Erasmus University
FEM 11061
—————————————————— Lecture Notes: Economics of Entrepreneurship ——————————————————
Christian Roessler
Fall 2009
Contents
1 Entrepreneurship and Change 1.1 The Nature of Entrepreneurs and Opportunities . . . 1.1.1 Schools of Thought . . . . . . . . . . . . . . . 1.1.2 Modeling Entrepreneurs . . . . . . . . . . . . 1.1.3 Identifying Entrepreneurs . . . . . . . . . . . 1.2 E ciency and Growth . . . . . . . . . . . . . . . . . 1.2.1 Arbitrage . . . . . . . . . . . . . . . . . . . . 1.2.2 Innovation . . . . . . . . . . . . . . . . . . . . 1.2.3 The Direction of Entrepreneurial Activity . . 1.2.4 Business Cycles . . . . . . . . . . . . . . . . . 1.3 Incomes . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.1 Entrepreneurship and the Income Distribution 1.3.2 Mobility . . . . . . . . . . . . . . . . . . . . . 2 Motives 2.1 Risk Attitude and Perception . . . . . . . . 2.1.1 Risk Bearing . . . . . . . . . . . . . 2.1.2 Entrepreneurship and Risk Aversion 2.1.3 Cognition . . . . . . . . . . . . . . . 2.2 The Private Equity Premium Puzzle . . . . 2.2.1 Low Return to Entrepreneurship . . 2.2.2 Explanations . . . . . . . . . . . . . 2.3 Status Preference . . . . . . . . . . . . . . . 2.3.1 "Own Boss" Bias . . . . . . . . . . . 2.3.2 Relative Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 3 6 7 9 9 12 14 15 17 17 18 18 19 19 19 22 23 23 24 26 26 27 28 28 28 29 30 31 33 33 34 35
3 Means 3.1 Talent and Quali…cations . . . . . . . . . . . . . . . . . . . . . 3.1.1 Return to Talent . . . . . . . . . . . . . . . . . . . . . 3.1.2 Adverse Selection into Employment or Entrepreneurship 3.1.3 Evolution of Firm Size . . . . . . . . . . . . . . . . . . 3.1.4 Education . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Social Barriers . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Ethnicity, Immigration and Self-Employment . . . . . . 3.2.2 Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 The Liquidity Constraints Debate . . . . . . . . . . . . . . . . 1
3.3.1 3.3.2
Evidence of Constraints . . . . . . . . . . . . . . . . . 35 The Course of Development with Capital Constraints . 37 . . . . . . . . . . . . . . . Finance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 38 38 41 41 41 43 44 45 45 46 49 50 50 50 52 53 55 55 56 56 58 58 59 60 61 61 61 61 63 64 66 66
4 Venture Finance 4.1 Self-Funding . . . . . . . . . . . . . 4.1.1 Entrepreneurial Saving . . . 4.1.2 Family Contributions . . . . 4.2 Information Issues and Relationship 4.2.1 Moral Hazard . . . . . . . . 4.2.2 Venture Capital . . . . . . . 4.2.3 Strategic Investors . . . . . 4.3 Security Choice . . . . . . . . . . . 4.3.1 Debt vs. Equity . . . . . . . 4.3.2 Mixed Securities . . . . . . 4.3.3 Staged Investment . . . . .
5 Entrepreneurial Firms 5.1 Leadership and Strategy . . . . . . . 5.1.1 Entrepreneurial Labor Supply 5.1.2 Leadership and Delegation . . 5.1.3 Innovation Strategy . . . . . . 5.2 Organization . . . . . . . . . . . . . 5.2.1 Legal Form . . . . . . . . . . 5.2.2 Networks . . . . . . . . . . . 5.2.3 Spin-O¤s . . . . . . . . . . . . 5.3 Survival Factors . . . . . . . . . . . . 5.3.1 Objectives . . . . . . . . . . . 5.3.2 Resources . . . . . . . . . . . 5.3.3 Technology . . . . . . . . . . 6 Public Policy 6.1 Institutions . . . . . . . . . 6.1.1 Employee Bene…ts . 6.1.2 Redistribution . . . . 6.1.3 Legal Infrastructure . 6.1.4 Financial System . . 6.2 Regulation . . . . . . . . . . 6.2.1 Market Imperfections
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . and Entry
2
6.2.2 6.3 Fiscal 6.3.1 6.3.2 6.3.3 6.3.4
Bankruptcy . . . Policy . . . . . . . Public Spending . Taxes . . . . . . Tari¤s . . . . . . Credit Subsidies .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
68 69 69 70 72 73
1
Entrepreneurship and Change
The introductory lecture is concerned with how entrepreneurs a¤ect market e ciency through arbitrage, long-run growth through innovation, and the income distribution through social mobility.
1.1
1.1.1
The Nature of Entrepreneurs and Opportunities
Schools of Thought Three fundamental schools of thought have dominated the scienti…c debate of entrepreneurship: the classical, Schumpeter and Austrian views. In the classical approach, entrepreneurs make calculated investments and are distinguished by their willingness to risk losses. Irishman Richard Cantillon, in his Essay on the Nature of Commerce (ca. 1730, in English 1755), was the …rst to write about the risk-taking role of business owners and the rewards they stand to gain from volatile markets in exchange for guaranteeing wages for workers. Cantillon viewed as entrepreneurs all economic actors who bear uncertainty by buying and selling in the market at …xed cost, in expectation of an unforeseeable revenue. These entrepreneurs expose themselves to demand uctuations and guarantee a stable income to resource owners: landlords, who collect …xed rents, and hirelings, who earn …xed wages. American economist Frank Knight (1885-1972), in his PhD dissertation Risk, Uncertainty and Pro…t (1921), later argued in the same spirit that investment decisions are made on the basis of risk-adjusted returns, where relatively risky entrepreneurial activity is favored by less risk-averse individuals. Knight distinguished various types of risk: at 3
one extreme, random outcomes have known probabilities, at the other, probabilities are ambiguous ("Knightian uncertainty"). To the extent that entrepreneurs operate in uncertain environments, they earn a return on their willingness to accept uncertainty, rather than conventional (probabilistic) risk. In Theory of Economic Development (1911) and subsequent works, Moravian-born Joseph Schumpeter (1883-1950) described the entrepreneur as someone who pioneers radical new technologies. Schumpeter' school of thought emphasizes the adventurous nature of the s entrepreneur - seeking out new technologies and unpredictable market environments. Entrepreneurial behavior upsets the status quo, it obliterates existing processes when their pro…t potential is exhausted, and replaces them with new technological regimes that create new opportunities for growth.This phenomenon Schumpeter famously encapsulated in the term "creative destruction." The Austrian tradition revolves around the view that opportunities are hidden, and it takes alert entrepreneurs to identify and exploit such opportunities. From a social standpoint, the entrepreneur enables the proper functioning of markets by providing arbitrage in exchange for rents. Here, the entrepreneur moves the market from a condition of disequilibrium toward one of equilibrium. This role of the entrepreneur was clearly formulated by English-born Israel Kirzner (1930-) in the 1960s and 1970s, but he was in uenced by the original Austrian school of Ludwig van Mises (1881-1973) and Nobel laureate Friedrich Hayek (1899-1992), who emphasized the role of imperfect information (about demands for goods and resources). Nobel laureate Theodore Schultz (1902-1998) also discussed the entrepreneur' function in reallocating s resources in response to disequilibria. Makowski and Ostroy [114] attempt a synthesis between neoclassical and Austrian perspectives by reinterpreting the neoclassical notion of perfect competition as the result of full appropriation of rents within the constraints of agents'bargaining powers in the market. From this perspective, entrepreneurs might discover large rents, but competition among entrants - in the absence of barriers or demand idiosyncracies ultimately diminishes or eradicates these rents. This view di¤ers from
4
ordinary perfect competition primarily in that it does not assume pricetaking behavior, but it often leads to the same outcome through the Austrian mechanism of entrepreneurial exploitation of opportunities. It di¤ers from Austrianism because it does not preclude the attainment of a perfectly competitive equilibrium (Austrians only admit a tendency toward the equilibrium, but argue that the economy remains in disequilibrium, with positive rents to be earned). Another approach to reconciling the di¤erent schools of thought is to see them as serving di¤erent goals - highlighting the various roles of entrepreneurship (all of which are present in reality), and describing the entrepreneur at various levels of detail. Endres and Woods [50] distinguish and compare the neoclassical, Austrian and behavioral schools of thought. They discuss central assumptions of each approach. Neoclassical theory is built around the notion that the economy is generally in equilibrium, where entrepreneurs are fully rational, maximizing agents who choose to operate …rms. While entrepreneurs have access to the same technologies and information as everyone else, they are distinguished by their preferences, such as greater tolerance for risk, and/or endowments (of wealth, skill, etc.). In the Austrian view, entrepreneurs have no special "group" characteristics, but exploit opportunities that present themselves to individuals who happen to be in the right place at the right time. These opportunities exist because the economy is typically in disequilibrium - information is far from perfect and public, leaving room for arbitrage when e.g. an individuals becomes aware that a good sells at di¤erent prices in di¤erent places. The Austrian entrepreneur is still rational and a maximizer, but he operates in an unpredictable environment that leads to opportunities he is uniquely positioned to exploit. In short, in neoclassical theory, individuals face the same opportunities, but di¤er in their ability or inclination to pro…t by them, whereas in Austrian theory, individuals face di¤erent opportunities, but are equally equipped to pro…t by them (except insofar as opportunity recognition is in itself a personal skill). Behavioral approaches relax the rationality assumption. They posit that the risk associated with opportunities cannot be made precise, that an entrepreneur cannot conceive of every possible development owing from an action, and moreover is unable to calcu5
late the optimal course, even if the problem is well-de…ned. Instead, entrepreneurs use heuristics - simple, routinized decision rules - that lead to imperfect decisions and systematic errors. Limited personal experience is likely to in uence judgements more heavily than historical facts, which the entrepreneur is not fully informed about and cannot process fully. Endres and Woods argue that the di¤erent approaches have di¤erent objects and should be evaluated with respect to these. Neoclassical theory is concerned with a "representative," rather than speci…cally existing, entrepreneur who, on average, makes unbiased, optimal decisions and, in the totality of his embodiments, has access to all information. The focus is on the behavior of the adjusted market, not the individual entrepreneur. The Austrian approach is concerned with the unadjusted market, while the behavioral approach is much more descriptive and "realistic" because its purpose is to understand the individual entrepreneur. 1.1.2 Modeling Entrepreneurs In one sense, economics seems ill-suited as a framework for studying entrepreneurs. The economic agent is a rational optimizer, aware of all available actions and their possible consequences. Nothing truly surprises or ba- es him, life is a well-de…ned maximization problem. I am not aware of any economist who believes in the existence of the economic agent. But much can be learned about real people, who try to the best of their abilities to achieve good outcomes for themselves, from studying this infallible individual. Real people are not likely to make systematic mistakes for long; they learn and correct because it is in their self-interest to emulate the economic agent. In the end, economic models of economic agents do tell us something about tendencies in the economic reality. They tell us what happens after some adjustments. What makes entrepreneurs di cult to study is that they operate during, not after, adjustment. They solve problems before they become well-de…ned. They recognize opportunities before they become apparent to everyone. Their decision-making process has little in common with that of the economic agent. Much of it is ad hoc, based on guess work and intuition, and is designed to be satisfactory rather than optimal. 6
The individual stories of entrepreneurs are unique, and economics might o¤er little insight on how they do what they do. But economics is quite well-suited to explain the choice to be an entrepreneur, which circumstances help and hinder, and how it impacts the economy. I.e. we can understand investment in entrepreneurship without knowing the speci…c content - just as we can study car purchases and their consequences without reference to car mechanics. Baumol [12] argues that economic models do a good job of describing the role of the "manager," who maintains the …rm' operations at a s static optimum. They do not capture the role of the entrepreneur, who is constantly looking for novel ways of doing things and reacts non-mechanically to economic problems. Nor can they, writes Baumol, because models are inherently about optimizing with respect to wellde…ned problems, and this is not what the entrepreneur faces and does. He advocates trying to understand the rewards of entrepreneurship, rather than the why and how, and direct policy toward mitigating risks and ensuring entrepreneurial pro…t. Shane and Venkatamaran [152] outline an entrepreneurship research program in the Austrian tradition, focused on why entrepreneurial opportunities exist, how they are discovered, and who exploits them. They see opportunities as arising from the subjective value of resources, due to (imperfect) individual judgement, technological innovation and the limited life span of existing uses (e.g. because suppliers raise prices over time). Discovery of such opportunities depends on private information as well as the ability to perceive its pro…t potential. The decision to exploit opportunities is in uenced by individual circumstances, such as opportunity costs (alternative income) and access to resources, and occurs either through existing …rms (as employees or via sale) or by creating new organizations. The authors suggest that entrepreneurship research should proceed along these lines. 1.1.3 Identifying Entrepreneurs Whom then do we mean by "entrepreneur" An often-cited heuristic de…nition emphasizes the decision-making function of the entrepreneur. Hebert and Link [79] suggest the following de…nition of entrepreneur7
ship based on a historical discussion of major schools of thought: "The entrepreneur is someone who specializes in taking responsibility for and making judgemental decisions that a¤ect the location, form, and the use of goods, resources, or institutions." The most detailed and pragmatic de…nitions arise from data collection e¤orts, where every individual has to be categorized. The leading entrepreneurship database is the Global Entrepreneurship Monitor (GEM), described in Reynolds et al. [147]. GEM currently covers over 40 countries: in each, at least 2,000 individuals are surveyed annually about their entrepreneurial attitudes and activities. GEM de…nes entrepreneurs as "adults in the process of setting up a business they will (partly) own and / or currently owning and managing an operating young business." This is intended as a catch-all criterion that incorporates narrower de…nitions in use. It includes independent businesses (younger than 42 months) as well as ventures that are sponsored by an established …rm (branches, subsidiaries), where the respondent has an ownership stake. While in the committed planning stage, individuals are classi…ed as "nascent entrepreneurs." When wages have been paid for three months, "…rm birth" is said to have taken place. For the …rst three-and-a-half years in which the …rm pays wages, it is considered "new," afterwards "established." Most …rms which survive this initial phase are viable and continue for much longer. The entrepreneurially active population (owners of nascent and new …rms) responds either to a perceived opportunity or to necessity (having no suitable alternative source of income). Roughly one-third tend to be self-identi…ed necessity entrepreneurs. Innovativeness is determined from questions about customer familiarity with the product and the availability of similar products in the market. One-third of all ventures are not innovative, a very small percentage (3%) is highly innovative. About half of all entrepreneurs expect to create more than ten jobs in the foreseeable future. GEM also polls about several indicators of entrepreneurial perceptions and reports results of interviews with local experts about conditions for entrepreneurship.
