|
Article Excerpt
1. Introduction
Where do entrepreneurs come from? Employees of established companies turn out to be one of the most important sources for entrepreneurship. The semiconductor industry, for example, has an impressive genealogy, where, generation after generation, employees left their parent company to launch the next entrant. (1) In a sample of fast-growing private companies, Bhide (1994, p. 151) finds that "71 percent of all founders had replicated or modified an idea encountered through previous employment." Similarly, Cooper (1985) reports that in a broad cross section of industries 70 percent of all founders pursued opportunities closely related to their previous employment.
A popular complaint is that established firms simply miss out on all these great opportunities, (2) yet this hardly constitutes a systematic explanation. This paper argues that the departure of employees to become entrepreneurs can be understood as a natural equilibrium outcome in the process of innovation. In addition, we stress that it is not the only possible equilibrium outcome. While some employees leave, others stay to develop their innovations internally, turning employees into intrapreneurs. (3) Other firms use their employees' innovations to found spin-offs. (4) To get a better understanding of the supply of entrepreneurs, we need to ask how companies manage employee innovations.
This paper provides a theory of how firms formulate polices that affect the generation and development of employee innovations. We consider a model where a wealth-constrained employee sometimes obtains a new idea that is unrelated to her assigned employment task. She faces a multitask choice of either focusing on her task, or exploring the new idea. The incentive to remain focused depends on how likely she is to succeed with the assigned task, and how well the task is compensated. The attraction of pursuing a new idea is that it may generate far greater returns (including private benefits) than the employee could ever hope to earn on her assigned task. However, pursing an idea is risky, and whether it generates returns depends on how it can be commercialized.
In managing the employee's incentives, the firm faces a fundamental trade-off between exploitation and exploration. The firm wants the employee to focus on the core task, yet the employee is attracted to idea exploration, which holds promise of greater private benefits. The firm can offer bonuses for core tasks and define policies for developing employee innovations. An important assumption is that the firm is able to commit to policies, even if they are ex post inefficient. The analysis generates four types of equilibria. First, in a focused equilibrium, the firm provides sufficient bonuses to keep all employees focused on their assigned tasks. The firm has a policy of refusing to develop any innovations that fall outside of the core activities, although in equilibrium employees never generate innovations. Second, the stubborn equilibrium is similar to the focused equilibrium, except that some employees (those with ideas and poor prospects for their assigned task) explore their ideas and sometimes generate profitable innovations. The firm sticks to its policy of refusing to support innovations. In equilibrium the employee leaves to become an entrepreneur (this requires that the employee owns the (intellectual property (IP)). In the first two equilibria, the firm actively discourages idea exploration. In the next two equilibria the firm accepts that it cannot "stem the tide" of its employee's ambitions, and decides to "go with the flow." In the intrapreneurial equilibrium, employees excessively explore their ideas, and the firm makes the best of their innovations by developing them as internal ventures. The entrepreneurial equilibrium is similar, except that development occurs through start-ups (if the employee owns the IP) or spin-offs (if the firm owns the IP).
The model generates several key insights. In the first two equilibria firms actively discourage employee innovation. In the stubborn equilibrium this even leads to a "dramatic" sequence of events, where employees prefer to develop their innovations inside the firm, the firm refuses (even though this is ex post inefficient), and the employees leave to become entrepreneurs. The model explains the curious empirical phenomenon that some entrepreneurs start their companies only after being rejected by their employers. While this may lead to a perception of "firms missing opportunities," the analysis shows that the firm's intransigence may be part of an optimal ex ante policy to promote greater focus on core tasks.
The model also generates new insights about IP rights and the importance of the external environment. If the employee owns the IP, as the external environment becomes increasingly attractive, the outcomes go from the focused, to the stubborn, to the intrapreneurial, to the entrepreneurial equilibrium. Contrast this with the case in which the firm owns the IP: For most parameter values the firm implements the focused equilibrium; only if the external environment becomes very attractive does the firm switch to the entrepreneurial equilibrium. Allocating IP rights to the employee thus increases employee innovation. The key intuition is that if the firm owns the IP, then the external environment only constitutes an opportunity for the firm. If the employee owns the IP, then the external environment also constitutes a threat, and the firm has to accommodate employee innovation more. (5)
The paper differs from the prior literature in several important respects. Most of the literature on innovation uses a single-task model (e.g., Aghion and Tirole 1994, Anand and Galetovic 2000, Gans and Stern 2000) and focuses on underinvestment in innovation. This paper considers a multitask model where incentives for innovation may interfere with incentives for core tasks. The prior literature explains employee departures as efficient outcomes (e.g., Cassiman and Ueda 2006, Lewis and Yao 2001, Pakes and Nitzan 1983), whereas our analysis can account for employee departures that are ex post inefficient. This is because the firm can commit to an optimal ex ante policy that trades off the benefits of greater employee focus with the costs of inefficient employee departures. Moreover, unlike some of the previous literature, the allocation of IP rights matters in this model. The reason that the Coase theorem does not apply is that firms can commit to ex post inefficient outcomes. The paper is related to several other literatures, not only economic theory, but also empirical research, ranging from ethnographic observation, to legal study, to econometric data analysis. Rather than attempting to summarize all this in the introduction, we defer the full literature discussion until after we have developed the theory.
