The plate before this named six gaps that keep a buyer and a seller apart, and called the bridging of them the middleman's work. This plate takes up the first and most fundamental of those gaps, the one that must be crossed before any of the others can matter: the gap of search — the plain difficulty that a buyer and a seller, scattered across a wide and crowded world, so often simply fail to find each other at all. The blackboard market of the textbook knows no such difficulty. There, buyers and sellers meet at once and for nothing, a single price is called out and heard by all, and every willing pair trades in an instant. The real market is nothing of the kind. Its traders are dispersed and ignorant of one another; the man with a thing to sell does not know who wants it, and the man who wants it does not know who has it; and the finding out is slow, laborious, and costly. The economics of this was opened by George Stigler, who saw that information is not free but must be hunted at a price, and that the failure to hunt it down completely leaves its mark in the scattering of prices for one and the same good. It grew, in the hands of others, into a theory of how the scattered are matched — a theory that recast unemployment itself as the time a worker spends searching for the job that wants him. And it bears directly on this volume's question, for the middleman who closes the gap of search, who makes himself the place where the scattered come to find each other, performs the clearest and least disputable of his services — and yet finds, in the very friction that makes him valuable, the seed of a power that can make him unavoidable.
It is worth dwelling on how much the frictionless market leaves out, for the omission is not a small one. The supply-and-demand diagram that opens every course in economics shows a single price at which the quantity offered equals the quantity wanted, and tacitly assumes that this price is known to everyone, that every buyer can reach every seller, and that the meeting of the two costs nothing. Relax those assumptions — as the world relentlessly does — and the tidy diagram dissolves. Buyers do not know all the prices on offer, and learning them takes time they could spend otherwise. Sellers do not know where their customers are, and reaching them takes effort and expense. The two sides are separated not only by the gap between what one will pay and the other will accept, but by the sheer difficulty of locating each other in the first place — a difficulty that grows with the size and dispersion of the market. This difficulty is what economists call a search friction, and it is not an exotic special case but the ordinary condition of nearly every real market. The frictionless market is the exception, achieved only in a few highly organized exchanges; the friction-ridden market, in which finding the other side is itself a costly undertaking, is the rule. And wherever the friction is real, there is room — and need — for someone whose business is to reduce it.
Stigler's insight was to treat the acquiring of information as an economic activity like any other, governed by cost and benefit at the margin. A buyer who does not know the prices on offer can find them out only by searching — by visiting another shop, soliciting another quotation, making another inquiry — and each act of search costs him something in time and trouble while promising him something in return: the chance that the next price he turns up will be lower than the best he has so far. The rational searcher continues, then, only so long as the expected saving from one more look exceeds its cost, and stops the moment it does not — which is, in general, well before he has found the very lowest price in the market, for the labour of certainty is rarely worth its price. From this simple logic Stigler drew a striking consequence. If buyers do not search all the way to the bottom, then sellers can charge different prices for the identical good and all still find custom, the dearer ones supported by the buyers who did not search far enough to discover the cheaper. Price dispersion — the same thing selling for different sums in the same market — is therefore not an anomaly to be explained away but the natural and permanent footprint of costly search: a standing measure, as Stigler put it, of ignorance in the market.
The same good sells for a dozen different prices, not because it differs, but because no buyer can afford to learn them all; price dispersion is the standing measure of the market's ignorance.
From Stigler's account of the searching buyer grew a fuller theory of how the two sides of a market are matched — and it began with a paradox that revealed how powerful even a small friction can be. Peter Diamond showed that if every buyer must pay even the tiniest cost to search out a second price, the only stable outcome is for every seller to charge the full monopoly price. The reasoning is stark: no buyer will pay a cost to look elsewhere when he expects all sellers to charge the same, and knowing this, each seller raises his price to the monopoly level, and the expectation is fulfilled. A search cost of a single penny, in the limit, hands every seller the power of a monopolist — a result so extreme that much of the field since has been devoted to escaping it, by allowing that some buyers search more than others, which restores a genuine spread of prices between the competitive and the monopolistic. The same machinery, turned upon the labour market, became the great matching theory of Mortensen and Pissarides, in which workers and jobs find one another not instantly but through a slow and costly process, and unemployment is reconceived not as a simple excess of workers over posts but as the population in transit through the matching at any moment — the searchers not yet matched. The friction, in every version, is shown to be both pervasive and consequential: it disperses prices, it idles willing workers, and it confers, on whoever can reduce it, a power worth having.
