ABSTRACT
SUCCESS IN THE PRE-START-UP PHASE
VARIABLES
DESIGN AND SAMPLE
FOLLOW-UP RESULTS AND FURTHER DATA CONSIDERATIONS
COMPARING THE THREE GROUPS
DISCUSSION
ACKNOWLEDGMENT
CONTACT
REFERENCES
TABLE 1
TABLE 2
TABLE 3
TABLE 4
TABLE 5

Why does one person actually succeed in starting a business, while a second gives up, and a third is still trying? To answer this question, a longitudinal study was set up in which 330 nascent entrepreneurs (people setting up a business) were followed over a one-year period. After one year, 47% actually started a business, 27% was still organizing, and 26% gave up the effort. In comparison to people who gave up starters are entrepreneur already, have more industry experience, start out with less start-up capital, use less of third party loans, and start out in manufacturing. In comparison to people still organizing starters are relatively often male, entrepreneur, and want to start full-time.

Explaining firm performance is an important part of entrepreneurship research (Cooper and Gascon, 1992; Lussier, 1995; Honig, 1998; Boden and Nucci, 2000; van Gelderen, Frese and Thurik, 2001). Most research deals with the success of existing firms. However, the first success of a firm is that it becomes one. Entrepreneurs have frequently been compared with non-entrepreneurs (Baron, 1999; Kaufman, 1999), but not often with persons who wanted to start a business but did not succeed in doing so. Why does one person actually succeed in starting a business, while a second gives up, and a third is still busy organizing? Answers to this question are directly relevant for practitioners who want to evaluate their own prospects, chances and behavior. For example, in one of the few studies on the subject, Carter, Gartner and Reynolds (1995) report that both individuals who started their business as well as individuals who gave up the start-up effort undertook more activities to make their business real. People who were still trying to set up their business had undertaken fewer activities than the other two groups. Therefore, the authors advice individuals considering a business start-up to pursue opportunities aggressively in the short term, in order not to find themselves perennially still trying. Comparisons of nascent entrepreneurs who start, still try, or give up are also relevant for governmental agencies that deal with nascent entrepreneurs. Research on pre-startup failure variables gives insight into the factors that hinder aspiring founders from realizing their plans. This knowledge can guide policy measures that improve the general conditions surrounding start-ups, thus enabling a more effective use of the nascent entrepreneurs’ potential (Chini, Frank, Korunka, and Lueger, 2000). Research by Chini et al. (2000) points to the importance of information use and availability. They found that people who had abandoned their start-up effort frequently indicated that information was unavailable or discouraging. Therefore, governmental agencies are heeded to make stimulating information and guidance available.

Finally, knowledge of the behavior of nascent entrepreneurs is important for those involved in creating and maintaining policy measures on a macro-economic level. The level of entrepreneurship, i.e., the number of business owners per work force, differs considerably across countries and periods (Thurik, 1999; Carree and Thurik, 1999). Both the causes and consequences of variation in the level of entrepreneurship are a matter of extensive scientific debate as well as of great policy importance. A high level of entrepreneurial activity is assumed and shown to contribute to innovative activities, competition, economic growth and job creation (Baumol, 1993; Thurik, 1996; Audretsch and Thurik, 2000 and 2001; Carree, van Stel, Thurik and Wennekers, 2001). For European countries in particular the fragile economic growth, coupled with the persistently high levels of unemployment, has fostered entrepreneurship (OECD, 2000). Many governments now seek to promote entrepreneurship, and high hopes are attached to entrepreneurship as a source of job creation and economic growth (Thurik, 1996). The exploitation of economies of scale and scope is no longer at the heart of modern economies (Teece, 1993; Wennekers and Thurik, 1999). The reason is that globalization and the ICT-revolution imply a need for a knowledge intensive economy. Such an economy emerges only after significant structural change, requiring a substantial reallocation and reorganization of resources. This induces an intense demand for entrepreneurship (Casson, 1995, Audretsch and Thurik, 2000 and 2001). When it comes to how the mechanisms work, little is known, either on how entrepreneurship can best be promoted or on how entrepreneurship influences economic performance. Promotion of entrepreneurship starts with insight in the motives and behavior of those seriously playing with the idea of becoming one.

