem, Baker et al. work in two steps. First, they
solve for the optimal investments in firm-specific and general human capital
for a given wage policy. Second, they explore the wage policy that maximizes
total firm profits.10
10
As either the number of outside firms increases, or the variance of their offers increases,
the value of the next best alternative increases such that it is more costly for the incumbent
firm to satisfy the worker’s participation constraint. In any case, Baker et al. show that the
worker can receive at least the value of her general human capital in alternative employment.
Therefore, the firm needs to pay above the value of the worker’s general human capital if it
wants the worker to stay and work at the firm.
28
The first result generated by the model is that efforts in general and firmspecific capital are substitutes in the worker’s expected utility function. The
intuition is simple: when the worker invests in more general human capital, he
reduces the likelihood that he will stay with the firm. This reduces the
marginal utility of investing in firm specific human capital. If he invests more
in firm-specific human capital, he lowers the probability of leaving, and
decreases the marginal utility of investment in general human capital.
The second result from the model is that there is a critical level at which
the employee’s investment in firm-specific human capital will jump from zero
to a positive level. Hence, there is a discontinuity in the worker’s optimization
problem: she either invests nothing in firm-specific human capital (and
decides to leave) or she invests a significant portion of her efforts in firmspecific human capital such that she may stay.
The third result derived from the authors’ one-period model is that it
might be the case that no profitable wage policy exists. If the value of firmspecific human capital investment is low and/or the probability of a large
outside wage offer is high, there is no profitable wage policy that can induce
any investment in firm-specific human capital. This result—that there is no
profitable wage policy—occurs when the returns to the employee of investing
in firm-specific human capital, in terms of both wages and the increased
probability of staying, are not worth the cost of investing in firm-specific
human capital. When this is the case, wages are equal to the employee's
accumulated stock of general human capital and turnover is very high.
The fourth result from the authors’ one-period model is that, for certain
parameters of the model, the optimal wage policy makes the employee strictly
better off investing in positive levels of firm-specific human capital than she
would be at zero investment. This occurs when the value of firm-specific
human capital investment is high, and the probability of large outside wage
offers is low.
Discussion and implications
The model makes some predictions about the stability of human capitalintensive firms. Baker et al. argue that in particular the non-convexity of the
employee’s investment problem, and the resultant lack of continuity in her
investment choice and the firm’s optimal wage policy, lead to the conclusion
that some firms might be quite fragile. In certain environments, the firm’s
optimal wage policy makes the worker just indifferent between investing in
positive levels of firm-specific human capital and investing in none, leading to
this fragility. In those firms where the optimal wage policy is bound by the
participation constraint, employees are constantly on the brink of defecting,
reducing their investments in firm specific human capital, and focusing on
their outside market value. Thus, firms will be vulnerable to these kinds of
changes and hence fragile.
In other environments, where the marginal value of investment in firmspecific human capital is higher or where the highest outside offer is lower,
due to fewer competing firms in the labor market or less variability of the
wage offers in the labor market, the firm will have a more robust wage policy.
In these cases, the firm’s optimal wage policy gives the employee more utility
than she would get from not investing in firm-specific human capital. In this
29
situation, small changes in the environment do not threaten the firm with mass
defections.
Hence, three different regimes can be indicated dependent on the marginal
value of investing in firm-specific human capital and on the number and
variability of outside offers arriving (stochastically) at workers. The first
regime occurs if the marginal value of investing in firm-specific human capital
is low, whereas many outside offers arrive and/or they are highly variable
such that it is more costly to satisfy the participation constraint and hence
keep the worker. In this case, no profitable wage policy will exist, workers
will not invest in firm-specific human capital and turnover will be high. The
second regime represents the other extreme in which the marginal value of
firm-specific human capital is high and outside offers are relatively cheap to
match. In that case, workers will certainly invest in firm-specific human
capital, turnover will be low and firms will be stable and offer robust wage
policies. The third regime represents the intermediary case: investments in
firm-specific human capital will be positive, turnover will be at the
intermediate level and firms can easily collapse by offering fragile wage
policies.
We now turn to discussing the authors’ multi-period version of this model
in order to understand the dynamics of these stocks, and the firm’s dynamic
wage policy. The situations with full and without commitment are discussed.
Full commitment means that the firm is able to commit to the complete path of
future marginal values of investments in human capital. It is assumed that both
the firm and the worker have perfect foresight, i.e. they can fully anticipate
the future effects of the investments made in firm-specific and general human
capital.
Multiple periods with full commitment
It is shown that a long horizon and the ability to commit to a wage policy
increase the incentives to invest in human capital substantially. Baker et al.
further show that the worker will either leave in period 1 or stay until the end.
The wage contract with commitment generates significantly higher levels
of investment, higher wages, and higher profits. However, the investment in
general human capital is lower than if the worker had quit. This highlights the
risk identified in the introduction: when an employee invests in firm-specific
human at the expense of general human capital, she lowers her outside
alternative and makes herself vulnerable to the firm. This risk to the
employee is exacerbated by the fact that this full commitment wage policy is
dynamically inconsistent. After the first period, the firm would like to reduce
the ‘payout ratio’ in order to “take back” some of the investment in firmspecific human capital that the worker makes in early periods.
Multiple periods without commitment
The distinction between credible commitment and no commitment is essential.
The firm would prefer to commit to a wage policy that induces the worker to
invest in firm-specific human capital. However, after the employee has made
these investments, the firm finds it tempting to change the wage policy to
capture the existing stocks of human capital. Baker et al. analyze that when
the worker's participation constraint will be binding, investments in firm
30
specific human capital are low and investments in general human capital are
high.
4.3. Conclusions and relevance for entrepreneurship11
Differences in labor market mobility affect company strategies, industry
structure, innovation et cetera. When turnover is high, firms are less likely to
reward for investments in firm-specific human capital and workers will be
less inclined to invest in it. Hence, many labor market movements result and
the circle is complete. This situation is likely to occur in markets where there
are many firms competing for workers based on their levels of general human
capital and generating highly varying outside offers. Hence, this situation can
be likely to be found in markets where there is much entry and exit of firms
and where entrepreneurship plays an important role. Moreover, if workers
invest more in general human capital this also increases the likelihood that
they will become entrepreneurs themselves. If firm-specific human capital is
not very valuable, the tendency to become an entrepreneur will increase. In
addition, if bankruptcy is more likely to occur, i.e. a higher likelihood of firm
exits, the marginal value of the worker’s investment in firm-specific human
capital will decrease proportionally with the probability of bankruptcy, leading
to less investment in firm-specific human capital in turn (and more turnover
and entrepreneurial spawning). Finally, if starting up a firm is a likely activity,
it is also a likely outside option. Again, adding more outside options leads to
less investment in firm-specific human capital.
On the contrary, if firm-specific human capital is very valuable, companies
are more likely to adopt wage policies that reward for firm-specific human
capital. Hence, turnover will be reduced and entrepreneurial spawning will be
lower. In addition, if firm-specific human capital is more important, the
likelihood of hold-up problems increases, i.e., firms cannot commit easily to
reward a worker’s investment in firm-specific human capital. Therefore,
vertical integration will be more likely, and again the role for
entrepreneurship will diminish.
Based on the models by Baker and his coauthors one can conclude that, in
general, there exists a quite strong relationship between the value of firmspecific human capital in a labor market (profession) and the occurrence of
entrepreneurial ventures. The higher the value of firm-specific human capital,
the less likely are entrepreneurial ventures for the various reasons set out.
This relationship might be worth of further empirical analyses.
References
Becker, G. S. (1964), “Human Capital”. NY: Columbia University Press.
11
Added by the editor based on George Baker’s presentation at the ACE/G2G conference
“Entrepreneurship and Human Capital”.
31
CHAPTER 5
Based on:
“Returns to Intelligence: Entrepreneurs versus Employees”
by
Justin Van Der Sluis
University of Amsterdam
&
C. Mirjam Van Praag
University of Amsterdam
ABSTRACT
In this study Van Der Sluis & Van Praag measure to what extent intelligence
is productive for entrepreneurs as compared to employees. They answer the
following three questions empirically and distinguish between entrepreneurs
and employees; (1) To what extent does an individual's general intelligence
level affect productivity? (2) Do different areas of intelligence (e.g. math
ability, language ability etc.) affect productivity differently? and (3) To what
extent does the spread in these areas of intelligence affect an individual's
labor market productivity? The latter question is related to and extends
Lazear's Jacks-of all-Trades (JAT) theory pertaining to entrepreneurs.
Productivity is measured as the average annual income derived from labor
market activities.
Van Der Sluis and Van Praag find that an individual's level of general
intelligence increases both entrepreneurs' and employees' incomes to the
same extent. Moreover, entrepreneurs benefit from specific areas of
intelligence that are partly different from the specific areas of intelligence
that are valuable for employees. The spread across the various areas of
intelligence is another determinant of earning power. The higher the spread,
the lower are earnings, in particular for entrepreneurs. This finding supports
Lazear's JAT theory.
32
5.1 Introduction
Previous empirical analyses have demonstrated that human capital is an
important determinant of entrepreneur performance. Both education and
specific sorts of labor market experience have significant effects on the
performance of entrepreneurs.
Besides these sorts of acquired human capital, a specific sort of human
capital, i.e. intelligence, might also affect the performance of entrepreneurs.
Little is known yet about the effect of intelligence on the performance of
entrepreneurs. More in particular, no empirical research has yet taken place
to measure the labor market value of (various areas of) intelligence for
entrepreneurs and compared them to the value for employees. Comparing the
returns to intelligence for entrepreneurs and employees is the aim of the
study by Van Der Sluis and Van Praag. They use individual incomes as the
performance measure for both labor market segments distinguished, i.e.
entrepreneurs and employees (as in Chapter 3).
This empirical study leads to a better understanding of how and to what
extent intelligence enhances entrepreneurial incomes and how this compares
to the effect of intelligence on the incomes of employees. In particular, the
authors analyze the effects of general intelligence, specific areas of
intelligence and the spread of intelligence across specific areas of intelligence
on the incomes of entrepreneurs and employees.12
These analyses enable answering the following three questions
empirically, due to the availability of measures of various sorts of intelligence
in the dataset used (the NLS youth sample, see Chapter 3). All three questions
are answered for both entrepreneurs and employees. (1) To what extent does
an individual's general intelligence level affect income? (2) Do different areas
of intelligence (like math, clerical, language, technical and social intelligence)
affect incomes differently?, and (3) To what extent does the spread in these
areas of intelligence affect an individual's income? Comparing the results from
the third analysis across employees and entrepreneurs will further our
understanding of Lazear's Jacks-of-all-Trades (JAT) theory, which indicates
that entrepreneurs, -in contrast to employees- need a wide variety of skills.
5.2.
Literature and hypotheses
The returns to general ability
According to psychologists the effect of intelligence on job performance is
mainly caused by the impact of intelligence on the acquisition of job
knowledge. As both entrepreneurs and employees are likely to require job
knowledge and dealing with complex situations, Van Der Sluis and Van Praag
expect a positive return to general ability for both labor market groups.
12
They account for the possible endogeneity in the decision to belong to either of the labor
market segments. Moreover, they perform the study upon including and excluding an
instrumented measure of education, thereby acknowledging that education is possibly an
endogenous variable in income equations.
33
The returns to specific abilities
Based on the available measurements of the relationship between verbal
ability and labor market performance, Van Der Sluis and Van Praag expect to
find a small positive non-linear effect of verbal ability on the performance of
both employees and entrepreneurs; they do not have any expectations about
the difference in returns to verbal ability across entrepreneurs and
employees.
The returns to math ability are mostly found to be significantly positive.
Therefore, Van Der Sluis and Van Praag expect math ability to have a positive
effect on labor market outcomes for both employees and entrepreneurs.
Again, they have no upfront expectation about the difference in returns to this
type of ability across entrepreneurs and employees.
