Theories of Growth, Innovation, and Entrepreneurship

Abstract

One of the most important—and most difficult—areas of research in economics concerns the mechanisms that cause higher growth and increased prosperity. Economists base their work on theoretical models that are expected to capture the complex relationships of real-world behavior. Policy conclusions are then derived from these simplified models. However, if a model is based on incorrect or over-simplified assumptions, these conclusions will likely prove to be just as flawed.

One of the most important—and most difficult—areas of research in economics concerns the mechanisms that cause higher growth and increased prosperity. Economists base their work on theoretical models that are expected to capture the complex relationships of real-world behavior. Policy conclusions are then derived from these simplified models. However, if a model is based on incorrect or over-simplified assumptions, these conclusions will likely prove to be just as flawed.

This is clear in the analysis of growth and transformation. The current growth paradigm is based on theories developed in the late 1980s and beginning of the 1990s, notably by Robert Lucas, Paul Romer, Philippe Aghion, and Peter Howitt. These models highlight investment in knowledge—measured as education and R&D—as the main source of growth and have had profound implications for economic policy. Among other things, the early ambition that three percent of the EU’s total budget should go to R&D can be linked to this theory, as well as Sweden’s “knowledge boost”—a government investment program for adult education in 1997–2002—and the massive expansion of regional universities and colleges.

Knowledge is undoubtedly crucial for economic growth. The great leaps forward in the material development of mankind—such as the first Industrial Revolution in the late 1700s and the second one a century later—were based on new knowledge, new technology, and transformative innovations. The same is true for the IT revolution, followed by the digitized technology of our own time. Nevertheless, econometric analyses of the effects that knowledge investments have on growth, measured in terms of R&D or education, do not show unambiguously positive results (Gordon 2012). A simple correlation between R&D investment and growth for OECD countries during the 2000s rather indicates a weakly negative covariation (see Fig. 2.1). However, at finer levels of disaggregation, such as the industry or firm level, the results are more robust and generally positive, especially for private sector R&D. It is more difficult to demonstrate positive effects of publicly funded R&D initiatives aimed at the business sector (Bergman 2012), although publicly funded basic research seems to have positive effects, even if the time lag can be sizeable.¹

Fig. 2.1

The relationship between annual GDP growth and R&D investment in OECD countries, 1981–2021, R&D lagged eight years. Source: Updated from Acs et al. (2009), OECD

One explanation for this relatively weak connection is that knowledge-based growth models primarily look at how knowledge is acquired and how much knowledge (measured in various ways) is produced, but they do not explain how it is disseminated and transformed into economically valuable goods and services. What is measured and included as explanatory factors in these models—R&D as a share of GDP serving as a typical example—are therefore not necessarily the most relevant factors for the questions we are trying to answer regarding forces that promote innovations. This has led to the emergence of empirical research which shows that investment in knowledge and research should be supplemented by, for example, entrepreneurship, competition, knowledge sourcing, and mobility between and within industries and firms in order for economic growth to ensue.² The institutions—laws, regulations, and norms—that are relevant for the transformation of knowledge into societal benefits are thus central to growth. These represent different mechanisms and policy areas than those that are emphasized in today’s dominant but narrow growth models, namely quantitative measures of R&D and education.

In this chapter, we discuss how research on economic growth has evolved since Schumpeter’s pioneering work in the early twentieth century, emphasizing the microeconomic foundations of growth. Significant progress has been made in recent decades, but a number of “unknowns” remain. We compare the knowledge-driven growth models with the evolutionary growth models that have been developed in parallel. The impact of the latter models in guiding economic policy varies, but knowledge-driven models are arguably the most influential. To a greater extent, the Schumpeterian and evolutionary models highlight the importance of institutions that influence competition, search costs, and industry-level routines, which in turn create the conditions for entrepreneurs and firms to engage in innovation. However, countries with similar formal institutions show large differences in growth. This indicates both that there may be substantial differences between de jure and de facto institutions and that informal institutions (norms) are important, but also that the design of institutions at a more detailed level affects incentive structures and growth.³ In Chap. 3, we will devote additional space to discussing the importance of both formal and informal institutions for innovation and entrepreneurship.

2.1 Development of Growth Models

What are the factors that drive economic growth? At a general level, there is consensus that credible institutions that promote property rights, transparency, and basic education are necessary—but not sufficient—conditions. The effect of other variables is even more uncertain. Some countries in Asia show strong growth but under different institutional conditions than many mature industrialized countries feature, often based on imitating technology from the leading economies. More generally, growth rates can also differ significantly among countries with comparable institutions.

Before we begin our examination of the various growth models, we begin by defining two key concepts—innovation and entrepreneurship—to which we will devote a great deal of attention hereafter⁴:

• Our definition of “innovation” is based on the widely implemented version presented in the Oslo Manual (OECD 2018), although we emphasize the market perspective. Hence, we define “innovation” as a new or improved product or service, a new form of organization, new inputs or new markets, or a combination of these aspects. Inventions, scientific findings, or technical discoveries do not necessarily have a market value—as a rule, an entrepreneur, who can identify a market opportunity, is required in order for these to attain a specific market value. Innovation spans most sectors, industries, and economic activities. At the same time, the difficulties of measuring innovation are obvious: R&D expenditure and patents are the most commonly used measures, but they are obviously difficult to apply, for example, to organizational changes, to the identification of new markets, and in the service sector (Nagaoka et al. 2010). In addition, R&D expenditure is an input measure, while we are interested in the return it provides in the form of innovation output. Novelty is not sufficient to “earn” the designation of an innovation; it must also be economically valuable.

• “Entrepreneur” and “entrepreneurship” can be defined on the basis of various characteristics and functions (Braunerhjelm et al. 2022; Hébert and Link 2007). The more well-known definitions include Joseph Schumpeter’s (an agent that disturbs the equilibrium state of the economy), Frank Knight’s (a bearer of uncertainty), and Israel Kirzner’s (one who drives the economy towards equilibrium). To these classic definitions, additional ones have been added that are based on the entrepreneur’s function/area of activity, for example intrapreneurs, social entrepreneurs, and gig economy entrepreneurs. All definitions emphasize the entrepreneur as an agent of change, a force that drives development. This view returns to Schumpeter’s original definition of the entrepreneur as a disrupter of economic equilibrium. We will use the following general definition: entrepreneurs are the agents of change in the economy whose actions result in restructuring, market experimentation, and dynamism.⁵

Entrepreneurship and innovation have undoubtedly played a crucial role in previous leaps in growth and economic development. This is illustrated by the second industrial revolution of the late nineteenth century, which was marked by major technological breakthroughs, including electricity and the internal combustion engine,⁶ paralleled by the emergence of new firms and industries based on these achievements. Characteristic for this period were reforms that included both knowledge upgrading (such as compulsory schooling) and improved opportunities for the transformation and dissemination of knowledge into public goods (for example, increased competition and limited risk-taking through incorporated businesses).⁷ Then, as now, this was preceded by increasing internationalization as trade volumes increased and cross-border investment grew, as did the cross-country mobility of labor. This was the environment that inspired Schumpeter (1934 [1911]) to launch his pioneering work on the entrepreneur as the primus motor of industrial transformation, dynamism, and growth. Much later, Baumol (2002, 2010) even claims that radical entrepreneurial innovations explain at least 90% of economic development since the onset of the Industrial Revolution. It is therefore self-evident, he argues, that entrepreneurship and innovation should be as much a part of economic theory and economics education as the role of markets and price mechanisms.

