Inside, Outside, Upside Down (or how Antitrust got itself Dr Seussed)
There are times when I feel like I am living in a Dr. Seuss book. Getting the kids ready for nursery (red shoe, blue shoe); juggling writing, family, house, life (ten apples up on top, we must not let them drop!); dreaming of making a better world (oh, the places you’ll go!). Recently there has been a question that has be plaguing me. It concerns the relationship between antitrust, markets, and the quintessential market participant — the company. The question is this: why doesn’t antitrust look inside the company? (Inside, outside, upside down).
Before we dive into this head-spinning question, a little personal context may be in order.
There was a time when I fully believed that the only thing needed to save the world (yes, my goal was — is — that ambitious) was freer markets. It’s why I became a competition lawyer. But in 2013 I had an epiphany, and it would not be hyperbole to say that it brought about an existential crisis. You see that year I found myself pulling all-nighters to, amongst other things, calculate the respective market shares of two merging fizzy drinks companies, which is referred to in the business rather graphically as the “share of throat”. Suddenly it struck me that it didn’t matter how competitive the market for fizzy drinks was. No amount of competition, lowering the price and increasing the output of carbonated sugar water, would be good for people, and a system solely designed to ensure such competition wasn’t doing enough. The externalities of the markets I was advising on were everywhere to see, although I had steadfastly ignored them. Fizzy drinks were the last straw. I couldn’t escape the obvious question: how can antitrust seriously care about prices when we are all busy destroying the world?
At the time I knew only one way to do competition law: you try to maintain free and fair competition on the promise, the hope, that efficiencies will ensue. You had to believe, you had to have faith. But I could no longer sleep easy, resting my conscience on the assumptions embedded within that form of antitrust. Firms seemed to be almost sociopathically designed to consume resources and spit out social harms. It was clear to me that this revelation was inconsistent with my continuing in my job because competition was at best a neutral factor and at worst a catalyst. I left Big Law and went in search of a better answer.
That journey took me into the world of responsible business and corporate governance (and ultimately B Corps). I found companies that were playing by a different rulebook, attempting to support their workers, bringing health and wealth to their customers and suppliers, and nurturing the environment, whilst also making money.
It started to become clear to me that the manual by which a company operates has a direct impact on its conduct, and therefore on the competitive dynamic and outcomes on the market. The imperative for the vast majority of firms to maximise shareholder returns drives them to seek and exploit market failures and to allocate the resulting rents in a particular way: to shareholders and (with shareholder complicity) to management, at the expense of workers, consumers, communities and the environment. But if that is the case then why don’t we look at the influence of shareholder value within the antitrust framework?
After some searching I have come up with two provisional answers to that question: (1) antitrust assumes firms behave in a certain way externally, on the market (i.e. they seek to maximise profits) and compete with other firms to do so, which is, according to the textbooks, consistent with the generation of efficiency. So if you have competitive markets then you don’t need to look at what is going on within the firm; and (2) antitrust assumes that firms will organise themselves internally in a way that will maximise value / minimise costs, so again if you have competitive markets then you don’t need to look inside the firm.
This means we can focus on maintaining competitive markets and we don’t have to worry about what firms actually do or why they do it — it’s the elegant invisible hand of self-interest at work. If there are externalities or distributional issues then this will be picked up by other regulation or taxation, it is of no concern for antitrust.
But this didn’t make sense — any system regulating the accumulation of power must have something to say about externalities and distribution. Surely? It was green eggs and ham, if ever I had seen it; something didn’t add up.
Firms don’t profit maximise, they maximise shareholder value
The first assumption mentioned above is that firms maximise profits, which drives efficiency as firms compete for customers by lowering their prices and improving their products. This may be correct, in a general sense, but not all profit-seeking is made equal, and how a firm chooses to generate profits will depend in part on how those profits are to be distributed. Profits could be allocated to training workers or to investing in plant and machinery or to investing in R&D or distributed to loyal customers or even to suppliers. But according to the seminal article by economist Milton Friedman, it is the only proper goal of business, and indeed its social responsibility, to maximise profits for the benefit of the shareholders — a principle known as “shareholder value”. This will dictate which path to profits a company chooses.
