Journal of Business Venturing 21 (2006) 866 – 885

New business start-up and subsequent entry into

self-employment
Lars Kolvereid *, Espen Isaksen 1
Bodø Graduate School of Business, N-8049 Bodø, Norway

Received 20 April 2004; received in revised form 1 June 2005; accepted 1 June 2005

Abstract

The present research focuses upon new businesses which are started from scratch. The theories of
reasoned action and planned behavior are used to formulate hypotheses concerning self-employment
intentions and subsequent entry into self-employment. The hypotheses are tested using longitudinal
data from 297 Norwegian business founders. The results suggest that salient beliefs concerning self-
employment determine attitudes toward self-employment, that attitude and subjective norm determine
intentions to become self-employed, and that intentions to become self-employed determine actual
entry into self-employment. The findings strongly support the theory of reasoned action, but provide no
support for the extension of the theory represented by the theory of planned behavior.
D 2005 Elsevier Inc. All rights reserved.

Keywords: Self-employment intentions; Self-employment; The theory of planned behavior; Longitudinal

evidence; Norway

1. Executive summary

This study draws upon the theory of reasoned action (TRA) and the theory of planned
behavior (TPB) to develop hypotheses concerning the antecedents of attitudes toward self-
employment, intentions to become self-employed and actual entry into self-employment.

* Corresponding author. Tel.: +47 75517200; fax: +47 75517268.

E-mail addresses: lars.kolvereid@hibo.no (L. Kolvereid)8 espen.john.isaksen@hibo.no (E. Isaksen).
1
Tel.: +47 75517200; fax: +47 75517268.

0883-9026/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.jbusvent.2005.06.008
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 867

Self-employment is defined as a continuous construct, measured by the number of average

weekly hours spent working for the business. This is likely to be a more accurate indicator
of self-employment than, for example, a simple self-reported yes/no answer.
Kolvereid (1996a) identified several reasons for preferring a career as self-employed to
a career as organizationally employed. According to the TRA and the TPB, these reasons
are salient beliefs concerning self-employment that determine individuals’ attitude toward
self-employment (Hypothesis H1). According to the TRA, intentions to become self-
employed are determined by attitude and subjective norm (Hypothesis H2a). The TPB is
essentially an extension of the TRA that include perceived behavioral control (PBC) as a
third antecedent of intentions and behavior. Perceived behavioral control is similar to self-
efficacy. However, self-efficacy may be the preferable construct, because it is more clearly
defined. It has also been found to accurately predict behavioral intentions (Armitage and
Conner, 2001). Four different entrepreneurial self-efficacy factors were identified and
expected to be positively associated with self-employment intentions (Hypothesis H2b):
(1) opportunity recognition, (2) investor relationships, (3) risk-taking and (4) economic
management. Both theories state that intention to become self-employed is related to entry
into self-employment (Hypothesis H3a). According to the TPB, behavior is not only
determined by intentions, but also by self-efficacy (PBC) (Hypothesis H3b).
Data was collected in a mail survey from 1048 businesses that in May–June 2002 recently
had been registered in the Norwegian Central Coordinating Register for Legal Entities. The
register was opened in 1995 and is continuously updated with new business registrations.
Virtually, every new business in the country has to register before it is allowed to start trading
or hiring. Legal registration can be used as an indicator of new business birth since there is a
very high correlation between self-reported start-up among nascent entrepreneurs and legal
registration in Norway (Magnussen, 1997). Follow-up data was collected from 651
respondents using telephone interviews about 19 months after the initial mailing. Removing
businesses that were no longer in operation, business that were not started and owned by the
respondent, businesses that were not started from scratch, businesses started by portfolio
entrepreneurs and respondents who failed to submit complete data reduced the final sample
to 297. Sample and response bias checks were carried out, and no significant differences
between participants and non-participants were found.
The hypotheses were tested using hierarchical regression. All the three hypotheses
derived from the theory of reasoned action received strong support. However, no support
was found for the extension of the theory represented by the theory of planned behavior.
The entrepreneurial self-efficacy factors did not contribute significantly to the explanation
of the variance in intentions to become self-employed or entry into self-employment. Two
different explanations of these findings are suggested. First, according to the theory of
planned behavior, if the behavior is under complete volitional control, entrepreneurial self-
efficacy should not be expected to be strongly associated with intention and behavior. It is
possible that entrepreneurs in Norway face a more munificent environment than
entrepreneurs in other nations. In other countries, where resources are scarce and
competition more hostile, becoming self-employed may be under less volitional control
and entrepreneurial self-efficacy more important. Another possible explanation of the
findings is found in the measures applied. While the measures of salient beliefs, attitude
and subjective norm concern self-employed specifically, the self-efficacy measures are
868 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

general indicators of entrepreneurial self-efficacy. Hence, there may be a need to develop

more specific measures of self-employment self-efficacy, where the focus is on the skills
needed to become and stay self-employed.
The present research has several implications. Salient beliefs, attitudes and self-efficacy
may be altered in education and training programs. The finding that subjective norm is
positively associated with self-employment intentions suggests that we should not only
consider the venture idea and the entrepreneur when evaluating or promoting a new
venture project. The entrepreneurs’ household and significant others should also be taken
into consideration. Future research should attempt to identify specific self-efficacy factors
that are associated with different kinds of entrepreneurial behavior.

2. Introduction

The present research focuses upon entry into self-employment among founders of new
businesses which are started from scratch. Based on the theories of reasoned action and
planned behavior, hypotheses are developed to identify antecedents of (1) attitude toward
self-employment, (2) self-employment intentions and (3) entry into self-employment.
While earlier cross-sectional surveys have investigated entrepreneurial intentions among
students (Kolvereid, 1996b; Tkashev and Kolvereid, 1999; Krueger et al., 2000), Shook et
al. (2003) urged researchers to investigate entrepreneurial intentions among venture
creators using a longitudinal design.
Krueger and Carsrud (1993) argued that entrepreneurial behaviors, such as becoming
self-employed or starting a business, are intentional and thus best predicted by intentions
toward the behavior. Among integrated intention theories, which include different
theoretical constructs to explain and predict behavior, the most widely researched are
the theory of reasoned action (TRA) (Ajzen and Fishbein, 1980) and the theory of planned
behavior (TPB) (Ajzen, 1988, 1991). According to the TRA, behavioral intentions are
determined by: (1) the attitude toward the behavior, i.e. the degree to which the individual
has a favorable or unfavorable evaluation of the behavior in question, and (2) subjective
norm, which refers to the perceived social pressure to perform (or not perform) the
behavior. The TPB is essentially an extension of the TRA that includes perceived
behavioral control (PBC) as an additional antecedent of intentions and behavior. Perceived
behavioral control refers to the perceived ease or difficulty of performing the behavior. The
rationale behind the addition of perceived behavioral control in the TPB was that it would
allow prediction of behaviors that were not under complete volitional control. When there
were constraints to action, intentions alone were not sufficient to predict behavior
accurately. Moreover, the relative importance of attitude, subjective norm and perceived
behavioral control in the prediction of behavioral intentions is expected to vary across
behaviors and situations (Ajzen, 1991). In general, individuals tend to have stronger
intentions when the behavior in question is believed to be achievable (Bandura, 1997). The
antecedents of attitude, subjective norm and perceived behavioral control are
corresponding beliefs reflecting the underlying cognitive structure. For example, the
antecedents of attitude toward self-employment are the salient or behavioral beliefs people
have about the benefits (and costs) of entering into self-employment.
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 869

