Capitalism Stripped Bare
Competition is the central pillar supporting modern American industrial policy. Consumers love having multiple suppliers compete for their business. But investors would rather own shares of monopoly suppliers who exploit their customers.
Since most of us are both consumers and investors, we get tugged in both directions; we adore the hardworking competitors with whom we trade and the profitable exploiters in our portfolios. So it’s not so strange that sometimes we get confused.Industrial organization and antitrust, the economic and legal studies of competition, the discontents who attempt to destroy it, and the regulation that prevents its destruction, can guide us through our confusing loyalties. IO studies the ways that industries are organized—how large the firms are, how many of them compete (or how concentrated the industry is), and the impact that size and concentration have on competition, prices, innovation, and consumers. These economic studies suggest that only some industries are highly competitive. In others, a small number of firms are disproportionately powerful. That’s where antitrust law comes in. Many branches of the law protect the small and weak from the large and powerful. Antitrust law fits neatly into that category. Over the years, IO economists have theorized and observed that some market structures provide dominant producers with enough power to exploit their many small customers. Antitrust law protects those consumers. Antitrust litigators, regulators, and enforcers police the free market system. They identify potentially dangerous concentrations of power, monitor the ways in which this power is used, terminate illicit exertions of power, and punish those who would callously exploit the consuming public.
The consuming public isn’t always grateful. We’d prefer for the market cops to enforce the antitrust laws vigorously only against companies whose products we buy, but to ease up on the companies in our investment portfolios.
The government, for example, sought to discipline Microsoft in the name of consumers of personal computers and software. Microsoft’s shareholders, many of whom were also consumers of personal computers and software, were less than thrilled. These same investors, however, also discovered a new uncharted investment arena called cyberspace, found parallels to Microsoft in each of its spaces, and ran out to invest in the Internet’s inevitable monopolists. The consuming public thereby abdicated its consumption role to recast itself as the investing public. This masquerade worked for a while and created enormous paper wealth; we valued small, weak companies as if they were huge, powerful, inevitable monopolists. But before long, most of us remembered that we actually consume more than we invest, and decided to treat all Internet companies outside our portfolios as if they were small and weak. Not surprisingly, that was the beginning of the end. It left us with regrets, remorse, and deflated portfolios. But more than that, it left us confused. And it’s that confusion that IO and antitrust can address.The story of the information sector is a story of capitalism, and competition is the cornerstone of capitalist systems. In the information sector, we’re all hard-core capitalists—innovative, productive, and above all competitive. The information sector is capitalism stripped bare, capitalism in a pure, raw, uncut, unadulterated form. Or at least, it would be if not for some powerful discontents.
Capitalism’s striptease for the information sector bares some of its most basic rules. First, profit is a powerful motivator. Second, buyers like to pay low prices for quality goods. They’re happiest when many sellers offer high quality goods at low prices, each hoping to make the sale. This competitive state of the world maximizes consumer welfare—it’s the best of all possible worlds for consumers. Third, sellers like to charge high prices for low-quality goods. They’re happiest when they can destroy their competitors, dominate their industries, and force consumers to overpay for substandard goods.
While consumers describe such over-charges with a more colorful term, IO calls this rent extraction. Antitrust economists and lawyers scrutinize such uncompetitive behavior very closely. Fourth, in most cases middlemen are between buyers and sellers, adding both costs and values to the transactions. Technology often allows producers and consumers to eliminate these middlemen; middlemen tend to fight against these advances. Fifth, and most remarkably, transactions close at the “right” price and resources flow to where they can be used most efficiently—as long as markets stay competitive.
So how realistic is it for markets to be “competitive,” in the sense needed to keep prices right and resource allocation appropriate? The two places that come closest to the competitive ideal may be freshman economics texts and the information sector. Elementary textbooks generally assume markets where many small buyers face off against many small sellers. Individual buyers and sellers are all helpless in their desires to shape the market. Prices fall where they should fall, and any deviations caused by odd buyer or seller behavior are ephemeral. Prices always return to where they should be.
Suppose that you’re a producer out to sell your goods at the highest price possible (that’s the profit motive in action). You look at your production costs, multiply them by ten, and launch your product. Not bad, you think. All I need to do is sell it, and I’ve cleared a 900% profit. Your competitor gets the same idea, but since he’d be willing to settle for an 800% profit, your customers all run to him. You counter by reducing your price below his, he counters back, and so on and so forth until you’ve dropped your price to a penny above your costs. Your competitor drops his price to equal production costs. Checkmate. The best that you can do is match him. If you try going lower, you’ll lose money on every sale. Prices simply can’t drop any lower—unless and until someone discovers a more efficient method of production.
Now let’s add a twist. Suppose that producers need to buy an expensive machine to enter the business, but that once they’ve got the machine, they can crank out finished products from inexpensive inputs. Now where should the price fall? Well, since the producer had to lay out the “fixed” cost of the machine just to get into the game, it’s mostly irrelevant to his thinking going forward—though he would like to be able to recoup his initial expense at some point. This need to recoup total costs differentiates the producer’s long-run thinking from his short-run calculations. So let’s think about the short run first and worry about the long run later.