To establish the differences in characteristics between these three groups, a number of independent variables were measured which are listed in Table 1. We consider pre-startup performance as a function of the personal characteristics of the entrepreneur (ENT), the environment (ENV), strategy (STR), and resources (RES). We propose the following functional relationship: pre-startup performance = f(ENT,ENV,STR,RES) (see also Chrisman, Bauerschmidt and Hofer, 1998). Specifically, demographics and experience are personal characteristics of the entrepreneur; industry sector and technology are considered characteristics of the environment; ambition and approach are part of strategy; and capital and third partly loans can be considered as resources. As few previous studies into success factors in the pre-start-up phase have been done, we made no predictions for the independent variables at this point. Due to space limitations, we will not further discuss theoretical development at this point.

Amongst other reasons, research of success and failure in the pre-start-up phase is scarce because of the lack of a representative sample (Reynolds and Miller, 1992; Reynolds, 1997). People walking around with an idea of starting a business are difficult to find. Of course, researchers may collect a sample of starting entrepreneurs and question them about their preparation phase retrospectively. However, in such an approach all people who did not succeed in getting a business started will be overlooked (survivor bias). Moreover, retrospective questioning may lead to biased memories (hindsight bias). To avoid survivor bias and hindsight bias, one has to collect a sample of nascent entrepreneurs, i.e., people who are in the process of setting up a business. For example, the researcher may collect a sample of nascent entrepreneurs from among people who take a course in setting up a business at the local Chamber of Commerce. However, the people who take part in such a course may form a biased sample. For example, ethnic minorities are less likely to participate in the regular information and guidance channels. Therefore, as a third desirable characteristic of a research design on success in the pre-start-up phase, one would not only want to avoid survivor and hindsight bias, but also draw a representative and random sample (Katz and Gartner, 1988). To this purpose, Paul Reynolds of Babson College has set up the Entrepreneurial Research Consortium (ERC). The ERC is an international research effort (joined among others by the United States, Sweden, Norway and the Netherlands) in which nascent entrepreneurs are collected by randomly calling phone numbers. The person who answers the phone is asked: are you currently, alone or with others, setting up a business? If the person answers affirmatively, two exclusions are made. First, it is essential to have an active and manifest desire to set up a business. If he or she is only dreaming about starting up a business, he or she is considered a potential entrepreneur instead of a nascent entrepreneur. Second, someone who has set up a business that is already operational, even though in a start-up phase, must be considered an entrepreneur instead of a nascent entrepreneur. By this design, a relevant, representative and random sample of nascent entrepreneurs is created avoiding the traps of survivor bias and hindsight bias.

In the fall of 1998, 49,936 phone numbers were dialed. These phone numbers were drawn randomly from our national phone directory. An interview was held with 21,393 persons (43%) aged between 18 and 65 years. The non-response was caused by refusals (14K), age constraints (below 18, above 65) (10K) and other reasons such as answering machines, infotones, etc (4,5K). Potential entrepreneurs were discerned from nascent entrepreneurs by asking whether one had been actively and seriously setting up the business in the past year. Out of the 21.393 interviewed, 526 were marked as nascent entrepreneurs. So the prevalence rate in our sample was 2.5%, which is indicative of the prevalence rate within the Dutch population between 18 and 65 years old. This prevalence rate is comparable with Scandinavian countries but much lower than in the United States (Delmar and Davidsson, 2000). This indicates that in comparison to the U.S. setting up a business in Scandinavia and in Holland is still to be considered a much more extraordinary event. In comparison with a control group (N=586) taken from the 21,393 persons who stated not to be currently setting up a business, the sample of nascent entrepreneurs was relatively male, young, had followed higher education and earned a higher income (van Gelderen, 1999).

In the follow-ups held among the sample of nascent entrepreneurs, the current status of the start-up effort was assessed. The actual question used is: How would you classify your firm? Is it (1) operational and running; (2) are you still setting up the business; (3) have you temporarily delayed your start-up effort; (4) have you completely abandoned your start-up effort. After one year, the current status of the start-up effort was known of 330 nascent entrepreneurs. Non-response was thus substantial. T-tests analyzing the differences with the response group showed no differences with regard to the independent variables. In the follow up sample of 330 persons, 47% started their business, 27% were still organizing, and 26% had abandoned the effort. The non-response causes these figures to be tentative, as we can not indicate whether these people dropped out, started or still organizing. Also it has to be taken into account that the sample is the unit of analysis. If we want to know what the success percentages are of people starting in a certain time period, for example a year, the numbers of people starting and quitting will be higher, and the number of people still organizing lower. This has two reasons: first, people still organizing are overrepresented in the sample as some of them already began setting up in previous time periods. Second, for a number of people the total period of being a nascent entrepreneurs is shorter than six months, so one would need to have several screening moments in a certain year.