Furthermore, they expect that technical ability is even more valuable for
entrepreneurs than for employees, since successful entrepreneurship often
depends on process or product innovation.
Finally, the authors expect social ability to have a positive and significant
effect on labor market outcomes for both employees and entrepreneurs.
Based on the suggestion by Baron and Markman (2003) they expect the effect
to be stronger for entrepreneurs than for employees.
The returns to a balanced set of abilities
A recent stream of papers pays attention to the combination of different
abilities instead of merely the level of abilities. Lazear's theory poses that,
individuals with a broad set of balanced competencies across different fields,
i.e. “Jacks-of-all-trades” (JAT), are more apt for entrepreneurship than those
who have a very unbalanced set of competencies, i.e. specialists (2005).
According to Lazear, the JAT characteristic determines the selection into
entrepreneurship since being a JAT is more valuable for entrepreneurs than
for employees. Van Der Sluis and Van Praag use an intelligence-based
measure of JAT, as was discussed in the introductory section. The advantage
of the intelligence measures, as compared to the measures used by Lazear, is
twofold. It is not influenced by the anticipated decision to become an
entrepreneur, whereas the curriculum choice or job role choice of a person
might be influenced by this anticipation, leading to unclear causality. Another,
but related, advantage of their JAT measure -measured at a relatively young
age- is that Van Der Sluis and Van Praag only measure the effect of being
endowed and raised as a JAT such that the effect is not mixed up with
possible effects of investments in schooling and labor market roles to become
a JAT. Second, they do not evaluate the effect of being a JAT on the selection
into entrepreneurship, but rather on the performance of entrepreneurs.
5.3.
Data
General ability, specific ability and spread in abilities
The intelligence test used is the Armed Service Vocational Aptitude
Battery (ASVAB). The ASVAB is a test developed by the U.S. Department of
Defense in the 1960s for the purpose of recruiting military personnel and is
still used for this purpose. In 1980 the ASVAB was added to the NLSY
questionnaire with the purpose of generating a benchmark sample for the
34
military that is representative of the total U.S. population and not only for the
military. Like most other intelligence tests, the ASVAB is built up of several
components. Unlike most intelligence tests, the ASVAB is administered at a
relatively young age of the respondents, i.e. between 14 and 23 years old,
such that the test outcomes are (almost) not affected by (future) labor market
choices.13
Van Der Sluis and Van Praag extract five specific ability levels from the
dataset: (1.) language ability, corresponding to `paragraph comprehension'
(2.) mathematical ability, corresponding to `mathematics knowledge' (3.)
technical ability, corresponding to `mechanical comprehension' (4.) clerical
ability, corresponding to `coding speed' and (5.) social ability.14
Moreover, and to test Lazear's JAT theory implying that a balanced set of
specific abilities would enhance entrepreneurial performance, a measure of
spread is required. The coefficient of variation measured across the individual
scores on the five types of specific ability included in their study will serve as
a measure of spread or `ability variation'.
The average general ability level is equal for entrepreneurs and
employees. The same is true for the five ability measures and the ability
variation.
Empirical methodology
A regression model is used to estimate returns on general ability, specific
abilities and the ability variation, while many control variables are included in
the model. The dependent variable is (log) income.
5.4. Results
Total population
General ability has a significant positive effect on income, also in case of
controlling for the differences generated by education levels. The results
further indicate that only at very low levels of verbal ability does an increase
in the level lead to a higher income. Moreover, the effect of mathematical
ability on income is only positive at the higher end of the mathematical ability
distribution: only if one is really good at mathematics as compared to other
people with the same schooling levels et cetera, does it affect incomes
positively. Technical ability has a significantly positive effect on income
irrespective of one’s level of technical ability. Furthermore, clerical ability has
a strong and positive effect on performance, whereas the effect of social
ability is significantly positive as well, but smaller. Van Der Sluis and Van
Praag conclude that the returns to general ability and all specific abilities are
positive and significant, at least for part of their distributions.
13
To remove the age and education effects from the ASVAB component scores Van Der Sluis
and Van Praag apply a simple but approved method.
14
These are as orthogonal as possible to each other.
35
The results show that ability variation is not significantly related to the
income of the average labor force participant. In other words, people do not
benefit from a more balanced set of abilities.15
Entrepreneurs versus employees
Entrepreneurs benefit as much from their general ability vis-à-vis employees.
The returns to education are on average higher for entrepreneurs than for
employees. As it turns out, there are substantial differences between
entrepreneurs and employees for all five specific abilities.
As was the case for the general population, only a (very) basic level of
verbal ability seems to be productive for each group. For entrepreneurs, a
much lower percentage of the verbal ability distribution –the bottom end- can
benefit from this ability in terms of income than for employees –the bottom
end and somewhat higher-. It turns out that mathematical ability is productive
only at the (for entrepreneurs very) high end of the sample distribution of
mathematical ability levels.
The returns to technical ability are positive and significant for both
entrepreneurs and employees, and more so for entrepreneurs. As expected,
clerical ability has positive returns for employees but not for entrepreneurs.
Finally, the results show that social ability renders much higher returns for
entrepreneurs than for employees, according to the expectations of the
authors.
Entrepreneurs have a very large and significant return to a balanced set of
abilities (i.e. low variation). Employees, on the other hand, have an
insignificant or even negative return to a balanced skill set. For them, being a
JAT is not beneficial in terms of generating additional income. On the
contrary, if anything, specialists seem to earn higher incomes as employees
than JATs. This result is consistent with Lazear's JAT theory: JAT's have a
comparative advantage in entrepreneurship.
5.5.
Discussion and conclusion
Van Der Sluis and Van Praag have measured the returns to intelligence for
entrepreneurs and employees. In what follows, I first discuss to what extent
their findings are consistent with the hypotheses. Second, I will discuss the
authors’ views on some of the limitations pertaining to this study in this novel
area, the questions that remain unanswered and recommendations for future
research. I then summarize the authors’ conclusions and recommendations as
to how their results can be used in practice.
Discussion of results
A comparison of the return to general intelligence between entrepreneurs and
employees shows that these are as high for entrepreneurs as for employees.
The effects of the five specific areas of intelligence are not the same for
entrepreneurs and employees. As expected, the effect of technical and social
15
The results show that the returns to education are on average 11 percent. Parental
education levels are not related to the income of their offspring.
36
intelligence is larger for entrepreneurs than for employees. The effects of
mathematical and verbal ability are stronger for employees: This rejects the
authors’ hypotheses that the returns to verbal and math ability would be
similar for entrepreneurs and employees.
Based on Lazear's Jacks-of-All-Trades theory, Van Der Sluis and Van
Praag also hypothesized that the return to being a JAT would be higher for
entrepreneurs than for employees. Their results strongly support this
hypothesis. They conclude that being born and raised as a JAT increases the
likelihood of successful entrepreneurship, apart from the possible effect that
people who intend to become entrepreneurs invest in schooling and labor
market experiences that turn them into JATs. All in all, the empirical results
are supportive of the (largely explorative) hypotheses that were based on the
somewhat scarce and scattered psychological and economic literature in this
area.
Limitations and recommendations for future research
Although the authors have an understanding of why entrepreneurs would
generate higher returns than employees from their social, technical and
clerical intelligence and their balanced skill sets, they have difficulties in
understanding their results pertaining to the remaining specific types of
ability, i.e. language and mathematical ability. To understand these
differences, research is probably needed that links the choice for
entrepreneurship versus employment with choices for specific professions.
Another remaining question is how the returns to some of the specific
areas of intelligence for some parts of their distributions can be actually
negative. Further research should indicate whether this is due to unobserved
characteristics that are correlated with specific abilities, or whether this
might relate to the fact that, for instance, a higher level of language ability
leads people to choosing professions with lower incomes.
A third remaining question the authors acknowledge is the relationship
between team entrepreneurship (or start-ups with personnel) on the one
hand, and the strength of the effect on entrepreneurs' incomes of being a JAT
on the other hand. One could expect that being a JAT is less valuable when an
enterprise is started up with a (multidisciplinary and/or complementary) team
of entrepreneurs. For instance, if many of the entrepreneurs have started up
their firms as part of a team, the measured effects of being a JAT might be
underestimates of the real effect of JAT on 'solo' entrepreneurship.
Conclusion and practical recommendations
The findings give a quite clear handle as to who should be stimulated to
become an entrepreneur and who should not. Especially, since entrepreneurs
render large returns to society in terms of growth, innovation and the creation
of labor, which often surpass their private returns, it is clear that attracting
the people with balanced skill sets, who are both technically and socially
sophisticated, would be beneficial for the economy (as well as for the
individuals themselves). Providers of equity, loans, subsidies, and licenses whether private or public authorities-, might use these insights to stimulate a
larger and more successful entrepreneurial economy.
37
References
Baron, R.A. and G.D. Markman (2003), “Beyond social capital: The role of
entrepreneurs' social competence in their financial success”. Journal of
Business Venturing 18(1), pp. 41-60.
Lazear, E.P. (2005), “Entrepreneurship”. Journal of Labor Economics, 23 (4),
pp. 649-680.
38
CHAPTER 6
Based on:
“Opportunity identification and pursuit:
Does an entrepreneur’s human capital matter?”
by
Deniz Ucbasaran
Nottingham University Business School
Paul Westhead
Warwick Business School
&
Mike Wright
Centre for Management Buy-Out Research
Nottingham University Business School
ABSTRACT
An entrepreneur’s ‘inputs’ relating to their general (i.e., education and work
experience) and entrepreneurship-specific human capital profile (i.e.,
business ownership experience, managerial capabilities, entrepreneurial
capabilities and technical capabilities) are presumed to be associated with the
entrepreneurial ‘outputs’ in the form of business opportunity identification and
pursuit. The authors derive hypotheses regarding the relationships between
these human capital inputs and outputs. They test the hypotheses on a
stratified random sample of 588 owners of private firms. With regard to the
number of business opportunities identified and pursued, entrepreneurshipspecific rather than general human capital variables ‘explained’ more of the
variance. Entrepreneurs reporting higher information search intensity
identified significantly more business opportunities, but they did not pursue
markedly more or less opportunities. The use of publications as a source of
information was positively associated with the probability of identifying more
opportunities, while information emanating from personal, professional and
business networks was not. Implications for practitioners and researchers are
discussed.
39
6.1 Introduction
Links between entrepreneurs’ human capital profiles and outcomes relating to
firm entry/exit and performance have been identified by many researchers.
Currently, opportunity oriented conceptualizations of entrepreneurship are
attracting attention but the human capital profiles of entrepreneurs who
identify and pursue more business opportunities are not well understood.
Shane (2003) argues that venture performance is determined by how
effectively the entrepreneur deals with entrepreneurial process activities
relating to opportunity identification, evaluation and exploitation.
By solely focusing on venture performance, the effects of the various
activities that lead to that performance may be confounded. There is,
therefore, a need to disentangle the activities involved in the entrepreneurial
process that can subsequently impact venture performance. Entrepreneurs
who identify more opportunities may select to pursue better quality
opportunities with greater wealth creating potential because they have more
opportunities to choose from. Studies that focus solely on traditional firm
performance outcomes (i.e., sales and employment growth and profitability)
do not provide an understanding of the entrepreneurial human capital drivers
that are linked to the identification of opportunities that create future venture
wealth.
Recently, several studies have empirically explored the links between
aspects of human capital and opportunity identification. These studies have
enhanced our understanding of opportunity identification and human capital
but they have focused on a narrow array of human capital variables. They
have also generally solely explored responses from students or nascent
entrepreneurs, rather than practicing entrepreneurs who in turn are
heterogeneous regarding their experience of the entrepreneurial process.
Further, a number of the studies have failed to operationalize the actual
extent of opportunity identification. Consequently, there is scant empirical
evidence relating to the links between practicing entrepreneurs’ human capital
profiles and their actual opportunity identification and pursuit activities. This
research gap is the focus of this study, which makes the following
contributions.
