Signaling drive over the long term

First observed by Fama (1980) and later shown by Holmström (1999) in a seminal paper, an economic agent's concern for themarket's perception of his productivity can create incentives to work hard even without any explicit incentives. Since then, career concerns models often assume that agents differ in their inherent ability, which affects their productivity by a constant in each period of employment (Gibbons andMurphy, 1992; Dewatripont et al., 1999). Althoughheterogeneity in ability is very important, another crucial dimension of agents' productivity is how strongly they respond to existing incentives: an agent is more productive if he is more motivated than others by a given incentive system. In a two-period model, Kőszegi and Li (2008) show that this heterogeneity in agents' responsiveness to incentives (“drive”), modeled as an agent's marginal utility of income, motivates them to work harder than that warranted by the existing incentives alone. Because more driven agents are expected to respond more strongly to existing incentives, a high output indicates high drive, which translates into expectations about harder work, and consequently higher future productivity and wages. These drive-signaling incentives induce the agents to work harder in the first period to signal their high valuation of the second period. In their model, however, the magnitude of existing incentives is important: the agent's effort supply dwindles to zero if the existing incentives are negligible. The current model studies how heterogeneity in drive affects the agent's effort supply over the long term, for instance, through an agent's career. In the career concerns tradition, this model assumes that there are some imperfect explicit incentives; and that agents are partly motivated by the implicit incentives provided by a competitive labor market. That is, the agents perceived to bemore productive are paidmore in each period. Although the main model focuses on drive-signaling incentives alone, the result extends easily into the casewhere the agent ismotivated by the implicit incentives provided by the market to signal his talent as in Holmström (1999). This paper finds that, in an infinite horizon model in which the agent's marginal utility of income evolves over time, drive-signaling incentives have a self-reinforcing feature, making it possible for them to build on negligible explicit incentives. Intuitively, agents with higher drive respond more strongly than others to any given explicit incentives. In addition to the explicit incentives, however, this now provides drive-signaling incentives as well, to which more driven