8
1.2
1.2.1
E ciency and Growth
Arbitrage Entrepreneurs play two very positive roles in the economy: arbitrage and innovation. Entrepreneurs "perfect" existing product markets by increasing competition for incumbents and thereby lowering rents. Similarly, entrepreneurship increases competition in the labor market: this eliminates employer rents by bidding up wages and providing an escape route from unemployment or "unfair" remuneration and promotion practices. Another type of entrepreneurial reallocation, besides price and resource demand arbitrage, is organizational arbitrage, where the entrepreneur recognizes that it is cheaper to internalize several production processes. Yu [164] describes the emergence of Hong Kong' small-scale manus facturing sector as a prototype of Kirzner-style entrepreneurship: constantly responding to market opportunities that open up through technological developments, trade policy and structural changes overseas. Many of these establishments stayed deliberately small in order to be able to …ll even small orders at a competitive cost and rapidly evolve their product and service menu. They focused on producing goods to be marketed by the clients, who bore the risk associated with new product introduction. Hence these …rms are responsive rather than innovative and not necessarily risk takers. The paper cites many cases of particular companies that exemplify the highly adaptable brand of Hong Kong entrepreneurship. Littlechild and Owen [109] model the Austrian entrepreneurial arbitrage function, assuming that markets for the same good initially differ in prices, due to dispersed information about markets, leading to market discovery and arbitrage by entrepreneurs who have varying tolerance thresholds for price di¤erentials (resulting from prices and arbitrage opportunities they are already aware of). They show that markets which are "linked" (there exists a chain of individuals such that any two adjacent members are initially aware of the same market) will converge to a uniform price, thus providing an underpinning of Walrasian equilibrium without the requirement that all markets and prices are public knowledge. (Basically, as prices converge, entrepreneurs' thresholds 9
are reduced and they are willing to arbitrage smaller price di¤erentials away, eradicating even the smallest di¤erences over time.) In general, markets (that are not necessarily linked) converge to a uniform price with probability 1 if entrepreneurs discover new markets with a probability that is proportional to their attractiveness (pro…t potential). A similar logic underlies this result. Hence entrepreneurs are able to smooth out disequilibria through their pro…t-seeking arbitrage behavior. Entrepreneurship counteracts unemployment by o¤ering a pro…table alternative in self-employment. In principle, it also has the potential to serve as an alternative to a traditional employment sector that o¤ers disadvantagous wages and promotion prospects. While ethnic minorities use entrepreneurship to escape from discrimination, little evidence has been found that women and African-Americans do. The reason could be that discrimination against them is no longer prevalent, or that it also a¤ects entrepreneurs, e.g. through unequal access to capital, suppliers and customers. Audretsch and Jin [6] refer to a discrepancy in the relationship between unemployment and …rm start-ups found through di¤erent methodologies. In time series analyses, the relationship tends to be positive: more unemployment pushes some people into entrepreneurship. In cross section analyses, the relationship tends to be negative: industries and countries with fewer entrepreneurs have more unemployment. They explain this phenomenon through shocks to price-cost margins that underlie the hiring and …ring decisions at existing …rms. A reduction in an industry' price-cost margin creates an incentive for laid-o¤ works ers to start a new …rm somewhere. Thus, in aggregate unemployment leads to more entrepreneurship, and this is re ected in time series studies where the unit of analysis is a country. On the other hand, the lower price-cost margin makes the industry less attractive for entrepreneurial activity. Prospective entrepreneurs go elsewhere, and this explains why more unemployment in a particular industry is associated with fewer start-ups in the industry. Hence cross section studies, where the unit of analysis is an industry, show a negative correlation. Clark and Drinkwater [34] observe increasing wage discrimination against British minorities between 1973-1979 and 1983-1989 in the British Gen10
eral Household Surveys (i.e. the earnings gap for given observable characteristics widened). The associated increase in self-employment among minorities documents that they seek out entrepreneurial activities to escape discriminatory wages. Moore [125] rejects, using the 1978 US Current Population Survey, that black/white and female/male earnings ratios are higher in selfemployment than in the wage sector (which could have indicated ight from discriminatory pay). One can rule out explanations based on nonrandom selection e¤ects (low earners among blacks and women going into self-employment) since blacks and women are also not overrepresented among the self-employed. There remains, however, the possibility of capital market and consumer demand discrimination that makes self-employment unattractive for these groups. From 1992 US census data (Characteristics of Business Owners), Bates [11] …nds that self-employment is often a temporary solution for collegeeducated Asian immigrants; they take salaried jobs after assimilation. This pattern is not observed among non-immigrant Asians. Apparently, immigrants initially face inferior employment opportunities and start relatively unpro…table (especially retail) businesses, until outside options improve. Chinese/Korean start-ups are less pro…table than Indian/Filipino start-ups, even with comparable education, which probably re ects the e¤ect of uency in English on opportunity costs (market wages). In his classic article on the theory of the …rm, Coase [36] asks why coordination through an entrepreneur replaces the price mechanism within a …rm. He advocates transaction costs, i.e. buying in the market incurs in practice information and other costs, against which must be weighed the disadvantages of centralized resource allocation: the entrepreneur' limited time and knowledge. For some purposes, and up s to a certain scale, the latter constraints are less costly than going to the market. Firms will carry out tasks internally as long as the technology is su ciently tractable, i.e. the optimal process is well-understood and requires limited decision making at the top. Coase criticizes Knight' s explanation of the …rm as a risk-sharing device, since risk-sharing does not require that the insured works under the insurer' direction. His s entrepreneur is an e cient organizer who is best able to substitute for 11
resource allocation that would otherwise be achieved at a transaction cost through the market. 1.2.2 Innovation Entrepreneurs create new product markets, which they initially monopolize. However, this is still an e ciency improvement, since it directs resources to an activity that consumers value enough to let the …rm realize a superior pro…t. Small entrepreneurial …rms are often the optimal vehicle for specializing in a single, risky activity, which exploits comparative advantage and, in aggregate, propels economic growth. Alternatively, entrepreneurial entry into an existing industry may be associated with cost-reducing process innovations. Either contributes to economic growth. Inventors often have entrepreneurial characteristics. In their study of great American inventors from the antebellum period (…rst half of the 19th century), Khan and Sokolo¤ [101] argue that inventive activity responded to market opportunities and did not seem to be driven primarily by innate genius. Important inventions tended to occur in high-demand periods and locations, rather than randomly. They focused on critical industries in times of war and breakdown in foreign trade. Inventors came from a broad cross-section of professions, not mainly technical ones, and they were often experienced and older when they started to patent. They were also more mobile than the general population, across places as well as industries, seeking out opportunities. Moreover, inventors in almost all cases made an e¤ort to pro…t through manufacturing, licensing, royalties or a combination thereof. Those who left their inventions unpatented were typically manufacturers who bene…ted from productivity increases or reputation gains. Hence, great inventors behaved in an entrepreneurial fashion, concentrating their energies on …nancially rewarding projects, and exhibiting exibility and a willingness to take risks in this pursuit. Audretsch [5] describes how scientists left their careers in genetics in medical research to enter the young biotechnology industry. Facilitating circumstances were a good commercialization infrastructure, minimal regulation and ready availability of …nancing. Biotechnology …rms 12
formed in regional clusters with complementary businesses in order to focus on very speci…c, risky tasks. Radical innovations have to be carried out in entrepreneurial …rms, rather than in subsidiaries of larger …rms. Wiggins [161] argues that activities which are very risky (yield a large pro…t with a small probability) and require unobservable personal e¤ort over a period of time with relatively little speci…c capital are di cult to carry out in large …rms, because of commitment problems. The …rm has an incentive to renege on any contractual promises it may make to the entrepreneur because the bene…ts are large, and reputation loss with potential future partners does not weigh heavily. Courts can often not enforce promises because the returns to the entrepreneurial e¤ort are di cult to verify. Hence, activities like "wild cat" oil exploration (high risk) and medical, legal and consulting practices (prolonged personal e¤ort) are carried out in small, entrepreneurial …rms dedicated to the purpose. Entrepreneurs are also needed as commercializers of inventions. In Michelacci' model [119], entrepreneurs do not invent, but commers cialize inventions that arise from research. Individuals choose to be either researchers or entrepreneurs. Researchers discover inventions and look for entrepreneurs to commercialize them. Entrepreneurs enter into a partnership with a researcher if the rewards are large enough. In this case, a …rm is created and the invention becomes an innovation. The ratio of researchers to entrepreneurs is in equilibrium when no researcher has an incentive to become an entrepreneur and vice versa. Whether this ratio is socially optimal depends on the relative bargaining powers of researchers and entrepreneurs. If these re exactly ect the respective contributions to social surplus, then the allocation is ef…cient, otherwise not. E.g. if entrepreneurs appropriate too little of the …rm' pro…t, then they e¤ectively bestow a positive externality on s researchers, leading to underinvestment in entrepreneurship (and vice versa for researchers who appropriate too little). This is a special case of a standard result from search theory. From a social point of view, there can be too much research or too much entrepreneurship, leading to low-growth equilibria. 13
1.2.3
The Direction of Entrepreneurial Activity Baumol [13] argues that the growth experience at di¤erent times and places can be in part explained by the allocation of entrepreneurial e¤ort to productive vs. unproductive activities, i.e. innovation in production vs. rent-seeking. In ancient Rome, respected avenues to wealth were through landholdings, usury, tax and other income from government positions, while commerce was left to the lower classes and came at the price of low prestige. Innovations were not necessarily encouraged and disseminated slowly (as re ected in the story of the inventor of an unbreakable glass, who was beheaded by the emperor Tiberius upon asking for a reward, because Tiberius was concerned that the invention might devalue gold). Entrepreneurship therefore concentrated on ways to extract more rent from political o ces, i.e. on redistribution rather than creation of wealth. In imperial China, advancement occurred through the examination system, which focused on memorizing and interpreting Confucian teachings and learning calligraphy, while commerce was frowned upon as a lesser pursuit. O cials who had passed the examinations had to complement their meager salaries by taking from their constituents, i.e. to engage in rent-seeking entrepreneurship similar to that of Roman o cials. In the early medieval ages in Europe, wealth was accumulated through land conquests, and much of the entrepreneurial energy went into improving weapons, which were then used to redistribute wealth through warfare. In the later medieval ages, when legal protections against con…scations and such were in place, Cistercian monasteries made innovations in rural commerce, while also seeking monopoly guarantees in some cases, for example for their water mills. Royal grants of legal monopoly became increasingly common, and were an important object of entrepreneurial activity, until the industrial revolution. In England, the custom of primogeniture (eldest son inherits the family' entire wealth) forced younger siblings s of noble families into commerce early on, making it more respectable and perhaps opening the door to the industrial revolution. Major inventions occurred in antique Rome (water mill, steam engine in Alexandria), China (paper, gunpowder), but found little or no industrial application. Today, rent-seeking entrepreneurship is found in litigation, takeovers, e¤orts at tax evasion and avoidance.
14
In [128], Murphy, Shleifer and Vishny relate the propensity of the most talented individuals to become rent-seekers vs. productive entrepreneurs to the elasticities of the production functions in the two sectors: if income from rent-seeking is more responsive to ability, then the most able become rent-seekers. One might conjecture that slower-growing economies at the technological frontier, i.e. developed countries, exemplify this situation and attract their brightest citizens into the legal, …nancial and political professions. The authors make the speculative point that discrimination against ethnic groups like the Jews in Europe, and excluding them from rent-seeking sectors such as the government bureaucracy and the military, has actually helped growth by forcing them to channel their entrepreneurial talents into the productive sector. They …nd that countries with greater fractions of engineering relative to law majors enrolled in college had higher growth rates, suggesting that a prevalence of rent-seeking occupations has a negative e¤ect on growth. Murphy, Shleifer and Vishny [129] consider the choice between entering productive activities and rent-seeking activities that expropriate income from the productive sector. They argue that rent-seeking is self-reinforcing: more rent-seekers make it less attractive for any individual to work productively, since a smaller share of the generated income remains in the productive entrepreneur' hands. This can lead s to bad equilibria with many rent-seekers and few entrepreneurs (vs. good equilibria that sustain many productive entrepreneurs because few rent-seekers live o¤ their income). The poor growth performance of some African economies is blamed on such circumstances, but redistributive taxation and expansive legal sectors in developed countries also …t the description. Weak property protection laws are in general likely to lead to high rent-seeking activity. 1.2.4 Business Cycles Entrepreneurs have an incentive to work harder in good times if profitability is more sensitive to e¤ort then. This means that the e¤ort choices of entrepreneurs over the business cycle tend to amplify it. The argument ignores the fact that the bank system and the stock market provides a kind of risk-sharing to entrepreneurs through the availabil15
ity of funds, which also adjusts to the business cycle (e.g. banks might provide cheaper credit in recessions, stockholders absorb some of the losses). However, such risk sharing must be consistent with e¤ort incentives. Rampini [143] examines the e¤ect of varying risk aversion over the business cycle on the cyclical supply of entrepreneurship. The wealth e¤ect (people are richer during booms) makes risk aversion countercyclical (people are less risk averse during booms). The agency cost e¤ect, or you might call it the "easy-money" e¤ect (a …rm is more likely to succeed at any level of entrepreneurial e¤ort during booms), also makes the risk-bearing required by …nanciers countercyclical (intermediaries require a lower equity stake, hence less personal exposure, to be held by entrepreneurs, to induce high e¤ort during booms). At any point in time, the wealthiest agents are least risk averse and become entrepreneurs. In a boom, individuals are wealthier and therefore willing to engage in entrepreneurship at greater risk. At the same time, they can bear a smaller risk and still exert high e¤ort, due to the high probability of success - high e¤ort is likely to lead to high pro…t, hence is attractive to entrepreneurs even if the share they get to keep is relatively small. Thus, …nancial intermediaries are willing to supply funds on less risky terms that are acceptable to many prospective entrepreneurs. In a recession, when people are less wealthy, they would only become entrepreneurs if the risk is relatively small. Yet, they must bear a greater risk in order to exert high e¤ort, since e¤ort is less likely to pay o¤ - unless the entrepreneur' share in pro…t is large, high e¤ort is s not worthwhile, given its low probability of success. Therefore, funding is available on riskier terms that are acceptable to fewer entrepreneurs. As a result, entrepreneurship increases during booms, when the entire population is willing to take more risks but can get funding at less risk. On the ipside, entrepreneurship decreases during recessions. Since entrepreneurship is a risky activity, its procyclical pattern contributes to business cycle volatility. Because entrepreneurs have incentives to work harder in good times, the supply of entrepreneurship is procyclical in the absence of risk sharing. With the risk sharing that can be achieved by …nancial intermediation, entrepreneurship remains procyclical. In a variation on the theme, entrepreneurs may also be expected to 16
work harder in booms because they are able to secure their loans with more collateral, hence have a greater stake in the survival of the …rm. This allows entrepreneurs to borrow and invest more (and, respectively, causes them to invest less in recessions), leading to a business cycle. Bernanke and Gertler [18] show that entrepreneurial investment can propagate a business cycle because the monitoring (auditing) costs of …nancial intermediaries cycle. In a boom period, when agents are wealthy and can provide more collateral, external funds are cheaper because less monitoring is necessary to induce high e¤ort. This increases entrepreneurial investment. In a recession, agents are poorer and have less collateral, so that lenders must monitor more frequently. The cost of external funds increases and depresses entrepreneurial investment. Hence a temporary increase in personal net wealth (say, because a small oil well is discovered) is ampli…ed by the falling cost of lending and resulting investment boom. After the shock has dissipated, lending costs rise and investment contracts.
1.3
1.3.1
Incomes
Entrepreneurship and the Income Distribution Friedman [60] theorizes that entrepreneurial risk-taking increases the spread of the income distribution. As an extreme example, if all members of the population choose to undertake the same project with a given probability distribution of returns, then this distribution of returns equals the expected income distribution. To the extent that entrepreneurship is riskier than other activities, the supply of entrepreneurship will increase income inequality. Kanbur [93] shows that, when Friedman' theory is placed in a general s equilibrium context, the implications for the income distribution are more subtle. When many agents choose the risky activity, its return diminishes, and therefore the population divides into entrepreneurs and non-entrepreneurs. When preference for risk-taking increases, the entrepreneurial group expands, its mean income decreases, and inequality within this group increases. However, this does not necessarily increase overall inequality, since a greater share of entrepreneurial incomes might resemble the incomes of non-entrepreneurs as a result. 17
Berkowitz and Jackson [17] compare the transition experience of Poland and Russia in the 1990s, and how income inequality was a¤ected by the di¤erent rates of new …rm entry. They …nd that entrepreneurship reduces income inequality and advance as explanations the role of entrepreneurial …rms in job creation (which lifts individuals at the low end of the income distribution out of unemployment), productivity increases (which generally raises wages) and e ciency improvement (which eliminates monopoly pro…ts and therefore lowers incomes at the top of the distribution). A 1% increase in the share of the workforce employed by new and small entreprises increases the income share of the bottom 40% by more than 1% in both Poland and Russia. While Poland' new s …rm share of the workforce more than doubled in the mid-1990s, Russia' share halved in the second half of the decade. Poland' income s s distribution remained stable in this period, whereas Russian income inequality increased signi…cantly. 1.3.2 Mobility Using the Panel Study of Income Dynamics, Holtz-Eakin, Rosen and Weathers [86] trace individual entrepreneurs' mobility between quintiles of the earnings distribution. They …nd that low-income individuals who became self-employed tended to move up, high-income individuals tended to move down. Hence entrepreneurship appears to work toward equality. It also generally increases income risk - i.e. the likelihood of moving to another quintile increases for entrepreneurs relative to salaried workers.
2
Motives
What distinguishes entrepreneurs from non-entrepreneurs Over the course of two lectures, we consider three factors, beginning with preference for risk and status (as well as subjective perception of risk, e.g. optimism). In each case, we examine the empirical evidence.
18
2.1
2.1.1
Risk Attitude and Perception
Risk Bearing The entrepreneur usually, or to a large extent, claims the residual profits of the …rm, after wages and capital expenses have been paid. Why is this so - why do not e.g. the workers have pro…t-share contracts, and the entrepreneur is paid a …xed sum The entrepreneur' role in s the …rm is the most di cult to monitor, so he needs to be given the strongest performance incentives. According to Barzel [8], the entrepreneur facilitates collaboration within a …rm by assuming all risks (including that associated with supplying capital) and thereby eliminating opportunities to gain at the expense of others, which his unique position in the …rm would otherwise a¤ord him. Since he earns the residual, he does not need to be monitored. The entrepreneur commits to monitoring employees, who are paid a …xed wage (else the entrepreneur has an incentive to understate the …rm' income in a risk-sharing arrangement). The role of the entrepres neur is the most di cult to monitor and should therefore be the one to assume all risks.
2.1.2
Entrepreneurship and Risk Aversion Since the entrepreneur is the natural risk bearer, it is plausible that the decision to become an entrepreneur re ects a relatively high tolerance for risk. Kihlstrom and La¤ont [102] model the choice between accepting employment at a …xed, competitive wage and engaging in a risky production process (where a random variable determines the output for any given number of workers) as an entrepreneur. Individuals are distinguished by their level of (Arrow-Pratt absolute) risk aversion. In equilibrium, the least risk averse individuals become entrepreneurs and normally (under plausible technological conditions) run the largest …rms. If everyone in the population becomes more risk averse, wages are depressed, since more individuals choose to be employed (on the ipside, entrepreneurial pro…ts increase at a given level of employment, but …rms are in fact smaller). The equilibrium is generally ine cient in the Pareto sense because (i) risk-averse entrepreneurs limit the scale 19
of their operations below the surplus-maximizing level (the more riskaverse tend to run smaller …rms), (ii) there is no risk-sharing between entrepreneurs and workers (entrepreneurs shoulder all risk, though it would still be e cient for workers to bear some), (iii) there are typically either too many or too few people engaged in entrepreneurship, since risk aversion distorts the incentives (its …rst-order e¤ect is to discourage entrepreneurship, although by lowering demand for labor and wages it makes entrepreneurship more pro…table as a second-order e¤ect). Only in the extreme scenario where all entrepreneurs are risk-neutral is the equilibrium e cient. There are problems with verifying the relationship between risk aversion and propensity for entrepreneurship empirically, since the experience of entrepreneurship could in uence a person' level of risk aversion, s either directly or through the e¤ect on personal wealth. Ideally, one would like to have data on subjects'risk aversion before they decided to enter salaried employment or entrepreneurship. (Risk aversion can be measured by letting people choose between lotteries with di¤erent degrees of risk exposure.) Absent these data, one can consider less direct indicators of risk aversion. Fairlie [52] uses past drug-dealing as a proxy for low risk-aversion (perhaps combined with entrepreneurial ability and desire for independence). He …nds drug-dealers in the 1980 National Longitudinal Survey of Youth are signi…cantly more likely to become self-employed, and the more so if they were relatively successful (higher sales frequency, higher income, less personal drug use). Of course, drug-dealing is a risky activity, given the exposure to criminal prosecution, territorial wars with other drug-dealers and loss of product through robbery or con…scation. The author rules out that the propensity to be self-employed can be explained by limited employment opportunities because of a jail history (controls for past incarceration) or reporting "self-employment" as a way to conceal continued drug-dealing (surveys are too detailed to do this e¤ectively), or by asset accumulation while a drug-dealer (controls for net worth). Things are more complicated when we leave the world of known risk and enter a world of Knightian uncertainty. Typically, individuals will be averse to uncertainty and demand a greater return to entrepreneurship 20
in these circumstances. Even this world is simple compared to one where entrepreneurs may be unaware of some possible events. Although they can make production decisions without worrying about situations they may have failed to anticipate (do the best they can given their state of knowledge), investment decisions are very much a¤ected. This includes the decision to become an entrepreneur in the …rst place unforeseeable contingencies, such as catastrophic events or pro…table inventions in the future of the …rm, are di cult to factor in. Kawamura [96] shows that entrepreneurs who know they may be unaware of some possible states nevertheless produce at the level that maximizes the expected pro…t with respect to states they are aware of. Unawareness can have complex implications for allocative decisions, such as how much to invest when an unknown product may be discovered in the future and how to price …nancial instruments when agents implicitly risk failure due to their unawareness of harmful events. If e¤ort is su ciently important in determining pro…tability, it may no longer be the wealthiest and least risk averse who become entrepreneurs. If poorer people are more risk-averse (because a loss would hit them harder), they have an incentive to put in more e¤ort and therefore increase the expected pro…t. This should make it easier for the poor than the wealthy to …nd investors who will provide pro…t sharing and reduce the risk to them. In this case, the Kihlstrom and La¤ont intuition may be overturned, and it could be the most (not the least) risk-averse who become entrepreneurs. In Newman' model [132], individuals - who di¤er only in wealth s choose to be workers (at a …xed wage) or entrepreneurs (earning the residual pro…t from production) or nothing (consuming initial wealth). E¤ort choice a¤ects the probability distribution of the …rm' output. s For entrepreneurs, it is always optimal to exert high e¤ort; workers must be given the incentive to exert high e¤ort. The risk to an entrepreneur (spread of the pro…t distribution) increases in the number of workers hired. Without insurance, as in Kihlstrom and La¤ont, the wealthiest agents become entrepreneurs, and the wealthiest entrepreneurs hire the most labor and have the highest expected pro…t. With the option to insure, e.g. through a stock market (where entrepreneur
21
can lower risk by selling shares), the poorest agents become entrepreneurs for a wide range of preferences, and …rms are of equal size and have the same expected pro…t. If the wealthy do hire the poor in some scenario with multiple e¤ort levels, then workers are expected to have higher incomes than entrepreneurs (since the poorest are expected to exert the greatest e¤ort). The intuition is as follows. With the option to insure, more insurance is available to entrepreneurs who are poorer, because they have a greater incentive to exert e¤ort to avoid comparatively small losses. This means …rms run by poorer entrepreneurs should be able to sell more stock, which makes it more attractive for poor people than for rich people to become entrepreneurs. This result depends on relatively fast decreasing absolute risk aversion (which increases the incentive problem), the fact that wealth levels are observable, and the impossibility of bankruptcy. Even though these are strong assumptions, it makes it problematic to see occupational choice only in light of risk preference. 2.1.3 Cognition Di¤erent people perceive risk di¤erently: there are optimists and pessimists. The perception of risk a¤ects, in conjunction with risk aversion, who chooses to be an entrepreneur. Casson [31] adds to the Austrian emphasis on di¤erential access to information one on di¤erential perception of information. The entrepreneur has a more optimistic view of the imperfect information available about volatile events, the response to which cannot be routinized. The entrepreneur becomes a market maker by convincing other parties, such as investors, employees and consumers, that a viable opportunity exists. Arabsheibani, de Meza, Maloney and Pearson [2] examine British Household Panel Study data from 1990-1996, where individuals are asked to evaluate their current economic situation and also their expected situation a year later. The self-employed were signi…cantly more likely to anticipate an improvement when they actually experienced a deterioration, hence were more optimistic. Optimism can also be detected indirectly. If we assume that banks 22
assess ventures fairly objectively, then high collaterization of loans indicates a lack of con…dence on the part of the bank relative to the entrepreneur' expectations. Similarly, high incidence of self-…nancing s can mean that overcon…dent entrepreneurs regard the risk component of the interest rate as unjusti…ably high. de Meza and Southey [42] cite evidence that banks tend to assess entrepreneurial risk more realistically than the entrepreneurs to whom they lend (e.g. businesses with highly collaterized loans - re ecting bank pessimism - are more likely to fail). Their model with well-informed banks and poorly informed entrepreneurs predicts that the average return to entrepreneurship is negative, that only optimists become entrepreneurs, and that they self-…nance as much as possible (since they believe the interest rate is excessive). The poorest optimists do not become entrepreneurs or enter at an ine ciently low scale. In equilibrium, the more collaterized the debt, the more likely the project will fail. Learning and declining optimistic bias account for initially large variation in entrepreneurial returns that diminishes over time. Fraser and Greene [57] make the point that entrepreneurial learning about …rm viability and rational exit decisions imply that the variance in entrepreneurial returns is greater for new start-ups than for established …rms. They …nd empirically, from the British Social Attitudes Survey (BSAS) for 1984– that entrepreneurs are more op99, timistic than wage workers, that more optimistic entrepreneurs start larger …rms, and optimism diminishes with entrepreneurial experience.