2.1.1 The Neoclassical School

During the 1930s and 1940s, Schumpeter’s entrepreneurial vision of growth was supplanted by more macro-oriented analyses. In the aftermath of extreme economic fluctuations and the disastrous depression of the early 1930s, Swedish economists of the Stockholm School (notably Gunnar Myrdal, Erik Lindahl, Erik Lundberg, and Bertil Ohlin) and John Maynard Keynes began to emphasize the role of the demand side. By using fiscal policy (taxes and public expenditure), monetary policy (interest rates), and foreign exchange policy (changes in the exchange rate), the government was able to influence overall demand in the economy and thereby temper cyclical fluctuations. The importance of entrepreneurship, business ownership, and other supply-related factors (such as technological progress) did not attract the same interest. For a long time, these Keynesian models also worked relatively well, especially when there were untapped resources—mainly labor—that could be employed in productive activities.

During the 1950s and 60s, growth models based on this thinking were further developed and formalized. Growth occurred as an interplay between investment, population growth, and consumers’ willingness to save. Consumers were prepared to refrain from consuming for a period—that is, to save—provided that the interest rate (defined by the marginal productivity of investing an additional unit of capital) at least corresponded to their rate of time preference (the discount rate). If the interest rate was higher, savings and investment would increase until the marginal productivity of capital returned to the level of the discount rate. Analogously, if the consumers’ discount rate exceeded the interest rate, consumption would increase while savings and investment would decrease. The growth rate stabilized when the net productivity of investment reached a certain level, i.e., when a steady state had been achieved. Given stable population growth, economies grew at a steady pace. In line with this model, growth policy was designed primarily to promote investment by lowering the cost of capital (possibilities for deductions, tax relief for internal investment funds, etc.) and to increase labor supply.

However, despite its elegance, this theory did not reflect actual development. Robert Solow (1956, 1957) demonstrated that the bulk of economic growth—perhaps as much as 80%—remained unexplained after the effects of increased investment and employment had been accounted for.⁸ The explanation was attributed to technological advances and knowledge enhancements, popularly known as the Solow residual, or as “the measure of our ignorance” (Abramovitz 1956, p. 11). But the mechanisms behind technological progress and knowledge growth remained a mystery. This was unsatisfactory, not least because the residual was often larger than what could be explained by the theory, i.e., growth became exogenous and not something captured or determined within the framework of the model.

The residual was assumed to contain primarily new and improved technology and better trained staff as well as innovations. Recent research has pointed out that organizational changes, changes in industry composition and markets, start-ups and the closure/exit of firms should also be included⁹—that is, much of what Schumpeter called “creative destruction.” Many of the variables omitted from the standard growth model of the 1950s and 1960s also have clear policy relevance, but since it was not possible to identify which ones that primarily affected growth, no definitive policy conclusions could be drawn.

A decade later, Jorgenson and Griliches (1967) proposed a model in which the factors of production were quality-adjusted (labor with respect to human capital and physical capital with respect to its level of technology), thereby eliminating large portions of the Solow residual. However, to be able to do this, they were forced to make some bold assumptions, notably that the stock market accurately values firm equity. The Jorgenson–Griliches model can be seen as a stepping stone towards the endogenous or knowledge-based growth models of the 1980s.

2.1.2 Endogenous, Knowledge-Based Growth Models

The next foundational step in the modeling of economic growth was pioneered by Paul Romer (1986, 1990) and Robert Lucas (1988), who developed the first knowledge-based growth models. These are referred to as endogenous growth models since knowledge and knowledge investment, which in the earlier models were treated as exogenous and part of the Solow residual, now become determined within the model.

Romer’s first model was built on three factors of production: capital, labor, and knowledge. Knowledge was assumed to be accumulated partly through R&D investments of firms, and partly as a result of spillovers from the aggregate stock of knowledge in the economy. Companies’ R&D investments were thus internal, but at the same time some of the new knowledge spilled over into the aggregate stock of knowledge. Goods production used knowledge, capital, and labor. Knowledge was assumed to be non-rival and thus gave rise to economies of scale. Even if labor and capital were kept constant, increases in the knowledge base would lead to increased production, higher productivity, and growth. Consequently, all firms would benefit from increased R&D investments. Asymmetric information and the risk that some firms could become “free riders” may then keep R&D investment at too low a level from a societal point of view, which opens the door for government policy to stimulate investment in R&D. In other respects, the model remains faithful to the original neoclassical growth model: firms are assumed to be price takers and make zero profits in the steady-state equilibrium.

In a subsequent article, Romer (1990) extended his model by positing a more realistic market structure. This model is based on four factors of production—labor, human capital, physical capital, and new knowledge (technology)—which are employed with different intensities in three sectors. In the production of new knowledge (i.e., R&D), only human capital is employed, which also utilizes the aggregate stock of knowledge, i.e., previously accumulated knowledge. The output is a new technology or design that is used together with capital to produce new capital goods (semi-finished products and other inputs). Finally, these differentiated capital goods are combined with labor and human capital to manufacture consumer goods. Production of the three types of goods—new productive knowledge, capital goods, and consumer goods—can take place within a single firm or be distributed across several firms.

Romer makes a number of strong assumptions to keep his model analytically tractable. Population is constant, as is the proportion that is highly educated; capital is assumed to be produced with the same technology as consumer goods; and knowledge is immediately available to all actors in the economy. New capital goods never become obsolete and are protected by perpetual patents. The development costs that businesses incur lead to a market structure of monopolistic competition. Companies cover these costs through a surcharge on the (given) price that a new product or quality makes possible. In equilibrium, the entry of new firms/products means that costs can be precisely covered. As a result, no profits are made.

