The role of innovation in the postentry performance of new small firms: evidence from Italy.

Vivarelli, Marco

1. Introduction

The aim of this paper is to relate the start-up decision to the postentry performance of new small firms. Hence, this paper is part of economic research in industrial organization, more specifically of the study of industrial dynamics.

In the literature focusing on industry dynamics, there are at least two different strands relevant to the object of this study. On the one hand, there is the literature devoted to entry and to the start-up decision; this line of the economic thought has typically examined the characteristics external to the firm, such as profit expectations in a given sector, entry barriers, and financial constraints. In this field, the basic research questions concern the "pull factors" that can stimulate the entry flow into a given industry. On the other hand, other studies have tried to link the decision to start a new enterprise to characteristics specific to the individual founder and his/her personal background, personal motivation, and sociological environment. In this field, the basic research issues regard the "push factors" that can stimulate entry into a given industry. A common thread running through both strands is that the decision-making horizon is essentially bounded by the point of entry. In other words, the object of the story is simply the entry into a given market.

More recently, a growing literature focusing on the postentry performance of firms has developed. Within this third strand of literature, research topics include the measurement and determinants of survival rates, early exit, and the firm's growth. Although the literature on entry does not pay attention to issues like early exit and chances of survival, papers on postentry performance generally take entry as an exogenous occurrence, in some way preliminary to the analysis. How and why the new firm entered the market is generally considered to be beyond the scope of such studies.

Nevertheless, one can suppose that the initial population of new entrants (entrepreneurs) is not likely to be homogeneous in terms of personal characteristics, motivation, and previous experiences, and it is plausible to suppose a possible link between these preentry features and the firm's subsequent postentry performance. In other words, this paper will address questions such as: does an innovative entrepreneur have the same chances of postentry profitability as a conservative one? Does a former blue collar worker have the same chances as a former manager? Does a defensive motivation for the start-up of a new firm - such as the avoidance of unemployment - tend to be related to an inferior postentry performance? Thus, the purpose of this paper is to provide some insight into the relationship between the decision to start a new firm and the subsequent postentry performance. In particular, we relate those factors that are decisive in leading an economic agent to leave a mother company to start a new firm to the postentry performance of the new-born enterprise. This enables us to distinguish between the types of motivation leading to the better or poorer postentry performance. Although policy conclusions will not be explicitly drawn in this particular study, the results presented in the following sections may be of some help in designing national and local industrial policies.

In the second section of this paper a survey of previous studies on entry and postentry performance will be presented. The dataset, which consists of a sample of 147 Italian spin-offs, is described in the third section. In the fourth section the hypothesis that different factors inducing new-firm start-ups have a disparate effect on the postentry performance is tested. Finally, the main findings of this study are briefly summarized.

2. Start-Up of Small Firms and Their Subsequent Early Performance: A Review of the Literature

Entry has an important role in the model of perfect competition since an excess level of profitability induces entry into the industry; that is, if excess profits occur - caused by the opening of a market niche, a cost-saving innovation, or product differentiation - additional agents are attracted into the market. In this view, a queue of well-informed potential entrepreneurs is supposed to be waiting outside the market, and the expected level of profit is the "pull factor" determining entry (Khemani and Shapiro 1986; Acs and Audretsch 1989a; Geroski 1991; Geroski and Schwalbach 1991).

In most of the conventional entry models, the maximization of expected profits is discounted by taking into account barriers to entry. In these models, entry is still driven by expected profits, but it is also hindered by factors such as initial sunk costs, minimum efficient size, industry-specific features such as innovation, and advertising expenditure (Mansfield 1962; Orr 1974; Baldwin and Gorecki 1987). More recently, "strategic barriers to entry" have attracted the attention of many scholars: in these models entry barriers play a preemptive role in a game between incumbents and potential entrants (Tirole 1989, ch. 8). The main shortcoming of this traditional approach to industrial organization is that by treating the industry market as the essential unit of observation, it is virtually impossible to take into account individual characteristics of the potential founder (push factors). In fact, while the diversification of an existing firm into a new market can be fully explained through the consideration of expected profits and (strategic) entry barriers, the study of the foundation of a new small firm by a single entrepreneur requires a wider perspective. Indeed, if the unit of observation is the level of the market, this obscures the decision-making process at the level of the individual (Winter 1991) and underestimates the factors shaping the entrepreneur's motivation in starting a new business (McClelland 1961).