First, a broader conceptualization of an entrepreneur’s human capital is
discussed. Previous studies have focused mainly on education and various
sorts of experience. An entrepreneur’s human capital relating to perceived
capabilities (i.e., self-efficacy) is considered in this study in addition to the
widely used human capital indicators.
Second, a distinction is made between an entrepreneur’s general (i.e.,
education and work experience) and entrepreneurship-specific human capital
profile (i.e., managerial capabilities, entrepreneurial capabilities, technical
capabilities and business ownership experience).
Third, two outcomes from the entrepreneurial process are considered.
Previous studies have generally focused on a single outcome. The authors
appreciate that the human capital profiles leveraged to identify business
opportunities (i.e., the number of opportunities identified) may not be exactly
the same as those required to pursue business opportunities (i.e., the number
of identified opportunities pursued). Since, entrepreneurs may decide to
40
pursue only a subset of the opportunities they identify, this selection decision
implies that they are perceived to be of greater quality.
Fourth, Ucbasaran and coauthors explore whether entrepreneurs who
search for more information and utilize particular types of information identify
and pursue more business opportunities. Their evidence will provide
additional insights to the debate surrounding whether opportunities are
identified as a result of information search, or the alertness of the
entrepreneur.
6.2 Conceptual framework
Context
In this study, a broad conceptualization of general and specific human capital
is presented to include widely used dimensions of human capital as well as
cognitive dimensions.16
An entrepreneur’s education and work experience are regarded as
surrogate measures of general human capital. Variables relating to prior
business ownership experience and self-perceived capabilities are considered
as surrogate measures of entrepreneurship-specific human capital.
6.3 Derivation of hypotheses
Six hypotheses are derived with reference to the human capital perspective
discussed above.
Human capital and opportunity identification
If opportunities are indeed circulating in the environment waiting to be
discovered, individuals with superior human capital may have greater
cognitive ability to be alert to opportunities, knowledge of where to look for
an opportunity, and/or knowledge of what an opportunity ‘looks like’.
Alternatively, if opportunities are imagined or created, entrepreneurs with
greater levels of human capital may have more ‘ingredients’ to work with to
identify/create an opportunity. This suggests the following hypothesis:
H1a Entrepreneurs reporting higher levels of human capital (i.e., general and
entrepreneurship-specific human capital) will identify more business
opportunities, in a given time period.
Entrepreneurship-specific human capital can provide an entrepreneur with the
knowledge of where to look for opportunities, as well as the ability to identify
an opportunity that might be ignored by entrepreneurs relying solely on their
general human capital.
16
General human capital relates to skills and knowledge that are easily transferable across a
variety of economic settings. Conversely, specific human capital relates to skills and
knowledge that are less transferable and have a narrower scope of applicability.
41
H1b Entrepreneurship-specific human capital will ‘explain’ more of the
variance in the number of business opportunities identified, in a given time
period, than general human capital.
Information search intensity and opportunity identification
Identifying opportunities may be related to the information and knowledge
entrepreneurs possess. If information facilitates the identification of an
opportunity, individuals may choose to increase their access to opportunities
by searching for new/current information. For a given level of prior
information (i.e., specific human capital), entrepreneurs may be able to access
opportunities by searching for up-to-date information.
H1c Entrepreneurs reporting higher levels of information search intensity
will identify more business opportunities, in a given time period.
Human capital and opportunity pursuit
Opportunity pursuit relates to the evaluation stage following the identification
of an opportunity that is reported prior to potential exploitation. An
entrepreneur’s human capital may be leveraged to screen opportunities and
select which opportunities should be pursued. Human capital may increase the
expected value from exploiting an opportunity and decrease the uncertainty
surrounding the expected value, because skills and knowledge affect the
probability of effectively exploiting the opportunity. This discussion suggests
the following hypothesis:
H2a Entrepreneurs reporting higher levels of human capital (i.e., general and
entrepreneurship-specific) will pursue more identified business opportunities,
in a given time period.
The knowledge and skills associated with entrepreneurship-specific human
capital may be more relevant to pursuing an opportunity than those associated
with general human capital. The opportunity costs associated with pursuing an
opportunity are, therefore, likely to be lower for entrepreneurs with higher
levels of entrepreneurship-specific human capital than those with higher
levels of general human capital. This suggests the following hypothesis:
H2b Entrepreneurship-specific human capital will ‘explain’ more of the
variance in the number of business opportunities pursued, in a given time
period, than general human capital.
Information search intensity and opportunity pursuit
The identified business opportunity may only be pursued if the information
suggests the opportunity is viable and/or valuable, that is, of higher quality
than other identified opportunities. Some entrepreneurs may choose not to
search for additional information. If too much time is spent acquiring
information, the brief window of opportunity for pursuing a venture idea may
close. Extensive information search may prevent some entrepreneurs from
pursuing a business opportunity. This may be desirable if the identified
business opportunity has a high probability of failure.
42
H2c Entrepreneurs reporting higher levels of information search intensity
will pursue fewer identified business opportunities, in a given time period.
6.4 Data and methodology17
Sample
The sampling frame of private firms was constructed by obtaining sampling
quotas for four broad industrial categories (i.e., agriculture, forestry and
fishing, production, construction and services) and the eleven Government
Official Regions in Great Britain in 1999 (Office for National Statistics, 1999).
In total, 588 respondents provided complete data for the selected variables
explored. Each key respondent was a founder and/or principal owner.
Dependent variables
Consistent with previous literature opportunity identification was
operationalized in terms of the number of opportunities identified. The
outcomes of the respondents to the question how many opportunities for
creating or purchasing a business they have identified within the past five
years, were collapsed into three broad categories; 0, 1 to 2 and 2+.
The number of opportunities pursued was ascertained as well. This
dependent variable solely relates to the evaluation stage of the
entrepreneurial process. The outcomes were collapsed into three broad
outcomes similar to the number of opportunities identified.
Independent variables
Two sets of general human capital variables were collected with regard to
education and work experience. Education was measured as years of
education reported by the respondents and the height of the education level
was created by the use of five dummy variables. The second set of general
human capital variables includes work experience, for which the number of
full-time jobs and the level of achievement are used as a measure. The level
of managerial attainment was ascertained as well.
The authors measure entrepreneurship-specific human capital as the level
of business ownership experience. Three component scores for perceived
capabilities were identified (i.e., entrepreneurial, managerial and technical
capability). Furthermore, an information search intensity measure was
operationalized. The number of information sources used was selected as an
indicator of the level of the information search. Because this does not
distinguish between different types of information sought, four distinct types
of information sources were used: ‘professional network’, ‘publications’,
‘business network’ and ‘personal network’.
17
Ucbasaran et al. face and acknowledge the same problems as were encountered in Chapters 3
and 5 with respect to the endogeneity in the context of human capital investments and
entrepreneurial performance. Ucbasaran et al. control for the role of ability by including the
selected capability measures (i.e., managerial, entrepreneurial and technical ability). Moreover,
they use econometric techniques (IV) to deal with the endogenous nature of education as well
as business ownership experience.
43
6.5 Results
Opportunity identification
Entrepreneurs with higher levels of education, work experience, business
ownership experience, managerial capability and entrepreneurial capability
were significantly associated with an increased probability of identifying more
opportunities. Hypothesis H1a indicating that entrepreneurs reporting higher
levels of human capital would identify more opportunities was, therefore,
supported.
Ucbasaran et al also find support for hypothesis H1b. Entrepreneurship
specific human capital is more strongly related to the number of opportunities
identified than general human capital. A one unit increase in each of the
general human capital variables together, increased the probability of an
entrepreneur identifying one or more opportunities by 12.9%. However, a one
unit increase in each of the entrepreneurship specific human capital variables
together, increased the probability of an entrepreneur identifying one or more
opportunities by 26.7%. This evidence provides support for hypothesis H1b.
Hypothesis H1c was tested using multiple information search variables.
Entrepreneurs who reported higher levels of information search intensity and
those who used more information sources identified significantly more
business opportunities. These findings render support for hypothesis H1c.
However, on a stand-alone basis, only publication-based information search
was significantly associated with opportunity identification. Information search
based on personal, professional and business networks were not significantly
associated with increased opportunity identification. Apparently, certain types
of information are more useful for opportunity identification than others.
Opportunity pursuit
Human capital has a positive association with the pursuit of opportunities.
Hypothesis H2a is, therefore, supported. The authors also find support for
hypothesis H2b: general human capital has a weaker relationship with
opportunity pursuit than entrepreneurship specific human capital. More in
particular, none of the general human capital variables were significantly
associated with the dependent variable. Entrepreneurship specific human
capital variables relating to business ownership experience, managerial,
technical and entrepreneurial capabilities were significantly associated with
the probability of opportunity pursuit.
Hypothesis H2c was tested with regard to the three information search
variables: information search intensity, number of information sources used
and the types of information. However, the results do not support hypothesis
H2c.
6.6
Conclusions and implications
Shane and Venkataraman (2000) have asserted that “to have
entrepreneurship, you must first have entrepreneurial opportunities”.
Irrespective of whether these opportunities exist in the environment or
emerge as a creative act, individuals are needed to identify and exploit them.
While opportunity identification is seen as a behavior distinguishing
44
entrepreneurs from non-entrepreneurs, we know little about why some
entrepreneurs identify and pursue more opportunities than other
entrepreneurs. In this study, this research gap has been explored with
reference to human capital theory.
The findings provide insights to assist the calls made by policy-makers
and practitioners to consider the entrepreneur rather than the firm as the unit
of analysis through all stages of the entrepreneurial process. Presented
evidence suggests that entrepreneurs leverage their entrepreneurship specific
human capital to a greater extent than their general human capital to identify
and pursue opportunities.
Policy to address attitudinal, resource and operational barriers to the
creation of new ventures may need to provide assistance that enhances the
entrepreneurship specific human capital of individuals. For example,
practitioners seeking to increase the pool of entrepreneurs, particularly in
areas with currently low levels of new firm formation, may consider
introducing initiatives that encourage nascent and practicing entrepreneurs to
hone their managerial capabilities. Rather than providing ‘blanket support’
encouraging individuals to consider new business creation, more targeted
assistance may be needed to address barriers to the pursuit of opportunities
identified. Initiatives that hone the technical and entrepreneurial capabilities of
entrepreneurs may increase the proportion of identified opportunities
converted into pursued opportunities. The pursuit of more identified
opportunities may lead to the creation of more wealth and jobs.
References
Office for National Statistics (1999), PA 1003 Commerce, Energy and
Industry: Size Analysis of the United Kingdom Businesses. London: Office for
National Statistics.
Shane, S. (2003), “A General Theory of Entrepreneurship”. Northampton, MA:
Edward Elgar Publishing.
Shane, S., and S. Venkataraman (2000), “The Promise of Entrepreneurship as
a Field of Research”. Academy of Management Review, 25, pp. 217-226.
45
Part IV
Work Experience
46
CHAPTER 7
Based on:
“Inferring the unobserved human capital of entrepreneurs”
by
Arnab Bhattacharjee
University of St Andrews
Jean Bonnet
Université de Caen
Nicolas Le Pape
Université de Caen
&
Régis Renault
Université de Cergy-Pontoise
ABSTRACT
The goal of the paper by Bhattacharjee et al. is to study the role of labour
market characteristics in entrepreneurial choice and its impact on the survival
of newly created firms. Their starting point is that, when starting a new
business, an entrepreneur’s labor market situation (e.g. employed or not –and
whether or not in the preferred sector-) reflects how his human capital may
be valuated through salaried labor, the alternative for entrepreneurship.
Hence, Bhattacharjee and coauthors view the labor market situation preceding
entrepreneurship as an indicator of the amount of unobserved human capital
people have. More in particular, when an individual sets up a firm in the same
sector of experience, this may indicate a high average level of unobserved
human capital due to a lack of depreciation motive. The authors test the
hypothesis that an individual’s previous labor market situation affects the
predictive value of observed human capital for surviving as an entrepreneur.