Seeking to maximise shareholder wealth, as Milton Friedman commanded, will tend to put the interests of shareholders in conflict with those of consumers and society. For example, shareholders may prefer initiatives that deliver short-term returns — like the rebranding of an existing product or the sale of existing assets. Consumers would benefit more from profit-seeking decisions that are based on developing new technologies and innovations, and their timeline for receiving those benefits may be more generous.
This is a gross simplification because shareholders, like consumers, are a diverse bunch with diverse and divergent interests. Nevertheless, it remains true that only some strategies for maximising shareholder wealth, or maximising profits, are consistent with lower prices for consumers. One way to maximise profits, for example, would be to build a business model based on the extraction of “free” natural resources — anything from fossil fuels to “free” human effort, secured through employment or forced labour under unfair, dangerous or exploitative conditions — and then to emit the negative externalities out of the corporation and onto society. This company could then maximise shareholder returns by channeling all their ill-gotten profits towards shareholders, and as little as possible towards investment in workers, compensation for society, tax to the public purse or research and development to the benefit of consumers. The sticker prices for consumers may be low, but only because they don’t reflect the true costs.
Another excellent way to maximise shareholder wealth would be to pursue monopoly rents. Mariana Mazzucato has written about how shareholder value drives wealth extraction (and / or destruction) and not wealth creation — the financial crisis being a poignant example. In particular, as we well know, market power delivers unearned rents to monopolists and their shareholders, representing a direct transfer from consumers to producers (or perhaps, getting into the flow of Mariana’s worldview we might call them not producers or “makers” but instead “takers”). Although many monopolists do create valuable products this in itself should warrant no more than a normal return. Supra-competitive prices and the accompanying rents are not the reward for value creation, but rather the loot from having exploited a market failure.
Joseph Stiglitz enumerates the costs monopoly and rent extraction: (1) rent-wealth crowds out capital formation; (2) for a monopolist, the marginal rate of investment is less than the average rate of investment, because if they produce more output then prices might decline, so they accordingly invest less; (3) it results in distortions in the allocation of resources; (4) it stifles innovation; (5) monopolists seem to find ways to avoid paying taxes; and (6) money moves from the bottom of the pyramid to the top, where less is spent by the relatively much smaller class of people, depressing aggregate demand. And, the big one, rent extraction and the transfer of wealth worsens inequality. Rent seeking dressed up as consumer-friendly profit seeking is as far from welfare maximisation as you can get.
Inside good, outside bad
So much for the first assumption of benign, welfare-enhancing profit maximisation. What about the second assumption: not just that competition between profit-maximising firms drives efficiency on the market but that what goes on inside the firm is efficient in itself?
It turns out that the boundary we use to distinguish between conduct that takes place inside the firm, and the positive assumptions we place around that conduct, versus conduct that takes place outside the firm, and the presumptions of harm associated with it, are porous concepts with arbitrary boundaries and possibly tautological inferences (if it’s good it must be happening inside the firm and if it’s happening inside the firm then it must be good).
Sanjukta Paul has attacked this question with fervour in her recently posted working paper “Antitrust as Allocator of Coordination Rights” (see also the conversation about her paper, transcript available at Forbes, which is as close as you may get to an antitrust brawl). I had an inkling that the answer would have something to do with Transaction Cost Economics and made a note to myself to look into it. Happily Paul has done all the heavy lifting for me.
We can start with Ronald Coase, and his theory of the firm, later developed by Oliver Williamson, as well as none other than Robert Bork. Coase’s “theory of the firm” has been used in different contexts to explain why firms exist at all. If people can meet in free markets and freely transact why should there be any need for the business corporation? Coase’s insight was that forming a company would have a distinct advantage over repeated, ad hoc interactions between individuals on the market: it would limit the transaction costs of entering into new contracts every time you wanted to get something done. What had to be done through contracting on the market could be smoothly achieved with a kind of command and control in the hierarchy of a company.
