We investigate the manifold posed question: “To what extent does investment in human and social capital, besides the effect of “talent,” enhance entrepreneurial performance?” We distinguish three different performance measures: survival, profits, and generated employment. On the basis of the empirical analysis of a rich Dutch longitudinal data set of firm founders, we conclude that specific investments indeed affect the three performance measures significantly and substantially. We explore some alternative explanations but find that there is no reason to alter this conclusion.

Many authors advocate that, besides talent and good luck, specific types of investment in human capital and/or social capital might enhance entrepreneurial performance (See for example Van Praag 1999, Brüderl and Preisendörfer 1998, Pennings et al. 1998, Cooper et al. 1994. Hence, the objective of this paper is to find a qualified answer to the important and manifold posed question: To what extent does investment in human and social capital, besides the widely believed determining effect of “talent,” enhance entrepreneurial performance? We distinguish three types of investment in both human and social capital: general, industry-specific and entrepreneurship-specific investment. We will empirically investigate which of these types of initial investment, if any, contributes mostly to the performance of the small business founder, our empirical equivalent of the entrepreneur. The empirical analysis is based on a representative panel survey amongst almost 1,000 new business founders in the Netherlands in the years 1994–1997. We employ three measures of performance: survival, profit, and generated employment.

Our main finding is that the endowed level of talent of a small business founder is not the unique determinant of performance. Rather, investment in industry-specific and entrepreneurship-specific human and social capital contributes significantly to the explanation of the cross-sectional variance of the performance of small firm founders. To formulate more precisely: industry-specific investments in human capital such as experience in the same industry enhance performance, whichever measure of the latter is used. In addition, human and social entrepreneurship-specific capital investments, i.e. former experience in starting up a business, the membership of an association for small business founders generate more promising start-ups too.

This paper attempts to further our understanding about effective general and specific investments in human and social capital of small business founders. We thereby seek to contribute to a policy-relevant ongoing debate in the entrepreneurship literature. Empirical studies into the effect of human (Cooper et al., 1994; Van Praag and Cramer, 2001) and social (Brüderl and Preisendörfer, 1998; Pennings et al 1994) capital on the performance of business founders are no novelty. The same holds for the distinction between the effect on performance of general and specific investments (Cooper et al., 1994; Pennings et al., 1994) in human and social capital. The key contribution of this study is twofold. First, we systematically test the effect of general, industry-specific and entrepreneurship-specific human and social capital on three distinct performance measures, while controlling for a rich set of other factors. Secondly, we explicitly analyze the meaning of our result by exploring alternative explanations that are often not considered (for example, Pennings, 1998; Cooper et al., 1994)

It is argued that the firm founder’s performance is partly determined by his human, social, and financial capital, and additionally by his talent, the circumstances and, good luck. We measure the impact of human, social and financial capital on the firm’s performance, while we control as much as possible for the other factors mentioned. Our focus is on the impact of human and social capital. We are particularly interested in the entrepreneurship-specific and industry-specific investments in these types of capital.

Human Capital

The supposedly positive impact of human capital on employee performance is well accepted and was formalized by Mincer (1974). Van Praag and Cramer (2001) made a first attempt to formalize this for the business founder’s case. Several authors, including Cooper et al. (1994), Pennings et al. (1998), Van Praag (2001, 1999, 1996) as well as De Wit and Van Winden (1993) have put forth empirical support for the theoretical foundation.

Hypothesis 1. Higher levels of human capital are associated with higher performance by business founders

Social Capital

The impact of social capital on a business founder’s performance has not yet been “theorized.” However its impact has empirically been supported by Blumberg and Pfann (1999), Brüderl et al. (1998), and Pennings et al 1998. A theory of investment in individual social capital has recently been developed by Glaeser et al. (2000). For the benefit of identifying determinants of social capital formation, Glaeser et al. first seek to understand the social capital investment decision of individuals. This understanding leads to seven propositions of which a subset is relevant: (i) a rational individual’s investment in social capital is higher in occupations with greater returns to social skills, (ii) social capital declines with expected mobility, (iii) people who invest in human capital also invest in social capital. Since small business founders need to have stronger or weaker ties with all the (prospective) stakeholders in their firm, such as clients, investors, debtors and, subcontractors, the expected benefits pertaining to social skills are high in the “occupation” of business founding. We furthermore know that business founders are less mobile than employees are (Blanchflower, 2000). This implies by Glaeser’s proposition (ii) that the expected returns to social capital are higher than average for business founders. The third of Glaeser’s propositions implies that individuals for whom the net return to human capital is high also might expect a relatively high net return to social capital. Together, these three propositions imply that social capital should enhance the performance of a business founder.

