Plate vii in print

The Embedded

Granovetter; the embeddedness of economic action in social networks, and how far a market is a web of standing relationships rather than a crowd of strangers.

The atom of economic theory — the actor who chooses alone, weighing his gains among people who are nothing to him but prices — is a fiction, and a misleading one. Real men do business with men they know: the same suppliers, the same customers, year upon year, bound by relations of acquaintance and trust that the theory wishes away. Economic action does not float free of society; it is sunk in it, and to see how it works one must look not at the lone calculator but at the relations in which his calculating is lodged.

— the argument of this plate, after Mark Granovetter, 1985

This volume asks how economic life is governed by rules that no one wrote down. The present plate asks a prior question, about the very stage on which that life is acted: is a market a crowd of strangers, each pursuing his own gain among people he will never see again, or is it a web of standing relationships, in which men deal again and again with the same partners, and what they do is shaped by whom they know? The reigning picture in economics is the first — the market as an arena of atomized actors, dealing at arm's length, their only bond the price. Mark Granovetter, a sociologist, argued that this picture is not merely incomplete but mistaken about the basic character of economic action, which is, he said, embedded: lodged in concrete, ongoing networks of personal relations that shape it at every turn. His claim was not that men are irrational, nor that interest does not move them, but that they pursue their interests from within a social structure — among partners they know, trust, and expect to meet again — and that this embeddedness is no friction to be assumed away, but the medium in which economic life actually occurs. The question is the plate's: how far is a market a web of relationships rather than a crowd of strangers?

Granovetter framed his argument against two opposite errors that, he noticed, arrive at the same place. The first is the undersocialized account, the economist's: the actor is an atom, a self-contained calculator who weighs costs and benefits in isolation, his ties to others shrunk to the prices at which he trades with them. The second is the oversocialized account, found in a certain sociology: the actor is a vessel of internalized values, so thoroughly stamped by his culture that he obeys its rules without a thought. These look like opposites — the one all cold calculation, the other all blind conformity — yet Granovetter saw that they share a buried premise: in both, the actor is atomized, cut off from the concrete relations in which he actually stands. The economist's calculator computes alone; the oversocialized conformist, having swallowed his norms whole, equally needs no one present to tell him how to act. Neither account leaves room for the thing Granovetter held decisive: the particular, continuing web of relationships — this supplier, that customer, the partner of twenty years — within which real economic action takes place, and from which it takes its character.1

1. Granovetter set embeddedness against two opposite errors that meet in the same place. The undersocialized account (the economist's): the actor is an atom, a self-contained calculator, his ties reduced to prices. The oversocialized account (a certain sociology): the actor is a vessel of internalized values, stamped by his culture to obey without thought. Both, he saw, share a buried premise — the actor is atomized, cut off from the concrete relations in which he actually stands.

The pivot of the argument is trust, and here embeddedness shows its force. How does economic order arise — how do people come to trust one another enough to trade and lend and cooperate, when each could gain by cheating? The undersocialized answer is institutional: trust comes from incentives — the standing one would lose, the contract one could be sued upon, the future custom one would forfeit. Granovetter did not deny that these matter, but he argued they are not enough, and that the order we observe rests far more on the concrete relations themselves. A merchant trusts his longtime supplier not because he has reckoned the supplier's incentive to stay honest, but because he knows him — has dealt with him for years, has the measure of his character, is bound to him by a relationship neither would lightly wreck. Trust, on this view, is not the output of an incentive sum but a property of a relationship, grown slowly and particular to the parties to it. The impersonal devices of the textbook — standing in the abstract, the enforceable contract — are a thin and often secondary layer over the thick reality of who knows whom.2

2. The pivot is trust. The undersocialized answer is institutional: trust comes from incentives — standing, the enforceable contract, future custom. Granovetter granted these matter but held they are not enough; order rests more on the concrete relations. A merchant trusts his longtime supplier not from reckoning the supplier's incentive but because he knows him. Trust is a property of a relationship, grown slowly, not the output of an incentive sum.
⁂

