A critical commentary on the critical comments on critical loss.

I. INTRODUCTION

In the last few years, applications of the critical loss methodology have been harshly criticized by a collection of economists, some academics, some consultants, some government regulators, and some all of the above. When stripped to its least common denominator, the basic criticism involves an objection to the alleged atheoretical application of the critical loss methodology. Although some of the characterizations are more colorful, the objecting economists basically believe that their colleagues, who aggressively practice critical loss analysis, are committing a grave analytical error. As the critical loss practitioners believe that the critical loss construct is a simple matter of arithmetic, it is reasonable to infer that they disagree. (1)

This article takes a step back and systematically reviews the underpinnings of critical loss analysis debate. First, we present an overview of the dispute to place the questions in context and identify the key shortcomings of the criticisms. Next, the analysis addresses the interactions between economic theory and the Merger Guidelines to set a background for the presentation. Then, we discuss the application of critical loss analysis to two distinct merger analysis questions: market definition and overall competitive effect.

The fundamental problem involves a difference of opinion on the market equilibrium concept. The standard critical loss construct starts with a market-level homogenous goods analysis and builds down to the firm, while the critical loss critics usually start with a firm-level presentation and build up to the differentiated market. To explain the relevance of the issues, our discussion will subdivide the presentation into two sections, one for relatively homogenous goods and the other for relatively differentiated goods.

Given the background on the theoretical dispute, the market definition question is addressed first and then the presentation moves on to the use of critical loss for competitive effects. The Merger Guidelines introduce the well-known hypothetical monopohst construct to aid in the definition of the market. Here, the idea is to find some way to quantify the hypothetical pricing decision facing the hypothetical monopolist competing in a hypothetical market. The standard critical loss methodology requires the analyst to first compute the critical loss, then predict the actual loss with market data and finally compare the two numbers. For homogenous goods analysis, the test is often straightforward; if the actual losses fail to exceed the critical loss, then the narrow market is accepted. In contrast, the differentiated goods case may be more complicated, because theoretical analysis uses data on actual margins to infer actual losses at the firm level. If the actual critical loss based on market-wide analysis differs substantially from the aggregate predictions of a firm-level model, the analyst would have to either reject the critical loss data or the specific implementation of the firm-level model.

Once the market is defined, the Merger Guidelines turn to an assessment of the merger's overall competitive effect. As merger policy generally focuses on price, it appears intuitive to quantify the post-merger pricing decision with some type of critical loss analysis. For relatively homogenous markets, the critical loss analysis simply focuses on the competitive responses (e.g., substitution to other products, fringe expansion, entry, cheating by mavericks) to a hypothetical market-wide anticompetitive price increase. The change in profits for the output-restricting group of firms is computed, and the effect of a price increase is deemed profitable (hence likely) or unprofitable (hence unlikely). Theoretical criticisms do not play a role in this analysis. In contrast, if the market is relatively differentiated, a specific demand structure, coupled with an equilibrium concept, allows the analyst to solve for the merger's potential impact on each firm in the market. However, some methodology must be developed to address the crucial issues of repositioning and entry to complete the competitive analysis. Section V presents a couple of suggestions in this area.

In conclusion, economics offers antitrust enforcement a wide range of techniques to aid in merger evaluation. No single construct trumps all the others. Standard critical loss techniques remain appropriate in some fact situations, but should be replaced with more complex critical loss analyses in other situations. Models are useful to structure data, but standing alone, only define a possible competitive outcome.