The R&D sector is thus central to this framework. Not only does this sector determine the growth rate; it also produces both firm-specific and generally available knowledge. The volume of new knowledge that is produced is assumed to be determined by the quantity of human capital in the R&D sector, the size of the aggregate stock of knowledge, and the productivity of R&D personnel. The size of the aggregate stock of knowledge will in turn affect productivity. To avoid an explosive increase in the growth rate, Romer also assumes that the growth effect of more research results is linear: If the number of researchers increases, the quantity of new knowledge also increases in the same proportion. Thus, given that the human capital share is constant in the R&D sector, the output of the R&D sector is assumed to be directly proportional to the aggregate stock of knowledge. This assumption has been questioned by Jones (1995a, b) among others, and this will be discussed later. The relationship between the size of the aggregate stock of knowledge and the productivity of R&D workers also means that the difference in growth and prosperity between industrialized and developing countries can be expected to increase, as the industrialized countries have a significantly larger aggregate stock of knowledge. Given these far-reaching and sometimes extreme assumptions, Romer argues that there exists a long-term stable growth path and that economic policy can be used to increase the steady-state rate of growth.

Far-reaching simplifications and strong assumptions notwithstanding, these knowledge-based models provided important insights into the role of knowledge in the growth process. First, investments in human capital and R&D are explained in terms of profit-seeking firms and individuals competing by means of a stronger knowledge base, higher quality products, and new goods and services. Second, investments in knowledge lead to large and sustained spillover effects that benefit other firms. Firms that invest in the creation and discovery of new knowledge will not be able to keep it entirely to themselves—some of it will “spill over” to other firms, thus increasing the aggregate stock of knowledge, which in turn boosts productivity and growth in all firms.

This argumentation led to an important policy conclusion. Since knowledge production (R&D) in the model is assumed to be privately financed, firms will underinvest in new knowledge because their own investment will partly benefit other firms including their competitors. At the same time, knowledge investments benefit society at large as they lead to a higher growth rate and rising incomes. Consequently, this version of Romer’s endogenous growth model also provides arguments for subsidies and tax incentives to stimulate investment in R&D.

2.1.3 Neo-Schumpeterian Growth Models

A new generation of knowledge-based growth models appeared in the early 1990s. Pioneers among these so-called neo-Schumpeterian model builders include Segerstrom (1991), Aghion and Howitt (1992, 1998) and Cheng and Dinopoulos (1992).¹⁰ Here, innovations are perceived as resulting from “competitions,” where the winner gains a temporary monopoly. At the same time, the innovation makes existing knowledge obsolete, and firms based on obsolete knowledge are eliminated. Innovation thus becomes a competitive tool that creates a willingness to pay for new, improved products.

These models claim to capture Schumpeter’s concept of creative destruction, which is partly correct. But at the same time, they focus on very specific and limited types of innovation and entrepreneurship originating in R&D that can most closely be likened to the activity of researchers at a large pharmaceutical firm. In these models, the entrepreneur appears exclusively as an agent who pursues R&D investments where it is assumed that returns on innovation investments follow an ex ante and objectively known probability distribution. The expected costs and returns of innovations are thus objectively calculable, and the value and economic uses of innovations are known once a new product or technology has been developed.

The entrepreneur is thus conceptualized as a decision-making agent who is responsible for allocating resources between two activities: goods production and R&D. As such, this model fails to capture the role of the entrepreneur within the firm, i.e., whether the role can be filled by a manager or whether “entrepreneur” refers to the owner(s). Hence, his or her activities differ substantially from entrepreneurship and innovation as construed by Schumpeter, Knight, and Kirzner. Consequently, the models’ policy conclusions are questionable. They not only underestimate the role of small firms and start-ups, but they are also unable to capture growth made possible through improved organizational forms and more rigorous competition.

2.1.4 Critique of Endogenous Growth Models

Knowledge-based growth models represented a significant step forward in the understanding of growth, insofar as Solow’s residual could—at least in part—be explained and integrated into the model (i.e., endogenized). Early on, however, several weaknesses were identified related to the assumptions that form the basis for these models. Some of the criticism was directed at the lack of realism in the assumptions regarding knowledge investment and knowledge development:

• Previous research showed that opportunities to assimilate knowledge appeared to be cumulative, i.e., existing levels of knowledge affected the development of new knowledge. The endogenous growth models partially account for this, but they subsequently imply that path dependence is prevalent. This may lead to lock-in effects that limit the spread of new knowledge as do varying levels of absorption capacities (Cohen and Levinthal 1990) and costs of absorbing new knowledge (Mansfield et al. 1981).

• Understanding economic growth implies considering the historical time path of economic development, which tends to be punctuated by “eras” when general purpose technologies emerge or techno-economic paradigm shifts occur (Bresnahan and Trajtenberg 1995; Freeman and Louça 2002). Similarly, processes of economic growth are embedded in and dependent on the institutional setup (Lundvall 1992; Nelson 1993; Braunerhjelm and Henrekson 2016). These aspects loom considerably larger in evolutionary approaches to economic growth (see below).

• Although knowledge can be spread across regions and countries (Coe and Helpman 1995), in principle there is agreement that the dissemination of knowledge is geographically limited. Even though technological advances in transmitting and sharing knowledge have facilitated its diffusion, the concentration of knowledge-dense areas is still a prevalent phenomenon (Andersson and Larsson 2022). In addition, the more advanced the new knowledge, the more difficult it is to interpret, codify, and apply commercially, and the more importance proximity assumes in assimilating such knowledge (Polanyi 1958). Innovation has proven to be even more geographically concentrated than both R&D and production (Ejermo 2009). This (as well as path dependence and lock-in effects) contrasts strongly with the typical model of endogenous growth in which the dissemination of knowledge takes place automatically and without cost.

• Weak incentive structures and low potential for organizational learning generally limit a company’s capacity for dynamism, i.e., its R&D development and the application of results in production. Larger companies are more risk averse in their respective technology and product areas (Christensen 1997), while radical innovations can be attributed to newer firms (Casson 2003; Baumol 2004).

• Smaller firms are generally more oriented towards the service sector and more focused on innovations that do not originate in R&D, while most knowledge-driven growth models have endogenized innovation solely through R&D investments. The dominant growth models lack the Schumpeterian entrepreneur who assimilates and exploits knowledge in ways that are not visible in the R&D statistics, but which still spills over to other firms. IKEA, Starbucks, and Ryanair are examples of innovative firms with little or no research in the narrow sense, although many of them invest significant resources in development and design.

• The somewhat weak connection between R&D and increases in growth/productivity may reflect the fact that chance and coincidence also play a role in producing successful innovations, that there is a high degree of imitation, and that R&D investment and innovation initiatives take place irregularly and even in firms on the verge of financial collapse. Previous research also shows the importance of continuity and sustainability in innovation initiatives; empirical studies need to span lengthy periods (Roper and Hewitt-Dundas 2008). In addition, and as mentioned earlier, innovation is associated with significant measurement problems where gradual (incremental) innovations rooted in learning by doing are seldom or never captured in the statistics. Finally, the rate of disruptive innovation seems to have slowed down, which may reflect the reliance on a narrower set of existing knowledge (Park et al. 2023).