Armitage and Conner (2001) carried out a meta-analysis of the results reported from
applications of the TRA and the TBP. They identified 161 journal articles and book
chapters containing 185 independent empirical tests of the theories. Their results indicated
that the average correlations of the antecedents and behavioral intentions were .49 for
attitude, .34 for subjective norms and .43 for perceived behavioral control. Moreover, the
average correlation between behavioral intention and behavior was .47, and between
perceived behavioral control and behavior .37. Overall, their findings provided support for
the efficacy of the TPB as a predictor of intentions and behavior.
While the TPB was developed to explain individual behavior in general, it has
subsequently been adopted by entrepreneurship scholars. It seems evident that much of what
we consider to be entrepreneurship actually is intentional planned behavior (Krueger et al.,
2000). New ventures are the direct outcome of individuals’ intentions and subsequent actions
(Bird, 1992). Further, entry into self-employment is a behavior that is determined by
intentions (Katz, 1992). Previous applications of the TPB to entrepreneurship, however,
have generally used student samples and applied a cross-sectional design (Shook et al.,
2003). With reference to a sample of 128 undergraduate Norwegian business students,
Kolvereid (1996b) found that attitudes, subjective norm and perceived behavioral control
were significantly related to intentions to pursue a self-employed career. These findings were
later replicated by Tkashev and Kolvereid (1999), with regard to a sample of 512 Russian
university students from different disciplines. In both studies, perceived behavioral control
was found to be the strongest predictor of behavior followed by subjective norm and attitude.
These studies also noted that having a family background in entrepreneurship, gender and
entrepreneurial experience, only indirectly influenced entrepreneurial intentions through
their effect on attitude, subjective norm and perceived behavioral control. Krueger et al.
(2000) tested the TPB’s ability to predict intentions to start a business in a sample of 97 senior
university business students. Attitude, subjective norm and perceived behavioral control
were significantly related to entrepreneurial intention, but only perceived behavioral control
and attitude were significant predictors of intentions in the subsequent multivariate analysis.

3. New business start-up

Norway has a new business registration regime that requires all new businesses to
register before they are allowed to start trading or hiring. Under such circumstances, new
business birth can be simply defined as legal registration of the firm. Magnussen (1997)
investigated the business gestation process of 131 nascent entrepreneurs in Norway. He
found that 89% of the 65 respondents who reported that they had started a business had
registered their firm and formed a legal entity. Hence, business registration corresponded
very closely with the entrepreneurs’ self-perception of business birth, indicating that
business registration is an appropriate indicator of new business birth in Norway.2

2
This is probably not the case in the USA where Carter et al. (1996) found that only 49% of those who reported
to had started a firm had formed a legal entity. Since a much higher percentage of those who stated that they had
started a business had received payment from sales of goods or services, Reynolds and Miller (1992) and Carter et
al. (1996) suggested using first sales as an indicator of new business birth.
870 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

A new business can be established from scratch–a so-called de novo entry, or acquired
in some way or another–a so-called acquisitive entry. Dennis (1997) estimated that, in the
USA, 78.5% of all entries are de novo entries, while 21.5% are acquisitive entries. While
de novo entries have much in common, acquisitive entries encompass several different
forms of entries, for example, businesses that are inherited, businesses that are purchased
and franchise operations. Even though different entry modes clearly represent different
phenomena, many entrepreneurship studies fail to specify what entry mode(s) they are
studying. Under the present circumstances, the focus is on de novo entries, since they are
the largest category of entries and since they represent a homogenous group of new
businesses with respect to the fact that they are all started from scratch.
Some new businesses registrations are re-registrations of old businesses. This can occur
if the initial business goes bankrupt or closes down for some other reason, or if the
business changes legal form. Since such re-registrations clearly are not de novo entries,
they should be excluded if the primary focus is to understand the gestation process of
businesses that are started from scratch.
New businesses can be established by novice founders, or by people who have
previously started one or more businesses. Multiple business founders can be further
divided into serial entrepreneurs, who previously have started businesses, but later have
sold them or closed them down, and portfolio entrepreneurs, who control two or more
businesses (Westhead and Wright, 1998a; Carter and Ram, 2003). Since most portfolio
business founders are full-time engaged in running their businesses, they should be
excluded from the analysis since it does not make much sense to study self-employment
intentions and entry into self-employment among them.

4. Self-employment

In a dictionary, self-employment is defined as bworking for oneself, as a freelance or

owner of a business, etc., not employed by and employerQ (The Concise Oxford Dictionary,
1995). This definition suggests that the individual is faced with two alternatives when
selecting a career—either as self-employed or employed in an organization. This choice of
employment status was defined by Katz (1992, p. 30) as bthe vocational decision process
in terms of the individual’s decision to enter an occupation as a wage or salaried individual
or as a self-employed oneQ. This dichotomization is clearly a simplification. First, it does
not open for the possibility that people can choose to be unemployed or unemployable.
Second, it is not clear how to categorize people who combine working for an employer and
running their own business. There is evidence to suggest that a large proportion of new
business founders start their business as a part-time operation while they continue to work
for their employer (Carter et al., 1996; Delmar and Davidsson, 2000). Some of these micro-
businesses will surely remain part-time operations (Vesper, 1990). Third, some scholars
reserve the term self-employment for those who are truly independents and have no
employees. For example, Westhead et al. (2005) collected data from business owners with
two or more employees and argued that their respondents were therefore not self-employed
at the time of the survey (emphasis added). Others define self-employed as business
owners (e.g. Wennekers and Thurik, 1999).
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 871

The legal status of the business has also been used as a criterion for determining
whether a person is to be categorized as employed, self-employed or something else. In a
sole proprietorship, there is little doubt that only the entrepreneur can be self-employed. In
a partnership, however, things start to become blurred. One may choose to define all active
partners in a business partnership as self-employed or only define one of the partners—the
lead entrepreneur—as self-employed. Things become even more complicated in a limited
liability company. In some government statistics, owners of companies with limited
liability are classified as employees, even if they own 100% of their business.
In order to overcome some of the problems mentioned above, self-employment is here
considered as a continuous construct. The degree to which the founder of the business is
self-employment can then be assessed by measuring the number of hours the entrepreneur
spends working for the business during a specific period of time. Given that people can
combine entrepreneurship with employment, defining self-employment as a continuous
construct is likely to a more valid and precise indicator of self-employment than for
example a simple self-reported yes/no answer.