The results show that the impact of education on the new firm’s survival is
indeed most pronounced for firms created by individuals salaried in their
preferred branch of activity while it is rather limited for other entrepreneurs.
47
7.1 Introduction
An individual’s choice for entrepreneurship implies that the entrepreneur
anticipates better returns on his human capital by running his own firm than
by selling his human capital in the labor market. Bhattacharjee et al. argue in
their paper that entrepreneurship and successful entrepreneurship is to a
large extent the result of inefficiencies in the labor market. They consider
two sorts of inefficiencies: first, actual human capital is usually imperfectly
rewarded by the labor market because of information asymmetries or
incentive problems and second, frictions in the labor market may prevent
individuals from allocating their human capital optimally, either because they
stay unemployed or they stay in a position with which they are poorly
matched.
Bhattacharjee et al. argue that even when human capital is perfectly
observable and there are no incentive problems, an individual may be
prevented from getting a job with which he has a good match due to various
labor market rigidities. Actual human capital may therefore be undervalued,
especially in the case of potential entrepreneurs who may have some unusual
and novel management, commercial or technological skills. Even if information
is perfect, human capital may still be rewarded differently in the labor market
than in entrepreneurship.
Market rigidities can be seen from both a static and dynamic point of view.
In a static sense, labor market rigidities give rise to lower wage earnings
given the observed level of human capital, since workers are unemployed or
in a badly matched position where their productivity is low. Hence such
rigidities give incentives to choose for entrepreneurship regardless of
potential asymmetries on actual human capital. There is also a dynamic impact
resulting from the potential depreciation in human capital for those who are
unemployed or working in a branch of activity that does not suit their skills.
Entrepreneurship may then be a way to keep working in the preferred sector,
thus preventing such depreciation.
7.2.
Data
Sample
The data are extracted from the SINE 94-survey conducted by the French
National Institute of Statistical and Economic Studies. A second survey carried
out in 1997 (SINE 97) gives information about the situation of the same firms
(closed down or still running four years later; when closed down, the date of
the discontinuation). The surveyed units belong to the private productive
sector in the field of industry, building, commerce and services. Only firms
are considered that were set up by an individual. Takeovers are excluded. In
order to ensure some homogeneity in labor supply behavior, the sample is
further narrowed down to French male middle aged (aged 30-50)
entrepreneurs who started a business in metropolitan France.
The database SINE 94 provides information about whether the individual
was employed or not. For unemployed individuals it indicates whether the
unemployment spell is short (less than one year) or long (beyond one year).
For those who were employed, the data provide information about the
48
entrepreneur’s experience in the branch of activity or the new business or in
some other branch. This information allows distinguishing four different subgroups: employed in the same branch, employed in a different branch
(mismatched), unemployed for less than one year, unemployed for more than
one year. For each of these sub-groups, Bhattacharjee et al. compare the
survival rates of newly created firms for two extreme populations of
entrepreneurs: those holding a degree obtained after two years of higher
education (whom they label as having a high education level) and those who
hold no degree at all (labeled as having a low education level). A sample size
of 6041 entrepreneurs is obtained.
Descriptive statistics
The statistics show that the survival rates for mismatched or unemployed
people are lower than those of people who were previously working in their
preferred branch of activity (48,33%, 54,87% and 67,63%, respectively). It is
not so much the difference between employed and unemployed individuals
that matters. The statistics rather show that having been employed in the
right branch of activity provides a significant advantage in terms of the
survival of the newly created firm.
Moreover, the returns to within-sector experience is significantly higher
for employed or recently unemployed individuals than for long term
unemployed individuals. Survival rates with and without within-sector
experience are 67.63% versus 48.33% for employed, 62.12% versus 41.92%
for short term unemployed and 54.23% versus 43.31% for long-term
unemployed. The data reveal furthermore that the positive impact of a higher
education on survival is very strong for entrepreneurs who were initially well
matched whereas it is rather limited for the other two groups.
The interesting new insight is that the extent of the positive impact of
human capital on survival strongly depends on previous the labor market
situation of the entrepreneur. Bhattacharjee et al. argue that these differences
may be explained by viewing entrepreneurship as a response to the
aforementioned labor market inefficiencies.
7.3.
Simple model
The model controls for two motives to choose for entrepreneurship:
circumventing the undervaluation of human capital and avoiding human capital
depreciation. The model considers an individual whose actual human capital is
either high or low. This human capital however may not be perfectly observed
by employers. The model takes this into account, both in a static and a
dynamic framework. Moreover, as explained above, the decision to become an
entrepreneur can be taken from three different situations: (1) unemployment,
(2) a salaried position in a sector where he is highly productive, or (3) a
salaried position in a sector where his productivity is poor. Though the
second situation is clearly preferable to the other two, the agent may be
unable to reach it because of frictions in the labor market.
The probability associated with creating a new business for an individual is
a function of actual human capital and the particular situation the prospective
49
entrepreneur is in. An individual will start a new business if the value of
creating a new business exceeds the value of staying in the current state.
A first important result of the theoretical model by Bhattacharjee et al. is
that the benefits from having a high level of human capital are larger for
entrepreneurs than for employees. This is consistent with the result of
Chapter 3. Their second result is that, if employers observe human capital
imperfectly, an entrepreneur who was well matched in his job when he started
a business should be expected to have a higher level of human capital than an
entrepreneur who was unemployed or stuck in a job where his productivity
was low. This is because one can infer that this individual has on average
some unobserved human capital and no depreciation motive to set up his firm.
Third, the incentives to start a business of an individual with a low level of
human capital are independent of his initial state since the labor market
rewards his human capital equally poorly in all situations. The fourth result
from the theoretical model is that, given the level of observed human capital,
the incentives to start a business are lower for individuals employed with a
good match than in the other two states, irrespective of the actual level of
human capital. This is because earnings in the labor market are independent
of actual human capital since they are fully based on observed human capital.
Fifth, in a world of extreme information asymmetry and given the level of
observed human capital, entrepreneurs who started out with a good match in
the labor market are expected to have a higher level of actual human capital
than those who were unemployed or badly matched.
7.4.
Empirical analysis
The theoretical results suggest that there may indeed be significant
differences in the relationship between observed and unobserved human
capital according to the initial situation of the entrepreneur in the labor
market.
In order to identify differences in the impact of observed human capital on
survival across initial situations of the entrepreneur, the authors run their
regressions explaining survival chances based on four different sub-samples:
(i) individuals employed in the same branch of activity; (ii) individuals
employed in different branches of activity; (iii) individuals unemployed for
less than one year; (iv) individuals employed for more than one year.
Results on the impact of education are consistent with the descriptive
statistics of Section 7.2. More education reduces the chance of failure for
individuals employed in the same sector or unemployed for more than one
year significantly. It has no significant impact or may even increase the
chance of failure for individuals employed in a different branch or unemployed
for less than one year: in particular for individuals employed in a different
branch, those with a high education level have a significantly higher chance of
failure. Education being significant for long term unemployed individuals may
be understood as reflecting a lack of depreciation concern for those who are
highly educated: this is because their human capital has already depreciated.
More generally, after such a long unemployment spell, the situation of the
individual on the labor market no longer depends much on education.
50
On the contrary, the results on the impact of experience are different from
those on education since a longer experience always significantly reduces the
chance of failure.18
7.5.
Conclusion
Bhattacharjee et al, have explored the relationship between an individual’s
labor market position at the time a venture is started and the effect of human
capital on start-up and/or survival chances. The basic idea is that due to labor
market rigidities some people find jobs in the (employee) labor market that
better match their human capital than others. For those who find a
‘mismatched’ job, or who are even unemployed and who become
entrepreneurs, the actual level of human capital is likely to be higher than the
level of observed human capital even if the difference in actual human capital
is higher for people well matched (setting up a firm is always good news). The
difference between the actual and observed levels of human capital is what
Bhattacharjee et al. refer to as the level of unobserved human capital. The
higher the fraction of one’s actual level of human capital remains unobserved,
the less attractive are the prospects of employment, and the higher is the
likelihood of entrepreneurship. Hence, the more labor market rigidities there
are, the more people who are poorly matched or unemployed start as
entrepreneurs. Then, the more labor market rigidities there are, the lower on
average is the unobserved human capital of entrepreneurs. This is the case,
since for individuals with a high level of human capital, there is a deprecation
motive. The weak propensity to set up a firm in France (in comparison with
entrepreneurial economies like USA or UK) is more due to the lack of
entrepreneurs who were previously well matched in a salaried position (pull
effect). This phenomenon can be explained by the difficulties to find again a
comparable new job if the new firm fails (cultural aspect, level of
unemployment and of course rigidities of the labor market).
The authors’ preliminary empirical results partly support the notion that
the effect of human capital on entrepreneurship (survival) depends on the
initial labor market situation: for education they find such a differentiated
effect, but not for experience.
More research into the effect of labor market institutions on the relative
returns to human capital for entrepreneurs and employees might enrich our
understanding of the relevance of labor market institutions to increase the
pool of (successful) entrepreneurs in societies that tend to become more and
more knowledge intensive.
18
However, these results are somewhat difficult to interpret since the authors only have
experience data for those individuals who have been working in the same branch of activity
(which explains why there are no results for entrepreneurs who changed branch when they
started a business).
51
CHAPTER 8
Based on:
“The role of initial and acquired human capital in the long- term survival and
performance of immigrant entrepreneurs”
by
Miri Lerner
Academic College of Tel-Aviv-Jaffa and Tel-Aviv University, Israel
&
Susanna Khavul
London Business School, UK
ABSTRACT
This paper examines how founder human capital compares to social capital
and institutional capital in improving the survival of immigrant owned
businesses. Hypotheses about these relationships are tested on a longitudinal
sample of immigrant business owners who received loans from a small
business assistance program administered by the government of Israel. The
study provides additional support for the notion that human capital plays a key
role in the survival of immigrant business owners, more than the role of the
institutional or social capital. The study also expands the literature on the
effect of human capital on performance by showing that acquired human
capital in the host country has a higher impact on performance than initial
human capital acquired in the country of origin.
52
8.1 Introduction
Immigrant small business has been viewed as an alternative and more viable
route to upward economic mobility for immigrant groups. The vulnerability of
immigrant owned businesses, especially during their early years, has attracted
some empirical and theoretical attention. Failure rates of immigrants have
been ascribed to limited tangible and intangible resources of the owners
including their limited human and social capital in the new cultural and
economic environment. The literature has long held that immigrant
entrepreneurs start their businesses out of necessity for employment rather
than from recognition of an opportunity.
The experience of immigrant entrepreneurs in Israel during 1990s
presents an interesting opportunity to compare the relative importance of
founder initial and acquired human capital versus institutional capital in terms
of their impact on firm survival. In the 1990s, Israel’s population grew by 20%
as a result of large scale immigration primarily from the former Soviet Union.
Between 1989 and 2005 more than 1.2 million immigrants arrived to Israel,
81% of them (about 962,000 immigrants) from the former Soviet Union, mainly
during the 1990s.19
Numerous governmental and non-govermental organizations try to fill the
void by providing a myriad of financial, advisory, and networking support
before and after immigrant businesses are established. While previous studies
have looked at the efficiency and effectiveness of government institutional
programs, relatively few studies have compared the relative importance of the
entrepreneur’s initial and acquired human capital with the institutional capital
available. Due in part to the lack of data, previous research has rarely
compared immigrant businesses that have survived with those that have
failed.
This paper uses a unique data set of firms that received small business
assistance loans from a government fund designed to support immigrants in
their efforts to establish or grow their own new business. The survey was
conducted approximately 5 years after the loans were received. While all the
firms in the sample had received loans, the firms varied in their use as well as
their access to a diverse set of other institutional support networks.
8.2 Background and hypotheses
The study of Lerner and Khavul is based on the typology of resources that an
entrepreneurial start-up firm needs in order to survive the liabilities of
newness. The dominant aspect of this resource-based view distinguishes
between physical, human, social, organizational, and financial resource
bundles. The focus of their present study is on comparing the importance of
human and social capital to that of governmental institutional capital in
improving the survival of immigrant businesses.
