Boyan Jovanovic has made pioneering research that advances our understanding of the competitive dynamics between incumbent firms and new entrants, entrepreneurial learning and selection processes, and the importance of entrepreneurship for the economy. His contributions to entrepreneurship research can be categorized into three areas:
(i) why some people become entrepreneurs, (ii) the competitive dynamics between incumbent firms and new entrants, and (iii) the importance of entrepreneurship for the economy.
Why do some people become entrepreneurs?—risk bearers and occupational choices
Why do some people become entrepreneurs? Boyan Jovanovic has contributed to our understanding by analyzing the labor market sorting of individuals with heterogeneous human capital. The basic premise is that the interplay of supply (individual choice) and demand (e.g., employers’ search for skilled personnel) in the labor market matches and sorts individuals on the basis of their human capital. Match quality is unknown ex ante to employment, i.e., match quality is an “experience good” in the sense that it needs to be experienced to be evaluated. In the short run, mismatching is possible which may lead to employee turnover (Jovanovic 1979a, b, 1984, 1994). At the same time, individuals learn and adapt to technologies (Jovanovic and Nyarko 1995, 1996; Jovanovic and Moffitt 1990), and the processes of matching, turnover, and learning lead to human capital formation. Within this context, some individuals become entrepreneurs. In one of his most cited papers, Boyan Jovanovic investigates the role played by liquidity constraints in the choice of becoming an entrepreneur.
Indeed, in Evans and Jovanovic (1989), the authors investigate whether liquidity constraints are binding and cause the number of workers who opt for self-employment to be sub-optimal. Until then, the occupational choice literature (i.e., Johnson 1978; Miller 1984) had implied that because of the risk attached to entrepreneurship, younger individuals would be more likely to enter entrepreneurship than older individuals. This, however, was inconsistent with findings by Evans and Leighton (1989) who found no risk-age relationship. In the attempt to reconcile this inconsistency, Jovanovic and his coauthor hypothesized that liquidity constraints could be a significant barrier for people trying to start a business. That being the case, the inconsistency in previous results would be explained because entrepreneurship would not be a good option for younger people who would not have had enough time to build up capital and would face difficulties in borrowing funds.
To make their case, Evans and Jovanovic develop a model of entrepreneurial choice where the tightness of the liquidity constraint is a key parameter. They then test the model with data and show that there exists a positive correlation between the probability of starting a business and assets, but only if the individuals are liquidity constrained. This means that a wealthier individual can start a business with a more efficient level of capital, thereby obtaining higher returns than a poorer individual. A direct implication of their results is also that the correlation between entrepreneurial earnings and initial assets is positive, since wealthier individuals will have started businesses with the appropriate (more efficient) levels of capital. Importantly, only individuals with high ability and low asset are affected by the wealth constraint. Unfortunately, however, these are the individuals more likely to want to switch to entrepreneurship since, given their high ability, they can earn more in self-employment than in paid employment. The more general implication of Evans and Jovanovic (1989) is that liquidity constraints are indeed binding and reduce the amount of capital flowing to entrepreneurship in two ways. First, they prevent some individuals from trying entrepreneurship to begin with. Second, being constrained, those individuals who do try entrepreneurship use less than the optimal amount of capital which, in turn, leads to less efficient businesses.
Furthermore, because liquidity constraints are shown to be binding, Evans and Jovanovic view their results in the light of the long-standing debate between Knight’s view of a risk-bearing entrepreneur and Schumpeter’s (1934) view that capital markets allow for the separation of the entrepreneurial and capitalist functions. In fact, they interpret their results as providing support for Knight’s (1921) argument that bearing risk is one of the essential characteristics of entrepreneurs who are forced by the market to internalize the costs of moral hazard and adverse selection problems.
Jovanovic’s work on occupational choice also includes the analysis of the heterogeneity of human capital and labor skills. For example, in a 1994 article, he analyzes how individuals allocate their talent between managerial and waged labor, and where alternative conditions in the labor market may cause a suboptimal allocation of talent to emerge. He shows that the best potential managers will end up as wage workers because of their inability to extract appropriate rents for their efforts. This is an important contribution because most of the analytical frameworks up to that point treated workers as interchangeable units of labor. After the idea of talent heterogeneity was introduced, economists began discussing the alternative role played by different types of skills and experiences which eventually lead to Lazear’s “Jack’s of all trade” argument.
The competitive dynamics between incumbent firms and new entrants: entry, exit, and industry dynamics
Jovanovic has also contributed significantly to our understanding of how competitive dynamics between firms in an industry as well as how technological knowledge and its diffusion emerge endogenously. His most cited paper, published in Econometrica 1982, addresses the issue of industry dynamics (Jovanovic 1982a). The paper is a theoretical contribution based on a mathematical model of firm entry and exit in which he provides a theory of selection with incomplete information in which efficient firms would grow and survive whereas inefficient firms would decline and fail.