Hypothesis 2. Higher levels of a business founder’s social capital are associated with greater performance of the firm.

Productivity and Signaling

The impact of human and social capital on the performance of the firm is likely to be caused by two underlying forces: productivity and signaling. (1) The more capital is embedded in the firm’s founder, the greater is the firm’s productivity, and (2) Prospective stakeholders such as clients, subcontractors, and investors are likely to have imperfect information about the type of firms they consider dealing with. Therefore, these parties evaluate firms (and the prices involved in for instance buying their services) based on observable characteristics that they presume to be correlated with the unobserved type/quality of the firm and its founder. Investment in social and human capital might be a signal (Spence, 1974) as long as a high investment in these types of capital is the preferred choice of talented people, whereas a low level of investment is the most preferred choice for people with less talent (Milgrom and Roberts, 1992). We shall come back to this interpretation when discussing the self-selection interpretation.

Specificity of Investments

We differentiate the firm founder’s capital investments with respect to the specificity of the investment. It is well known both from human capital theory (Becker, 1964) and from resource-based theories of the firm (Montgomery, 1995) that the more specific an investment is to its current use, the higher the expected returns, i.e. its contribution to firm performance, should be. According to the human capital theory, the returns to a deliberate specific investment in a current activity should be sufficiently large to outweigh the cost attached to the investment. Contrarily, the returns to an investment in a more general asset might accrue to the investor during a longer period of time while performing various activities. Moreover, according to resource-based theories of the firm, success of firms is determined by the extent to which Ricardian rents are earned by the accumulation and deployment of non-imitable resources. The probability of resources to be non-imitable is much higher when these are obtained through specific investment than when they are obtained by more general investment. We contrast general investments with two types of specific investments in human and social capital: industry-specific next to entrepreneurship-specific investments. Industry-specific investments lose (part of) their value outside the industry in which the business venture is started, whereas entrepreneurship-specific investment loses its return outside the entrepreneurial environment. Our next hypothesis results:

Hypothesis 3. Industry-specific and entrepreneurship-specific investments in the business founder’s human and social capital are more influential on firm performance than are general investments in human and social capital.

Self-Selection and Unobserved Heterogeneity

Theoretically, investment behavior is in line with the comparative advantages hypothesis: individuals who experience more than average net gains from a particular investment, are more likely to choose that investment option. In other words, individuals self-select themselves into certain “treatment groups.” The comparative advantage underlying the type of investment decisions that we study depends on individual characteristics that partly remain unobserved. The unobserved part of these characteristics about which the investor has private information, such as talent and intelligence, however, probably not only affects the outcome of the investment decision, but also the performance of the business venture.

From a more technical point of view, the above implies that the unexplained part of the cross-sectional variation in business founders’ performance is correlated with the levels that these individuals invest in human and social capital. This correlation can be positive or negative. Most probably, the cost pertaining to the investment is lower to higher ability business founders than to lower ability business founders. Therefore, given a fixed difference in benefits accruing to the business founder of either ability level, a positive correlation between the error term and the investment levels in human and social capital is expected. If the level of benefits has the same association with the unobserved factors as business performance has, the correlation is surely positive. However, if lower ability business founders expect higher benefits from their investment than do higher ability entrepreneurs, because the latter group might judge that they “don’t need” these investments and, if this effect is stronger than the opposite effect at the cost-side, a negative correlation between the error term and the investment might result.

Summarizing, failing to implement the deliberate investment decision in the empirical model could generate a misspecified model with biased estimates.

Hypothesis 4. The choice of firm founders for specific investments is partly dependent on unobserved talent of the firm founders. This would lead to biased estimates of the effect of these investments.

“Knowledge Industries”

The relative contribution to production of human capital and knowledge has rapidly increased over the past decade (Audretsch and Thurik, 2000). This is particularly the case in the so-called “knowledge industries,” such as the ICT industry. Hence, one might argue that specific investments in human and social capital are particularly relevant in these “knowledge industries.”

Hypothesis 5. Industry-specific and entrepreneurship-specific investments in the business founder’s human and social capital are more influential in the “knowledge industries” than they are in other industries.

