Abstract
This paper explores the role of entrepreneurial human capital in the post-entry performance of firms in high- and low-tech sectors. Using a dataset from the Japanese manufacturing industry, we examine the determinants of new-firm survival, taking into account exit routes to differentiate ‘failure’ (bankruptcy) and ‘nonfailure’ (voluntary liquidation and merger) outcomes. Our results show that entrepreneurial human capital, measured as educational background, is important in reducing the probability of bankruptcy in high-tech sectors, although it does not help significantly in this regard in low-tech sectors. By contrast, we provide evidence that entrepreneurs with high levels of human capital are more likely to voluntarily close businesses both in high- and low-tech sectors. Furthermore, we find that firms managed by entrepreneurs with high levels of human capital are more likely to exit via merger than others, particularly in high-tech sectors. We provide evidence that entrepreneurs with scientific backgrounds are less likely to voluntarily exit than those with humanistic backgrounds, particularly in low-tech sectors.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
It is widely recognized that entrepreneurial human capital helps to improve the post-entry performance of firms. Some scholars have highlighted entrepreneurial human capital as an essential resource for new firms, and have found that firms managed by entrepreneurs with high levels of human capital outperform other firms (e.g., Bates 1990; Cressy 1996; Kato et al. 2015). Meanwhile, the role of entrepreneurial human capital may vary with industry. As argued by Unger et al. (2011), higher levels of knowledge or skills are more important for better and faster decisions in high-tech sectors than in low-tech sectors, because managers must continually adapt to new developments in these sectors. At the same time, entrepreneurial human capital may be a more important signal of the capabilities of firms to external stakeholders in high-tech sectors than in low-tech sectors, because information asymmetries tend to be especially severe in high-tech sectors. In these respects, industry environments affect the role of human capital in the post-entry performance of firms, which has been ignored in the literature.
This paper explores the role of entrepreneurial human capital in the post-entry performance of firms in different industry environments, specifically high- and low-tech sectors. Using a dataset from the Japanese manufacturing industry, we estimate a discrete-time hazard model for the determinants of new-firm survival as post-entry performance. Although many firms exit the market shortly after establishment in different ways, the existing literature tends not to differentiate between different forms of exit. Therefore, this paper differentiates ‘failure’ outcomes, such as bankruptcy, from ‘nonfailure’ outcomes, such as voluntary liquidation and merger, which enables us to evaluate the post-entry performance of firms more precisely. Moreover, our findings indicate that entrepreneurial human capital plays different roles in high- and low-tech sectors. More specifically, we find that while highly educated entrepreneurs are less likely to suffer bankruptcy in high-tech sectors, they tend to voluntarily exit both in high- and low-tech sectors. Furthermore, we find that highly educated entrepreneurs are more likely to exit via merger, particularly in high-tech sectors.
The remainder of this paper is organized as follows. In Section 2, we discuss the background and related literature, and develop our hypotheses. Section 3 explains the data and methods employed in the analyses. Next, the determinants of new-firm survival are discussed in Section 4. The empirical results and discussion are then presented in Section 5. Finally, the last section provides conclusions.
2 Backgrounds and hypothesis development
The role of entrepreneurial human capital in the survival of new firms has been addressed in a rich stream of literature on the economics of entrepreneurship. In this paper, we shed light not only on whether entrepreneurial human capital affects the survival of new firms, but also on how its effects differ between industries, while considering heterogeneity in exit routes. In order to clarify our contributions, we provide background discussion and review the related literature on survival and exit routes, entrepreneurial human capital, and the interplay between entrepreneurial human capital and industry environments. Then, we present our hypotheses.
2.1 Survival and exit routes
Numerous studies have examined the determinants of new-firm survival (e.g., Audretsch 1991; Mata and Portugal 1994; Wagner 1994; Audretsch and Mahmood 1995; Mata et al. 1995).Footnote 1 While many of these studies highlighted firm size and age as factors affecting survival, they treated exit as a homogeneous event in their studies. In this line of studies, survival and exit have been considered to be a measure of performance.
Unlike the above studies, several studies on the survival of firms have considered different exit routes. For example, Schary (1991), in a study of the cotton textile industry, distinguished among bankruptcy, voluntary liquidation, and merger as exit routes. Harhoff et al. (1998) also examined the determinants of firm duration in West German firms, and in so doing distinguished between bankruptcy and voluntary liquidation. Grilli et al. (2010) examined the effects of firm size and age on the survival and exit of new firms by distinguishing between the closure and acquisition exit routes.Footnote 2 Additionally, some studies have examined the duration of self-employment by distinguishing between failure and transition to alternative employment, although these studies on self-employment may differ from studies on new-firm survival (e.g., Taylor 1999; Van Praag 2003; Cueto and Mato 2006).
While it is often considered that survival and exit indicate good and bad outcomes, respectively, recent studies have argued that there are successful cases among firms’ exits. For example, Bates (2005) highlighted successful closure that represents the owner’s decision to redeploy the knowledge gained in the entrepreneurial venture in another small business. By contrast, some entrepreneurs may voluntarily dissolve their businesses to take advantage of higher wages by working elsewhere as employees. Other entrepreneurs may close their businesses because they are approaching retirement age and lack successors. In these respects, such a ‘voluntary’ exit should be distinguished from a ‘compulsory’ exit.
In addition, Arora and Nandkumar (2011) pointed out that an entrepreneur with a high opportunity cost values the option embodied in survival less than an entrepreneur with a low opportunity cost. They argued that the hazard of cash-out rises faster with opportunity cost in high-quality ventures, whereas failure rises more slowly. This suggests that exit through merger should be differentiated from ‘survival’ in evaluating the post-entry performance of firms more precisely.
2.2 Entrepreneurial human capital
To date, many studies have emphasized the importance of entrepreneurial human capital to the post-entry performance of firms. For example, Bates (1990) argued that entrepreneurial input of human capital affects small business longevity. Cressy (1996) emphasized that human capital is the true determinant of firm survival. Åstebro and Bernhardt (2005) argued that increased human capital increases the ability of founders to create and manage viable enterprises. From the competence-based perspective of the firm, Colombo and Grilli (2005) indicated that individuals with greater human capital are likely to have better entrepreneurial judgment, and firms whose founders have greater human capital outperform others because of their founders’ unique capabilities. These studies have suggested that entrepreneurial human capital is a valuable resource for new firms and critical to firm performance, partly because it can compensate for a lack of business experience and resources.
There are several reasons why entrepreneurial human capital plays an important role in the post-entry performance of firms. First, entrepreneurial human capital relates to knowledge or skills for managerial decisions, which will yield better performance. Entrepreneurs with high levels of human capital may be able to learn more about market conditions after starting their businesses, thereby achieving better judgment in their operations. In addition, if entrepreneurs themselves engage in developing new products or technologies, the probability of new technological achievement may also increase with their levels of human capital. Therefore, it has been widely recognized that entrepreneurs with high levels of human capital are more likely to perform better than those with low levels of human capital. Then, entrepreneurial human capital provides an important signal of capabilities of firms to external stakeholders. As information asymmetries are severe in the start-up period, entrepreneurial human capital, such as educational background, serves as valuable information for external providers of capital by signaling the firm’s capabilities, which may lead to lower financing costs.