19
Among the Jewish immigrants from the former Soviet Union to Israel, around one third of
the migrants reported a scientific and academic occupation in the country of origin, including
professions in medicine, engineering, law, the natural sciences, the social sciences and the
humanities.
53
Founder human capital
The authors make a distinction between: (i) general human capital
(represented by the entrepreneur’s education, gender and race); (ii)
management know-how; (iii) industry specific know how and; (iv) financial
capital. Based on the human capital perspective it is reasonable to expect –
and supported empirically in many studies- that those immigrants with higher
levels of human capital will own the better-performing businesses among the
immigrants.
Since time is required to accumulate experience, skills and material
resources, it can be expected that those immigrants who have longer
residence period in Israel and more years of experience in its labor market
will own the better performing businesses among the immigrants.20
Hypothesis 1: Businesses owned by immigrants with higher levels of human
capital will be more likely to survive than similar firms whose owners have
lower human capital. The probability of survival should be higher for
businesses started by immigrants with higher levels of human capital acquired
in the host country.
.
Motivation for business ownership
While the balance of opportunity and necessity entrepreneurship varies from
country to country, Israeli data shows that there are approximately four times
more opportunity entrepreneurs in Israel than there are necessity
entrepreneurs. Opportunity entrepreneurs have higher expectations of growth
in terms of rates of employment and foreign exports and are generally viewed
as adding higher value to the economy. Therefore, hypothesis 2 is stated as
follows:
Hypothesis 2: Immigrants who start businesses as a response to a market
opportunity will be more likely to have their businesses survive than
immigrants who start their businesses out of necessity.
Social capital
For immigrant entrepreneurs suffering from a shortage of relationships to
mainstream entrepreneurs and clients, the importance of social relationships
within the immigrant community increases in importance. This brings Lerner
and Khavul to their third hypothesis:
Hypothesis 3: The social capital acquired through business ownership will
improve the chances of firm survival.
Institutional capital
Public policy and initiatives is another major topic that has an impact on
immigrant entrepreneurs directly and indirectly. Governments have major
impact on the structure of opportunities of immigrants in many areas of their
20
Given that businesses owned by immigrants are often family businesses, the human capital
of the spouse is another facet of immigrants businesses. The family facilitates the pooling of
labor and financial resources.
54
lives: residence, occupation, and economic well-being. Access to ownership
is also affected by governmental policies on the terms of country entrance as
well as the policies affecting the ease with which businesses can be started.
A study of immigrants from the former Soviet Union in Israel, Menahem
and Lerner (2001) revealed the importance of governmental support for the
incorporation of immigrants into society. Their findings indicate that when
dealing with large waves of immigrants with relatively high human capital in
the original society, the difference between the more and less successful
immigrants is the adaptability of human capital to the host society. Retraining
programs may contribute to such adaptability and thus enhance human capital
transferability. This forms the basis for hypothesis 4:
Hypothesis 4: Immigrant businesses whose owners received a diversity of
institutional support will be more likely to survive than businesses whose
owners received less support.
8.3
Methodology
Data
The research population consists of a cohort of immigrants from the former
Soviet Union to Israel who received business assistance loans from a
government fund five years ago. Tracking a cohort of immigrant business
entrepreneurs over a five year period enables the long-term impact of
institutional capital on the performance of immigrant businesses to be
assessed. It also allows the immigrant businesses that have survived to be
compared to those that have failed. From the total population of 892
immigrants who received loans, 200 were randomly selected for participation
in the study. The firms under study represent 19% of the population and have
cooperated at the rate of 85%. Of the respondents, 72% of the businesses
were still in operation five years after receiving the loan.
Variables
Business survival is the dependent variables in the study. Concerning the
independent variables, human capital included traditional variables such as
business founder’s age and experience, as well as variables specific to
immigrants such as length of residence in Israel and length of experience as a
salaried employee in Israel. The number of financial and advisory support
services the business owners received before the loan was approved
measured the diversity of institutional capital. The acquired social capital
variable measured the contribution of business ownership to absorption in
Israel. Motivation was examined according the main motive for establishing
the business – economic necessity or opportunity identification.
55
8.4
Findings
Human capital
The first hypothesis is supported. Results of the logistic regression indicate
that human capital variables play an important role in predicting the odds of
survival for immigrant businesses. The length of residence in Israel and length
of experience as a salaried employee in Israel both positively and significantly
affect the odds of business survival. Moreover, the occupational status of the
founders after immigration was significantly higher for owners of operating
businesses than for owners of closed businesses. Owners of operating
businesses are also shown to be significantly less likely to report a lack of
experience in the industry of the business they have founded. The results also
show that a working spouse outside the started business is associated
negatively with the likelihood of the business to survive.
Motivation for business ownership
The second hypothesis is not supported: Business survival rates are not
significantly higher for entrepreneurs who started in response to an
opportunity than those who started as a result of necessity.
Social capital
The third hypothesis is supported. The more social capital the entrepreneur
has in terms of the contribution of the business to absorption in Israel, the
higher is the likelihood of the firm to have survived its first five years.
Institutional capital
There was no significant difference between operating and closed businesses
in the number of different financial or advisory support services received
prior to the loan; this implies that the fourth hypothesis is rejected.
8.5
Conclusions
The results showed that human capital in general and, especially the human
capital acquired in the host country, significantly improve the odds of survival
for immigrant owned business. Immigrants accumulated this human capital
through the length of their residence in Israel. The length of residence in
Israel and length of experience as a salaried employee in Israel both
positively and significantly affect the odds of business survival. Moreover,
immigrants with a higher occupational status in Israel are more likely to
survive. In contrast, the occupational status of the immigrant prior to
immigrating to Israel does not lessen nor increase the odds of survival. In
short, acquired rather than initial (foreign) human capital is critical to the
survival of immigrant businesses.
The second implication of this study is that the social capital immigrants
acquire through business ownership significantly improves the odds that their
businesses will survive; their absorption into Israeli society was also
facilitated. Contrary to expectation, no statistically significant effect was
found for either financial or advisory support prior to business founding on the
likelihood of firm survival.
56
References
Menahem, G. and M. Lerner (2001), “An Evaluation of the Effect of Public
Support in Enhancing Occupational Incorporation of Former Soviet Union
Immigrants to Israel: A Longitudinal Study”. Journal of Social Policy, 2, pp.
307-331.
57
Part V
Scientific Entrepreneurship
58
CHAPTER 9
Based on:
“Scientist entrepreneurship: Human capital or social capital?”
by
David B. Audretsch
Taylor Aldridge
&
Alex Oettl
Max Planck Institute of Economics
ABSTRACT
The study of Audretsch et al. examines the prevalence and determinants of
the commercialization of research by university scientists funded by grants
from the National Cancer Institute (NCI). Because the two publicly available
modes of scientist commercialization – patents and Small Business Innovation
Research (SBIR) grants – do not cover the full spectrum of commercializing
activities undertaken by university scientists, the study also includes two
additional measures obtained from a survey – licensing of intellectual property
and starting a new firm. These measures are used to assess the prevalence
and impact of human capital and social capital on the propensity for scientists
to commercialize their research. In particular, the empirical findings suggest
that scientists receiving funding from the National Cancer Institute exhibit a
robust propensity to commercialize their research. However, the prevalence
of commercialization depends highly upon the actual mode of
commercialization. Some modes of commercialization, such as patents, are
more prevalent, while other modes, such as funding by the SBIR program are
rarely used. Scientist entrepreneurship may be the sleeping giant of university
research. More than one in four patenting NCI scientists reveals that they
have started a new firm. This suggests a considerably more vigorous and
pervasive mode of commercialization than has been reflected in studies
restricted to the commercialization efforts of Technology Transfer Offices
(TTOs). The results suggest that both human capital and social capital
enhances the propensity for scientists to commercialize their research.
59
9.1 Introduction
In the new era of globalization, the US’ unrivaled investment in research and
knowledge to generate economic growth, employment and competitiveness
has been looked upon by both scholars and policy makers in internationally
linked markets for continued prosperity. However, this investment need not
be a guarantee for economic growth. Rather, knowledge investment should
penetrate what Audretsch et al. refer to as “the knowledge filter”. It is this
knowledge filter that stands between investments in research and its
commercialization through innovation, leading ultimately to economic growth.
The
mechanism
or
instrument
attributed
to
facilitating
the
commercialization of university scientist research has been the university
Technology Transfer Office (TTO).21 By most accounts, the impact of
facilitating this commercialization has been impressive. However, there are
compelling reasons to suspect that measuring and analyzing the
commercialization of university research by relying solely upon the
intellectual property disclosed to and registered by the TTOs may lead to a
systematic underestimation of commercialization and innovation emanating
from university research. The mandate of the TTO is not to measure and
document all of the intellectual property created by university research along
with the subsequent commercialization. Rather, what is measured and
documented is the intellectual property and commercialization activities with
which the TTO is involved. This involvement is typically a subset of the
broader and more pervasive intellectual property being generated by
university research and its commercialization.
If the spillover of knowledge generated by university research is viewed as
essential for economic growth, employment creation, and international
competitiveness in global markets, the systematic underreporting of
university spillovers resulting from the commercialization of scientist
research concomitantly may lead to severe policy distortions. Thus, rather
than relying on commercialization reported by the TTO to measure and
analyze the commercialization of university research, this study instead
develops alternative measures based on the commercialization activities
reported by scientists. In particular, the purpose of this study is to provide a
measure of scientist commercialization of university research and identify
which factors are conducive to scientist commercialization and which factors
inhibit scientist commercialization. To this end, the authors develop a new
database measuring the propensity of scientists funded by grants from the
National Cancer Institute (NCI) to commercialize their research as well as the
mode of commercialization.
21
While the TTO was not an invention of the Bayh-Dole Act, its prevalence exploded
following passage of the Act in 1980. Not only does the TTO typically engage in painstaking
collection of the intellectual property disclosed by scientists to the university but also the
extent of commercialization emanating from the TTO.
60
9.2.
Measurement
The commercialization activity of university scientists was measured by
starting with those scientists awarded a research grant by the National
Cancer Institute (NCI) between 1998 and 2002. Of those research grant
awards, the largest twenty percent, were taken to form the database used in
this study. The NCI awarded a total of $5,350,977,742 to these 1,693
awardees from 1998 to 2002.
Since the focus of this paper is on the propensity for scientists to
commercialize their research, commercialization must be operationalized and
measured. Based on the literature identified in the previous section, five main
measures of scientist commercialization are used, which reflect five different
modes by which scientists can and do commercialize their research. These
are (1) patenting inventions, (2) issuing licenses, (3) receiving an SBIR grant
to obtain funding for an innovative small business, (4) starting a new firm, and
(5) selling a patent.
Based on these five different measures reflecting distinct modes of
scientist commercialization of research, an NCI awardee database was created
to answer the question, “Why do some scientists commercialize while others
do not?”
Patents
The first measure of commercialization of research by an NCI award scientist
is inventions which are patented. Patent data werer obtained from the United
States Patent and Trademark Office (USPTO).
Survey implementation
A survey instrument was designed probing four subgroups of issues:
licensing, entrepreneurship, social capital and the role of the TTO.22 The 398
patenting scientists were “Googled” to obtain their e-mail and telephone
information. Of those 398 scientists identified in the database, 140 valid
responses were obtained.
Small business innovation research (SBIR)
One of the most striking insights to emerge in this study is that use of the
SBIR is not a prevalent or even common mode of commercialization by
scientists receiving NCI awards.
9.3.
Determinants of scientist commercialization
In addition to characteristics, which have already been probed in a number of
studies into the determinants of commercialization activities, such as age,
22
The question in the licensing section asked if the scientist has licensed. The question
contained in the entrepreneurship section identified whether the scientist started a new firm.
