This paper, availabl, here, argues that, while profit margins should (and do) play a role in the assessment of the potential price effect of a horizontal merger, there is no justification for the adoption of a policy that targets high-margin markets. Such a policy is bound to produce false negatives (Type II errors) and false positives (Type I errors) because: (i) accounting profits are not necessarily in line with economic profits, (ii) comparing accounting profits across firms, industries and countries is a notoriously complex exercise, bound to produce misleading conclusions, and (iii) mergers between profitable and not so profitable firms facilitate the efficient reallocation of resources and are, therefore, likely to have positive microeconomic and macroeconomic implications.
Section II looks at the relationship between profit margins and market concentration.
Economists have debated the relationship between profit margins and market concentration for years. Based on some cross-section industry studies in the USA, industrial organisation economists believed for a long period of time that market structure fully determines firms’ strategies and market performance. According to this structure–conduct–performance (SCP) hypothesis, market concentration would cause firms to compete less aggressively and, therefore, lead to higher prices and profit margins. The SCP prediction about the relationship between profit margins and market concentration was later formalised through a model of oligopolistic competition where equilibrium profit margins are given by the Lerner Index. This Index links profit margins and market structure at the firm level.
The views of the SCP school were criticised in two ways. First, economists noted that, while the Lerner Index establishes a relationship between margins and concentration, it is not (necessarily) a causal relationship. Firms’ market shares are not exogenously determined and do not drive profitability, as the SCP paradigm sustains, but rather they are the result of the strategies of competing firms. In other words, firms’ conduct jointly determines market structure and performance; it is not predetermined by market structure. Secondly, some economists noted that the Lerner Index is only valid when firms compete by setting quantities (or choosing capacities) and their products are homogeneous. The link between profit margins and concentration may also exist when firms compete in prices and/or products are differentiated, but that is not necessarily the case. The theoretical consensus is therefore that competition tends to be fiercer the greater the number of firms in the market, but that not all firms impose the same competitive constraint on their competitors and that the strength of the competitive constraint exerted by a given firm need not be directly associated with its market share.
Section III discusses profit margins and merger control.
The discussion above provides support for the ‘structural presumption’ that underpins horizontal merger control throughout most of the world. Mergers that materially increase concentration are presumed to raise prices and reduce consumer welfare. Yet, merger control should also investigate profit margins, and especially focus on mergers involving high-margin firms in the right circumstances. The price effect of a merger depends on the extent to which the sales lost by one merging party from a price rise are diverted to the other and vice versa (i.e. the parties’ diversion ratios), and their relative margins. The reason why margins matter is straightforward: when margins are higher, the recapture of the diverted sales from a price increase will be more valuable and thus the merger is more likely to make a price increase profitable. In other words, the higher the merging parties’ margins in a given case, the more likely a traditional merger analysis, based on market shares alone, will underestimate the price effects of the merger (all else equal).
However, the adoption of a stricter approach when assessing mergers involving highly profitable firms or industries is not justified in general. The merging parties may not be close competitors after all. Further, not all firms operating in a high-margin industry will be equally profitable; in fact, there are many examples of mergers in profitable industries involving industry laggards seeking to gain scale and efficiency in order to be able to compete with market leaders. Finally, such an approach ignores the possibility of efficiencies, in particular that a firm’s high margin may be driven by its relatively high efficiency.
A well-functioning economy must be able to reallocate resources away from its relatively inefficient firms to its most efficient ones. While inefficient firms may exit markets in different ways, there are reasons to believe that the merger route plays a fundamental role in many markets. Larger firms may have incentives to close down relatively efficient plants before a smaller firm would close down its less efficient plants, since larger firms can internalise the positive price effect caused by a reduction in capacity to a greater extent. Mergers may thus facilitate the retirement of inefficient capacity, since larger firms may find it optimal to buy and close the assets of the relatively inefficient firms instead of closing down their more efficient plants. Another reason to refrain from targeting mergers in high-margin markets is that they may not only facilitate efficient exit levels but also encourage the entry of new firms. Start-ups face considerable risks, and the most efficient way to monetise their investments may be to sell their businesses to a firm with the necessary cash flows or capitalisation—i.e. to a profitable incumbent.
Section IV asks what profit margin we should take into account.
A merger will be more likely to result in a price increase when the value of the recaptured sales is large. It follows, therefore, that the relevant margin for assessing the price effect of a merger is whatever measure best reflects the profits made on the recaptured sales. In some cases, the relevant profit margin will be given by the difference between the average price and the short-run average variable cost. In others, however, this will not be the case. For example, in some instances, the relevant profit margin will be the incremental profit of additional sales in the longer term.
Identifying the right margin measure is only part of the story though; one still needs to be able to measure it accurately. The problem in practice is that the only information available to the researcher is accounting information, which may not be consistent from firm to firm or industry to industry, and may not correspond to economists’ definitions of profits. Accounting profit margins exclude opportunity costs that are relevant to understanding the pricing and investment decisions of firms; economic profits and accounting profits differ in the measurement of capital costs; and accounting profit margins are particularly inappropriate to measure long-term margins, especially in multi-product firms with significant fixed costs that are common to many of their products.
This does not mean that profit margins cannot be identified. However, attention should be devoted to the identification of the relevant profit margin measure in each case. Secondly, because calculating an economically meaningful profit margin using accounting data is not easy, it is important to perform various sensitivity analyses. Thirdly, comparing accounting profit margins across industries is notoriously complicated. Thus, a merger control policy geared towards high-margin industries may prove not only unfair but inefficient, causing both Type I errors (in the industries incorrectly classified as high-margin) and Type II errors (in the industries incorrectly classified as low-margin).
Section V deals with merger control and the decline of the labour share.
The share of labour in GDP has fallen both in the USA and in many European economies since the 1980s and, particularly, since the 2000s. The two standard explanations for this are globalisation and the reduction in the cost of capital relative to the cost of labour as a result of technological progress. Recent papers, however, have found that the decline of the labour share is related to increased mark-ups and profit margins.
Against that background, some economists in the USA blame the increase in concentration and market power on a lenient merger control policy: too many anticompetitive mergers have been cleared and the remedies imposed have proved ineffective. This may be true, but the author doubts it explains it all – structural elements seem to be at play. In short, the author considers that the evidence that the observed decline in the labour share and the corresponding increase in inequality are the result of a lenient merger control policy is not robust enough to support a radical change in merger control policy and, in particular, increase scrutiny in high-margin industries.
This is a relatively old paper, but I thought it would be interesting to review it here because it presents a balanced argument on how to take high-profit margins into account in the context of prevailing merger control regimes. In a way, it is a riposte to the argument in the previous paper that higher profit margins justify a more interventionist merger control. I would note that while that paper focused on protecting market contestability, this one focuses much more – as it typical of the authors’ papers on the matter – on allowing the most productive companies to prosper as a way to enhance consumer welfare, even if that means that less efficient companies are acquired or exit the market.
At the same time, I am not sure that the papers are as much at odds as one would think on a superficial reading. After all, both agree that mergers involving high-margin companies can soften competition in a problematic way, and deserve special attention. And while the papers disagree on whether stricter enforcement is justified in high-margin instances, none of the papers argues for specific changes to the substantive assessment of mergers.