A++ Work on time, quality work

CareerSuccess.pdf

Home >Science homework help >A++ Work on time, quality work

Drivers of freelance career success

ARJAN VAN DEN BORN1* AND ARJEN VAN WITTELOOSTUIJN1,2,3 1School of Economics, Utrecht University, Utrecht, The Netherlands 2University of Antwerp, Antwerp, Belgium 3Tilburg University, Tilburg, The Netherlands

Summary Recent evidence shows that the frequently proclaimed collapse of the traditional career model is actually not supported by job tenure data. This paper argues that the observed stability of job tenure might be explained by an increasing number of shamrock organizations. This organizational form has three types of workers: core employees, professional freelancers, and routine workers. In such an organization, two very different career models coexist. The organization largely determines the career of the core employee, whereas the individual essentially shapes that of the professional freelancer. This paper studies extensively the career of this second group: the professional freelancer, a growing phenomenon in many developed countries but not yet the focus of many career studies. We develop a freelance career success model on basis of the intelligent career framework augmented by insights from literature on entrepreneurship. Data are from a web survey with responses from about 1600 independent professionals in the Netherlands, in combination with 51 in-depth interviews. We provide two main contributions. First, we report findings from the first large-scale quantitative study into freelance career success. Second, this study enhances our understanding of the success of the modern career by building bridges between career and entrepreneurship literatures. We conclude that the external environment in which an individual freelancer operates is the most important factor determining career success. The study therefore suggests that more work needs to be performed on the relationship between the environment and individual career success. Copyright © 2012 John Wiley & Sons, Ltd.

Keywords: freelancers; self-employed; career success; portfolio workers

Introduction

Is the traditional career of the 20th century becoming rare? Since the 1980s, many authors have argued that the responsibility for the career increasingly resides in the individual (Arnold, 2001; Raabe, Frese, & Beehr, 2006) and transcends any employer (Arthur, Khapova, & Wilderom, 2005). But recently, Rodrigues and Guest (2010) showed for multiple countries that the collapse of the traditional career model is not supported at all by actual job tenure data. Does this imply that the commonly accepted view of people having increasingly multiple-organization careers is a myth? This paper argues that the observed stability of job tenure may be caused by the emergence of a new type of worker: the skilled independent professional. This new type of worker contracts out her or his skills to various organizations (Barley & Kunda, 2004; Barley, Kunda, & Evans, 2002; Connelly & Gallagher, 2004; Kirkpatrick & Hoque, 2006). Avant la lettre, Handy (1985) coined the term “portfolio worker” for those workers who create a portfolio of work

for themselves. In current times, terms such as freelancer, independent professional, or contractor are more frequently used. Although it is hard to estimate the exact number of these workers, it is clear that their incidence is growing since the 1980s. According to Arum and Müller (2004, p. 1), “[s]elf-employment can no longer be dismissed as an economic

*Correspondence to: Jan van den Born, School of Economics, Utrecht University, Utrecht, The Netherlands. E-mail: [email protected]

Revised 18 January 2012, Accepted 22 January 2012

Journal of Organizational Behavior, J. Organiz. Behav. 34, 24–46 (2013) Published online 9 March 2012 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/job.1786Research

A rticle

activity on the verge of withering away in response to processes of capital accumulation or in competition with large firms.” Marler, Barringer, and Milkovic (2002) showed that these boundaryless workers can be distinguished from traditional “temps” by their preference for temporary work in combination with their high level of skill and experience. The claim is that these “contractors of choice” are especially likely to report positive outcomes about job and career satisfaction (Anderson, 2008; Ajayi-Obe & Parker, 2005; Benz & Frey, 2008; Guest, 2004; Guest & Clinton, 2006). As Blanchflower (2004, p. 21) summarized this argument, “These self-employed work under a lot of pressure, find their work stressful and come home exhausted. How- ever, they are especially likely to say they have control over their lives as well as being highly satisfied with their lives.” The increase of the number of independent knowledge workers fits nicely with a prediction of Handy

(1989), who argued that the organizations of the future will have three types of workers: (i) professional employees representing the core competencies of the organization; (ii) professional freelancers contractually hired on a project-by-project basis; and (iii) a contingent workforce doing routine jobs. Handy introduced the term shamrock organization for this new mode of organizing. Growth in the density of this shamrock type of organizations could explain why overall job tenure does not decrease, because the freelancer is officially not an employee of the organization but is self-employed. Although these free agents hop from one organization to the next, they never officially change their employer. In a way, they are employees turned into entrepreneurs. The shamrock organization model is associated with the coexistence of two career types. On the one hand, the

organization largely determines the career of the professional core. Hence, Lips-Wiersma and Hall (2007) correctly argued that the role of the organization is not over. On the other hand, the individual largely determines the career of the growing number of independent professionals—or freelancers. Consequently, modern career concepts, such as the protean career model (Hall, 1976, 2002) and the intelligent career framework (Parker, Khapova, & Arthur, 2009), apply nicely to this new type of workers. This paper studies extensively the determinants of the career success of this second group: the professional freelancer, a growing phenomenon in many developed countries but not yet the focus of many career studies.

Building a Freelance Career Success Model

Process

Freelancers can be considered as a hybrid of employees and entrepreneurs. On the one hand, they are employees because they are almost always hired by (large) firms to work for a period selling nothing else but their intangible professional knowledge, which is different from other entrepreneurs and self-employed selling tangible products to customers. On the other hand, they are entrepreneurs because they work for their own risk and reward without any organizational guarantee or support. Any freelance career success model should therefore build on insights from both the modern career literature and the literature on entrepreneurship. To develop a testable model, we first created a draft career success model based on the available academic literature. This first draft model consisted of a number of building blocks that included variables that might predict freelance career success, with five to 10 variables for every building block. Second, we piloted this first conceptual model with experts from the field by interview- ing 51 practitioners (i.e., 33 freelancers and 18 outside experts from intermediary agencies and professional associations). On the basis of their comments, we created a final freelance career success model in our third step by (i) adding groups of variables that were considered to be missing and by (ii) converting our variables into operationalized measures. In the fourth step, we developed an online survey and administered it to test this final freelance career success model.

DRIVERS OF FREELANCE CAREER SUCCESS 25

Individual career perspective

Our first draft model was based both on the career and entrepreneurship literatures, starting from the intelligent career framework (Parker et al., 2009) as the core. We consulted studies in the entrepreneurship literature to add variables to our core model that may be related to entrepreneurial success. The freelance career can be seen as quite similar to the boundaryless career. DeFillippi and Arthur (1996) defined boundaryless careers as “sequences of job opportunities that go beyond the boundaries of single employment setting.” A freelancer is probably the archetypical job hopper going from one project and employer to the next, never staying for very long in a single organization. Therefore, the boundaryless career concept provides a good starting point to build a freelance career success model. The boundaryless career concept was already developed in the 1970s at the Massachusetts Institute of Technology (Tams & Arthur, 2010). However, the boundaryless career perspective was really developed further in the 1990s, especially through the efforts of DeFillippi and Arthur (1994, 1996) and Arthur and Rousseau (1996). In the mid-1990s, Arthur, Claman, and DeFillippi (1995) pointed to strong links between Quinn’s (1992) intelligent enterprise and the boundaryless career concept. They argue that the creation of the intelligent enterprise requires intelligent careers, which are built upon principles from the boundaryless career concept. This intelligent career concept evolved into the intelligent career framework (Parker et al., 2009). The intelligent career framework includes three interrelated classes of variables, referred to as “ways of knowing,”

that are argued to predict career development. The first way of knowing is knowing why, which involves career motivation, personal meaning, identity, and personality. Knowing why is associated with an individual’s capability to understand herself or himself, to explore different possibilities, and to adapt to constantly changing work circumstances. The second way of knowing is knowing how, which reflects career-relevant skills and job-related knowledge. Knowing how is closely related to established ideas on individual knowledge, skills, and abilities (Schneider & Konz, 1989). The third way of knowing, knowing whom, involves relevant personal and business networks. This has to do with career-related networks and contacts, including business relationships and personal connections (Parker & Arthur, 2000). These three classes of variables are not independent; there are strong links between these concepts, with a range of different theories explaining these linkages (Parker et al., 2009). Despite the theoretical attractiveness and practical relevance of the intelligent career framework, much empirical

work continues to focus on career success as evaluated from an organizational perspective (e.g., in terms of organizational position and promotion). This seems outdated because hierarchies are continuously flattening (Littler, Wiesner, & Dunford, 2003) and because external labor markets generate an increasing influence over today’s employment landscape (Cappelli, 1999). Arthur et al. (2005) therefore called for further rapprochement between career theory and empirical research. As the employment landscape is changing, career research should reflect the “new deal” that views the career actor as concerned more with individual rather than organizational goals and which involves the kind of “meta-competencies” that allow for easier mobility between successive employers or temporary contracts.

Entrepreneurial perspective

The intelligent career framework is originally based on organizational models of firm competencies and firm success (DeFillipi & Arthur, 1994, p. 308). This is why the intelligent career framework offers an appropriate platform for developing our draft freelance career success model. We cross the same bridge as DeFillipi and Arthur did and review empirical results reported in studies of firm success in search for variables to be added to the model. Adding such business aspects to our freelance career success model will create a better fit with freelancing challenges and dilemmas. After all, a modern freelancer is a self-employed entrepreneur as well and not only a “boundaryless employee.” Hence, we are especially interested in research into small entrepreneurs such as the self-employed and small businesses and not so much in large businesses where the founder influence is limited (Boone, De Brabander, & van Witteloostuijn, 1996) and other qualities are needed to be successful (Scott & Bruce, 1987).

26 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

In the succeeding texts, we highlight the core drivers of entrepreneurial success as identified by empirical research into the self-employed and small businesses (see Parker, 2004, for an overview). First, the human capital of the entrepreneurial founder has a strong positive impact on firm performance (e.g., Bosma, van Praag, Thurik, & de Wit, 2004; Pennings, Lee, & van Witteloostuijn, 1998). Being well educated or highly experienced, for instance, contributes to the success of entrepreneurial ventures. Second, much evidence relates to the positive effect of social capital on firm success (e.g., Bosma et al., 2004; Chiesi, 2007; Hoanga & Antoncic, 2003; Witt, 2004). Inherently, entrepreneurship is a process of building bridges in a network, and hence of developing and maintaining social capital, which provides broad and early access to information, and offers control over the distribution and interpretation of information. In the freelancing context, social capital can be expected to (i) generate a broad base of referrals; (ii) help the freelancer identify promising opportunities; and (iii) increase the probability that the freelancer knows how to pitch a project. Third, the study of the impact of what may be coined personal capital on entrepreneurial success has a long tradition. Entrepreneurs are argued to “engage the energies of everyone,” “involve many people inside and outside the organization,” “create and sustain networks of relationships,” and “make the most of the intellectual and other resources people have to offer” while “helping those people to achieve their goals as well” (McMillan & Gunther-McGrath, 2000, p. 3). These qualities highlight the desirability of a specific type of personality or other personal features, such as self-insight and leadership style. In their overview of the literature on entrepreneurial personality, Amit, Glosten, and Muller (1993) suggested

that the four personality traits most commonly associated with self-employment are as follows: (i) need for achievement (McClelland, 1965); (ii) internal locus of control (Sexton & Bowman, 1986); (iii) above-average risk-taking propensity (Brockhaus, 1980); and (iv) tolerance for ambiguity (Frenkel-Bruswik, 1948). Note, however, that although research has shown that personality traits are important for entrepreneurial success, personality traits have typically produced very weak relationships with entrepreneurial performance (Begley & Boyd, 1987; Low & MacMillan, 1988; Parker, 2004; Stam et al., 2012). The search for a psychological explanation for business success has led to the development of tailor-made multifaceted “entrepreneurial” personality traits, such as Chen, Greene, & Crick’s (1998) entrepreneurial self-efficacy.