For the reduction of the search friction is exactly the middleman's first and clearest service, and a great deal of intermediation exists for no other reason. The broker is the man who knows, at once, who is selling and who is buying, and spares each the labour of finding the other. The exchange and the marketplace gather a multitude of buyers and sellers into a single place, so that the search which would have ranged across a whole dispersed market collapses into a single visit. The dealer who stands always ready to trade abolishes search altogether for the man in a hurry: he need not find a counterparty at all, but simply comes to the dealer, who is reliably there. Economists have modelled the middleman precisely as the creature of the search friction — the party who arises, in a market where buyers and sellers struggle to meet, to hold himself out as the fixed point that both can find — and have shown that such a middleman can make both sides better off, taking for himself a cut that is less than the search he has spared them. This is the earned middleman in his purest form. He produces, in the narrow sense, nothing; but he converts a market in which willing pairs went unmatched into one in which they meet, and the value of that conversion is real, large, and his to claim. The matchmaker's whole service is to be the one place the scattered can reliably find each other, and his cut is the honest price of the search he saves them all.
The matchmaker produces nothing one can hold; but he turns a market where willing pairs went unmet into one where they meet, and that turning is worth its wage.
But here the volume's recurring shadow falls across even this purest of services, for the very friction that makes the matchmaker valuable is also the source from which his power grows. The matchmaker is most useful when he is the place where the most others can be found — and so his usefulness feeds upon itself: buyers come to him because that is where the sellers are, and sellers because that is where the buyers are, until the gathering he created becomes the gathering no one can afford to leave. This is the network effect, and it converts the search friction from the matchmaker's justification into the matchmaker's moat. Once he is the one place where the two sides reliably meet, the alternative to dealing through him is not a rival matchmaker but a return to the very search he abolished — and that alternative is precisely the friction he was paid to remove. The Diamond paradox, in this light, is not merely a curiosity about pennies and monopoly prices; it is the general law of the successful intermediary. For if even a small cost of searching elsewhere can hand a seller the power of a monopolist, then the matchmaker who has made searching elsewhere prohibitively costly — by gathering everyone under his own roof — has handed himself that power in full. The resolver of the search friction tends, by the very success of his resolving, toward the keeper of a toll on the only road that remains.
So the search friction is, for this volume's question, a double-edged thing, and the honest verdict must hold both edges at once. On the one side, the friction is the first and most legitimate justification of the middleman, and the resolving of it the first and most genuine of his earned services. Buyers and sellers truly do fail to find each other, truly do pay for the failure in dispersed prices and unmatched pairs and idle hands, and the one who gathers the scattered and matches them truly does create value where none was being realized — value for which his cut is, in the competitive case, the honest wage. To deny this is to retreat into the parasite charge the previous plate refuted. But on the other side, the friction that justifies the middleman is the same friction from which his power grows, for the gathering that makes him useful makes him, in time, unavoidable, and the unavoidable can charge for more than it does. The line between the matchmaker who earns his cut by sparing a search and the toll-keeper who levies a cut by controlling the only place a search can end is therefore not a line between two kinds of intermediary but a passage that a single intermediary travels, from the one condition into the other, as his very success at matching hardens into a position none can bypass. Which side of that passage any given middleman stands on is a question of how easily he can be gone around — and that is the question the volume's later plates, on the platform and the toll, are kept to press.
The gap of search, then, is where the middleman's case is strongest and his danger is born in the same breath — the clearest service he renders and the first source of the power he can come to abuse. The volume will follow that power to its fuller forms, in the platform that prices its own indispensability and the toll levied from a chokepoint held; here it is enough to have seen that the friction which calls the middleman into being is the friction that can, in the end, enthrone him. The matter belongs, by old right, to the price of labour, for it was in the search for work that this friction was studied most closely and felt most keenly — the worker idle not because no job exists but because he and it have not yet found each other, and the wage itself bent by how hard the finding is. The next plate turns from the gap of search to the dealer who stands in it, and to the two prices, one to buy and one to sell, between which he quietly keeps the difference. As the table says of every price, first to last: the argument remains open.