Steve Reads

(Attention conservation notice: 1800 words, 1900 if you count a block quote in the middle, on Joe Stiglitz’s economic treatise. If you know Stiglitz’s work, consider Whither Socialism? a summing-up — though only a temporary one; the man seems to be everywhere these days, and will probably churn out economics until the sun goes nova. If you’re unfamiliar with Stiglitz, he’s famous for exploring what happens when one party to a contract knows more than another, and what effects this might have on all sorts of things. What happens in a company, for instance, when you know how much effort you put in every day but your boss doesn’t? Scale out: what effect might this have on an entire economy? Thar be the doctrine of efficiency wages. Stiglitz won the Nobel for spinning out this work.)

This book lacks a subtitle. If it had such a thing, it might be A History Of The Last Thirty Years Of Joe Stiglitz, Written From The Perspective of Me, Joe Stiglitz. I didn’t count Stiglitz citations in the bibliography, but just the ones whose sole author or first coauthor is Stiglitz run to several pages.

Perhaps this is excusable and expected. Whither Socialism? is an extended attack on perfect-competition models, delivered by a fellow who’s been poking holes in those models for most of his career. I suppose he has the right to draw from the deep well of his own research. I must note, though, that it’s off-putting to see lots of objective-sounding phrases like “as has been shown,” then to look back in the bibliography and see that it “has been shown” by Joe Stiglitz himself.

Whither Socialism? spends the vast bulk of its time spearing a set of economic theorems known as the fundamental theorems of welfare economics. These are really quite beautiful constructs, are completely right within the set of assumptions that they make, and bear almost no relation to a world that any human being has ever inhabited or will ever inhabit. Still, they’re quite worth understanding, for a few reasons. First: mistakes are valuable. We tend to make fun of overthrown doctrines; I pity poor Jean-Baptiste Lamarck, whose work was false but which Charles Darwin saw fit to refute. At least Lamarck had the decency to be rebuttable. Much of what we believe is not even wrong. So while I think that two of the main pillars of free-market economics — the efficient-market hypothesis and the fundamental theorems of welfare economics — are obviously wrong [1], I applaud them for giving us a starting point for discussion. If they’re wrong, they’re wrong because their assumptions are wrong. It then behooves us to explain why those assumptions are wrong, and to replace them with more-realistic assumptions from which we can derive more-realistic stories.

The fundamental stories of welfare economics tell us two things. First, a free market — suitably defined — is Pareto-optimal. That is, under certain assumptions, it is impossible to make anyone within a free market better off without making someone else worse off. “But wait!” you might be saying; “Does Pareto-optimality hew to our notions of fairness? I believe that we ought to make the poorest people better off before we make the wealthy even better off than they already are. In your free market, it may be true that we can’t make anyone better off without making someone worse off — but distribution matters: I’d gladly make a wealthy person worse if it means making a poor person better.”

The second fundamental story rides in to your rescue! It says that in a free market — suitably defined — we can achieve any distribution of wealth that we care to achieve. So the free market not only distributes resources optimally; it also empowers us to obtain the fair world that we want.

So what’s wrong with these stories? First, they assume a “complete set of markets.” That is, you should be able to place bets on any future state of the world. But as Stiglitz points out, this completely ignores the role of innovation: how can you bet on future states of the world when you don’t know what future states of the world will be? In one state of the world, for instance, Apple sells 10 million iPhones. Could I have bet on Apple’s selling 10 million iPhones back in 1990?

The other major pillar of the fundamental theorems is price-taking behavior: everyone involved in a market must believe that he — as a buyer or a seller — can have no effect on prices. In particular, every seller must believe that if he raises his prices by the tiniest amount, everyone will leave his company and buy from his cheaper competitor. This is called “perfect competition.” It obviously doesn’t obtain in the real world at all, except possibly in markets for a few agricultural commodities. It doesn’t hold for a number of reasons. First would be legal constraints: if I want to make a product exactly like Coca-Cola, and compete with it on a level playing field, I will need to package my Fake Coke in cans that look exactly like Real Coke’s. The Coca-Cola Corporation, and the federal government, will allow this to continue for approximately an hour before my company is burned and plowed under, and the land where it once lay is sown with salt.

Perfect competition is unlikely to arise in any industry where there are significant economies of scale. If by investing a significant amount of capital, I can reduce my per-unit costs below those of my competitors, I gain a significant advantage that can quickly snowball. I don’t have the arsenal to argue it at the moment, but it seems clear that perfect competition assumes constant returns to scale.

Much more fundamentally, and pervading Whither Socialism?: according to Stiglitz, prices just aren’t a significant factor in most corporate decisions. Transactions at large corporations happen almost entirely within the firm, rather than between the firm and its suppliers or customers. Prices are not involved in internal communications at all. General Motors doesn’t ask its final-assembly division to buy engines from another division and bodies from a third; it commands the construction from a position above all of them. (It’s worth wondering, by the way, why no one seems to take that third option. The normal dichotomy we hear about is the-firm versus the-market. Why not the-market-within-the-firm? Stiglitz tells us that there are few if any examples of this working in practice.)

