Market Failure 1: Ideal Markets (Why you should care about spherical cows)

To the naive mind that can conceive of order only as the product of deliberate arrangement, it may seem absurd that in complex conditions order, and adaptation to the unknown, can be achieved more effectively by decentralizing decisions and that a division of authority will actually extend the possibility of overall order. Yet that decentralization actually leads to more information being taken into account.

Friedrich Hayek, The Fatal Conceit: The Errors of Socialism

To understand how markets fail it is important to understand how markets work. Without that, any discussion of market failure will degenerate into subjective arguments about what the market should or shouldn’t be doing. Before we get into market failure we need to have a concrete definition of market success.

To get an understanding of what properly-functioning markets do, read I, Pencil. If you discuss politics on the Internet you have probably been pointed at this before (it’s a favourite of libertarians), but even though its writing style might best be described as “overwrought” I’d like you to read it, not as a weapon in the rhetorical war between markets and government (although that is clearly what the author intended it as) but rather as a description of what the market for a relatively simple good has to account for when establishing a price:

• How many pencils people want, and how much they are willing to pay for them.
• The availability of each of the natural resources that go into pencils, and the time and effort required to extract them.
• The time and effort taken to turn these resources into pencils.
• The time, effort and resources required to transport the pencils and pencil-precursors from extraction, to manufacture to retail.

Note that markets have to solve this problem simultaneously across all goods since wood and aluminium and graphite all have alternative uses they may be put to, so you can’t just solve for pencil production and go home. For that matter, it needs to be able to solve for goods that don’t exist, but could.

Also note that markets have to solve this problem without any central direction. Nobody on Earth knows all of the things listed above. Somehow we are getting a suitable number of pencils being produced without anyone having the requisite knowledge to make an informed decision.

So how is the trick performed? How do you make a system that is smarter than its constituent parts? The key is to take global problems and turn them into local ones. Termites manage to make complex nests and yet no one termite has the cognitive capacity to understand the architecture of what it is creating. They do it by applying a set of very simple rules and signals that ensure each termite only needs to worry about the part of the nest it is working on – the global problem of designing a nest is reduced to a set of simple, local problems by using a signal to coordinate the disparate actors. This is also how markets work and the signals markets use markets are prices.

Prices are important because talk is cheap. If you just ask everyone how much they want something, basically everyone will say “a lot” and that gets you nowhere. First off, how do you weigh one person’s statement of desire against another’s? Secondly, how do you determine if they are inflating their statement of desire? To be useful as a resource allocator, a statement of desire must be credible (it can be expected to reflect the person’s actual preferences) and commensurable (you can meaningfully compare them). Prices fulfil both of these requirements very well.

If you are a rubber producer you don’t need to balance the global desire for pencils against tyres or other rubber products, you just sell your rubber to whoever will pay the most for it. In turn, pencil makers balance the prices of their inputs for the price they can sell their pencils. And the pencil retailers determine how much they can pay for pencils, based on how many pencils they can expect to sell, and for how much.

In addition, since the market participants are trying to make a profit, the “do what makes you the most money” rule is incentive-compatible. It’s a rule the market participants are inclined to follow anyway, so there’s no need to try and enforce compliance. This is what Adam Smith meant by the Invisible Hand – the market takes self-interest, a problem in most social systems, and co-opts it to improve the operation of the market.

OK, that may sound good in theory, but how do we evaluate it in practice? If the output of the market is so complex that we can’t independently check what it does are we stuck with just shrugging off every anomalous result by saying “the market moves in mysterious ways”? For economics to be more than ideology (or that matter theology), we need to be able to evaluate market performance in practice.

The key is to evaluate the process instead of the output, if the process is working fine then we can most likely rely on the outputs. If the process is flawed, then identifying those flaws is the key to working out what the problem is and fixing it.

The tool used to do this is the First Welfare Theorem, a mathematical description of the circumstances under which markets would deliver perfectly Kaldor-Hicks efficient outcomes. Kaldor-Hicks efficiency is what economists normally mean by efficiency (except when we’re talking about the Efficient Market Hypothesis). For a situation to be Kaldor-Hicks efficient there has to be no possible action that could make anyone in the society better off without making others sufficiently worse off that the winners couldn’t compensate the losers. This makes Kaldor-Hicks a utilitarian criterion for evaluating a social system, and while it isn’t necessarily the only thing you should care about, it is very important as it reflects the preferences of the people in your society.

The conditions required by the First Welfare Theorem area little technical, but in broad terms for markets to be perfectly efficient the following things need to be true:

• The costs and benefits of a transaction accrue to those who are engaged in the transaction – that anyone who is gaining or losing from another’s activity has a chance to bargain with them.
• No one player in the market, be they buyer or seller, may have enough influence to materially affect prices through their own decisions.
• People need to have some baseline level of information and decision-making ability.

Having read that list, you may be ready to object that the real world doesn’t look much like that. And you’d be right. The point of the First Welfare Theorem is to give us a model of a perfect market, so we can identify the flaws in real-world markets. We call these flaws Market Failures, and they are going to be the topic of discussion for most of the remainder of this series.

Next week, we’ll get into the first of these Market Failures – externalities.