5. Hypotheses

There seems to be agreement that attitude toward an object is determined by beliefs

about that object (Ajzen and Fishbein, 1980). The bobjectQ under the present circumstances
is self-employment. Previous research has identified a number of reasons for entrepreneurs
to start their business. Asking a sample of business owner/managers in different countries,
the Society for Associated Researchers of International Researchers (SARIE) identified
motives for starting a business such as the need for personal development, independence,
approval, to follow the example of others and financial success (Scheinberg and
MacMillan, 1988; Shane et al., 1991; Birley and Westhead, 1994). As pointed out by
Carter et al. (2003), the SARIE studies (and similar studies of business owner/managers)
suffer from a severe shortcoming, in that they surveyed people years after their
occupational choice have been made. Such a retrospective procedure is likely to yield
biased and inaccurate results. Kolvereid (1996a) asked a sample of 250 business graduates
in their early career to state whether they preferred running their own business or being
employed by someone. This measure of choice intention (Ajzen and Fishbein, 1980) was
immediately followed by an open-ended question: bWhat is your main reason for
preferring one career path over the other?Q The findings suggested that the following
factors were important reasons for preferring a career as self-employed: economic
opportunity, authority, autonomy, challenge, self-realization and participation in the whole
process. Subsequent studies by Kolvereid (1996b) and Tkashev and Kolvereid (1999)
investigated the relationship between these salient beliefs concerning self-employment and
self-employment intentions among students. In both studies, the beliefs that were most
strongly associated with self-employment intentions were autonomy, authority, self-
realization and economic opportunity. The relationship between beliefs concerning self-
employment and entrepreneurial intentions has also been investigated in other surveys.
Douglas and Sheperd (2002), for example, found that the intention to be an entrepreneur to
be stronger for those with more positive attitude toward risk and independence.
872 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

Carter et al. (2003, p. 17) argued that breasons that individuals offer for getting into
business (or not) matter, because reasons are traditionally considered to be the basis of
intentionsQ. However, according to the TRA and the TPB, salient beliefs concerning the
object in question determine attitude toward the object, not intentions. Consequently,
behavioral beliefs concerning self-employment, autonomy, authority, self-realization and
economic opportunity, are under the present circumstances hypothesized to explain a
significant proportion of the variance in attitudes toward self-employment.

H1. The more important autonomy, authority, self-realization and economic opportunity is
for the individual in the choice of employment status, the more positive the attitude toward
self-employment.
According to the TRA and the TPB, attitude toward self-employment and subjective
norm with regard to becoming self-employed are central antecedents of intentions. Hence,
the second hypothesis to be tested is:
H2a. The more favorable the attitude and subjective norm with respect to self-
employment, the stronger the individual’s intention to become self-employed.
In the TPB, perceived behavioral control is included as an additional antecedent of
intentions and behavior. Perceived behavioral control is related to, but not similar to self-
efficacy. According to Bandura (1986, 1997), self-efficacy is concerned about cognitive
perceptions of control based on internal factors, whereas perceived behavioral control also
include external and more general factors. Armitage and Conner (2001, p. 479) distinguish
between three different, but related concepts: (1) self-efficacy defined as bconfidence in
one’s own ability to carry out a particular behaviorQ (measured by statements such as bI
believe I have the ability to. . .Q); (2) perceived control over behavior defined as bperceived
controllability of behavior (measured by statements such as bWhether or not I do X is
entirely up to meQ); and (3) perceived behavioral control defined as bthe perceived ease or
difficulty of performing behaviorQ (measured by mixed questions concerning the perceived
ease or difficulty associated with performing the behavior in question). This categorization is
also used and further enhanced by Ajzen (2002), who pointed out that the term bperceived
behavioral controlQ in the TPB should be read as bperceived control over performance of a
behaviorQ. However, based on their meta-analysis, Armitage and Conner (2001) concluded
that self-efficacy is the preferred construct, since self-efficacy is more clearly defined and
more strongly correlated with intention and behavior than perceived behavioral control.
After having reviewed the literature, the present authors identified five different
entrepreneurship self-efficacy factors relevant for entering into self-employment: Coping
with unexpected challenges (DeNoble et al., 1999), risk-taking (Chen et al., 1998),
developing new product and market opportunities (DeNoble et al., 1999), economic
management (Anna et al., 2000) and initiating investor relationships (DeNoble et al., 1999).
To test whether these entrepreneurial self-efficacy factors have any influence on intentions
to become self-employed as the TPB suggests, the following hypothesis was developed:

H2b. Entrepreneurial self-efficacy factors, along with attitude and subjective norm toward
self-employment, explain a higher proportion of the variance in self-employment
intentions than attitude and subjective norm alone.
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 873

While the TPB and its predecessor have been applied to predict intentions to
become self-employed and intentions to start a business, no previous study has as
yet applied the theories to predict business start-up or entry into self-employment
using a longitudinal design (Shook et al., 2003). Hence, the present study is the
first to test the link between self-employment intentions and subsequent entry into
self-employment:
H3a. The stronger the intention to become self-employed, the more likely the individual is
to enter into self-employment.
According to the theory of planned behavior, perceived behavioral control or
entrepreneurial self-efficacy is not only associated with intentions to become self-
employed (as suggested in Hypothesis 2b), but also with subsequent entry into self-
employment. Hence, the last hypothesis to be tested under the present circumstances is:
H3b. Entrepreneurial self-efficacy factors, along with intentions to become self-employed,
explain a higher proportion of the variance of entry into self-employment than self-
employment intentions alone.
The research model with hypotheses is illustrated in Fig. 1, H1, H2a and H3a can be
derived from both the TRA and the TPB. Hypotheses H2b and H3b concern the effect of
entrepreneurial self-efficacy on intentions and behavior, respectively, and represent tests of
the extension of the TRA represented by the TPB.

6. Method

The Norwegian Central Coordinating Register for Legal Entities was established in
1995. As the name suggests, this register coordinates information that exist in other
government registers, including (1) the register of employers, (2) the register of business
enterprises and (3) the value added tax register. Hence, the Central Coordinating Register
contains all businesses that have employees, all limited liability companies and

Salient H1
Attitude
beliefs H2a

Subjective H2a H3a

Intention Behavior
norm

H2b
H3b
Self-efficacy
(PCB)

Fig. 1. The research model with hypotheses.