2.2
2.2.1
The Private Equity Premium Puzzle
Low Return to Entrepreneurship If entrepreneurs are risk-averse (as they were, to some extent, in all of the above explanations), they should demand a premium over what they could earn in salaried employment. Empirically, this premium does not appear to exist. Hamilton [75] …nds that self-employment income, measured in a number of ways in the Survey of Income and Program Participation (19831986), after 10 years in business falls short of predicted alternative earn23
ings under employment by about 35% on average and exhibits lower earnings growth. (The median present value of an entrepreneurial …rm surviving for 25 years is over 25 percent less than the present value of holding a paid job for 25 years.) This does not include employer fringe bene…ts (such as health insurance), which would further widen the gap. The median of entrepreneurial earnings is, at any length of operation, less than the average starting wage, and the distribution of entrepreneurial pro…t exhibits greater spread (risk) than the wage distribution (a few entrepreneurs with very large earnings are responsible for a signi…cant share of the average entrepreneurial income). The three measures of entrepreneurial income are net pro…t (which is often understated for purposes of tax evasion), draw (the income that the entrepreneur pays himself as a wage), and draw plus change in the value of the …rm' equity (which re s ects reinvested income and capital gains). The author rejects the possibility that low-wage earners select themselves disproportionately into entrepreneurship (wages of entrepreneurs before self-employment were at, if not a little above, average). Moskowitz and Vissing-Jorgensen [126] discuss the "private equity premium puzzle," namely that entrepreneurs in the 1990s invested much of their wealth in the …rm they run, exposing themselves to substantial risks, while the average return is no higher than that from a diversi…ed portfolio of publically traded equity. This is also a problem because entrepreneurs receive their wage income from the same …rm, so are doubly exposed to failure. In fact, one would expect investors to demand a private equity premium of 10% or more. The "private equity premium puzzle" is even more striking given the highly publicized "equity premium puzzle," which is concerned with excessive returns to stock vs. safe bonds. There, the puzzle is opposite: investors would have to be extremely risk-averse for the di¤erence in returns to make sense. Here, they would have to be risk-neutral - the coexistence of both return gaps is even more puzzling. 2.2.2 Explanations In order to reconcile the relatively low returns to entrepreneurship with the fact that people nevertheless choose to be entrepreneurs, economists have looked for entrepreneurial bene…ts that have not been accounted 24
for. One possibility are tax advantages or capital gains. Another is that entrepreneurs implicitly "purchase" a valuable option. Christensen [33] considers corporate and non-corporate rates of return to capital in the US between 1929 and 1969. Unlike studies that found much higher corporate rates of return, he uses after-tax rates that are adjusted for capital gains. These rates, which are the relevant ones for investors, are similar (7.6% in the non-corporate vs. 7.3% in the corporate sector). The technique is to impute labor share in entrepreneurial income, with the assumption that annual earnings of the self-employed should on average equal those of the employed (ignoring possible differences in average education, and also that the reported number of self-employed workers may exclude contributing family members and include entrepreneurs who do not devote all their time to the business). The remaining income is attributed to entrepreneurial capital and can be compared to the corporate sector. From this income taxes are substracted (which were substantially higher in the corporate sector over the period), and capital gains are calculated from replacement costs of the assets (using average prices for broad capital categories, like structure, land, durables and inventories). The non-corporate sector bene…ts from capital gains relative to the corporate sector in part because land is a much larger share of its assets, and land has appreciated much faster than other categories. Christensen notes that if the corporate and non-corporate returns to capital in fact o¤set each other, then the disutility from risk aversion and the positive utility from being one' s own master must approximately o¤set each other. Polkovnichenko [141] argues that educated entrepreneurs have a high tolerance for low average returns because they have the option to reenter paid employment if their business fails. I.e. they do not put at risk their future earnings potential. Taking this into account (by pricing the option and imputing it into the return to entrepreneurship), he …nds a very modest private equity premium remains, which remains to be explained by other bene…ts of entrepreneurship. Bohacek [25] sees early-stage operation of a business as an investment in future growth. People compare their "career incomes" as entrepreneurs or workers and might forego a better-paying salary job to take advantage of learning e¤ects in entrepreneurship that improve productivity. 25
The increasing optimal scale of the …rm provides a key motivation for entrepreneurs, who face borrowing constraints, to save more than workers. Entrepreneurship is expected to eventually pay o¤ by inducing a steeper income path. Finally, it is possible that incomes of entrepreneurs and employees are not directly comparable: if employees have di¤erent characteristics from entrepreneurs, then it is di cult to predict counterfactually what an entrepreneur would earn in employment. Rees and Shah [145] examine data for self-employed men from the four quarterly samples of the 1978 UK General Household Survey. They identify positive selection bias in employment: i.e. for given individual characteristics, earnings are higher in employment than predicted earnings in self-employment. On the other hand, there is no selection bias for the self-employed: an entrepreneur does not earn more than an employed worker with the same individual characteristics. (These statements can both be true because in one case we consider the individual characteristics that are prevalent among the employed, and compare them to a hypothetical self-employed individual based on the estimated model, whereas in the other case, we compare on the basis of the individual characteristics that are found in entrepreneurs.) The interpretation is that those in paid employment have a comparative advantage in being employed (but the same is not true for the selfemployed). They also …nd that higher education, age, and good health increase the likelihood of being self-employed.
2.3
2.3.1
Status Preference
"Own Boss" Bias What remains of the Private Equity Premium Puzzle must be attributable to an inherent preference for being one' own boss. This prefers ence is well-documented empirically and can also be given a theoretical explanation in terms of a preference for greater relative (not only absolute) income. Benz and Frey [14] …nd empirically that self-employed people report greater job satisfaction. Data come from the German Socio-Economic 26
Panel Survey (GSOEP), the British Household Panel Survey (BHPS) and the Swiss Household Panel Survey (SHP). The authors carefully rule out that this is because happier people select themselves into selfemployment. They determine, …rst in their dataset and second in the "natural experiment" that occurred when the Berlin Wall fell (and a relatively large population suddenly switched to self-employment) that individuals whose status has changed between employment and selfemployment are more satis…ed with their jobs in the period where they are self-employed. Blanch ower, Oswald and Stutzer [22] …nd that many more people would like to be self-employed than actually are, which is evidence of constraints to starting a business and, given the magnitudes, indicative of a strong preference for self-employment in most countries. E.g. more than 70% would prefer to be self-employed in Poland, Portugal and the US; in most countries in the sample, the number exceeds 30%. Everywhere, signi…cantly more people want to be self-employed than actually are. 2.3.2 Relative Income Clemens [35] considers status preference: individuals care (1) about their income relative to the universal mean or (2) about their profession' average income relative to the universal mean (e.g. while individs ual income is not publically observable, occupation and occupational income are). Status e¤ects are two-fold: they increase the rewards from a more pro…table occupation (more so because higher income imposes an externality on others), but they exacerbate the impact of riskiness (a potential loss is more catastrophic because it also involves a loss of status). Hence, whether status preference increases or decreases the number of entrepreneurs depends on whether risk aversion is generally low or high if status is tied to own income (1). If it is tied to the average income of the occupation (2), which is less risky than individual income, the rewards e¤ect predominates and status preference increases the number of entrepreneurs, partially o¤setting the ine ciently low supply due to risk aversion.
27
3
Means
We entertain theories that entrepreneurs possess personal or material advantages such as greater talent, education, social privilege or wealth.
3.1
3.1.1
Talent and Quali…cations
Return to Talent How well does the market reward those who have a superior talent in creating a product Rosen [149] shows that small advantages in talent can have disproportionately large payo¤s. If an individual can produce a slightly superior good at the same cost as others, then it is possible to extract a higher price while selling a larger quantity. This means that the optimal revenue increases in a convex fashion with talent (the "price of talent" increases nonlinearly in talent), and that the distribution of returns is right-skewed relative to the distribution of talent: i.e. a few people at the top of the talent distribution make most of the money and become "superstars." MacDonald [113] studies a two-period model where agents decide initially whether to enter the industry considered by Rosen, and then whether to stay on. An agent' talent is initially unknown, but a track s record of "good" or "bad" performance is revealed to the public through end-of-period reviews; second-period talent is stochastically increasing in …rst-period talent. Since the product ("tickets") is sold ahead of each period, earnings are necessarily low for new entrants, but if they perform well, they may achieve superstar rewards in the second period. This can be seen as an analogy to entrepreneurial entry in search of abnormally high returns. A characteristic of the model is that higher superstar pro…ts lower the pro…ts required at the initial stage, since the greater potential prize increases entry and forces down "ticket" prices for new players. Hence people may choose to become entrepreneurs even if initial pro…ts are very low, and exit if their products do not garner su cient praise early on. Demand divides between products from established agents at high prices (more quality-conscious consumers)
28
and products from new entrants at low prices (less quality-conscious consumers). 3.1.2 Adverse Selection into Employment or Entrepreneurship When talent is more perfectly observable in employment or in entrepreneurship, there is an added incentive for more talented individuals to move into the more informed sector, and similarly an incentive for less talented individuals to move out. I.e. there is adverse selection, leading to an ine ciently high or low number of entrepreneurs. Laussel and Le Breton [105] develop a model in which individual ability is private knowledge and can therefore only a¤ect entrepreneurial payo¤s, but not wages. The ablest individuals therefore become entrepreneurs, preferring their income in pro…ts, which increase in ability, to the uniform wage. This adverse selection into the employment sector can lead to an ine ciently large number of entrepreneurs because those with su ciently greater ability than the average employee have an incentive to switch to entrepreneurship, thereby further reducing the expected ability and wage of the employed and causing more employees to switch. Decreasing returns to scale counterbalance this e¤ect by raising the e cient number of entrepreneurs. Analogously, increasing returns to scale reinforce the tendency for there to be too many entrepreneurs. Adverse selection into entrepreneurship has "peculiar" policy implications, as Ghatak, Morelli and Sjostrom [67] showed. Interest rates and product prices depend on the expected talent of the entrepreneur, hence implicitly on the wage level, since it determines the talent of the marginal entrepreneur. An argument can be made that wage increases bene…t not only workers, but also (those who stay) entrepreneurs, by improving the beliefs of creditors and consumers about the typical entrepreneur' quality, which leads to better contracting terms (lower s interest, higher prices). Policies that raise wages (such as a higher minimum wage or investments in labor productivity) therefore also create jobs and may turn out to be e cient. This conclusion is weakened, but can still hold, when a screening device exists (such as wealth that entrepreneurs can use as collateral in borrowing). Then, an increase in wages will not bene…t entrepreneurs directly, but it does reduce the 29
cost of screening (less talented entrepreneurs have a lesser incentive to put up a given collateral to obtain credit). Risk-averse entrepreneurs bene…t indirectly because less collateral, i.e. less risk taking, is required of them. Parker [138] shows how adverse selection into entrepreneurship can lead to underinvestment. Suppose workers can signal their talent, while entrepreneurs cannot. Then the least able individuals become entrepreneurs and borrow from banks, who charge high interest rates because of the high risk of default. The pooling of entrepreneurs with di¤erent abilities in the debt market implies that the more able would have to pay excessive interest on loans (given their low default risk). This biases them toward employment, where they can earn relatively good wages. Some socially desirable entrepreneurial ventures are not carried out. On the other hand, if entrepreneurs could be distinguished by talent (but not workers), overinvestment in entrepreneurial ventures would result, as high-skill workers escape low wages that result from adverse selection into employment. 3.1.3 Evolution of Firm Size Entrepreneurs may not perfectly know their talents, hence the prospects of their venture. This explains changes in the size of …rms over the life cycle: …rms grow if entrepreneurs observe that they are successful and are willing and able to invest more capital. Jovanovic [92] argues that relatively high failure rates among small …rms, coupled with high growth rates on average, re learning about ect e ciency. New …rms enter with imperfect information about their costs and grow quickly if they turn out to be competitive. Large …rms have been shown "…t" to survive, while ine cient new …rms frequently go out of business. Frank [56] explains the empirical pattern of the …rm life cycle through learning about entrepreneurial talent. Initially, the most con…dent individuals enter and invest the most. Over time, observations of actual …rm pro…tability provide feedback and prompt the entrepreneur to revise his opinions. In general, it takes some time to gather enough negative information to convince the entrepreneur to exit, given the high 30
initial con…dence that is required for entry. In practice, the largest wave of exits occurs after 3-4 years. Since …rms with greater investments are run by more con…dent entrepreneurs, they are slower than others to exit, which can create the illusion that entrepreneurs irrationally stay in the market due to high sunk costs. Newer …rms tend to be smaller than established …rms, since a …rm needs to be certain of its viability, through experience in the market, before it will invest in large scale operations. Lucas [112] analyzes a model where individuals have di¤erent managerial skills and the returns to management are decreasing (every additional worker under a given manager is less productive). The most able managers become entrepreneurs, run the largest …rms and collect the largest rents, proportional to …rm size. As wealth and therefore the capitalization per worker increases, the managerial "time" constraint is e¤ectively relaxed (wages rise more than entrepreneurial rents), and marginal entrepreneurs have an incentive to become workers. Hence the …rm size distribution tends toward fewer and bigger …rms over the course of development. Oi [134] assumes that entrepreneurs must divide their time between productive activities and monitoring. More able entrepreneus have the capacity to monitor more workers, and can thus build larger …rms, but they are also more productive, so that their opportunity cost of monitoring is greater. The author shows how di¤erences in monitoring cost a¤ect the …rm' operations with respect to organization, worker s productivity and product design. 3.1.4 Education Can education improve entrepreneurial prospects Schultz [150] views entrepreneurs, in the Austrian spirit, as arbitrageurs who respond to the pro…t opportunities that arise from disequilibria. He emphasizes the role of education in endowing people with "allocative ability:" the perceptiveness and exibility to take advantage of changing economic circumstances. In Lazear' [106], [107] theory of balanced skills, the entrepreneur must s be familiar with all tasks that are carried out in a business (in order to 31
hire and monitor people), while workers are employed and rewarded for only one task. That is, workers are specialists with incomes constrained by their strongest skill, and entrepreneurs are generalists with incomes constrained by their weakest skill. The theory predicts that entrepreneurs are more likely to invest broadly in all skills, whereas specialists invest in a single skill. This is supported by course choices of Stanford MBA students: those who go on to be entrepreneurs tend to take more courses outside the primary …eld. Also, entrepreneurs are expected to have generalist, rather than high-tech, backgrounds (which would be di cult to attain while also mastering management skills). Lazear cites evidence of more entrepreneurs in non-technical occupations. Friar and Meyer [58] distinguish between micro businesses (that do not innovate and are not major players in their industry, at national or local level) and high-growth ventures (innovative, pro…t- and growthoriented). They report results of an annual business plan competition at Northeastern University in Boston, where entries are judged by an industry expert panel and classi…ed as micro businesses or high-growth ventures. In a sample of ninety semi-…nalists, submitted over four years, plans for high-growth ventures were typically created by teams involving students with work experience or advanced training relevant to their ideas. On the other hand, it has been argued that education can discourage entrepreneurship, e.g. by leading to other opportunities or through a signaling e¤ect. Nicholas [133] collected information about roughly one-thousand notable British entrepreneurs, born between 1789 and 1937 and deceased between 1868 and 1993, from business dictionaries and probate records. He uses inheritances to deduce lifetime wealth accumulation. Being a (second- or third-generation) heir to an entrepreneur' fortune and s having a high-status education (from a Clarendon school or Oxford or Cambridge University) correlate negatively with wealth accumulation. These entrepreneurs earned little more than the return to land (the equivalent of putting money in the bank today). Region and type of industry (staple vs. new) where the entrepreneur was active, as well as religion, have no signi…cant e¤ect. If entrepreneuerial talent in a family is not correlated across generations, it makes sense that the founder' s 32
fortune is typically "lived down" by the descendants. Prestigious education is to some extent explained by inheritance, but also plays an independent role: possibly by opening up non-business pursuits in art or politics. Orzach and Tauman [135] explain why entrepreneurs sometimes choose to drop out of higher education. In their model, there are ordinary and gifted individuals; both types have the ability to acquire education that increases the probability of succeeding as an entrepreneur. However, while the expected entrepreneurial pro…t is always positive for gifted individuals (regardless of the level of education), it is negative for ordinary individuals unless they get educated. Although education increases the ordinary individual' likelihood of success, it never exceeds s that of the gifted. In order to attract external funding, gifted individuals want to signal their type. They do this (in a separating equilibrium) by not getting educated, thereby demonstrating that they can a¤ord to do so because of their innate likelihood of succeeding as entrepreneurs. Ordinary individuals, however, require education in order to be entrepreneurs, but this signals their types to lenders, who prefer to fund the gifted. Hence ordinary individuals become paid employees.
3.2
3.2.1
Social Barriers
Ethnicity, Immigration and Self-Employment Membership in an ethnic group bestows problems and opportunities for entrepreneurs: while they may have di culty accessing demand of prejudiced groups, they may be in a better position to meet the demand of their ethnic group. Borjas and Bronars [27] rule out, on the basis of 1980 US census data, that the self-employment rates of blacks and Hispanics are determined in the same way as for whites: judging by population characteristics, they should be similar, but far fewer blacks and hispanics are actually self-employed. Selection (by skill) into self-employment is positive for whites, zero for blacks, negative for Hispanics and Asians. (I.e. among whites, but not among minorities, the self-employed tend to be more skilled). The authors present a costly search model where consumers and entrepreneurs meet randomly and discover their reservation and 33
o¤er prices, as well as the entrepreneur' race. Whites are assumed s to dislike buying from minorities (e.g. when the product is a personal service). More able entrepreneurs can produce at a lower cost, hence have a greater incentive to serve all races, instead of just their own. The implication is that returns to ability are lower for minority than for white entrepreneurs, since a minority entrepreneur must charge a lower price in order to attract white consumers. While minorities may also be discriminated against in wage employment, this does not a¤ect the premium a worker can get for greater ability. Since returns to ability are depressed for entrepreneurs only, skilled minorities are more likely to seek employment, and unskilled minorities will be overrepresented as entrepreneurs. I.e. the model predicts the negative selection we see in the data. Moreover, it predicts that the mean (after controlling for skill) and variance of entrepreneurial incomes is smaller for minorities than for whites, and this is also borne out by the data (except for self-employed Asians, who on average earn more than whites). Borjas [26] cites statistics according to which the incidence of selfemployment is, for all races, greater for immigrants than US-born Americans. Immigrants are relatively concentrated in retail and cater speci…cally to customers from their own ethnic groups. The author …nds that the likelihood that an immigrant is self-employed is higher after "assimilation" (longer time in the country) and if the immigrant resides in an enclave (a region with a large population of the same ethnic origin). This con…rms that enclaves provide a demand that immigrants have a comparative advantage at meeting, after some period of settling in and accumulating capital. The incidence of self-employment is greater in recent waves of immigrants than earlier ones, which might be explained by "family reuni…cation" laws that permit spouses to follow and therefore provide an additional source of motivated labor for the entrepreneur. 3.2.2 Gender Women are underrepresented among nascent entrepreneurs relative to men. Minniti and Nardone [124] show empirically that the di¤erence is not explained by socio-economic variables (age, household income, empoyment status, education, country-speci…c economic factors), but 34
is largely due to perceptual idiosyncracies (women are less likely to feel quali…ed, have a greater fear of failure and judge opportunities more pessimistically). Their study is based on international data from the Global Entrepreneuship Monitor.