Expert piloting

After the creation of our draft freelance career success model, on the basis of the review of the literature, we piloted this draft in interviews with practitioners. These interviews were semi-structured. They started with open-ended questions (e.g., How would you define freelance career success? and What are the key factors that determine freelance success?), followed by a series of partly closed and partly open questions on the overall draft model and each variable in the model (e.g., Do you think this is an important determinant of freelance career success or not, and why do you think so?). Our interviewees defined two major topics that they considered to be missing. Additionally, they expressed a variety of smaller suggestions about missing variables and suggested many improve- ments in defining variables and operationalizing measures. One example is partner support. We included this variable in the final model, as many interviewees convincingly argued that a freelancer could not be successful without the proper support of her or his partner. The first and largest deficiency in our draft model was the lack of reference to the external market. All 51

interviewees noticed that we ignored market factors in our initial model design. This is especially important because we know that the environment is crucial for firm performance (Porter, 1980). A stylized fact in the business literature is that organizational and industry variables dominate over individual-level variables as drivers of venture success (e.g., McGahan & Porter, 1997; Sandberg & Hofer, 1987). In contrast, the external product or service market in which the employer organization operates often plays a minor role in the employee career literature. In fact, career studies generally neglect such forces. The studies that link career outcomes with characteristics of internal labor markets, which were quite popular in the 1980s (e.g., Baron, Davis-Blake, & Bielby, 1986), are perhaps the most relevant for our study. However, internal labor markets are very different from their external product or service

DRIVERS OF FREELANCE CAREER SUCCESS 27

market counterparts. In recent years, the issue of context has gained, again, prominence in career research (Arnold & Cohen, 2008; Cohen & Mallon, 1999; Mayrhofer, Meyer, & Steyrer, 2007). But as far as we know, quantitative empirical studies that estimate the impact of the external product or service market on career factors or outcomes are absent. Clearly, our expert and freelance interviewees all consider the external market to be a very important driver of

freelance career success. Cyclical and structural market conditions of demand and supply, transparency, industry structure, and industry institutional arrangements are likely to largely determine career outcomes and may co- determine key success factors. For instance, in the Netherlands, the traditional free professions (i.e., accounting, law, notary, and medicine) are associated with strict education and learning requirements that effectively regulate and limit supply and necessitate pre-entry and post-entry investments in human capital for all independent professionals in the market (for an example, see Maijoor & van Witteloostuijn, 1996). In other markets (e.g., interpreters and some technical occupations), there are only a limited number of companies that employ the specialized services of freelancers, implying an oligopsony that drives fees down (Bhaskar & To, 2003). Our interviewees pointed out that some markets had only a limited number of potential clients; in these markets, managing reputation was crucially important to retain business. In other markets, the number of potential clients was much larger; in such markets, creating visibility by branding, marketing, and networking was said to be of much greater importance than reputation management. The second major aspect that was lacking in our initial freelance career success draft framework, according to our

interviewees, was business strategy. The essence of the strategic management literature is the assumption, for which there is ample evidence (e.g., McGahan & Porter, 1997), that strategy matters, which is also echoed in the entrepre- neurship tradition (e.g., Parker, Storey, & van Witteloostuijn, 2010). Hence, we decided to add a number of business strategy variables in our model to reflect the fact that a freelancer is both an individual employee and an entrepre- neurial “organization,” in the latter capacity pursuing business strategies just like any other business venture.

Crafting a Testable Freelance Career Success Model

Model

In Figure 1, we summarize our final freelance career success research model. The central concepts of the intelligent career framework remain largely in place, where human capital can arguably be seen as representing knowing how,

(Individual) career

literature

Draft freelance career model

Entrepre- neurship literature

Interviews (51)

Final detailed model

Online survey

I II III IV

Analysis

Figure 1. Stepwise approach of this study

28 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

social capital as reflecting knowing whom, and both motivation and personality measures, which can jointly be coined personal capital, as mirroring knowing why. However, compared with the intelligent career framework, our freelance career success model is extended by adding two clusters of variables that represent the external environment and business strategy. This is precisely the consequence of the hybrid nature of the freelance status, implying that the individual operates as both an employee and an entrepreneur. Of course, an empirical model is never complete and is therefore always restricted in one way or the other.

Inevitably, developing a measurement model, even one that is designed to be so comprehensive as ours, implies that decisions have to be made about what not to take on board. For instance, not taking any variables into consideration that refer to organizational support—a decision based on the observation that, in the end, freelancers are not employees linked to a single organization—means that hypotheses regarding this aspect cannot be tested. As a rule, the selection of variables in our final model is based on the academic literature and suggestions from practitioners, bounded by practical considerations about the maximum length of a survey. In this context, the exploratory nature of our study is key. That is, the series of hypotheses introduced in the succeeding texts is very broad in an attempt to assess which potential drivers of career success are more important than others. Our aim is not to test any specific theory but rather to broadly explore what does and what does not matter in this new freelance world. In so doing, for instance, our study offers the opportunity to evaluate the value added of variables derived from the entrepreneurship literature, such as the market environment and business strategy, relative to those inspired by the career literature. A final remark relates to our choice of career success measures. We used two measures of success: objective

career success (OCS) and subjective career success (SCS). Moreover, our model explicitly incorporates the interrelationship between OCS and SCS. The distinction between OCS and SCS is important in the freelancer context because independent professionals self-select into self-employment for a variety of reasons: not only monetary motives but also arguments relating to autonomy, flexibility, and work–life balance are potentially impor- tant. Indeed, the extant literature emphasizes this reinforcing feedback loop, where career success not only produces happiness but happiness also enhances further career success (Boehm & Lyubomirsky, 2008). Where we believe that we can expect different effects of our independent variable upon OCS vis-à-vis SCS, we will develop two separate sub-hypotheses; if we see no reason to expect a differential effect, we simply refer to freelance career success in our hypotheses.

Hypotheses

We introduce our hypotheses by systematically discussing all Figure 1’s clusters of variables. We start with the well-established standard human capital (knowing how) determinant of success, as this set of capabilities is strongly supported by the literature on career success as well as that on entrepreneurial performance. Moreover, on the basis of our interviews with experts and freelancers, we have learned that freelancers are very aware of the need to continuously develop their human capital. They are always scanning knowledge sources (e.g., internet, books, and magazines) to be up-to-date on the latest industry trends (Barley & Kunda, 2004).

Hypothesis 1: Human capital is positively related to freelance career success.

Social capital (knowing why) is another cluster of variables that figures prominently in both the career and entrepreneurship literatures. In traditional career research, measures of organizational sponsorship, which reflects social capital in the context of a traditional career, often represent social capital. The meta-analysis of Ng, Lillian, Eby, Sorensen, and Feldman (2005) shows that organizational sponsorship variables (e.g., career support, training and skill development opportunities, and mentoring) demonstrate a weak relationship with salary and promotion and a strong relationship with career satisfaction. As organizational support is largely absent for freelancers, with

DRIVERS OF FREELANCE CAREER SUCCESS 29

the notable exception of support of former employers (Cohen & Mallon, 1999), most organizational sponsorship measures are not very applicable to freelancers. One concept that is closely related to organizational sponsorship is applicable to freelancers, though: agency sponsorship. Support of staffing agencies is very important for freelancers (Barley & Kunda, 2004). Many freelancers develop strong relations with external staffing agencies, which are important partners for freelancers in their search for new assignments. In the entrepreneurship literature, network characteristics commonly measure social capital (e.g., Chiesi, 2007;

Witt, 2004). Such measures of social capital are applicable in the freelancer context, too, as a strong network is required to provide a steady flow of assignments, especially in the absence of supporting agencies. Hence, the characteristics of these networks, such as network size and tie strength (Granovetter, 1973), offer a second perspective to measure the social capital of freelancers. A third perspective regarding social capital inspired by the entrepreneurship literature involves the amount of support that an entrepreneur receives. Such support could come not only from a business club or business network, such as Lions and Rotary (Barbieri, 2003; Davidson & Honig, 2003), but also from the freelancer’s partner. Our interviewees strongly argued that freelancing is stressful and that partner support is therefore critical.

Hypothesis 2: Social capital is positively related to freelance career success.

To measure knowing why, we focus on personality traits and motivational drivers. In selecting personality- related variables that can represent this personal capital of freelancers, we largely followed the example of Eby, Butts, and Lockwood’s (2003) and used career insight, pro-activeness, and openness to represent the freelancers’ personal capital (Figure 2). These measures of personal capital are distinct from the traditional entrepreneurial personality measures such as internal locus of control, risk-taking propensity, and need for achievement. We have three main reasons to do so. First, we needed to limit the number of personality-related variables to a maximum of three because personality measures often encompass lengthy item lists that are very time-consuming to complete. Second, these traditional entrepreneurial personality measures have never been shown to be very powerful predictors of performance in the entrepreneurship literature (Stam et al., 2012), and there is no reason to assume that these personality characteristics are more important in a freelance context, where professional skills are arguably at least as important as entrepreneurial capabilities. Third, our interviewees

Subjective Succes

(Satisfaction)

Objective Succes

(Revenue)

Business strategies

Market factors

Social Capital

(Knowing -whom)

Human Capital

(Knowing -how)

Personal Cap & Motivation

(Knowing -why)

Figure 2. The final freelance career model

30 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

emphasized the importance of career insight, pro-activeness, and openness. For instance, career insight involves knowing which assignments to accept and which assignments to reject to build a clear profile and a strong résumé, which was frequently reported as crucial to career success. Another important aspect of knowing why is motivation. Not all freelancers pursue a freelance career because

of the same underlying motivational driver. Monetary reasons reflect, of course, a viable and often expressed argument, but intrinsic motivation as reflected in the wish for increased autonomy, flexibility, and work–life balance may well be even more important. It is highly likely that the type of motivation influences the outcome of a career. Heslin (2003) suggested that career motivations and goals are the main yardsticks against which to evaluate career success. From our interviews with freelancers, we learned that a trade-off might be at play here. That is, those freelancers motivated by non-financial objectives related to autonomy, flexibility, and work–life balance emphasize SCS over OCS yardsticks and vice versa, where subjective yardsticks involve “soft” non- monetary satisfaction and objective yardsticks “hard” economic achievements.