And this insight played out in the real world in the 1950s, 60s and 70s. The 1950s were marked by the height of managerialism in the corporate world, with management taking the role of the central planner, mimicking the role of the government in the war effort. As Jeffrey N. Gordon writes: “The war was waged and won by huge centralized bureaucracies that were able to surmount logistical and planning challenges. These lessons could be applied to the private firm in shaping and managing its environment”. This led Arnold J. Toynbee to write in 1958 “[T]he connotation of the word ‘business’ is changing. Instead of its original association with the notions of enterprise and profit, it is coming to be associated in our minds more and more with the very different notions of administration and organization” (from Thinking Ahead, HARV. BUS. REV. Sept.-Oct. 1958, quoted at fn 183 at the previous link).
By the 1990s, firms began to disaggregate from bloated conglomerates into vast networks of supply chains, managed by the central firm and aided by the swift improvements in computing and communications that accompanied the 1980s.
Against this background, the relevance to antitrust of the theory of the firm can be seen most clearly in the treatment of vertical integration. Price theory had previously held vertical integration as suspect because of the potential impact on market power — for example the possibility of foreclosing a rival by blocking them from an essential input. For price theorists the primary economic benefits of vertical integration were any technological efficiencies that resulted from bringing production under one roof. But this couldn’t be used to excuse many other types of integration, of the sort taking place under the direction of kingdom-enlarging managers, particularly partial integration, where such technological efficiencies were not available.
As Paul explains, Bork revived Coase’s notion of transaction costs, and particularly the idea that hierarchical business firms are able to reduce the costs of contracting on the market by bringing such activities inside the firm, to support the idea that what happens within the single firm’s sphere of influence must be efficient. So we don’t need to look inside the firm to interrogate any arrangements within it because we will find only an efficient machine of resource organisation. Firms organise themselves to lower their costs — if they must consolidate their external activities and bring them inside the firm to do so, then so be it.
Paul pursues this line of thought with a different agenda to my own (although I am very sympathetic to the campaign she wages). I gather that she is interested in the differential treatment of conduct that occurs outside the firm as opposed to that which occurs inside, from the perspective that this principle is used to justify targeting otherwise beneficial cooperation. For example, there is no reason that the transaction cost reasoning could not be applied to organisations that are otherwise branded as “cartels”, like cooperatives and unions, to show that other forms of coordination, outside the firm, also reduce costs that may be passed on to consumers. Instead the transaction cost rationale is almost exclusively applied to mergers or in defence of intra-firm conduct.
Paul characterises this as part of the Chicago School campaign to remove fairness, dispersal of power, and a commitment to small enterprise from the purpose of antitrust in order to clear space for the Borkian model of focusing only on competitive free markets. By definition this means that any coordination outside the firm must be a market distortion with a presumed accompanying cost in terms of inefficiency.
The result is that we get rules like this: Uber is free to coordinate drivers’ prices but drivers cannot coordinate to bargain with Uber on pay. Coordination within the firm benefits from what Paul calls a “firm exemption”, and is only illegal if it satisfies the strict criteria of monopolisation or abuse of dominance.
If a firm chooses to expand its dominion then it is assumed that this must be for the pro-competitive purpose of reducing costs — just as big firms are given the benefit of the doubt that they must be using their economies of scale to benefit the consumer. The normal suspicions against firms expanding their control of aspects of the market are relaxed because the activities are either taking place within the firm or are vertical.
The overall impact on antitrust is that we have become credulous of the intra-firm efficiencies of coordination resulting from integration through consolidation or integration into the firm’s vertical hierarchy through contracting, but highly skeptical of (i) other coordination, especially if horizontal, inter-firm or inter-personal, as between separate market actors; and (ii) the possibility of monopoly, which is presumed to be rare because of contestability: a monopolist’s market position is only maintained in so far as they can maintain the efficiencies of integration and they must go out onto the market always in the shadow of the potential competitor.