Sample

The panel consists of annual questionnaires conducted on a sample of Dutch entrepreneurs that started their business in 1994. The sample was taken from all newly registered firms in the first quarter of 1994 as reported in the database of the Dutch Chamber of Commerce. Initially 10,627 firms were contacted by telephone. A total of 3,000 firm founders agreed to participate in the survey. Approximately 2,000 firm founders finally completed the 1994 questionnaire. From 1995, a questionnaire was sent to the remaining group of business founders. The 1997 questionnaire was completed by over 1,100 respondents, implying a cumulative attrition rate of 45%. The firm size and sector distribution of the 1994 and 1997 respondents were comparable to those of the initial sample.

As nearly half of the firm founders left the panel between 1994 and 1997 one could suspect that this could lead to biased results. We constructed an explicit model to investigate possible biases and found no significant ones.

The measures for entrepreneurial performance are constructed from the questionnaires in 1995, 1996 and 1997, whereas the possible determinants are derived solely from the 1994 questionnaire. In this way we prevent problems of reversed causality.

Measures of Entrepreneurial Performance

The dataset used provides three performance measures. The first one considers profit and is equated to the profit made in 1997. The entrepreneur has then been active for three years. The profit may especially in the first two years be somewhat misleading, as initial (sunk) costs often have to be gained back, which reduces profit. For entrepreneurs that are known to have terminated their businesses the profit variable is equated to zero. The second measure used is the cumulative employment created in the period 1994–1997.² While profit is mainly an individual performance measure, the employment created by an entrepreneur can be seen as a society performance measure. The third performance measure is the simple matter of survival. Is the firm still in business in 1997? And if not, how many months has it taken before the entrepreneur quit? In the firm founders panel, information is available on the survival time of the firms. We have constructed a variable measuring the number of months that a firm has been active. A survival model will be applied when investigating this performance measure.

Determinants of Entrepreneurial Performance

From the 1994 firm founders questionnaire, we derived possible determinants of performance as set out in Table 1. The variables are grouped horizontally by measure of specificity and vertically by the form of capital. Following the arguments made earlier, we especially focus on the five variables listed in italics in Table 1.

We include seven human capital variables. Experience of the business founder is measured in different dimensions: experience in business ownership itself, experience in activities related to business ownership (e.g. experience in leadership), and experience in the industry in which the founded business is active. For analyzing the effect of age (proxy for “knowledge of the world” as in Van Praag, 1999), the inclusion of age and age squared enables a nonlinear relation. Education enters the analyses as a dummy variable, differentiating the high-educated business founders from the less educated ones. Like in many other studies, a dummy indicating whether the respondent has experience as an employee is also included.

The three social capital variables are the following. If the business founder is engaged in an organized network of entrepreneurs (like a rotary club) this is indicated. Also, the emotional support of the business founder’s spouse may be important for entrepreneurial performance. Finally, the way the entrepreneur plans his way of information gathering may be influential for his performance. In the questionnaire ten possible actions were shown to the entrepreneurs. The respondents indicated whether they frequently, sometimes or never made use of these actions. Some of these actions are closely related. Factor analysis revealed four major strategies with respect to planned information gathering among the Dutch business founders:

1.  Focus on the branch in general (main actions: keep up with literature; attending congresses, courses and branch exchanges);
2.  Focus on direct business relations (main actions: information from customers and information from suppliers);
3.  Focus on commercial relations (main actions: information from banks; commercial cooperation; advise from experts);
4.  Focus on fellow entrepreneurs.

From the financial angle, five variables are distinguished. Three dummies indicate from which source the start-up capital was derived. First, a dummy is available indicating which firm founders financed their businesses 100 percent by themselves. Second, a dummy indicates if a loan was received from family members for setting up the business. A third dummy in this category reveals whether a business partner contributed financially in setting up the business. Also, the amount of the other income available in the household of the business founder may affect his performance. Thus, we include the amount of other income received by the business owner, as well the wage income of his spouse.

To complete the hypothesized performance indicators, six control variables are included. The inclusion of gender is common practice. Another dummy variable indicates whether the business founder considers his/her job as fulltime. Two control variables listed in Table 1 reveal performance expectations by the business owner at the start. One picks out the business founders who explicitly state that one of their goals is to generate employment. The other retrieves the entrepreneurs answering that attaining a higher income than their expected wage income was one of the motives to start the firm. Finally, there is a dummy variable measuring whether having entrepreneurs in the family was one of the reasons for founding the business and another dummy indicating whether boarding out of activities to external parties is planned.