In this particular paper it is the entrepreneur himself who defines whether his business is actually started or still in the start-up phase. This implies that entrepreneurs can use different criteria to judge whether they consider themselves started or not. In fact, the question why a nascent entrepreneur considered himself started gave rise to a plethora of answers. In Table 2 these answers are classified using the properties of emerging organizations given by Katz and Gartner (1988). So when interpreting the results, one has to bear in mind that there is an underlying heterogeneity in the performance measure. We choose to use the subjective (nascent entrepreneurs) definition in this paper just because of this heterogeneity. Using a uniform objective (researchers) definition has the disadvantage of forcing unity on a situation that is in fact very diverse. In a different study using this data set, we did work with theory driven measures of whether a business actually started, resulting in partly different explanatory success factors (van Gelderen, 2001).

In our analyses, groups (2) and (3) are taken together and classified as the group “still organizing” because of the reasons people gave for classifying themselves as pausing their start-up efforts (like waiting for a license). The frequencies of the independent variables are given in Table 3. Five continuous variables (personal income, work experience, management experience, industry experience, and desired start-up capital) were recoded into categories to mitigate the effects of very large numbers. Also, the categories become larger as the average value of the categories increases in order to reflect diminishing marginal returns. Age was recoded into categories to obtain insight into the relations of the different age categories with the other variables. As can be seen in Table 3, most independent variables had some missing data, most notably personal income and desired number of personnel in five years. For the multivariate analyses, which were done using a multinomial logistic regression technique, an expected maximization procedure was executed to replace missing data based on underlying data patterns, while keeping means and standard deviations constant. The EM procedure did not drive the results, as is indicated by calculations on the dataset including missing values as well as by the similarity between univariate and multivariate results. Industry sector (manufacturing, trade, business services, consumer services) and daily activity status (employee, entrepreneur, social welfare, student) were recoded into dummy variables. Employee status was left out of the model because it took up 65% of the variable “daily activity” (see Table 3), leaving identification problems for the other dummies representing daily activity.

Very few non-biased samples of entrepreneurs in the pre-startup phase exist. Table 3 provides detailed descriptive statistics on what was until now unknown territory. It is striking that while in comparison to a control group the nascent entrepreneurs are relatively highly educated and earn a high income, only a minority of them prefers to grow large, to become rich, to start full time, and to use a third party loan. These findings point to a tendency of people of higher social strata to start a business next to their former activities. The relationships of the independent variables with performance (started, still organizing, abandoned effort) are analyzed both in a univariate framework and a multivariate one. Univariate analyses are done using simple chi-square analyses, as the dependent variable consists of three categories. Table 3 gives the results of the chi-square statistics as well as the frequencies per success category. We find significantly more females and people with the intention to start part-time that are still busy organizing. These two groups are highly correlated, as can be seen in Table 3. Moreover, the two dummy variables “manufacturing” and “regarding oneself as an entrepreneur” are highly discriminative between the categories of “actually started” and “abandoned”. Industry experience is a success factor, as opposed to other types of experience, but only up to a certain amount of years. Starting out without making a loan is a highly significant success factor, as opposed to wishing to start out with a large start-up capital.

Most of these results emerge also in a multinomial logistic model presented in Table 4. This type of regression is similar to logistic regression but more general because the dependent variable is not restricted to two categories. The vector coefficients do not represent an absolute effect but the marginal effect of an explanatory variable on the probability of “abandoned” and “still organizing” relative to the probability of “actually started” (Cooper, Gimeno-Gascon, and Woo, 1994; Long, 1997). In Table 4 the nascent entrepreneurs that actually started serve as a benchmark group for the persons who gave up (first column) and for the persons who were still setting up their business (second column). A comparison between the nascent entrepreneurs that abandoned their start-up effort and entrepreneurs still organizing is not presented, as no significant differences are found. When distinguishing between nascents that actually started and nascents still organizing, we again find females and part-timers still setting up, and entrepreneurs being less likely to be still organizing. As a success factor of nascent entrepreneurs who finally started, again the following factors emerge: manufacturing, regarding oneself as entrepreneur, industry experience and using own money. Start-up capital loses its significance due to the non-linearity of its relationship with performance, as can be seen in Table 3.