874 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

partnerships, and all sole proprietorships obliged to pay VAT.3 Businesses that register are
assigned a unique organization number that identify the business for government
authorities and others. The four most common legal forms of new businesses in the
register are sole proprietorships, partnerships with mutual responsibility, partnerships with
shared responsibility and unlisted limited liability companies. Since a total of 98.6% of
Norwegian new registrations in 2002 chose one of these four legal forms (Statistics
Norway, 2004), other less common legal forms were disregarded. All new businesses that
registered during weeks 21–24 2002 were approached. In other words, the whole
population of new business registrations during these 4 weeks constituted the sampling
frame. With only a week delay, the register delivered lists of new businesses that were
registered each of these weeks. Within a week after the register supplied these lists, a
questionnaire was mailed out in four rounds to 603 businesses registered in week 21, 866
businesses registered in week 22, 747 businesses registered in week 23 and 905 businesses
registered in week 24. A reminder with a new copy of the questionnaire was sent out in
four rounds 3 weeks after the initial mailings. A total of 3121 businesses were approached.
Of the questionnaires mailed out, 126 were returned unreachable, while we received 1048
competed questionnaires—a response rate of 35.0%.
The second round of data collection took place during week numbers 5–8 in 2004 (i.e.
about 19 months after the initial mailing). A professional survey agency was engaged to
telephone the respondents to the mail survey in order to find out what had happened to the
businesses since the first round of data collection. Among the 1048 businesses that
responded to the mail survey, 29 businesses were excluded since they had been deleted
from the register. Another six businesses were excluded because they had more than 50%
missing data on the first round of data collection. Finally, 33 respondents were excluded
since the business or the contact person was not listed in any of the available telephone
directories. The survey agency attempted to reach all the remaining 980 respondents.
Among these, 275 persons were inaccessible and 54 others refused to participate. A total
of 3924 telephone calls were made in order to collect follow-up data from a total of 651 of
the business founders, 66% of the 980 founders on the list.
Among the 651 respondents from the second round of data collection, 557 reported that
the business was still in operation. Removing respondents who stated that (1) they were
not responsible for the new business start-up, (2) not owner of the business, (3) the start-up
was an acquisitive entry or (4) the businesses was started by a portfolio entrepreneur,
further reduced the sample to 363 cases. Finally, cases containing incomplete data were
removed, reducing the sample to 297 businesses started from scratch by novice or serial
business founders still operating at the time of the follow-up telephone survey. To check
for sample selection and response bias, the final sample of 297 was compared to the
remaining 2698 cases that received the questionnaire. No difference with regard to legal
status or localization (county and urban/rural setting) was found at the .01 level of
statistical significance. Subsequently, the 297 respondents who are included in the final
sample were compared to the 751 respondents who replied to the mail questionnaire

3
At the time of the initial data collection in 2002, this included, with few exceptions, all sole proprietorships (as
well as other businesses) with an annual turnover of NOK 30,000 or more. 1 NOK = approx. 0.14 USD.
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 875

survey, but did not participate in the follow-up telephone survey or failed to comply with
the sample requirements specified above. Cross tabulations and t-tests showed no
significant differences at the .01 level of statistical significance between these groups with
regard to any of the independent variables used in the present analysis including self-
employment intentions. These tests indicate that sample selection and response bias are not
serious problems and suggest that the sample is representative of the population.
In the final sample, the average age of the respondents was 37.85 years (range 18–72).
A total of 24.7% were females. Among respondents, 19.0% reported that they had one or
more business partners and 14.7% had previous start-up experience. The proportion of
businesses by industry was as follows: 19.7% in production, 13.3% in trade and 67.0% in
services.

6.1. Measures

Salient beliefs were measured using four belief-based measures of self-employment

identified by Kolvereid (1996a): autonomy, authority, economic opportunity and self-
realization. Two to four items were used as indicators of each dimension and respondents
were asked about the extent to which these factors were important for them to consider in
their choice of job. A principal component analysis was carried out in order to ensure the
independence of scales. Orthogonal rotated scales were used in order to avoid possible
multicollinearity problems in the subsequent multivariate testing of hypotheses. The
results are shown in Table 1.
As shown in Table 1, the principal component analysis produced four components as
expected. The communalities are satisfactory and there are no side loadings to warrant any
concern. Component scores were calculated using the regression method. Cronbach’s a for

Table 1
Principal components analysis with varimax rotation of salient beliefs concerning self-employment (n = 297)
Autonomy Authority Economic opportunity Self-realization Communality
Eigenvalue 2.65 2.28 1.95 1.66
Cumulative % 22.10 41.11 57.33 71.17
Economic opportunity .05 .07 .80 .05 .65
Merit-based compensation .23 .01 .75 .25 .68
Profits/profit-sharing .23 .08 .76 .03 .63
Freedom at work .81 .01 .02 .07 .66
Independence at work .89 .09 .13 .05 .81
To be your own boss .68 .26 .30 .17 .65
Self-direction .73 .13 .15 .20 .62
Have authority .09 .86 .11 .14 .78
Have the power to .04 .88 .02 .16 .80
make decisions
Have full control .21 .77 .07 .09 .65
Self-realization .12 .13 .06 .89 .83
To realize your dreams .21 .23 .08 .82 .78
Cronbach’s a .82 .82 .70 .76
Eigenvalues and the percent variance explained are after varimax rotation. Kaiser–Meyer–Olkin measure of
sampling adequacy = 0.750. Bartlett test of sphericity approx. chi-square = 1403, df = 66, p V .0000.
876 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

autonomy was .82, for authority .82, for economic opportunity .70 and for self-realization
.76. While these values generally are considered satisfactory (Hair et al., 1998), there
seems to be some room for improvement of the economic opportunity scale.
Attitude toward self-employment was measured using a scale developed by Gundry and
Welch (2001). This measure refers to the degree to which the founder is committed to the
new business and how much he or she is willing to sacrifice in order to become and to stay
self-employed. Respondents were asked along a seven-point scale about the degree to
which they agreed to the following statements: (1) bI would rather own my own business
than earn a higher salary employed by someone elseQ. (2) bI would rather own my own
business than pursue another promising careerQ. (3) bI am willing to make significant
personal sacrifices in order to stay in businessQ. (4) bI would work somewhere else only
long enough to make another attempt to establish my businessQ. A self-constructed fifth
item was also added: bI am willing to work more with the same salary in my own business,
than as employed in an organizationQ. All items loaded on the same factor in a principal
component analysis, but reliability analysis indicated that the fourth item should be
excluded since the Cronbach’s a of the composite index increased by removing this item.
Responses to the four remaining statements were added together and divided by four in
order to obtain a summary measure of self-employment attitude (Cronbach’s a = .81).
The measure of subjective norm was adopted from Kolvereid (1996b). The new
business founders were asked along a seven-point scale if (a) bmy closest familyQ, (b) bmy
closest friendsQ and (c) bpeople who are important to meQ think that the respondent should
not (= 1) or should (= 7) pursue a career as self-employed. Subsequently, in order to assess
motivation to comply, respondents were asked along a seven-point scale (from 1 = not at all
to 7 = very much) about the extent to which they cared about the opinion of (a) bmy closest
familyQ, (b) bmy closest friendsQ and (c) bpeople who are important to meQ in their choice
of employment status. The belief items were re-coded into a bi-polar scale (1 =  3 to
7 = + 3), multiplied with the respective motivation to comply item, and the scores averaged
in order to obtain an overall measure of subjective norm (Cronbach’s a = .77). Similar
measures of subjective norms have been used in a number of studies (Ajzen and Fishbein,
1980; Ajzen and Driver, 1992; Madden et al., 1992). The use of such a multi-item measure
to assess subjective norm is strongly recommended over less reliable single-item measures
(Armitage and Conner, 2001).
A total of 18 items were included in the original entrepreneurial self-efficacy scales.
One self-constructed item concerning investor relationship was added. Following the
recommendations made by Betz and Hackett (1998) and Bandura (2001) concerning the
measurement of self-efficacy, respondents were asked to indicate their degree of
confidence in performing the various tasks successfully along a 11-point scale from
0 = bno confidence at allQ, through 5 = bsome confidenceQ, to 10 = bcomplete confidenceQ.
Based on principal component and reliability analyses, one item (btaking responsibility for
ideas and decisionsQ) was dropped from the risk-taking index. The results from a principal
component analysis with the remaining 18 items are shown in Table 2.
As Table 2 shows four components emerged. These components were subsequently
labelled (1) opportunity recognition, (2) investor relationships, (3) risk-taking and (4)
economic management. All six items suggested by DeNoble et al. (1999) to measure new
products and markets loaded on the first component, opportunity recognition (Cronbach’s
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 877

