The "market failure" idea in policy debates.

At the Silicon Flatirons conference at the University of Colorado Law School last weekend, some panelists made the commonly-heard point that regulators should only regulate in cases of “market failure.” This argument has irritated me for years but when I have tried to explain this I get the impression that people think that the alternative to a “market failure” framework is that regulators ought to regulate for the sheer joy of it. Then my boss, Public Knowledge legal director Harold Feld wrote this blog post on the tradeoffs involved in spectrum policy (namely, competition versus efficiency/fairness) which inspired me to get off my virtual duff and write this. Now when people disagree with me I can point them to what exactly it is they’re disagreeing with.

A “market failure” can be any number of things. It may be an instance where parties to a transaction are causing harms to third parties, or an instance where parties arrive at a result that is not Pareto efficient–that is, where at least one of the parties could be made better off without making any of the other parties worse off. It might be the failure of the free market to provide the kind or quantity of goods that consumers demand. (Although how would you know?) Or perhaps it is a failure of a free market to provide some public good.

Proposed fixes to market failures can include regulation, government provision of services, or the elimination of whatever economic or legal glitch is causing the usually bug-free market to fail in this instance. Or the favorite of many–creating some kind of pseudo-maket around some kind of pseudo-property (“market mechanisms”) to try to align people’s self-interest with the common good (e.g., cap and trade, intellectual property laws).

“Market failure” is so broad a concept that it’s not necessarily useful in policy discussions (particularly those involving lawyers, politicians, and engineers, as is common in Internet and telecommunications policy debates) whatever its usefulness to economists. Generally advocates on one side of any given side of a policy debate will point to some market outcome they do not like and deem it a “failure.” The other side will counter, “That’s just the market working.” The discussion goes nowhere. There is no way to determine whether or not a given market failure actually is a market failure and, even if there is agreement on that point, what to do about it.

Policy advocates might do well to consider alternate framings. For instance, I do not consider a failure of parties to a transaction to provide sufficient positive externalities to third parties to be an instance of “market failure.” Nor is a party charging a profit-maximizing price or otherwise behaving in self-interested ways a market failure–for example, how is it a market failure when a profit-seeking company declines to lose money by providing broadband service to a market with too few potential customers to provide a return on investment? This is a failure of markets, perhaps, which seems to me to be distinct from a “market failure.” (A “market failure” assumes that the market is the best way to provision some kind of good, but it just didn’t work out this time. But a “failure of markets” framing would allow that there are some goods that markets are not suited to provide.)

You could chop logic and say that a failure to provide a desired good is equivalent to a harm, and I reserve the right to argue this under a different framework if that’s the one that the relevant decision-makers buy into. But I don’t think that most of the “consumer protection” issues I am concerned with are best conceptualized as market failures. Again, I’m not denying that they can be, but that the steps required to do so make the framing cumbersome and more appropriate to academics than to lawyers or activists.

Rather, I think it makes more sense to simply accept that markets don’t deliver all of the things we want them to, if left alone. In particular, relevant to the things I am concerned with, a healthy, functioning telecommunications market will tend toward concentration, and will lead to dominant providers behaving in ways that users would rather they didn’t. Economies of scale, network effects, and so forth mean that most telecommunications markets are natural monopolies or something close to it. This is what you’d expect to happen, not a “failure” caused by some subtle flaw in the legal system.

Actually I need to digress on the “something close to it” point. If a given provider were a true monopoly then the “one monopoly profit rule” would kick in. The monopoly provider could extract its profit-maximizing rents from simply charging a high price and would have no incentive to, for example, muck about with network traffic to extract extra tolls. But there are no “perfect” monopolies just as there is no “perfect” competition. DSL is slower than fiber but will do in a pinch if the fiber provider goes crazy. Or you can check your email on your phone. These products are not perfect substitutes for each other but they provider a check on some of the most egregious abuses that a given provider may engage in. Thus in a highly concentrated market different competitors and partial competitors have to get fancy if they want to make the largest possible profit. This is how we end up with people paying “usage” charges for a non-scarce resource, net neutrality violations, early termination fees, and a myriad other nickel-and-diming schemes that generally fail to rise to the level of outright consumer revolt but which, in the aggregate, suck. (Additionally, in concentrated markets coordinated effects are easier–different competitors are more likely to silently agree that eliminating early termination fees is in none of their own best interests, for example.)

This situation is not necessarily a “market failure” as I described above but simply the expected market outcome given the nature of telecommunications. Nevertheless I don’t like that outcome and think it should be different. I don’t particularly care if the telecommunications industry is as “efficient” as it can be, and I want certain positive externalities–such as neutral networks and simple, fair billing schemes–to happen, even if those who would provide them have no incentive to make them happen. The failure of the market to provide these good things is not a “market failure.”

I know that the objection to this would be, “Who are you? The market reflects the aggregate revealed preferences of society, not some wonk’s vision of how Things Ought to Be. But if society’s revealed preference is for concentrated telecommunications markets, this is only in the same sense that my revealed preference is to eat a lot of candy and cake. It’s not good for me and on an "executive” level it’s not my preference, yet it’s my true revealed preference when you look at my caloric intake. Similarly if you actually ask most people–most consumers, most voters, and most politicians–what outcome they want for communications markets the answer will likely be either “competition” or at least all of the good things that competition is supposed to provide. Many people profoundly disagree on this point, but I think it is appropriate for a democratic government to fiddle with the market, within reason, to try to achieve some result that people want but otherwise can’t get.

In telecommunications, at least, the best way to get to the outcome that I think most people (apart from telecom incumbents and their investors) want is to boost competition beyond its natural, “efficient” level, through measures such as line-sharing, structural separation between infrastructure and retail, and regulated interconnection between networks. The theory behind this is that if competition between and diversity among providers is greater than the market would provide on its own, these competitors will have an even greater incentive to make their customers happy than competitors in a concentrated market. Bad practices may continue but at least the people who find them truly irritating can more easily switch providers, and not only to some kind of close substitute. The second-best way is to make various bad practices illegal.

(Of course even a perfectly competitive market is no guarantee that every last social good will happen–the whole point of a positive externality is that neither the provider nor the user benefit from it and thus neither of them have an incentive to seek it out. But at least with telecommunications all of the benefit of a neutral network does not accrue to edge providers themselves–users benefit when edge companies benefit, and in a competitive market a user are in a better position to demand even the indirect benefits caused by a provider behaving neutrally toward even those services a user may not use.)

There are similar arguments to be made in places other than telecom. One might argue that fragility in the financial system caused by excessive concentration is a “market failure” that inflicts harms on those future people who have to suffer through the next collapse–or simply say that a stable financial system is a good thing whether or not the market provides it.

Instead of a “market failure” framework, then, I think it would be better if policymakers and non-economist advocates stopped using fancy concepts in imprecise ways and simply focused on: (1) What do we want, (2) How can we get it, and (3) What are the tradeoffs. This gets to the crux of the matter and does not buy into the idea that the only goods society may have are ones provided by functioning markets, and that the only job of government is as a market mechanic spot-fixing various “failures.”