Goodness-of-fit is measured in a manner similar to Cooper, Gimeno-Gascon and Woo (1994). For logistic regression models, a straight R² statistic is not available. Some alternatives pseudo R² measures have been calculated. The Nagelkerke R² equals 0.306, whereas the McFadden equals 0.148. A common measure for determining the fit of the model in these kinds of applications is the Hosmer and Lemeshow test (Hosmer and Lemeshow, 1989), where the probability of an outcome is specified rather than the actual occurrence of an outcome. For all three categories the test did not point at rejection of the hypothesis that the model fits well (the cases were divided into 10 subgroups of 33 observations each). The p-values associated with the chi-square test were 0.58, 0.24 and 0.22 for respectively abandoned, still trying and started. Given that the nascent entrepreneurs who are still trying are placed in a temporary category (every person in this category should ultimately belong to the category “abandoned” or “started” and the timing for transfers into one these two categories may therefore be important), we conclude that our model fits the data reasonably well.

The variables connected with “giving up,” “abandoned” respectively “still organizing” do not necessarily coincide with the reasons given by the respondents when asked why they had given up their business respectively what remained to be done before they would get started (Table 5). The main reason given for abandoning the start-up effort was the opportunity offered by a job. Of course, the choice for another job might be influenced by difficulties in the start-up process. Obtaining appropriate finance seems the major bottleneck of the people still busy organizing.

Characteristics of nascents, i.e., people who are in the process of setting up a business, are hardly dealt with in the area of entrepreneurship research. Our results must be seen as an empirical step that needs to be followed up by a more in-depth theoretical approach that investigates the entire underlying process. Apart from generating a large number of descriptive statistics, the present study sheds light on the impact and relative importance of some explanatory variables connected with the pre-start-up phase. Our results lead to some intriguing questions. We give three examples. First, women need more time to actually start up a business. Is this a question of difficulties in obtaining access to resources or of differing values (Brush, 1992; Fischer, Reuber, and Dyke, 1993; Verheul and Thurik, 2001)? The strong correlations between being male and management and industry experience, respectively point to the first position, while the strong correlations between between being female and part-time business ownership point to the second position.

Second, we find that a third party loan and a higher start-up capital are variables connected with failure in the nascent phase. This indicates a difference between the pre-startup phase and the post-startup phase, as it has repeatedly been shown that capitalization is an important success factor in the post-startup phase. The question is whether the selection process that takes place in the pre-startup phase is healthy or unhealthy. Does the group of nascents that want to start out large consist of relatively many dreamers, who are rightfully rejected by banks and other financiers? Or do these people calculate their prospects carefully and then either start or back off (Carter, Gartner, and Reynolds, 1995)? Or do the financial markets in the Netherlands lack opportunities for nascent entrepreneurs? In any case, for many nascent entrepreneurs it is beneficial to start out modestly.

Third, a striking dissimilarity between pre-start-up and post-start-up has to do with experience. It is puzzling that industry experience is a success factor, while work experience, management experience, and experience in setting up a business as well as education are not. Particularly management experience has been repeatedly shown to affect post-start-up performance (Lussier, 1995). Can the result that having knowledge of the industry and a network in the market is decisive be replicated, and why would this result emerge? Perhaps knowledge of an industry and a network in a market are crucial for actually starting a business, while after start-up management experience takes over in importance. As industry experience is significantly correlated with age, it might be that industry experience opens a strategic window for older people to set up a business (Harvey and Evans, 1995).