The crucial feature that allows Jovanovic’s model to characterize empirical observation more accurately than previous models is that costs are randomized and, therefore, differ across firms. In his model, firms do not know exactly what their true costs are until they are operating in the market. Once in the market, firms are able to update their beliefs as new information becomes available. If a firm revealed costs are low, it is likely that that firm will survive. If they are high, the firm will exit. At the beginning, firms operate in a competitive environment and, as a result, prices are known. Thus, entry, production (i.e., size), and exit decisions are made on the basis of efficiency (i.e., revealed costs). In the model, firms differ in size because some are more efficient than others and the varying distribution of efficiency gives rise to entry, growth, and exit.
Since the efficient survives while the inefficient fails, the average efficiency of the survivors improves over time. A further important implication of the model is that firm size and concentration are positively related to rates of returns, and that a higher concentration is associated with higher profits for larger firms, but not for smaller firms. This is the case because concentration is an indicator of high efficiency variance. Thus, larger firms that survive earn higher profits and smaller firms that survive have higher and more variable growth rates but are more likely to fail. In general, firms that fail are exactly those that would have grown more slowly. These results are consistent with the empirical observation that, within industries, smaller firms tend to grow faster than large firms but are also more likely to fail.
The 1982 paper lacks innovation in the system, which implies that an important function of smaller and newer firms is neglected (Acs and Audretsch 1988). In Jovanovic and MacDonald (1994a), the authors build on the industry life-cycle work by Gort and Klepper (1982) and Klepper and Graddy (1990). They posit that dramatic industry shakeouts that tend to occur in the life cycle of many industries emerge from the introduction of a radical innovation that firms try to implement in the attempt to remain competitive. Those that succeed survive and grow, those that do not, exit. Technological improvements reduce production costs but, in doing so, cause a shakeout and contribute to increasing the optimal size of the firm. Over time, the number of firms in the industry decreases. The technology-based explanation for shakeouts is compelling because it shows that the early adoption of innovations may not only offer great rewards but also create dramatic discontinuities. Such a result is also clearly important in pointing out the important role played by entrepreneurial firms throughout the industrial cycle.
The important role played by entrepreneurial firms in industry dynamics emphasized in Jovanovic and MacDonald (1994a) is further investigated in Jovanovic and Rousseau (2014) and in Jovanovic (2001). In the latter, in particular, Jovanovic presents some descriptive historical evidence and notes that the average age of companies listed on the stock market (including very large ones) has been declining and that, as for any population, the birth rate of new capital needs to increase if we are to maintain the existing levels of capital stock. On p. 55, he writes “It is clear that we are entering the era of the young firm. … The small firm will thus resume a role that, in its importance, is greater than it has been at any time in the last seventy years or so.”
The importance of entrepreneurship for the economy
Boyan Jovanovic’s research has also significantly improved our understanding of technological change and the entrepreneurial function in the wider economy. Specifically, Jovanovic has analyzed the impact of technology diffusion and human capital allocation across different types of employment, and on how alternative joint distributions of skills and technology explain income and growth differences across countries. A main perspective in this work is that one important role of smaller firms and entrepreneurs rests with their role as vehicles for the allocation of human capital and technological change. Those, in turn, produce the industrial churning which determines the rate of economic growth (Greenwood and Jovanovic 1999; Hobijn and Jovanovic 2001; Benhabib and Jovanovic 1991; Jovanovic and Lach 1997; Jovanovic 2009; Eeckhout and Jovanovic 2012).
In Jovanovic (2009), he argues that the distribution of skills is heterogeneous across agents and that agents with low skill levels prefer to use old technologies because they are less costly for them. Specifically, Jovanovic develops a model which accounts for a variety of technologies of different vintage. Learning by doing (primarily in the research sector) produces new technologies. The complementarity between skills and technological sophistication matches agents with low skills to old technologies, and agents with high skills to new technologies. As time goes by, technologies become old and less valuable until they are abandoned. In this set-up, agents with low skills have no incentives to learn new skills. The model endogenizes both technology and skills and shows how their joint distribution explains the cross-section relation between a country’s income per capita and the average age of the technologies used by its workers.
In a related paper, Eeckhout and Jovanovic (2012) build and expand upon Jovanovic’s earlier research and show that global welfare gains depend positively on the skill heterogeneity of the labor force and on the opportunity for individuals to switch employment and sort themselves into the type of employment that best fits their skills’ level. Jovanovic and his coauthor also emphasize the important distinction between managerial and wage work functions broadly defined.
The paper shows how gains from trade relate to the global dispersion of skills, the global diffusion of technological knowledge, and the distribution of workers across occupations. The basic intuition behind the argument rests, in the authors’ terminology, on a “span-of-control production technology,” where the distribution of skills between managers and workers determines the firm’s productivity. High-skill managers are more productive if they command a given set of workers than low-skill managers. Since managerial skills are heterogeneous, but workers’ skills are homogenous, the compensation schedule for managers is non-linear in manager skill and linear in worker skill. This compensation structure leads to sorting of higher-skilled individuals into managerial occupations. Because higher-skilled managers generate higher output with the same set of workers, a high-skill economy has a comparative advantage in managerial occupations. With increased openness and economic integration, and consistently with empirical observations, this leads to a disproportionately high-occupational choice of managerial jobs in high-skilled economies.
This stream of work illustrates that in Boyan Jovanovic’s research, the allocation of human talent is at the core of all economic phenomena. Nothing happens without the application of human capital to technological improvements and the latter do not propagate without entrepreneurial firms.