3.3
3.3.1
The Liquidity Constraints Debate
Evidence of Constraints The question whether wealth is essential to entrepreneurship concerns the perfection of capital markets. A fully functional capital market channels funds to any venture that is expected to turn a pro…t above the average market return. In practice, information problems may prevent this, and entrepreneurs then have to own collateral to guarantee loans with. Does a sudden increase in wealth increase the probability of switching into self-employment Holtz-Eakin, Joulfaian and Rosen [85].conclude from US tax return data for 1981 and 1985 that receiving a larger inheritance signi…cantly increases the likelihood that an individual quits wage employment and becomes an entrepreneur, as well as the start-up capitalization. Blanch ower and Oswald [21] infer liquidity constraints from the "natural experiment" a¤orded by inheritances in the British National Child Development Study (a longitudinal dataset surveying all UK residents born in a particular week in 1958, at birth and ages 7, 11, 16, 23, 33). They …nd that the likelihood of entering self-employment (in 1981 and 1991) is greater when an individual has received a larger inheritance in the past. They also report that, when individuals are asked directly what keeps them from starting a business, in the British Social Attitudes Survey (1983 to 1989), they most commonly mention di culties in obtaining capital. Black, de Meza and Je¤reys [20] …nd evidence that house appreciation in the UK signi…cantly increases new start-ups, and these start-ups have average survival prospects. This suggests the entrepreneurs are credit-constrained and are able to use the additional collateral to secure loans to get started. 35
Conclusively tracing capital market imperfections in data is complicated because, as we have already seen, the observation that wealthier people are more likely to start …rms has explanations other than credit constraints. It can re that greater wealth reduces risk averect sion (hence the required return to entrepreneurship), or that wealthier individuals value the option to work part-time that comes with selfemployment. A consensus is emerging that credit constraints do not tend to prevent entry, but make it di cult for entrepreneurs with highreturn projects to operate at the e cient scale. Hurst and Lusardi [87] amass empirical evidence against liquidity constraints from the Panel Study of Income Dynamics (1984-1994) and National Survey of Small Business Finances (1987). They do not …nd evidence that entrepreneurs from poorer households tend to venture into sectors with relatively low start-up costs. Start-ups are correlated with past as well as future inheritances, suggesting that the relationship between wealth accumulation and entrepreneurship is not causal, but rather driven by a third factor. Housing appreciation, a wealth increase beyond the entrepreneur' control, did not increase the probability of s starting a new business in their analysis. The authors suggest that the correlation between personal wealth and entrepreneurship has largely reasons unrelated to …nancial constraints: risk aversion, the same factors drive income and entrepreneurial aspirations, entrepreneurship is a luxury good. However, while the entry decision does not seem to depend on wealth, the start-up size could (i.e. businesses might operate at suboptimal scale because of limited funds). Evans and Jovanovic [51] determine empirically, from the National Longitudinal Survey of Young Men for 1976 and 1978, that entrepreneurs are …nancially constrained: on average, entrepreneurs only have access to about one-and-a-half times the value of their personal assets. In the Evans and Jovanovic model, as in others, the optimal scale at start-up increases in the entrepreneur' ability. They argue, based on s typical asset endowments, that entrepreneurs of ordinary ability tend to be able to start out on the scale that is compatible with their ability. However, the ablest entrepreneurs are often …nancially constrained. This is re ected in the observation that an increase in capital on average increases pro…tability, hence that some entrepreneurs are forced to 36
operate below the optimal scale. 3.3.2 The Course of Development with Capital Constraints Capital constraints are, of course, a much more likely proposition when we look at economies in the early phases of development. The very poor have no collateral to o¤er and therefore will not qualify for loans. The central issue is whether the constraints are going to persist or diminish, i.e. whether countries that start out poor will create jobs in which ordinary individuals accumulate enough wealth to eventually have access to capital markets. Banerjee and Newman [7] model a world where individuals choose one of three investment options: a riskless asset (which allows them to work in a factory), cottage industry (then they are self-employed, operating a single-person risky technology), a factory (then they are large-scale entrepreneurs, operating a technology that monitors workers who are hired at a market-determined wage). Running a factory requires the largest investment, a multiple of the cottage industry investment, proportional to number of workers employed. The equilibrium has the property that those with wealth above a certain level become entrepreneurs (invest in factories), the next group becomes self-employed, and the poorest group chooses wage employment or subsistence (unemployment) if there is not enough demand for labor. There are many possible wealth distribution scenarios (depending on the precise nature of the production risks), which determine inter-generational mobility between wealth levels. In one, we have (i) cottage industry and factories with little social mobility between workers, self-employed and large-scale entrepreneurs. In another (ii), we have mostly cottage industry, again stable over time. There are also parameterizations where the economy either stagnates in a bad equilibrium with most individuals not working and simply investing in the safe asset (when there are initially many poor and few wealthy), or a good equilibrium with frequent inter-generational transitions between worker, self-employed and entrepreneur roles (when there are initially few poor). We get (i) when there are initially many people with little wealth, who necessarily become workers. Due to resulting low wages, they cannot bequeath enough to their descendants for them to become self-employed, while 37
entrepreneurs earn enough to stay entrepreneurs. If few individuals are relatively poor to begin with (ii), then workers are in short supply / too expensive, and everyone is self-employed. Lloyd-Ellis and Bernhardt [110] focus on the interaction of the entrepreneurial skill and wealth distributions along the development path. Initially, credit constraints only make it possible for the wealthy (and their descendants) to be entrepreneurs. Competition for workers ultimately bids up wages and enables the initially poor to become entrepreneurs. This process leads to continuous increases in productivity and sustained growth if there are su ciently many people with high entrepreneurial ability. However, rising wages drive less talented entrepreneurs - who initially started …rms because of their access to capital, and because returns were high at a low wage level - out. If entrepreneurial talent is scarce, they may not be replaced by su ciently many new entrepreneurs, and the economy may actually shrink as a result - until wages fall to a point where less talented entrepreneurs reenter. Then we observe boom and bust cycles.
4
Venture Finance
We delve more deeply into the issue of funding an entrepreneurial venture. What constrains bank lending How is the capital structure (debt / equity) determined
4.1
4.1.1
Self-Funding
Entrepreneurial Saving If entrepreneurs are liquidity-constrained, then we expect them to save more than non-entrepreneurs. This is true empirically and can be given more speci…c theoretical support. Gentry and Hubbard [65] report that entrepreneurial households save more than others, invest the savings in their own business, and hold a substantial share of overall wealth in the economy. From the 1983 and 1989 Federal Reserve Board Surveys of Consumer Finances, the authors estimate that new entrepreneurs (who started a business between the 38
two surveys) save on average 35% more of their income than salaried workers, while continuing entrepreneurs save about 16% more. As a result, between 1982 and 1988, the median wealth-income ratio increased for new entrepreneurs from 2.5 to 4.0 and for continuing entrepreneurs from 6.1 to 7.9, compared to 1.3 to 1.7 for non-entrepreneurs and a falling ratio from 6.5 to 4.0 for exiting entrepreneurs. In 1989, less than 10% of individuals were entrepreneurs, but they held almost 40% of all wealth. Nearly every third household in the top 5% by income was entrepreneurial. New entrepreneurs gained about 85%, and continuing entrepreneurs 67%, of their additional wealth through an increase in the value of their business. Saving makes start-ups possible, and can also reduce the cost of funds and allow the …rm to operate at the optimal scale. Quadrini [142] calibrates a model of entrepreneurial saving to match observations from the Panel Study of Income Dynamics (1984, 1989, 1994) and the Survey of Consumer Finances (1989, 1992): (i) business-owning households have much higher wealth-to-income ratios than salary-earning households (on average twice as high), and (ii) business-owning households tend to maintain or improve their positions in the wealth distribution, whereas salary-earning households tend to maintain or deteriorate their positions. The author shows that these and other empirical facts can be explained by the greater incentive for business owners to save: initially, they must generate the wealth that is required for starting a venture, then they save in order to lower interest payments (by substituting for loans and building up collateral). Morever, because the income of entrepreneurs is more exposed to risk, they have a greater precautionary motive to save than workers do. Wealth accumulation by business-owning households is reinforced by their persistence as entrepreneurs (the typical business-owning households has been in this role for considerable time and proven itself, so that temporary setbacks will not cause them to quit entrepreneurial pursuits). Cagetti and De Nardi [29] successfully match a model to the data where entrepreneurs save more in order to expand their business and take advantage of superior returns. Moreover, there is an incentive to pass family wealth on to the next generation, in order to facilitate any entrepreneurial aspirations children might have. This type of behavior 39
is evidence of capital constraints, since entrepreneurs could otherwise borrow and make consumption choices like wage workers, who are unconstrained by the need to …nance a business. Capital market imperfections not only necessitate saving for prospective entrepreneurs but also create rents that make entrepreneurship more attractive. Furthermore, they induce precautionary saving, since entrepreneurs are unable to borrow in low-income periods. Ghatak, Morelli and Sjostrom [66] build a dynamic model where young agents work hard and save in order to become entrepreneurs later, in the style of the "American dream." Entrepreneurial production requires ownership of an asset that one can either self-…nance through wage income or buy with borrowed funds. Credit market imperfections imply that borrowing is expensive; not all workers manage to save enough to purchase the asset (even with much e¤ort, earnings under an incentive contract partly depend on luck). Hence, entrepreneurs are in limited supply and are able to o¤er wage contracts that leave them with a rent. This provides the incentive to try to maximize saving when young by committing full e¤ort. The extra motivation mitigates the moral hazard problem between entrepreneurs and workers, is self-reinforcing because it increases entrepreneurial rents, and improves e ciency. Implicitly, capital market imperfections have a social advantage: without them, entrepreneurial rents would disappear, and e¤ort incentives would be reduced. Covas [38] highlights the strength of the precautionary savings motive in an environment with borrowing constraints and uninsurable production risks. High demand for safe assets lowers their interest rates and forces entrepreneurs to generate income by investing in risky projects. This in turn increases risk exposure and necessitates more precautionary saving. As a result, entrepreneurs are exptected to "oversave" dramatically relative to a world with unrestricted borrowing and insurance. Meh [118] points out that, due to the entrepreneurial "saving for my business" motive, revoking the corporate income tax could have a counterintuitive e¤ect on saving behavior. Since the corporate income tax amounts to double taxation of business income (on top of the personal 40
income tax), one would ordinarily expect that its removal raises pro…ts and therefore overall saving. However, it would also diminish the incentive for unincorporated entrepreneurship that previously o¤ered a way out of double taxation. If such entrepreneurship is capital-constrained and therefore requires substantial personal saving, then switching to corporate status with its associated opportunities for raising capital would eliminate high savers. This dampens, or even reverses, the increase in saving absent the corporate income tax. 4.1.2 Family Contributions Basu and Parker [9] model informal lending to entrepreneurs by family and friends as a rational decision. Over half of the South Asian owned start-ups in the UK rely, at least to some extent, on family loans. The authors model, and …nd empirically, that both altruistic and "selfish" motives play a role in family lending: there is an expectation of reciprocity (i.e. access to funds if the lenders themselves have entrepreneurial aspirations later). If family members are purely sel…sh, then a feasible arrangement is to borrow a proportion of a relative' wealth s in exchange for the entrepreneur' implicit promise to lend the same s proportion later (in case the relative wants to start a business). For the entrepreneur this is preferable to a bank loan if the interest rate is lower and the probability of having to reciprocate is su ciently small. For the relative, it is preferable to earning the market interest rate if the entrepreneur is expected to become wealthy, so that the promised reciprocal loan is substantial. However, an arrangement where a …xed sum is lent in exchange for the promise to lend the same sum is not feasible, unless the relative is altruistic (or the entrepreneur cannot borrow from the bank).
4.2
4.2.1
Information Issues and Relationship Finance
Moral Hazard Why is access to capital constrained Fundamentally, entrepreneur and investor have to overcome asymmetric information problems that arise because the entrepreneur cannot be completely monitored. The entrepreneur has no inherent interest to protect the investment and
41
might take excessive risks or misreport pro…ts or abandon the venture. These concerns can lead to limited availability of capital or rule out certain types of contracts. Paulson, Townsend and Karaivanov [140] make the observation that the source of the credit constraint is either limited liability (entrepreneur is not fully exposed to risk of failure) or moral hazard (entrepreneur does not work as hard as he can because pro…ts are shared with external investors). The entrepreneur' demand for funding as wealth increases s (the constraint is relaxed) depends on which source is dominant. If limited liability drives the capital constraint, then a wealthier entrepreneur will take advantage of the ability to borrow more. If moral hazard is key, then more wealth will lead the entrepreneur to substitute personal assets for borrowed funds in order to maximize his incentive to work hard. Based on these relationships, Thai data from a survey in 1997 suggest that moral hazard is the primary source of capital constraints. Stiglitz and Weiss [155] explain credit rationing by the fact that banks cannot continue to increase interests rate to clear the market (attract deposits) without changing default risks. At a higher interest rate, safe projects face the greatest expected interest cost, since they are the most likely to be repaid. Hence, entrepreneurs respond to a rising interest rate by increasing the riskiness of their investment: accepting a higher probability of default (in which case, they do not repay fully) for a smaller probability of large gains (in which case, they pro…t substantially despite high interest payments). Raising collateral requirements excludes poorer, more risk-averse entrepreneur and therefore also leads to more risk-taking, reducing the expected repayment to the bank. At some level, it is optimal for banks to ration credit randomly, rather than increase the interest rate or collateral and su¤er adverse selection, through the exit of the most capable entrepreneurs from the borrower pool, and moral hazard, in the form of high-risk project selection. Hart and Moore [76] show how lending terms are a¤ected by the possibility that an entrepreneur, who cannot be perfectly replaced, might withdraw from the project in the future. Because the entrepreneur can attempt to renegotiate terms, under threat of quitting, the debt can never become so large that the entrepreneur would prefer liquidation
42
(and lose his own funds in the …rm). On the other hand, the entrepreneur cannot be forced to repay too early, since he initially needs to be able to invest. These constraints - deferred repayment and limited accumulation - place a cap on the amount the entrepreneur can be allowed to borrow. Innes [89] notes that, whereas in practice pro…t-sharing between an entrepreneur and investors is such that investor earnings are increasing in the …rm' pro…tability (e.g. straight debt or equity), it may be optimal s for the entrepreneur to retain all pro…t when it is su ciently high (and get a constant share otherwise). The entrepreneur' increased e¤ort s (and higher resulting pro…t in any state of the world) can more than compensate investors for the absence of claims in a highly favorable, but relatively unlikely scenario. The reason such contracts are not normally observed may be that the entrepreneur then faces moral hazard in reporting results (an incentive to overstate pro…t in some cases). 4.2.2 Venture Capital Venture capitalists take a fairly active role in the business they invest in. In order to capitalize on their strengths, they invest when monitoring is important due to severe information asymmetries, i.e. when the relationship cannot be arm' length. s Kerins, Smith and Smith [97] estimate, for high tech IPOs in the late 1990s, that entrepreneurs have two to four times as high an opportunity cost as well-diversi…ed investors, such as venture capitalists, in supplying funds. I.e. injecting private assets is, in the economic sense, far more costly than sourcing through the market. The authors derive the entrepreneur' private cost of capital from the required return s in the capital asset pricing model, where systematic risk for privately held …rms is derived from the systematic risk of young publically traded …rms. This is compared to the investor' cost of capital, which re s ects the average market return. Winton and Yerramilli [162] compare debt and venture capital …nancing for a new start-up. While the venture capitalist requires a higher return for its investors (to o¤set its greater liquidity needs, given the responsive and risky nature of its investments), it is able to quickly 43
furnish capital when the …rm encounters growth opportunities. The venture capitalist monitors the entrepreneur more actively and can immediately verify such opportunities, whereas a bank con…nes itself to ensuring that the conditions of the debt contract are observed. Hence, …nancing through venture capital is optimal when the business is operating in a fast-paced, high-risk environment, where extreme returns are possible, but not very likely (else, the entrepreneur would incur a su ciently large debt from the outset). Kaplan and Stromberg [95] analyze the properties of over 200 investments by 14 venture capitalists. The allocation of control rights (board seats, voting rights, liquidation rights) is conditional on performance: the venture capitalist claims more powers if the …rm does poorly, otherwise the entrepreneur gains more independence and the venture capitalist retains only cash rights. Venture capitalists often impose ow provisions that bar the entrepreneur from competing against the venture, in order to minimize the hold-up problem (the entrepreneur' s incentive to bargain for a greater share in pro…t by threatening to leave and start a similar business). All these empirical features are consistent with the key theoretical predictions. Hellmann [80] shows that it is in the interest of entrepreneurs to give venture capitalists the right to …re them, even if the entrepreneur has an intrinsic desire to retain control of the …rm. Since a venture capitalist will sta¤ the …rm so that it maximizes expected return, it will …nance the …rm more cheaply if granted the power to replace the entrepreneur when needed. To an extent, the entrepreneur is willing to trade o¤ the risk of losing control against a lower cost of capital. 4.2.3 Strategic Investors Strategic investors are …rms that have more than a direct …nancial interest in the success of the entrepreneurial …rm: they may be competitors, business customers, suppliers of intermediate goods or complementary products. Such investors may be willing to lend on more favorable terms. Hellmann [81] considers the entrepreneur' preference for a strategic s investor (a …rm in a related industry). If the products are complements, 44
then the strategic investor is chosen. If the products are substitutes, then either venture capital is preferred (if they are weak substitutes) or a venture capitalist becomes the active investor and the strategic investor takes on a passive role (if products are highly substitutable). The latter arrangement occurs because the incumbent will bid to be involved in order to dilute the venture capitalist' interest in the new s …rm. Trade credits (deferred payment for shipments) are a rather expensive alternative to bank debt (implicit interest rates can be 40%-60%). Nevertheless, Huyghebaert, Van de Gucht and Van Hulle [88] …nd, for Belgian start-ups established in 1992, that trade credits are common in industries that are characterized by high failure rates and where …rm assets have high liquidation value. They argue that suppliers have a lesser incentive to collect debt and are more likely to extend additional credit (since they intrinsically bene…t from the survival of a customer). Hence they are preferred lenders for entrepreneurs who anticipate cash problems, especially when the assets have high resale value, so that ow creditors can bene…t from liquidation.
