Plate iii in print

The Moat

why the monopoly is not stormed: the barriers that keep rivals out — the vast scale that fits but one firm, the patent, the network that swells with each user, the brand and the cost of switching — and the natural monopoly no entrant can challenge.

What shields an established firm from the rivals its profits invite is the height of the wall a newcomer must climb to reach the market — and where that wall is high, the incumbent may charge above his cost for as long as it stands.

— the principle of this plate, after Joe S. Bain, 1956

A monopolist's profit is bait. It tells every other firm in the world that here, in this market, price stands above cost, and money is to be made by entering and underselling him. In a market truly open, the bait is taken: rivals enter, the price falls to cost, and the monopoly dies of its own attractiveness. That monopolies nonetheless endure means something must be keeping the rivals out — and that something is the subject of this plate, the barrier to entry, the moat around the monopolist's castle.

Joe Bain, who made the study of barriers his life's work, defined the matter by its effect: a barrier to entry is whatever lets an established firm hold its price above cost without drawing in the rivals such a price should summon. The barrier is the measure of how much a monopolist can gouge before entry finally checks him — low where a newcomer can match him easily, so that even a small overcharge brings a flood of competitors; high where a newcomer cannot, so that the incumbent may charge well above cost and rest secure. The height of the wall sets the depth of the power. And the great achievement of the structural study of industry was to catalogue the walls — to see that market power is not a mysterious property of bigness but the shadow of specific, nameable obstacles to entry, each of which can be sought, measured, and sometimes torn down.1

1. Joe Bain (Barriers to New Competition, 1956) defined a barrier to entry as any factor letting incumbents hold price above cost without attracting entry. Its height sets the durability and extent of market power: low barriers make power transient (overcharging draws entrants); high barriers make it durable. The structural tradition catalogued barriers, treating market power as the consequence of specific obstacles to entry rather than of size as such.

The walls are of several kinds. The first and most fundamental is scale: where producing cheaply requires producing in vast quantity — because the fixed cost of the factory, the network, the research is enormous and is spread thinner the more units carry it — a large incumbent makes each unit for less than any small newcomer could, and the newcomer, unable to match his price, dares not enter. Where this is extreme, one firm can serve the whole market more cheaply than two, and we have a natural monopoly, a wall no entrant can scale. The second is law: the patent that forbids copying an invention, the license that rations entry to a trade, the regulation whose cost only an incumbent can bear — barriers erected by the state itself. The third is the network: the telephone, the marketplace, the standard that grows more valuable the more people use it, so that an incumbent with many users offers something a newcomer with few cannot match, however good his product. And the fourth is the bond between a firm and its customers: the brand they trust, the habit they have formed, the cost they would bear to switch — each a wall not of production but of attachment.2

2. Principal barriers to entry: (1) economies of scale — large fixed costs spread over volume let incumbents underprice entrants; in the extreme, natural monopoly (one firm serves the market most cheaply — see The Grid); (2) legal barriers — patents, licenses, costly regulation; (3) network effects — value rising with users advantages the incumbent (cf. the Externality's Network, and The Gatekeeper); (4) customer attachment — brand, habit, switching costs. Each shelters market power by raising the cost of entry.
⁂

Here a refinement enters that sharpens the whole idea, and points toward the flagship. What disciplines a firm, the theory of contestable markets observed, is not how many rivals it has today but whether it could be challenged tomorrow — not the actual presence of competitors but the credible threat of entry. A single firm in a market with no barriers at all — where a rival could enter instantly, compete, and exit without loss — would behave as if it had a hundred competitors, charging no more than cost, because any attempt to gouge would summon the entry its own profit invited. Such a market is contestable, and a contestable monopoly is no monopoly in its conduct, however lonely in its structure. Conversely, even several firms, if walled in behind high barriers, may rest comfortably above cost, secure that no newcomer can punish them. So market power lives not in the count of firms but in the height of the barriers — not in structure but in contestability — and the number of rivals a market happens to contain is a symptom, sometimes misleading, of the thing that matters, which is whether the door is open.3

3. The theory of contestable markets (Baumol, Panzar, Willig, 1982) holds that potential entry, not actual rivalry, disciplines firms: a monopoly facing costless hit-and-run entry prices competitively, while firms behind high barriers need not. Market power thus resides in barriers (contestability), not in the number of firms (structure). Concentration is at best a symptom; the open door — the possibility of entry, the buyer's alternative — is what disciplines. This anticipates the flagship's account of power as the absence of an alternative.

If power lives in barriers, the cure for power is to lower them — and this reframes the whole project of competition policy. It suggests that the most powerful pro-competitive act is often not to break up a large firm, which may be large because scale makes it efficient, but to remove the barrier that shelters it: to strike down the license that rations entry, to open the network to rivals, to shorten the patent, to forbid the conduct by which incumbents raise the wall against newcomers. It suggests, too, that the gravest market power is the kind erected or fortified by the state — for a wall built by law is the hardest for competition to breach, and the easiest for the incumbent to defend by lobbying. And it warns that not every large firm is a problem: the firm that is large because it is efficient, in a market a rival could still enter, holds no real power, while the small firm sheltered by a high wall may gouge with impunity. The barrier, not the size, is the thing to watch.4

4. Locating power in barriers reframes the cure: lowering barriers (striking anticompetitive licensing, opening networks, shortening patents, forbidding exclusionary conduct) may do more than breaking up firms, since size can reflect efficient scale. Barriers erected by the state are especially durable and lobby-defended. Efficient large firms in contestable markets hold little real power, while small firms behind high walls may gouge freely — so barriers, not size, are the proper target. The Chicago critique of structural antitrust (The Decree) builds on this.

The contestability idea is a corrective, not a dispensation, and its limits matter as much as its insight. Entry is rarely as costless as the pure theory imagines: the would-be entrant must usually sink real money into plant or research or reputation before he can compete, money he cannot recover if he fails, and the prospect of that loss — especially against an incumbent who can cut his price savagely the moment a rival appears, and raise it again once the rival is gone — deters the entry that was supposed to discipline. Sunk costs, predatory threats, the incumbent's control of an essential input, the network already locked in: these make many markets far less contestable than their barriers, naively counted, would suggest. So the open door can be, in practice, a door held shut by subtler means than a formal wall — and the assessment of market power becomes the assessment of whether entry is truly possible, which is harder than counting firms and harder than listing barriers, but is the question that decides whether a price is disciplined or free.5

5. Contestability has limits: entry usually requires sunk (unrecoverable) costs, which deter hit-and-run entry; incumbents can deter entry by predatory pricing (cutting price when rivals appear), control of essential inputs, or locked-in networks. So markets are often far less contestable than a naive barrier-count suggests, and the door may be held shut by subtle means. Assessing market power means assessing whether entry is genuinely possible — the practical core of the flagship's question of whether the other side has an alternative.

The moat is the answer to why the monopolist is not simply stormed: the wall of scale, of law, of network, of attachment, that keeps the rivals his profit invites from reaching his market — and the height of that wall, not the loneliness of his position, is the true measure of his power. Where the wall is low, or the door open, even a sole seller is disciplined by the rivals who could enter; where it is high, even a few sellers may gouge in comfort. So market power is the shadow of barriers to entry, and the contest over it is the contest to raise or lower walls. The plates that follow take up the power that the merely few contrive by agreement among themselves, and the law that would tear the walls down or break the firms behind them — before the flagship finds, beneath barrier and contestability alike, the single fact they both come down to: whether the party who must trade has anywhere else to go. The next plate turns from the one seller to the few, and to the temptation that besets them: to act, by agreement, as the one they are not.