Table 2
Principal components analysis with varimax rotation for entrepreneurial self-efficacy (n = 297)
Opportunity Investor Risk Economic Communality
recognition relationships taking management
Eigenvalue 3.83 3.22 2.85 2.83
Cumulative % 21.28 39.20 55.04 70.76
See new market opportunities for .69 .35 .21 .02 .65
new products/services
Discover new ways to improve .82 .10 .13 .07 .70
existing products/services
Identify new areas for potential .79 .23 .23 .09 .73
growth
Design product/services that solve .79 .03 .22 .09 .68
current problems
Create product/services that fulfill .78 .05 .17 .07 .64
unmet customer needs
Bring a product concept to a .70 .30 .13 .06 .61
market in a timely manner
Be able to obtain sufficient funds .13 .68 .05 .42 .66
for future growth
Develop and maintain favorable .20 .86 .13 .19 .84
relationships with potential investors
Develop relationships with key .22 .87 .13 .16 .85
people who are connected to
capital sources
Identify potential sources of funding .21 .84 .13 .18 .79
for investments
Work productively under continuous .15 .19 .70 .11 .57
stress, pressure and conflict
Tolerate unexpected changes in .14 .26 .73 .02 .62
business conditions
Persist in the face of adversity .25 .16 .63 .18 .51
Take calculated risks .17 .05 .72 .15 .58
Make decisions under uncertainty .20 .04 .79 .10 .68
and risk
Manage expenses .03 .28 .08 .88 .86
Control business costs .13 .19 .10 .91 .90
Manage cash flows .08 .17 .06 .92 .88
Cronbach’s a .89 .91 .80 .94
Eigenvalues and the percent variance explained are after varimax rotation. Kaiser–Meyer–Olkin measure of
sampling adequacy = 0.839. Bartlett test of sphericity approx. chi-square = 3649, df = 153, p V .0000.

a = .89). The three items suggested by DeNoble et al. (1999) as indicators of investor
relationships as well as the self-constructed item loaded on the second component
(Cronbach’s a = .91). The three items used by DeNoble et al. (1999) to measure coping
with challenges and the three items used by Chen et al. (1998) to measure risk-taking
loaded together on the third component (Chronbach’s a = .80). Finally, the three items
suggested by Anna et al. (2000) to measure economic management loaded on the fourth
component (Cronbach’s a = .94). All communalities are satisfactory and none of the side-
loadings warrants serious concern. Component scores were calculated using the regression
method to be used in the subsequent testing of hypotheses.
878 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

The following demographic control variables were included: age, gender (coded as
0 = female, 1 = male), education (coded as 0 = no college education, 1 = at least some
college education), parental background in entrepreneurship (parental role model) (0 = no,
1 = yes) and a measure of past behavior indicating whether the founder is a novice or serial
entrepreneur (coded as 0 = novice, 1 = serial). In addition, two aspects of the business were
controlled for: urban localization, defined as businesses located in municipalities with
10,000 or more people4 (coded as rural = 0 and urban = 1) and industry. Industry is in this
study categorized into three groups: production, trade and services. The first group,
production, consists of industries such as agriculture, forestry, fishery, fish farms,
manufacturing and construction. The second group, trade, consists of industries such as
retail, wholesale, hotels and restaurants. The last group, services, consists of industries
such as transportation, financial services, personal services, professional services and other
services.
Intention to become self-employed was measured by a single item: bHow likely are you
to be working full-time for the new business in one year from now? Responses were given
along a seven-point scale from 1 = very unlikely to 7 = very likely. This is a self-predictive
measure of intentions, somewhat different from measures of behavioral desires (for
example, bI would prefer a career as self-employed to a career as employed in an
organizationQ) and behavioral intentions (for example, bI intend to become self-employed
within a yearQ). With regard to these different measures of intentions, Armitage and
Conner (2001) concluded that: (1) The multiple correlation of attitude subjective norm and
PBC with desire is generally significantly stronger than with intention, self-prediction or
the mixed measure. (2) Self-predictions and intentions are more strongly related to
behavior than desires. Finally, (3) PBC is a less important determinant of behavior when
measures of intentions or self-predictions are employed rather than measures of desire.
To check the convergent validity of the intention measure, the applied self-prediction
measure was correlated with a measure of behavioral intentions: bOne year from now you
intend to be. . .Q The options were indicated on a five-point scale from 1 = only engaged as
employed, through 3 = engaged as both self-employed and employed (on a 50/50 basis), to
5 = only engaged as self-employed. The correlation between the applied self-predictive
measure of intention and the behavioral intention measure was .66, indicating a high
degree of consistency between the two measures.
Behavior or entry into self-employment was measured in the second round of data
collection by a single item: bHow many hours do you currently on average work for the
businessQ. As shown in Table 3, respondents in average reported to work 27.10 weekly
hours for the business in question (S.D. = 20.78). This is a self-reported measure of
behavior. Ideally, behavior should be measured by observation rather than by self-reports.
However, self-reported measures are typically more strongly related to behavioral
intentions than are measures of observed behavior. The explanation is probably that the
measurement correspondence is higher for self-reported measures of behavior than for
measures of observed behavior.