The remainder of this paper is structured as follows. Section 2 derives the theory model. Section 3 provides an extensive discussion, including a literature review. It is followed by a brief conclusion. All proofs are given in the online appendix (provided in the e-companion). (6)
2. The Theory
2.1. The Base Model
Suppose there is a firm and a wealth-constrained employee. At date an employment contract is drafted. At date 1 the employee may get a new idea that is unrelated to her main employment task. She has a private choice about whether to explore her idea or to focus on her assigned task. If she sticks to her task, she generates some verifiable value at date 3. If she explores her idea, she discovers at date 2 whether or not the idea is feasible. If it is feasible, the idea can be developed inside or outside the firm, or else it can be shelved. All returns accrue at date 3. All parties are risk-neutral profit maximizers. There is no discounting. For convenience, Table EC.1 in the online appendix provides a time line and Table EC.2 summarizes the notation used in the paper.
Let us begin with the idea exploration stage at date 1. As part of her regular job activities, the employee may serendipitously get a new idea that falls outside her main task. Let [lambda] be the probability that she obtains such an idea. When she gets an idea, she can choose to ignore or explore it. We model this as a multitask problem, where she can focus either on Project A or B.
Project A denotes the assigned employment task, which may include a large variety of activities that improve the firm's current core businesses. If successful, the assigned task generates a value a for the firm. At the time of receiving the idea, the employee also receives a private signal about the likelihood of success: Either the prospects for the assigned task are "strong," so that success is certain; or they are "weak," so that the probability of success is given by [phi][member of] (0, 1). The probability that the core prospects are strong is denoted by [gamma] [member of] (0, 1). (7)
Project B represents exploring a new idea. There are two stages to turning an idea into a successful venture: exploring the idea and developing the innovation. If the employee explores her idea, she finds out, at date 2, whether it is feasible (which occurs with probability p) or not. If it is not feasible, all parties receive zero utility. If it is feasible, we call the idea an innovation. A development effort is then required to generate value from the innovation. (8) The employee is indispensable for development. (9)
If the employee has an idea, she has to choose between focusing on her assigned task (A) or exploring her idea (B). We distinguish three levels of exploration:
[[sigma].sup.-]: No exploration (employee never explores ideas)
[sigma]*: Efficient exploration (employee explores ideas only if core prospects weak)
[[sigma].sup.+]: Excessive exploration (employee explores ideas irrespective of core prospects)
Condition (1) below justifies the use of the word "efficient" and "excessive."
Consider now the events at date 2. If the employee has chosen A, then nothing happens until date 3. If the employee has chosen to pursue an idea that turns out to be feasible, there are three possible development choices:
[[sigma].sup.0]: No Development (Shelving)
[[delta].sup.y]: Internal Development (Internal ventures)
[[delta].sup.x]: External Development (Start-ups, Spin-offs)
If the innovation is not developed ([[delta].sup.0]), there are no returns. If the innovation is developed internally ([[delta].sup.y]), it generates a total utility y. If the innovation is developed externally ([[delta].sup.x]), it generates a total utility x. We assume that x and y are exogenous and symmetrically known. (10)
The utilities x and y include not only financial returns, but also any private benefits that the employee derives from developing her innovation. Private benefits might represent an increase in the value of the employee's human capital, and/or the satisfaction of developing her innovation, as well as monetary returns that the employee is always able to extract. (11) Private benefits allow us to capture the notion that employees are attracted to exploring ideas, rather than performing core tasks. This introduces a natural source of disagreement between the employee and the firm. Let [[beta].sub.x] be the fraction of the total utility x that constitutes a private benefit, i.e., [[beta].sub.x]x is the expected (nontransferable) private benefit and (1 - [[beta].sub.x])x is the expected (transferable) financial return. It is the same for [[beta].sub.y]. For expositional convenience, we focus on the case where [[beta].sub.x] = [[beta].sub.y]. The online appendix shows that relaxing...
|