As is often argued, individuals, including entrepreneurs, with high levels of human capital are more likely to have higher opportunity costs than those with low levels of human capital because of their better employment alternatives (e.g., Evans and Jovanovic 1989; Gimeno et al. 1997; Taylor 1999; Bates 2005; Cassar 2006). While poorly educated individuals have few attractive alternatives to self-employment, this is not the case for highly educated individuals. Therefore, entrepreneurs with high levels of human capital are less likely to stay at their businesses if they receive lower profits than expected, because they incur high opportunity costs. In contrast, entrepreneurs with low levels of human capital, who incur low opportunity costs, are more likely to stay at their businesses even if their profits are not high enough.
On the other hand, entrepreneurs with high levels of human capital may want to leave their businesses, even if profitable. Arora and Nandkumar (2011) indeed found that entrepreneurs with higher opportunity costs are more likely to adopt strategies that hasten a cash-out. In addition, entrepreneurs with high levels of human capital may be more likely to voluntarily liquidate their own firms than those with lower levels of human capital, because they have more attractive alternatives. In these respects, entrepreneurial human capital affects survival and exit as measures of the post-entry performance of firms.
Additionally, the effects of entrepreneurial human capital on exit through merger have been overlooked until now. Firms managed by entrepreneurs with high levels of human capital are more likely to become merger targets, because they are seen to have growth potential. Therefore, it is worth while to examine the effects of entrepreneurial human capital on new-firm survival by distinguishing between bankruptcy, voluntary liquidation, and merger simultaneously.
2.3 Entrepreneurial human capital and industry environments
While entrepreneurial human capital matters for the post-entry performance of firms, its role may depend on the environments in which firms operate. In the field of evolutionary economics, which dates back to Schumpeter (1934, 1943), it has been argued that the roles of entrepreneurs and new firms in innovation differ between industries. As described by Malerba and Orsenigo (1997), entrepreneurs and new firms are very important to innovation in industries characterized by ‘creative destruction’ with technological ease of entry.Footnote 3 In contrast, they described how, in industries characterized by ‘creative accumulation,’ large established firms play a key role and create entry barriers for new entrepreneurs and small firms. These studies clearly indicate that the role of entrepreneurs differs between industries.
In practice, for example, in industries characterized by rapid technological change and consequent short product cycles, firms need to keep up with changing technologies, and frequently develop new products or services to compete. Entrepreneurs in such industries require better and faster managerial judgment. In particular, as has already been argued, higher levels of knowledge or skills may be more important for managerial decisions in high-tech sectors than in low-tech sectors (e.g., Unger et al. 2011). Malerba et al. (2007) also pointed out that because new firms need to find a market that keeps them alive long enough to develop new technologies, the structure of demand in industries plays an important role in the likelihood of the survival of new firms. This suggests that the ability of entrepreneurs to find such a market is more important to survival in high-tech sectors with frequent and rapid technological changes. However, to our knowledge, there has been little research on how the role of entrepreneurial human capital differs between industries.
At the same time, as mentioned earlier, entrepreneurial human capital may be a more important signal of the capabilities of firms to external stakeholders in high-tech sectors than in low-tech sectors. Given the existence of information asymmetries in imperfect capital markets, for example, external providers of finance may take entrepreneurial human capital, such as educational background, as a signal of firm capabilities, and hence of likely firm performance. In addition, entrepreneurial human capital will act as a signal of capabilities to potential business or research partners (e.g., Okamuro et al. 2011). The asymmetric information problem tends to be especially severe in high-tech sectors, because firms are reluctant to disclose their research and development (R&D) projects (e.g., Guiso 1998; Hall and Lerner 2010). As suggested by Himmelberg and Petersen (1994), high-tech firms tend to be more credit-rationed than low-tech ones, partly because information asymmetries are severe and therefore adverse selection problems may be pronounced in high-tech sectors.Footnote 4 These studies suggest that entrepreneurial human capital is more important for post-entry performance in high-tech sectors than in low-tech sectors because of its signaling effect under information asymmetries.
2.4 Hypothesis development
Taking into account the above arguments, we present three sets of hypotheses in turn. The first set is related to bankruptcy as a result of business failure. Based on the competence-based perspective, firms whose founders have greater human capital will outperform others because of their founders’ unique capabilities (e.g., Colombo and Grilli 2005). In addition, because entrepreneurial human capital acts as an important signal of capabilities of firms to external stakeholders under capital market imperfections, firms managed by entrepreneurs with high levels of human capital may have a greater advantage in accessing capital and labor markets. Thus, we postulate the following hypothesis:
Hypothesis 1a: Firms managed by entrepreneurs with high levels of human capital are less likely to go bankrupt.
More importantly, the role of entrepreneurial human capital should vary across sectors with different characteristics. As already discussed, entrepreneurs in high-tech sectors need to keep up with changing technologies and frequently develop new products or services to compete. In addition, higher levels of knowledge or skills may be more important for managerial decisions in high-tech than in low-tech sectors (e.g., Unger et al. 2011). Moreover, entrepreneurial human capital may be a more important signal of the capabilities of firms to external stakeholders in high-tech sectors than in low-tech sectors, because of more asymmetric information and adverse selection problems in the former sectors. In these respects, entrepreneurial human capital may be extremely important for success in high-tech sectors. Thus, we test the following hypothesis:
Hypothesis 1b: The negative association between the levels of human capital of entrepreneurs and the likelihood of bankruptcy for firms is greater in high-tech sectors than in low-tech sectors.
The second set of hypotheses is related to voluntary liquidation that does not necessarily indicate business failure. When entrepreneurs perceive higher opportunity costs than their expected income by staying in their current businesses, they may exit quickly and voluntarily. As already mentioned, entrepreneurs with high levels of human capital tend to have higher opportunity costs than those with lower levels of human capital because they have better employment alternatives. Therefore, we postulate the following hypothesis:
Hypothesis 2a: Firms managed by entrepreneurs with high levels of human capital are more likely to voluntarily liquidate their businesses.
Meanwhile, entrepreneurs may be subject to different opportunity costs in high- and low-tech sectors. Entrepreneurs with high levels of human capital have more alternative employment opportunities in high-tech sectors than in low-tech sectors, because superior capabilities and richer knowledge and skills are more valuable in the former sectors. In contrast, entrepreneurs with high levels of human capital may incur higher opportunity costs in low-tech sectors than in high-tech sectors, because such entrepreneurs may be able to find alternative employment opportunities in high-tech sectors. Therefore, we test the following hypothesis:
Hypothesis 2b: The positive association between the levels of human capital of entrepreneurs and the likelihood of voluntary liquidation for firms is greater in low-tech sectors than in high-tech sectors.
The third set of hypotheses is related to exit through merger, which is sometimes considered a result of business success. It is plausible that entrepreneurs with high levels of human capital, which can be regarded as a valuable resource, have superior managerial and technological capabilities than those with low levels of human capital. In this respect, firms managed by entrepreneurs with high levels of human capital are more likely to be the target of a merger, because merging firms value the targeted firms as an essential resource for further growth. We therefore test the following hypothesis:
Hypothesis 3a: Firms managed by entrepreneurs with high levels of human capital are more likely to exit through merger.