As a complement to the simplified view of entry discussed so far, a different tradition can be singled out in the history of economic thought: Schumpeter (1911), Knight (1921) and Oxenfeldt (1943) drew attention to the subjective characteristics of the actual founder of a new firm. Their well-known definitions of entrepreneurship opened the way to a more general framework where pull factors are studied together with some push factors concerning the environment and the particular situation of the potential founder (for a discussion of push factors contrasted with pull factors, see Kilby 1971).

This literature is mainly empirical in nature and gives interesting insights into the actual process of entry. For instance, according to the results of the surveys carried out by Storey (1982) and Johnson (1986), the founder of a new firm is heavily influenced by his/her own background, with particular reference to his/her previous job experience (it has to be taken into account that the vast majority of new firms originate through spin-off from a mother company: 60% in Storey [1982] and 68% in Vivarelli [1991]). In other words, his/her entrepreneurial project is dependent on technical and managerial competence acquired in the previous job: technical savoir faire, organizational skills, knowledge of the market, and so on (Bates 1990). Thus, spin-off results can be affected by the previous hierarchical job position of an ex-employee (for instance, successful entrepreneurs are more likely to have been managers involved in supervisory functions). As regards the personal characteristics of the founder, family background is also singled out as a key factor by econometric estimates explaining new firm formation as an act of self-employment (Evans and Leighton 1989; De Wit and Van Winden 1989). However, other empirical studies tend to play down the relative importance of this factor in shaping the start-up decision (Vivarelli 1991).

An interesting way to model entry decision to encompass both pull and push factors is the so-called "income choice" approach (Creedy and Johnson 1983; Storey and Jones 1987; Blanchflower and Oswald 1990; Evans and Leighton 1990; Blanchflower and Meyer 1994). In this theory, a potential founder compares his present income and prospects as an employee with the expected income from the independent activity; if this difference is more than a given threshold, whose level depends on the individual's risk aversion and on particular psychological aptitudes such as a strong desire to be independent, the new firm will be founded. While this approach encompasses the traditional view whereby entry is pulled by expected profits, it also explains some situations where the founding of the new firm is induced either by uncertainty about future career perspectives or by a strong psychological motivation to be independent (for an application of this model to the Italian case, see Foti and Vivarelli 1994; for a general view of entry and survival processes in Italy, see Contini and Revelli 1992; Santarelli and Sterlacchini 1994). Obviously, this model is of particular interest in modeling the spin-off of previously dependent workers who opt for self-employment (for theoretical models based on the income choice, see Lucas 1978; Kihlstrom and Laffont 1979; Blau 1987; Holmes and Schmitz 1990).

In addition, the income choice approach can be adapted to some situations where the decision to start a new firm is induced either by unemployment or by fear of becoming unemployed; that is, by a very low expected income from the present employment. Indeed, this kind of start-up has been called "escape from unemployment" (Oxenfeldt 1943, cb. 10). Storey and Jones (1987) discovered a percentage of previously unemployed founders varying between 25 and 50%, whereas Highfield and Smiley (1987) found evidence for a counter-cyclical feature of new firm formation (for critical reviews of the escape from unemployment hypothesis, see Hamilton 1989; Storey 1991).

The literature on the entry process has roots in the history of economic thought, but a new strand of literature on the postentry performance of firms has emerged in the last few years (see, for instance, Reid 1991; Boeri and Cramer 1992; Dunne and Hughes 1994; Audretsch and Mahmood 1995; Baldwin and Rafiquzzaman 1995; Mata, Portugal, and Guimaraes 1995). These studies have drawn attention to issues of industry dynamics such as postentry patterns of survival, growth, and early exit. In these analyses, cohorts of new firms are tracked over time and their postentry performance is theoretically modeled and empirically observed. Here it will be sufficient to focus attention on those studies that have attempted to relate postentry performance to some newly founded firms' initial characteristics. For instance, Dixit (1989) and Hopenhayan (1992) both argue that postentry performance may be affected by the level of sunk costs in the industry: higher sunk costs should reduce the likelihood of early exit since precommitment can be seen as a signal of superior entrepreneurial capabilities. Other studies are more concerned with industrial rather than firm characteristics; for example, Audretsch (1995) focuses on the degree of scale economies in a given industry, arguing that new small firms not able to grow and to approach the industry's minimum efficient scale will presumably be characterized either by bad postentry performance or by early exit (see also, Acs and Audretsch 1989b).