Hypothesis 3: Career insight, pro-activeness, and openness are positively related to OCS and SCS.

Hypothesis 4a: Autonomy, flexibility, and work–life balance are negatively related to OCS.

Hypothesis 4b: Autonomy, flexibility, and work–life balance are positively related to SCS.

As said, we added business strategy to our freelance career success model because of the suggestions made by our interviewees and for theoretical reasons as strategy is the linking pin between the external environment and internal capabilities (Venkatraman & Camillus, 1984). As far as we know, studies on the business strategies of freelancers are missing altogether. We decided to apply the well-established resource-based view of the firm so as to develop a typology of freelance business strategies (Barney, 1991; Maijoor & van Witteloostuijn, 1996). The resource-based view argues that sustainable competitive advantage is derived from resources that are valuable and rare and which are protected from imitation, transfer, or substitution. This implies that freelancers must try to construct a rare, valuable, and protected combination of resources (knowledge, skills, abilities, networks, etc.), which together create a sustainable competitive advantage. Such a strategy is very hard to pursue, as even complex combinations of resources are subject to imitation and substitution in the low barrier world of freelancing. Therefore, freelancers tend to reveal a strong tendency to continuously develop themselves to stay ahead of competition. We built our business strategies upon the insights of Ostgaard and Birley (1996), who distinguish the following

six strategies for businesses with less than 50 employees: marketing differentiation, product innovation, broad market/product range, many distribution channels, growth through capital, and differentiation through quality. These small-business strategies were discussed with our interviewees, which resulted in a somewhat adopted set of seven freelance business strategies. In our final set of freelance business strategies, Ostgaard and Birley’s (1996) small-business strategies of many distribution channels and growth through capital were replaced by (i) price leadership, (ii) industry specialization, and (iii) product specialization. Subsequently, on the basis of the theories of Porter (1980), Baum, Locke, and Smith (2001), and Boone and van Witteloostuijn (2004), not all these resulting seven freelance business strategies can be expected to improve freelance career success. That is, only those business strategies that aim at low cost, focus, or differentiation are hypothesized to lead to freelance career success by developing a competitive position in a freelancer’s niche. The other strategies are “stuck-in-the-middle” and doomed to produce failure (Porter, 1980, p. 42).

Hypothesis 5: Low-cost, focus, and differentiation strategies are positively related to freelance career success.

On the basis of insights from the management literature, we hypothesize that features of the external market are the most important determinant of OCS. In contrast, on the basis of earlier empirical results from the traditional

DRIVERS OF FREELANCE CAREER SUCCESS 31

career literature (Ng et al., 2005) and related to the arguments earlier, we hypothesize that personal capital has the strongest impact on SCS.

Hypothesis 6a: Market features dominate all other factors in explaining OCS.

Hypothesis 6b: Personal capital dominates all other factors in explaining SCS.

Method

Participants

We collected data using a customized Internet survey in the Netherlands, which ran in the period from 1 January to 1 April 2008. At the time of the research, the Netherlands is still experiencing an extensive period of robust economic growth, although doubts about the economic outlook were starting to emerge in the financial press as the sub-prime banking crises in the U.S. already started to affect consumer confidence and the stock exchange, albeit not yet in the Netherlands. A variety of professional associations that represent freelancers in a range of occupations or professions supported the Internet survey. In total, 19 professional associations agreed to inform their members about the Internet survey and to send their membership a link to the online survey. As a result, a total of about 40 000 independent professionals received an email with information on the survey, including a hyperlink to the survey. Of this population, 3146 persons clicked on this hyperlink to an introduction page explaining the goal of the research and the time needed to complete the questionnaire. From this number, 1981 individuals actually started to work on the online survey. In the end, 1612 respondents (88 per cent of those who started) finished the entire survey of 24 pages and completed 99 questions. Our sample might reflect an overrepresentation of individuals who are members of the sponsoring profes-

sional organizations. However, the large number of respondents who copied the web link of the questionnaire so as to forward this link to their friends and colleagues somewhat mitigated this effect. Moreover, a few very broad Internet sites offered their services by providing information to unspecified independent professionals, putting the link to the online questionnaire high up on their Internet pages. All in all, less than 50 per cent of the respondents in this study are members of a professional organization. Moreover, we cannot exclude the distinct possibility that some professions are more eager to respond to this type of requests than other professions, which implies another source of possible overrepresentation of specific professions. In the succeeding texts, we present the occupational distribution of our sample. Of course, this distribution must be kept in mind while interpreting the evidence.

Variables and measures

We have two measures of career success. We assessed SCS using a slightly modified version of the widely used (e.g., Boudreau, Boswell, & Judge, 2001; Judge & Ferris, 1993; Seibert, Kraimer, & Liden, 2001) career satis- faction scale of Greenhaus, Parasuraman, and Wormley (1990). Heslin (2005) earlier used this 6-item scale. Item examples are How satisfied are you with the income you have attained, relative to your career aspirations? and How satisfied are you with the autonomy you have attained, relative to your career aspirations? We measured OCS using revenue (logged, to produce normality). In most research into firm success, a definition of income is used instead of revenue. In the freelancer context, however, revenue is superior to income as an OCS measure for two critical reasons. On the one hand, income is subject to tax manipulation. As a consequence, although in

32 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

cross-industry entrepreneurial analysis income is often used, this is arguably a very noisy measure in practice. On the other hand, usually, revenue cannot be used as a measure for entrepreneurial success, because the cost structure of one entrepreneur (e.g., a car dealer with a large stock) can be totally different for another entrepre- neur (e.g., an owner of an industrial company). However, because all respondents of our survey are highly educated knowledge professionals, they all have similar (and relatively limited) costs. In the Netherlands, they all need a laptop, a (mobile) telephone, an Internet connection, and a car to perform their profession. They do not have cost of capital or extreme procurement costs that are dominant in companies dealing with tangible goods. Of course, up-market professionals have higher absolute costs than professionals working in down-market segments. In our sample, however, the costs reported by the respondent-freelancers are consistently between 20 and 30 per cent of total revenue. Our interviewees confirmed this percentage, and it is in line with the figures noted in more popular research on the Internet. We do not think revenue that is a perfect measure, as some professionals may have higher relative costs than others, but we do believe that, in the case of the independent professionals, revenue is an appropriate measure of monetary success that is less biased than income. To measure human capital, we used a number of variables. First, to measure experience, we included total work

experience and total freelance experience. We measured these by using the same scale, with eight answer categories: 0–6 months, 7–12 months, 1–2 years, 3–5 years, 6–10 years, 11–15 years, 16–20 years, and 21 years or more. Second, we asked for the highest educational level of the freelancer, including the year of graduation, which is a scale running from primary school (“Lagere school”) to a post-university degree (“post-doctoral”). Third, on the basis of van der Heijden (2006), we asked the respondents to estimate the number of training days in the last two years. We split this up in three types of training: training in core skills, training in developing new skills, and training in adjacent or supporting skills (e.g., administration or personal effectiveness). To measure social capital, we used network characteristics and a few proxies for support. First, to measure

the network characteristics, we first asked respondents to estimate the size of their personal and business networks. The network definition given follows Witt’s (2004). We defined strong contacts in line with Granovetter (1973) as those personal contacts that are family or friends. On the basis of Barbieri (2003) and Davidson and Honig (2003), we divided the network in higher and lower value contacts. We defined higher value network ties as contacts at the senior management level or above (i.e., director, vice-president, senior vice-president, or CEO). Second, we included a question about how much time the respondent invested in networking activities. On the basis of Aldrich and Reese (1993), we asked respondents to indicate the time per week (in hours) that they engaged, on average, in networking, with reference to Forret and Dougherty’s (2001) list of networking activities. Specifically, we asked individuals to rate how frequently they engaged, on a 6-point scale ranging from never to almost every day, in giving out business cards; sending thank you notes or gifts to people who have helped you in your work or career; sending cards, newspaper clippings, faxes, or emails to keep in touch; phoning business contacts to keep in touch; and having lunch with business relations. Third, to measure the amount of support the freelancer received, we asked whether the freelancer was a member of a business club or business network such as Lions and Rotary (Barbieri, 2003; Davidson & Honig, 2003). Fourth, we measured the influence of agency support by asking for the number of employ- ment agencies a freelancer was registered with as well as the frequency of their communications with these agencies. Fifth and finally, we assessed the support of the partner by using Greenhaus and Friedman’s (2000) 4-item measure, using a 5-point Likert-type scale. To measure personal capital, we included personality as well as motivational factors. In selecting variables,

we largely followed the example of Eby et al. (2003), using their measures of career insight, pro-activeness, and openness. We included Noe, Noe, and Bachhuber’s (1990) 6-item measure to assess career insight. We measured the pro-active personality trait by using Kickul and Gundry’s (2002) version of the 5-item scale of Bateman and Crant (1993). We assessed openness to experience through Saucier’s (1994) so-called Mini-Markers Set. We captured motivation by asking respondents to indicate the reasons why they had opted for a freelance career. We formulated eight predefined answers in our survey on the basis of our interview results: “more autonomy,” “increased professionalism,” “more work variety,” “more money,” “more flexibility