• Another aspect of the critique is more model oriented. Jones (1995a, b) observed that early endogenous growth models included a scale factor, which implied that technological change and innovations were proportional to R&D investments and that population was assumed to be constant. The proportion of researchers (and the proportion employed in manufacturing) was also assumed to be constant. All other things equal, this means that if R&D costs (researchers) are doubled, growth will also double. However, Jones pointed out that this is inconsistent with observable facts: the number of researchers has increased sharply in recent decades with no corresponding hike in the growth rate. Instead, Jones suggests that productivity in the R&D sector should stand in an inverse relationship to the level of accumulated knowledge. By positing a declining rate of return in the R&D sector, the model becomes more realistic. As R&D becomes more difficult, the rate of technological change, the pace of innovation, and the aggregate rate of growth decline.¹¹

Criticism of endogenous growth models thus takes various forms but is mainly directed at how knowledge is disseminated and transformed. It should be emphasized that while the weakness of the earlier neoclassical model was that knowledge was perceived as “manna from heaven,” knowledge-based models fail to explain how knowledge is spread. At present, its conversion into commercial goods is based on abstract assumptions and consequently becomes exogenous in the model. Moreover, the utilization of new knowledge is not associated with any costs for firms. This shortcoming is probably one reason why the empirical literature does not find unequivocal support for the notion that investment in R&D, and to some extent education, has positive effects on growth.

However, recent empirical research has identified certain mechanisms as particularly important for disseminating and transforming knowledge into economically valuable goods and services. These include labor mobility, entrepreneurship, and advanced clusters.¹² These factors are in turn affected by the institutional (largely politically determined) framework within which they operate. The causes of different growth rates across countries and regions should thus be sought, among other things, in how the conditions under which entrepreneurs and firms transform and develop knowledge vary over time.

Even more importantly, by depicting the entrepreneur as an actor whose economic function is to invest in calculable outcomes, the role of the neo-Schumpeterian entrepreneur is relegated to that of a routine decision-maker in pursuit of discoverable business opportunities. However, as emphasized by Knight (1921), many—perhaps most—innovations are undertaken without full information on their potential value, not even in a probabilistic sense. The activity is marked by genuine uncertainty. This implies that innovations lack strictly objective benefits against which their costs can be weighed. Instead, they can be expected to be wholly or partly pursued based on the subjective valuations and judgment-based decisions of individual entrepreneurs (e.g., Bylund and Packard 2021). Thus, given elements of genuine uncertainty, entrepreneurs cannot solely rely on objective knowledge regarding the final economic uses of ideas to determine their expected economic value. Instead, they must maintain an active role in identifying the economic uses of innovations if they wish to appropriate their economic value.

At the same time, introducing incalculability and subjectivity into the economic models does not imply that innovation outcomes are driven solely by chance and subjectivity. On the contrary, several determinants of innovation success can likely be incorporated to increase both the causal interpretability and predictive power of existing frameworks. Notably, Knight (1921) stresses the central role of the knowledge, experience, and innate abilities of entrepreneurs in the selection and outcome of disruptive innovations, i.e., what he refers to as “judgment.”¹³ For example, it is likely that the tacit knowledge gained from past experiences of creating and exploiting innovations is a core element of entrepreneurial acumen.

Given that innovations are, at least partly, associated with genuine uncertainty, this implies that extant neo-Schumpeterian growth models run the risk of providing misleading guidance to policymakers aiming to stimulate economic growth. A potential counterargument is that neo-Schumpeterian growth models seek to explain and predict the macroevolution of the economy, and at the aggregate level, it may be fair to abstract from the genuine uncertainty of innovative outcomes at the microlevel.¹⁴ Although the validity of this assertion is debatable per se (Frydman et al. 2019), this line of reasoning is also likely to be controversial in this specific context for at least two reasons. First, given that economics seeks to explain the causes of economic growth, a deeper causal understanding is called for. Second, economists aspire to provide reliable policy advice and the adequacy and precision of policy proposals hinge on a good causal understanding of the growth process and its microeconomic foundations.

2.2 Evolutionary Growth Models: Schumpeter’s Legacy

As mentioned before, Schumpeter considered the entrepreneur to be the agent that transformed knowledge into innovation. By developing and combining both new and existing knowledge in new ways or in new contexts, the entrepreneur contributes to creative destruction and economic development. Sometimes the researcher/inventor/entrepreneur can be one and the same person, but this seems to be the exception rather than the rule. On the other hand, the outcome tends to be successful when researchers or inventors collaborate with entrepreneurs, because such collaborations increase possibilities for commercialization (Braunerhjelm and Svensson 2010).

What Schumpeter did not anticipate was how small and new businesses can collaborate with large incumbent firms—something that has been facilitated by new information and communication technologies. On the contrary, he argued in his later work (Schumpeter 1942) that investment in R&D and innovation by large firms would disadvantage smaller ones—which, he feared, would hobble capitalism and undermine it in the long run. However, new research suggests that large firms can create a market for entrepreneurial ideas and thereby contribute to innovation and entrepreneurship (Norbäck and Persson 2012).¹⁵ As we pointed out above, technological developments are also likely to have diminished economies of scale in several areas.

2.2.1 The Role of Entrepreneurship

The idea that knowledge and skills are spread over a large number of individuals and firms dates back to Menger (1871) and Hayek (1945) and characterizes the older Austrian school.¹⁶ At both the individual and firm level, opportunities for renewal and innovation therefore differ radically from one situation to another, as does the expected outcome of such initiatives. Based on this view of the economy—decentralized knowledge and the spontaneous confluence of individuals and ideas with the surrounding economic policy environment—it becomes much more difficult to formulate an economic policy that promotes innovation in a targeted manner.

A complex, non-linear economy that deviates from the traditional equilibrium model always features unexploited opportunities and inefficiencies which have consequences for how the economy functions and develops. Continuous experimentation is required—to test, alter, innovate, and imitate—to identify both business opportunities and workable methods of production and distribution (Eliasson 2009; Dosi and Nelson 2009). Information is not only important and scarce (and therefore precious) but also dispersed. Individuals have different information about different things, and their interpretations of that information may also differ. Not even the most knowledgeable expert, economist, or entrepreneur can be well informed about more than a fraction of any country’s industries and sectors.

As information is scattered and fragmented, economic decision-making needs to be decentralized. Centralized states are finding it increasingly difficult to manage an economy consisting of millions of employees and consumers and hundreds of thousands of firms as they become more sophisticated and knowledge intensive. In the same way, large, centrally controlled firms will find it difficult to focus effectively on more than a few specific markets. In an advanced economy, it is therefore vital that its main actors—each with disparate fragments of knowledge but no full perception of the whole—can act on the basis of their own information. In business, this is done through reorganization and decentralization within firms, and through entry and exit.

