This paper – which can be found here – is yet another attack on the current foundations of antitrust and one more call for a more interventionist antitrust.
The reason why I circulate it here is that this is not the typical paper of antitrust enforcement practice – which tends to argue that antitrust should go beyond its focus on consumer / total welfare and look at other criteria, such as market structure, competitive process or what have you. Instead, the attack is on the very foundations of consumer welfare or efficiency as suitable standards for antitrust analysis.
The paper begins by identifying three goals– consumer surplus, allocative efficiency, and productive efficiency – which are commonly put forth as being the basis of antitrust. It then proceeds to attack each in turn.
Regarding consumer surplus, the main question is whether it – i.e. consumer surplus understood in terms of prices and quantities – bears any relationship to actual welfare. Part of this critique relates to findings of behavioural economics , and to how antitrust models disregard effective utility and look instead at expected utility. Another element of the critique is that consumer surplus leads to different (intuitive) welfare assessments depending on the market we are speaking about. A good example of this is that, from an enforcement perspective, practices affecting consumer surplus in the yachts and baby food market are unlikely to attract the same attention – and rightly so, it seems, because for the same level of consumer surplus it can hardly be said that the gains in consumer welfare are identical.
Regarding allocative efficiency (i.e. total welfare), this is a legal standard that raises serious distributional issues regarding how to allocate the relevant surplus. Inasmuch as consumer welfare is favoured, this is a normative choice which has no basis in economic theory. I would reinforce this point by reminding that Bork in his “The Antitrust Paradox” favoured total welfare, even though he thought that higher overall welfare would lead ultimately to increased consumer welfare as well. Another point (which, curiously, is only made in a footnote), is that such redistributional choices can only be anchored in terms of welfare on the basis of interpersonal comparisons of utility which are impossible to make – a point which would seem to be valid for assessments of consumer surplus in different markets as well. More importantly, all the issues regarding the dubious correlation between consumer surplus and welfare discussed in the paragraph above also apply to scenarios of allocative efficiency. Lastly, measures of allocative efficiency do not take into account externalities – which means that, absent internalisation of social costs, allocative efficiency for the parties may very well lead to production in excess of the social optimum, and hence to allocative inefficiency.
The last goal pursued by antitrust – productive efficiency – is only a proximate goal. According to the author, productive efficiency is used only for a small number antitrust cases with a view to identifying exclusion (e.g. where pricing below cost is at issue). Hence, productive efficiency is ultimately a tool for identifying practices which have the potential to reduce consumer or total welfare, one would assume. His main criticism is, again, that antitrust policy refuses to consider externalities beyond considerations about price and quantities.
It is hard to be harsh on such a short polemic, but I have some problems with this piece:
First, it ignores the role of non-price standards in antitrust. Now, I’m not saying that they are important in practice. However, the law requires that non-price elements – such as quality, variety or innovation – be taken into account (in theory). On the other hand, the fact that we really don’t have any off-the-shelf (or even just normatively coherent) mechanisms for assessing and balancing all relevant (price and non-price) antitrust considerations is something that would seem to reinforce the argument in this piece that antitrust says one thing but does another (if it knows what it is doing at all).
My second concern relates to the first: the reasons why consumer and total surplus is used in lieu welfare, and why antitrust favours consumers over producers, are not based on economic logic. Instead, it is because these may well be the best proxies that we have. This state of affairs raises a number of issues: (i) if we have no better proxy for welfare than surplus, than criticising surplus as a standard seems spurious; (ii) however, it is evident that the basis for considering that welfare is the same as surplus have been increasingly challenged by findings in behavioural economics; (iii) we could avoid this problem (in practice) if we were applying per se rules (but these rules can then be attacked as not being reasonable, as they have been in the past 50 years). Given this state of affairs, and the fact that virtually all practices other than naked cartels are subject to a rule of reason, if surplus cannot be used as a valid metric then we have nothing on which to base a finding of infringement. While the author is clearly concerned with the reduction in the levels of antitrust enforcement in the US (one recent study of rule-of-reason cases decided during 1999–2009, for example, revealed that modern “plaintiffs almost never win under the rule of reason.” During that decade, defendants won 221 of 222 rule-of-reason cases…), he advances no alternative approach.
And therein lies the problem: this paper debunks the use of traditional tools without presenting any better (or even valid) alternatives. Thus, the author seems to have unwittingly placed himself between a bigger rock and a harder place: if the economic foundations falter and we have no better (economic) tools to measure consumer welfare, then we need to look outside of economics – in which case it is unclear why we should keep focusing on consumer welfare at all. But if we do decide to keep using consumer welfare dissociated from economics, it is unclear how we can avoid that “consumer welfare” be construed to encompass such varied concerns that it becomes a mere slogan.
I would point out I tend to sympathise with the author and some of his arguments, as I think that our “economic priors” are inevitably biased. Further, to a lawyer this state of affairs is not necessarily problematic – rules are, of necessity, somewhat arbitrary, and there is a trade-off between the extent of a rule and its normative coherence. From where I stand – and despite the criticisms above – the virtue of the article is that the author brings very much into the open how limited the pretensions to normative coherence of antitrust are – and does this not by arguing that antitrust should have other goals (as is usually the case) but by exposing the potential internal incoherence of antitrust under its own terms.