On the whole, although these studies are very useful in representing the role of entry in determining evolutionary patterns at industry level (industry differentiation and innovation, Gibrat's law, patterns of early exit and population renewal, etc.), they say little about the postentry performance of the single new-born firms. In this paper, however, the focus of the analysis will be the microperformances of newly founded firms and how their performances can be linked to their initial characteristics.

Jovanovic's studies (1982, 1994) are based on the hypothesis that the income decision resulting in a new entrepreneurial firm and subsequent postentry performance are formed under a "veil of ignorance." In this paper the hypothesis is that the determinants of the starting of a new firm do not have a neutral influence on the postentry performance. In Jovanovic's framework it is the postentry learning that matters. In our analysis, postentry performance is partially predetermined on the basis of the founder's preentry features. For example (and taking into account both conventional entry models and the income choice model), if the underlying motivation to start a new firm is explicitly linked to innovative projects, then a better postentry performance may be expected than if a new firm is started on the basis of a purely defensive motivation, such as the fear of becoming unemployed (for the role of innovation in making a new small firm successful, see Audretsch 1991; Acs and Audretsch 1990). Similarly, some deeply rooted psychological motivation, such as the strong desire to be independent, can hinder a rational and objective consideration of actual profit expectations for the new firm and jeopardize future chances of business success. Moreover, a spin-off originated by a managerial position may have better chances than a spin-off originated by a low-skilled worker.

More generally (and taking into account that our sample is made up of new-born firms originated through spin-off), it is important to see that an individual economic agent's choice to start a new firm rather than remaining employed in an incumbent organization can be influenced by many different factors related to the founder's previous job experience and learning within the mother company (Porter 1980; Storper and Salais 1997, 164 and ff.).

3. Data and Measurement Issues

A database of 147 new manufacturing firms originating through spin-offs in the Italian province of Milano was constructed on the basis of returned questionnaires collected in 1996. In order to avoid possible bias, inconsistencies, and exaggeration which typically affect mail surveys, questionnaires were filled in during direct phone interviews conducted by interviewers previously trained by the authors.

The database includes only entirely new firms set up in 1989 and 1990 and surviving up to the time of the interview (1996). Thus, it consists solely of surviving firms; this creates a particular bias in that early failures are excluded from the analysis. However, a variance in business performance within surviving firms was sufficiently large to ensure that the survivor bias would not be particularly worrying for the purposes of this study. We will not be able to address the issues concerning early exit versus survival, but we will focus on performance differentials among surviving spin-offs.

The firms included in the database were randomly selected according to the local industrial structure with the assistance of lists from the local Chamber of Commerce. From an initial sample of 439 new-born firms, any start-up that was either a new branch of an incumbent enterprise or simply a nominal legal transformation of an existing firm was eliminated from the database. Next, the interviewers selected only the new founders who were previously employed in existing firms (spin-off). The resulting subsample consisted of 226 cases, out of which 147 entrepreneurs allowed interviews.

As previously mentioned, the entire sample was found in the Milano province, which is the most industrialized and advanced local economic system in Italy. The province industrial structure is deeply diversified and does not exhibit idiosyncratic features such as a specific sectoral specialization (in contrast with most industrial districts located in northwest and central Italy), the presence of a dominant large firm (as in Torino with the presence of FIAT), or the prevailing weight of the public sector (as in the south of Italy). In addition, national and local industrial policy, although not completely absent, does not appear to have significantly affected the structure and the performance of the local economy. Hence, the degree of generalizability of this particular empirical study appears to be sufficiently high, though limited to industrialized, advanced local economic systems.