Economic growth is thus driven by the identification (or generation), commercialization, and selection of successful business opportunities:

• The identification process is characterized by the ability to identify (generate) new ideas and innovations.

• The commercialization process is characterized by the will and ability to introduce these to the market.

• In the final selection process, inferior innovations are screened out and replaced by better ones.

In this way, the economy is in perpetual motion, continuously exposed to pressure to adapt and transform (Acs et al. 2009). In such a dynamic economy, products, firms and sometimes entire markets disappear and are replaced by novel, better products and more efficient firms. New markets or niches function as experimental workshops where new ideas are tested against old ones; the most successful survive, while those without a future are discontinued and thereby free up resources that can be used elsewhere.

2.2.2 The Importance of the Entrepreneur for Growth

Figure 2.2 schematically illustrates the market process and the importance of entrepreneurship for growth and economic development, as these appear within the evolutionary growth framework described above. New entrepreneurial discoveries are identified (generated) and commercialized in the market, where a selection process takes place. This market process leads to both direct and (more long-term) indirect effects.

Fig. 2.2

Entrepreneurship and economic growth. Source: Further developed from Fritsch and Mueller (2004)

There are two direct effects. First, if entrepreneurial commercialization is successful, new capacity and new structures are created, either by the founding of new firms or the expansion of existing ones. The second direct effect is the exclusion of capacity or business-stealing effect. Old operations lose profitability and are replaced by new ones, which can also turn out to be poor investments, become unprofitable, and need to be liquidated.

In addition to these direct effects, at least six indirect effects can occur that affect output: higher efficiency, more rapid structural change, an increased propensity to innovate, a greater variety of goods and services, new skills (increased entrepreneurial human capital), and the creation of role models. These indirect effects occur mainly through tougher competition and are crucial for the long-term development of an economy. The entrepreneur plays a decisive role in this process, functioning here as an active agent of change.¹⁷

Furthermore, entrepreneurship often has a self-reinforcing effect. New discoveries and products generate new opportunities. In Fig. 2.2, this is marked by the arrow from “growth” back to “new entrepreneurial opportunity,” as entrepreneurship itself gives rise to new opportunities. An influx of new entrepreneurs can also have a “demonstration effect,” i.e., a new business can act as a signal to other potential entrepreneurs to take the step of starting a business themselves (Verheul et al. 2001).

Aggregate data make changes in economic growth seem fairly small. In developed countries, economic growth rarely exceeds three percent, but the aggregate figure conceals a more tumultuous reality. Economic growth is not primarily about firms growing by a similar percentage or productivity rising in existing jobs because of technological change and more capital per worker. More accurately, growth emanates mainly from churning (firm and job turnover) and restructuring—primarily shifts in production from less to more successful firms within narrowly defined industries, rather than from declining to growing sectors (Caballero 2007).

Growth requires some firms to fail or contract so that resources can shift to entering and expanding firms. Growth presupposes structural transformation: new firms manufacturing new products in new ways and old ones innovating and reorganizing or liquidating. In order to achieve growth, substantial turnover of companies and jobs is required. This is indeed the essence of the creative destruction process envisaged by Schumpeter (1934 [1911]).

As shown in Table 2.1, over the four decades 1977–2016, new jobs averaged 16.4% of total jobs, a third of them in new firms; 14.5% of jobs were lost annually through closures and contractions. The net result was an aggregate annual job growth of 1.9%. As a consequence, this 1.9% net gain was associated with a gross job reallocation rate of 30.9% (16.4 + 14.5) and thus with an excess job reallocation rate—the amount of job-churning beyond the minimum required to accommodate the net employment change—of 29%.

Table 2.1 Job creation and destruction in the U.S. economy, annual average, 1977–2016

Although churning is higher in the United States, extensive churning is pervasive in all OECD countries and more so in the wealthiest ones.¹⁸ At least 80% of the reallocation of workers in developed countries takes place within narrowly defined sectors.¹⁹ This reallocation has two main drivers: adjustment among firms with different technologies, and experimentation with improved products, management, and other production systems. Excess job reallocation rates are higher for newer plants because of greater uncertainty, experimentation, and variability in the quality of goods produced.

2.2.3 The Importance of New Firms

The indirect effects referred to above are often linked to new firms. In practice, an influx of new entrepreneurial firms is essential for an economy’s development, renewal, and transformation. Although entrepreneurship can take place within incumbent firms and among employees,²⁰ new and (at least to begin with) small firms are required to maintain a sufficiently high level of innovation pressure. Moreover, the objectives and effects of innovative activities differ between young and small firms on the one hand, and more established incumbents on the other. A dynamic innovative environment needs both types of innovators.

New firms expose existing ones to competition and encourage them to become more efficient while contributing to structural change and innovation. Incumbent firms are often tied to existing technologies through extensive investments in physical and human capital, which can become obsolete in the face of radically new innovations. This applies not only to investors (who have invested capital in a certain technology and business plan) but also to employees (who master a certain technology and production process). Thus, it is not only investments in financial capital that are threatened by new challengers, but also old investments in human capital.

An incumbent firm that develops new products thus competes with itself, as its new products can erode the profit made on its established products. This may weaken the firm’s motivation to further innovation. An innovation may also require a completely new organizational or compensation structure (Cullen and Gordon 2006). As a result, genuinely new products and production methods may be difficult to introduce in large, mature firms. Incumbent firms instead tend to safeguard and exploit their already existing markets, while new products are best produced in new firms, which are often established precisely for this purpose.²¹

Hence, a division of labor between large and small firms seems to have emerged. Large ones are relatively better at R&D focused on improving existing products, while radical innovations often emerge in smaller ones. The latter, in turn, are often spin-offs from larger firms (Andersson and Klepper 2013; Klepper 2016). New technology is thereby developed, implemented, commercialized, and often disseminated in the form of new entrepreneurial firms. William Baumol has shown the importance of small businesses for the emergence of many revolutionary American innovations, which have since in many cases been further developed and reached their full potential in large firms. Baumol (2004) speaks of a “symbiosis between David and Goliath.”²² Many incumbent firms acquire other firms precisely to gain access to new technology.

2.2.4 Entrepreneurship as a Factor of Production

To equate business with entrepreneurship is to downplay the special skills that are necessary for innovative entrepreneurship (see also the Appendix to this chapter for a discussion of how entrepreneurship should be measured). Most firms are neither innovative nor growing, and most entrepreneurs do not have, nor will they ever have, a single employee in addition to themselves. It is thus important to distinguish between business owners as a group and the smaller number of fast-growing firms where entrepreneurship is more prominent. Potentially innovative entrepreneurs are few, not possible to identify ex ante and not easily interchangeable. They also tend to already have secure and well-paid jobs in the career hierarchies of existing businesses, which they must relinquish if they wish to engage in independent entrepreneurship.