4.3
4.3.1
Security Choice
Debt vs. Equity If contracts have to be incomplete because of unforeseeable contingencies, then control matters, and the …nancial structure determines the control structure. In many respects, debt is an optimal arrangement (as compared to equity) because it makes control rights e¤ectively conditional on the performance of the …rm - assigning them to the entrepreneur as long as the …rm does not default, and assigning them to creditors in the event of default. Further advantages of debt are that it leaves incentives for the entrepreneur intact (because, subject to paying back the …xed debt, her income varies one-for-one with pro…t), and it minimizes information requirements (the investor does not need to monitor cash ows to verify that the entrepreneur adheres to the contract). A disadvantage is that debt may lead to bankruptcy when the …rm cannot repay on time, but the threat of bankruptcy can be a screening device for pro…table ventures. Equity is either voting (in which case, the entrepreneur shares control with creditors at all times) 45
or non-voting (in which case the entrepreneur retains all control rights). In either case, equity dilutes the entrepreneur' incentives, since pro…t s is shared. Krasa and Villamil [104] show that the (Pareto-) optimal equilibrium contract between entrepreneur and investor is a debt contract, if the investor cannot perfectly commit to enforcing repayment (e.g. because court action is costly). The reason is simply that the entrepreneurs' ability to repay a …xed debt (or alternatively the liquidation value of the …rm) is easiest (cheapest) to verify, hence enforcement cost is minimized, and the entrepreneur has the greatest incentive to repay as agreed. de Meza and Webb [43] argue that entrepreneurs'willingness to enter into a debt contract, rather than equity …nancing, signals to uninformed investors that they have a su ciently high pro…t expectation to cover repayment. The signal plays a role when there are entrepreneurs whose expected pro…t does not exceed the market interest rate (they are still looking for funds because bankruptcy would shift losses to the bank). While debt is the equilibrium contract under these circumstances, competition between banks still accommodates too many entrepreneurs. A reduction in interest rates attracts both high- and low-pro…t entrepreneurs, and is therefore pro…table, until successive undercutting lets banks only break even, so that some entrepreneurs are funded who are expected to default. 4.3.2 Mixed Securities The e cient …nancial structure depends on who should optimally control the …rm (and possibly this ought to change with intermediate outcomes). If the investors receive equity, then they have control of the …rm; if they get non-voting equity (preferred stock), then they have senior claims to payouts (after debt), but the control rights remain with the entrepreneur. If investors hold debt, they have control conditional on default; or, in the case of convertible debt, also in case they choose to turn their debt into equity. Convertible preferred stock may be turned into debt, in which case the investor obtains control rights in the case of default, i.e. it gives the investor priority over stock owners in claiming the liquidation value of the …rm. An option or warrant 46
allows an investor to buy additional equity in the …rm at a …xed price during a certain period (the key di¤erence is that, in case a warrant is exercised, the …rm issues new equity - diluting other owners'shares whereas an option transfers existing equity to a new owner). Aghion and Bolton [1] consider the optimal contract when the entrepreneur has non-pecuniary interests in the venture. The e cient …nancial structure trades o¤ the wish of the entrepreneur to retain control against the need to guarantee performance for the investor. It may be in the entrepreneur' interest to o¤er control rights, in order to obtain s the best combination of cost of funds and non-pecuniary bene…ts. The less the objectives of the entrepreneur and investor are compatible, the more control rights should be assigned to the investor. Voting equity gives the investor maximum control, non-voting equity gives the least, and debt or convertible securities give conditional control. Entrepreneur and investors may have a particular interest in controlling the …rm either when it performs well or badly. If strong performance implies that the entrepreneur is well-suited to continue to operate the …rm, then debt should be issued (giving the investor control only if performance is weak). If, on the other hand, the …rm is best served by replacing the entrepreneur with professional management once it grows to a certain point, then convertible debt is preferable (giving the investor control when she wants it, i.e. if performance is strong). In some cases, it may be e cient for the entrepreneur and investor to renegotiate: e.g. if the …rm is successful and the investor converts debt to equity, the entrepreneur might o¤er a payment (buy back stock) to be able to stay in charge. Because the resale value of a …rm' assets may not cover its debt, creds itors have an incentive to renegotiate their claims if the …rm cannot repay, and knowing this gives the entrepreneur adverse incentives to risk default. Berglof and von Thadden [16] show that this problem can be addressed by issuing both senior and junior claims (e.g. debt and equity). Senior creditors will not also hold junior claims, since doing so would diminish their resolve to liquidate in case of default (stock holders typically su¤er losses in bankruptcy). Their aggressiveness sets a minimum performance target for the entrepreneur. Equity and junior debt holders are more willing to defer claims, since they may 47
otherwise go empty-handed. The option to renegotiate junior claims (e.g. not paying out dividends) gives the entrepreneur a grace period in case the …rm temporarily underperforms. Senior and junior claims are complementary in that they balance performance incentives for the entrepreneur with avoidance of premature liquidations and thus provide a combination of low-cost capital and resilience to cash ow uctuations. Ravid and Spiegel [144] consider an environment where the entrepreneur can make unobserved high-risk choices after receiving …nancing. In this case, they derive that equilibrium …nancing is "linear," i.e. composed of riskless (fully collaterized) debt (the …xed component) and equity (the component that varies one-for-one with pro…ts). In other words, risky debt (which is only partially secured) is not feasible because it gives the entrepreneur limited exposure to losses (and disproportional participation in gains). The entrepreneur might then take risky actions with negative expected value. (Equity …nancing reduces the entrepreneur' share in pro…t proportionately, and thereby the ins centive for excessive risk-taking.) Bergemann and Hege [15] note that the entrepreneur' incentive to s divert capital for private consumption increases with the life of the venture, and pro…t sharing with the venture capitalist must therefore be dynamic. Diversion is a real option the entrepreneur can secretly exercise, at the cost of transmitting negative information to the venture capitalist about the performance of the …rm. Initially, such negative information can lead to early liquidation, hence the entrepreneur tends to put the capital to its best use. Later, when positive information has accumulated, a bad period has a smaller e¤ect on the venture capitalist' beliefs, and diversion becomes increasingly attractive to the entres preneur. As a result, relationship …nancing (where venture capitalists monitor and have the option to replace the entrepreneur) works better than arm' length …nancing through the market, and it is optimal for s venture capitalists to hold either a combination of debt and equity or convertible securities, giving them claims to the liquidation value of the …rm. This reduces the entrepreneur' incentive to divert by making s liquidation attractive to the venture capitalist in case of continued bad news, while leaving little for the entrepreneur to recover in that case. In the model presented by the authors, the venture has a …xed reward 48
in case of success (the timing of which is uncertain and depends on the timing of investments). To counteract moral hazard, the entrepreneur' s expected share of pro…t must diminish with the length of the venture (lest the entrepreneur prolongs it by diverting funds toward the end). Dessi [46] argues that, in an optimal collusion-proof contract between entrepreneur and venture capitalist, the intermediary holds convertible debt as well as the right to liquidate the venture. This arrangement gives the intermediary the incentive to monitor the entrepreneur and continue the venture only if it has a good chance of success. If the venture does well, then the intermediary bene…ts from continuation through the option to convert debt to equity interest (or because it already owns preferred stock); if not, then it bene…ts from liquidation through priority claims to revenues. To maximize the value of these options, the intermediary monitors diligently. 4.3.3 Staged Investment Especially venture capitalists often prefer to fund a start-up in phases: conditional on meeting intermediate performance goals (such as developing and patenting a product), the next infusion of capital is approved. Neher [130] discusses staged …nancing as a response to the entrepreneur' incentive to renegotiate debt midway through the project, since s the investor may not be able to recover her funds at this point. Through staged …nancing, the initial exposure is small (and thus the likelihood that the liquidation value of the …rm exceeds the investment greater). Over time, the …rm creates tangible assets (that are valuable independently of the entrepreneur' involvement), such as patents and explicit s strategies: these function as additional collateral that securitizes subsequent investment and reduces the entrepreneur' bargaining power. s Under stage …nancing, the entrepreneur has an incentive to practice "window dressing," i.e. to manipulate the observable performance in the short run, in order to induce further investment. Cornelli and Yosha [37] demonstrate that this problem can be solved if the investor holds convertible securities. Then, if the entrepreneur window dresses, the investor is more likely to convert debt into equity and reduce the entrepreneur' share of future pro…ts. While window dressing overstates s 49
performance, it is only possible if the …rm is doing well enough, in which case the entrepreneur wants to maintain the greatest possible stake. From the entrepreneur' point of view, the advantage of window s dressing - reducing the threat of liquidation - is therefore o¤set by the disadvantage that positive information reaches the investor. Convertibility of debt plays a similar role as an option to renegotiate with a mix of straight debt and equity. An entrepreneur may not want to continue investing in a viable venture because he bears the full cost of capital (i.e. the interest), but only some of the costs of failure (creditors share in the losses). Snyder [153] shows that the optimal solution to this problem, known as debt overhang, is a loan commitment, where the entrepreneur pays a …xed fee to the bank in exchange for the right to borrow exible amounts at a discounted interest rate in the future. This arrangement lowers the costs of continuation for the entrepreneur, while giving the bank a positive expected return. Loan commitments are common in practice.
5
Entrepreneurial Firms
We turn to the form and behavior of the entrepreneurial business, exploring issues such as governance, strategy and cooperation. And we ask which attributes distinguish survivors from failures.
5.1
Leadership and Strategy
We start at the individual level - how much time does the entrepreneur invest in the business, to what extent is decision-making power shared, and how is innovation achieved
5.1.1
Entrepreneurial Labor Supply For most goods, we expect supply to increase with price. Not so with labor. The self-employed, who have control over how much they wish to work, may well work less if their income per hour increases: with every additional dollar of wealth, the value of leisure time goes up.
50
Thornton and Eakin [158] observe that doctors reduce their work hours if fees go up, but the medical practice nevertheless provides more service by hiring additional support sta¤. Parker, Belghitar and Barmby [139] …nd that past volatility in income causes the self-employed to work more hours today, in the Panel Study of Income Dynamics (1968-1993) sample of self-employed American men. There is no e¤ect from wage itself (i.e. the self-employed do not systematically work more in response to higher income). Instead, the self-employed adjust their labor supply to insure themselves under circumstances of greater uncertainty. Brown and Snow [28] derive "price" e¤ects of an increase in risk from product price volatility. The substitution e¤ect in the labor-leisure choice is negative (more risk in production lowers expected utility from production income, makes leisure more attractive). The income effect is negative if labor supply increases in income (more risk makes the individual feel "poorer") and positive if labor supply decreases in income. Typically, the entrepreneur can improve the likelihood that the venture is pro…table by investing more e¤ort in the …rm. However, e¤ort is costly and will be weighed against the bene…t (utility) the entrepreneur derives from higher expected pro…ts. Wealth reduces the marginal bene…t of income. An implication is that entrepreneurs may work less in times when the …rm is pro…table (and they are relatively wealthy). If this e¤ect is too strong, it is optimal to sell the …rm to an external owner. De Fraja [41] models the choice of an entrepreneur, whose personal effort is an essential input, between retaining ownership of a …rm (and borrowing funds) or selling it to an investor and be employed as an agent on an incentive contract to manage the …rm. The advantage of being the owner is that the entrepreneur can reduce e¤ort in good times and sit back to enjoy the income from pro…ts. An advantage of selling the …rm is that it commits the agent to a more productive allocation of e¤ort between good and bad times, i.e. the …rm' value is greater when s it is managed under contract. The optimal employment contract involves more pro…t-sharing in good times, whereas the insurance motive 51
(investor bears most of the risk) predominates in bad times. Hence, it is best to own the …rm if its pro…tability is less sensitive to e¤ort in good times than in bad times. 5.1.2 Leadership and Delegation According to Minkler [122], whether an entrepreneur makes decisions directly or delegates decision-making to workers depends on the value of knowledge in the workforce. If workers are experts at their tasks compared to entrepreneurs (especially in the sense that they have a better idea of how to get a complex job done), monitoring is relatively costly and workers are not very averse to e¤ort (so that limited pro…t sharing produces strong incentives), then the entrepreneur will practice participatory decision making. Mukhtar [127] conducted a survey among more than 5,000 small …rms in the UK to compare managerial styles of male and female entrepreneurs. She discovered substantial di¤erences. Women are typically more "autocratic;" they prefer informal decision-making processes in which they exercise as much direct control as possible, and they tend to put personal objectives (such as asserting power) ahead of business performance. Men are more inclined to delegate and create formal procedures, especially as the …rm gets larger; they prioritize business performance. One explanation might be that women are better able to multi-task and keep "the big picture" in mind; this allows them to take more responsibility. A company culture that supports the entrepreneur' vision for the busis ness can reduce the need for monitoring. Witt [163] describes the entrepreneur as a cognitive leader in a world of boundedly rational individuals who perceive opportunities selectively. The entrepreneur' vision leads to the formation of the …rm (he anticis pates the pro…ts that make it viable to hire workers) and then focuses the employees' attention on the business model and process. There is a lesser role for governance (monitoring) in the enterpreneurial …rm because employees share the entrepreneur' vision and ignore oppors tunities to take advantage of trust. When the …rm transforms into a larger organization, high-level responsibilities must be shared and other 52
cognitive leaders may emerge who challenge the entrepreneur' vision. s Employees begin to explore alternative (including self-serving) behavior. This creates a need for stricter rules and monitoring, at the cost of sti innovative thinking. ing 5.1.3 Innovation Strategy One reason innovativeness is attractive in a venture is that it limits exposure to competition from established …rms. Such competition might also express itself in attempts to acquire the venture, which may not be in the entrepreneur' interest, if she wants to keep control. s Schwienbacher [151] takes as his starting point that venture-capital backed …rms will either be sold to another …rm or by initial public o¤ering (IPO) to shareholders (typically after 4-7 years). In the …rst case, the entrepreneur is likely to be replaced; in the second, the entrepreneur can stay in control (which he will typically prefer). A sale to another …rm is likely if the start-up is competing aggressively against an incumbent …rm' product, since the incumbent …rm then gains a des gree of monopoly power. Hence, entrepreneurs attempt to di¤erentiate excessively, so that the IPO is likely to generate more value (the author assumes acquisition by a company is less e cient, since the …rm has to be integrated into an existing organization). In a series of articles, Gi¤ord [68], [69], [70] studies the entrepreneur' s allocation of time between product innovation and improvement. The entrepreneur can, at each point in time, evaluate new products or ways to update the existing product line (restore pro…tability of an obsolete product). There are two optimal orientations. An "innovative entrepreneur" only evaluates new products and never updates. She may introduce a new product and may discard an old product if its pro…t potential is exhausted (it has become too old). The "managerial entrepreneur" prioritizes product improvement and evaluates new products only if all current products are performing. Individuals with the highest entrepreneurial talent (to identify new pro…t opportunities) relative to managerial ability (to maintain the pro…tability of ongoing operations) become innovative entrepreneurs. Those with high managerial relative to entrepreneurial ability become managerial entrepreneurs. Workers
53
come from the population with the least entrepreneurial and managerial talent. How does one search for innovative ideas, and how can the entrepreneur ascertain the viability of an idea without giving it to someone else Research by Fiet and Patel [54] suggests that entrepreneurs who look for ideas in a business area they are familiar with have a much better chance of identifying feasible ventures. They conducted an experiment where control and treatment groups received training over several months in search through, respectively, alertness (to any kind of opportunity) and thinking about market gaps within the con…nes of the individual' professional experience. The treatment group, which used s constrained search, was far more successful in generating ideas that were deemed valuable and su ciently hard to emulate. This group tended to focus on tapping existing market demand, whereas the control group tended to base its ideas on personal preferences. Biais and Perotti [19] note that several aspects of an innovative idea have to be evaluated (technical feasibility, market demand, legality, patentability). To do this, the …rm needs to work with experts, who have an opportunity to steal the idea in the process. If they steal the idea, then it will be screened incompletely, since other experts would have the same incentive to steal it, so it cannot be passed on to them for evaluation. If other experts'information is critical, then stealing is discouraged, and the entrepreneur can have the idea evaluated. Ideas that are too valuable cannot be shared and must be pursued by the entrepreneur with very little information. Gans, Hsu and Stern [62] analyze the potential for cooperation between innovators and established …rms (as is common in the biotechnology industry), compared to the alternative to go it alone and compete with incumbents (common in the electronic industry). Cooperation involves some foreclosure of competition (the …rms will want to avoid duplicating the incumbent' product and preserve market power), since the s start-up and the incumbent share the gains from trade. Three factors are predicted, and found empirically (in a commercialization strategy survey of 118 US start-ups in 1999) to improve the potential for cooperation: intellectual property protection (then the entrant can credibly 54
threaten long-lived competitive entry, without having its idea expropriated, and therefore has enough bargaining power); availability of intermediaries (e.g. venture capitalists) that make the connections; …xed costs of market entry that the start-up would have to bear if it introduces its product.
5.2
Organization
Here our interest turns to the level of the organization: formal and informal (in the sense of linkages), and also to the question of the …rm' stability. s
5.2.1
Legal Form Two choices between legal forms that the entrepreneur has are between limited liability (e.g. corporations) and unlimited liability (e.g. sole proprietorships, partnerships), and between for-pro…t and not-for-pro…t status. Chamley [32] points out that the limited liability form allows entrepreneurs of lesser ability to share some risk, and ability is therefore revealed to investors through the choice of liability form (more able entrepreneurs choose unlimited liability). This is an e ciency improvement over a world where only the unlimited liability form is possible. Glaeser and Shleifer [72] explain why completely self-interested entrepreneurs may start not-for-pro…t …rms. These …rms are restricted to distributing their pro…ts in non-monetary bene…ts (less e¤ort, larger o ces, etc.) instead of income. A lesser incentive to maximize pro…ts can garner the trust of consumers in markets where …rms make unobserved quality choices (e.g. the quality of teachers hired in education or child care, the e¤ort invested into identifying the best uses of funds in charitable organizations). Hence the not-for-pro…t status can give …rms a competitive advantage. If the demand e¤ect from this commitment exceeds the disadvantages from the non-monetary receipt of pro…ts, not-for-pro…t status is chosen.
55
5.2.2
Networks Empirically, we observe both local entrepreneurship (people operating start-ups in their home region) and agglomeration of entrepreneurs in particular urban areas (people moving to where the "action" is). Both are indicative of the importance of networks: in one case, friends and relatives; in the other, informal or organized professional contacts and links with complementary industry. Michelacci and Silva [120] observe that the fraction of entrepreneurs who work in the region where they were born is signi…cantly higher than the corresponding fraction of salaried workers. This e¤ect is strongest in regions that are well-developed and have a good …nancial infrastructure; moreover, local-operated …rms are larger, better capitalized and have better access to external funding. Thus, local roots seem to enhance entrepreneurial prospects, particularly through access to …nance. (Presumably, the access comes through the personal network, but rootedness may also make it costlier for the entrepreneur to move, so that the costs of failure are greater.) Support is found in US census data from 2000 as well as Italian data rom the Survey of Household Income and Wealth (1991, 1993 and 1995 waves). Minniti [123] argues that entrepreneurs respond to opportunities because they have a competitive advantage in exploiting them, but they also incur uncertainty that they do not feel better equipped to deal with than others. Therefore, they derive a bene…t from imitating the behavior of other entrepreneurs (who presumably have made these choices based on knowledge or observation of others). In areas with a high concentration of entrepreneurs, this minimizes uncertainty and makes entrepreneurship least costly. The logic implies increasing returns to scale in, and predicts agglomeration of, entrepreneurial activity.
5.2.3
Spin-O¤s The entrepreneurial …rm' business model and culture becomes familiar s to employees, especially after they are given signi…cant responsibility. Preventing harmful spin-o¤s that compete with the mother …rm is a critical challenge for many ventures, but ultimately spin-o¤s transfer
56
successful practices and make innovative products available to more consumers. Ziegler [166] discusses the experience of an innovative venture that spawned numerous spin-o¤s founded by former employees and co-owners. He argues that spin-o¤s occur when the entrepreneur institutionalizes his integrative role between production and marketing by appointing managers who deal with both aspects. These managers acquire a wellrounded knowledge of what makes the venture feasible and unique: they absorb the entrepreneurial culture and gain access to the network of bankers, suppliers and customers. This exposure enables them to play the role of the entrepreneur and as a result strengthens their bargaining positions vis a vis the entrepreneur. Unless the entrepreneur is willing to essentially make them partners in the …rm, they have an incentive to leave. They will take essential employees with them, by convincing them that they do not share su ciently in the venture' pro…tability, s and this can eventually harm the original venture' competitiveness and s even force it out of business. Arend [3] considers how established …rms try to prevent employees from forming spin-o¤s. Incumbent …rms may not wish to adopt an innovation because it would cannibalize their existing products. But this is not a consideration for a potential defector, who therefore has a higher valuation for the innovation. To discourage a spin-o¤ (if it is worth preventing from the incumbent …rm' point of view), the employer makes s side-payments to employees or pursues the innovation anyway, in order to deter the spin-o¤. Spin-o¤s are themselves are in themselves a form of entrepreneurship that serves the purpose of reallocating resources (in this case, the mother …rm' former employees'ideas and skills) to socially desirable s projects, for which there is a demand (but in which the mother …rm does not necessarily have a private interest). Holmes and Schmitz [82] interpret business transfers (…rm sales, employee movements) as a way to achieve better division of labor. (I.e. entrepreneurs move on to a more productive activity that they have recently become aware of.)