As is repeatedly discussed, higher levels of knowledge or skills are more important in high-tech sectors, while they are less important in low-tech sectors. In contrast, it might not be easy for firms constantly to generate new ideas by themselves in high-tech sectors with fast-changing technologies. Therefore, acquiring new ideas from external organizations can be an important strategy for established high-tech firms. In this sense, firms managed by entrepreneurs with high levels of human capital may be more likely to be a merger target, especially in high-tech sectors, because of their superior capabilities. We thus formulate a related hypothesis to the one above:
Hypothesis 3b: The positive association between the levels of human capital of entrepreneurs and the likelihood of exit through merger for firms is greater in high-tech sectors than in low-tech sectors.
To test the above three sets of hypotheses, in the following sections we present the data and model used in the empirical analyses.
3 Data and methods
3.1 Data sources
The dataset employed in this paper comes from the TSR Data Bank compiled by Tokyo Shoko Research (TSR), one of the major credit investigation companies in Japan.Footnote 5 The dataset consists of manufacturing firms founded between 1997 and 2004, and includes information on the survival and exit of such firms from their year of entry to 2009. The data provide information not only on whether a firm exits but also on its exit route. Besides information on survival and exit, this source provides data on entrepreneur- and firm-specific characteristics.
As explained, this paper examines the different roles of entrepreneurial human capital in explaining the survival of new firms in high- and low-tech sectors. To do this, we divided the full sample into high- and low-tech subsamples using the classification of the Organisation for Economic Co-operation and Development (OECD) (2011).Footnote 6 Table 6 of the Appendix lists the industries in the high- and low-tech sectors.
Regarding other industry-specific characteristics, data on industry concentration and growth at the two-digit Standard Industrial Classification (SIC) level are obtained from the Japan Industrial Productivity (JIP) Database by the Research Institute for Economy, Trade and Industry (RIETI). Furthermore, data on capital intensity were obtained at the three-digit SIC level from the Report by Industry, Census of Manufactures compiled by the Minister of Economy, Trade and Industry (METI). Finally, annual unemployment data for each prefecture came from the Labor Force Survey by the Ministry of Internal Affairs and Communications (MIC).
Meanwhile, some data problems arose in relation to dataset construction. First, information on entrepreneur-specific characteristics, such as education and age, was unavailable for some firms in the TSR Data Bank. Therefore, we dropped these firms from the sample.Footnote 7 Second, while industrial classifications changed considerably for some industries during the observation period, we endeavored to construct a balanced panel of industry-level data for each data source throughout the observation period, 1997–2009. However, when data were unobtainable for a year because of confidentiality, we substituted the industry values for the same observation period from the previous year.Footnote 8
As a result, we obtained a sample of 7,868 small Japanese manufacturing firms founded during the period 1997–2004, and observed them from their year of entry to 2009.Footnote 9
3.2 Distinct exit routes
As explained above, we classify firm exits into three routes—bankruptcy, voluntary liquidation, and merger—using the classifications in the TSR Data Bank.Footnote 10 Bankruptcy is the situation in which firms cannot repay their debts and thus cease operations, and includes firms that apply for court protection under the Bankruptcy Law, as well as those that apply for it under the Corporate Rehabilitation Law and the Civil Rehabilitation Law enacted in Japan in April 2000. Additionally, firms whose banks stop providing credit to service bills payable are considered bankrupt even in the absence of a court judgment. Here we define bankruptcy to include not only firms legally declared bankrupt, but also those that are inactive economically.
In contrast, voluntary liquidation indicates the situation where firms voluntarily dissolve their businesses without insolvency. There are a number of reasons for voluntary liquidation, although identifying them precisely can be difficult. Some entrepreneurs may dissolve their businesses because they recognize that their firms are performing poorly and insolvency is likely. Finally, merger describes the situation in which a firm disappears because of a merger with another firm.Footnote 11
However, a problem arises when we identify exit route and timing, because the TSR Data Bank does not allow the identification of the month and year of exit in the cases of voluntary liquidation or merger. Additionally, in a few bankruptcy cases—those of firms with a total deficit of less than 10 million yen—the month and year of exit cannot be identified, although for most bankruptcies these details are available. According to TSR, its researchers collect information on firms by telephone, postal questionnaires, and field surveys several times a year. This information is not updated for exited firms. Therefore, using information on the accounting period of the last statement of accounts before exit, we identify the exit year for firms exiting via voluntary liquidation and merger, along with that for bankruptcies that involve a total deficit of less than 10 million yen. For these firms, the year following the reporting of the final statement of account is regarded as the exit year.Footnote 12
With respect to distinct exit routes, 418 (5.3 %) of the 7,868 firms in the sample exited the market through bankruptcies, 587 (7.5 %) voluntarily ceased operations, and 165 (2.1 %) disappeared through mergers. The exited firms in the sample totaled 1,170 (15.0 %).Footnote 13
3.3 Methods
Our interest is to estimate the probability of a new firm exiting and to identify the factors that determine its exit route. However, some firms did not exit during the observation period; that is, their duration to exit is right-censored. For this reason, previous literature has applied the duration model—specifically, the proportional hazards model—to the survival and exit of new establishments or firms over time (e.g., Audretsch and Mahmood 1991, 1995; Mata et al. 1995; Honjo 2000b). As already mentioned, the duration model has an advantage in that it can accommodate right-censored observations.
In this paper, the post-entry performance of firms is classified as either survival or exit by one of three routes: bankruptcy, voluntary liquidation, or merger. Although her analysis did not focus on new firms, Schary (1991) assumed these three exit routes to be inherently ordered as follows: merger, nonfailure, and failure. However, this order cannot be reasonably assumed to hold for all situations. As our dataset records just three exit routes, each of which excludes the others, a competing-risks proportional hazards model (CPH model, hereafter) is used to deal with the presence of competing events that impede the event of interest. Furthermore, we use a discrete-time duration model to examine the factors that affect new-firm duration and how they vary according to exit route.Footnote 14
As discussed above, we consider three exit routes—bankruptcy, voluntary liquidation, and merger, so the number of exit routes, m, is set to three (m=3). Let x i j denote a vector of covariates affecting the probability of firm i exiting via route j(=1,…,m). To model the transition from survival to exit through bankruptcy, voluntary liquidation, or merger, we define the hazard function h i j (t), which represents the conditional probability of a transition to route j in period t for surviving firm i.