DRIVERS OF FREELANCE CAREER SUCCESS 33

in time management,” “better work–life balance,” “more challenge,” and “I had no choice.” Of course, alternatively, we could have asked them about their current instead of their past motivations, but only the former captures career choice motives. Moreover, we feel that the measure of past motivation is probably highly correlated with current motivation because most respondents started their freelance career only three to five years ago. To measure business strategy, we included two variables. A first variable measures the distinctiveness of the

respondent-freelancer’s business proposition vis-à-vis the competition (using a 7-point Likert-type scale). A second set of measures assesses the business strategy of the freelancer by using a number of statements based on our seven freelance business strategies: that is, “I focus on a single industry or a single organization,” “I focus on a single product/service,” “I offer better service,” “I offer my products/services at lower costs than the competition,” “I have a broad product/service portfolio,” and “I offer innovative products/services.” Although our interviewees strongly suggested to incorporate market factors into our measurement model,

this was not easy because objective information on these freelance markets and their many niches is completely missing in the Netherlands. Regrettably, reliable information on market factors such as numbers of suppliers, number of competitors, market transparency, and intellectual property rights is simply not avail- able. However, if we would not take account of the environment, this is likely to produce an omitted-variable bias. The potentially strong impact of the market combined with the adverse effect of not including market factors made us decide to use the profession of the freelancer [interim manager, information technology (IT) professional, etc.] to act as a dummy proxy for all market factors. It is debatable whether this occupa- tional dummy can act as a market dummy. Alternatively, this dummy can simply be taken as a control variable (or even a human capital variable, as occupation tends to be highly correlated with certain specific skills). However, our interviewees made strongly arguments in favor of a direct link between their occupation or profession and the market they operate in. That is, they convincingly argued that the characteristics of the market they operate in are very much

defined by their occupation or profession. In some professions, there is excess supply, being associated with low and declining fees; in other professions, there is a shortage of skilled labor, resulting in high and increas- ing fees. Therefore, we decided to follow our interviewees by assuming that the occupational or professional dummy signals market conditions more than anything else, although this dummy also captures other aspects such as specific intangible skills. Additionally, influence of the external market was also measured with a second indicator: the geographical location of the freelancer based on postal code. We translated the latter into a binary variable that distinguished between the most urbanized part of the Netherlands—“Randstad,” which is the heavily populated area in the Amsterdam–The Hague–Rotterdam–Utrecht triangle—and the rest of the country. Finally, we added a number of control variables to our freelance career success model, such as age (in years),

gender (with female coded as 0 and male coded as 1), and health status. We measured health status using a self-assessed Likert-type scale ranging from 1 = poor health condition to 7 = excellent health condition. To conclude, we included a dummy variable that indicated whether the freelancer had another source of income (e.g., through a pension or a part-time job).

Designing the survey

We started with converting our final research model, variables, and corresponding measures into an Internet survey. Sixty freelancers extensively tested this survey to ensure that all questions were clear and easy to answer and that ambiguous and vague terms were avoided. The Internet survey was in Dutch. We first trans- lated validated scales adopted from the international literature from the original English into Dutch and then another person back-translated it from Dutch into English. If there were differences between the original

34 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

English text and the English text that was twice translated, both translators and a third independent person (all fluent in Dutch and English) decided on the final text in Dutch. Common-method variance (CMV) has been identified as a major problem in organizational research on the

basis of self-reports (Chang, van Witteloostuijn, & Eden, 2010; Podsakoff & Organ, 1986; Spector, 2006). Our data are mostly self-reports of objective data and therefore only minimally affected by CMV issues (Spector, 2006, p. 229). Nevertheless, we adopted the procedural remedies suggested by Podsakoff, MacKenzie, Lee, and Podsakoff (2003) to overcome any remaining CMV issues. We assured respondents of anonymity and confidentiality and asked them to answer honestly. We assured individuals that right and wrong answers did not exist. Also, the fact that the survey was sponsored by a total of 19 professional freelance organizations that were all considered very trustworthy decreased the tendency of respondents to give socially desirable answers. Moreover, we promised all respondents a benchmark report on their position in relation to other independent professionals, which motivated respondents to give the “true” score, as they were anxious to see how they performed within their peer group of freelancers, given that wrong answers would distort their own benchmark report. We further designed the survey tool in a way to prevent CMV by adopting a random order of questions, including different types of scales, and by providing clear information on the survey completion process to prevent boredom.

Data Reduction and Analysis

Given the large number of survey questions, we performed data reduction analyses. For every cluster of multi- item independent variables (i.e., human capital, social capital, personality, motivation, and business strategy), we performed a factor analysis on all variables to reduce data complexity. Multicolinearity was not an issue in our data set (all variance inflation factor scores are lower than 1.6), with one exception: the colinearity between age, year of graduation, and total work experience is too high. This is why we excluded both total work experience and year of graduation from the list of human capital variables (but we kept age as a control variable) on which the factor analysis was performed. The factor analysis generated four human capital variables with (a) eigenvalues larger than 1 and (b) an appropriate pattern of loadings: (i) total freelance experience; (ii) higher education level; (iii) university education level; and (iv) recent training (the sum of training days in core professional, new professional, and supporting and adjacent skills in 2006 and 2007). The original survey includes 18 items that attempt to measure social capital. We measured not only network size

but also the total number of weak and strong links, the number of persons new in the network (refreshment), and the seniority or quality of the network. In a factor analysis, the original and validated scales of network activity (the five questions from Forret & Dougherty, 2001) and partner support (the four questions from Greenhaus & Friedman, 2000) came out as separate variables, given the eigenvalues and loading pattern. This is why we performed a second factor analysis without the nine network activity and partner support questions. This resulted in a clear 3-factor outcome, applying the usual eigenvalue threshold (>1) and clean loading pattern criterion: size of the network, business club membership, and actively managing employment agencies. The survey includes 15 personality questions, relating to three validated scales: four on openness, five on pro-

activity, and six on career insight. From the factor analysis, all three scales emerged as specific factors with eigen- values larger than 1 and a clean factor loading pattern. We thus decided to work with the original test scores rather than the factor loading sum. On the basis of the usual eigenvalue and loading pattern criteria, three mo- tivational factors emerged from the factor analysis with the eight motivation-related items: (i) motivated by work–life balance and flexibility; (ii) motivated by professionalism and autonomy; and (iii) motivated by challenge, variety, and money. This is a somewhat peculiar finding. It seems logical that work–life balance and flexibility are in the same grouping, and the grouping of professionalism and autonomy is also

DRIVERS OF FREELANCE CAREER SUCCESS 35

understandable because a high level of professionalism might also require a high level of autonomy. It is less understandable that variety loads on the same factor as challenge and money. Apparently, the same under- lying motivational factor, which may be related to ambition and the need to feel challenged, drives these three aspects. We factor analyzed 11 business strategy items. On the basis of the eigenvalues and the loading pattern, four busi-

ness strategy variables emerged: (i) freelancers who carry innovative and/or differentiating products or services; (ii) freelancers who specialize in one industry; (iii) freelancers who offer a broad product/service range; and (iv) freelancers whose value proposition is based on providing better service and/or lower cost. Finally, we used correlations and logical reasoning to create a limited set of occupational dummies as proxies

for the external market, next to the geographical location dummy distinguishing the urbanized Randstad from the rest of the country. In our data set, it was possible for freelancers to have more than one profession. So, a freelancer could be a coach and a trainer or a software programmer and a functional analyst. This allowed us to see which professions are very much linked. We used this information about overlap between professionals together with logical reasoning to create six occupational variables. First, an interim manager is someone who temporarily heads an organization or department frequently to turn this organization around. Second, an interim professional is someone with specific knowledge of supporting business processes (e.g., finance, human resource management, legal, or IT). Third, a journalist or media professional is someone working in the media industry (e.g., editor or journalist). Fourth, a technical professional is someone who works as an engineer in electronics, construction, and so on. Fifth, coaches and trainers are involved in educat- ing and coaching individuals and groups in certain skills. The sixth and last group defined was called other free agents but consisted primarily out of persons working in facility management. This last group was the only group with significant lower education levels with only 34 per cent having a master’s degree (overall, this is 44 per cent). The control variables included in this study are age, gender, having another source of income (e.g., a pension

scheme or a long-term employment contract), and health status. We excluded from further analysis a number of other variables that are available from the survey because they are endogenous to freelance income (e.g., percentage of freelance income of total income and size of the financial buffer in months). We added age and age squared to explore the potentially parabolic relationship between age and freelance revenue, following standard labor econom- ics. We included SCS in the OCS equation, and vice versa, to estimate the interrelationship between both measures of freelance career success. To investigate whether, despite all the preventive measures, CMV in the data might bias the estimates as reported

in the succeeding texts, we conducted a single overarching Harmon one-factor test. On the basis of these results, there is no evidence that CMV poses a problem. The factor with the largest eigenvalue explains less than eight per cent of the total variance, and more than 20 factors had eigenvalues above 1. In our study, we included six scales that are based on multiple items: openness, pro-activeness, career insight, network activity, partner support, and SCS. We checked the internal consistency of these scales. All tests proved reliable with satisfactory Cronbach’s alphas (a > 0.7), and all reflected a single factor with an eigenvalue above 1.

Results

We show means and standard deviations of the study variables in Table 1. Tables 2 and 3 give the estimation results of the freelance career success model. We estimated both equations using maximum likelihood with robust variance estimators to tackle heterogeneity. We also applied a number of other estimators (such as an instrumental variables estimation to deal with endogeneity issues), but all these robustness analyses generated very comparable results (available upon request).

36 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

T ab le

1 . M ea n s, st an d ar d d ev ia ti o n , an d co rr el at io n s o f d ep en d en t an d in d ep en d en t v ar ia b le s.

M S D

1 2

3 4

5 6

7 8

9 1 0

1 1

1 2

1 3

1 4

1 5

1 6

1 7

1 8

1 9

2 0

2 1

2 2

2 3

2 4

2 5

2 6

2 7

2 8

2 9

3 0

3 1

1 O b je ct iv e ca re er

su cc es s

(l o g )

1 0 .7 7

1 .1 5

1 .3 4

�. 1 2

.2 0

.0 5

�. 0 3

.2 0

.0 1

.1 2

.2 5

.0 5

.0 9

.1 1

.0 7

�. 0 4

.0 5

.0 8

.1 2

.1 5

.0 0

�. 0 4

.3 2

.2 7

�. 1 5

�. 0 8

.0 4

�. 2 4

.0 3

.1 6

.2 7

�. 3 4

2 S u b je ct iv e ca re er

su cc es s

3 1 .9 6

6 .5 0

— 1

�. 0 4

.0 8

.0 3

.1 0

.1 9

.1 8

.1 0

.0 5

.0 6

.2 5

.2 3

.2 9

.0 6

.1 3

.2 8

.0 5

�. 0 4

.0 7

.0 4

.0 2

�. 0 4

.0 2

�. 0 6

.0 2

.2 3

�. 0 2

�. 1 5

3 B ac h el o r’ s d eg re e

0 .4 1

0 .4 9

— —

1 �.

7 4

�. 0 3

.0 4

.0 3

�. 0 4

�. 0 3

.0 4

.0 3

.0 7

�. 0 3

.0 4

�. 0 5

.0 3

�. 0 1

�. 0 6

.0 7

.0 4

�. 0 2

�. 0 8

.0 2

.0 8

�. 0 0

.0 1

�. 0 6

�. 0 5

.0 5

4 M as te r’ s d eg re e

0 .4 4

0 .5 0

— —

— 1

.0 0

�. 0 2

�. 0 0

.0 8

.0 2

�. 0 0

�. 0 7

.0 2

�. 0 4

.0 7

�. 0 3

.0 6

.0 3

.1 0

�. 1 0

�. 0 6

.0 4

.1 5

�. 0 3

�. 0 1

�. 0 9

�. 0 8

� . 0 7

.0 8

�. 0 0

�. 0 6

5 F re el an ce

ex p er ie n ce

(y ea rs )

1 8 .8 0

6 .8 7

— —

1 �.

