Coase: The problem of social cost

Coase. 1960. The problem of social cost. Journal of Law and Economics 3 (October): 1-44.

Y: Why some externalities are internalized (thus, rights, payments, and resource allocation)

X: Transactions costs and assignment of property rights

The classic (Pigou's) approach in economics to externalities has been something like this: If a plant's pollution causes $100 in damage every year, tax it $100 until it fixes it. If installing a scrubber will cost $90, the plant will install the scrubber to end the taxation.

Coase points out the problem: if A [the plant] hurts B [the surrounding community], then compelling A to repay B simply allows B to hurt A. Pigou is not socially efficient.

So Coase would add this to the story with the polluting plant: Perhaps the people living around the plant could move away from the pollution for only $40 total [for everybody]. Then the socially optimal solution [i.e. efficient, not equitable] is to have the people move (which costs only $40). Just as continued pollution imposes a cost of $100 on the community, forcing the plant to pay this cost unilaterally allows the community to overcharge the plant (by 100-40 = $60). So A hurts B, but forced restitution allows B to hurt A. In a world without transactions costs, the parties will negotiate some sort of side payment or something [This solution assumes no transactions costs.]

RESULT without transaction costs: If rights are assigned and there are not transactions costs, then the externalities will be internalized (e.g. a side payment in the above example). You will renegotiate the property rights since there aren't transactions costs for doing so. The externality cost is priced and dealt with.

RESULT with transactions costs: Well, now it starts to matter who has the property rights. You'll look at your options: bargain as above, use the courts to enforce your right, or internalize by putting A and B into a single firm. As transactions costs rise, you become far less likely to internalize the externality on your own. Transactions costs can rise in many ways: too many people involved, too hard to identify who is involved, too hard to determine the cost of the externality, etc. WHEN TRANSACTIONS COSTS RISE, YOU MUST MOVE TOWARD HIGHER LEVELS OF CENTRALIZATION IN ORDER TO DEAL WITH EXTERNALITIES. You might need to move up from the market to the firm, or to the state.

He also spends quite a bit of time at the beginning illustrating how non-coercive (non-governmental) solutions can arise. In many ways, this discussion is logically prior to that which I've summarized above.

He makes this argument by way of analogy involving a rancher living next to a farmer. See the highlighted sections. This corresponds to what Stevens (ch. 5) calls "Coase's theorem."