The questionnaire has a qualitative/quantitative nature: together with personal qualitative characteristics and background, the founder interviewed was asked to give a score (from 0 to 10) to motivations and factors that had been relevant to shaping his/her start-up decision. Most of the interviewed entrepreneurs had set up very small firms after leaving larger but still small mother firms; 90.4% of those interviewed had fewer than five employees at start-up, while 52.9% of the mother companies had fewer than 10 employees (although 34.8% had more than 20 employees). All those interviewed had at least one employee, so pure self-employment has been excluded from the analysis and the results focused on the spin-off of a new (although small) firm from an incumbent firm. In addition, mean start-up investment costs proved very low (about U.S. $60,000) and only partially sunk (on average, 2/3 of the initial investment was considered irreversible). On the basis of these results, one can say that barriers to entry do not play an important role in deterring spin-offs in Italian manufacturing.

The database was used to construct, through a factor analysis, a joint measure of postentry performance. This measure takes into account both annual postentry employment growth and the relative profitability of the new-born firm (in comparison with the sectoral average). Although the second measure is the standard way of appraising a firm's performance, annual employment growth has been chosen as a second objective measure of new-born firms' performance, since these firms generally start very small and they decide to hire new employees only in the case of very positive expectations regarding current and future incomes. Although virtually all firms within this sample entered the market at suboptimal size, and so all of them "needed to grow," one can argue that in different sectors the specific levels of the minimum efficient scale call for different paces of employment growth for new entries that want to survive. These considerations support the decision to check for sectoral fixed effects when a newly founded firm's performance is analyzed. Thus, the regressions explaining firm performance with different determinants will be characterized by the inclusion of sectoral dummies (see below).

The empirical tests carried out in the following section will follow a three step procedure.

(i) Answers to the questionnaire will be processed and resulting rankings will be commented on, trying to underline the main determinants of new firm formation with regard to both personal background and motivation and to learning and competence accumulated in the mother firm.

(ii) All factors will be grouped by factor and cluster analyses, with the purpose of singling out some dominant profiles that may help to construct a taxonomy of different categories of spin-offs.

(iii) Factors and clusters derived by steps i and ii will be used, together with other isolated characteristics and checking by means of sectoral dummies, as regressors of an econometric analysis for the explanation of postentry performance across the sample of 147 spin-offs.

4. Empirical Results

Each entrepreneur ranked each factor from a scale of 0 (indicating a minimal influence on the start-up decision) to 10 (indicating a maximal influence on the start-up decision). The median ranking of those factors influencing the decision to start a new firm is reported in Table 1 together with means, standard deviations, and variation coefficients for each answer (the ranking according to the average scores would prove very similar to that shown in Table 1).

Although it is not explicitly the purpose of this study, it can be seen that the ranking of factors shaping the entrepreneurial decision is consistent with the findings in the literature discussed in section 2. As Storey (1982), Johnson (1986), and Vivarelli (1991) found, the desire to be independent turns out to be the most decisive factor underlying the decision to start a new firm. Indeed, at the top of the ranking we find the desire to manage one's working time and the refusal to carry out tasks for others.

Although the perceived potential for income increase is somewhat less important (median = 6), it still plays an important role; interestingly enough, profit expectations are more related to demand factors (the openness of a market niche) than to supply factors (belief in introducing an innovation).

In fact, innovative motivation ranks in the middle of the list (marketing innovation = 5; process innovation = 4; product innovation = 2). This result is consistent with previous research on the start-up of Italian small firms (Vivarelli 1991, p. 220). However, although innovative determinants are not dominant within the population of new founders, they might be relevant to the superior postentry performance of the subsample of better entrepreneurs (see below).

Turning attention to personal motivation connected with previous job experience, the attempt [TABULAR DATA FOR TABLE 1 OMITTED] to create a situation where an economic agent's human capital is more fully deployed seems to play an important role in the entrepreneurial decision. Disagreement about the general managing criteria of the mother firm is the fifth factor in the ranking (median = 5), and undervaluation of skills is the seventh (median = 4). Other more specific forms of disagreement with the management of the mother firm follow: the undervaluation of technological opportunities (median = 3) and the waste of market opportunities (median = 1).