0 4

�. 1 0

�. 0 5

.0 0

�. 1 1

.0 0

�. 0 5

�. 0 6

�. 1 4

�. 0 7

�. 1 6

.0 3

.1 9

�. 0 7

�. 0 4

.0 3

�. 1 0

�. 1 3

.1 5

�. 0 6

�. 0 4

.1 5

.4 1

�. 0 1

.0 1

�. 0 5

6 T ra in in g d ay s (p er

y ea r)

5 .8 7

5 .4 8

— —

— 1

.2 0

.0 2

.1 2

.0 1

.1 1

.0 9

.1 5

.2 3

.0 2

.0 7

.1 3

�. 0 0

.0 2

�. 0 1

.0 5

.0 4

�. 1 3

.2 7

�. 0 5

�. 0 6

.0 2

.0 1

�. 0 7

.0 4

7 N et w o rk

ac ti v it y (s co re )

1 6 .0 8

3 .7 9

— —

1 .0 8

.3 1

.1 1

.2 5

.2 9

.2 4

.3 3

.0 0

.0 3

.0 5

.2 5

.0 2

.0 5

�. 0 3

.1 7

.1 3

.0 1

.1 2

�. 1 1

�. 1 4

.0 1

.0 8

�. 0 2

�. 0 6

8 P ar tn er

su p p o rt (s co re )

1 4 .9 4

3 .9 9

— —

— 1

.0 1

�. 0 4

.0 5

.1 4

.1 9

.0 6

.0 1

.0 6

.1 0

.0 1

�. 0 1

.0 4

�. 0 3

�. 0 4

.0 4

.0 3

�. 0 4

.0 5

�. 1 0

.1 1

�. 1 6

�. 0 3

9 S iz e o f n et w o rk

(f ac to r)

0 .0 0

1 .0 0

— —

1 .0 0

.0 0

.1 5

.1 4

.1 1

�. 0 1

.0 6

.1 9

�. 0 3

.0 4

� . 0 4

.1 1

.0 8

.0 1

.0 8

�. 0 4

.0 5

.0 3

�. 0 3

1 0

M an ag in g ag en ci es

(f ac to r)

0 .0 0

1 .0 0

— —

— 1

.0 0

.0 6

.0 7

.1 2

.0 0

.1 1

�. 0 3

�. 1 2

.1 1

.0 8

�. 0 4

.4 1

.2 0

�. 1 4

�. 0 6

�. 1 0

.0 1

.0 7

.1 8

�. 1 2

1 1

B u si n es s cl u b m em

be rs h ip

(f ac to r)

0 .0 0

1 .0 0

— —

1 .1 1

.1 0

.1 5

.0 2

.0 0

.0 2

.0 9

�. 0 1

.0 1

.0 6

.0 7

.0 5

.0 6

�. 0 8

�. 0 3

.0 7

.0 4

�. 0 8

�. 0 3

1 2

O p en n es s (s co re )

1 9 .0 7

3 .2 1

— —

— 1

.4 6

.3 3

�. 0 2

.0 3

.1 0

.2 1

�. 0 1

.0 6

.0 4

.1 6

.0 5

�. 0 5

.1 1

�. 0 1

�. 0 5

�. 0 1

.1 9

�. 0 9

�. 0 4

1 3

P ro -a ct iv en es s (s co re )

2 1 .9 1

4 .0 0

— —

1 .3 4

�. 0 1

.0 6

.1 5

.3 5

�. 0 2

.0 7

�. 0 4

.2 2

.1 1

�. 0 4

.0 5

.0 3

�. 1 1

�. 0 2

.1 2

.0 7

�. 0 1

1 4

C ar ee r in si g h t (s co re )

2 1 .8 5

3 .9 9

— —

— 1

.0 1

.0 9

.0 6

.2 9

.0 6

�. 0 0

.0 4

.1 2

.0 4

�. 0 7

.1 4

.0 0

�. 0 6

.0 2

.1 2

�. 0 6

.0 1

1 5

W o rk – li fe

b al an ce

(s co re )

0 .0 0

1 .0 0

— —

1 .0 0

.0 0

�. 0 3

�. 0 5

�. 0 2

.0 5

�. 0 7

�. 0 4

.0 7

�. 0 0

�. 0 5

�. 0 1

�. 1 8

.0 3

�. 1 0

�. 0 6

1 6

C h al le n ge

an d p ro fi t

(f ac to r)

0 .0 0

1 .0 0

— —

— 1

�. 0 0

.0 3

.0 6

.0 8

.1 0

.0 2

.0 0

�. 0 3

.0 5

�. 0 7

�. 1 3

.0 6

.0 2

.0 1

1 7

A u to n o m y an d

m as te ry

(f ac to r)

0 .0 0

1 .0 0

— —

1 .1 6

�. 0 1

�. 0 2

.0 5

�. 0 1

.0 5

�. 0 2

.1 3

�. 0 9

�. 0 6

�. 0 7

�. 0 0

�. 0 6

�. 0 5

1 8

D if fe re n ti at e an d in n o v at e

(f ac to r)

0 .0 0

1 .0 0

— —

— 1

.0 0

�. 0 0

.0 0

�. 0 2

.0 5

.0 8

.0 3

.0 0

.1 8

.0 7

.0 3

�. 0 3

1 9

In d u st ry

sp ec ia li za ti o n

(f ac to r)

0 .0 0

1 .0 0

— —

1 �.

0 0

�. 0 0

.1 2

�. 2 5

�. 0 2

�. 0 6

�. 0 2

.0 3

.0 8

�. 0 3

2 0

P ro d u ct

ra n g e (f ac to r)

0 .0 0

1 .0 0

— —

— 1

�. 0 0

.0 9

.0 6

�. 0 1

�. 0 6

.0 3

.0 2

�. 0 5

.0 1

.0 2

�. 0 1

2 1

P ri ce

le ad er sh ip

(f ac to r)

0 .0 0

1 .0 0

— —

1 �.

0 9

�. 0 0

�. 1 0

.0 0

.0 2

.0 6

�. 0 1

�. 0 3

.0 3

2 2

In te ri m

m an ag er

0 .2 7

0 .4 5

— —

— 1

.2 6

�. 2 0

.0 6

�. 0 3

�. 2 1

.1 2

.1 0

.2 8

�. 0 4

2 3

In te ri m

p ro fe ss io n al

0 .4 6

0 .5 0

— —

1 �.

3 3

.0 2

�. 1 2

�. 2 9

�. 0 1

.0 4

.3 0

�. 0 3

2 4

Jo u rn al is t

0 .1 9

0 .3 9

— —

— 1

�. 2 3

�. 1 2

�. 0 8

�. 0 9

�. 0 5

�. 1 6

�. 0 4

2 5

T ra in er /c o ac h

0 .3 6

0 .4 8

— —

1 �.

1 4

�. 1 8

.1 4

�. 0 0

�. 1 1

.0 9

2 6

T ec h n ic al

0 .0 6

0 .2 5

— —

— 1

�. 0 8

.0 5

.0 2

.2 1

.0 2

2 7

O th er

fr ee la n ce r

0 .1 5

0 .3 5

— —

1 �.

0 4

�. 0 5

�. 1 9

.0 3

2 8

A g e

4 6 .2 9

8 .7 1

— —

— 1

.0 0

.1 9

.1 5

2 9

H ea lt h st at u s

6 .0 7

0 .8 9

— —

1 .0 4

�. 0 7

3 0

G en d er

0 .5 0

— —

— 1

.0 0

3 1

O th er

in co m e

0 .2 4

0 .4 3

— —

DRIVERS OF FREELANCE CAREER SUCCESS 37

Objective career success

Table 2 shows that the adjusted R2 is 44 per cent. This implies a very good model fit for this type of analyses, given the stylized fact in the entrepreneurship literature that an (adjusted) R2 of 15 per cent is already exceptional in studies of small-firm performance (Parker et al., 2010). Almost all control variables are significant and have the expected sign. This is also the case for most external market and human capital variables, with only one remarkable exception within the group of human capital variables: the significantly negative relationship between recent training effort and OCS. This may indicate reverse causality. First, a freelancer engaging in training cannot charge clients for services. Second, independent professionals without a current assignment have much more spare time to invest in training. Within the group personality variables of the personal capital cluster, none of the variables are significant. Neither career insight nor openness nor a pro-active personality seems to be related to OCS. A number of motivation

Table 2. Testing the freelance career model—objective career success.

Factor Coefficient SE z-statistic

Control variables Constant 6.980609 0.7493958 9.311

Age 0.0812761 0.0286227 2.841

Age2 �0.0009651 0.0003046 �3.171 Gender (male = 1) 0.4200948 0.0580151 7.241

Health (self-assessment score) 0.0472644 0.0286808 1.65†

Receiving other income (e.g., regular job) �0.589795 0.0666048 �8.861 Market factors Living in Randstad area (Randstad = 1) 0.0799059 0.0480397 1.66†

Interim manager (dummy) 0.402606 0.0606881 6.631

Interim professional (dummy) 0.1432689 0.0589014 2.43* Journalist and/or media professional (dummy) �0.3227554 0.0780846 �4.131 Technical professional (dummy) 0.1039439 0.1071783 0.97 Trainer and/or coach (dummy) �0.2152659 0.0572763 �3.761 Other free agent (dummy) �0.461215 0.0811657 �5.681

Human capital University education (dummy) 0.3556629 0.0790103 4.501

“HBO” education (dummy) 0.1380998 0.0805207 1.72†

Freelance experience (log) 0.1622809 0.0229957 7.061

Recent training participation (sum) �0.0073522 0.0025106 �2.931 Motivation capital Motivated by autonomy and professionalism 0.0415792 0.0244645 1.70†

Motivated by challenge and money 0.0196196 0.0237121 0.83 Motivated by work–life balance and flexibility �0.0535313 0.0235548 �2.27*

Personality capital Insight in career (sum) �0.0042271 0.0074761 �0.57 Open personality (sum) �0.0034226 0.0070108 �0.49 Pro-active personality (sum) �0.0149058 0.0092242 �1.62

Social capital Business club member (factor) 0.0149219 0.024884 0.60 Managing the agent (factor) 0.0838912 0.0248262 3.381

Size of network (factor) 0.0338625 0.0267804 1.26 Network activity (sum) 0.029306 0.0074711 3.921

Partner support (sum) 0.0018221 0.0058532 0.31 Business strategy Better service or low-cost strategy (factor) �0.0101224 0.0236016 �0.43

Innovative and differentiation strategy (factor) 0.02476 0.0282662 0.88 Industry specialization strategy (factor) 0.0562731 0.0235155 2.39* Broad product range (factor) �0.0468089 0.0247748 �1.89† Subjective career success (sum) 0.0441134 0.004511 9.781

Adjusted R2 0.4393 Est. method ML (robust) Log likelihood �1754.97 Observations (n) 1385 Akaike information criterion 2.581907

†significant at 10% level, *significant at 5% level; and **significant at 1% level.