The questions concerning social capital asked the scientist if she sat on any industry science
advisory boards (SAB) or board of directors, the extent to which the NCI grant award
facilitated commercialization, along with other sources of major funding received from a
governmental agency.
61
gender, experience, reputation, region and university context, Audretsch et al.
also include a number of additional factors into their analysis. These include
not just scientific human capital, but social capital as well, along with the role
of the TTO, and the commercialization route selected by the scientist.
Scientific human capital
Audretsch et al. hypothesize that measures of the quality of the scientist, or
her scientific reputation as measured by citation, be linked to
commercialization. This is because the commercialization of scientific
research is particularly risky and uncertain and hence a strong scientific
reputation provides a greatly valued signal of scientific credibility and
capability to any anticipated commercialized venture or project. Scientific
human capital is empirically measured by the authors as scientist citations and
prior patents. The latter variable is included to control for previous
experience with commercialization activities. A positive coefficient would
suggest that previous commercialization experiences elevates the propensity
of a scientist to engage in commercialization activity.
Social capital
According to the literature, entrepreneurial activity should be enhanced where
investments in social capital are greater. Thus, the social capital of a scientist
is posited to be conducive to the commercialization of research.
Social capital is operationalized through the following measures: (i) Copatents – This variable reflects the extent of linkages between scientists by
measuring the number of patents where two NCI scientists shared a patent;
(ii) Board – This is a binary variable taking on the value of one if the scientist
has sat on a scientific advisory board or the board of directors of a firm; (iii)
Industry co-publications – This variable reflects linkages between university
scientists and their counterparts in industry and is measured as co-authorship
between a university scientist and an industry scientist; (iv) Industry Copublication Asia -This variable reflects linkages between university scientists
and their counterparts located in Asia.
TTO
The TTO has a mandate to facilitate and promote the commercialization of
university science. The TTO is characterized empirically by means of the
following measures: (i) TTO Helpful – This is a binary variable taking on the
value of one for scientists who responded to the survey that their TTO
directly helped them commercialize their research and zero otherwise; (ii)
TTO Age – This variable reflects the year in which the TTO was founded at
the particular university; (iii) TTO Employees – This variable measures the
mean number of employees per year responsible for license and patent
acquisitions. A positive relationship would suggest that a greater commitment
of TTO employee resources yields a higher propensity for scientists to
commercialize their research; (iv) TTO Licensing Commitment – Dividing the
number of employees dedicated to licensing technology by the number of
administrative employees reflects the commitment of the TTO to licensing
relative to other TTO functions. A positive relationship would suggest that
allocating a greater share of TTO employees to licensing would increase
scientist commercialization; (v) TTO Efficiency – The mean number of patents
62
applied for is divided by the number of issued patents. A positive coefficient
would reflect that a higher yield of patent applications resulting in patents
granted lead to greater scientist commercialization.
Scientist commercialization route
In commercializing their research, scientists can choose two routes: the TTO
route, assigning their patents to the university’s TTO, or, alternatively, they
can choose what Audretsch et al. term the entrepreneurial route of
commercialization. The entrepreneurial route to scientist commercialization
refers to those scientists who do not assign all of their patents to the
university’s TTO. Of the NCI patenting scientists, 70 percent assigned all of
their patents to their university TTO and 30 percent chose the entrepreneurial
route to commercialize their research.
Resources
It is argued that the propensity for a scientist to engage in commercialization
activity is positively related to the amount of the award, on the grounds that a
greater award amount, ceteris paribus, represents a greater investment in
new knowledge. Measures included are: (i) NCI Grant – This variable is the
mean total NCI grant awarded to a scientist between 1998 and 2002; (ii)
Government Funding – This binary variable takes on the value of one for
scientists responding to the scientist survey that they received additional
funding in excess of $750,000 from government sources and zero otherwise.
Scientist lifecycle
Scientist life-cycle models suggest that early in their careers scientists invest
heavily in human capital in order to build a scientific reputation. Scientist Age
was obtained from the scientist survey. It is expected that more mature
scientists have a higher propensity to engage in commercialization activities.
Location and institutional contexts
Scientist location can influence the decision to commercialize for two reasons.
First, knowledge tends to spill over within geographically bounded regions,
which implies that scientists working in regions with a high level of
investments in new knowledge can more easily access and generate new
scientific ideas. This suggests that scientists working in knowledge clusters
should tend to be more productive than their counterparts who are
geographically isolated.
A second component of externalities involves behavioral knowledge as
opposed to technical knowledge. The commercialization behavior and
attitudes exhibited by the chair, supervisors and peers in the relevant
department is likely to affect the commercialization behavior of a scientist.
Locations distinguished in the research by Audretsch et al. are the North
East, California and the Great Lakes. Those regions, which tend to have
greater investments in research and science, and also have developed a culture
more encouraging of university and scientist commercialization, such as
California and the North East, might be expected to have a positive coefficient.
Institutional context is measured by means of three additional variables: (i)
NCI Center – This is a binary variable taking on the value of one if the scientist
is employed at one of the 39 nationally-recognized comprehensive cancer
63
centers, and zero otherwise; (ii) Ivy League – A binary variable taking on the
value of one for all scientists employed at Brown University, Cornell
University, Columbia University, Dartmouth College, Harvard University,
Princeton University, the University of Pennsylvania and Yale University, and
zero otherwise; (iii) Public Institution – A binary variable taking on the value of
one for scientists employed at public universities and zero otherwise. Because
they are at least partially financed by the public, state universities tend to have
a stronger mandate for commercialization of research. This may suggest a
positive coefficient.
9.4.
Results
To shed light on the question; “Why do some scientists commercialize their
scientific research while others do not?” a probit model was estimated where
the dependent variable takes on the value of one if a NCI scientist has
commercialized over the time period 1998-2004 and zero if she has not. As
the previous section emphasized, there is no singular mode for scientist
commercialization. Rather, scientists select across multiple modes of possible
commercialization. Thus, the (probit) model was estimated for each of the
main modes of commercialization.
Start-up
The results estimating the likelihood of an NCI scientist starting a firm provide
consistent and compelling evidence that social capital promotes scientist
entrepreneurship, while having a helpful TTO and assigning the patent to the
TTO are associated with a lower propensity for scientists to become
entrepreneurs.
Licensing
The results reported from estimating scientist-licensing reveal several striking
similarities but also differences
from those
estimating
scientist
entrepreneurship. First, the impact of social capital is positive for both
entrepreneurship and licensing. Co-patenting with other academic scientists as
well as sitting on a scientific advisory board or board of directors of a private
company increases the likelihood of a scientist both starting a business and
licensing her intellectual property. However, co-publishing with scientists in
industry spurs scientist entrepreneurship, while it has no impact on licensing
behaviour.
Second, the particular commercialization route that is chosen by the
scientist influences the mode of commercialization. Those scientists choosing
the TTO commercialization route exhibit a higher likelihood of licensing but a
lower propensity to start a new firm. By contrast, scientists choosing the
entrepreneurship route to commercialize their research have a greater
propensity to start new firms rather than license their intellectual property.
Entrepreneurship in the form of a new firm start-up apparently serves as a
substitute for licensing when scientist commercialization is through the
entrepreneurial route and not the TTO route.
64
Patenting
The authors find evidence that measures of social capital do not only increase
the scientist propensity to license and become an entrepreneur, but also the
propensity to patent. Furthermore, the location of the co-author apparently
influences the propensity to patent. If the co-author is located in Asia, the
likelihood of a U.S. based scientist patenting in the U.S. is greater. The other
consistent result involves TTO Efficiency. Those scientists working at
universities with a more efficient TTO exhibit a higher propensity to patent.
There is also at least some evidence suggesting that older and more
established TTOs and larger TTOs, as measured by employment, tend to be
associated with a higher scientist propensity to patent.
Hence, overall, there are at least some indications suggesting that social capital
promotes all modes of commercialization, but perhaps entrepreneurship the
strongest. By contrast, the TTO seems to be most effective in promoting first
and foremost patents and then licensing, but much less start-ups.
9.5.
Conclusion
A consequence of globalization in the most developed countries, such as the
United States, has been to shift the comparative advantage away from
traditional manufacturing industries and towards new knowledge-based
economic activity. But where is this knowledge to come from? Research
undertaken by universities is sure to play a prominent role. The massive
investment in university research can impact economic growth only if
knowledge can be transformed into actual innovations and new and better
products through the commercialization process.
Thus, a large literature has emerged trying to gauge and analyze the
extent to which university research spills over into commercial activity. Much,
if not most, of this previous research has been restricted to focusing on the
activities emanating from Technology Transfer Offices.
This study has taken a different approach. Rather than focus on what the
TTOs do, it instead focuses on what university scientists do. The results are
revealing. In particular, while all modes of commercialization are important,
scientist entrepreneurship emerges as an important and prevalent mode of
commercialization of university research. More than one in four patenting NCI
scientists has started a new firm. This is a remarkably high rate of
entrepreneurship for any group of people, let alone university scientists.
Scientist entrepreneurship may prove to be the sleeping giant of university
commercialization.
How are scientists able to start a business without TTO support? There is
at least some evidence indicating that social capital can serve as a mechanism
to compensate for lack of TTO help when starting a new firm. This would
suggest that university governance and public policy facilitating participation in
scientific networks may be a valuable investment accruing positive returns in
terms of knowledge spillovers and technology transfer, ultimately leading to
commercialization, innovation and economic growth.
65
Whatever the answers are to the many remaining questions and other
crucial questions future research can uncover, the sleeping giant of scientist
entrepreneurship may prove to be one giant that is worth waking up.
66
Part VI
Guidelines for Policy
67
CHAPTER 10
Based on:
“Implementing the Community Lisbon Programme:
Fostering entrepreneurial mindsets through education and learning”
by
Simone Baldassari
Commission of the European Communities
68
10.1
Introduction
In February 2005, the Commission proposed a new start for the Lisbon
Strategy focusing the European Union’s efforts on two principal tasks –
delivering stronger, lasting growth and providing more and better jobs. The
new Partnership for Growth and Jobs stresses the importance of promoting a
more entrepreneurial culture and of creating a supportive environment for
SMEs.
Research suggests that there is a positive correlation between
entrepreneurship and economic growth, particularly in high-income countries.
Sustainable growth based on innovation and excellence requires an increasing
number of start-ups, which are likely to provide more and better jobs. If
Europe wants to successfully maintain its social model, it needs more
economic growth, more new firms, more entrepreneurs willing to embark in
innovative ventures, and more high-growth SMEs.
Entrepreneurship can also contribute to social cohesion for less-developed
regions and to putting unemployed or disadvantaged people into work.
Moreover, it can contribute to unlocking the entrepreneurial potential of
women, which has yet to be further exploited.
There is a need to create a more favorable societal climate for
entrepreneurship, based on an integrated policy with a view to not only
changing the mindset but also improving the skills of Europeans and removing
obstacles to the start-up, transfer and growth of businesses.
Research suggests that cultural support (e.g. through education programs,
promotional campaigns, etc.) is positively linked with the amount of
entrepreneurial activity in the EU. Promoting entrepreneurship among young
people is a key element of the European Youth Pact adopted by the European
Council in March 2005.
Entrepreneurship is a key competence
Entrepreneurship refers to an individual’s ability to turn ideas into action. It
includes creativity, innovation and risk taking, as well as the ability to plan
and manage projects in order to achieve objectives. Entrepreneurship is a key
competence for all, helping young people to be more creative and selfconfident in whatever they undertake and to act in a socially responsible way.
Developing generic attributes and skills that are the foundations of
entrepreneurship is complemented by imparting more specific knowledge
about business according to the level of education.
Traditionally, formal education in Europe has not been conducive to
entrepreneurship. However, as attitudes and cultural references take shape at
an early age, the education systems can greatly contribute to successfully
addressing the entrepreneurial challenge within the EU. Therefore, while
recognizing that the entrepreneurship competence should be acquired
throughout lifelong learning, this Communication focuses on education from
primary school to university, including also secondary level vocational
education (initial vocational training) and technical institutions of tertiary
level. It aims to support Member States of the European Union in developing a
more systematic strategy for entrepreneurship education.