Following previous studies, we use a complementary log-log model (cloglog model, hereafter) (e.g., Jenkins 2005). The hazard function for the cloglog model can be expressed as follows:
where h 0j (t) is the baseline function at the tth interval with spell duration, x i j is a vector of covariates (some time varying) that affect the survival and exit of new firms, and β j denotes the parameters to be estimated.Footnote 15
4 Determinants of new-firm survival
4.1 Entrepreneur-specific characteristics
In this paper, we consider entrepreneurial human capital to be the most important factor affecting new-firm survival. Entrepreneurs’ educational background is used as a covariate for entrepreneurial human capital.Footnote 16 Educational background has often been used as a measure of human capital in research on the determinants of new-firm survival. Bates (1990) found that firms with highly educated entrepreneurs are more likely to survive. Highly educated entrepreneurs are likely to have more knowledge and skills useful for managing their firms and developing new products. Higher educational background may also signal the superior capabilities of firms to external stakeholders, which helps mitigate problems due to information asymmetries, including financial constraints. A dummy for entrepreneurs with university education (UNIV) is thus introduced as a covariate representing educational background.
While the dummy for university education (UNIV) captures the length of education received by entrepreneurs, the quality of education is another important dimension of human capital. In this paper, a dummy for education at top-ranked universities (TOPUNIV) is introduced to measure the quality of the education received by entrepreneurs. There are 87 national universities, 76 municipal universities, and 570 private universities across 47 prefectures in Japan. From these, we select 12 top-ranked universities (Hitotsubashi, Hokkaido, Keio, Kobe, Kyoto, Kyushu, Nagoya, Osaka, Tohoku, Tokyo, Tokyo Institute of Technology, and Waseda), based on the difficulty of entry at the undergraduate level as of April 2006.Footnote 17 The coefficient of this covariate indicates the hazards of exit faced by elite entrepreneurs educated at top-ranked universities compared with those educated at other universities.
In addition to the ‘level’ of education of entrepreneurs, the ‘field’ of education, such as scientific or humanistic backgrounds, may be another aspect of entrepreneurial human capital. Therefore, the field of education received by entrepreneurs is included as a control in the model. In this respect, we can say that the effects of level of education are identified after controlling for differences in the field of education between entrepreneurs. In practice, entrepreneurs with scientific backgrounds may have advantages in developing new products and technologies, rather than in managing their businesses. In contrast, entrepreneurs with humanistic backgrounds may have disadvantages in technological decision-making, even if they have superior management skills. To capture the effects of the field of education received by entrepreneurs, we construct a dummy (SCI) indicating if the school (e.g., university, junior college, and high school) has scientific departments or courses only and a dummy (COMPRE) indicating if the school has both scientific (e.g., natural sciences, engineering, agriculture, medical, and pharmaceutical) and humanistic departments or courses (including humanities and social sciences). Both of these dummies are zero if the school has only humanistic departmentsor courses.Footnote 18
Besides educational background, the model includes entrepreneurs’ age. Some entrepreneurs who are approaching retirement age and lack successors may be more likely to voluntarily close their firms even when successful. Dummies for entrepreneurs’ age (E A G E2,…,E A G E5) are used as covariates.Footnote 19 Gender differences among entrepreneurs are controlled in the model. Several studies have examined gender differences in the post-entry performance of firms (e.g., Kalleberg and Leicht 1991; Carter et al. 1997; Harada 2003).Footnote 20 A dummy for male entrepreneurs (MALE) is included in the model.
4.2 Other characteristics
With respect to firm-specific characteristics, the logarithm of paid-in capital and its squared term are used as a measure of firm size. Unlike total assets, paid-in capital excludes liabilities or retained profits. While total assets may better represent asset size, they include liabilities, which can increase the probability of bankruptcy if too large. In the model, paid-in capital is included as a control, because numerous studies have provided evidence that firm survival probability increases with size (e.g., Audretsch 1991; Audretsch and Mahmood 1991, 1995; Honjo 2000a, b).Footnote 21
Additionally, the logarithm of firm age and its squared term are included in the model. As indicated by Evans (1987), firm survival and exit depend heavily on firm age, and firms with a longer history are likely to perform differently than newer firms. Older firms may be less likely to experience business failure because of advantages such as learning and scale, but simultaneously are more likely to become merger targets. Furthermore, the covariate for joint-stock companies with limited liability is included to control for differences in legal forms across new firms. Generally, as discussed by Harhoff et al. (1998), differences exist between corporate and noncorporate firms. While entrepreneurs of corporate firms with limited liability are protected in bankruptcies, those of noncorporate firms are fully and personally liable.
While we examine differences in the role of entrepreneurial human capital between high- and low-tech sectors, other industry-specific characteristics should be controlled in the model. Industry concentration is first included to measure competition intensity, which may affect the probability of failure for new firms. As is often argued in the field of industrial organization, high concentration may represent a lack of competition between firms. The Hirshman–Herfindahl Index (HHI), calculated as the sum of the squared market shares for each firm in the industry, is used to measure concentration.
Industry growth is expected to provide a better environment that can help new firms survive and grow. Bradburd and Caves (1982) found that industry growth tends to elevate price–cost margins, and Audretsch and Mahmood (1995, p. 98) pointed out that such margin elevation allows establishments to survive despite operating at suboptimal scale. In this respect, high industry growth may reduce the failure probability of new firms. Regarding capital intensity, firms in capital-intensive industries tend to incur large fixed costs at start-up, because they must establish and operate large plants. In industries where new firms must establish and operate specialized plants and machinery, they may face cost disadvantages relative to large established firms. In this sense, capital intensity may be positively related to the probability of failure of new firms.Footnote 22
The covariate for unemployment rate by prefecture is included to control for regional economic conditions. The unemployment rate has often been used as a measure of regional economic distress (e.g., Storey 1994; Acs et al. 2007), as it negatively affects the performance of region-specific businesses. Thus, new firms in regions with higher unemployment may be more likely to be forced into bankruptcy. In contrast, because entrepreneurs may be unable to find alternative employment in economically distressed regions, they are less likely to voluntarily dissolve their firms. Dummies for firms founded in different years are included to control for different conditions on firm entry.
The definition of these covariates is shown in Table 1. While industry- and region-specific covariates are time varying, entrepreneur- and firm-specific characteristics (except for firm age) are time invariant. A one-year lag is applied to the former covariates to avoid reverse causality.
5 Results
5.1 Descriptive statistics
Before estimating the model for the determinants of new-firm survival, we show and discuss some descriptive statistics for the covariates. As shown in Table 2, the descriptive statistics for the covariates are shown for both the full sample and the subsamples (high- and low-tech subsamples). Regarding the entrepreneur-specific characteristics, about half of the entrepreneurs had a university education, and 6 % of entrepreneurs graduated from a top-12 ranked university in the full sample. While 13 % of entrepreneurs were educated at schools with scientific departments only, 77 % graduated from schools with both scientific and humanistic departments in the full sample. The remaining entrepreneurs (10 %) graduated from schools with humanistic departments only.
An interesting finding is that there are some differences in these covariates between the high- and low-tech subsamples. The share of entrepreneurs with a university education is higher in high-tech sectors (53 %) than in low-tech sectors (47 %). Similarly, the share of entrepreneurs who attended a top-12 ranked university is higher in high-tech sectors (9 %) than in low-tech sectors (5 %). Regarding the field of education, there are substantial differences in the shares of entrepreneurs with scientific backgrounds between high-tech (19 %) and low-tech sectors (9 %).