57
5.3
Survival Factors
Which characteristics of …rms or their owners account for medium-term survival
5.3.1
Objectives Some people, so-called opportunity entrepreneurs, set up businesses because they believe in the commercial potential of an idea; others, necessity entrepreneurs, start a business (often from unemployment) to apply themselves somewhere, even in an activity that is expected to yield little more than subsistence. In addition, some choose selfemployment for lifestyle reasons (needing exible work hours). Not surprisingly, it is the …rst type of venture that is more likely to survive, but when the latter go out of business, it may also be voluntary (the formerly unemployed entrepreneur got a new job; the child has grown up and the parent can rejoin the workforce ...). Reid [146] emphasizes that entrepreneurs need to make many critical choices simultaneously, which distinguishes them from managers of established …rms, who mainly focus on supply decisions. In a sample of 150 randomly selected small …rms in Scotland, he …nds that survival over a three-year time horizon (1994-1997) depends on an entrepreneur who has a business vision (sacri…cing quick pro…t for sustained growth) rather than a personal objective (alternative to unemployment, getting rich, being one' own boss), retires debt early and operates on a s exible scale (many employees, but not all on full-time contracts). Taylor [157] obtains evidence for the UK that many exits are not to bankruptcy, but to employment (this was especially true for individuals with a prior history of unemployment). Those who survive in business typically have little or no unemployment history and some work experience and capital. The non-survivors appear to be resourceconstrained and temporarily forced into self-employment ("necessity entrepreneurs"). He uses data from waves 1 to 5 (1991 to 1995) of the British Household Panel Survey (BHPS), which is nationally representative, including 5,500 households and 10,000 individuals. Headd [78] refers to the 1996 US Census Bureau' survey of Characters istics of Business Owners (CBO), in which owners of closed …rms were 58
asked whether their business was "successful at closure." He …nds that a third of all …rms going out of business are successful at closure and that successful closed …rms have similar characteristics to unsuccessful closed …rms (little or no start-up capital, small size, inexperienced owner). He concludes that some "failures" occur because entrepreneurs execute rational exit strategies. Although …rms generally need to be focused on performance in order to survive, they do not have to be strict pro…t maximizers. The following suggests that pro…t-maximizing …rms may be the exception, rather than the norm, in reality. Dutta and Radner [48] show that pro…t-maximizing …rms fail in …nite time: due to diminishing marginal return to capital, they only accumulate up to a …nite level of capital and, over short or long, a run of losses will drive working capital down to zero, i.e. the …rm goes bankrupt. Non-pro…t-maximizing …rms (that allow the capital stock to grow the above e cient level) have positive probability of not failing and can get …nanced in equilibrium (provided the required interest rate is smaller than the highest potential return, which is given by pro…t maximization). Ultimately only non pro…t-maximizing …rms are expected to survive. 5.3.2 Resources Human capital, encompassing formal education, work experience in the business, and the experience of growing up in an entrepreneurial household, is a robust predictor of better survival chances. It may also, to a large extent, determine access to …nancial resources. Holtz-Eakin, Joulfaian and Rosen [84] …nd that survival rates increase with owner wealth, presumably because businesses cannot operate at the optimal scale without su cient capital. In 1981 and 1985 US income tax returns, the authors …nd that receipt of a $150,000 inheritance in 1982 and 1983, conditional on survival, increases the business income of a sole proprietor by almost 20%. Cressy [39] studies 2,000 entrepreneurs who opened business accounts with the National Westminster Bank in the UK in 1988. He considers the …rm to have survived if the account was still open in 1992. 59
In the author' analysis, debt-…nancing is "spuriously" correlated with s survival because entrepreneurs with more human capital (work experience, education, quali…cations) borrow more money, and also have a greater probability to survive (through their skill). Hence human capital is the primary determinant of survival. Similarly, Bates [10] observes, in a random sample of American men, who entered self-employment between 1976 and 1982, that the highly educated are more likely to create …rms that survive (through 1986), and are also likely to attract more …nancial capital (although larger investments also increase the probability of survival independently). van Praag and Cramer [159] test a model in which equilibrium …rm size increases in the entrepreneur' talent for a Dutch population surveyed s in 1952, 1983 and 1993 about background factors and labor market outcomes. The talent implicit in observed …rm sizes was then related to background variables such as parents' education and careers, own education, childhood IQ and risk aversion. Coming from an entrepreneurial family, having a strong education and IQ and low risk aversion all predict creating a relatively large entrepreneurial …rm. Lentz and Laband [108] …nd that start-ups run by second-generation proprietors (whose parents had a family business) are signi…cantly more likely to be successful. In the US, roughly 50% of all self-employed proprietors are second-generation. They were found to start their businesses at a signi…cantly younger age, on average, than …rst-generation proprietors, but with more relevant experience (their informal home training substitutes for formal education). These observations are based on a sample of 514 self-employed proprietors drawn randomly from the 1979 members of the National Federation of Independent Businesses (NFIB). 5.3.3 Technology By selling an innovative product, entrepreneurs can avoid tough competition and gain some time to grow competitive. Nerkar and Shane [131] analyze a dataset of 128 …rms founded to commercialize technologies licensed from MIT between 1980 and 1996. 60
They …nd that the introduction of radical technologies (measured by the count of patent class citations outside of the technology' own class) s reduces probability of failure, but was more successful in industries that were fragmented, rather than concentrated. Concentration may make it more di cult to …nd strategic partners (since the incumbents'incentive tends to be to keep the entrant out). It also brings with it the need to compete against large-scale marketing activities and distribution infrastructure, and is an indication of economies of scale (which makes small-scale entry unpro…table and demands rapid growth).
6
Public Policy
The …nal lecture discusses how the institutional environment and government policy a¤ect the supply of entrepreneurship.
6.1
6.1.1
Institutions
Employee Bene…ts In exiting salaried employment and entering self-employment, entrepreneurs may have to give up bene…ts such as discounted family health insurance. The existence of such bene…ts creates an additional opportunity cost that should bias people against becoming entrepreneurs. In support of this argument, Wellington [160] …nds that people are more likely to move into self-employment if they have access to health insurance through a spouse. Based on data from the 1993 Current Population Survey, spousal health insurance raises the probability that the partner becomes self-employed by up to 5%, and universal health insurance coverage would increase the fraction of entrepreneurs in the US workforce by 2% to 3.5%.
6.1.2
Redistribution A government policy to redistribute from the wealthy to the poor limits the potential pro…t from successful entrepreneurial ventures. But it can also encourage entrepreneurship by providing capital to talented
61
but wealth-constrained individuals, as well as a form of insurance to the risk-averse. Kanbur [94] points out that there is no inherent con between tarict geting equity in the income distribution and encouraging entrepreneurship. Although progressive taxation limits the pro…ts of successful entrepreneurs, it also limits the losses, thus provides insurance to the risk-averse and attracts newcomers with lower expected returns into self-employment. This adjustment reduces the average incomes of entrepreneurs, which are higher than wage incomes by the risk premium entrepreneurs require. Therefore, progressive taxation can reduce income inequality while also increasing the number of entrepreneurs. Gruener [74] demonstrates how (ex ante) redistribution of wealth can be Pareto-improving for a population di¤ering in abilities that are private knowledge, by creating more equal opportunities. Without redistribution, entrepreneurs self-select primarily on the basis of wealth, rather than ability; raising capital through the market is expensive for those who cannot eliminate moral hazard by taking a substantial position in their own venture. With redistribution, access to reasonably priced capital broadens, and the most talented individuals become entrepreneurs, while the less talented lend their money to the entrepreneurs. Then, the expected return to capital increases since entrepreneurs are of higher quality. At the same time, entrepreneurs can borrow at lower cost, because self-selection on the basis of ability implies endogenously better information about the success probabilities of entrepreneurial ventures. Hence even wealthy individuals, who initially have to make transfers to others, may bene…t: either by lending at high rate of return (if they have relatively low ability and thus choose not to be entrepreneurs after redistribution) or by borrowing at lower cost as entrepreneurs. Garcia-Penalosa and Wen [63] show that (ex post) redistribution can increase the number of entrepreneurs and improve welfare (make everyone better o¤) in a risk-averse population. Redistribution provides insurance to entrepreneurs by reducing potential losses and gains. Skilled employees have to pay a greater fraction of their incomes in taxes under a progressive tax system, but if su ciently many of them switch to entrepreneurship, skilled wages increase and potentially more than o¤set the tax increase. Since redistribution certainly makes entrepreneurship 62
more attractive, it leads to more innovation and growth. Income inequality can decrease or increase, depending on whether the primary redistribution e¤ect (which raises the incomes of the unskilled) or the induced wage increase for the skilled dominates, i.e. whether there is enough redistribution. An alternative view is explored by Meh [117], who considers the e¤ect of switching from progressive to proportional taxation. Although lower marginal taxes on the incomes of the wealthy encourage greater productive activity and therefore lead them to accumulate more wealth, there is an opposite secondary e¤ect on the income distribution, which arises from the entrepreneurial role. Given capital constraints, which make the scale of businesses dependent on the entrepreneur' personal s wealth, business owners respond by expanding their operations and hiring more people, thus increasing competition for labor and ultimately the wages of the employed. In the author' calibration, if proportional s taxes were introduced in the US, the direct e¤ect on income inequality through wealth accumulation would be approximately o¤set by the indirect e¤ect through demand for labor. 6.1.3 Legal Infrastructure Rule of law makes contracts enforceable and therefore facilitates economic relationships over time, such as borrowing or supply of intermediate goods (in principle, the supplier is exposed to "hold ups," renogitiation of terms after irreversible investments have been made). Since entrepreneurial ventures require investments (…nancial and otherwise), they depend on being able to make binding commitments that are legally guaranteed. Transition economies are unique in that deregulation can open up sudden and and very pro…table opportunities that are accessible to many individuals. Corruption then often emerges as a price mechanism that allocates these opportunities to entrepreneurs. As Johnson, McMillan and Woodru¤ [91] report, corruption has a signi…cant dampening effect on entrepreneurial activity. In the transition economies in their study with the most corruption (Russia and Ukraine), entrepreneurs reinvested only half the fraction of earnings (about 30%) that their
63
counterparts put back into their businesses where bribes were least common (Poland and Romania). McMillan and Woodru¤ [116] discuss how entrepreneurs cope in environments where property rights are weak and courts provide limited enforcement. They stress reliance on repeated interaction as an informal mechanism to provide incentives to pay debts to suppliers. 6.1.4 Financial System The …nancial system allocates investment funds to the most valuable projects. Under ideal conditions, it will select the most promising entrepreneurial ventures and provide the required capital to them. If the …nancial system does not ful…ll this task, then the wealthiest, rather than the most talented, individuals become entrepreneurs, and poorer prospective entrepreneurs must defer entry and save …rst, or enter on an ine ciently small scale. Entrepreneurship increases variety in …nancial markets - it gives investors the opportunity to diversify their portfolios, which is valuable in itself. Since this is a positive externality, entrepreneurship may be undersupplied. On the other hand, entrepreneurship is to some extent associated with experimentation, so it may divert resources to activities that turn out to be less productive. Because of the social demand for investment alternatives and overly optimistic assessment of risks by individuals, there may be overexperimentation. King and Levine [103] argue that a …nancial system ful…lls several important roles in facilitating entrepreneurship. Besides supplying funds, it evaluates proposed ventures and screens out the ones that are least likely to succeed (intermediaries can do this more e ciently than individual investors). It attracts capital by o¤ering risk-averse investors the opportunity to hold diversi…ed portfolios. And the prices of …nancial assets re expected future pro…t ect ows, hence signal to society the rewards to innovation. The authors cite empirical evidence, ranging from case studies to panel estimation, that the …nancial system acts on economic growth through its e¤ect on productivity. I.e. countries with better …nancial systems are wealthier because they allocate
64
their resources more e ciently, in particular to promising innovative ventures. Pagano [136] discusses the role of entrepreneurial stock issues in providing risk-sharing opportunities to the market. This is a positive externality, resulting in too few stock issues, especially in countries with small stock markets. A small stock market may in turn make entrepreneurial activities less attractive, since it is di cult to diversify the associated risk away. Hence potential entrepreneurs in developing countries face a two-fold disincentive: the private bene…t is smaller than the social bene…t, and it is di cult to spread risk. If entrepreneurs' otation decisions are correlated (as in the case of credit constraints for purchasing shares - so that fewer issues imply less wealth available for stock purchases, hence less demand for another issue - or with costly issue), then countries can be trapped in small-market or large-market equilibria. Michelacci and Suarez [121] point to the importance of scarce "informed capital." This refers to intermediaries specialized in monitoring entrepreneurial start-ups before enough information about their pro…tability becomes available publically for regular investors to step in and fund them. Reducing costs for start-ups to go public (allowing entrepreneurs to raise funds through equity earlier) frees up informed capital that becomes available for new start-ups, thus increases entrepreneurial activity. A recent innovation that makes capital accessible to the poor with entrepreneurial aspirations are microcredit programs, such as the Grameen Bank in Bangladesh. Such programs lend to groups of (typically about …ve) individuals who are jointly liable for the loans and therefore have an incentive to screen and monitor each other. They also bundle training services with the loans. McKernan [115] …nds from a 1991-1992 survey of 1,798 rural Bangladeshi households that borrowers at Grameen Bank had higher-than-average entrepreneurial pro…ts and repayment rates. An approximate 175% increase in self-employment income is attributable to the provision of credit by three programs, and a further 125% increase to the provision of non-credit educational services (the author isolates this e¤ect by controlling for capital). More functional capital markets level the playing …eld for prospective 65
entrepreneurs by reducing dependence on private …nancial resources. There are winners and losers in this development. Gine and Townsend [71] calibrate a model with capital constraints to a sample of Thais in their twenties from the 1976-1996 Socio-Economic Survey (SES), as well as a 1997 survey of about 3,000 households in two Thai provinces. From the …nancial liberalization that occurred over this period, talented but poor would-be entrepreneurs gained and wealthy incumbent entrepreneurs lost, as steadily rising wages eroded their formerly high pro…ts.
6.2
6.2.1
Regulation
Market Imperfections and Entry Barriers to entry may lead to an insu cient amount of entrepreneurship from a social point of view. But powerful incumbents do not always hamper entrepreneurial entry. Such …rms have incentives to extract rents to the fullest extent possible and therefore place consumers on the threshold to "defection," i.e. create demand for entrants. They may also open up pro…table markets for suppliers and other related businesses. Ekelund, Hebert and Tollison [49] discuss the rise of the Protestant church as an entrepreneurial entrant into the "spiritual services" industry, formerly monopolized by the Catholic church. This occurred as a result of the Catholic church' attempts to extract maximal rents s from its monopoly, by introducing innovative prohibitions and costly exemptions, while price-discriminating between the wealthy and the poor. Examples are found in the sale of indulgences (o¤ering a tradeo¤ between a "sinful" lifestyle and expensive absolution) or matrimonial dispensations (o¤ering nobility special permissions to practice "endogamy," i.e. marriage to kin). Protestantism o¤ered "cheaper" direct salvation by faith, eliminating some of these rent-seeking opportunities. It emerged initially in urban areas and regions not characterized by the custom of primogeniture (eldest son is the sole heir), to which Catholicism was in some ways complementary (by providing alternative careers for younger siblings in the nobility). I.e. Protestantism sprung
66
up where wealth was spreading to larger portions of the population, who had an interest in escaping price discrimination. Large discount retailers are often criticized for destroying "mom and pop" stores, and there is clear evidence that small retail competitors close down in large numbers where Walmart stores open. But Sobel and Dean [154] call attention to the Schumpeterian prediction that the emergence of a new retail format like Walmart' not only drives some s retailers out; it should also create markets for new start-ups. They collect data for all continental US states on the number of Walmarts.as well as self-employment rates, number of small businesses and employment share of small retail shops (one to four and …ve to nine employees). Their innovation is to include in the analysis small businesses of all kinds, and state- and nationwide, not just local retailers of similar products. Overall, the authors …nd no support for the claim that Walmart signi…cantly a¤ects the prevalence small businesses in the US. Entrepreneurs contribute to the emergence of competitive markets in which prices re costs and e cient production choices result. Entry ect is facilitated by low barriers and costs for start-ups as well as antitrust regulations that prevent incumbents from monopolizing markets. Djankov et al. [47] inquire into the motives of entry regulation through the observed quality of government services. In data on 85 countries, compiled from publications by governments and international organizations, they identify stricter entry regulations with provision of inferior public goods (health), more negative externalities (pollution), weaker competition, a larger uno cial economy, greater corruption and less democratic political institutions. The extent of regulation is assessed on several dimensions: number of procedures necessary to register a new start-up (ranging from 2 in Canada to 21 in the Dominican Republic), time required (from 2 business days in Australia and Canada to 152 in Madagascar), cost incurred in fees (from less than 0.5% of GDP per capita in the US to close to 500% of GDP per capita in the Dominican Republic). In most countries, it is by all measures very expensive to start a new …rm, and the authors conclude that regulation primarily tends to bene…t the regulators, rather than the public, in accordance with the public choice perspective that policymakers are rent-seekers.
67
Not surprisingly, Fonseca, Lopez-Garcia and Pissarides [55] argue, and …nd support in basic correlations for the EU, US and Japan, that higher start-up costs translate into diminished job creation and employment via reduced entrepreneurial activity. 6.2.2 Bankruptcy An important factor for whether entrepreneurship is worthwhile, given the substantial risk that is attached, is the extent to which personal assets are protected from any losses in businesses. I.e. more lenient bankruptcy provisions encourage entrepreneurship. One needs to keep a mind, however, that a moral hazard issue arises from the bankruptcy option: entrepreneurs may fully realize pro…ts, but not be fully responsible for losses. Hence, entrepreneurs privately face risks that di¤er from the risks to society, leading to rational overexperimentation. When an entrepreneur …les for bankruptcy in the US, he loses all business assets and, if the …rm is not incorporated, all personal assets beyond a …xed exemption that varies by state. Fan and White [53] take advantage of this variance in exemption levels to test how lenient bankruptcy laws a¤ect the prevalence of entrepreneurship. Households are more likely by 35% to own businesses in states where all personal assets are exempt, compared to states where the exemption is low. The positive e¤ect of leniency on entrepreneurial activity is corroborated at the country level by Armour and Cumming [4], who observe that entrepreneurial demand for venture capital and private equity in 15 Western European and North American countries between 1990 and 2003 decreases in the number of years before individuals can make a fresh start after bankruptcy. Their data from venture capital association databases and other sources also show that venture capital supply increases in an index of a country' "investor friendliness" (a combinas tion of legal and …scal attributes) and is crowded out by government programs that provide publically funded venture capital. Zazzaro [165] demonstrates that mandating a stricter accounting standard (which simpli…es auditing of entrepreneurs) improves e ciency, but strengthening contract enforcement could be ine cient, since it causes intermediaries to invest e¤ort into identifying and attracting 68
better entrepreneurs, which lowers the average quality of entrepreneurs while increasing their numbers. In many countries, particularly in Asia, bankruptcy proceedings are lengthy and tedious, and creditors ultimately do not recover much of the money they have lent (i.e. entrepreneurs can "tunnel" funds out of the …rm). Yet, Friedman, Johnson and Mitton [59] observe that the debt to assets ratio is especially high in these countries. They argue that entrepreneurs have an incentive to "prop" by injecting their own wealth into the …rm when it has di culty repaying, since they would otherwise be unable to borrow in the future (investors are extremely wary of past defaulters, given the di culties in recovering debt). The expectation of propping makes borrowing feasible in the presence of weak bankruptcy laws. Although entrepreneurs prop in response to a moderate bad shock, they take advantage of the opportunity to tunnel (divert wealth to other businesses they own) when the shock is severe. The authors apply this logic to explain the Asian …nancial crisis of the late 1990s.