4
This statistical definition has been used by several authors, including, for example, Westhead and Wright
(1998b).
Table 3
Descriptive statistics and Spearman correlations among the analysis variables (n = 297)
Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Control variables

L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

1. Age 37.85 10.51 1
2. Gender 0.75 0.43 .06 1
3. Education 0.55 0.50 .12 .05 1
4. Parental role model 0.34 0.47 .05 .03 .07 1
5. Serial entrepreneur 0.15 0.35 .33 .15 .07 .07 1
6. Urban location 0.73 0.44 .04 .02 .20 .03 .01 1
7. Production 0.20 0.39 .15 .15 .29 .06 .01 .08 1
8. Trade 0.13 0.34 .06 .02 .20 .03 .01 .03  .19 1

Salient beliefs
9. Autonomy 0.00 1.00 .03 .14 .06 .11 .02 .04 .13 .02 1
10. Authority 0.00 1.00 .09 .00 .10 .02 .09 .02 .04 .08 .07 1
11. Economic opportunity 0.00 1.00 .07 .00 .19 .09 .15 .02 .06 .14 .05 .06 1
12. Self-realization 0.00 1.00 .24 .18 .02 .00 .18 .03  .03 .00 .06 .03 .02 1

Attitude and subjective norm

13. Attitude 4.97 1.30 .01 .04 .24 .02 .00 .10 .08 .11 .28 .20 .26 .23 1
14. Subjective norm 6.20 5.85 .01 .08 .17 .04 .05 .06 .03 .03 .25 .20 .18 .13 .28 1

Self-efficacy (PBC)
15. Opportunity recognition 0.00 1.00 .01 .12 .13 .04 .00 .02  .06 .06 .11 .12 .11 .18 .21 .11 1
16. Investor relationships 0.00 1.00 .00 .02 .12 .02 .08 .09 .05 .08 .03 .17 .21 .01 .19 .15 .01 1
17. Risk taking 0.00 1.00 .07 .09 .02 .03 .11 .06 .02 .02 .10 .05 .08 .12 .26 .06 .05 .05 1
18. Economic management 0.00 1.00 .04 .08 .05 .18 .01 .08 .01 .01 .10 .16 .11 .02 .13 .13 .02 .03 .06 1

Intentions and behavior

19. Self-employment intention 4.75 2.30 .01 .04 .08 .01 .13 .06 .07 .06 .07 .13 .21 .11 .39 .33 .18 .10 .11 .01 1
20. Entry into self-employment 27.10 20.78 .04 .14 .14 .02 .08 .07 .10 .10 .07 .16 .21 .02 .24 .21 .11 .09 .11 .04 .63
Correlations z.12 are statistically significant at the .05 level (two-tailed tests).

879
880 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

To check the convergent validity of the applied measure of behavior, responses were first
divided into two categories (1) those who reported to work at least 35 h a week for the
business in question and (2) those reporting to work less than 35 h/week in their newly
founded business. Then this measure was cross tabulated with a measure of the
respondent’s self-perceived employment status also obtained in the follow-up telephone
survey: bAre you currently full-time self-employed?Q (0 = no, 1 = yes). As expected, the
two measures were highly associated with each other (chi-square = 111.11, significant at
the .001 level, classifying 80.7% of the cases correctly). Thus, the applied measure of self-
employment appears to be a valid indicator of employment status.

7. Results

Descriptive statistics and correlations among the analysis variables are shown in
Table 3.
The correlation matrix offers preliminary support for several of the hypotheses. The
correlations between the four measures of salient beliefs and attitude are all statistical
significant and in the expected direction. Moreover, attitude and subjective norm are
significantly correlated with intentions in the expected direction. Intentions are also
strongly correlated with behavior in the expected direction. A disappointing finding is that
only one of the four measures of self-efficacy is significantly correlated with intentions
and behavior.
The hypotheses were tested using hierarchical regression. Before each analysis,
multicollinearity checks were conducted. The maximum VIF value found was less than
1.3. Hence, multicollinearity was not a problem in the analyses. In order to test the first
hypothesis concerning attitude toward self-employment, the control variables were first
entered into a regression followed by a second regression that included the control
variables as well as the four measures of salient beliefs. In support of Hypothesis H1, all
four measures of salient beliefs were significantly related to intentions (see Table 4, second
column). Moreover, the regression that included the salient beliefs explained a significant
higher proportion of the variance than the base model, which contained the control
variables only.
To test Hypothesis 2a, which concerned the prediction of intentions using TRA, a base
model containing the control variables only was first run. Subsequently, attitude,
subjective norm and salient beliefs were entered into the regression. In support of
Hypothesis 2a, both attitude and subjective norm were significantly associated with
intentions (see Table 4, fourth column). As expected, salient beliefs in this model only
represent noise, since beliefs explain attitude, not intentions. Moreover, the regression that
included attitude and subjective norm explained a significant higher proportion of the
variance in intentions than the base model.
In order to test Hypothesis 2b, which concerned the effect of self-efficacy on intentions,
the measures of self-efficacy along with the control variables, salient beliefs, attitude and
subjective norm were entered into a regression with self-employment intentions as the
dependent variable. Only one of the four self-efficacy measures, opportunity recognition,
was marginally significantly related to intentions in the expected direction (see Table 4,
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 881

Table 4
Prediction of attitude toward self-employment, self-employment intentions and entry into self-employment
(n = 297)
Attitude Intentions Behavior
Base Beliefs Base TRA TPB Base TRA TPB
Control variables
Age .01 .05 .06 .06 .06 .04 .08 .08
Gender .04 .10y .01 .00 .01 .12* .13** .13**
Education .20** .14* .06 .06 .07 .09 .05 .05
Parental role model .03 .00 .03 .04 .04 .06 .05 .06
Serial entrepreneur .02 .20 .14* .14* .14* .03 .06 .06
Urban location .06 .06 .07 .10y .09 .04 .09y .10*
Productionz .00 .02 .05 .06 .07 .08 .04 .05
Tradez .09 .04 .07 .04 .04 .10 .05 .05

Salient beliefs
Autonomy .21*** .09 .10y .07 .06
Authority .15** .04 .04 .07 .07
Economic opportunity .26*** .04 .04 .11* .11*
Self-realization .21*** .02 .01 .01 .01

Attitude and subjective norm

Attitude .34*** .33*** .07 .09
Subjective norm .25*** .25*** .18 .16

Self-efficacy (PBC)
Opportunity recognition .10y .04
Investor relationships .02 .05
Risk taking .01 .06
Economic management .04 .07

Intentions
Self-employment intentions .62*** .62***
Multiple R .257 .484 .186 .497 .508 .238 .663 .671
Adjusted R 2 .040 .202 .008 .209 .210 .030 .410 .412
DR 2 – .168*** – .212*** .011 – .383*** .010
F value 2.549 7.244 1.291 6.595 5.363 2.164 14.694 11.912
Significance of F .011 .000 .248 .000 .000 .030 .000 .000
The coefficients reported are standardized b’s. Level of statistical significance: y indicates p V.10; *indicates
p V.05, **indicates V.01 and *** indicates p V .001 (2-tailed).
z
The reference group is service.

