9. The Effects of Uncertainty and Sunk Costs on Firms’ Decision-Making: Evidence from Net Entry, Industry Structure and Investment Dynamics

Pakes and Ericson (1998), Hopenhayn (1992), among others, study firm dynamics under firm-specific uncertainty and evaluate models of firm dynamics under active v. passive learning. These class of models can be better subjected to empirical evaluation using micro-datasets. Since our data is at the industry level, we are not in a position to evaluate the predictions of these models.

Caballero and Pindyck (1996) examine the intertemporal path of a competitive industry where negative demand shocks decrease price along existing supply curve, but positive shocks may induce entry/expansion by incumbents, shifting the supply curve to the right and dampening price increase. Their evidence from a sample of U.S. manufacturing industries shows that sunk costs and industry-wide uncertainty cause the entry (investment) trigger to exceed the cost of capital.

The above models assume perfect competition. Models of oligopolistic competition (e.g. Dixit and Pindyck, p. 309–315) highlight the dependence of outcomes on model assumptions and difficulties of arriving at clear predictions. As in models of perfect competition, the entry price exceeds the Marshallian trigger due to uncertainty and sunk costs, preserving the option value of waiting. But, for example, under simultaneous decision making, neither firm may wants to wait for fear of being preempted by its rival and losing leadership. This could lead to faster, simultaneous, entry than in the leader-follower sequential entry setting. Thus fear of pre-emption may necessitate a faster response and counteract the option value of waiting.

In Audretsch (1995, p. 73–80), mean size of the entering firm is seven employees, varying from 4 to 15 across 2-digit industries. Audretsch (p. 159) finds 19% of exiting firms have been in the industry less than 2 years with mean size of 14 employees; for exiting firms of all ages, the mean size is 23. Dunne, Roberts and Samuelson (1988, p. 503) note that about 39% of firms exit from one Census to the next and entry cohort in each year accounts for about 16% of an industry’s output. While the number of entrants is large, their size is tiny relative to incumbents. Data indicate similar pattern for exiters.

See Ghosal (2007) for construction of the TFP measure. This is the standard TFP measure corrected for cyclical factor utilization (Basu, 1996).