Several stylized facts appear well established: (i) there is significant churning of firms even in mature industries; (ii) entrants are typically small compared to incumbents and have low survival probabilities; (iii) typical exiting firm is small and young; and (iv) larger firms tend to be older with higher survival probabilities. Given these findings, identifying the forces that drive industry dynamics and the evolution of firm size distribution has taken on renewed interest. The relatively recent literature has tended to focus on technological change as the key driving force. The primary objective of this paper is to assess the role played by uncertainty and sunk costs on the intertemporal dynamics of industry structure.
Uncertainty and sunk costs imply an option value of waiting which alters the entry and exit trigger prices (Dixit, 1989; Dixit and Pindyck, 1994; Caballero and Pindyck, 1996). This suggests that the option value channel may be an important determinant of entry, exit and industry dynamics (Section II(i)). A second channel via which uncertainty and sunk costs may affect industry dynamics is financial market frictions (e.g., Greenwald and Stiglitz, 1990; Williamson, 1988; Cooley and Quadrini, 2000; Cabral and Mata, 2001). This literature suggests that uncertainty and sunk costs exacerbate financing constraints, affecting decisions of entrants and incumbents (Section II(ii)). Our study is motivated by the fact that while the theory linking uncertainty and sunk costs to industry dynamics is relatively well developed, empirical evaluation of these models appears limited. Finally, since the literature has shown innovation to be a key determinant of industry dynamics, our empirical analysis also examines the role played by technological change (Section III).
We assemble an extensive database covering 267 U.S. manufacturing industries over a 30-year period and containing information on the number of firms and establishments (by size class), output concentration, alternate proxies for sunk capital costs, and use time-series data to measure uncertainty and technical change (Section IV). A dynamic panel data model is used to estimate the impact of uncertainty and sunk costs on the intertemporal dynamics of various industry structure characteristics (Section V). Our key findings (Section VI) are that greater uncertainty (i) reduces the number of small establishments and firms primarily in high sunk cost industries; (ii) has virtually no impact on large establishments; (iii) results in a less skewed size distribution of firms and establishments in high sunk costs industries, and (iv) marginally increases industry concentration. Second, technological progress has an adverse impact on the number of small establishments and firms in an industry. Third, our estimates suggest that uncertainty and sunk costs have a greater quantitative impact on industry dynamics than technical change.
Our findings on uncertainty and sunk costs could be useful in several areas. First, they provide guidance for competition policy since analysis of entry and other forces that regulate competition are an integral part of DOJ and FTC merger guidelines. The guidelines contain extensive discussion of sunk costs as a barrier to entry, but uncertainty is de-emphasized. Our results suggest that uncertainty compounds the sunk cost barriers, lowers the probability of survival of smaller incumbents and retards entry. This implies that mergers, for example, ought to receive closer scrutiny in markets with greater uncertainty. Second, evaluating the determinants of merger and acquisition activity has long been an important area of research; see Jovanovic and Rousseau (2001) and the reference there. If uncertainty reduces the probability of survival, it would have implications for reallocation of capital; e.g., do the assets exit the industry or are they reallocated to other firms within the industry via merger or acquisition? Given our broad findings, it would be interesting to explore whether uncertainty in combination with sunk costs helps explain part of M&A waves.
Third, while skewed firm size distribution has been well documented, empirical analysis quantifying it’s evolution is somewhat limited. Our results suggest that uncertainty and sunk costs are important determinants of the evolution of firm size distribution. Fourth, Davis, Haltiwanger and Schuh (1996) find that job destruction and creation decline sharply with firm size/age; Cooley and Quadrini (2000) shed more light on this and show that small/young firms have greater exits (destruction) due to financial market frictions. Our results provide specific insights: uncertainty and sunk costs significantly contribute to the turnover of firms. If entry and exit reflect the bigger picture of economic activity, then our results imply that uncertainty and sunk costs influence key variables like job turnover and investment spending. Finally, they could provide insights into the evolution of specific industries: e.g., the electric industry is undergoing deregulation and we are observing numerous mergers involving firms of different sizes. Also, uncertainty about prices and profits with sharp fluctuations during 1998 and 1999, and 2000 in specific regions in the U.S., are well documented resulting in many utilities experiencing financial distress. While a detailed study would be required to examine this industry’s evolution, our findings could be used to predict a future path that leads to weeding out of smaller firms and greater concentration.
Download
Impact of Uncertainty and Sunk Costs on Firm Survival and Industry Dynamics