The mean entrepreneur age at start-up is about 47 years. With respect to gender, 96 % of entrepreneurs in the sample are male. No large differences in these covariates exist between the high- and low-tech subsamples. The correlation coefficients between entrepreneur-specific characteristics and exit probability (pooled exit, bankruptcy, voluntary liquidation, and merger) are shown for both the full sample and the subsamples in Table 3.
5.2 Full-sample results
To examine how the role of entrepreneurial human capital differs between high- and low-tech sectors, we estimate the cloglog model for the determinants of new-firm survival, according to exit route. Our sample consists of 7,868 manufacturing firms founded in Japan during the period 1997–2004, with data monitored until 2009. As a result, the number of firm–year observations is 66,286. We present the marginal effects (dF/dx), instead of the coefficients (β), to evaluate the direction and magnitude of the effects of the covariates. In the following sections, we discuss the results for the full sample, followed by those for the high- and low-tech subsamples.
The estimation results for the full sample are shown in Table 4. As for entrepreneurial human capital, while university education (UNIV) positively affects the probability of pooled exit in equation (i), the signs of the coefficients differ across equations (ii)–(iv). The effect of this covariate on the probability of bankruptcy is negative and significant in equation (ii). By contrast, its effects on the probabilities of voluntary liquidation or merger are positive and significant in equations (iii) and (iv), respectively. The results indicate that while firms with highly educated entrepreneurs are less likely to experience bankruptcy, they are more likely to exit via voluntary liquidation or merger. It also suggests that higher educational background is taken as an important signal of superior capabilities of firms to external stakeholders, which may mitigate the asymmetric information problem.
The dummy for education at top-ranked universities (TOPUNIV) is not significant throughout the full-sample results, and so no difference is found between top-ranked and other universities. This suggests that the quality of education is less important for the post-entry performance of firms. Regarding the field of education, the dummy for education at schools with scientific departments only (SCI) is negative and significant in the full sample. This indicates that entrepreneurs with scientific backgrounds are less likely to voluntarily close businesses. Entrepreneurs with scientific backgrounds tend to acquire technological skills specific to the current businesses by playing a role as engineers as well as managers; therefore, they may face difficulties in switching to other jobs. In contrast, entrepreneurs with humanistic backgrounds may be able to switch to other jobs more easily, because they have invested not in firm-specific skills but in general skills.
Regarding entrepreneur age, E A G E4 and E A G E5 tend to significantly and positively affect both voluntary liquidation and pooled exit, and the marginal effects increase with age in equations (i) and (iii) in Table 4. The findings generally indicate that elderly entrepreneurs are more likely to voluntarily close their businesses than are younger entrepreneurs, consistent with Harhoff et al. (1998).
5.3 Subsample results
Next, we show and discuss the estimation results for the subsamples. The results for the high- and low-tech subsamples are presented in equations (i)–(viii) in Table 5.
With respect to bankruptcy, university education (UNIV) has a significantly negative impact on bankruptcy for the high-tech subsample in equation (ii), while the coefficient of this covariate on bankruptcy is negative but insignificant for the low-tech subsample in equation (vi). The results indicate that firms managed by entrepreneurs with high levels of human capital are less likely to go bankrupt in high-tech sectors, which supports Hypothesis 1. However, the dummy for education at top-ranked universities (TOPUNIV) is not significant, both for the high- and low-tech subsamples.
The results suggest that knowledge and skills generated by university education, regardless of university quality, is more important in high-tech sectors than in low-tech sectors. This may indicate that entrepreneurs subject to frequent market and technological changes require superior capabilities to make better and faster decisions, particularly in high-tech sectors. It may also indicate that higher educational backgrounds are taken as an important signal of superior capabilities of firms to external stakeholders, which may mitigate the asymmetric information problem, particularly in high-tech sectors.
The field of education (SCI and COMPRE) does not have a significant effect in both subsamples. Regarding entrepreneurs’ age and gender, the results of the subsamples resemble those of the full sample. As for other covariates, the results do not differ significantly between the high- and low-tech subsamples, although industry-specific characteristics have different effects to those found in the full sample.
Regarding voluntary liquidation, as shown in equations (iii) and (vii) in Table 5, the dummy for university education (UNIV) has a positive sign in both the high- and low-tech subsamples, but this dummy is significant only in the low-tech subsample. Meanwhile, the dummy for education at top-ranked universities (TOPUNIV) has a significantly positive sign only in the high-tech subsample. Hypothesis 2 is partly supported, and these results indicate that highly educated entrepreneurs tend to have more alternative employment opportunities, and therefore incur higher opportunity costs both in high- and low-tech sectors.
With respect to the field of education, the dummy for education at schools with scientific departments only (SCI) is negative and significant in the low-tech subsample, indicating that entrepreneurs educated at such schools are less likely to voluntarily close businesses in low-tech sectors. As already discussed, entrepreneurs with scientific backgrounds face difficulties in switching to other jobs, while entrepreneurs with humanistic backgrounds can switch to other jobs more easily. Meanwhile, as suggested by Audretsch et al. (1996), asset specificity, including skills and technologies, tends to be more prominent in low-tech than in high-tech sectors. Given that technological change occurs less frequently in low-tech sectors than in high-tech sectors, entrepreneurs may accumulate specific skills particularly in the former sectors. This may be the reason why entrepreneurs with scientific backgrounds are less likely to switch to other jobs voluntarily, especially in low-tech sectors.
As for exit via merger, equations (iv) and (viii) in Table 5 show that the dummy for university education (UNIV) is positive and significant only in the high-tech subsample. The result indicates that highly educated entrepreneur are more likely to exit via merger in high-tech sectors, which supports Hypothesis 3. This result suggests that firms managed by entrepreneurs with high levels of human capital are more likely to be the target of a merger in high-tech sectors than those entrepreneurs in low-tech sectors. However, the dummies for education at top-ranked universities (TOPUNIV) and the field of education (SCI and COMPRE) are not significant, both in the high- and low-tech subsamples.
In summary, our estimation results generally support the hypotheses discussed in Section 2, although Hypothesis 2.b is only partly supported in the sense that entrepreneurial human capital (UNIV or TOPUNIV) matters for voluntary liquidation both in the high- and low-tech subsamples. The most important finding presented in Tables 4 and 5 is that, while entrepreneurial human capital matters in both the high- and low-tech sectors, the role of human capital differs between these sectors.
5.4 Robustness checks
We have used the cloglog model to estimate the determinants of new-firm survival according to exit route for both the full sample and the subsamples. However, unobserved heterogeneity (frailty) may exist among firms, which affects the duration of new firms, such as firm-specific management abilities, skills, or culture. Such unobserved heterogeneity affects the probability of exit via a particular route, and thus induces significant correlation between exit routes. Neglecting the existence of unobserved heterogeneity (when relevant) biases the estimated-duration dependence of the hazard rate, and may attenuate the proportionate response of the hazard variation in each regressor at any survival time (e.g., Jenkins 2005).