Contracting, also, can’t carry all the weight normally ascribed to it. Here I’ll hand the mic to Joe:

An example of the complexity of the product space was recently provided by the U.S. Defense Department, when it put up for bidding a standard, white T-shirt, the kind of commodity that can be purchased in any clothing store for a few dollars. The specifications were thirty small-print pages. Even then I suspect that the product was incompletely specified. Of course most consumers do not have to articulate completely what it is that they are buying when they buy a T-shirt — suggesting that there is a fundamental difference between the way actual markets work and how they are envisaged to work in the market socialist model.

You might ask, “Socialism? I thought we were talking about capitalism here.” Whither Socialism treats them within the same framework, because the formulators of market socialism — people like Oskar Lange — did the same. Market socialism could work as well as capitalism, they said, because all capitalism needed was an adequate set of price signals. Market socialists, that is, were guided by the same fundamental stories that capitalists were. The central planning bureau, said the socialists, could work out demand schedules as well as any capitalist firm, because it would be receiving the same price signals from the market. Stiglitz deals a blow to any naïve view of socialism or of capitalism: price signals alone won’t do the trick. A market-socialist central planner relying only on price signals would soon find itself writing 30-page contracts for t-shirts.

Well, then, how do firms communicate? How do consumers buy their t-shirts? That’s where the theory gets interesting, and that’s where Stiglitz has spent his career. In a world where at least one side of a business transaction can’t know whether the other is lying to him, reputation counts for a lot. The economic discussion of reputation is probably very old, but for our purposes it may have begun with George Akerlof’s article on the “market for lemons”. Why is it, asked Akerlof, that a brand-new car, driven right off the lot, instantaneously loses a sizable fraction of its value? The problem, as Akerlof explained it, is information asymmetry. Imagine that 50% of used cars are “lemons” (low-quality) and 50% are “mint,” and that only the seller actually knows which type of car it is. Let’s say lemons are worth $10,000, while mint cars are worth $20,000. A buyer who comes to a used-car lot with no other knowledge of the seller’s reputation should assume the same split of lemons and mint cars. So the average car in front of him is worth $15,000. He will, accordingly, pay no more than $15,000. Hence no mint cars will sell, and they’ll be systematically driven from the market. Now the only cars left are the $10,000 lemons. Down we go, through more and more lemony used cars.

Of course this doesn’t quite happen in reality. One reason is that we’ve formed institutions to get around these obvious market failures. One such institution is the warranty: you’ll fix my car for the next few years, no matter how lemony it is. But then who warrants the warrantor? How do I know, that is, that your warranty is worth the paper it’s written on? We have laws that help with this, and of course there’s word of mouth.

The point is that reputations and trust pervade everything we do, and that these trust relations cannot be conveyed through prices alone. Any economic model that rests entirely atop prices doesn’t capture the living, breathing world of economic relations that we live in.

If the fundamental stories of welfare economics, then, don’t capture what makes capitalism such a powerful, dynamic force, what does? Because surely there is a difference between market socialism — socialism of any form, really — and capitalism. Competition is surely a hugely important part of the difference: multiple companies fighting one another to win your money are likely to treat you better than a single government agency to which you’re compelled to pay taxes.

Here again, though, we have to temper the classic story. Competition is never perfect competition. When companies face nontrivial returns to scale, we’re likely to get substantial concentration. Then there’s John Sutton’s book Sunk Costs and Market Structure, which I’ve only gotten a little ways into, arguingthat advertising becomes a sunk cost which new entrants to a market are forced to match. (Imagine a new cereal company competing with Kellogg’s and Post.)

Private companies also have the not-so-small matter of the “hard budget constraint”. In a word: if they lose money consistently, they go out of business. The great danger of government-run industry is the “soft budget constraint”, which is to say: if they lose money, the government just prints more.

Readers who’ve been awake at any point in the past six months might be raising their hands here. CitiCorp and AIG don’t seem to be living and dying at the razor-edged hand of the hard budget constraint. Quite so. And here lies a point (theorem?) of Stiglitz’s that particularly grabbed me, in the light of recent news: a government cannot credibly commit to not bailing out a firm. The assumptions behind such a theorem would have to be important. What if the antitrust division of the government ruthlessly forced banks to stay below a certain size? We’ve recently heard a lot about “systemically important nonbank financial firms” or “large, interconnected, non-depository financial institutions” or, more colloquially, banks that are “too big to fail.” What if we made it a policy never to let a company become too big to fail? A commitment to such a policy may, itself, not be credible, but in any case: if it’s a theorem that the government can’t credibly commit to letting companies fail, I need a bit more argument.

Despite its title, then, and a lot of nuggets of advice for newly democratized Eastern European countries, Whither Socialism? is actually a careful, multifaceted reworking of fundamental economic doctrine. Think of it as Econ 101 for the new century.

[1] — then again, this may be the mathematician’s definition of “obviousness”:

A mathematics professor was lecturing to a class of students. As he wrote something on the board, he said to the class “Of course, this is immediately obvious.” Upon seeing the blank stares of the students, he turned back to contemplate what he had just written. He began to pace back and forth, deep in thought. After about 10 minutes, just as the silence was beginning to become uncomfortable, he brightened, turned to the class and said, “Yes, it IS obvious.”