What is interesting about this differential treatment of conduct inside and outside the firm is the selective vision of free markets that it supports: these markets are free from government intervention, state funded provision, and horizontal coordination but not free of monopoly, corporate consolidation and vertical restraints, which are all presumed to be compatible with free markets.
Paul points out that it is peculiar that we argue for transaction cost savings within the firm from first principles, but with little evidence that they ever or always exist. But instead of maintaining a healthy skepticism we have built a presumption of cost savings into our legal rules, whether they exist in real life in a particular situation or not.
The research on the existence or otherwise of scale economies is enlightening, and may or may not hint towards a similar pattern with supposed transaction cost savings. Zephyr Teachout quotes Walter Adams and James Brock: it is “the quintessential myth of America’s corporate culture that industrial giantism is the handmaiden of economic efficiency”, adding that “[a]fter decades of research, the virtues of very large size are unproven”. She cites various sources of research (see page 43) showing that firms operate either with flat cost curves or that diseconomies of scale set in quite quickly as size increases. Our assumption, settled in law, that economies of scale exist and can defend a merger or justify firm size ought to be reconsidered, and so too with the assumption that the firm organises itself in manner that reduces transaction costs, more so than can be achieved with external coordination.
Escaping the Upside Down
Going back to my original quest, we have the beginnings of an answer as to why antitrust is complacent about what goes on inside the firm: it assumes firms profit maximise for the benefit of consumers and it assumes that whatever happens within the bounds of the firm is efficient. In fact we can see that profit maximisation may be directed towards the benefit of the shareholders, not consumers, and the boundaries of the firm, and what may or may not constitute efficient activity, have been arbitrarily drawn. The assumption that firms organise themselves at all times to minimise transaction costs and increase efficiencies is dubious. Therefore the discipline of corporate governance, and its influence over how firms share the rents that they manage to acquire as a result of the competitiveness, or lack of competitiveness, on the market, has considerable relevance for antitrust. We continue to ignore the inner workings of the firm at our peril.
We can argue about the origins of antitrust law, and this seems to be a favourite pastime of American scholars in particular, but I believe that antitrust today should be about controlling the accumulation of power and the extraction (and therefore initial distribution) of rents. Paul mentions that the first corporations were formed under public charter; they were in effect granted a monopoly over coordinating certain assets and resources but this was tied to some explicit public purpose. In antitrust, that sense of public purpose has been reduced to a responsibility to compete on the markets, including those which the firm dominates, with no responsibility for the outcomes. But we can see that not only will this not necessarily of itself bring about low prices but low prices are not all that matter.
It turns out that we have our analysis completely upside down and inside out. We have been looking at monopoly and market power purely in terms of allocative efficiency, assuming that any cost savings resulting from a merger or other consolidation will be passed on to the customer. But in fact these “efficiencies” (i.e. profits) are earmarked for shareholders, and the drive for shareholder wealth maximisation drives rent extraction. In the pockets of shareholders these rents multiply and a mischievous alchemy takes place whereby market power and economic rents are transformed into economic power and political power.
So in the Uber example, we assume that Uber coordinating drivers’ prices is fine for consumers whereas if Uber drivers were to unionise consumers would suffer from higher prices. But an equally plausible outcome would be that shareholders would lose out because drivers would wrest control of those rents from shareholders’ hands without necessarily harming consumers. Meanwhile Uber accumulates substantial VC investment off the back of growing market power, rents, and political power, becoming such an entrenched market player that users protest when authorities attempt to regulate it.
This analysis has implications for the kinds of factors that should be relevant to antitrust analysis and the kinds of remedies that the competition authorities might legitimately impose. There has been relatively little research at the interface between competition and corporate governance but both are concerned with the impacts of the firm’s conduct and, critically, in whose interests the firm must act and the law must protect. We cannot continue to separate a company’s market power from its governance responsibility; doing so has thus far well and truly Seussed us.
