Postscript to Coase

concerning costs

In my earlier post I observed that the organizational costs of hierarchical coordination become larger as the size of the firm, the number of people being coordinated, increased. It occurred to me when I was writing it that there was also a pattern to the transaction costs of market coordination, that they decreased as the size of the market, the number of alternative sellers or buyers of the good being produced, increases. I did not have my thoughts on that subject well enough worked out at that point to include them in the post, hence this postscript.

The pre-Coasian version of that point is the efficiency theorem for perfect competition, the demonstration that with an infinite number of firms and consumers it is in the interest of each firm to produce the quantity of each good it produces at which price equals its marginal cost of production, in the interest of each consumer to consume the quantity of each good he consumes at which his marginal value for it equals its price, hence that every unit of every good is produced and consumed if and only if its value to the consumer is at least its cost of production, which maximizes economic efficiency.

In the traditional account, the efficiency of perfect competition is contrasted to the imperfect efficiency of imperfect competition, the fact that a monopoly or monopsony maximizes its profit producing or consuming less than the socially optimal quantity. As Coase famously pointed out in his other famous article, The Problem of Social Cost, that would not be the case in a zero transaction world, since it would be in the interest of buyer and seller to agree to any change in the quantity produced and consumed that produced a net gain that could be divided between them. It follows that the inefficiency traditionally blamed on imperfect competition can instead be seen as due to transaction costs.

What determines how large they are?

For part of the answer, consider a firm or individual trying to obtain a good or service on the market. There will be a range of prices at which it is cheaper to buy the good than to produce it. If there is only a single seller, the seller can try to hold out for his estimate of the highest price in the range while the buyer tries to purchase at his estimate of the lowest the seller can accept. Both parties can consume substantial time and effort in the bargaining and face a risk of bargaining breakdown, the transaction not happening because one of the parties misestimated what terms the other would accept. As the number of sellers increases, that problem decreases. If one seller holds out for a high price another can undercut him. If the buyer finds the price he is offering is too low for not only one seller but all he can be reasonably confident that it is below the bottom of their bargaining range and raise his offer. With multiple buyers and sellers in the market for the good, past transactions will provide information to narrow the bargaining range for future transactions.

It follows that the transaction costs of market transactions are lower the larger the number of buyers and sellers of the good transacted over, hence that firms are more likely to produce an input in-house the fewer the firms that produce it.

This provides a second reason, in addition to their small size, to structure families as firms rather than markets; for many of the goods and services family members provide each other there are no, or very few, alternative providers who can provide them at a similar cost or quality. There are few babysitters who can get our children to sleep as well as we can, and a baby sitter has to come and go; we live here. We know what dishes go into the dishwasher and where they go when they come out; the housecleaners who come once a week do not. The firm may include employees, such as a nanny or housekeeper, as well as family members. Some services and many goods are outsourced, purchased on the market — restaurant meals, electric power, groceries — but many are produced within the family/firm.

Costs Within the Firm: Monitoring

One of the things determining how large a firm is, how many people are being coordinated, is how hard it is to monitor them, to know whether the workers are doing what the boss wants them to do. Some jobs are easier to monitor than others, either because output is measurable — paying fruit pickers by weight of fruit picked or salesmen by amount sold — or because effort is measurable. The latter comes up in the literature on slavery as an explanation of why slave production could outcompete production by free workers for some crops but not all. For crops such as cotton that were produced by gang labor, a lot of slaves all doing the same thing at the same time with one supervisor making sure they did it, monitoring effort was relatively easy. It was easier still to monitor a crew of galley slaves, all rowing together.

The same logic applies to monitoring employees. Workers on an assembly line can be monitored for effort, fruit pickers for output, so a firm employing them can get large with only a small percentage of the labor force monitoring and managing, few layers between top and bottom of the organization. A firm whose employees are lawyers or high school teachers has a harder problem; effort is hard to observe; output, cases won or students educated, is harder to measure and has a larger random component. If follows that we would expect smaller firms for those enterprises, more inputs purchased instead of produced.

Coase’s Project: The Future of Economics

A very long time ago, I wrote an article trying to explain the implications of Coase’s argument in “The Problem of Social Cost.” I later based Chapters 3 and 4 of my book Law’s Order on the article, spent Chapter 5 working out in some detail how, if we knew enough about transaction costs, we could use Coase’s ideas to design liability law.

I was, at the time, a faculty fellow at the University of Chicago Law School; Coase was a colleague of mine. His comment on my article was that you never really understand your ideas until someone else explains them to you. Knowing Coase I suspected the comment was not to be taken literally.

Many years later I agreed to write an article on Coase’s final, and very interesting, book, How China Became Capitalist (coauthored with Ning Wang). To do so, I first reread much of Coase’s writing.

I believe I now understand his comment. I had interpreted “The Problem of Social Cost” as a guide to designing the law. He had intended it as a critique of current economic theory, a demonstration that the assumption of zero transaction costs implicit in most of price theory, taken seriously, led to conclusions inconsistent with that theory, hence that economics needed to be reconstructed in a way that paid attention to transaction costs, a project that would require paying more attention to real world markets and transactions than economic theorists were in the habit of doing.

As he wrote, in a talk I only came across while writing this post:

For me, ‘The Problem of Social Cost’ was an essay in economics. It was aimed at economists. What I wanted to do was to improve our analysis of the working of the economic system. Law came into article because, in a regime of positive transaction costs, the character of the law becomes one of the main factors determining the performance of the economy. … I pointed out that the judges in their opinions often seemed to show a better understanding of the economic problem than did many economists even though their views were not always expressed in a very explicit fashion. I did this not to praise the judges but to shame economists. (Coase, “Law and Economics at Chicago“).

Think of this post as a very small contribution to a theory of transaction costs.

---

My web page, with the full text of multiple books and articles and much else

Past posts, sorted by topic

A search bar for past posts and much of my other writing