fifth column). The inclusion of the four self-efficacy measures did not explain a significant
higher proportion of the variance of intentions. Hence, Hypothesis 2b is not supported.
Hypothesis 3a concerned the ability of intentions to predict subsequent behavior. The
base model, which contained the control variables only, was compared with a model that
contained salient beliefs, attitude, subjective norm and intentions. The correlation between
intention and behavior was .63, indicating that intentions alone explain 39.7% of the
variance in the measure of behavior. As expected, the other variables included in the model
were essentially noise. These findings support Hypothesis 3a.
882 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

To test the final hypothesis, concerning the effect of self-efficacy on behavior, the four
self-efficacy measures were entered into a regression along with all the other variables.
None of the self-efficacy measures were significantly related to behavior and adding the
self-efficacy measures did not increase the variance explained significantly (see Table 4,
eighth column). Hence, Hypothesis 3b is not supported.5

8. Discussion

The findings presented here strongly support the theory of reasoned action. All four
salient beliefs that concern self-employment were significantly associated with attitude
toward self-employment. Attitude and subjective norm were significant predictors of self-
employment intentions. Intentions to become self-employed were strongly related to actual
entry into self-employment. The inclusion of salient beliefs to predict intention and
behavior represented essentially noise, since beliefs only have an indirect effect through
attitude. The findings, however, do not support the extension of the theory represented by
the theory of planned behavior. The four measures of entrepreneurial self-efficacy did not
add to the explanation of the variance of self-employment intentions or behavior.
The failure of the present survey to find support for the hypothesized relationships
between self-efficacy, intentions and behavior can be explained in at least two different
ways. First, if the behavior is under complete volitional control, self-efficacy (or PBC)
should not be expected to be strongly associated with intention or behavior (Ajzen, 1991),
especially when a self-predictive measure of intentions is applied (Armitage and Conner,
2001). In some countries, entry into self-employment may be a complete voluntary choice.
It has been suggested that entrepreneurs in Norway face a more munificent, less complex
and less hostile environment than entrepreneurs in other nations (Shane and Kolvereid,
1995). It is possible that entrepreneurial self-efficacy is more strongly related to entry into
self-employment in countries where entrepreneurs are faced with less favorable national
environments.
The present authors, however, hesitate to embrace the explanation of the findings
mentioned above. It is difficult to conclude that entrepreneurial self-efficacy is not related
to entry into self-employment when PBC previously has been found to be strongly
associated with entrepreneurial intentions in student samples from Norway (Kolvereid,
1996b), Russia (Tkashev and Kolvereid, 1999) and the USA (Krueger et al., 2000). A
more realistic explanation is therefore found in the operationalization of the constructs.
While the measures of salient beliefs, attitude and subjective norm applied under the
present circumstances concern self-employed specifically, the self-efficacy measures
included are general indicators of entrepreneurial self-efficacy. A more specific measure of
self-employment self-efficacy is likely to have produced better results (Bandura, 1997).

5
An additional test of Hypotheses 3a and 3b was carried out using logistic regression with full time
employment in the business (i.e. more/less than 35 h/week) as the dependent variable. The results obtained from
this analysis were practically identical to those reported in Table 4, but two additional independent variables
became significant in the equation (at p V .05): Education with a negative sign and risk-taking with a positive
sign.
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 883

Ironically, a more general measure of self-efficacy is also likely to have produced better
results. Previous research suggest that general self-efficacy is a common personality
characteristic among self-employed people compared to others and that self-employed
individuals report higher levels of self-efficacy than others (Bradley and Roberts, 2004).
However, specific measures are preferred, since only such measures can help us identify
critical aspects to be focused on in education and training of potential entrepreneurs.

8.1. Recommendations for future research

Future research should investigate the possibility that the potential of self-efficacy is
influenced by other variables, such as environmental munificence. It is also possible that
the term entrepreneurial self-efficacy is a too broad concept to accurately predict the
various tasks associated with entrepreneurship. Hence, there is a need to develop better
measures of self-efficacy designed to predict more specific entrepreneurial intentions and
behaviors associated with, for example, new business start-up, entry into self-employment,
business growth and wealth creation. This implies that different tasks normally associated
with entrepreneurship may require very different sets of skills. For example, a critical
competence in order to become self-employed may be the ability to work independently. A
critical competence in order to become an employer may be the ability to manage
employees. And critical competences in order for the business to grow may be the ability
to deal with investors and the ability to manage risk and uncertainty.

8.2. Conclusion

Some people may argue that trying to become self-employed is not the most interesting
entrepreneurial behavior. Entrepreneurship is ultimately about wealth creation, not about
the creation of a job for the founder. However, starting a business and entering into self-
employment is in most cases the first step of an entrepreneurial career. To really learn
about entrepreneurship, it is not sufficient to study success stories, we also need
knowledge about the process entrepreneurs go through on their way to success or failure.
The present research illustrates that the theory of reasoned action and the theory of
planned behavior are well suited for research into entrepreneurial behaviors. Salient beliefs
concerning self-employment determine attitude toward self-employment. Attitude and
subjective norm concerning self-employment determine intentions to become self-
employed, and self-employment intentions determine actual entry into self-employment.
The other variables included in the regressions were essentially noise. However, the
finding that male entrepreneurs are significantly more likely to enter into self-employment
than their female counterparts causes some concern and may be used to justify training
programs that target female entrepreneurs.
The findings reported here show a strong, highly significant relationship between
self-employment intention and entry into self-employment, suggesting that entry into
self-employment is an intentional behavior. Self-employment intention, in turn, is
determined by attitude and subjective norms concerning self-employment. Attitudes may
be altered in education and training programs. The finding that subjective norm is
positively associated with self-employment intentions suggest that we should not only
884 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885

consider the venture idea and the entrepreneur when evaluating or promoting a new
venture project. The entrepreneurs’ household and significant others should also be
taken into consideration. There is reason to believe that the measures of entrepreneurial
self-efficacy included in the present research were wrong, given that the dependent
variable was entry into self-employment and not growth. Future research should attempt
to identify specific competences that are associated with different kinds of entrepre-
neurial behavior. When we can demonstrate that a specific competence is central for
entrepreneurial progress, we should incorporate them in our entrepreneurship education
and training programs.