In line with Audretsch and Keilbach (2005) and Baumol (2010), we find it fruitful to treat entrepreneurship as a separate factor of production. What the entrepreneur does in the start-up phase of a business is precisely this: he or she creates more capital, both using the firm’s existing capital and by means of his or her own specific entrepreneurial labor. This capital can be based on science or technology, and it may also be organizational or structural. In the case of a successful start-up, the economic value of this new capital is many times greater than the financial resources invested. Companies such as Moderna, Skype, or Tesla serve as striking examples. We would argue that in economic models that seek to achieve a deeper understanding of innovation, dynamism, and growth, it is necessary to include entrepreneurship as a separate factor of production that includes unique characteristics providing a distinct contribution to the production result.

In market transactions, prices and volumes can be measured, which means we can distinguish the return on labor and capital, respectively. For entrepreneurial activities, such measurement is impossible, as the return is a result of the value generated through the combination of the entrepreneurs’ own labor, their entrepreneurial input, and financial resources. Entrepreneurship interacts with other input factors and can thus be described as an indivisible bundle of these inputs. An entrepreneurial firm whose founder does not reinvest a high proportion of the return will generally not be able to grow. Entrepreneurship is largely about building companies that can generate future returns, i.e., creating capital by means of one’s own labor and previously built-up capital.

Another important argument for treating entrepreneurship as a separate factor of production is that, empirically, entrepreneurs seem to behave differently than employees (Baumol 2010; Hurst et al. 2014). For example, their behavior is more sensitive to financial incentives than that of hired workers. Comparisons generally show that the incomes of the self-employed are affected more by taxes (more tax-elastic) than those of employees, perhaps because the self-employed have greater control over their working hours and how they report their income (Chetty et al. 2011; Kleven and Schultz 2011; Harju et al. 2022). This is an argument for taxing entrepreneurs differently than employees in certain contexts; for more on this aspect, see Chap. 5.

Braunerhjelm and Lappi (2023) provide additional evidence that entrepreneurs should be viewed as a separate factor of production, although from a somewhat different angle. By introducing a new and hitherto neglected measure of human capital, defined as employees’ former involvement in entrepreneurship, they investigate the influence of such entrepreneurial human capital (EHC) on firm performance. Based on longitudinal register data for Sweden over the period 1993–2018, they construct a stock variable of EHC for all private incorporated firms. The results strongly support the observation that higher EHC among employees is associated with higher levels of productivity and innovation. More precisely, a ten percent increase of employees who are former entrepreneurs increases firm-level productivity by 3.9%. The results are shown to be robust to adding control variables, estimation techniques, alternative definitions of EHC, and other performance measures.

To summarize here, it seems that the entrepreneur fulfills an important function in converting a scientific discovery or an invention into an innovation that can be commercialized and introduced in the marketplace. The entrepreneur is thus the missing link in knowledge-driven or endogenous growth theory, responsible for transforming knowledge into innovation.²³ Based on this insight, the concept of the entrepreneur—the agent of change in the economy—becomes strategically decisive and a starting point for economic policy. Thus, exclusive investment in R&D and education, without further analysis of how knowledge is disseminated and how entrepreneurs can use it to bring about change, risks becoming sterile or sub-optimally exploited.

2.2.5 The Evolutionary Approach to Economic Growth

Schumpeter coined the term “creative destruction” to describe an evolutionary market dynamic characterized by selection, dynamism, and growth. He expressed it in this way: “The essential point to grasp is that in dealing with capitalism we are dealing with an evolutionary process” (Schumpeter 1942).²⁴ In recent decades, an evolutionary growth approach has been developed in parallel with the endogenous growth models. It emphasizes conditions and opportunities at the microlevel, i.e., the opportunities for individuals and firms to exploit new and existing knowledge for innovation purposes. This perspective also underlines the importance of diversity, variety, and selection. Small firms and start-ups are significant because they can be expected to work with different varieties and combinations of new and existing knowledge, testing them on the market.²⁵ These innovative activities are characterized by experimentation, uncertainty, and risk-taking, where a product’s commercial potential is ultimately decided in the marketplace.²⁶

Evolutionary models emphasize disequilibrium dynamics as a general feature in the search for new production methods, new products, and economic behavior in the broader sense. This process entails trial and error, gross mistakes, and unexpected successes, as firms persistently search for and adopt new technologies as well as new organizational forms and new behavioral patterns in order to gain advantages over their competitors. Markets are characterized by experiments and uncertainties about how new knowledge is best combined and applied, which generates an influx of new firms, firm growth, and corporate failures. This is what Metcalfe (2000) refers to as “restless capitalism.” Different abilities to innovate and imitate are central aspects and drivers of industrial evolution, shaping the patterns of growth, decline and exit over populations of competing firms, as well as the opportunities for entry of new businesses. The dynamics of evolutionary processes are then driven by the twin forces of idiosyncratic learning by persistently heterogeneous firms, on the one hand, and (imperfect) market selection delivering prizes and penalties—in terms of profits, growth opportunities, and survival probabilities—on the other across heterogeneous corporate populations. These dynamic processes rhyme less well with the stereotype entrepreneur in the neo-Schumpeterian endogenous growth models.

Another key component is that replication and adoption of technological knowledge concerning processes, organizational arrangements, and products are associated with costs and uncertainty linked to the tacit elements involved in technological know-how (Mansfield et al. 1981; Dosi and Nelson 2009; Maurseth and Svensson 2020). This creates lumpiness, retards the diffusion and application of new technology, and strengthens path dependence. As knowledge about the new technology is accumulated, recipes—that is, coded programs—are increasingly used to implement even newer technologies. Meanwhile, Winter (2016) stresses the importance of including alternative strands of science, e.g., psychology, to better understand how human nature and entrepreneurial behavior can be explained. In particular, entrepreneurial behavior may be less constrained than presumed by the forces of habit and fear of uncertain outcomes and ultimately failure, thereby diminishing the risks that the entrepreneur will be daunted or delayed by path dependence.²⁷

The roots of these insights originate in work by Hayek (1945) and von Mises (1949). They argue that unevenly distributed individual abilities and capacities play a central role in transformative processes and growth. Subsequently, it is important to examine how this dynamic is affected by the institutional framework within which entrepreneurs operate. Nelson and Winter (1982) were the first to present an evolutionary growth model incorporating several of the features discussed above. One starting point was that firms are generally reluctant to change their operations, which, in combination with endless possibilities for change and a finite ability to rationally review these possibilities (bounded rationality), creates a need for rules of thumb, or more precisely, routines. Firms are assumed to be continuously involved in a search process either to develop new routines themselves (R&D), which Nelson and Winter call innovation (process innovation), or to imitate other firms. All search behaviors are associated with costs; the probability of discovering an improvement increases as R&D or other search costs increase. Innovation thus requires more resources but can also generate higher returns. Finally, it should be noted that Nelson and Winter assume that the resources invested in searching for new routines depend on a firm’s profitability. This tends to lead to the gradually increasing domination of the economy by large firms as they succeed in attaining higher profits.