But embeddedness is no romance of harmony, and Granovetter took care to insist on its darker face. The same concrete relations that breed trust also lay the ground for its betrayal. Precisely because we trust those we know, we are exposed to them as we never are to strangers: the great frauds are worked not by anonymous predators but by trusted insiders — the partner, the advisor, the member of one's own circle — who turns the gathered trust of a relationship into the instrument of its abuse. The swindler's first labor is to become embedded: to enter the network, win a place within it, and only then exploit the lowered guard that belonging brings. Disorder and malfeasance, then, are not simply the absence of social relations, to be cured by adding more trust; they frequently require relations, and feed upon them. This is why the account is not a sentimental one. Embeddedness explains at once the high cooperation of much economic life and the spectacular betrayals that punctuate it — the two as the joint issue of a single fact, that we deal, for good and ill, with people we know.3

3. Embeddedness is no romance. The same relations that breed trust lay the ground for betrayal: the great frauds are worked by trusted insiders, not anonymous predators. The swindler's first labor is to become embedded — to enter the network and win a place before exploiting the lowered guard that belonging brings. Malfeasance is often not the absence of relations but their product; trust and betrayal are the joint issue of dealing with people we know.

The argument bears directly on a question the institutions take up elsewhere: why some economic activity is organized within a firm and some left to the open market. The transaction-cost account, which the volume on the firm pursues, holds that activity is drawn inside a firm when dealing in the market grows too hazardous — when the parties are bound so tightly that either lies open to the other's bad faith. Granovetter's reply was that this account, fixed on the stark choice between the anonymous market and the authoritative firm, overlooks the vast middle ground where most real business is done: the dense networks of ongoing relations between firms, the personal ties that run across organizational lines, the long alliances that are neither arm's-length trades nor commands within a hierarchy. The enduring tie between a manufacturer and its parts suppliers — independent firms, yet bound by years of dealing into something closer than any spot market and looser than any merger — is the rule of industrial life, not the exception. Order in economic life, he argued, comes far less often from the firm's authority than from this embeddedness — the web of relations that lets independent parties deal with one another, repeatedly and on trust, without either a perfect contract or a single master over them. The boundary between firm and market, so sharp in the theory, dissolves on inspection into a continuous fabric of relationships.4

4. The argument bears on why some activity is organized in firms. The transaction-cost account holds that activity moves inside a firm when market dealing grows too hazardous. Granovetter replied that this overlooks the middle ground — the dense networks of ongoing relations between firms, the ties that cross organizational lines — where most business is done. Order comes far less from the firm's authority than from embeddedness, the web of relations among independent parties.

What follows is a reproof to the founding abstraction of the discipline. The market of the textbooks — the arena of anonymous strangers, meeting once, knowing nothing of one another but the price — is not the general case of economic life but a rare and special limit, approached in a handful of impersonal exchanges and receding everywhere else into a thicket of acquaintance. The market for grain, for labor, for capital: examine any one of them closely and the strangers resolve into people who know each other, deal again and again, hire whom they have heard of, lend to whom they trust. The social structure that economics sets aside as noise — the network of who knows whom — proves to be the very thing that makes the market run, supplying the trust that prices alone cannot, and the order that contracts alone cannot secure. Granovetter's embeddedness does not overthrow the economics of self-interest; it relocates it, setting the calculating actor back among the relationships from which the theory had lifted him, and insisting that to understand what he does, one must know not only what he wants but whom he knows. A market, in the end, is far more a web of relationships than a crowd of strangers — and the unwritten rule beneath all the others is that economic life is lived among people one knows.5

5. The market of the textbooks — anonymous strangers meeting once, knowing only the price — is not the general case but a rare limit. Examine any real market and the strangers resolve into people who know each other and deal again and again. The social structure economics sets aside as noise is what makes the market run. Embeddedness does not overthrow self-interest; it relocates the calculating actor among his relationships.