Collaborative Research: Empirical Implications of Sunk Entry and Exit Costs

Despite the important role that sunk costs potentially play in explaining market structure and firm performance there is relatively little empirical evidence on their magnitude, how they vary across different products or across different types of firms, and whether they are useful in explaining the persistent heterogeneity in firm performance within markets. The purpose of this research is to estimate the magnitude of sunk entry and exit costs for a number of U.S. manufactured products and test if they are an important determinant of firm turnover and performance patterns. This project will utilize data from the U.S. Census of Manufactures for the period 1963-1992 to study the entry and exit process for a number of products that are sold in relatively small geographic markets. The substantial movements in population across geographic areas over this 30 year period have produced changes in the structure of local markets as firms have entered and exited in response to these changing demand patterns. The first part of this project will measure entry and exit rates across geographic markets and use these turnover patterns to estimate the magnitude of sunk entry and exit costs faced by firms in specific product markets. The work will recognize that there are different types of entrants, including new firms and those diversifying from other product and geographic markets, and that firms can enter through several channels, including constructing new plants or altering the product mix of their existing production facilities. Each of these types and methods of entry is likely to involve very different levels of sunk investments and thus lead to different firm turnover rates. By estimating sunk costs that vary by type of firm and method of entry, it should be possible to determine if sunk costs are a significant factor generating the differences in turnover patterns and firm lifetimes. The second part of the project will utilize the estimated patterns of sunk cost s across products and entry types to determine if sunk costs contribute to the amount of producer heterogeneity observed within different product markets. Predictions will be tested about the relationships among sunk costs, turnover, and the level and persistence of firm heterogeneity that are derived from recent theoretical models of industry evolution. The Census of Manufactures data are also uniquely suited to this endeavor because they allow the characterization of the amount of heterogeneity among the entire set of producers in a market.