Studies of firms and industries.

Studies of Firms and Industries

About 40 economists met in Cambridge on July 11-12 for an NBER conference on Studies of Firms and Industries. Research Associates Timothy F. Bresnahan, Stanford University; R. Glenn Hubbard, Columbia University; and Ariel Pakes, Yale University, organized

Ricardo J. Caballero, Columbia University; and

Richard K. Lyons, NBER and Columbia University,

"The Role of External Economies in U.S.

Manufacturing" and "Internal versus External Economies

in European Industry" (This paper is described in

"International Seminar on Macroeconomics.")

Frank R. Lichtenberg, NBER and Columbia University;

and Donald Siegel, State University of New York at

Stony Brook, "The Effects of Leveraged Buyouts

on Productivity and Related Aspects of Firm

Behavior" (NBER Working Paper No. 3022)

Steven N. Kaplan, University of Chicago,

"Management Buyouts: Evidence on Post-Buyout Operating

Changes"

William P. Rogerson, Northwestern University,

"Profit Regulation of Defense Contractors and Prizes

for Innovation"

Anil Kashyap and David W. Wilcox, Federal Reserve

Board, "Production Smoothing at the General

Motors Corporation during the 1920s and 1930s"

Thomas J. Holmes, University of Wisconsin at

Madison; and James A. Schmitz, Jr., State University of

New York at Stony Brook, "A Theory of

Entrepreneurship and Its Applications to the Study of

Business Transfers"

Timothy F. Bresnahan, and Peter C. Reiss, NBER and

Stanford University, "How Much Does Entry Change

Competition?"

Tito Boeri, New York University, "Product Choice,

Growth of Incumbent Firms, Entry, and Exit"

Caballero and Lyons develop a method for joint estimation of internal returns to scale and external economies. They then estimate indexes of returns to scale for U.S. manufacturing industries at the two-digit level. Overall, they find that only three of the 20 industry categories show any evidence of internal increasing returns: primary metals, electrical machinery, and paper products. However, there is very strong evidence of external economies, defined as external to a given two-digit industry and internal to the United States. They estimate that if all manufacturing industries simultaneously raise their inputs by 10 percent, aggregate manufacturing production will rise by 13 percent, of which about 5 percent is caused by external economies. Thus, when an industry increases its inputs in isolation by 10 percent, its output rises by no more than 8 percent.

Based on 1100 manufacturing plants involved in leveraged buyouts (LBOs) during 1981-6, Lichtenberg and Siegel find that plants involved in LBOs had significantly higher rates of total factor productivity (TFP) growth than other plants in the same industry. The impact of LBOs on productivity is much larger than the authors' previous estimates of the impact of ownership changes in general on productivity. Management buyouts appear to have a particularly strong positive effect on TFP. Labor and capital employed tend to decline (relative to the industry average) after the buyout, but at a slower rate than they did before the buyout. The ratio of nonproduction-to-production labor cost declines sharply, and wage rates for production workers increase, following LBOs. Plants involved in management buyouts (but not in other LBOs) are less likely to close subsequently than other plants. The average R and D intensity of firms involved in LBOs increased at least as much from 1978 to 1986 as did the average R and D intensity of all firms responding to the NSF/Census Survey of industrial R and D.

Kaplan presents evidence on changes in operating cash flows for a sample of 76 large management buyouts of public companies completed between 1980 and 1986. In the three years after the buyouts, these companies had increased operating income (before depreciation), decreased capital expenditures, and increased net cash flow (the difference between operating income and capital expenditures). Consistent with the operating changes, the (median) combined market adjusted return to pre-buyout public shareholders and post-buyout investors is 77 percent. Kaplan considers three explanations for the post-buyout changes: employment cuts, informational advantages held by managers, and the new incentives created by the buyout. The evidence is strongest for the incentive explanation.

Rogerson argues that constraints on information and incentives require that regulatory institutions create prizes for innovation. Since the quality of an innovation is difficult to describe objectively or to measure, the most natural method for awarding prizes is to allow firms to earn positive economic profit on production contracts. Rogerson calculates the value of the prizes offered on 12 major aerospace systems. The prizes clearly are large enough to support the contention that their existence is an important aspect of the current regulatory structure.

Kashyap and Wilcox examine the development and implementation of production control methods at the General Motors Corporation during the 1920s and 1930s. They show that GM's senior management understood the costs and benefits of production smoothing and implemented an aggressive program of production control roughly 30 years before the economics profession had formally studied the problem. Using new data for 1922 to 1940, Kashyap and Wilcox show that production often was smoother than sales, especially prior to the Great Depression. Critical to this finding, however, is explicit recognition of the importance of model changeover. Production smoothing became less evident after 1932, coincident with a major revision in corporate policy that had the effect of granting greater autonomy to the various divisions of the corporation, and limiting central control. In comparing these policies with current practices, two important lessons emerge. First, both then and now the production planning horizon is tied to the model year. Therefore, there is no direct link between business cycle conditions and inventory positions. Second, to the extent that seasonal variation in demand is an important factor in production and inventory planning, the planning problem has become easier over time because seasonal swings are now much less pronounced than in the prewar era.

Holmes and Schmitz formalize a view of entrepreneurship in which entrepreneurs respond to the opportunities for creating new products that arise because of technological progress. The theory has implications for entry and exit, specialization of labor, and business transfers. These business transfers correspond to individuals changing jobs and sales of firms, among other things. Transfers are seen as a mechanism facilitating division of labor.

Bresnahan and Reiss estimate the equilibrium number of producers in oligopolistic markets, recognizing the importance of scale economies and allowing for heterogeneity in entrants' costs. They show how firms' incentives to enter a market, because of an increase in market demand, depend on the strength of post-entry competition. Using this framework to estimate how entry affects competition in geographically concentrated retail and professional markets, Bresnahan and Reiss find that almost all of the variation in competitive conduct in markets with five or fewer firms occurs in monopolies and duopolies. By the time the market has three to five firms, the next entrant has little effect on competitive conduct.

Studies based on longitudinal samples of business units have found that: 1) job gains from entry of business units systematically exceed job losses caused by exit; 2) growth of incumbent units is unstable; and 3) the growth rate of continuing firms is heterogeneous within sectors. To explain this, Boeri focuses on the interaction between incumbent firms and entrants in a market with product differentiation and uncertainty about the evolution of consumers' preferences over varieties. In the presence of adjustment costs to changing the design of products, entrants might find a more favorable location than incumbent units and erode their market share. Because entry does not occur immediately after a shock, continuing firms temporarily "overshoot" the level of output that can be sustained by their choice of location.