6.3
6.3.1
Fiscal Policy
Public Spending A large public sector, which is to say a large share of government spending relative to the total size of the economy, is often found to come at the expense of entrepreneurship. A main reason is that large bureaucracies tend to create more red tape, and therefore greater costs, for start-ups. A more subtle reason has to do with the di¤erences between public and private demand. Private demand, which is driven by personal tastes, tends to change faster and create markets for innovation; public demand is less exible, therefore faces higher prices and attracts resources away from potential entrepreneurs. When the government is a major spender in the economy, the price level increases and private demand is crowded out. Takii [156] points out that, if government demand is static and private demand is dynamic, …rms respond to …scal expansion by investing less in "entrepreneurship," which is understood here (in an Austrian sense) as the prediction of changing demand patterns. Fiscal expansion can even 69
result in a decline in real GDP because it comes at the expense of a contraction in the more competitive private sector. 6.3.2 Taxes A progressive income tax schedule makes entrepreneurship less attractive (relative to a proportional tax) because success is e¤ectively "punished" (through the higher marginal tax rate). In general, tax differentials between localities, sources of income, etc. distort incentives for entrepreneurs and harm e ciency, unless they correct an existing e ciency. Papke [137] shows, in a panel study of …rm births across 22 manufacturing industries and …ve US states from 1975 to 1982, that entrepreneurial entry (plant building) is deterred by high marginal tax rates on business income after controlling for industry and state e¤ects. The analysis reveals that entry into di¤erent industries responds to di¤erent state characteristics, hence accounting for this heterogeneity is important in order to isolate the e¤ects of tax rates. Gentry and Hubbard [64] demonstrate on a 1979-1992 sample from the Panel Study of Income Dynamics (PSID), which covers exposures to di¤erent tax regimes, that proportional taxation discourages entrepreneurial entry. Hence, it appears that the insurance that is implicit in progressive taxation, which should appeal to risk-averse individuals, is not valued by prospective entrepreneurs. Robson and Wren [148] study the di¤erent e¤ects of marginal and average taxes. They conclude from panel data from 15 OECD countries (and observations in 3-4 year intervals between 1978 and1992) that entrepreneurial activity decreases in the marginal tax rate and increases in the average tax rate. They argue that the former acts on the incentive for e¤ort and the latter on the incentive to evade. This ows from the plausible assumption that the return to e¤ort and opportunities to evade are greater in self-employment than in salaried employment. Blau [23] observes that the non-agricultural self-employment rate has trended up again in the US, Western Europe and Japan since the 1970s, after a long decline. From annual time series data on US men, extracted 70
from the 1948-1982 Current Population Surveys, he attributes a role to rising marginal tax rates that favored self-employment, perhaps because self-empolyment presents greater opportunities to underreport. However, the primary explanation appeared to be that the industry pro…le shifted toward industries that are no longer characterized by substantial economies of scale, so that smaller …rms became more competitive (and entry thus had a greater chance of succeeding). A third factor were changes in the labor market: a rise in lay-o¤s, on the one hand, and a rise in demand for more exible work hours, on the other, led to involuntary and voluntary exists from salaried employment. Long [111] con…rms that higher marginal tax rates increase the selfemployment rate, using data on 25-64 year-old men (working in nonagricultural professions) from the 1970 US Census of Population. He points out that occupational choices will tend to equalize after-tax earnings in wage- and self-employment, o¤setting the e¤ective tax di¤erential that arises from evasion opportunities in self-employment. This is another source of ine ciency introduced by the income tax: not only do individuals reduce productive e¤orts (since they are only partially rewarded for them), but they also (in some cases) become self-employed even though they would be more productivity in the wage sector. Boadway, Marchand and Pestieau [24] derive the optimal linear tax (in income) for various scenarios, involving heterogeneity in ability and risk aversion, taking account of the fact that the tax system a¤ects not only e ciency and equity, but also introduces distortions into occupational choice. Keuschnigg [98] and Keuschnigg and Nielsen [100] argue that the capital gains tax is particularly harmful to e ciency because it exacerbates the "double moral hazard" problem between entrepreneurs and venture capitalists: the success of the start-up depends both on unobservable entrepeneurial e¤ort and on unobservable e¤ort of the venture capitalist in its advisory role. In order to resolve the moral hazard, each side needs to share as much as possible in marginal pro…t, but the capital gains tax takes a further cut out of any additional dollar earned, resulting in a large welfare loss. However, if time spent advising a failing …rm is tax deductible, as in [99], then an increase in the capital gains tax may encourage venture capital e¤ort. 71
de Meza and Webb [44], [45] argue that an interest rate tax, which increases the cost of borrowing for entrepreneurs, may improve e ciency in the presence of asymmetric information in the credit market. The pooling of entrepreneurs under the same borrowing terms e¤ectively subsidizes those with the lowest expected project returns; a tax can o¤set the "subsidy" and cause the least pro…table projects to be dropped or reduced in scale. Cullen and Gordon [40] point out that taxation a¤ects not only the supply of entrepreneurship, but also the amount of risk entrepreneurs take. Because business losses can be o¤set in the calculation of the personal income tax, the relationship between the personal and corporate income tax determines the relative impact of entrepreneurial losses and gains. Based on 1964-1993 tax return data, the authors predict that a 5% reduction in the personal income tax (without a corresponding reduction in the corporate tax rate) would reduce risk-taking by about 40%. Conversely, implementing negative taxes (i.e. the government reimburses net losses in annual income) would increase risk-taking. 6.3.3 Tari¤s Free trade can hinder the emergence of local entrepreneurs. This is true for imports as well as exports; exports may allocate the economy' s resources too narrowly on a few industries. In developing countries, Hausmann and Rodrik [77] argue, it is the role of entrepreneurs to discover the local cost of entering modern industries - in particular, whether a particular product can be produced at lower cost locally than in the world market. Since such discoveries are not protected by patents, they are easily emulated and therefore provide the entrepreneur with an ine ciently low expected return. Experimentation is underprovided, and when it happens entry is overly concentrated in the successful industries. The authors argue that the Asian miracle can partly be understood by barriers to entry and effective monopoly guarantees, which allowed some enterprises, such as the large Korean and Japanese business groups, to explore costs widely and fully realize the country' comparative advantage. By contrast, in s Latin America deregulation may have sti discovery because future ed competition would quickly erode the original entrepreneurs'returns. 72
Grossman [73] shows that trade and foreign direct investment diminish the domestic supply of entrepreneurship in an open developing economy; however, neither reduces the country' welfare in an appropris ate sense. Gains from trade can, in principle, be redistributed from bene…ciaries (landlords) to fully compensate losers (entrepreneurs and workers), while foreign investment lowers national income, but leaves the population' expected utilities unchanged. Protectionist policies s (import tari¤s or taxes on agricultural output to encourage industrial production at home) can lead to more entrepreneurship, but will reduce welfare. In order to increase entrepreneurial activity and welfare, it would be necessary to insure potential entrepreneurs against losses from industrial ventures (something trade policy instruments do not achieve). Holmes and Schmitz [83] explain how lowering tari¤s increases the incentive of domestic entrepreneurs to engage in productive (research) rather than unproductive activities (seeking regulatory protection from competitors). With lower tari¤s, imports from foreign …rms (that are not subject to regulation) pose a greater threat, and entrepeneurs need to focus on maintaining competitiveness; keeping domestic rivals is of little value. Moreover, if tari¤s are also lowered abroad, competitiveness pays o¤ even more through exports. 6.3.4 Credit Subsidies Through interest rate subsidies, the goverment can o¤set some of the costs that accrue to entrepreneurs due to asymmetric information (if entrepreneurs can judge the viability of their venture better than the lender), which increases their borrowing costs. O¤ering targeted government subsidies to some entrepreneurs (e.g. in certain industries) may raise unsubsidized interest rates and crowd out other entrepreneurs. In a model proposed by Innes [90], banks cannot distinguish between high- and low-ability entrepreneurs, but since the former can make better use of funds, they are able to signal their type, and get better …nancing terms, by borrowing excessively. In order to avoid this ine ciency, the government can o¤er a subsidized credit that meets the needs of low-ability entrepreneurs, who therefore have no incentive 73
to go to the private credit market. As a result, only high-ability entrepreneurs borrow from banks; they can be identi…ed and o¤ered the e cient loan amount at prices that re their default risk. ect Fuest and Tillessen [61] examine why credit subsidies are often for a …xed loan amount (closed-end subsidies), rather than proportional to the full amount of the loan the entrepreneur wishes to borrow (openended subsidies), which would a¤ect investment decisions at the margin. In their model of the credit market, without subsidies, high-ability entrepreneurs overinvest in order to dinstinguish themselves from lowability entrepreneurs, whose investments are less productive and ideally small-scale, from mimicking their behavior. The goal of a subsidy is to reduce the overinvestment required for separation by making lowvolume loans (that carry higher interest rates, since only low-ability entrepreneurs are expected to apply for them) cheaper. A closedended subsidy discounts only the loan amount low-ability entrepreneurs plan to borrow, and therefore makes mimicking less attractive than an open-ended subsidy, which also discounts additional borrowing, would. Hence closed-ended substitutes achieve type separation with less overinvestment by high-ability entrepreneurs than open-ended subsidies can.
References
[1] Aghion, P.; Bolton, P. (1992). An Incomplete Contracts Approach to Financial Contracting. Review of Economic Studies, v. 59, iss. 3, pp. 473-94. [2] Arabsheibani, G; de Meza, D.; Maloney, J.; Pearson, B. (2000). And a Vision Appeared unto Them of a Great Pro…t: Evidence of SelfDeception among the Self-Employed. Economics Letters, v. 67, iss. 1, pp. 35-41. [3] Arend, R. J. (2001). When Rivals Originate from Within. Small Business Economics, v. 16, iss. 3, pp. 205-22. [4] Armour, J.; Cumming, D. (2006). The Legislative Road to Silicon Valley. Oxford Economic Papers, v. 58, iss. 4, pp. 596-635.
74
[5] Audretsch, D. B. (2001). The Role of Small Firms in U.S. Biotechnology Clusters. Small Business Economics, v. 17, iss. 1-2, pp. 3-15. [6] Audretsch, D. B.; Jin, J. (1994). A Reconciliation of the Unemployment– New Firm Startup Paradox. Small Business Economics, v. 6, iss. 5, pp. 381-85. [7] Banerjee, A.; Newman, A. (1993). Occupational Choice and the Process of Economic Development. Journal of Political Economy, v. 707, iss. 2, pp. 274-98. [8] Barzel, Y. (1987). The Entrepreneur' Reward for Self-policing. Ecos nomic Inquiry, v. 25, iss. 1, pp. 103-16. [9] Basu, A.; Parker, S. C. (2001). Family Finance and New Business StartUps. Oxford Bulletin of Economics and Statistics, v. 63, iss. 3, pp. 333-58. [10] Bates, T. (1990). Entrepreneur Human Capital Inputs and Small Business Longevity. Review of Economics and Statistics, v. 72, iss. 4, pp. 551-59. [11] Bates, T. (1999). Exiting Self-Employment: An Analysis of Asian Immigrant-Owned Small Businesses. Small Business Economics, v. 13, iss. 3, pp. 171-83. [12] Baumol, W. J. (1968) Entrepreneurship in Economic Theory. American Economic Review Papers and Proceedings, v. 58, iss. 2, pp. 64-71. [13] Baumol, W. J. (1990). Entrepreneurship: Productive, Unproductive, and Destructive. Journal of Political Economy (pt. 1), v. 98, iss. 5, pp. 893-921. [14] Benz, M.; Frey, B. S. (2008). Being Independent Is a Great Thing: Subjective Evaluations of Self-Employment and Hierarchy. Economica, v. 75, iss. 298, pp. 362-83. [15] Bergemann, D.; Hege, U. (1998). Venture Capital Financing, Moral Hazard, and Learning. Journal of Banking and Finance, v. 22, iss. 6-8, pp. 703-35.
75
[16] Berglof, E.; von Thadden, E.-L. (1994). Short-Term versus Long-Term Interests: Capital Structure with Multiple Investors. Quarterly Journal of Economics, v. 109, iss. 4, pp. 1055-84. [17] Berkowitz, D.; Jackson, J. E. (2006). Entrepreneurship and the Evolution of Income Distributions in Poland and Russia. Journal of Comparative Economics, v. 34, iss. 2, pp. 338-56. [18] Bernanke, B.; Gertler, M. (1989). Agency Costs, Net Worth and Business Fluctuations. American Economic Review, v. 79, iss. 1, pp. 14-31. [19] Biais, B.; Perotti, E. (2008). Entrepreneurs and New Ideas. Rand Journal of Economics, v. 39, iss. 4, pp. 1105-25. [20] Black, J.; de Meza, D.; Je¤reys, D. (1996). House Price, the Supply of Collateral and the Enterprise Economy. Economic Journal, v. 106, iss. 434, pp. 60-75. [21] Blanch ower, D. G.; Oswald, A. J. (1998). What Makes an Entrepreneur Journal of Labor Economics, v. 16, iss. 1, pp. 26-60. [22] Blanch ower, D. G.; Oswald, A.; Stutzer, A. (2001). Latent Entrepreneurship across Nations. European Economic Review, v. 45, iss. 4-6, pp. 680-91. [23] Blau, D. M. (1987). A Time-Series Analysis of Self-Employment in the United States. Journal of Political Economy, v. 95, iss. 3, pp. 445-67. [24] Boadway, R.; Marchand, M.; Pestieau, P. (1991). Optimal Linear Income Taxation in Models with Occupational Choice. Journal of Public Economics, v. 46, iss. 2, pp. 133-62. [25] Bohacek, R. (2006). Financial Constraints and Entrepreneurial Investment. Journal of Monetary Economics, v. 53, iss. 8, pp. 2195-2212. [26] Borjas, G. J. (1986). The Self-Employment Experience of Immigrants. Journal of Human Resources, v. 21, iss. 4, pp. 485-506. [27] Borjas, G. J.; Bronars, S. G. (1989). Consumer Discrimination and Self-employment. Journal of Political Economy, v. 97, iss. 3, pp. 581605. 76
[28] Brown, D. M.; Snow, A. (1990). The E¤ects of Product Price Uncertainty on the Owner-Managed Firm: Theory and Evidence. Southern Economic Journal, v. 56, iss. 3, pp. 705-17. [29] Cagetti, M.; De Nardi, M. (2006). Entrepreneurship, Frictions, and Wealth. Journal of Political Economy, v. 114, iss. 5, pp. 835-70. [30] Carrasco, Raquel (1999). Transitions to and from Self-employment in Spain: An Empirical Analysis. Oxford Bulletin of Economics and Statistics, v. 61, iss. 3, pp. 315-41. [31] Casson, M. (2005). Entrepreneurship and the Theory of the Firm. Journal of Economic Behavior and Organization, v.58, iss. 2, pp. 327-48. [32] Chamley, C. (1983). Entrepreneurial Abilities and Liabilities in a Model of Self-Selection. Bell Journal of Economics, v. 14, iss. 1, pp. 70-80. [33] Christensen, L. R. (1971). Entrepreneurial Income: How Does It Measure Up American Economic Review, v. 61, iss. 4, pp. 575-85. [34] Clark, K.; Drinkwater, S. (1998). Ethnicity and Self-Employment in Britain. Oxford Bulletin of Economics and Statistics, v. 60, iss. 3, pp. 383-407. [35] Clemens, C. (2006). Status Concerns and Occupational Choice under Uncertainty. Advances in Theoretical Economics, v. 6, iss. 1, pp. 1-27. [36] Coase, R. H. (1937). The Nature of the Firm. Economica, v. 4, iss. 16, pp. 386-405. [37] Cornelli, F.; Yosha, O. (2003). Stage Financing and the Role of Convertible Securities. Review of Economic Studies, v. 70, iss. 1, pp. 1-32. [38] Covas, F. (2006). Uninsured Idiosyncratic Production Risk with Borrowing Constraints. Journal of Economic Dynamics and Control, v. 30, iss. 11, pp. 2167-90. [39] Cressy, R. (1996). Are Business Startups Debt-Rationed Economic Journal, v. 106, iss. 438, pp. 1253-70.
77
[40] Cullen, J. B.; Gordon, R. H. (2007). Taxes and Entrepreneurial RiskTaking: Theory and Evidence for the U.S. Journal of Public Economics, v. 91, iss. 7-8, pp. 1479-1505. [41] De Fraja, G. (1996). Entrepreneur or Manager: Who Runs the Firm Journal of Industrial Economics, v. 44, iss. 1, pp. 89-98. [42] de Meza, D.; Southey, C. (1996). The Borrower' Curse: Optimism, s Finance and Entrepreneurship. Economic Journal, v. 106, iss. 435, pp. 375-86. [43] de Meza, D.; Webb, D. (1987). Too Much Investment: a Problem of Asymmetric Information. Quarterly Journal of Economics, v. 102, iss. 2, pp. 281-92. [44] de Meza, D.; Webb, D. C. (1989). The Role of Interest Rate Taxes in Credit Markets with Divisible Projects and Asymmetric Information. Journal of Public Economics, v. 39, iss. 1, pp. 33-44. [45] de Meza, D.; Webb, D. C. (1999). Wealth, Enterprise and Credit Policy. Economic Journal, v. 109, iss. 455, pp. 153-63. [46] Dessi, R. (2005). Start-Up Finance, Monitoring, and Collusion. Rand Journal of Economics, v. 36, iss. 2, pp. 255-74. [47] Djankov, S.; La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A. (2002). The Regulation of Entry. Quarterly Journal of Economics, v. 117, iss. 1, pp. 1-37. [48] Dutta, P. K.; Radner, R. (1999). Pro…t Maximization and the Market Selection Hypothesis. Review of Economic Studies, v. 66, iss. 4, pp. 769-98. [49] Ekelund, R. B., Jr.; Hebert, R. F.; Tollison, R. D. (2002). An Economic Analysis of the Protestant Reformation. Journal of Political Economy, v. 110, iss. 3, pp. 646-71. [50] Endres, A. M.; Woods, C. R. (2006). Modern Theories of Entrepreneurial Behavior: A Comparison and Appraisal. Small Business Economics, v. 26, iss. 2, pp. 189-202.
78
[51] Evans, D. S.; Jovanovic, B. (1989). An Estimated Model of Entrepreneurial Choice under Liquidity Constraints. Journal of Political Economy, v. 97, iss. 4, pp. 808-27. [52] Fairlie, R. W. (2002). Drug Dealing and Legitimate Self-Employment. Journal of Labor Economics, v. 20, iss. 3, pp. 538-67. [53] Fan, W.; White, M. J. (2003). Personal Bankruptcy and the Level of Entrepreneurial Activity. Journal of Law and Economics, v. 46, iss. 2, pp. 543-67. [54] Fiet, J. O.; Patel, P. C. (2008). Entrepreneurial Discovery as Constrained, Systematic Search. Small Business Economics, v. 30, iss. 3, pp. 215-29. [55] Fonseca, R.; Lopez-Garcia, P.; Pissarides, C. A. (2001). Entrepreneurship, Start-Up Costs and Employment. European Economic Review, v. 45, iss. 4-6, pp. 692-705. [56] Frank, M. Z. (1988). An Intertemporal Model of Industrial Exit. Quarterly Journal of Economics, v. 103, iss.. 2, pp. 333-44. [57] Fraser, S.; Greene, F. J. (2006). The E¤ects of Experience on Entrepreneurial Optimism and Uncertainty. Economica, v. 73, iss. 290, pp. 169-92. [58] Friar, J. H.; Meyer, M. H. (2003). Entrepreneurship and Start-Ups in the Boston Region: Factors di¤erentiating High-Growth Ventures from Micro-ventures. Small Business Economics, v. 21, iss. 2, pp. 145-52. [59] Friedman, E.; Johnson, S.; Mitton, T. (2003). Propping and Tunneling. Journal of Comparative Economics, v. 31, iss. 4, pp. 732-50. [60] Friedman, M. (1953). Choice, Chance, and the Personal Distribution of Income. Journal of Political Economy, v. 61, iss. 4, pp. 277-90. [61] Fuest, C.; Tillessen, P. (2005). Why Do Governments Use Closed Ended Subsidies to Support Entrepreneurial Investment Economics Letters, v. 89, iss. 1, pp. 24-30.