References

Ajzen, I., 1988. Attitudes, Personality and Behavior. Open University Press, Milton Keynes, UK.
Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50 (2),
179 – 211.
Ajzen, I., 2002. Perceived behavioral control, self-efficacy, locus of control, and the theory of planned behavior.
Journal of Applied Social Psychology 32 (4), 665 – 683.
Ajzen, I., Driver, B.L., 1992. Application of the theory of planned behavior to leisure choice. Journal of Leisure
Research 24 (3), 207 – 224.
Ajzen, I., Fishbein, M., 1980. Understanding Attitudes and Predicting Social Behavior. Prentice Hall, Englewood
Cliffs, NJ.
Anna, A.L, Chandler, G.N., Jansen, E., Mero, N.P., 2000. Women business owners in traditional and non-
traditional industries. Journal of Business Venturing 15 (3), 279 – 303.
Armitage, C.J., Conner, M., 2001. Efficacy of the theory of planned behavior: a meta-analytic review. British
Journal of Social Psychology 40 (4), 471 – 499.
Bandura, A., 1986. Social Foundation of Thought and Action: A Social Cognitive Theory. Prentice Hall,
Englewood Cliffs, NJ.
Bandura, A., 1997. Self-Efficacy. The Exercise of Control. W.H. Freeman and Co., New York.
Bandura, A. 2001. Guide for constructing self-efficacy scales (revised). Available from Frank Pajares, Emory
University.
Betz, N.E., Hackett, G., 1998. Manual for the occupational self-efficacy scale. (Downloaded 31 October 2001).
Available from http://seamonkey.ed.asu.edu/gail/occse1.htm.
Bird, B.J., 1992. The operation of intentions in time: the emergence of a new venture. Entrepreneurship Theory
and Practice 17 (1), 11 – 20.
Birley, S., Westhead, P., 1994. A taxonomy of business start-up reasons and their impact on firm growth and size.
Journal of Business Venturing 9 (1), 7 – 31.
Bradley, D.E., Roberts, J.A., 2004. Self-employment and job satisfaction: investigating the role of self-efficacy,
depression and seniority. Journal of Small Business Management 42 (1), 37 – 58.
Carter, S., Ram, M., 2003. Reassessing portfolio entrepreneurship: towards a multi-disciplinary approach. Small
Business Economics 24 (4), 371 – 380.
Carter, N.M., Gartner, W.B., Reynolds, P.D., 1996. Exploring start-up event sequences. Journal of Business
Venturing 11 (3), 151 – 166.
Carter, N.M., Gartner, W.B., Shaver, K.G., Gatewood, E.J., 2003. The career reasons for nascent entrepreneurs.
Journal of Business Venturing 18 (1), 13 – 39.
Chen, C.C., Greene, P.G., Crick, A., 1998. Does entrepreneurial self-efficacy distinguish entrepreneurs from
managers? Journal of Business Venturing 13 (4), 295 – 316.
Delmar, F., Davidsson, P., 2000. Where do they come from? Prevalence and characteristics of nascent
entrepreneurs. Entrepreneurship and Regional Development 12 (1), 1 – 23.
Dennis Jr., W.J., 1997. More than you think: an illusive estimate of business entries. Journal of Business
Venturing 12 (3), 175 – 196.
 L. Kolvereid, E. Isaksen / Journal of Business Venturing 21 (2006) 866–885 885

DeNoble, A.F., Jung, D., Ehrlich, S.B., 1999. Entrepreneurial self-efficacy: the development of a measure and its
relation to entrepreneurial action. In: Reynolds, P.D., Bygrave, W.D., Manigart, S., Mason, C.M., Meyer,
G.D., Sapienza, H.J., Shaver, K.G. (Eds.), Frontiers of Entrepreneurship Research. Babson College,
Wellesley, MA, pp. 73 – 87.
Douglas, E.J., Sheperd, D.A., 2002. Self-employment as a career choice: attitudes, entrepreneurial intentions, and
utility maximization. Entrepreneurship Theory and Practice 26 (3), 81 – 90.
Gundry, L.K., Welch, H.P., 2001. The ambitious entrepreneur: high growth strategies of women-owned
enterprises. Journal of Business Venturing 16 (5), 453 – 470.
Hair Jr., J.F., Anderson, R.E., Tatham, R.R., Black, W.C., 1998. Multivariate Data Analysis, 5th ed. Prentice Hall,
Upper Saddle River, NJ.
Katz, J., 1992. A psychological cognitive model of employment status choice. Entrepreneurship Theory and
Practice 17 (1), 29 – 37.
Kolvereid, L., 1996a. Organizational employment versus self-employment: reasons for career choice intentions.
Entrepreneurship Theory and Practice 20 (3), 23 – 31.
Kolvereid, L., 1996b. Prediction of employment status choice intentions. Entrepreneurship Theory and Practice
21 (1), 47 – 56.
Krueger, N.F., Carsrud, A.L., 1993. Entrepreneurial intentions: applying the theory of planned behavior.
Entrepreneurship Theory and Practice 5 (3), 315 – 330.
Krueger, N.F., Reilly, M.D., Carsrud, A.L., 2000. Competing models of entrepreneurial intentions. Journal of
Business Venturing 15 (5/6), 411 – 432.
Madden, T.J., Ellen, P.S., Ajzen, I., 1992. A comparison of the theory of planned behavior and reasoned action.
Personality and Social Psychology Bulletin 18 (1), 3 – 9.
Magnussen, R.Å. 1997. Etableringssukess: En funksjon av etablereren eller etableringsprosessen? (Entrepre-
neurial success: a function of the entrepreneur or the start-up process?) Unpublished Master Thesis, Bodø
Graduate School of Business, Norway.
Reynolds, P., Miller, B., 1992. New firm gestation: conception, birth, and implications for research. Journal of
Business Venturing 7 (5), 405 – 417.
Scheinberg, S., MacMillan, I.C., 1988. An 11-country study of motivations to start a business. In: Kirchhoff,
B.A., Long, W.A., McMullan, W.E., Vesper, K.H., Wetzel, W.E. (Eds.), Frontiers of Entrepreneurship
Research. Babson College, Wellesley, MA, pp. 669 – 687.
Shane, S., Kolvereid, L., 1995. National environment, strategy, and new venture performance: a three country
study. Journal of Small Business Management 33 (2), 37 – 50.
Shane, S., Kolvereid, L., Westhead, P., 1991. An exploratory examination of the reasons leading to new firm
formation across country and gender. Journal of Business Venturing 6 (6), 431 – 446.
Shook, C.L., Priem, R.L., McGee, J.E., 2003. Value creation and the enterprising individual: a review and
synthesis. Journal of Management 29 (3), 379 – 399.
Statistics Norway, 2004. (Online) (Downloaded 17 March 2004). Available from http://www.ssb.no.
The Concise Oxford Dictionary. Oxford University Press, Oxford, UK.
Tkashev, A., Kolvereid, L., 1999. Self-employment intentions among Russian students. Entrepreneurship and
Regional Development 11 (3), 269 – 280.
Vesper, K.H., 1990. New Venture Strategies. Prentice Hall, Englewood Cliffs, NJ.
Wennekers, A.R.M., Thurik, A.R., 1999. Linking entrepreneurship to economic growth. Small Business
Economics 13 (1), 27 – 55.
Westhead, P., Wright, M., 1998a. Novice, serial and portfolio founders: are they different? Journal of Business
Venturing 13 (3), 171 – 204.
Westhead, P., Wright, M., 1998b. Novice, portfolio, and serial founders in rural and urban areas. Entrepreneurship
Theory and Practice 22 (4), 63 – 100.
Westhead, P., Ucbasaran, D., Wright, M., 2005. Experience and cognition: do novice, serial and portfolio
entrepreneurs differ? International Small Business Journal 23 (1), 72 – 98.