To take into account unobserved heterogeneity, we estimate a random-effects cloglog model for the determinants of new-firm survival according to exit route.Footnote 23 We also apply a likelihood-ratio (LR) test to verify whether the panel-level variance component is unimportant and whether the panel estimator is the same as the pooled estimator (cloglog).Footnote 24 We find that the effects of the covariates resemble those in Tables 4 and 5. In fact, the LR tests indicate that the panel estimator does not differ from the pooled estimator for the three distinct exit routes. The results for the random-effects cloglog model for the full sample and subsamples are shown in Tables 7 and 8, respectively, in the Appendix. Footnote 25
6 Conclusions
This paper has explored the role of entrepreneurial human capital in the post-entry performance of firms in high- and low-tech sectors. Using a dataset from the Japanese manufacturing industry, we examined the determinants of new-firm survival, taking into account exit routes to differentiate ‘failure’ (bankruptcy) and ‘nonfailure’ (voluntary liquidation and merger) outcomes. Our results showed that entrepreneurial human capital, measured as educational background, is important in reducing the probability of bankruptcy in high-tech sectors, although it does not significantly help in this regard in low-tech sectors. By contrast, we provided evidence that entrepreneurs with high levels of human capital are more likely to voluntarily close businesses both in high-and low-tech sectors. Furthermore, it was found that firms managed by entrepreneurs with high levels of human capital are more likely to exit via merger than others, particularly in high-tech sectors. We provided evidence that entrepreneurs with scientific backgrounds are less likely to voluntarily exit than those with humanistic backgrounds, particularly in low-tech sectors.
However, this paper has some limitations. Although we classified firm exits into three distinct routes, heterogeneity may still exist within these exit routes. For example, while a merger is a successful outcome for merged firms, it may be driven by poor performance (e.g., Wennberg et al. 2010; Coad 2013). Additionally, although voluntary liquidation occurs for various reasons, we could not classify it into more detailed routes. With respect to entrepreneurial human capital, while we focused on educational background, we did not examine other entrepreneur-specific characteristics, such as managerial and work experience before start-up. As suggested by previous literature in evolutionary economics, including Fontana and Nesta (2010), post-entry learning may be another important aspect of human capital that affects the survival of firms, because entrepreneurs should learn about the market and build their knowledge and skills after start-up. Furthermore, we could not control firm-specific characteristics, especially those associated with financial conditions. Some studies, including Fotopoulos and Louri (2000) and Huynh et al. (2010), found such characteristics to be more important than others in determining the survival and exit of new firms. As indicated by Buddelmeyer et al. (2010) and Wagner and Cockburn (2010), intangible resources, such as patents and trademarks, may also affect firm duration. Moreover, extending this research to other industries may be interesting because, for example, service industries rather than manufacturing industries have recently attracted the attention of entrepreneurs.
Despite the limitations of this study, our findings indicate that the role of entrepreneurial human capital differs between industries, suggesting that we should simultaneously pay heed to both external factors, such as industry-specific characteristics, and internal factors, such as entrepreneur-specific characteristics. These findings differ from those of the previous literature, which ignored the interplay between entrepreneurial human capital and industry environments. Our study revealed heterogeneity in the determinants of survival and exit via bankruptcy, voluntary liquidation, or merger among new firms, focusing on the role of entrepreneurial human capital. We found clear differences in the role of entrepreneurial human capital between exit routes.
From the perspective of economic policies, our findings suggest that more attention should be paid to the interplay between entrepreneurial human capital and industry environments in defining a target support policy for entrepreneurs, because human capital relates to business success, particularly in high-tech sectors. It has been argued that public policies for new businesses should focus more on entrepreneurs and new firms with growth potential (e.g., Santarelli and Vivarelli 2002, 2007; Shane 2009). This is because such firms contribute disproportionately to job creation and innovation, and thus to economic growth. By contrast, developing public policies that are equally attractive for all potential entrepreneurs may be problematic, because such policies tend to promote the entry of low-productivity and noninnovative ‘revolving door’ firms. In practice, Branstetter et al. (2014) found that regulation reform that reduces barriers to entry creates fringe firms with low productivity. In a sluggish economy with a limited budget, such as that of Japan, the government should seek a more efficient way of achieving economic growth by placing high priority on public support for entrepreneurs with high potential. Further investigation of this topic is warranted in order to help develop future economic policies.
Notes
Regarding empirical studies using Japanese data, Doi (1999) examined the determinants of firm exit at the industry level. Honjo (2000a, b) examined the determinants of business failure of new firms using a proportional hazards model. Harada (2007) also examined the determinants of small-firm exit in Japan by distinguishing exits forced by economic factors from other exits. See, for example, Storey and Greene (2010) for a cross-country survey of the evidence on new-firm survival and exit.
Malerba and Orsenigo (1997) also described how new entrepreneurs enter an industry with new ideas and innovations, launch new enterprises that challenge established firms, and continuously disrupt the current ways of production, organization, and distribution, thus wiping out the quasi rents associated with previous innovations.
Honjo et al. (2014) found that entrepreneurial human capital, such as educational background, is positively related to the amount of funds for R&D at start-up, while R&D-oriented start-ups suffer from a funding gap between required and actual investment in R&D.
As a public data source, the Establishment and Enterprise Census reports data, such as numbers of entries and exits, at the individual establishment level, for individual industries or regions. However, it is difficult to obtain data for individual firms from public data sources, and generally we could not use these sources to identify which establishments (or firms) have become active or extinct. Additionally, reliance on these sources is accompanied by the possibility that the relocation of an establishment to another region is recorded as an exit even if the establishment remains in the market. These sources thus create difficulties in identifying whether a firm actually exited the market.
In this paper, although the OECD (2011) classified manufacturing industries into four groups, namely high-tech, medium-high-tech, medium-low-tech, and low-tech, we defined the first two and last two groups as high- and low-tech sectors, respectively. In addition, we divided the full sample into subsamples based on R&D intensity (industry R&D expenditures divided by sales). However, we do not report the results using this methodology because they were generally consistent with those obtained using the definition based on the OECD classification.
We checked whether the results remained consistent if these firms were included in the sample by introducing dummies for firms whose entrepreneurs’ backgrounds are unknown. This test revealed similar results before and after dropping these firms from the sample.
Additionally, we dropped 19 observations in one industry for which we could not match three-digit SIC classifications for data on capital intensity between the periods before and after the changes in SIC.
While we dropped firms with 100 or more employees from the sample as outliers, the exclusion of these firms from the sample had little impact on the results. Moreover, the results generally held even when we tried alternative cutoff points.
While some studies have paid attention to business exit, in this paper ‘exit’ means the disappearance of a firm.
In this paper, a merging firm is regarded as surviving if it continues to operate as the same entity. On the other hand, a merging firm is regarded as exiting through merger if a new entity is created. A merged firm is also regarded as exiting through merger. With respect to acquisition cases, an acquiring firm and an acquired firm are regarded as surviving firms, because neither firm disappears, although ownership is transferred.
However, this assumption regarding the exit year may contain bias. Therefore, we estimated the exit year for all exit routes, including firms with total deficits equal to or greater than 10 million yen, based on the year of the last reported statement of account, and also estimated the determinants of exit. The estimation results changed little, regardless of the method used to identify the exit year.