69
10.2
Entrepreneurship in school education
Entrepreneurship in framework curricula for schools
Including explicit objectives in curricula, together with guidelines for putting
them into practice, provides a more solid basis for entrepreneurship
education. Particularly in secondary education, there are subjects – although
now often extra-curricular – that can be used for entrepreneurship learning.
Entrepreneurship competence is developed in both formal and non-formal
settings (e.g. youth work and various forms of participation in society). Tools
for the recognition and validation of entrepreneurship-related skills acquired
in non-formal learning should be further developed.
Entrepreneurship in primary education (pupils below the age of 14)
Awareness should be raised of the benefits of basic entrepreneurship learning
to society at large and to learners themselves, even at the early stages of
education. As for all competences leading to better management of one’s own
life, the foundations are laid in the early years of education. At primary level,
nurturing qualities such as creativity and a spirit of initiative helps develop
entrepreneurial attitudes. This is best done through active learning based on
children’s natural curiosity. In addition, learning about society should also
include early knowledge of and contact with the world of work and business,
and lead to a better understanding of the role of entrepreneurs in the
community.
In a number of Member States, curricula already encourage schools to
guide children towards taking initiative and responsibility. However, examples
of more explicit entrepreneurship education are few. In general terms,
coherent initiatives or programs led by education authorities are still rare in
primary education; activities are often led by external actors, such as nonprofit organizations supported by the private sector. Nonetheless, there are a
number of good practices that should be disseminated to public authorities,
schools, teachers and parents.
Entrepreneurship in secondary education (from the age of 14)
Secondary education should raise students’ awareness of self-employment
and entrepreneurship as option for their future career. Entrepreneurial
mindsets and skills can best be promoted trough learning by doing and
experiencing entrepreneurship in practice, by means of practical projects and
activities.
In most European countries, curricula have broad objectives and include
subjects that would allow learning about entrepreneurship. However,
implementation often relies on the initiative of schools and teachers and the
support of the local business community. In a few Member States, practical
experience of entrepreneurship is embedded into the established courses.
Within vocational secondary education (initial vocational training), specific
training on how to start a company can be particularly effective, as students
are close to entering working life. However, with exceptions in some
countries, a real focus on entrepreneurship is missing in most cases, since the
main task is seen as being to train skilled workers.
70
There is a perception that secondary school curricula do not provide
sufficient motivation to teachers and schools to develop entrepreneurship
education. It is therefore crucial to offer them support and incentives.
Measures to support schools and teachers
Schools should be given support and incentives to encourage take-up of
entrepreneurship activities and programs, as many concrete examples of how
to do it exist already. The support should include providing initial and inservice training for teachers. The commitment of heads of schools and school
boards is crucial, as is parental involvement. Public authorities should take the
initiative and actively promote education for entrepreneurship to schools,
heads of schools and teachers.
An important first step at a national level is establishing formal cooperation
between different departments of the administration, given the horizontal and
interdisciplinary nature of entrepreneurship education. This cooperation can
lead to launching a national strategy or action plan.
One major obstacle is that teachers lack motivation and specific training.
The efforts that teachers devote to practice-based activities, sometimes even
outside their normal working hours, should be recognized as an official school
task. Supporting the efforts of dedicated organizations is an effective method
of spreading the entrepreneurial spirit in schools and of encouraging
partnerships with the business world.
Furthermore, the involvement of private actors should be seen by firms as
a long-term investment, and as an important aspect of their corporate social
responsibility. Private-public partnerships are crucial to the development of
entrepreneurship education. The establishment of school-businesscommunity links is a key element of successful programs. This process needs
to be encouraged further.
Numerous organizations are currently disseminating entrepreneurship
education across Europe by means of partnerships with the business world,
with a certain degree of public support. They promote programs based on
learning in practice, and provide training for teachers, and may act as drivers
of change in national educational policies. Their contribution to
entrepreneurship education is significant in most European countries.
10.3
Entrepreneurship in higher education
Universities and technical institutes should integrate entrepreneurship as an
important part of the curriculum, spread across different subjects, and require
or encourage students to take entrepreneurship courses. Combining
entrepreneurial mindsets and competence with excellence in scientific and
technical studies should enable students and researchers to better
commercialize their ideas and new technologies developed. Entrepreneurship
education provides specific training on how to start and run a business, and
encourages and supports business ideas from students.
There are few Chairs in Entrepreneurship in Europe, which lags behind the
US by a factor of four. Universities should integrate entrepreneurship in
different subjects of their study programs, as it may add value to all degree
courses. Also, in order to tackle the shortage of specialized professors,
71
entrepreneurship should be more broadly recognized as a specialization field
for doctoral programs.
Case studies and other interactive teaching methods are under-utilized, as
is the involvement of business people in the learning process. To encourage
entrepreneurial behavior, a supportive environment is needed. Higher
education establishments committed to entrepreneurship provide or facilitate
access to risk capital, management capacity building and networking. Business
plan competitions are an effective way to expose students to investors. The
presence of incubators and science parks also clearly signals universities’
commitment, through the practical supply of services.
Special attention should be paid to systematically integrating
entrepreneurship training in scientific and technical studies and within
technical institutions, in order to better enable spin-offs and innovative startups, and as a means to help researchers to acquire entrepreneurial skills.
Business schools and technical/scientific faculties should collaborate more.
More focus is needed on developing the skills and competencies necessary
for fully exploiting innovation and knowledge transfer activities in combination
with the commercialization of new technologies.
Academic spin-offs are increasingly seen as important means of enhancing
local economic development. However, in order to accomplish their new roles,
scientists as well as universities must build business and managerial
competencies.
There are some internal barriers, which prevent an entrepreneurial path to
be seen as a credible option. Problems also seem to pertain to labor mobility
in and out of academia, and to the ability to flexibly and strategically recruit
personnel within universities. Inter-sectoral mobility of researchers at all
stages of their careers (including at the level of doctoral training) should
become a normal component of a researcher career path. Such mobility should
also help to develop the necessary skills and competences for enhancing the
entrepreneurship culture and attitude within universities.
Finally, it is vital to create a critical mass of entrepreneurship teachers,
and to step up cross-border collaboration. Sharing of practices should be
increased.
10.4
The way ahead
The following recommendations for concrete action are based on evidence
and good practice found in Europe. Most of the action needs to be taken at
national or local level. The proposals aim to help formulate more systematic
approaches to entrepreneurship education and to enhance the role of
education in creating a more entrepreneurial culture in European societies.
A coherent framework
• National and regional authorities should establish cooperation between
different departments, leading to developing a strategy with clear objectives
and covering all stages of education in the context of the Lisbon national
programs.
• Curricula for schools at all levels should explicitly include entrepreneurship
as an objective of education, accompanied by implementation guidelines.
72
Support for schools and teachers
• Schools should be given practical support and incentives to encourage takeup of entrepreneurship activities and programs, through a range of different
instruments.
• Special attention should be given to training teachers, through initial and inservice training as well as practical experience, and to raising the awareness
of heads of schools and school boards.
Participation by external actors and businesses
• Cooperation between educational establishments and the local community,
especially businesses, should be encouraged. Involvement in formal and nonformal education should be seen by firms as an investment and as an aspect of
their corporate social responsibility.
• The use of student mini-companies at school should be further promoted. In
that context, the activity of organizations promoting these programs, such as
NGOs, should be recognized, and their initiatives more systematically
supported.
Fostering entrepreneurship in higher education
• Higher education institutions should integrate entrepreneurship across
different subjects and courses, notably within scientific and technical studies.
• Public authorities’ support is especially needed to provide high-level
training for teachers and to develop networks that can share good practice.
• Teacher mobility between university and the business world should be
encouraged, together with the involvement of business people in teaching.
The Commission will continue to support Member States’ actions on more
comprehensive policies, through coordination activities and specific projects.
It will disseminate good practice and raise the visibility of entrepreneurship
education through a wide range of actions, including the follow-up to the
Recommendation on key competences. From 2006, work on entrepreneurship
in higher education will be intensified. From 2007, the proposed new
Community Integrated Programme on Lifelong Learning will support
innovative projects with a European dimension, aiming to foster
entrepreneurial attitudes and skills and to promote links between educational
establishments and enterprises. The European Social Fund will continue to
support initiatives at European, national and local level.
73
CHAPTER 11
So what?
Policy Recommendations resulting from the Workshop
“Entrepreneurship and Human Capital”
Based on:
All the previous chapters
by
Mirjam van Praag
University of Amsterdam
Amsterdam Center for Entrepreneurship
74
11.1 Introduction
Most research papers about entrepreneurship start with the premise that
entrepreneurship generates high economic value. The value pertaining to
entrepreneurship is not limited to starting up new businesses. Entrepreneurial
attitudes and behaviour in large and incumbent firms is valuable too. But what
are the returns to the research on entrepreneurship? Today, academic
research in entrepreneurship is not much used by public and private decision
makers in practice. Is research on entrepreneurship a ‘sleeping giant’, whose
proceeds just need to be communicated to policy makers and other parties
involved in the practical process of making entrepreneurs (successful)? Or
should academic researchers themselves be more prone to undertake
research that contributes to (public) policy making?
In general, public authorities might expect valuable outcomes from
academic research in entrepreneurship with respect to fiscal policies (taxes,
subsidies, loan programs), capital markets (debt and equity), labor markets
(including targeted policies towards the unemployed, new immigrants, or the
emancipation of females or labor market participants of specific age groups),
bankruptcy laws, governance, education policies, the optimal organization of
the commercialization process of university research, intellectual property
rights, and other areas such as for instance legislation with respect to the
entrepreneurs’ administration and accounting processes. Moreover, capital
providers of debt and equity and, corporations active in corporate venturing
or stimulating corporate entrepreneurship (‘intrapreneurship’) might benefit
from the outcomes from academic research into the determinants of
successful entrepreneurship. Last but not least, the outcomes from academic
research might be insightful for entrepreneurs themselves at the various
possible engagement levels, to improve their strategies and/or organizations.
In particular, this bundle of research results has addressed the effects of
various sorts of human capital on entrepreneurs’ outcomes at various stages.
In what follows, I first discuss the implications resulting from the research and
from a discussion with academics, policy makers (and one entrepreneur) at
the workshop “Entrepreneurship and Human Capital”. I will then address to
what extent and in which areas such research can be better targeted towards
generating useful implications for public and private decision makers in
practice.
11.2 Policy implications
Human capital is a very influential factor of entrepreneurial productivity.
Improving the entrepreneurs’ relevant aspects of human capital, as well as
increasing the likelihood that individuals with valuable human capital will
become entrepreneurs will enhance economic outcomes. The following human
capital factors are addressed: family background, (entrepreneurship)
education, experience, (entrepreneurship) specific and general human capital,
imported versus ‘home made’ human capital, and culture. Entrepreneurial
outcomes considered are start-up, takeover, opportunity identification and
pursuit, survival, growth, income and profits. What policy measures can be
taken or developed further to improve entrepreneurship at its various
75
engagement levels by means of a better development or utilization of human
capital?
Family background
Entrepreneurs coming from business-owning families are distinct from
entrepreneurs coming from families without a business. Approximately half of
the entrepreneurs come from business-owning families. However, although
those born into families owning a business are more likely to become
entrepreneurs, they are not more likely to become successful.
Entrepreneurs coming from business-owning families who have worked in
the family business are, however, associated with more successful
entrepreneurship. This effect is measured on top of the effect of management
experience and within-industry experience. This means that working in a
family business itself generates valuable experience. Hence, in order to
stimulate more successful entrepreneurship, it is useful to give the possibility
and incentives to people to work in their (or another) family business, before
starting a business. This could be done by offering internships in small family
businesses within the educational system, or by stimulating people to work in
a family business at a young age if they want to become entrepreneurs.
