VoxEU Column Industrial organisation

The economic ideas of Ronald Coase

Ronald Coase, an economics Nobel Prize winner, passed away on 2 September 2013. He is best known for stating that transaction costs explain many puzzles in the organisation of society, and that pricing for durable goods presents a particular worry since even a monopolist selling a durable good needs to compete with its future self. This column reviews how these ideas, though often misinterpreted, deeply influenced the thinking of economists.

Ronald Coase passed away on 2 September 2013. He was still working at the incredible age of 102 – often on the Chinese economy. Coase is best known to economists for two statements:

  • That transaction costs explain many puzzles in the organisation of society.
  • That pricing for durable goods presents a particular worry since even a monopolist selling a durable good needs to compete with its future self.

Both of these statements have influenced the thinking of virtually all living economists, but they have also been interpreted and misinterpreted in many ways.

Transaction costs

First, consider transaction costs, as in “The Nature of the Firm” (1937) and “The Problem of Social Cost” (1960), two papers which have each received over 20,000 citations.

The Problem of Social Cost leads with its famous cattle-versus-crops example. A farmer wishes to grow crops, and a rancher wishes his cattle to roam where the crops grow.

Should the rancher be liable for damage to the crops, or ought we to restrain the farmer from building a fence where the cattle wish to roam?

Coase points out that in some sense both parties are causally responsible for the externality, that there is some socially efficient amount of cattle grazing and crop planting, and that if a bargain can be reached costlessly, then there is some set of side payments where the rancher and the farmer are both better off than having the crops eaten or the cattle fenced.

Further, this bargain is theoretically identical whether you give grazing rights to the cattle and force the farmer to pay for the right to fence and grow crops, or whether you give farming rights and force the rancher to pay for the right to roam his cattle.

This basic principle applies widely in law, where Coase had his largest impact. He cites a case where confectioner machines shake a doctor’s office, making it impossible for the doctor to perform certain examinations. The court restricts the ability of the confectioner to use the machine. But Coase points out that if the value of the machine to the confectioner exceeds the harm of shaking to the doctor, then there is scope for a mutually beneficial side payment whereby the machine is used (at some level) and one or the other is compensated. A very powerful idea indeed.

Powerful, but widely misunderstood. I deliberately did not mention property rights above. Coase is often misunderstood (and, to be fair, he does at many points imply this misunderstanding) as saying that property rights are important, because once we have property rights, we have something that can 'be priced' when bargaining. Hence property rights plus externalities plus no transaction costs should lead to efficiency if side payments can be made.

This is the famous Coase Theorem – if trade in an externality is possible and there are no transaction costs, bargaining will lead to an efficient outcome regardless of the initial allocation of property.

Dan Usher (1998) famously argued that this is “either tautological, incoherent, or wrong”. Costless bargaining is efficient tautologically; if we assume people can agree on socially efficient bargains, then of course they will. But these side payments can exist even when there are no property rights at all. Coase (1960) says that “[i]t is necessary to know whether the damaging business is liable or not for damage since without the establishment of this initial delimitation of rights there can be no market transactions to transfer and recombine them.” Usher is correct; that statement is wrong. In the absence of property rights, a bargain establishes a contract between parties with novel rights that needn’t exist ex-ante.

But all is not lost for Coase. Because the real point of his paper begins with Section VI, not before. Here, he notes that the case without transaction costs is not the interesting one. The interesting case is when transaction costs make bargaining difficult. His foundational point is that social efficiency can be enhanced by institutions (including the firm!) which allow socially efficient bargains to be reached by removing restrictive transaction costs, and particularly that the assignment of property rights to different parties can either help or hinder those institutions. That is, rather than finding interesting theorems in a world without transaction costs, Coase instead argues that efficient side payments are generally hindered by transaction costs, hence the need for institutions which minimise them. He is clear about this in his Nobel lecture (1992), arguing his essential point that “there [are] costs of using the pricing mechanism.” It is these costs that explain why, though markets in general have many amazing features, even in capitalist countries, large firms are run internally as something resembling a command state.