The exit rate for our sample is much lower than that in some previous studies (e.g., Dunne et al. 1988; Audretsch 1995; Bartelsman et al. 2005). One reason is that the TSR Data Bank, on which our sample is based, comes from the company register, which does not include sole proprietorships. Therefore, the sample may exclude tiny firms, which would naturally exit the market faster than others.
While some previous studies have used the continuous-time duration model to examine firm duration, others have used the discrete-time duration model (e.g., Fontana and Nesta 2009; Cefis and Marsili 2011; 2012). Because the timings of survival and exit are observable only to the year, we use the discrete-time duration model, following Fontana and Nesta (2009) and Cefis and Marsili (2011, 2012).
In this paper, t corresponds to calendar years, which implies that the baseline function is determined by macroeconomic conditions.
Although some firms may be established by multiple entrepreneurs, because of data unavailability, we assume the president to be the entrepreneur.
According to Kato and Odagiri (2012), the difficulty of entry is the best proxy for measuring the quality of universities in Japan. To identify top-ranked universities, we used the score book published by Benesse (formerly Fukutake Shoten), one of the major firms selling services to university entrance examinees. It is well recognized in Japan that these 12 universities have been top ranked for a long time. While we checked whether the results are sensitive to the identification of top-ranked universities by trying other cutoff points between top-ranked and the other universities, the results are generally consistent with those using the dummy for the 12 top-ranked universities.
We classified the schools where entrepreneurs were educated into these three groups, based on information from their official websites and other sources. For high schools, agricultural, fisheries, and technical high schools were classified as having scientific courses only, while commercial high schools were regarded as having humanistic courses only. For junior colleges, technical and commercial junior colleges were included in the former and latter groups, respectively. Both for high schools and junior colleges, the schools with general courses were classified as having both scientific and humanistic courses. To classify universities into the groups, we used a data source, Zenkoku Daigaku Ichiran (List of Universities in the Nation), published annually by Bunkyo Kyokai, which listed all the educational and research organizations in Japanese universities and colleges.
We use the dummies instead of a covariate for continuous ages, because there is the possibility that the effects of age are not linear.
Fairlie and Robb (2009) suggested that female-owned businesses have lower survival rates because of less start-up capital. They also concluded that female business owners have different preferences in terms of goals for their businesses.
Instead of paid-in capital, we used the number of employees as a measure of firm size. However, the results are generally consistent with those using paid-in capital. Data on paid-in capital and the number of employees are not measured for the year of entry, because the TSR Data Bank provides information at the latest available year.
Additionally, we examined the effects of entry rates by industry. As is well known, entry rate is positively correlated with exit rate (e.g., Dunne et al. 1988; Geroski 1995; Caves 1998; Disney et al. 2003). Furthermore, entry rate is considered to be positively correlated with industry growth, because the latter induces the former. To avoid reverse causality and multicollinearity, we excluded the covariate for entry rates, despite it having positive effects on each exit route.
Additionally, we estimated our model using a multinomial logit model. The results are generally consistent with those obtained using the cloglog models.
For more details, see the Stata Manual.
Furthermore, we estimate the cloglog model by restricting the observation period to a fixed amount of time for each firm (e.g., 5 or 7 years), in order to take into account the possibility that the role of entrepreneurial human capital changes after firm foundation. However, we do not report the results, because they are generally consistent with those of Tables 4 and 5.
References
Acs ZJ, Armington C, Zhang T (2007) The determinants of new-firm survival across regional economies: the role of human capital stock and knowledge spillover. Pap Reg Sci 86:367–391
Arora A, Nandkumar A (2011) Cash-out or flameout! Opportunity cost and entrepreneurial strategy: theory, and evidence from the information security industry. Manage Sci 57:1844–1860
Åstebro T, Bernhardt I (2005) The winner’s curse of human capital. Small Bus Econ 24:63–87
Audretsch DB (1991) New-firm survival and the technological regime. Rev Econ Stat 68:520–526
Audretsch DB (1995) Innovation, growth and survival. Int J Ind Organ 13:441–457
Audretsch DB, Mahmood T (1991) The hazard rate of new establishments: a first report. Econ Lett 36:409–412
Audretsch DB, Mahmood T (1995) New firm survival: new results using a hazard function. Rev Econ Stat 64:97–103
Audretsch DB, Menkveld AJ, Thurik R (1996) The decision between internal and external R & D. J Inst Theor Econ 152:519–530
Bartelsman E, Scarpetta S, Schivardi F (2005) Comparative analysis of firm demographics and survival: evidence from micro-level sources in OECD countries. Ind Corp Change 14:365–391
Bates T (1990) Entrepreneur human capital inputs and small business longevity. Rev Econ Stat 72:551–559
Bates T (2005) Analysis of young, small firms that have closed: delineating successful from unsuccessful closures. J Bus Ventur 20:343–358
Bradburd R, Caves RE (1982) A closer look at the effect of market growth on industries profits. Rev Econ Stat 64:635–645
Branstetter L, Lima F, Taylor LJ, Venâncio A (2014) Do entry regulations deter entrepreneurship and job creation? Evidence from recent reforms in Portugal. Econ J 124:805–832
Buddelmeyer H, Jensen PH, Webster E (2010) Innovation and the determinants of company survival. Oxford Econ Pap 62:261–285
Buehler S, Kaiser C, Jaeger F (2006) Merge or fail? The determinants of mergers and bankruptcies in Switzerland, 1995–2000. Econ Lett 90:88–95
Carter NM, Williams M, Reynolds PD (1997) Discontinuance among new firms in retail: the influence of initial resources, strategy, and gender. J Bus Ventur 12:125–145
Cassar G (2006) Entrepreneur opportunity costs and intended venture growth. J Bus Ventur 21:610–632
Caves RE (1998) Industrial organization and new findings on the turnover and mobility of firms. J Econ Lit 36:1947–1982
Cefis E, Marsili O (2011) Born to flip. Exit decisions of entrepreneurial firms in high-tech and low-tech industries. J Evol Econ 14:1167–1192
Cefis E, Marsili O (2012) Going, going, gone. Exit forms and the innovative capabilities of firms. Res Policy 41:795–807
Coad A (2013) Death is not a success: reflections on business exits. Int Small Bus J. forthcoming
Colombo MG, Grilli L (2005) Founders’ human capital and the growth of new technology-based firms: a competence-based view. Res Policy 34:795–816
Cressy R (1996) Are business startups debt-rationed? Econ J 106:1253–1270
Cueto B, Mato J (2006) An analysis of self-employment subsidies with duration models. Appl Econ 38:23–32