38 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

variables are significant, though: freelancers primarily driven by flexibility and/or work–life balance motives have lower revenue than professionals who are in freelancing for reasons other than that. Two social capital variables proved to be significant: managing agencies, and the network activity score of

Forret and Dougherty (2001). Actively managing employment agencies (i.e., being registered with employment agencies and actively visiting them) adds to the OCS of the independent professional. The positive and significant estimate of network activity suggests that being an active networker is important. Calling people, visiting business contacts, handing out business cards, and sending mails, cards, and gifts to individuals in the network contribute to financial success. Neither the size of the network nor whether an individual is a member of a business club adds to OCS. Also, partner support proved not to be related to OCS. Finally, the business strategy variables also seem to add to OCS. Industry specialization is important for freelancers,

Table 3. Testing the freelance career model—subjective career success.

Factor Coefficient SE z-statistic

Control variables Constant �1.356558 4.407242 �0.31 Age �0.2052013 0.173482 �1.18 Age squared 0.002439 0.001864 1.31 Gender (male = 1) �0.799514 0.3471066 �2.30* Health (self-assessment score) 0.984084 0.1931404 5.10** Receiving other income (e.g., regular job) �0.4763586 0.4079105 �1.17

Market factors Living in Randstad area (Randstad = 1) 0.0150572 0.306105 0.05 Interim manager (dummy) �1.173891 0.398087 �2.95** Interim professional (dummy) �0.7440611 0.3739605 �1.99* Journalist and/or media professional (dummy) �0.5157591 0.5009407 �1.03 Technical professional (dummy) 0.434244 0.6271601 0.69 Trainer and/or coach (dummy) �0.2692279 0.3636077 �0.74 Other free agent (dummy) �0.0336539 0.5867987 �0.06

Human capital University education (dummy) 0.7539263 0.4749756 1.59 “HBO” education (dummy) 0.4834854 0.4773371 1.01 Freelance experience (log) �0.003937 0.141182 �0.03 Recent training participation (sum) 0.0242919 0.0153892 1.58

Motivation capital Motivated by autonomy and professionalism 0.3427641 0.1552853 2.21* Motivated by challenge and money 0.1804451 0.1503154 1.20 Motivated by work–life balance and flexibility 0.3705541 0.1561106 2.37*

Personality Capital Insight in career (sum) 0.225339 0.0479622 4.70** Open personality (sum) 0.1392696 0.049208 2.83** Pro-active personality (sum) 0.130006 0.0616591 2.11*

Social capital Business club member (factor) �0.1613818 0.1536743 �1.05 Managing the agent (factor) 0.0023039 0.1864347 0.01 Size of network (factor) �0.0089879 0.1558321 �0.06 Network activity (sum) 0.0369041 0.0511501 0.72 Partner support (sum) 0.1274405 0.0467177 2.73**

Business strategy Better service or low-cost strategy (factor) �0.2922321 0.1466776 �1.99* Innovative and differentiation strategy (factor) 0.9313536 0.1882798 4.95** Industry specialization strategy (factor) 0.0075344 0.1495288 0.05 Broad product range (factor) 0.4305484 0.162117 2.66** Objective career success (log) 1.769593 0.1813919 9.76** Adjusted R2 0.2868 Est. method ML (robust) Log likelihood �4311.5 Observations (n) 1385 Akaike information criterion 6.273649

†significant at 10% level; *significant at 5% level; and **significant at 1% level.

DRIVERS OF FREELANCE CAREER SUCCESS 39

significantly improving OCS. A broad product/service range is almost significant, leading to less OCS. This sug- gests that focuses on an industry and on a small range of products/services are important financial success factors.

Subjective career success

Looking at the results in Tables 2 and 3, we observe that the adjusted R2 is considerably lower for the SCS (0.29) than for OCS (0.44), albeit still far above the top level of 0.15 in the small-business literature. This suggests that our model is better in predicting OCS than SCS. Moreover, the estimates reveal that a set of very different variables is important in determining SCS when compared with OCS. In effect, the variables that are important for SCS are those variables that are irrelevant to OCS and vice versa. Among the control variables, only self-assessed health and gender are significantly predicting SCS, in the expected direction. Men are less satisfied with their career than women, and freelancers with good health are more satisfied with their career than those struggling with their health. By and large, external market and human capital indicators, although crucial for OCS, fail to affect SCS. Only the

dummies for interim managers and professionals are significant. Freelancers in these occupations are not as satisfied with their career as one would expect. All personality variables from the personal capital cluster are strongly signif- icant. Especially, career insight stands out, but the personality traits of openness to experience and pro-activeness are essential drivers of SCS as well. Motivation is important for SCS, too: professionals who freelance for reasons of flexibility and work–life balance and professionals who freelance because of autonomy and professionalism are more satisfied with their career. Only one social capital variable has a positive impact on SCS: partner support. Although not hypothesized about, for lack of theory on this, we explore the link between business strategy

variables and SCS. Indeed, business strategy turns out to be very important for SCS. Especially, independent professionals who distinguish themselves through a strategy that is innovative and independent professionals with a broad product/service range have highly satisfying careers (i.e., high SCS scores). It might be that a large variety of different assignments and being innovative provides intrinsic utility to independent professionals, which subsequently increases their SCS. Independent professionals with a “low-cost” or “better-service” strategy have lower SCS levels. In the last row of Tables 2 and 3, the positive cross-relationship between OCS and SCS is evident. Although causality cannot be established in cross-sectional analysis, OCS and SCS are definitely positively related (r = 0.34, with p < 0.001). This correlation is somewhat higher than the average correlation of 0.30 between salary and career satisfaction (with a 95 per cent confidence interval of 0.28–0.32) as reported by Ng et al. (2005) in their meta-analysis of OCS and SCS for employees. The fact that some freelancers in our sample are largely motivated by monetary rewards may perhaps explain the somewhat higher correlation between OCS and SCS when compared with employees. In Table 4, an overview is given of our findings about the first five hypotheses. For the most part, empirical

evidence supported these hypotheses. Nevertheless, there are some interesting observations. Human capital is important for OCS but not for SCS. In contrast, personal capital is important for SCS but not for OCS. Only social capital is important for both OCS and SCS. All in all, our hypotheses are fully or partly supported. Our hypotheses are arguably not very detailed, but we feel that these broadly formulated hypotheses are appropriate in the context of exploratory analyses of a relatively new and unexplored subject. Our final two hypotheses (6a and 6b) are not listed in Table 4, as we need complementary analyses for this pair of

hypotheses. We applied Budescu’s (1993) dominance analysis, based on generalized least squares estimation, in this respect. Although dominance analysis is a very labor-intensive method, it is very much suited to explore our two remaining hypotheses. For every possible combination of variable classes, we calculated the R2 measure. In total, this generates 64 measures of fit for OCS and SCS. For all combinations, we calculated the extra fit of adding a new class of variables. Subsequently, we computed the average extra fit. We reported the outcomes of this exercise in Tables 5 and 6. K is the number of variable classes already in the equation. In the cells, the extra R2 measured the

40 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

effect on fit. For example, if an equation includes only one class of variables (K = 1)—that is, human capital, social capital, personality measures, motivational variables, or business strategy—the addition of the external market proxies will increase the fit of that equation (i.e., R2) with, on average, 0.0819. Similarly, if we add the external market dummies to an equation with two other classes of variables included (K = 2)—for example, human and social capital—the fit implied by the extra R2 will increase with an average of 0.0744.

Table 6. Dominance analysis of subjective career success.

External market Human capital Personality Motivation Social capital Business strategy

K = 0 0.0070 0.0104 0.0720 0.0155 0.0410 0.0614 K = 1 0.0081 0.0076 0.0613 0.0121 0.0358 0.0503 K = 2 0.0086 0.0055 0.0519 0.0097 0.0315 0.0406 K = 3 0.0087 �0.0001 0.0437 0.0078 0.0220 0.0240 K = 4 0.0085 0.0032 0.0346 0.0063 0.0254 0.0256 K = 5 0.0083 0.0028 0.0311 0.0052 0.0236 0.0202 Mean 0.0085 0.0038 0.0481 0.0090 0.0283 0.0346 Relative percentage 6.4 2.9 36.4 6.8 21.4 26.2

Table 4. Overview of hypotheses.

Hypothesis Outcome

1. Human capital is positively related to OCS and SCS

Partly accepted. Yes, for OCS (p < 0.001); no for SCS (p = 0.271)

2. Social capital is positively related to OCS and SCS

Completely accepted. Yes, for OCS (p < 0.001); yes for SCS (p = 0.034)

3. Career insight, pro-activeness, and openness are positively related to OCS and SCS

Partly accepted. No, for OCS (p = 0.18); yes for SCS (p < 0.001)

4a: Autonomy, flexibility, and work–life balance are negatively related to OCS

Partly accepted. Only flexibility and work–life balance are motivations that negatively affect OCS (p = 0.02). Autonomy is not negatively related to OCS

4b: Autonomy, flexibility, and work–life balance are positively related to SCS

Fully accepted. Flexibility and work–life balance are motivations that positively affect OCS (p = 0.02). Autonomy is also positively related to SCS (p = 0.03)

5. Low-cost, focus, and differentiation strategies are positively related to OCS and SCS

Partly accepted. Focus strategies [i.e., product/service focus (p = 0.06) and industry focus (p = 0.02)] are positively related to OCS. Differentiation leads to SCS (p < 0.001). Low-cost strategy leads to less SCS (p = 0.04)

OCS, objective career success; SCS, subjective career success.

Table 5. Dominance analysis of objective career success.