Fear of unemployment and scepticism about the future prospects of the mother firm turn out to be minor motivations (median = 1), although concern about future career developments scores higher (median = 3). On the whole, defensive motivations are not very common, although they may be relevant to a minority of new potential entrepreneurs. Finally, these spinoffs were not supported by the mother firms (median score = 0).

As can be deduced from the reported standard deviation values, the population of new founders shows a quite large variability in the motivation leading to start-up. In other words, although motivation can be ranked, new founders seem to be quite heterogeneous. It could therefore be useful to construct a typology of entrepreneurs.

This exercise has been conducted with the help of a factor analysis, with results being reported in Table 2. The results of the statistical inference are both satisfactory (the explained variance is more than 60%) and interesting.

The first factor (F1) is undoubtedly connected with innovative motivation, expressed in all its different forms (four questions in the questionnaire). Moreover, innovative factors are related to actual market opportunities (two questions). F2 is made up of a group of motivations related to the previous frustration within the mother company and to the desire to better exploit different opportunities that were inadequately valued in the mother firm. F3 is a defensive factor both in terms of fear of losing one's job and of scepticism about future career opportunities. Finally, factor F4 is a residual factor mainly connected with the desire to be independent coupled with profit expectations.

Table 2. Factor Analysis of Motives for Founding the New Firm
(Factor Scores [greater than] 0.40)

 F1 F2 F3 F4

V11H 0.793
V11C 0.782
V11I 0.775
V11O 0.560
V11Q 0.518 0.445
V11E 0.824
V11B 0.781
V11D 0.488 0.691
V11L 0.881
V11N 0.776
V11G 0.747
V11P 0.810
V11F 0.776
V11A 0.598
V11M -0.458

Total variance explained: 60.1%; % of variance: F1 = 27.8;
F2 = 12.9; F3 = 11.8; F4 = 7.8.

Key: V11A, perceived potential for income increase; V11B,
disagreement about the general managing criteria of the mother firm;
V11C, belief in introducing a marketing innovation; V11D,
underestimation of technological opportunities by the mother firm;
V11E, underestimation of skills by the mother firm; V11F, desire to
manage one's working time; V11G, concern about future career
developments; V11H, belief in introducing a product innovation;
V11I, belief in introducing a process innovation; V11L, fear of
becoming unemployed; V11M, support from the mother firm; V11N,
scepticism about prospects of the mother firm; V11O, openness of a
market niche; V11P, refusal to carry out tasks for others; V11Q,
waste of market opportunities by the mother firm.

Table 3 is similar to Table 1, but it concerns the learning process that occurred in the mother firm and that may have played a role in shaping the start-up decision and in determining postentry performance.

As can be clearly seen, acquired technical capabilities are ranked as the most important preconditions for start-up. On the other hand, and consistent with the results shown in Table 1, innovative learning is not so common among the founders interviewed (the four innovative forms of learning are all among the last five items in the ranking list). In the mid-ranking, there are both managerial learned competences (organization, accounting, finance) and entrepreneurial abilities (information, personal contacts, knowledge of the market and the cost-benefit approach).

All these determinants linked to previous job experience are grouped into three factors, according to the results of a second factor analysis shown in Table 4.

The outcome of the factor analysis explains almost 55% of the entire variance and singles out three factors. FF1 groups together all technical capabilities and entrepreneurial skills, FF2 is clearly defined and picks up the four forms of learning connected with innovation, and FF3 consists of managerial skills.

In general, the two factor analyses (Tables 2, 4) are sufficiently clear-cut and the seven factors can be used in an attempt to explain postentry performance. The regression analysis is threefold: the first regression relates postentry performance (as defined in the previous section) to the four factors (F) derived from personal motivation; the second regression relates postentry performance to the three factors (FF) derived from learning in the mother firm; the third regression relates postentry performance to three dummy variables, which have been built on the basis of a cluster analysis run on the seven factors F and FF. The resulting clusters are the [TABULAR DATA FOR TABLE 3 OMITTED] innovative firms (19 firms, where F1 and FF2 are the dominant factors), the experience firms (47 firms, where F2 and FF3 are dominant), and the defensive plus residual firms (81 firms, where F3, F4, and FF1 are dominant).