Nelson and Winter’s approach explains both variation and selection and how knowledge is preserved and transferred across periods. Their theory led to extensive research that modified and further developed variants of their original model.²⁸ Of particular interest is Winter’s own (1984) extension of the model to include entrepreneurs and start-ups. Two dominant innovation activities are postulated—one entrepreneurial and one traditional. The former, which is assumed to be more dependent on external knowledge, is dominated by entrepreneurs and newly established firms, while the latter is assumed to be associated with the in-house R&D of existing larger firms.

Other models were to a greater extent based on the dominant general equilibrium paradigm. Jovanovic (1982) presents a model for industrial development based on learning. The model assumes an infinite number of small firms that take price as a given. They have perfect information about the equilibrium structure but are ignorant of their own performance (productivity); however, they learn after market entry. These small firms are at greater risk of failure and are also assumed to have poorer growth opportunities.²⁹

Acs et al. (2004) and Braunerhjelm et al. (2010) take one of Romer’s models as their point of departure and show how entrepreneurs who are not involved in research also contribute to innovation and growth. In the model, the business community consists of incumbents that invest in research, as well as entrepreneurs who do not contribute to it. The ability of entrepreneurs to innovate is based on their capacity (unevenly distributed) to draw on previous research investment and to use that knowledge to launch new goods and services. In this way, the entrepreneur becomes an instrument for disseminating knowledge; he or she contributes a mechanism for knowledge to be commercialized. In a model that includes entrepreneurs, the opportunities for sustainably higher growth increase.

In a slightly different model, Acs et al. (2005, 2009) show how entrepreneurship can be endogenized on the basis of knowledge investment and institutional conditions (regulations, well-functioning financial markets, etc.). Given an environment that promotes entrepreneurship, knowledge investment will result in individuals with different entrepreneurial abilities choosing entrepreneurship over employment. There is thus a complementarity between existing and new firms that leads to the testing and exploitation in the market of a larger proportion of an economy’s knowledge base.³⁰

Several of these models seek to incorporate more of the evolutionary elements into a general equilibrium structure, which has occasionally led to drastic assumptions regarding prices, information, transaction costs, distribution of profits, exogeneities, and more. Other models are so complex that it becomes difficult or impossible to calculate a solution, so that simulation methods must be used instead (this includes Nelson and Winter and their successors).³¹ In many cases, Schumpeterian creative destruction is not satisfactorily modeled. The more aggregated the data, the more difficult it becomes to distinguish which components are driving the processes.

Nevertheless, the contribution of these newer models is significant. First, they show how variation and selection under competition characterize market economies and are crucial for business sector development. Second, these phenomena take place in dynamically adaptive systems where learning and feedback take place continuously. One conclusion is that change tends to materialize slowly and is dependent on several factors that affect both knowledge building and knowledge dissemination and commercialization. These in turn are affected by institutions and norms. Finally, the evolutionary system is adaptive, complex, and partly self-organizing, while a state of (traditional) equilibrium is usually an exception rather than the rule.

2.3 Small Businesses and Entrepreneurs: The Empirical Picture

An increasing number of studies point to the importance of new and small firms for the development and commercialization of knowledge, even though they invest relatively modest sums in R&D. Instead, they contribute through their efforts to apply knowledge. However, according to Cooper’s (1964) analysis based on case studies, when small firms undertake R&D, they manage it more efficiently than large firms. He suggests two major explanations: (i) smaller firms have an advantage in exploiting employees’ abilities and (ii) there are different attitudes as well as more direct communication among R&D personnel in small firms. Acs and Audretsch (1987, 1990), drawing on more extensive data, concluded that even though larger firms accounted for the bulk of R&D investment, smaller ones were significantly more innovative in certain industries, such as computers and machine tools, while the reverse was true in the automotive industry. For manufacturing as a whole, the rate of innovation was significantly higher in smaller firms. Similar findings were presented by Baldwin and Johnson (1999) for electronics, instrumentation, medical equipment, steel, and biotechnology. Other studies show that smaller firms are more skilled at producing radically new products. Based on both a theoretical model and an empirical analysis, Michelacci (2003) shows that relatively weak commercialization of research can be explained by too few entrepreneurs.³²

Regarding the importance of new and smaller firms for economic development, a number of empirical analyses have also found a positive relationship between small businesses and growth when other factors such as investment, employment, R&D, and internationalization are taken into account. Already in the early 1990s, Levine and Renelt (1992) argued that there was a strong positive relationship between the share of small businesses in an economy and economic growth.³³

Haltiwanger et al. (2013) show that it is young firms, rather than small ones, that create a disproportionately large number of jobs. Controlling for age, they find that there is no longer any relationship between size and job creation. This has major implications for policy: job creation should be supported by targeting young firms rather than small firms. If the latter are old, they should not be expected to create many jobs. Young firms with superior capabilities and routines move “up” in terms of size and performance, while young ones in which inferior capabilities are discovered are more likely to decline and exit from the market (Huber et al. 2017). Indeed, while the majority of new firms will fold in their first five years, the remaining survivors will show considerable growth.

Later studies on both Sweden and the United States confirm these findings; here, the net contribution of jobs can be disproportionately attributed to young and small firms. On the other hand, productivity seems to have been dominated by larger firms. These differences are explained by the division of labor where low-educated workers can increasingly be found in small and medium-sized enterprises (SMEs) whereas the share of highly educated workers increases with the size of the company. Furthermore, productivity growth is determined in large part by the composition of an industry; aggregate productivity growth will be higher if the share of industries where fast-growing and technology-intensive firms are particularly important. As a corollary, this finding indicates that labor mobility across firms of different sizes is important for the diffusion of knowledge.³⁴ At the same time, the probability of survival is lowest in these firms, especially in technology-intensive industries (Audretsch 1995; Parastuty 2018; Braunerhjelm and Halldin 2022). An influx of new firms and the testing of new ideas—but also their exclusion—is nevertheless a critically important component of dynamic economies.³⁵

To establish a causal relationship between the entry of new firms and aggregate growth implies further difficulties in tracing how different variables interact and whether any actual impact on growth can be identified. However, all growth models emphasize the role of innovation for growth and given that small, young, and new firms contribute a disproportionate share of innovation, entrepreneurial ventures should have either a direct or indirect effect on growth. Several studies also report a correlation between entrepreneurship and growth, which seems to have been reinforced over time (Thurik 1999; Acs et al. 2004; Salgado-Banda 2005; Block et al. 2009; Braunerhjelm et al. 2010; Galindo and Méndez 2014; Urbano et al. 2019). Klapper et al. (2010) note that entrepreneurship is a necessary condition for a dynamic market and that it leads to both tougher competition and greater growth.