Estimation Methods

Both the profit and the employment measure have zero as the lower bound. Negative profits are not observed, while negative employment is non-interpretable. Therefore, both equations are estimated using tobit regressions. For the expected survival time, we apply a different model than for profit and employment, viz. a survival model. In this model the (logarithm of the) expected survival time is modelled as a function of the characteristics of the entrepreneurs. For the underlying expected distribution of the probability to exit, we assume a log-logistic distribution (Lancaster, 1992, p. 8).

Estimation results are depicted in Table 2. From this table we will draw some conclusions concerning hypotheses 1–3 of section 2. We will also make some additional observations. Section 5 will deal with additional analyses meant to test the remaining hypotheses.

Human capital appears to influence our complete set of performance measures. To be more specific, former experience of the business founder in the industry in which he starts his business appears to improve all performance measures. Moreover, experience in business ownership itself leads to higher profits and employment, while experience in activities relevant to business ownership (e.g. experience in leadership) increases the firm’s survival time. Age appears to affect all performance measures. Finally, high-educated people make more profits, while those who have experience as an employee create more employment. Leaving out the human capital variables results in a model that according to the likelihood ratio tests performs significantly worse for all three performance measures. We may thus conclude that hypothesis 1—human capital influences the performance—is confirmed, though the specific aspects of human capital that influence each performance measure are not exactly alike.

Social capital also appears to influence performance. If business owners plan to gather their information via commercial relations, this improves our three performance measures. Furthermore, information gathering via general channels further increases the survival time and the generated employment, while information gathering via direct business relations further increases profits. Contact with other entrepreneurs in networks, e.g. rotary clubs, has a positive effect on the employment the business founder generates. The effect of these formal networks is insignificant on the other performance measures. Finally, the emotional support of a spouse appears also to be of importance: those who get it earn approximately 45% more than their fellow entrepreneurs who have to do without. All in all, we can conclude that there is sufficient support for hypothesis 2: social capital positively affects entrepreneurial performance.

The specific investments in human and social capital are printed in italics in Table 2. It appears that making specific investments matters in explaining performance. Exclusion of variables related to specific investments results in a model with a significantly lower likelihood than the model with these variables included. However, as can be seen from the table, general investments in human and social capital are also of importance—be it somewhat less.³ Thus, there is some support for hypothesis 3: specific investments are more influential than general investments.

With respect to the influence of financial capital, we find that those who used only their own capital for setting up their businesses generate approximately 70% less employment than their fellow entrepreneurs. Moreover, those using external capital due to the contribution of a business partner perform even better with respect to profits as well as with respect to generated employment. Finally, those business owners who already have another source of income perform worse. Clearly, they are not as dedicated to their business than their fellow entrepreneurs.

With respect to our controls, the following results are worth mentioning. Gender matters, at least for this particular sample of 1994 Dutch business founders. Male business founders perform better on all performance measures. Fulltime dedication to self-employment works out positive: it increases the survival time as well as the profits generated. Finally, people indicating that one of their goals is achieving employment growth, indeed generate more employment and people who indicated that a higher expected income was an important motive to become self-employed indeed made more profits.

In the previous section we found that specific investments of firm founders in human or social capital enhance their performance. However, as explained in section 2, without further explorations one cannot be sure whether this positive effect is solely due to the investment itself or partly due to the fact that more talented firm founders invest more in their human and social capital. In the latter case it would be wrong to assign the credits of the better performance solely to the investment. In other words, we would have found an upwardly biased effect.

This problem, widely recognized in the schooling literature (for example Ashenfelter et al., 2000), can be solved with instrumental variables estimation techniques. In this technique, the model is enhanced with an extra equation describing the investment decision of the firm founders in a particular form of human or social capital. The extra investment equation should contain at least one variable, the “instrument,” which (i) has a significant effect on the investment decision, while it (ii) cannot influence the performance of the firm founder on theoretical grounds. In this way it is possible to sort out the “real” investment effect. For, if the instrument is found to have an effect in the performance equation, it is then clear that this effect can only be due to the investment (because by assumption the instrument has no direct effect on performance).

Unfortunately, as is commonly known, it is notoriously difficult to find appropriate instruments. Although we tried various ones, we could not find satisfying instruments. So we cannot rule out potential biases in this way. Nevertheless, we do not suspect large biases because of the following two reasons. First, due to the fact that we have a rich data set we already control in our regression with a lot of proxies for talent. This leaves less room for unobserved talent correlated with the investment decision of the firm founder. Second, on theoretical grounds we do not suspect large biases for most of our investments variables. For example, we find that firm founders with experience in the industry in which they start their businesses, perform better. We think it is unlikely that this effect would be largely due to unobserved talent from these firm founders. Hence, though we are not able to reject hypothesis 4—there are biases because of unobserved talent—unreservedly, we are not afraid for large biases in our results because of it.