Table 4. Factor Analysis of Competence Accrued in the Mother Firm
(Factor Scores [greater than] 0.40)

 FF1 FF2 FF3

V25B 0.815
V25A 0.797
V25E 0.762
V25D 0.718
V25C 0.688
V25G 0.681
V25H 0.517
V25F 0.436
V25L 0.816
V25P 0.736
V25Q 0.682
V25I 0.679
V25O 0.794
V25N 0.733
V25M 0.648
V25S 0.426 0.489

Total variance explained: 54.9%; % of variance: FF1 = 34.2;
FF2 = 11.8; FF3 = 8.8.

Key: V25A, knowledge of the final market; V25B, specific technical
skills, V25C, personal contacts; V25D, knowledge of competing firms;
V25E, knowledge of the market of raw materials and intermediate
goods; V25F, knowledge of unexploited opportunities; V25G, technical
knowledge; V25H, production process management; V25I, experience in
developing new products; V25L, experience in new equipment
engineering; V25M, general management competence; V25N, marketing
skills; V25O, administrative skills; V25P, involvement in advanced
technological projects; V25Q, opportunity to check product/process
innovation; V25S, cost-benefit analysis of new entrepreneurial
projects.

[TABULAR DATA FOR TABLE 5 OMITTED]

All regressions have been controlled for both sectoral effects (dummy variables included in all the regressions reported in Table 5, except e) and objective characteristics of the newborn firm: initial size, size of the mother firm, number of business partners, years of previous founder's job experience, founder's level of education, founder's previous hierarchical job position. Among all these variables, only the last was positively and significantly connected with postentry performance and has been retained in the final regression analysis. The status of the previous job position of the entrepreneur (position) has been measured on a scale from 1 (low-skilled worker) to 6 (top manager). All regressions have been submitted to heteroscedasticity tests, and White's correction has been used when necessary.

Regressions a and b refer to the four factors derived from the revealed motivation leading to the start-up. A clear-cut result is that innovative motivation (F1) is positively and significantly (99%) connected with superior postentry performance. Although the other three factors were not significant, it is of some interest to notice that defensive motivation (F3) is characterized by the lowest coefficients and degrees of significance. The inclusion of the previous job position does not affect the results and confirms the positive role of previous managerial experience in supporting better postentry performance (although the coefficient is just barely significant).

Regressions c and d are instead based on factors that have been derived from the revealed importance of different job experience within the mother firm. Again, learning connected with innovative activities (FF2) is strongly linked with better postentry performance and shows a high degree of statistical significance. On the other hand, the other forms of learning, although characterized by positive coefficients, do not exhibit a sufficient degree of significance. The inclusion of the additional regressor "position" does not change the results and consistently shows the positive influence of previous job experience in a higher hierarchical position.

Interestingly enough, both regressions a and c throw some light on what is commonly known as managerial efficiency. Indeed, factor analyses presented in Tables 2 and 4 are built on the different features that may cause different degrees of managerial efficiency within the spin-off firm. Instead of treating managerial effects as dummy variables, regressions a and c directly relate factors to the firm's performance. However, one can argue that these regressions might be affected by the omission of an additional direct measure of managerial effort. In order to check for this possible bias, regressions a and c have been rerun including an additional regressor measuring the weight of managerial and coordination expenditure as a percentage of the firm's total expenditure. Both regressions are robust regarding this additional control: results reported in Table 5 remain virtually unchanged, both in terms of coefficient values and in terms of their statistical significance, while the additional regressor turns out to be of the expected sign (positive), but not significant (t = 1.39 in the modified regression a and t = 1.11 in the modified regression c).

Regression e is based on the cluster analysis described above and shows, once again, the positive and significant effect of an innovative background (innovative motivation plus innovative experience). This result is not surprising and is consistent with the positive effect of F1 and FF2, the "constituent factors" of the innovative cluster. In addition, the defensive cluster shows a negative, though barely significant, influence on postentry performance. This result gives some support to the hypothesis that when avoidance of unemployment or career concern are motivating forces behind the entrepreneurial decision, the subsequent postentry performance tends to be worse.

On the whole, the most unambiguous result in Table 5 concerns the positive roles of innovative motivation and experience in promoting a better postentry performance of newly founded firms.