External market Human capital Personality Motivation Social capital Business strategy

K = 0 0.0895 0.0368 0.0000 0.0059 0.0465 0.0134 K = 1 0.0819 0.0391 0.0011 0.0053 0.0411 0.0110 K = 2 0.0744 0.0405 0.0018 0.0047 0.0356 0.0090 K = 3 0.0669 0.0411 0.0020 0.0042 0.0298 0.0073 K = 4 0.0598 0.0413 0.0023 0.0038 0.0248 0.0055 K = 5 0.0531 0.0412 0.0014 0.0035 0.0200 0.0040 Mean 0.0707 0.0405 0.0018 0.0045 0.0328 0.0082 Relative percentage 44.6 25.6 1.1 2.8 20.7 5.2

DRIVERS OF FREELANCE CAREER SUCCESS 41

By inspecting all possible combinations, we can conclude that the average extra fit of adding external market dummies is 0.0707. This is higher than the average extra fit of human capital (0.0405), personality (0.0018), motivation (0.0045), social capital (0.0328), and business strategy (0.0082) variables. External market features are, therefore, the most important determinant of OCS (44.6 per cent of explained variance), followed by human capital (25.6 per cent) and social capital (20.7 per cent), respectively. Business strategy (5.2 per cent), motivation (2.8 per cent), and personality (1.1 per cent) are all of minor importance in driving OCS. We followed the same procedure to determine the relative importance of factors determining SCS. We summa-

rized the findings in Table 6. Personality (0.0481; 36.4 per cent of explained variance) is the most important driver of SCS, followed by business strategy (0.0346; 26.2 per cent) and social capital (0.0283; 21.4 per cent), in that order. The other success drivers are all less important. The average extra fit of market features is 0.0085 or 6.4 per cent, of motivation 0.0090 or 6.8 per cent, and of human capital 0.0038 or 2.9 per cent. These drivers are arguably very important in determining OCS, but they are almost irrelevant in affecting SCS. Social capital is the only factor that has a considerably positive impact on both OCS and SCS. On the basis of the evidence provided by the dominance analysis, we accept our final two Hypotheses 6a and 6b.

Discussion

We would like to start with emphasizing the two major benefits of this study. First, it is one of the first large-scale quantitative studies that specifically targets freelancers as the sample group. If Handy’s (1989) assessment about modern organizations is correct, the number of freelancers will continue to increase, as current evidence strongly suggests indeed. Second, as the freelance career is the quintessence of the individual career, as is argued in this paper, freelance careers provide a great opportunity to test the individual career concept. We believe that this empirical study of freelancers provided perhaps one of the best settings for testing insights from the individual career perspective. A first important finding of this study is that a model based on the intelligent career framework augmented with

factors from the entrepreneurship literature is largely suited to explain freelance career success, as is demonstrated by the support for many of our hypotheses. Hence, this study confirms the usefulness of individual career models. A second finding of this study is that subjective and objective career yardsticks are characterized by very different underlying processes. Individual career makers continuously make trade-offs between SCS, on the one hand, and OCS, on the other hand. More research into the trade-off between SCS and OCS is needed, to deepen our understanding of the options open to freelancers. Additionally, a third key result of this study is that business aspects such as the external market and business strategy are critical elements driving the freelance career. This confirms that freelancers reflect a hybrid, with elements from the small-business world and aspects from individual career theory. A better understanding of the external market and its implications for individual careers is needed if we want to make better career models. Here, we find that career models can learn from strategic management, as both perspectives address the question about what drives competitive advantage, but strategic management especially addresses the environment as an element in the vector of variables that determine success. In the current study, only a first step is taken. We also think that there is much practical merit in all the small insights provided by this empirical research that is

relevant to freelancers. This paper shows that there is a parabolic influence of age on OCS and that the maximum revenue of freelancers lies between 45 and 50 years of age; after which, utilization rates and professional fees diminish. With respect to freelance networks, the study shows that network size is not important, but that building strong relations with agents and putting substantial effort in building and maintaining a network are essential elements that drive both OCS and SCS. Moreover, the paper demonstrates clearly the importance of a focus strategy, where the freelancer specializes in a product/service or sector, as well as the adverse effects of a low-cost strategy. These and other insights are interesting for academics but of great practical importance to freelancers, too.

42 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

As any study, ours suffers from drawbacks, which leave room for further study. We would like to emphasize seven of these. First, our final model as depicted in Figure 1 does not incorporate the concept of strategic fit (Parker & van Witteloostuijn, 2010). Since the 1960s, contingency theory has argued convincingly that there is no universally best way to achieve organizational success. Instead, the performance-maximizing course of action is contingent upon the organization’s internal context and external environment. The concept of strategic fit is part of the intelligent career framework, too, as interrelationships between the three ways of knowing may affect career outcomes. Although modeling these interrelations is probably warranted from a theoretical perspective, we have not modeled these complex interrelations at this exploratory stage of our research (see also Zajac, Kraatz, & Bresser, 2000) but set this aside as a task for future work. Second, this study is cross-sectional, which implies that we cannot study career changes over lifetimes. As the

modern career becomes more fluid, individuals leap increasingly from one assignment to another and from a freelance contract to a traditional employment job and back. Third, our cross-section only includes freelancers, implying that we cannot investigate any selection effect. In future work, it would be interesting to see what causes an individual to become a freelancer, instead of an employee or an employer. Fourth, this study’s design does not include signaling (Jones, 2002) and stretch work (O’Mahony & Bechky, 2006), although we know that these strategies are crucial to the success of freelancers. A future study might—for instance, by studying the curriculum vitae of freelancers—focus on such signaling and stretch work behavior, to determine how these strategies impact the success rate of freelancers. Fifth, this study focuses on the Netherlands. A comparative international study of freelancers, either generally or focusing on a number of specific freelance markets (e.g., IT and media), is needed to see whether our findings are generalizable to the economies of other developed and non-developed societies. Last but not least, this study was restricted to occupational dummies as proxies for market factors. This is far from ideal, as we discussed at length earlier. We should try to include direct measures of market features (size, transparency, seller versus buyer markets, etc.) in future work. Additionally, as career choices are always shaped by the whole context in which the individual resides, we should look not only into developments in the marketplace but also at other aspects that may guide the interests, goals, and actions of freelancers (Lent, Brown, & Hackett, 2000).

Author biographies

Arjan van de Born is assistant professor in Economics and Management at the Universities of Antwerpen and Utrecht. His research interests range from entrepreneurship, network organizations and organizational change. Arjen van Witteloostuijn is full professor in Economics and Management at the Universities of Antwerpen, Tilburg and Utrecht. His research interests range from the effect of language on individual decision-making to the evolution of political party systems. He has published widely in such journals as the Academy of Management Jour- nal, Academy of Management Review, American Journal of Political Science, American Sociological Review, Journal of International Business Studies, Management Science, Organization Science and Strategic Management Journal.

References Ajayi-Obe, O., & Parker, S. C. (2005). The changing nature of work among the self-employed in the 1990s: Evidence from Britain. Journal of Labor Research, 26, 501–517.

Aldrich, H. E., & Reese, P. R. (1993). Does networking pay off? A panel study of entrepreneurs in the research triangle. In N. C. Churchill, S. Birley, J. Doutriaux, E. J. Gatewood, F. S. Hoy, & W. E. Wetzel Jr. (Eds.), Frontiers of entrepreneurship re- search (pp. 325–339). Babson College Entrepreneurship Research Conference (BCERC). http://digitalknowledge.babson. edu/fer/

DRIVERS OF FREELANCE CAREER SUCCESS 43

http://digitalknowledge.babson.edu/fer/

Amit, R., Glosten L., & Muller, E. (1993). Challenges to theory development in entrepreneurship research. Journal of Management Studies, 30, 815–834.

Anderson, P. (2008). Happiness & health: Well-being among the self-employed. Journal of Socio-Economics, 37, 213–236. Arthur, M. B., & Rousseau, D. M. (1996). A career lexicon for the 21st century. The Academy of Management Executive, 10, 28–39. Arthur, M. B., Khapova, S. N., & Wilderom, C. P. M. (2005). Career success in a boundaryless career world. Journal of Organizational Behavior, 26, 177–205.

Arthur, M. B., Claman, P. H., & DeFillippi, R. J. (1995). Intelligent enterprise, intelligent careers. The Academy of Management Executive, 9, 7–22.

Arnold, J. (2001). Career and career management. In N. Anderson, D. S. Ones, & H. K. Sinangil (Eds.), Handbook of industrial, work and organizational psychology: Organizational psychology, Thousand Oaks, CA: Sage.

Arnold, J., & Cohen, L. (2008). The psychology of careers in industrial and organizational settings: A critical but appreciative analysis. International Review of Industrial and Organizational Psychology, 23, 1–44.

Arum, R., & Müller, W. (2004). The return of self-employment: A cross-national study of self-employment and social inequality, Princeton, NJ: Princeton University Press.

Barbieri, P. (2003). Social capital and self-employment a network analysis experiment and several considerations. International Sociology, 18, 681–701.

Barley, S. R., Kunda G., & Evans, J. (2002). Why do contractors contract? The experience of highly skilled technical professionals in a contingent labor market. Industrial & Labor Relations Review, 55, 234–261.

Barley, S. R. & Kunda, G. (2004). Gurus, hired guns, and warm bodies: Itinerant experts in a knowledge economy. Princeton, NJ: Princeton University Press.

Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17, 99–120. Baron, J. N., Davis-Blake, A., & Bielby, W. T. (1986). The structure of opportunity: How promotion ladders vary within and among organizations. Administrative Science Quarterly, 31, 249–273.

Bhaskar, V., & To, T. (2003). Oligopsony and the distribution of wages. European Economic Review, 47, 371–399. Bateman, T. S., & Crant, J. M. (1993). The proactive component of organizational behavior: A measure and correlates. Journal of Organizational Behavior, 14, 103–118.

Baum, J. R., Locke, E. A., & Smith, K. G. (2001). A multi-dimensional model of venture growth. Academy of Management Journal, 44, 292–303.

Begley, T. M., & Boyd, D. P. (1987). Psychological characteristics associated with performance in entrepreneurial firms and smaller businesses. Journal of Business Venturing, 2, 79–93.

Benz, M., & Frey, B. S. (2008). Being independent is a great thing: Subjective evaluations of self-employment and hierarchy. Economica, 75, 362–383.

Blanchflower, D. G. (2004). Self-employment: More may not be better. Swedish Economic Policy Review, 11, 15–74. Boehm, J. K., & Lyubomirsky, S. (2008). Does happiness promote career success? Journal of Career Assessment, 16, 101–116. Boone, C., De Brabander B., & van Witteloostuijn, A. (1996). CEO locus of control and small firm performance: An integrative framework and empirical test. Journal of Management Studies, 33, 667–699.

Boone, C., & van Witteloostuijn, A. (2004). Towards a unified theory of market partitioning: Integrating resource partitioning and sunk cost theories of dual market structures. Industrial and Corporate Change, 13, 701–726.

Bosma, N., van Praag, M., Thurik, R., & de Wit, G. (2004). The value of human and social capital investments for the business performance of startups. Small Business Economics, 23, 227–236.

Boudreau, J. W., Boswell, W. R., & Judge, T. A. (2001). Effects of personality on executive career success in the United States and Europe. Journal of Vocational Behavior, 58, 53–81.

Brockhaus, R. H. (1980). Psychological and environmental factors which distinguish the successful from the unsuccessful entrepreneur: A longitudinal study. Proceedings of the Academy of Management of the 40th Annual Meeting, 368–372.