Based on an endogenous growth model that includes entry and exit of firms, Akcigit and Kerr (2018) demonstrate the impact that different types of innovations (explorative versus exploitative) have on economic growth. The classification of innovation strategies on explorative and exploitative paths was suggested by March (1991). The former refers to innovation having a more general scope of search whereas exploitative innovation implies prioritizing search depth, i.e., improvement of current products, services, and processes. According to Tushman and Smith (2002), there is a link between these and previous concepts where exploitative innovations can, to a greater extent, be associated with process innovation. In Akcigit and Kerr’s model firms invest in explorative R&D to acquire new product lines and in exploitative R&D to improve their existing product lines. They show that explorative R&D does not correspond as strongly with firm size as exploitative R&D, suggesting that smaller and younger firms are important for the latter type of innovations. They also find some empirical evidence that new firm entry together with SMEs has relatively higher growth spillover effects.

At the regional level—where the analysis is facilitated by the fact that the formal institutions are the same—there are a large number of studies concluding that entrepreneurship and knowledge levels both contribute strongly to higher growth and prosperity.³⁶ Several studies of U.S. states show that entrepreneurship (approximated by inflows and outflows to the market) has a positive effect on productivity and employment. In Europe, similar results have been found, for example, in Spain and Germany (Audretsch and Keilbach 2005; Habersetzer et al. 2021). The results have been interpreted as evidence that international convergence towards more entrepreneurship-led growth is underway, despite differences in institutions and regulations. However, there are still significant variations between countries.

Given the role attributed to new, young, and small firms in the above-mentioned studies, the recent decline reported in entrepreneurship in several countries is reason for concern (Hathaway and Litan 2014; Decker et al. 2017; Naude 2019; Salgado 2020). This seems to be a trend in most developed countries, Sweden being an exception. According to Heyman et al. (2019), this is due to the reforms introduced in Sweden since the 1980s, in particular in the aftermath of the severe crisis in the early 1990s. This is consistent with new and young firms having played a prominent role in job creation during the period, spurred by labor market reforms such as permitting staffing agencies, the allowance of temporary contracts, informally coordinated wage negotiations, and the Industry Agreement adopted in 1997 (see Chap. 6).

Simultaneously, product markets were deregulated (transportation, education, healthcare, etc.) and a new competition act in 1993 replaced the one from 1925, while the lifting of all foreign exchange controls exposed Swedish firm owners to international competition. As a result, Sweden experienced the highest labor productivity growth rate among the OECD countries between 1995 and 2011 paired with improved allocative efficiency and firm-level productivity as the forces of creative destruction were unleashed. This was a huge step away from the interventionist policies pursued in the 1970s and 1980s, an era characterized by lagging productivity and decreased efficiency.

2.4 In Sum

New theoretical and empirical research within the evolutionary approach that we consider to be the most promising shows how a number of institutional factors interact to promote knowledge dissemination and entrepreneurship, thereby paving the way for important innovations. Within this approach, the challenge is much more complex compared to the traditional knowledge-driven growth model, where policy implications tend to be limited to subsidies or tax breaks for R&D, and measures to expand and prolong formal education. The difference in policy conclusions compared with those derived from endogenous growth models is straightforward: If knowledge and entrepreneurial abilities are decentralized and spread across a large number of actors, policy should be more general and ensure that the measures implemented cover everyone in a non-discriminatory manner and embrace several policy areas. The government should thus refrain from engaging in active “industrial” policies by identifying certain sectors or technological niches.

In the next chapter, we will take a closer, more concrete look at these factors and analyze how their design affects innovation activities.

Appendix: Measuring Entrepreneurship

Economists have often chosen to measure the degree of entrepreneurship in a country or region as the share of total employment constituted by the self-employed.³⁷ This is a crude simplification because the definition of “entrepreneur” as a self-employed worker fails to capture all facets of entrepreneurship. In addition, a variety of reasons may lie behind the choice to start one’s own business. This is shown in Table 2.2, which presents various motives for entrepreneurship. The term “self-employed” includes both those who are sole proprietors and those who run their business as a limited liability firm.

Table 2.2 Entrepreneurial and non-entrepreneurial motives for self-employment

Motivations for becoming a self-employed person can include the prospect of a better life or realizing the full entrepreneurial potential of an idea. Often, however, self-employment is a way of circumventing restrictive regulations and obstacles within the company where a person works. Alternatively, the labor market may be too regulated for employees. A third—less agreeable—reason is that self-employment can increase opportunities for rent seeking and improper tax evasion.

The prevailing conceptual confusion regarding entrepreneurship, where completely different kinds of entrepreneurs are lumped together, sometimes gives inconsistent research results and policy recommendations. According to Schumpeter’s (1934) definition—the one to which we adhere—most new businesspeople and self-employed are not entrepreneurs, as they lack the ambition to be innovative or growth-oriented.³⁸ Most self-employed people in Sweden do not have and will never have a single external employee. It is thus important to distinguish between self-employed people as a group and a small number of—often fast-growing—firms where entrepreneurship is more prominent.³⁹ The latter group accounts for a disproportionately high share of restructuring and job creation in the economy.⁴⁰ The category of firms with ambition and potential for innovation and rapid growth is significantly smaller than the broad group of entrepreneurs and differs markedly from the majority.

It is sometimes claimed that these are all the same kind of firm, and that it is only chance that determines who manages to grow. Of course, there is a large gray area here, for example firms that originally had no ambition to grow but which changed their targets when their growth unexpectedly took off. However, we believe that there are also ex ante fundamental differences between different types of entrepreneurs. The overwhelming majority of self-employed people in Sweden state that they do not define themselves as entrepreneurs based on ambitions to grow, a large number of employees, or innovative ideas. There are other important differences between these separate business owner categories. While self-employed people are similar to wage earners in terms of education, entrepreneurs in the United States who have received venture capital funding are 20 times more likely than the average American to have a doctorate (Bengtsson and Hsu 2010).

We dare say that there is a strong international trend among researchers and experts towards using the terms entrepreneur, entrepreneurial activity, and entrepreneurship for people, actions, and phenomena where the intention is to create valuable benefits through (growth-oriented) economic activity based on the identification and/or creation of new products, processes, and markets. Due to its complexity, this concept is impossible to capture in a single indicator. Those who wish to delve into the problems of measurement are referred to OECD (2008), which offers 18 different indicators of the degree of entrepreneurial activity.