5. Conclusions

The start-up of new small firms is not a homogeneous phenomenon: different motivation and different previous job experience lead to different forms of spin-offs, which can be characterized by very different postentry business performances.

One of the main conclusions of this study is that the postentry performance of new firms originated by spin-off does not seem neutral regarding the motivation and job experience that led the economic agent to the decision to start a new firm. In other words, the questions posed in the introduction of this study turn out to be relevant; that is, innovative motivation and job experience in innovative activities tend to be linked to a superior postentry performance.

In contrast, there is some evidence that defensive motivation such as the avoidance of possible or actual unemployment seems to be a predictor of a poorer postentry performance. Finally, a high position in the previous job tends to be positively correlated with postentry performance.

Since the sample used in this study does not appear to be significantly biased by idiosyncratic local and institutional factors (see section 3), these conclusions could be applied to other advanced and industrialized local economic systems. However, this work opens the way to further research on a subject that seems to be of some importance in explaining industry dynamics.

We thank Elisabetta Arcaini for careful research assistance; the financial contribution from the Chamber of Commerce of Milano is gratefully acknowledged.

References

Acs, Zoltan J., and David B. Audretsch. 1989a. Small-firm entry in US manufacturing. Economica 56:255-65.

Acs, Zoltan J., and David B. Audretsch. 1989b. Births and firm size. Southern Economic Journal 56:467-75.

Acs, Zoltan J., and David B. Audretsch. 1990. Innovation and small firms. Cambridge, MA: MIT Press.

Audretsch, David B. 1991. New-firm survival and the technological regime. Review of Economics and Statistics 73:441-50.

Audretsch, David B. 1995. Innovation and industry evolution. Cambridge, MA: MIT Press.

Audretsch, David B., and Talat Mahmood. 1995. New firm survival: New results using a hazard function. Review of Economics and Statistics 77:97-103.

Baldwin, John R., and Paul K. Gorecki. 1987. Plant creation versus plant acquisition: The entry process in Canadian manufacturing. International Journal of Industrial Organization 5:27-41.

Baldwin, John R., and Mohammed Rafiquzzaman. 1995. Selection versus evolutionary adaptation learning and post-entry performance. International Journal of Industrial Organization 13:501-22.

Bates, Timothy. 1990. Entrepreneur human capital inputs and small business longevity. Review of Economics and Statistics 72:551-59.

Blanchflower, D., and B. Meyer. 1994. A longitudinal analysis of young entrepreneurs in Australia and the United States. Small Business Economics 6:1-20.

Blanchflower, D., and A. Oswald. 1990. What makes an entrepreneur? NBER Working Paper No. 3252.

Blau, David M. 1987. A time series analysis of self-employment in the United States. Journal of Political Economy 95:445-67.

Boeri, Tito, and Ulrich Cramer. 1992. Employment growth, incumbents and entrants: Evidence from Germany. International Journal of Industrial Organization 10:545-66.

Contini, Bruno, and Riccardo Revelli. 1992. Imprese, occupazione e retribuzioni al microscopio. Bologna: il Mulino.

Creedy, J., and Peter S. Johnson. 1983. Firm formation in manufacturing industry. Applied Economics 15:177-85.

De Wit, Gerrit, and Frans A.A.M. Van Winden. 1989. An empirical analysis of self-employment in the Netherlands. Small Business Economics 1:263-72.

Dixit, Avinash. 1989. Entry and exit decisions under uncertainty. Journal of Political Economy 97:620-38.

Dunne, Paul, and Alan Hughes. 1994. Age, size, growth and survival: UK companies in the 1980s. Journal of Industrial Economics 42:115-40.

Evans, David S., and Linda S. Leighton. 1989. Some empirical aspects of entrepreneurship. American Economic Review 79:519-35.

Evans, David S., and Linda S. Leighton. 1990. Small business formation by unemployed and employed workers. Small Business Economics 2:319-30.

Foti, Alessandro, and Marco Vivarelli. 1994. An econometric test of the self-employment model: The case of Italy. Small Business Economics 6:81-93.

