In the church of economic theory, as in that other church, the central symbol of the faith is a cross. Only this one is tilted and looks like an "x" not a "t." As any communicant learns early on, the x represents supply and demand "curves" (usually portrayed as straight lines). Demand slopes down to reflect the fact that people tend to want less of something as it gets more expensive. Supply slopes up because … because, well, we'll get to that. But where the two lines cross is the blessed point of "equilibrium," where the price is exactly what is needed for supply to equal demand. At that point markets clear, utility is maximized, lions are beaten into plowshares, bread walks on wine, and so on. And it only takes a few more curves and diagrams to show how free markets inexorably lead you to that point.
Having been inducted into the faith in college, I'm a fairly devout believer in the basic doctrines. But I have always been troubled by doubts on one item: In my innermost heart, I wonder if the supply curve really slopes upward. (There, I've said it.) These doubts affect many other members of the church. Most suffer in silence, while some join schismatic sects such as the notorious Galbraithian heresy.
The problem is reconciling the upward-sloping supply curve with another key doctrinal concept known as "marginal cost." Marginal cost is the cost of producing one more unit. Doctrine holds that in a free market, competition will make price equal to marginal cost, because anyone attempting to charge more will get undercut by a rival, and anyone who charges less will lose money. Each producer increases production until the cost of producing the next unit starts to exceed the price she can get. Essentially, rising marginal costs determine the supply curve.
Now, I yearn for market-clearing equilibrium as much as the next fellow. But the whole doctrinal contraption involves two difficult assumptions. First, marginal costs have to be increasing (to avoid the embarrassment of supply and demand curves that never cross). Second, at the point where supply and demand do cross, marginal costs have to be above average costs. That is, if the price you get for every widget you produce is determined by the extra cost of producing one more widget, that price has got to be high enough to cover all your costs.
Why am I dragging you down these dank doctrinal corridors? Because the news these days is full of controversies where the basic problem is that marginal costs do not rise. They start out below average cost and stay that way. This is why airlines, for example, charge $105 or $501 for the same seat on the same route. The marginal cost of filling an empty seat on a plane that's going anyway is almost nothing. (To be precise, it's 53.7 cents, which is the cost of shredding used boarding passes and molding them into "vegetarian lasagna.") So $105 covers your marginal cost on the last seat, but you need $501 from some people to cover your average cost.
Prescription drug prices are another hot example. The cost of producing the billionth pill is nothing compared with the cost of developing the first one. Some people must pay a lot more than marginal cost or new drugs will never be developed. At the same time, any particular sale even slightly above marginal cost is profitable—and, in fact, reduces the price for everybody by absorbing some of the development costs. So it's irritating but not necessarily bad for the United States that the same drugs sell for less in Canada and Mexico.
In a speech last May, Treasury Secretary Lawrence Summers identified low marginal costs as a central feature of the "New Economy." Developing a computer program is expensive; putting it on a disc is cheap. In their purest form, information and knowledge—the basic products of the new economy—cost nothing at all to reproduce and distribute, and never get used up. You can give information to me and still have just as much of it yourself. In fact, because of the so-called "network effect"—the more people who use a software program, the more valuable it is to each user—some products of the new economy may even have negative marginal costs.
In such a world, the basic tenets of economic theory can no longer tell you what prices will be or what they should be. Nor can church doctrine guarantee that an invisible hand will produce the most efficient result. Summers noted that if marginal costs are below average costs, competition that drives prices down to marginal cost is ruinous. Therefore, the only way a firm can make money is to have some degree of monopoly power. Critics of antitrust have seized on Summers' analysis, generally failing to note that it undermines the case for free markets at least as much as the case for government intervention.
And now that it's fashionable to admit that marginal costs don't always go up, maybe it's permissible to ask aloud how often they ever do. Even an Old Economy business like the auto industry spends billions of dollars designing and preparing to produce a new model. Can the marginal cost of producing one more car, leaving aside these startup costs, actually be more than the average cost of all cars of this model, startup costs included? If not, what is left of the one true faith? And yet somehow it's hard to believe.