One final misunderstood idea about the Coase Theorem. In his arguments, Coase often implicitly refers to Pareto efficiency, but since property rights are an endowment, we know from the Welfare Theorems that benefits exceeding costs is not a sufficient condition for maximising social welfare (e.g. Arrow 1983). The benefit of a trade to each party is a concept that fundamentally depends on the initial endowment since preferences are maximised conditionally, subject to a budget constraint. Nothing in standard welfare economics requires us to prefer, on the social level, a Pareto dominant allocation.

The Coase conjecture

Coase’s second famous theoretical statement is the Coase conjecture, from a very short 1972 paper, "Durability and Monopoly". The idea is simple and clever.

Let a monopolist own all of the land in the US. If there was a competitive market in land, the price per unit would be P and all Q units will be sold. Surely a profit-maximising monopolist would sell a reduced quantity Q2 less than Q at price P2 greater than P? But once that land is sold, the monopolist still has Q-Q2 units of land. Unless the monopolist can commit to never sell that additional land, buyers will realise he will try to sell it sometime later, at a new maximising price P3 which is greater than P but less than P2. The monopolist then still has some land left over, which he will sell even cheaper in the next period. Hence, why should anyone buy in the first period, knowing the price will fall (and note that the seller who discounts the future has the incentive to make the length between periods of price cutting arbitrarily short)?

The monopolist with a durable good is thus unable to earn rents. Now, Coase essentially never uses mathematical theorems in his papers, and you game theorists surely can see that there are many auxiliary assumptions about beliefs and the like running in the background here. (And Coase was aware of this need to eventually formalise ideas, despite his reputation for being averse to math. From his Nobel Prize lecture (1992): “My remarks have sometimes been interpreted as implying that I am hostile to the mathematisation of economic theory. This is untrue. Indeed, once we begin to uncover the real factors affecting the performance of the economic system, the complication interrelations between them will clearly necessitate a mathematical treatment, as in the natural sciences, and economists like myself, who write in prose, will take their bow. May this period come soon.”)
It is no surprise, given the importance of this conjecture to pricing strategies, antitrust, and auctions, among many others, that there has been much formal work on the conjecture since 1972.

  • Nancy Stokey (1981) showed that the conjecture only holds strictly when the seller is capable of selling in continuous time and the buyers are updating beliefs continuously, though approximate versions of the conjecture hold when periods are discrete.
  • Gul, Sonnenschein and Wilson (1986) flesh out the model more completely, generally showing the conjecture to hold in well-defined stationary equilibria across various assumptions about the demand curve.
  • McAfee and Wiseman (2008) show that even the tiniest amount of “capacity cost”, or a fee that must be paid in any period for X amount of capacity (i.e. the need to hire sales agents for the land), destroys the Coase reasoning.

The idea is that in the final few periods, when there are few remaining customers with inverse demand above cost, even a small capacity cost is large relative to the size of the market, so the firm won’t pay it; backward-inducting agents in previous periods know it is not necessarily worthwhile to wait, and hence they buy earlier at the higher price. It goes without saying that there are many more papers in the formal literature.


McAfee, Preston and Thomas Wiseman, Capacity Choice Counters the Coase Conjecture, Review of Economic Studies 75.1 (2008)

Arrow, Kenneth, Paul Samuelson’s Contributions to Welfare Economics, in Paul Samuelson and Modern Economic Theory (Brown and Solow, eds.), McGraw Hill (1983)

Coase, Ronald H (1937), "The Nature of the Firm", Economics 4:16.

Coase, Ronald H (1960), "The Problem of Social Cost", Journal of Law & Economics 3:1.

Coase, Ronald H (1972), "Durability and Monopoly", Journal of Law & Economics 15:1.

Coase, Ronald H (1992), "The Institutional Structure of Production", The American Economic Review 82:4

Gul, Farouk, Hugo Sonnenschein and Robert Wilson(1986), "Foundations of Dynamic Monopoly and the Coase Conjecture", Journal of Economic Theory 39.

Stokey, Nancy (1981), "Rational Expectations and Durable Goods Pricing", Bell Journal of Economics 12:1.

Usher, Dan (1998), "The Coase Theorem is Tautological, Incoherent, or Wrong", Economic Letters 61.

8,966 Reads