Is Cost-Cutting Evidence of X-Inefficiency?
探讨企业削减成本公告能否作为研究X-低效率变化的证据，区分削减“赘肉”与正常重新优化，并检验财务压力下企业更可能削减赘肉的假设。
X-inefficiency is surely among the most important topics in microeconomics. Yet, economists have found it difficult to study. If a given level of X-inefficiency were inevitable and changeless, it would be of little interest (indeed, would not really deserve to be called Xinefficiency at all). Thus, our attention should focus on actual and potential changes in Xinefficiency: that is, on causes of changes, internal to a firm, that shift the firm’s cost function. We explore the use of firms’ “costcutting” announcements to study the causes of changes in X-inefficiency. Cost-cutting announcements by large corporations are made frequently and are reported in the business press. One might be tempted to interpret these announcements as indicating efforts to reduce X-inefficiency, and indeed we think that a substantial number of them are just that: our discussions with managers confirm that finding and trimming “fat” within the corporation is an important, yet somewhat intermittent, activity. It is also clear, however, that not all costcutting announcements concern X-inefficiency. Some, perhaps many, may instead be reoptimizing (input and/or output) quantity responses to changes in exogenous factors, such as the prices of inputs or outputs. The announcements seldom make it clear whether the activity is reoptimization or fat-trimming. We aim to explore the existence and nature of fat-trimming within a firm, and how one might distinguish this from normal reoptimization. In particular, we seek evidence bearing on a central hypothesis in the informal theory of Xinefficiency (Harvey Leibenstein, 1966), with very broad support in news reports, in popular belief, and in our interviews with managers. This fat hypothesis is that a firm is most apt to cut costs to reduce X-inefficiency when it is under financial pressure. This hypothesis, if correct, has implications both for firm strategy and for competition policy. While Olivier Blanchard et al. (1994) studied the effects of idiosyncratic cash shocks, it seems desirable to find a more systematic source of wealth shocks. One such source is exogenous changes in the prices of competitively supplied inputs or outputs. In Borenstein and Farrell (1999), however, we explain that not only must the price changes be out of the firm’s control, they must also leave the firm’s production possibilities unchanged. For instance, if the price of oil increases due to political instability, that raises the expected profits of U.S. oil companies on the oil they will be able to sell in any case, but it might simultaneously indicate reduced opportunity to explore for oil in the future. Likewise, if the technology for gold mining improved, the price of gold would fall, but the availability of the new technology to firms under study would offset the price change and could result in a net positive wealth shock (and increase their optimal production quantities). One might assess the source of price changes by examining quantity changes (potentially differentiating between supply and demand shocks), news reports, or the technology involved. Drawing inferences from cost-cutting announcements faces another problem, because a price shock that (say) lowers the firm’s overall profitability is likely also to lower its marginal * Borenstein: Haas School of Business, University of California, Berkeley, CA 94720-1900 and NBER (e-mail: borenste@haas.berkeley.edu); Farrell: Department of Economics, University of California, Berkeley, CA 94720-3880 (e-mail: farrell@econ.berkeley.edu). For research support, we are grateful to the Alfred P. Sloan Foundation, via a grant to the Industrial Technology and Productivity Project of the National Bureau of Economic Research. Erin Mansur, David O’Neill, and Evan Rose provided outstanding research assistance. We thank many managers in the goldmining and oil industries for discussing cost-cutting with us. 1 There are other data problems, including the following: (i) the division(s) in which the cost cutting is to occur are usually not precisely reported; (ii) magnitudes are not systematically reported (and usually are projections); and (iii) multiple reports of a single cost-cutting effort, sometimes months apart, are common.