For exploring the role of the knowledge industries in explaining performance (hypothesis 5), dummies for these industries (i.e. business services and other services) were implemented. For the specific investment variables, cross products were generated and included in the analysis. However, the hypothesized industry effects did not emerge from the estimation results. On these grounds, we reject hypothesis 5.

We conclude that specific investments in human and social capital enhance entrepreneurial performance substantially. This is true for all three distinguished performance measures: survival, profits, and generated employment.

1. This paper has been written in the framework of the research programme SCALES (Scientific AnaLysis of Entrepreneurship and SMEs), which is financed by the Dutch Ministry of Economic Affairs. We acknowledge the support of Niels Bosma, Roy Thurik, and Hessel Oosterbeek.

2. Other employment measures have also been investigated. These include employment growth and the employment in 1997. Both measures produced results that did not differ significantly the result pertaining to the performance measure “cumulated employment.” We therefore included cumulated employment as the single employment measure in this paper.

3. More precise, exclusion of the variables related to general investments decreases the likelihood less than exclusion of the variables related to specific investments.

CONTACT: Mirjam van Praag, University of Amsterdam, The Netherlands; vanpraag@fee.uva.nl

Ashenfelter, O., H. Colm and H. Oosterbeek (1999) “A review of estimates of the schooling/earnings relationship, with tests for publication bias.” Labour Economics (6)4: 453–470.

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

Audretsch, D.B. and A.R. Thurik (2001) “What is new about the new economy: sources of growth in the managed and entrepreneurial economies.” Industrial and Corporate Change, forthcoming.

Becker, G. S. (1964), Human Capital, New York: Columbia University Press.

Blanchflower, D.G. (2000). “Self-Employment in OECD countries.” NBER Working Paper No. 7486.

Blumberg, B.F. and G.A. Pfann (1999) “Social Capital and the Self-Employment Decision,” unpublished manuscript Maastricht University, Business Investment Research Center.

Brüderl, J. and P. Preisendörfer (1998) “Network support and Success of Newly Founded Businesses.” Small Business Economics 10: 213–225.

Cooke, P. and D. Wills (1999) “Small Firms, Social Capital and the Enhancement of Business Performance Through Innovation Programs.” Small Business Economics 13: 219–234.

Cooper, A.C., F.J. Gimeno-Gascon, and C.Y. Woo (1994) “Initial Human and Financial Capital as Predictors of New Venture Performance.” Journal of Business Venturing 9: 371–395.

De Wit, G. and F. van Winden (1989) “An Empirical Analysis of Self-Employment in the Netherlands.” Small Business Economics 1: 263–272.

Glaeser, E.L., D. Laibson and B. Sacerdote (2000) “The Economic Approach to Social Capital,” NBER Working Paper No. 7728.

Lancaster, T. (1992) “The econometric analysis of transition data.” Cambridge: Cambridge University Press.

Le, A.T. (1999) “Empirical Studies of Self-Employment.” Journal of Economic Surveys 13(4): 381–416.

Maddala, G.S. (1987) “Limited-Dependent and Qualitative Variables in Econometrics.” Cambridge: Cambridge University Press.

Milgrom, P. and J. Roberts, J. (1992) “Economics, Organization and Management.” Englewood Cliffs.

Mincer, J. (1974) “Schooling, Experience, and Earnings.” New York: Columbia University Press.

Montgomery, C. A. (1995) “Resource-Based and Evolutionary Theories of the Firm: Towards a Synthesis.” Kluwer Academic Publishers, Boston/Dordrecht/London.215.

Pennings, L., L. Lee & A. van Witteloostuijn (1998) “Human capital, social capital, and firm dissolution.” Academy of Management Journal, pp. 425–440.

Spence, Michael A. (1974) “Job Market Signaling: Informational Transfers in Hiring and Related Processes.” Cambridge: Harvard University Press.

Van Praag, C.M. (1996) “Determinants of Successful Entrepreneurship.” dissertation, Thesis Publishers, Amsterdam.

Van Praag, C.M. (1999) “Some Classic Views on Entrepreneurship.” De Economist 147(3), pp. 311–335.

Van Praag, C.M. and J.S Cramer. (2001) “The roots of entrepreneurship and labour demand: individual ability and low risk aversion,” Economica 269, 2001, pp. 45–62.