Disney R, Haskel J, Heden Y (2003) Entry, exit and establishment survival in UK manufacturing. J Ind Econ 51:91–112
Doi N (1999) The determinants of firm exit in Japanese manufacturing industries. Small Bus Econ 13:331–337
Dunne T, Roberts MJ, Samuelson L (1988) Patterns of firm entry and exit in U.S. manufacturing industries. Rand J Econ 19:495–515
Esteve-Pérez S, Sanchis-Llopis A, Sanchis-Llopis J (2010) A competing risks analysis of firms’ exit. Empir Econ 38:281–304
Evans DS (1987) The relationship between firm growth, size, and age: estimates for 100 manufacturing industries. J Ind Econ 35:567–581
Evans DS, Jovanovic B (1989) An estimated model of entrepreneurial choice under liquidity constraints. J Polit Econ 97:808–827
Fairlie RW, Robb AM (2009) Gender differences in business performance: evidence from the characteristics of business owners survey. Small Bus Econ 33:375–395
Fontana R, Nesta L (2009) Product innovation and survival in a high-tech industry. Rev Ind Organ 34:287–306
Fontana R, Nesta L (2010) Pre-entry experience, post-entry learning and firm survival: evidence from the local area networking switch industry. Struct. Change. Econ Dynam 21:41–49
Fotopoulos G, Louri H (2000) Determinants of hazard confronting new entry: does financial structure matter? Rev Ind Organ 17:285–300
Geroski PA (1995) What do we know about entry? Int J Ind Organ 13:421–440
Gimeno J, Folta TB, Cooper AC, Woo CY (1997) Survival of the fittest? Entrepreneurial human capital and the persistence of underperforming firms. Adm Sci Q 42:750–783
Grilli L, Piva E, Lamastra CR (2010) Firm dissolution in high-tech sectors: an analysis of closure and M&A. Econ Lett 109:14–16
Guiso L (1998) High-tech firms and credit rationing. J Econ Behav Organ 35:39–59
Hall B, Lerner J (2010) The financing of R&D and innovation. In: Hall B., Rosenberg N. (eds) Handbook of the economics of innovation, vol 1, North-Holland, pp 609–639
Harada N (2003) Who succeeds as an entrepreneur? An analysis of the post-entry performance of new firms in Japan. Jpn World Econ 15:211–222
Harada N (2007) Which firms exit and why? An analysis of small firm exits in Japan. Small Bus Econ 29:401–414
Harhoff D, Stahl K, Woywode M (1998) Legal form, growth and exit of West German firms—empirical results for manufacturing, construction, trade and service industries. J Ind Econ 46:453–488
Himmelberg CP, Petersen BC (1994) R&D and internal finance: a panel study of small firms in high-tech industries. Rev Econ Stat 76:38–51
Honjo Y (2000a) Business failure of new software firms. Appl Econ Lett 7:575–579
Honjo Y (2000b) Business failure of new firms: an empirical analysis using a multiplicative hazards model. Int J Ind Organ 18:557–574
Honjo Y, Kato M, Okamuro H (2014) R&D investment of start-up firms: does founders’ human capital matter? Small Bus Econ 42:207–220
Huynh KP, Petrunia RJ, Voia M (2010) The impact of initial financial state on firm duration across entry cohorts. J Ind Econ 58:661–689
Jenkins SP (2005) Survival analysis. unpublished manuscript
Kalleberg AL, Leicht KT (1991) Gender and organizational performance: determinants of small business survival and success. Acad Manage J 34:136–161
Kato M, Odagiri H (2012) Development of university life-science programs and university–industry joint research in Japan. Res Policy 41:939–952
Kato M, Okamuro H, Honjo Y (2015) Does founders’ human capital matter for innovation? Evidence from Japanese start-ups. J Small Bus Manage 53:114–128
Malerba F, Nelson R, Orsenigo L, Winter S (2007) Demand, innovation, and the dynamics of market structure: the role of experimental users and diverse preferences. J Evol Econ 17:371–399
Malerba F, Orsenigo L (1997) Technological regimes and sectoral patterns of innovative activities. Ind Corp Change 6:83–118
Mata J, Antunes A, Portugal P (2007) Borrowing patterns, bankruptcy and voluntary liquidation. Working paper, Universidade Nova de Lisboa
Mata J, Portugal P (1994) Life duration of new firms. J Ind Econ 42:227–245
Mata J, Portugal P, Guimarães P (1995) The survival of new plants: start-up conditions and post-entry evolution. Int J Ind Organ 13:459–481
Organisation for Economic Co-operation and Development (2011) Technology intensity definition: classification of manufacturing industries into categories based on R&D intensities. Directorate for Science, Technology and Industry, Paris
Okamuro H, Kato M, Honjo Y (2011) Determinants of R&D cooperation in Japanese start-ups. Res Policy 40:728–738
Santarelli E, Vivarelli M (2002) Is subsidizing entry an optimal policy? Ind Corp Change 11:39–52
Santarelli E, Vivarelli M (2007) Entrepreneurship and the process of firms’ entry, survival and growth. Ind Corp Change 16:455–488
Schary MA (1991) The probability of exit. Rand J Econ 22:339–353
Schumpeter JA (1934) The theory of economic development: an inquiry into profits, capital, interest, and the business cycle. Harvard University Press, Cambridge
Schumpeter JA (1943) Capitalism, socialism and democracy. Harper, New York
Shane S (2009) Why encouraging more people to become entrepreneurs is bad public policy. Small Bus Econ 33:141–149
Storey DJ (1994) Understanding the small business sector. Routledge, London
Storey DJ, Greene F (2010) Small Business and Entrepreneurship. Pearson Education, Upper Saddle River
Taylor M (1999) Survival of the fittest? an analysis of self-employment duration in Britain. Econ J 109:C140–C155
Unger JM, Rauch A, Frese M, Rosenbusch N (2011) Human capital and entrepreneurial success: a meta-analytical review. J Bus Ventur 26:341–358
Van Praag CM (2003) Business survival and success of young small business owners. Small Bus Econ 21:1–17
Wagner J (1994) The post-entry performance of new small firms in German manufacturing industries. J Ind Econ 42:141–154
Wagner S, Cockburn I (2010) Patents and the survival of internet-related IPOs. Res Policy 39:214–228
Wennberg K, Wiklund J, DeTienne DR, Cardon MS (2010) Reconceptualizing entrepreneurial exit: divergent exit routes and their drivers. J Bus Ventur 25:361–375
Acknowledgments
We extend our thanks for comments from Alex Coad, Kim Huynh, Francine Lafontaine, Jose Mata, Jose Maria Millan, Masayuki Morikawa, Sadao Nagaoka, Hiroyuki Odagiri, Hiroyuki Okamuro, and the participants in seminars at Hitotsubashi University, Erasmus University Rotterdam, the University of Groningen, and the University of Frankfurt, and in the EARIE Annual Conference (Istanbul), the JEA Autumn Meeting (Hyogo), the CAED Conference (London), the RENT Annual Conference (Maastricht), the Competition Policy Research Center Conference (Tokyo), and the Japan Productivity Center Workshop (Tokyo). We also thank the editors (Uwe Cantner and Roberto Fontana) and two anonymous referees for their useful comments. Financial supports from Kwansei Gakuin University Special Grant for Individual Research (A) for the first author and Grant-in-Aid for Scientific Research (B) (No. 26285060) for the first and second authors are gratefully acknowledged. Needless to say, any remaining errors are our own.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kato, M., Honjo, Y. Entrepreneurial human capital and the survival of new firms in high- and low-tech sectors. J Evol Econ 25, 925–957 (2015). https://doi.org/10.1007/s00191-015-0427-3
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00191-015-0427-3