79
[62] Gans, J. S.; Hsu, D. H.; Stern, S. (2002). When Does Start-Up Innovation Spur the Gale of Creative Destruction Rand Journal of Economics, v. 33, iss. 4, pp. 571-86. [63] Garcia-Penalosa, C.; Wen, J.-F. (2008). Redistribution and Entrepreneurship with Schumpeterian Growth. Journal of Economic Growth, v. 13, iss. 1, pp. 57-80. [64] Gentry, W. M.; Hubbard, R. G. (2000). Tax Policy and Entrepreneurial Entry. American Economic Review, v. 90, iss. 2, pp. 283-87. [65] Gentry, W. M.; Hubbard, R. G. (2004). Entrepreneurship and Household Saving. Advances in Economic Analysis and Policy, v. 4, iss. 1, pp. 1-55. [66] Ghatak, M.; Morelli, M.; Sjostrom, T. (2001). Occupational Choice and Dynamic Incentives. Review of Economic Studies, v. 68, iss. 4, pp. 781-810. [67] Ghatak, M.; Morelli, M.; Sjostrom, T. (2007). Entrepreneurial Talent, Occupational Choice, and Trickle Up Policies. Journal of Economic Theory, v. 137, iss. 1, pp. 27-48. [68] Gi¤ord, S. (1992). Innovation, Firm Size and Growth in a Centralized Organization. Rand Journal of Economics, v. 23, iss. 2, pp. 284-98. [69] Gi¤ord, S. (1992). Allocation of Entrepreneurial Attention. Journal of Economic Behavior and Organization, v. 19, iss. 3, pp. 265-84. [70] Gi¤ord, S. (1993). Heterogeneous Ability, Career Choice and Firm Size. Small Business Economics, v. 5, iss. 4, pp. 249-59. [71] Gine, X.; Townsend, R. M. (2004). Evaluation of Financial Liberalization: A General Equilibrium Model with Constrained Occupation Choice. Journal of Development Economics, v. 74, iss. 2, pp. 269-307. [72] Glaeser, E. L.; Shleifer, A. (2001). Not-for-Pro…t Entrepreneurs. Journal of Public Economics, v. 81, iss. 1, pp. 99-115. [73] Grossman, G. M. (1984). International Trade, Foreign Investment, and the Formation of the Entrepreneurial Class. American Economic Review, v. 74, iss. 4, pp. 605-14. 80
[74] Gruener, H. P. (2003). Redistribution as a Selection Device. Journal of Economic Theory, v. 108, iss. 2, pp. 194-216. [75] Hamilton, B. H. (2000). Does Entrepreneurship Pay An Empirical Analysis of the Returns to Self-Employment. Journal of Political Economy, v. 108, iss. 3, pp. 604-31. [76] Hart, O.; Moore, J. (1994). A Theory of Debt Based on the Inalienability of Human Capital. Quarterly Journal of Economics, v. 109, iss. 4, pp. 841-79. [77] Hausmann, R.; Rodrik, D. (2003). Economic Development as SelfDiscovery. Journal of Development Economics, v. 72, iss. 2, pp. 603-33. [78] Headd, B. (2003). Rede…ning Business Success: Distinguishing between Closure and Failure. Small Business Economics, v. 21, iss. 1, pp. 51-61. [79] Hébert, R. F.; Link, A. N. (1989). In Search of the Meaning of Entrepreneurship. Small Business Economics, v. 1, iss. 1, pp. 39-49. [80] Hellmann, T. (1998). The Allocation of Control Rights in Venture Capital Contracts. Rand Journal of Economics, v. 29, iss. 1, pp. 57-76. [81] Hellmann, T. (2002). A Theory of Strategic Venture Investing. Journal of Financial Economics, v. 64, iss. 2, pp. 285-314. [82] Holmes, T. J.; Schmitz, J. A., Jr. (1990). A Theory of Entrepreneurship and Its Application to the Study of Business Transfers. Journal of Political Economy, v. 98, iss. 2, pp. 265-94. [83] Holmes, T. J.; Schmitz, J. A., Jr. (2001). A Gain from Trade: From Unproductive to Productive Entrepreneurship. Journal of Monetary Economics, v. 47, iss. 2, pp. 417-46. [84] Holtz-Eakin, D.; Joulfaian, D.; Rosen, H. S. (1994). Sticking It Out: Entrepreneurial Survival and Liquidity Constraints. Journal of Political Economy, v. 102, iss. 1, pp. 53-75. [85] Holtz-Eakin, D.; Joulfaian, D.; Rosen, H. S. (1994). Entrepreneurial Decisions and Liquidity Constraints. Rand Journal of Economics, v. 25, iss. 2, pp. 334-47. 81
[86] Holtz-Eakin, D.; Rosen, H. S.; Weathers, R. (2000). Horatio Alger Meets the Mobility Tables. Small Business Economics, June 2000, v. 14, iss. 4, pp. 243-74. [87] Hurst, E.; Lusardi, A. (2004). Liquidity Constraints, Household Wealth, and Entrepreneurship. Journal of Political Economy, v. 112, iss. 2, pp. 319-47. [88] Huyghebaert, N.; Van de Gucht, L.; Van Hulle, C. (2007). The Choice between Bank Debt and Trace Credit in Business Start-Ups. Small Business Economics, v. 29, iss. 4, pp. 435-52. [89] Innes, R. D. (1990). Limited Liability and Incentive Contracting with Ex-ante Action Choices. Journal of Economic Theory, v. 52, iss. 1, pp. 45-67. [90] Innes, R. (1991). Investment and Government Intervention in Credit Markets When There Is Asymmetric Information. Journal of Public Economics, v. 46, iss. 3, pp. 347-81. [91] Johnson, S.; McMillan, J.; Woodru¤, C. (2002). Property Rights and Finance. American Economic Review, v. 92, iss. 5, pp. 1335-56. [92] Jovanovic, B. (1982). Selection and the Evolution of Industry. Econometrica, v. 50, iss. 3, pp. 649-70. [93] Kanbur, S. M. (1979). Of Risk-Taking and the Personal Distribution of Income. Journal of Political Economy, v. 87, iss. 4, pp. 769-97. [94] Kanbur, S. M. R. (1982). Entrepreneurial Risk Taking, Inequality, and Public Policy: An Application of Inequality Decomposition Analysis to the General Equilibrium E¤ects of Progressive Taxation. Journal of Political Economy, v. 90, iss. 1, pp. 1-21. [95] Kaplan, S. N.; Stromberg, P. (2002). Financial Contracting Theory Meets the Real World: An Empirical Analysis of Venture Capital Contracts. Review of Economic Studies, v. 70, iss. 2, pp. 281-315. [96] Kawamura, E. (2005). Competitive Equilibrium with Unawareness in Economies with Production. Journal of Economic Theory, v. 121, iss. 2, pp. 167-91. 82
[97] Kerins, F.; Smith, J. K.; Smith, R. (2004). Opportunity Cost of Capital for Venture Capital Investors and Entrepreneurs. Journal of Financial and Quantitative Analysis, v. 39, iss. 2, pp. 385-405. [98] Keuschnigg, C. (2004). Taxation of a Venture Capitalist with a Portfolio of Firms. Oxford Economic Papers, v. 56, iss. 2, pp. 285-306. [99] Keuschnigg, C.; Nielsen, S. B. (2003). Tax Policy, Venture Capital, and Entrepreneurship. Journal of Public Economics, v. 87, iss. 1, pp. 175-203. [100] Keuschnigg, C.; Nielsen, S. B. (2004). Start-Ups, Venture Capitalists, and the Capital Gains Tax. Journal of Public Economics, v. 88, iss. 5, pp. 1011-42. [101] Khan, B. Z.; Sokolo¤, K. L. (1993). ' Schemes of Practical Utility' En: trepreneurship and Innovation among ' Great Inventors'in the United States, 1790-1865. Journal of Economic History, v. 53, iss. 2, pp. 289307. [102] Kihlstrom, R. E.; La¤ont, J.-J. (1979). A General Equilibrium Entrepreneurial Theory of Firm Formation Based on Risk Aversion. Journal of Political Economy, v. 87, iss. 4, pp. 719-48. [103] King, R. G.; Levine, R. (1993). Finance, Entrepreneurship, and Growth. Journal of Monetary Economics, v. 32, iss. 3, pp. 513-42. [104] Krasa, S.; Villamil, A. P. (2000). Optimal Contracts When Enforcement Is a Decision Variable. Econometrica, v. 68, iss. 1, pp. 119-34. [105] Laussel, D.; Le Breton, M. (1995). A General Equilibrium Theory of Firm Formation Based on Individual Unobservable Skills. European Economic Review, v. 39, iss. 7, pp 1303-19. [106] Lazear, E. P. (2004). Balanced Skills and Entrepreneurship. American Economic Review, v. 94, iss. 2, pp. 208-11. [107] Lazear, E. P. (2005). Entrepreneurship. Journal of Labor Economics, v. 23, iss. 4, pp. 649-80.
83
[108] Lentz, B. F.; Laband, D. N. (1990). Entrepreneurial Success and Occupational Inheritance among Proprietors. Canadian Journal of Economics, v. 23, iss. 3, pp. 563-79. [109] Littlechild, S. C.; Owen, G. (1980). An Austrian Model of the Entrepreneurial Market Process. Journal of Economic Theory, v. 23, iss. 3, pp. 361-79. [110] Lloyd-Ellis, H.; Bernhardt, D. (2000). Enterprise, Inequality and Economic Development. Review of Economic Studies, v. 67, iss. 1, pp. 147-68. [111] Long, J. E. (1982). The Income Tax and Self-Employment. National Tax Journal, v. 35, iss. 1, pp. 31-42. [112] Lucas, R. E. (1978). On the Size Distribution of Business Firms. Bell Journal of Economics, v. 9, iss. 2, pp. 508-23. [113] MacDonald, G. M. (1988) The Economics of Rising Stars. American Economic Review, v. 78, iss. 1, pp. 155-66. [114] Makowski, L.; Ostroy, J. M. (2001). Perfect Competition and the Creativity of the Market. Journal of Economic Literature, v. 39, iss. 2, pp. 479-535. [115] McKernan, S.-M. (2002). The Impact of Microcredit Programs on SelfEmployment Pro…ts: Do Noncredit Program Aspects Matter Review of Economics and Statistics, v. 84, iss. 1, pp. 93-115. [116] McMillan, J.; Woodru¤, C. (2002). The Central Role of Entrepreneurs in Transition Economies. Journal of Economic Perspectives, v. 16, iss. 3, pp. 153-70. [117] Meh, C. A. (2005). Entrepreneurship, Wealth Inequality, and Taxation. Review of Economic Dynamics, v. 8, iss. 3, pp. 688-719. [118] Meh, C. A. (2008). Business Risk, Credit Constraints, and Corporate Taxation. Journal of Economic Dynamics and Control, v. 32, iss. 9, pp. 2971-3008. [119] Michelacci, C. (2003). Low Returns in R&D Due to the Lack of Entrepreneurial Skills. Economic Journal, v. 113, iss. 484, pp. 207-25. 84
[120] Michelacci, C.; Silva, O. (2007). Why So Many Local Entrepreneurs Review of Economics and Statistics, v. 89, iss. 4, pp. 615-33. [121] Michelacci, C.; Suarez, J. (2004). Business Creation and the Stock Market. Review of Economic Studies, v. 71, iss. 2, pp. 459-81. [122] Minkler, A. P. (1993). Knowledge and Internal Organization. Journal of Economic Behavior and Organization, v. 21, iss. 1, pp. 17-30. [123] Minniti, M. (2005). Entrepreneurship and Network Externalities. Journal of Economic Behavior and Organization, v. 57, iss. 1, pp. 1-27. [124] Minniti, M.; Nardone, C. (2007). Being in Someone Else' Shoes: The s Role of Gender in Nascent Entrepreneurship. Small Business Economics, v. 28, iss. 2-3, pp. 223-38. [125] Moore, R. L. (1983). Employer Discrimination: Evidence from SelfEmployed Workers. Review of Economics and Statistics, v. 65, iss. 3, pp. 496-501. [126] Moskowitz, T. J.; Vissing-Jorgensen, A. (2002). The Returns to Entrepreneurial Investment: A Private Equity Premium Puzzle American Economic Review, v. 92, iss. 4, pp. 745-78. [127] Mukhtar, S.-M. (2002). Di¤erences in Male and Female Management Characteristics: A Study of Owner-Manager Businesses. Small Business Economics, v. 18, iss. 4, pp. 289-311. [128] Murphy, K. M.; Shleifer, A.; Vishny, R. W. (1991). The Allocation of Talent: Implications for Growth. Quarterly Journal of Economics, v. 106, iss. 2, pp. 503-30. [129] Murphy, K. M.; Shleifer, A.; Vishny, R. W. (1991). Why Is RentSeeking So Costly to Growth American Economic Review, v. 83, iss. 2, pp. 409-414. [130] Neher, D. V. (1999). Staged Financing: An Agency Perspective. Review of Economic Studies, v. 66, iss. 2, pp. 255-74. [131] Nerkar, A.; Shane, S. (2003). When Do Start-Ups That Exploit Patented Academic Knowledge Survive International Journal of Industrial Organization, v. 21, iss. 9, pp. 1391-1410. 85
[132] Newman, A. F. (2007). Risk-Bearing and Entrepreneurship. Journal of Economic Theory, v. 137, iss. 1, pp. 11-26. [133] Nicholas, T. (1999). Clogs to Clogs in Three Generations Explaining Entrepreneurial Performance in Britain since 1850. Journal of Economic History, v. 59, iss. 3, pp. 688-713. [134] Oi, W. Y. (1983). Heterogeneous Firms and the Organization of Production. Economic Inquiry, v. 21 iss. 2, pp. 147-171. [135] Orzach, R.; Tauman, Y. (2005). Strategic Dropouts. Games and Economic Behavior, v. 50, iss. 1, pp. 79-88. [136] Pagano, M. (1993). The Flotation of Companies on the Stock Market: A Coordination Failure Model. European Economic Review, v. 37, iss. 5, pp. 1101-25. [137] Papke, L. E. (1991). Interstate Business Tax Di¤erentials and New Firm Location: Evidence from Panel Data. Journal of Public Economics, v. 45, iss. 1, pp. 47-68. [138] Parker, S. C. (2003). Asymmetric Information, Occupational Choice and Government Policy. Economic Journal, v. 113, iss.. 490, pp. 86182. [139] Parker, S. C.; Belghitar, Y.; Barmby, T. (2005). Wage Uncertainty and the Labour Supply of Self-Employed Workers. Economic Journal, v. 115, iss. 502, pp. C190-207. [140] Paulson, A. L.; Townsend, R. M.; Karaivanov, A. (2006). Distinguishing Limited Liability from Moral Hazard in a Model of Entrepreneurship. Journal of Political Economy, v. 114, iss. 1, pp. 100-44. [141] Polkovnichenko, V. (2003). Human Capital and the Private Equity Premium. Review of Economic Dynamics, v. 6, iss. 4, pp. 831-45. [142] Quadrini, V. (2000). Entrepreneurship, Saving, and Social Mobility. Review of Economic Dynamics, v. 3, iss. 1, pp. 1-40. [143] Rampini, A. A. (2004). Entrepreneurial Activity, Risk, and the Business Cycle. Journal of Monetary Economics, v. 51, iss. 3, pp. 555-73. 86
[144] Ravid, S. A.; Spiegel, M. (1997). Optimal Financial Contracts for a Start-Up with Unlimited Operating Discretion. Journal of Financial and Quantitative Analysis, v. 32, iss. 3, pp. 269-86. [145] Rees, H.; Shah, A. (1986). An Empirical Analysis of Self-Employment in the U.K. Journal of Applied Econometrics, v. 1, iss. 1, pp. 95-108. [146] Reid, G. C. (1999). Complex Actions and Simple Outcomes: How New Entrepreneurs Stay in Business. Small Business Economics, v. 13, iss. 4, pp. 303-15. [147] Reynolds, P. D. et al. (2005). Global Entrepreneurship Monitor: Data Collection Design and Implementation 1998-2003. Small Business Economics v. 24, iss. 3, pp. 205-31. [148] Robson, M. T.; Wren, C. (1999). Marginal and Average Tax Rates and the Incentive for Self-Employment. Southern Economic Journal, v. 65, iss. 4, pp. 757-73. [149] Rosen, S. (1981). The Economics of Superstars. American Economic Review, v. 71, iss. 5, pp. 845-58. [150] Schultz, T. W. (1975) The Value of the Ability to Deal with Disequilibria. Journal of Economic Literature, v. 13, iss. 3, pp. 827-46. [151] Schwienbacher, A. (2008). Innovation and Venture Capital Exits. Economic Journal, v. 118, iss. 533, pp. 1888-916. [152] Shane, S.; Venkatamaran, S. (2000) The Promise of Entrepreneurship as a Field of Research. Academy of Management Review, 2000, vol. 25, iss. 1, pp. 217-26. [153] Snyder, C. M. (1998). Loan Commitments and the Debt Overhang Problem. Journal of Financial and Quantitative Analysis, v. 33, iss. 1, pp. 87-116. [154] Sobel, R. S.; Dean, A. M. (2008). Has Wal-Mart Buried Mom and Pop The Impact of Wal-Mart on Self-Employment and Small Establishments in the United States. Economic Inquiry, v. 46, iss. 4, pp. 676-95.
87
[155] Stigliz, J.E,; Weiss, A. (1981). Credit Rationing in Markets with Imperfect Information. American Economic Review, v. 71, iss. 3, 393– 411. [156] Takii, K. (2008). Fiscal Policy and Entrepreneurship. Journal of Economic Behavior and Organization, v. 65, iss. 3-4, pp. 592-608. [157] Taylor, M. P. (1999). Survival of the Fittest An Analysis of SelfEmployment Duration in Britain. Economic Journal, v. 109, iss. 454, pp. C140-55. [158] Thornton, J.; Eakin, B. K. (1997). The Utility-Maximizing SelfEmployed Physician. Journal of Human Resources, v. 32, iss. 1, pp. 98-128. [159] van Praag, C. M.; Cramer, J. S. (2001). The Roots of Entrepreneurship and Labour Demand: Individual Ability and Low Risk Aversion. Economica, v. 68, iss. 269, pp. 45-62. [160] Wellington, A. J. (2001). Health Insurance Coverage and Entrepreneurship. Contemporary Economic Policy, v. 19, iss. 4, pp. 465-78. [161] Wiggins, S. N. (1995). Entrepreneurial Enterprises, Endogenous Ownership, and the Limits to Firm Size. Economic Inquiry, v. 33, iss. 1, pp. 54-69. [162] Winton, A.; Yerramilli, V. (2008). Entrepreneurial Finance: Banks versus Venture Capital. Journal of Financial Economics, v. 88, iss. 1, pp. 51-79. [163] Witt, U. (1998). Imagination and Leadership: the Neglected Dimension of an Evolutionary Theory of the Firm. Journal of Economic Behaviour and Organization, v. 35, iss. 2, pp. 161-177. [164] Yu, T. F.-L. (1998). Adaptive Entrepreneurship and the Economic Development of Hong Kong. World Development, v. 26, iss. 5, pp. 897-911. [165] Zazzaro, A. (2005). Should Courts Enforce Credit Contracts Strictly Economic Journal, v. 115, iss. 500, pp. 166-84. [166] Ziegler, C. A. (1985). Innovation and the Imitative Entrepreneur. Journal of Economic Behavior and Organization, v. 6, iss. 2, pp. 103-21.
88