Geroski, Paul A. 1991. Domestic and foreign entry in the United Kingdom: 1983-1984. In Entry and market contestability: An international comparison, edited by Paul Geroski and Joachim Schwalbach. Oxford: Basil Blackwell.

Geroski, Paul A., and Joachim Schwalbach (eds.). 1991. Entry and market contestability: An international comparison. Oxford: Basil Blackwell.

Hamilton, R. T. 1989. Unemployment and business formation rate: Reconciling time-series and cross-section evidence. Environment and Planning 21:249-55.

Highfield, Richard, and Robert Smiley. 1987. New business starts and economic activity: An empirical investigation. International Journal of Industrial Organization 5:51-66.

Holmes, Thomas J., and James A. Schmitz, Jr. 1990. A theory of entrepreneurship and its application to the study of business transfer. Journal of Political Economy 98:265-94.

Hopenhayan, Hugo. 1992. Entry, exit and firm dynamics in long run equilibrium. Econometrica 60:1127-50.

Johnson, Peter S. 1986. New firms: An economic perspective. London: Allen & Unwin.

Jovanovic, Boyan. 1982. Selection and evolution of industry. Econometrica 50:649-70.

Jovanovic, Boyan. 1994. Firm formation with heterogeneous management and labor skills. Small Business Economics 6:185-91.

Khemani R. S., and D. M. Shapiro. 1986. The determinants of new plant entry in Canada. Applied Economics 18:1243-57.

Kihlstrom, Richard E., and Jean-Jacques Laffont. 1979. A general equilibrium entrepreneurial theory of firm formation based on risk aversion. Journal of Political Economy 87:719-48.

Kilby, Peter (ed.). 1971. Entrepreneurship and economic development. London: Freeprint.

Knight, Frank H. 1921. Uncertainty and profit. New York: Houghton Mifflin.

Lucas, Robert E., Jr. 1978. On the size distribution of business firms. Bell Journal of Economics 9:508-23.

Mansfield, Edwin. 1962. Entry, Gibrat's law, innovation and the growth of firms. American Economic Review 52:1023-51.

Mata, Jose, Pedro Portugal, and Paulo Guimaraes. 1995. The survival of new plants: Start-up conditions and post-entry evolution. International Journal of Industrial Organization 13:459-82.

McClelland, D. 1961. The achieving society. Princenton: D. Van Nostran.

Orr, Dale. 1974. The determinants of entry: A study of the Canadian manufacturing industries. Review of Economics anti Statistics 56:58-66.

Oxenfeldt, A. R. 1943. New firms and free enterprise: Pre-war and post-war aspects. Washington: American Council on Public Affairs.

Porter, Michael. E. 1980. Competitive strategy: Techniques for analysing industries and competitors. New York: Free Press.

Reid, Gavin C. 1991. Staying in business. International Journal of Industrial Organization 9:545-56.

Santarelli, Enrico, and Alessandro Sterlacchini. 1994. New firm formation in Italian industry: 1985-1989. Small Business Economics 6:95-106.

Schumpeter, Joseph A. 1911. Theorie der wirtschaftlichen Entwicklung. Eine Untersuchung uber Unternehmergewinn, Kapital, Kredit, Zins und den Konjunkturzyklus. Berlin: Duncker und Hublot (The Theory of Economic Development. Cambridge, MA: Harvard University Press, 1912).

Storey, David J. 1982. Entrepreneurship and the new firm. London: Croom Helm.

Storey, David J. 1991. The birth of new firms: Does unemployment matter? Small Business Economics 3:167-78.

Storey, David J., and A. M. Jones. 1987. New firm formation - A labor market approach to industrial entry. Scottish Journal of Political Economy 34:37-51.

Storper, Michael, and Robert Salais. 1997. Worlds of production. Cambridge, MA: Harvard University Press.

Tirole, Jean. 1989. The theory of industrial organization. Cambridge, MA: MIT Press.

Vivarelli, Marco. 1991. The birth of new enterprises. Small Business Economics 3:215-23.

Winter, Sidney G. 1991. On coarse, competence, and the corporation. In The nature of the firm: Origins, evolution and development, edited by Oliver E. Williamson and Sidney G. Winter. Oxford: Oxford University Press, pp. 179-95.