In the short run, the “variable” input costs needed to produce each additional unit at the margin dominate the producer’s forward thinking—leading to a model known as “marginal cost pricing.” Any finished unit that the producer can sell above his marginal input costs represents a profit on that unit. In many industries, marginal costs differ with the size of the production run. But there are exceptions. The cost of filling an additional seat on an airliner about to take off, or an additional container on a cargo ship set to sail, is close to zero—as long as there are vacancies on the plane or ship. Information products thus follow a pattern more common in transportation than in manufacturing. It’s quite expensive to generate the first copy of a program, but every copy thereafter is virtually cost-free. And unlike planes or ships, information products face no capacity constraint; it’s meaningless for them to be “full.” This observation actually leads to the special-case model of freshman textbooks—the constant variable cost model. For large numbers of information-sector producers, marginal costs equal average variable costs. Under any case of the marginal-cost pricing model, though, market dynamics set prices. Competition pushes prices down toward marginal cost, and the profit motive keeps prices from falling below marginal cost.
These considerations dictate that a product’s price should equal its marginal cost of production—at least as a theoretical matter in a perfectly competitive market and until we start thinking about the long run.From the other side, buyers determine sales volumes. Think about lining up buyers in order of their enthusiasm. The most eager buyers are willing to pay the most for the product; the least eager are willing to pay the least. The seller, of course, doesn’t known which buyers are most eager, and therefore fixes a single price for all potential buyers. And since the seller wants to produce one unit for each buyer willing to cover his input costs on that marginal unit, everything just falls into place. The price will drop to the marginal cost of production, and every buyer willing to pay that price will be satisfied.
Let’s take a concrete example. A good software-development team requires at least a bit of equipment and many smart people who could be earning good money elsewhere. Other information products, such as digitized songs or movies, can be even more expensive to produce. The fixed costs of these products are large, but the creation of a second copy of a bit string is close to free. Information-sector products thus tend to have high fixed costs and zero marginal costs. And therein lies the first key to the economics of the information sector: digital products should all be free.
Now, that prediction can’t survive for long, but the simplified textbook model has more to add before either the real world or the long run impose a touch of realism on its less-plausible predictions. Even theoretical prices don’t quite fall to the producer’s marginal costs. While buyers (or consumers) and sellers (or producers) are always essential, they typically need help transacting their business. Sometimes, for example, the buyer calls the seller to place her order, and the seller ships the product ordered. In that case, a phone company and a trucking company entered the fray, and both deserve to be compensated.
Where will the money come from? Since competition already drove the producer’s price to its bare minimum, the consumer must foot the bill. In other words, the consumer’s price just rose above the producer’s marginal cost. But the least willing consumers—those who were willing to pay only the producer’s marginal cost of production—would prefer to let the deal fall through than to absorb this extra cost. In the presence of such transaction costs, consumers lose because prices rise and producers lose because sales volumes decline.The myriad transaction costs prevent our real world from achieving the competitive ideal. Phone bills and shipping costs are but two of the more obvious. Many transaction costs stem from the difficulty of collecting information. Producers may not know exactly what features consumers want. Consumers may not know which vendor is offering the best price. The necessary information costs can make transactions much more expensive. But these costs all drop—dramatically—in the information sector. And therein lies the second key: the information sector reduces transaction costs.
Ronald Coase won the 1991 Nobel Prize in Economics for pushing transaction costs onto economists’ radars. His famous Coase theorem suggests that in the absence of transaction costs, all resources would flow to where the overall economy achieved maximum efficiency.1 So any reductions in transaction costs should be great for the economy and should make everyone happy. But then again, nothing ever makes everyone happy. While eliminating transaction costs would certainly help producers and consumers, the phone company and the trucking company would be less than thrilled. After all, they provide important services, they employ a fair number of people, and they live for transaction costs. Reduced transaction costs reduce their revenues. And therein lies our third key: middlemen will fight against information technology to reimpose transaction costs.
But let’s not get too upset at these callous middlemen. After all, many of them exist for valid reasons, provide valuable services, and are pillars of our economy. Many of them are even pillars of the information sector. Because a basic problem lingers in the world beyond the textbook—the theoretical prediction of free digital goods. If charging for information products is impossible, information producers must find a new way to recoup their investment in product development. Somewhere along the line, producers must recoup their fixed costs. Otherwise, no one would ever invest, and we’d never get any good information products. That’s where we bring producers’ long-run thinking back into the equation. Transaction costs provide a way out of the dilemma. We’ve used the law to impose an artificial transaction cost known as an IP right. We gave information producers the right to sue anyone who uses their creations without permission. That legally imposed transaction cost means that everyone but the innovator must face an additional cost: either a license or a potential lawsuit. IP rights thus give innovators a distinct cost advantage over their rivals, and they make it impossible for competitors to bid the innovator’s prices down to the innovator’s marginal costs. They thus allow at least some information producers to keep their prices above their marginal costs—and suggest that information products need not be free, after all. Legally induced transaction costs, in the form of IP licenses, may enable the entire information sector.
But we passed these laws long ago, back when copying and circulating information products was still expensive. Technology protected information producers from most people, and the law protected them from a few rich competitors. The information sector devastated their technological protection. While large corporations can still sue each other, they hate hailing small buyers into court; it’s rarely worth the trouble, and it makes them look bad (though they’ll do it when backed into a corner). And that’s precisely how the information sector strips capitalism bare. It creates a world of information products that come as close to the textbook model as anything that we’ve ever seen. The information sector points toward a world without transaction costs, where prices should fall at the marginal cost of production, and where that marginal cost of production is zero. It also points us toward a world in which the producers whose business models make sense only because the law protected their IP rights are likely to become increasingly unhappy with technological progress.