The present study has a number of weaknesses and limitations. In improved and extended versions of this paper we will present an extended theoretical framework as well as results based on uniform performance measures. Other limitations are likely to remain. First, in survey research one is limited to variables that are easily accessible. This does not mean that these variables are necessarily the most important variables (Cooper, 1993). The skills, knowledge and motives of nascents are not directly accessed. Also the so-called “how” variables (vanderWerf, 1989) are not taken into account, for example how resources are developed, how relationships are maintained, and how information is gained (Cooper, 1993). Second, as Table 5 indicates, there is only a partial connection between the success and failure factors in our model on the one hand and reasons actually given by people themselves as to why they have abandoned or why they were still busy organizing on the other. Of the four reasons that are usually given for why people abandon their start-up effort, three are not measured in our model. A good job offer, unfavorable outcomes of market research, and private reasons could be taken into account in further modeling of pre-startup performance. The same reasoning applies to the actual reasons given by people why they were still busy organizing. Third, our analyses of success and failure factors provide a general picture only. This limits the practical relevance, as it is well known that there is a large variety in types of ventures and types of entrepreneurs. So when analyzing specific types of entrepreneurs, more detailed pictures of factors connected with success and failure emerge that might very well deviate from the general picture. Of course, analyses of the success factors for specific types of entrepreneurs would require a larger or more specific sample.

Government policy in the old, managed economy was largely about control. High certainty dictated that it was known what to produce, how it should be produced, and who would produce it. The role of government was to constrain the power of large corporations, which were needed for efficiency under mass-production, but posed a threat to democracy through their concentration of power (Chandler, 1977 and 1990). Under the old, managed economy the policy debate centered on competition policies (antitrust), regulation and public ownership of business (Teece, 1993). In the new, entrepreneurial economy these constraining policies have become increasingly irrelevant. The central role of government policy in the new, entrepreneurial economy is enabling in nature. The focus is to foster the production and commercialization of knowledge. Rather than focus on limiting the freedom of firms to contract through antitrust, regulation and public ownership, government policy in the new, entrepreneurial economy targets education, increasing the skills and human capital of workers, and facilitating the mobility of workers and their ability to start new firms (Audretsch and Thurik, 2001). Knowledge of their motives and behavior in the pre-start-up phase is essential for creating a portfolio of new enabling policies. Therefore, we believe that efforts to understand predictors of pre-start-up performance will become an important part of entrepreneurship research. The present study is one of the first to contribute to this new area. We hope the simple model described here will encourage the work yet to be done.

The design of this research was created by the Entrepreneurial Research Consortium (ERC) under direction of Paul Reynolds.

CONTACT:Marco van Gelderen, Vrije Universiteit Amsterdam, Faculty of Economics, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands; (T) +31 20 4446123; mgelderen@econ.vu.nl

Audretsch, D.B., and Thurik, A.R. 2000. Capitalism and democracy in the 21st century: from the managed to the entrepreneurial economy. Journal of Evolutionary Economics, 10(1): 17–34.

Audretsch, D.B., and Thurik, A.R. 2001. Sources of growth: the entrepreneurial versus the managed economy. Industrial and Corporate Change, forthcoming.

Baron, R.A. 1999. Counterfactual thinking and venture formation: The potential effects of thinking about “what might have been.” Journal of Business Venturing, 15: 79–91.

Baumol, W.J. 1993. Entrepreneurship, Management and the Structure of Payoffs. Cambridge, MA: MIT-Press.

Bijleveld, C.C.J.H., and van der Kamp, L.J.Th. 1998. Longitudinal Data Analysis: Designs, Models, and Methods. London: Sage.

Boden, R.J., and Nucci, A.R. 2000. On the survival prospects of man’s and women’s new business ventures. Journal of Business Venturing, 15: 7–362.

Carree, M.A., Stel, A.J. van, Thurik, A.R., and Wennekers, A.R.M. 2001. Economic development and business ownership: an analysis using data of 23 OECD countries in the period 1976–1996. Small Business Economics, forthcoming.

Carree, M.A., and Thurik, A.R. 1999. Industrial structure and economic growth. In: D.B. Audretsch and A. R. Thurik eds., Innovation, Industry Evolution and Employment. Cambridge, UK: Cambridge University Press, 86–110

Carter, N.M., Gartner, W.B., and Reynolds, P.D. 1995. Exploring start-up event sequences. Journal of Business Venturing, 11: 151–166.

Casson, M. 1995. Entrepreneurship and Business Culture; Studies in the Economics of Trust, Vol. 1. Cheltenham, UK: Edward Elgar.

Chandler, A.D. 1977. The Visible Hand: the Managerial Revolution in American Business, Cambridge: Harvard University Press.

Chandler, A.D. 1990. Scale and Scope: the Dynamics of Industrial Capitalism, Cambridge: Harvard University Press.