Governmental programs providing mentoring, internships or apprenticeshiptype training may help to reduce historical inequalities in business ownership
patterns. A program might be developed in which national associations for
SMEs cooperate with institutes of education and/or Chambers of Commerce
led by the Ministry of Economic Affairs to offer (subsidized) internships or
short-term labor contracts within family firms. Ministries of Social Affairs
might also want to embark on such a collaborative project because this kind of
experience might also be useful for unemployed prospective entrepreneurs.
Entrepreneurs coming from business-owning families are more inclined to
enter the entrepreneurial market by takeover of a (family or non-family) firm,
rather than by starting up a firm from scratch. Even taking into account that
entrepreneurs who do not come from business-owning families have no
possibility of taking over a family firm, entrepreneurs born into family-firms
are more inclined to taking over a (non-family) firm. Moreover, entrepreneurs
born into family firms have lower levels of education than entrepreneurs born
in families without firms. In addition, the higher one’s education level is, the
lower is the likelihood that an entrepreneur has entered through takeover
rather than start-up.23
This has implications for public policy to increase the likelihood that
economic value embedded in existing family firms is preserved. Today, there
are many family firms owned by European post-war baby boomers for which
successors are required. Since the inclination to take over a firm within the
family has decreased tremendously over the past decades, successors should
be found outside of the family.
Therefore, while minimizing transaction costs, demand and supply of family
firms should be brought together in an effective manner. Demand is likely to
23
It should perhaps not be a problem that entrepreneurs born into business-owning families
pursue lower levels of formal education. In half of the cases, they have valuable human capital
in the form of work experience in the family business that others don’t have.
76
be highest among lower educated offspring from business-owning families.
Thus, markets for the demand and supply of family firm can best be
advertised through channels that reach business-owning families and their
(lower educated) offspring. Culturally, families should accept that if the
takeover of the family firm is no option for either side for whatever reason,
the family might be useful to find alternative businesses for the offspring of
business-owning families. Business-owning families might set up multigenerational trade fairs of family businesses. Governmental policy makers,
accountants and/or banks might have incentives to help organize such
markets.
Education
The returns to formal education are high for entrepreneurs. Although the
likelihood of becoming an entrepreneur is not affected by one’s education
level, the proceeds from the individual’s entrepreneurial venture are affected
positively by education. The higher is the education level, the more likely is
survival, higher growth, higher profits and higher incomes for entrepreneurs.
Even better, comparing the returns to education for entrepreneurs and
employees in terms of the private incomes an additional year of education
generates, entrepreneurs benefit more from formal education than employees.
Hence, entrepreneurship should be stimulated in schools of higher
education. Especially in Europe, students do hardly consider entrepreneurship
as a career option. Awareness of this career option should be created, and
also of the higher private benefits of education in this career option than in
the employment situation. Students and policy makers are not yet aware of
this higher return to education for entrepreneurs that has been measured for
the US and Europe (The Netherlands only). Of course, public authorities
should be aware that private returns to education are not the same as social
returns to education. The latter includes spill over effects and is in fact the
rationale for public policy measures. More research should be done to
investigate whether the higher private returns to education for entrepreneurs
go together with higher social returns to education.
Hardly any research has provided correct measures of the effectiveness of
entrepreneurship courses at schools (whether primary, secondary or tertiary
education). Recently, a rapid increase in entrepreneurship education has been
observed, across the discipline spectrum. Many programs have been
implemented in many countries. However, participation rates are low (less
than one percent on average) and usually only a self-selected group of
students is involved in these programs. This raises even more difficulties for
a proper assessment of their effectiveness. Existing research supports the
effectiveness of entrepreneurship education programs by means of ‘minienterprises’: For instance, in Norway and Sweden, 20 percent of the children
who participated in a program turned out to be an entrepreneur in the long
term. However, the research frameworks used do not identify whether the
higher percentage of start-ups is caused by the program or whether the selfselected group of students participating in the program would have a higher
propensity to become an entrepreneur even without participation. Evaluation
research, including using pre- and post-measurement of entrepreneurial
77
competencies and attitudes among a treatment and a control group is badly
needed.
Moreover, not only should the effectiveness of entrepreneurship education
be evaluated, it should be compared to the alternative of making formal
education
in
general
more
entrepreneurship
oriented.
Perhaps,
entrepreneurship education should not be a separate field. Education itself
should perhaps be innovated to become more entrepreneurial and creative.
That would be an alternative manner –to entrepreneurship education as a
field- in which students could be taught to be creative and to be a leader,
rather than to reproduce and learn how to obey rules.
Another unexplored research area is at what stage entrepreneurship
education or more entrepreneurship-oriented education should be
implemented in the educational system. Should it be at the level of primary,
secondary or tertiary education? Should it also be provided to older people to
create better long-term prospects by life long learning? Nowadays, many
(unemployed) older Europeans start up businesses.
The view of Mr Baldassari from the Commission of the European
Communities is as follows. Entrepreneurship should be implemented at all
school levels, both as separate experience workshops (mini-enterprises) and
by making the education program in general more entrepreneurship oriented.
He admits that his views are not yet based on adequate research, which the
Commission of the European Communities stimulates to be developed soon.
Specific versus general work experience, schooling and skills
Research into the relative value of specific and general human capital
sketches the following picture: entrepreneurship-specific human capital like
business ownership experience, entrepreneurial and management capabilities
all enhance the identification and pursuit of entrepreneurial opportunities
more strongly than general work experience. Moreover, specific types of
experience affect entrepreneurial business success also more strongly than
general experience. Internships in entrepreneurial firms (in the same industry)
and other programs might be likely candidates to increase the number and
quality of the pool of entrepreneurs. These programs could also enlighten the
situation of immigrants. Immigrants seem to benefit more from human capital
acquired in the host country than from human capital in the country of origin.
Especially work experience in the host country is beneficial for their survival
chances. The acquisition of entrepreneurship-specific human capital in host
countries is therefore a good instrument for stimulating absorption.
Successful entrepreneurship requires a balanced set of skills, whereas
social and technical skills as such are beneficial for successful
entrepreneurship. Hence, a diversified education curriculum as well as multifacetted labor market roles and a special orientation towards social and
technical skill acquisition would be beneficial for entrepreneurial performance.
However, not all capabilities have to be in one person, and entrepreneurial
teams might be a good way of developing Jacks-of-all-Trades. These
findings should be incorporated into the discussion of adequate
entrepreneurship education at schools and through internships or labor market
experience programs. More research on team entrepreneurship and
complementary entrepreneur skills needs to be undertaken.
78
Culture: Europe versus the United States
Why is it that the start-up, growth and innovative power of firms in Europe
lag behind the United States? Besides the usual suspects, such as legislation,
labor market rigidities, taxes, governance, and financial markets, culture or
the perception of entrepreneurship might as well explain the difference. In
Europe, for instance, students have a different and less supportive attitude
towards entrepreneurship. They usually view their education as preparing
them for the labor market of employees. As has been said, they hardly
consider entrepreneurship as a career option, No wonder that the (voluntary)
participation rates in entrepreneurship courses are so low across the board in
Europe whereas in the United States the demand for such courses exceeds
the supply (by entrepreneurship professors) by far. In the US, the majority of
students wish to become an entrepreneur. The cultural difference might be
caused by different attitudes towards risk, different views on the social status
of entrepreneurs, or alternatively, by different perceptions of failure. Perhaps,
these attitudes and perceptions should be changed. Research should be
initiated to provide guidelines as to whether this is a vital route to pursue
(changing culture), to assess what is the underlying problem and how can that
be solved? Changing attitudes might lead to increased numbers of (fastgrowing) start-ups. Perhaps, besides Labor Day, ‘Entrepreneurship Day’
should be an annual event to create an atmosphere that stimulates
entrepreneurship (esteem).
11.3 The value of academic research in entrepreneurship and human capital
Academics and policy makers should be aware that researchers use many
different human capital variables to understand many different measures of
entrepreneurial performance at various engagement levels. The human capital
factors that benefit the growth of an entrepreneurial venture can, of course,
be very different from the human capital factors that are beneficial for
opportunity identification or for start-up. Likewise, research is mostly based
on one country (with a specific cultural background and institutional
arrangement). It should be acknowledged that policy implications based on the
research pertaining to one country, for instance the US, must not be
considered a source of reference for other countries, such as those
constituting the European Union. Moreover, we should actually always
acknowledge the specificities of research. For instance, the human capital
factors contributing to successful entrepreneurship for immigrants from the
FSU to Israel during a massive wave of immigration cannot be attributed to
immigrants from or to other countries at different stages. Therefore, as the
research based on American datasets dominates, more research based on the
(diverse) countries constituting the European Union should be pursued.
A second concern is that academics and policy makers should be more
aware of the value of research to evaluate programs, for instance in
entrepreneurship education. In order to evaluate programs effectively,
researchers and institutions where the programs are implemented should
cooperate closely. Programs should be set up, from the outset, in such a way
that evaluation of the differences between a (randomly selected) treatment
and control group by means of pre- and post-measurement is possible. This
79
implies that some individuals (students) will temporarily not obtain the best
possible education, but a ‘placebo’ as in the medical science. The culture in
schools and universities is opposed to such ‘discrimination’ of the control
group. However, to really test whether the new program is better and how it
can be improved, such temporary ‘hardships’ are really needed in the form of
field experiments.
A third concern is related to the lack of field experiments. The
econometric measurement of ‘effects’ of for instance schooling on
entrepreneurship success should be up-to-date and not just based on
conditional correlations.
Fourth, it seems that there are so many topics that are really policyrelevant and that have hardly been studied. Properly measuring the effects of
entrepreneurship education, of culture, of various sorts of general or
vocational education are just starting points. Academics should really remain
in touch with policy makers such that we know what type of research is
relevant at an early as possible stage.24
Finally, academics have already done much research that is relevant for
policy. Communication of research results should improve. Better developed
networks combining academics, entrepreneurs and the relevant private and
public policy makers might be useful to this end. A two-way flow of contacts
and ideas is essential
24
To stimulate academics to get out of their ivory tower and communicate with public and
private policy makers, their rewards should perhaps be based on different performance
measures. The measure that is most heavily used nowadays to evaluate the performance of
academics is the number and quality of publications in top academic journals.
80
Main Sponsor: Fortis Bank Nederland
Entrepreneurship and Human Capital
What is the effect of initial and acquired human capital on the likelihood that people become
entrepreneurs, either through start-up or through takeover? What is the effect on the
likelihood that people identify and pursue entrepreneurial opportunities? How can human
capital influence the performance of entrepreneurs such as their survival, growth and
incomes? Is the return on general human capital investments higher or lower than the return
on (entrepreneurship) specific human capital? Do immigrants benefit less from the human
capital acquired in their original country than in their host country? How can universities
commercialize their human capital embedded in the many researchers and what is the role of
entrepreneurship? And how do rigidities in the labor market affect the relationship between
entrepreneurship and education? These are the questions addressed in this book.
Based on the answers, implications and recommendations for researchers, policy makers and,
last but not least, entrepreneurs are discussed.
ACE
ACE provides a forum for dialogue and networking between internationally high standing
academics and entrepreneurs, corporate managers and policy makers, who are influential in
the Netherlands. Thus, ACE aims at bridging the gap between academics and the corporate
world. In all, the ambition of ACE to influence the entrepreneurship climate in the
Netherlands, will be realized by shaping the research agenda and communicating its output to
corporate and governmental policy makers. ACE is sponsored by Fortis Bank Nederland and
cooperates with MKB-Nederland, the association for SMEs in the Netherlands.
Mirjam van Praag
Mirjam van Praag is the founding director of ACE and Professor of Entrepreneurship and
Organization at the University of Amsterdam School of Economics and School of Business.
This book has been written with the financial support of
the Gate2Growth Academic Network in Entrepreneurship, Innovation and Finance
July 2006