Budescu, D. V. (1993). Dominance analysis: A new approach to the problem of relative importance of predictors in multiple regression. Psychological Bulletin, 114, 542–551.

Cappelli, P. (1999). Career jobs are dead. California Management Review, 42, 146–167. Chang, S. J., van Witteloostuijn, A., & Eden, L. (2010). Common-method variance in international business research, letter from the editors. Journal of International Business Studies, 41, 178–184.

Chen, C. C., Greene, P. G., & Crick, A.(1998). Does entrepreneurial self-efficacy distinguish entrepreneurs from managers? Jour- nal of Business Venturing, 13, 295–316.

Chiesi, A. M. (2007). Measuring social capital and its effectiveness. The case of small entrepreneurs in Italy. European Sociological Review, 23, 437–453.

Cohen, L., & Mallon, M. (1999). The transition from organisational employment to portfolio working: Perceptions of boundary- lessness. Work, Employment & Society, 13, 329–352.

Connelly, C. E., & Gallagher, D. G. (2004). Emerging trends in contingent work research. Journal of Management, 30, 959–983.

44 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

Davidson, P., & Honig, B. (2003). The role of social and human capital among nascent entrepreneurs. Journal of Business Venturing, 18, 301–331.

DeFillipi, R. J., & Arthur, M. B. (1994). The boundaryless career: A competency-based perspective. Journal of Organizational Behavior, 15, 307–324.

DeFillippi, R. J., & Arthur, M. B. (1996). Boundaryless contexts and careers: A competency-based perspective. In M. B. Arthur, & D. M. Rousseau (Eds.), The boundaryless career. New York, NY: University Press.

Eby, L. T., Butts M., & Lockwood, A. (2003). Predictors of success in the era of the boundaryless career. Journal of Organiza- tional Behavior, 24, 689–708.

Forret, M. L., & Dougherty, T. W. (2001). Correlates of networking behavior for managerial and professional employees. Group Organization Management, 26, 283–311.

Frenkel-Brunswik, E. (1948). Intolerance of ambiguity as emotional perceptual. Personality of Assessment, 40, 67–72. Granovetter, M. S. (1973). The strength of weak ties. The American Journal of Sociology, 78, 1360–1380. Greenhaus, J. H., Parasuraman, S., & Wormley, W. M. (1990). Effects of race on organizational experiences, job performance evaluations, and career outcomes. Academy of Management Journal, 33, 64–86.

Greenhaus, J. H., & Friedman, D. (2000). Work and family—allies or enemies? What happens when business professionals confront life choices. Oxford: Oxford University Press.

Guest, D. (2004). Flexible employment contracts, the psychological contract and employee outcomes: An analysis and review of the evidence. International Journal of Management Reviews, 5, 1–19.

Guest, D., & Clinton, M. (2006). Temporary employment contracts, workers’ well-being and behaviour: Evidence from the UK. King’s College Department of Management Working Paper # 38.

Hall, D. T. (1976). Careers in organizations. Pacific Palisades, CA: Goodyear. Hall, D. T. (2002). Careers in and out of organizations. Thousand Oaks, CA: Sage. Handy, C. B. (1985). The future of work. Cambridge, MA: Blackwell Publishers. Handy, C. B. (1989). The age of unreason. Cambridge, MA: Harvard Business School Press. van der Heijden, B. I. J. M. (2006). Age differences in career activities among higher-level employees in the Netherlands: A comparison between profit sector and non-profit sector staff. International Journal of Training and Development, 10, 98–120.

Heslin, P. A. (2003). Self- and other-referent criteria of career success. Journal of Career Assessment, 11, 262–286. Heslin, P. A. (2005). Conceptualizing and evaluating career success. Journal of Organizational Behavior, 26, 113–136. Hoanga, H., & Antoncic, B. (2003). Network-based research in entrepreneurship: A critical review. Journal of Business Venturing, 18, 165–187.

Jones, C. (2002). Signaling expertise: How signals shape careers in creative industries. In M. A. Peiperl, & M. B. Arthur (Eds.), Career frontiers: New conceptions of working lives. Oxford: Oxford University Press.

Judge, T. A., & Ferris, G. R. (1993). Social context of performance evaluation decisions. Academy of Management Journal, 36, 80–105.

Kickul, J., & Gundry, L. (2002). Prospecting for strategic advantage: The proactive entrepreneurial personality and small firm innovation. Journal of Small Business Management, 40, 85–97.

Kirkpatrick, I., & Hoque, K. (2006). A retreat from permanent employment? Accounting for the rise of professional agency work in UK public services. Work, Employment and Society, 20, 649–666.

Lent, R. W., Brown, S. D., & Hackett, G. (2000). Contextual supports and barriers to career choice: A social cognitive analysis. Journal of Counseling Psychology, 47, 36–49.

Lips-Wiersma, M., & Hall, D. T. (2007). Organizational career development is not dead: A case study on managing the new career during organizational change. Journal of Organizational Behavior, 28, 771–792.

Littler, C. R., Wiesner, R., & Dunford, R. (2003). The dynamics of delayering: Changing management structures in three countries. Journal of Management Studies, 40, 225–256.

Low, M. B., & MacMillan, I. C. (1988). Entrepreneurship: Past research and future challenges. Journal of Management, 14, 139–151.

Maijoor, S. J., & van Witteloostuijn, A. (1996). An empirical test of the resource-based theory: Strategic regulation in the Dutch audit industry. Strategic Management Journal, 17, 549–569.

Marler, J. H., Barringer, M. W., & Milkovic, G. T. (2002). Boundaryless and traditional contingent employees: Worlds apart. Journal of Organizational Behavior, 23, 425–453.

Mayrhofer, W., Meyer, M., & Steyrer, J. (2007). Contextual issues in the study of careers. In H. Gunz, & Peiperl M. (Eds.), Handbook of career studies. Los Angeles, CA: Sage.

McClelland, D. C. (1965). Achievement and entrepreneurship: A longitudinal study. Journal of Personality and Social Psychology, 95, 389–392.

McGahan, A. M., & Porter, M. E. (1997). How much does industry matter, really? Strategic Management Journal, 18, 15–30. McMillan, I. C., & Gunther-McGrath, R. (2000). The entrepreneurial mindset: Strategies for continuously creating opportunity in an age of uncertainty. Cambridge, MA: Harvard Business School Press.

DRIVERS OF FREELANCE CAREER SUCCESS 45

Ng, T., Lillian, W. H., Eby, T., Sorensen, K. L., & Feldman, D. C. (2005). Predictors of objective and subjective career success: A meta-analysis. Personnel Psychology, 58, 367–408.

Noe, R. A., Noe, A. W., & Bachhuber, J. A. (1990). An investigation of the correlates of career motivation. Journal of Vocational Behavior, 37, 340–356.

O’Mahony, S., & Bechky, B. A. (2006). Stretchwork: Managing the career progression paradox in external labor markets. Academy of Management Journal, 49, 918–941.

Ostgaard, T. A., & Birley, S. (1996). New venture and personal growth networks. Journal of Business Research, 36, 37–50. Parker, P., & Arthur, M. B. (2000). Careers, organizing, and community. In Peiperl, M., M. B. Arthur, Goffee, R., & Morris, T. (Eds.), Career frontiers: New conceptions of working lives. Oxford: Oxford University Press.

Parker, P., Khapova, S. N., & Arthur, M. B. (2009). The intelligent career framework as a basis for interdisciplinary inquiry. Journal of Vocational Behavior, 75, 2009, 291–302.

Parker, S. C. (2004). The economics of self-employment and entrepreneurship. Cambridge, CA: Cambridge University Press. Parker, S. C., Storey, D., & van Witteloostuijn, A. (2010). What happens with gazelles? The role of dynamic management strategies. Small Business Economics, 35, 203–226.

Parker, S. C., & van Witteloostuijn, A. (2010). General framework for multidimensional contingency fit. Organization Science, 21, 540–553.

Pennings, J. M., Lee, K., & van Witteloostuijn, A. (1998). Human capital, social capital, and firm dissolution. Academy of Man- agement Journal, 41, 425–440.

Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and prospects. Journal of Manage- ment, 12, 531–544.

Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903.

Porter, M. E. (1980). Competitive strategy. New York, NY: Free Press. Quinn, J. B. (1992). Intelligent enterprise: A knowledge and service based paradigm for industry. New York, NY: Free Press. Raabe, B., Frese, M., & Beehr, T. A. (2006). Action regulation theory and career self-management. Journal of Vocational Behavior, 70, 297–311.

Rodrigues, R. A., & Guest, D. (2010). Have careers become boundaryless? Human Relations, 63, 1157–1175. Sandberg, W. R., & Hofer, C. W. (1987). Improving new venture performance: The role of strategy, industry structure, and the entrepreneur. Entrepreneurship Theory and Practice, 16, 73–90.

Saucier, G. (1994). Mini-markers: A brief version of Goldberg’s unipolar big-five markers. Journal of Personality Assessment, 63, 506–516.

Schneider, B., & Konz, A. M. (1989). Strategic job analysis. Human Resource Management, 28, 51–64. Scott, M., & Bruce, R. (1987). Five stages of growth in small business. Long Range Planning, 20, 45–52. Seibert, S. E., Kraimer, M. L., & Liden, R. C. (2001). A social capital theory of career success. Academy of Management Journal, 44, 219–237.

Sexton, D. L., & Bowman, N. B. (1986). Validation of a personality index: Comparative psychological characteristics analysis of female entrepreneurs, managers, entrepreneurship students, and business students. In R. Ronstadt, J. A. Hornaday, R. Peterson, & K. H. Vesper (Eds.), Frontiers of entrepreneurship research, (pp. 40–51).Babson College Entrepreneurship Research Con- ference (BCERC). http://digitalknowledge.babson.edu/fer/

Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9, 221–232.

Stam, S., Bosma, N., van Witteloostuijn, A., Bogaert, S., Edwards, N., Jaspers F., & de Jong, J. (2012). Ambitious entrepreneur- ship: A review of the state of the art. Brussels/The Hague: VRWI/AWT.

Tams, S., & Arthur, M. B. (2010). New directions for boundaryless careers: Agency and interdependence in a changing world. Journal of Organizational Behavior, 31, 629–646.

Venkatraman, N., & Camillus, J. C. (1984). Exploring the concept of “fit” in strategic management. Academy of Management Review, 9, 513–525.

Witt, P. (2004). Entrepreneurs’ networks and the success of start-ups. Entrepreneurship and Regional Development, 16, 391–412. Zajac, E. J., Kraatz, M. S., & Bresser, R. K. F. (2000). Modeling the dynamics of strategic fit: A normative approach to strategic change. Strategic Management Journal, 21, 429–453.

46 A. VAN DEN BORN AND A. VAN WITTELOOSTUIJN

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.