Chini, L., Frank, H., Korunka, C., and Lueger, M. 2000. Fouding a business—A configurational perspective. Proceedings of RENT conference, Prague, 23/24–11.

Chrisman, J.J., Bauerschmidt, A., and Hofer, C.W. 1998. The determinants of new venture performance: An extended model. Entrepreneurship Theory and Practice, 23(1): 5–22.

Cliff, J.E. 1998. Does one size fit all? Exploring the relationships between attitudes towards growth, gender, and business size. Journal of Business Venturing, 13: 523–542.

Cooper, A.C. 1993. Challenges in predicting new firm performance. Journal of Business Venturing, 8: 231–253.

Cooper, A.C., and Gimeno-Gascon, F.J. 1992. Entrepreneurs, processes of founding, and new-firm performance. In: Sexton, D.L., and Kasarda, J.D. ed.. The State of the Art in Entrepreneurship. Boston: PWS-Kent.

Cooper, A.C., Gimeno-Gascon, F.J., and Woo, C.Y. 1994. Initial human and financial capital as predictors of new venture performance. Journal of Business Venturing, 9: 371–395.

Delmar, F., and Davidsson, P. 2000. Where do they come from? Prevalence and characteristics of nascent entrepreneurs. Entrepreneurship and Regional Development, 12: 1–23.

Fischer, E.M., Reuber, A.R., and Dyke, L.S. 1993. A theoretical overview and extension of research on sex, gender, and entrepreneurship. Journal of Business Venturing, 8: 151–168.

Gelderen, M.W. van 1999. Ontluikend Ondernemerschap. Zoetermeer: E.I.M.

Gelderen, M.W. van 2001. Ondernemers vóór de start: definitiekwesties en succesfactoren. Bedrijfskunde, 73, 50–56.

Gelderen, M.W., Frese, M. and Thurik, A.R. 2001. Strategies, uncertainty and performance of small startups. Small Business Economics, 15: 165–181.

Harvey, M., and Evans, R. 1995. Strategic windows in the entrepreneurial process. Journal of Business Venturing, 10: 331–347.

Honig, B. 1998. What determines success? Examining the human, financial, and social capital of Jamaican microentrepreneurs. Journal of Business Venturing, 13: 371–394.

Hosmer, D.W., and Lemeshow, S. 1989. Applied Logistic Regression. New York: John Wiley and Sons.

Katz, J., and Gartner, W.B., 1988. Properties of emerging organisations. Academy of Management Review, 13: 429–441.

Kaufman, P.J. 1999. Franchising and the choice of self-employment. Journal of Business Venturing, 14: 345–362.

Long, J.S. 1997. Regression Models for Categorical and Limited Dependent Variables. London: Sage.

Lussier, R.N., 1995. A nonfinancial business success versus failure prediction model for young farmers. Journal of Small Business Management, 33: 8–20.

OECD 2000. OECD Employment Outlook. OECD: Paris.

Reynolds, P., and Miller, B. 1992. New firm gestation: Conception, birth, and implications for research. Journal of Business Venturing, 7: 405–417.

Reynolds, P.D. 1997. Who starts new firms?—Preliminary explorations of firms-in-gestation. Small Business Economics, 9: 449–462.

Teece, D.J. 1993. The dynamics of industrial capitalism: perspectives on Alfred Chandler’s scale and scope. Journal of Economic Literature, 31: 199–225.

Thurik, A.R. 1996. Small firms, entrepeneurship and economic growth. In P.H. Admiraal, ed. Small Business in the Modern Economy. Oxford: Basil Blackwell Publishers, 126–152.

Thurik, A.R. 1999. Entrepreneurship, industrial transformation and growth. In G.D. Libecap, ed.. The Sources of Entrepreneurial Activity: Vol. 11, Advances in the Study of Entrepreneurship, Innovation, and Economic Growth. Stamford, CT: JAI Press, 29–65.

VanderWerf, P.A. 1989. Achieving empirical progress in an undefined field. Entrepreneurship Theory and Practice, 14: 45–58.

Verheul, I., and Thurik, A.R. 2001. Start-up capital: differences between male and female entrepreneurs: “does gender matter?” Small Business Economics, forthcoming.

Wennekers, A.R.M., and Thurik, A.R. 1999. Linking entrepreneurship and economic growth. Small Business Economics, 13: 27–55.