Man who lifted the lid on hidden costs of buying

The man who first thought that governments should auction off rather give away the rights to such things as broadcast spectrum or taxi licences, and who started the thinking that led to the invention of emission trading schemes, died last week at the age of 102.

He also inspired the joke economists tell each other as a warning against reading too much into statistics: "If you torture the data long enough, it will confess."

He was British-American economist Ronald Coase (rhymes with rose), of the University of Chicago, who in 1991 was awarded the Nobel prize for his trouble.

The first of his discoveries came in 1937 and launched a whole sub-field of economics, but now seems pathetically obvious. He asked why firms exist. Why do capitalists employ people to make or do lots of things for them, when most of those things could be bought from the market?

Why are many of the things produced by a "market economy" actually produced inside firms - some of them employing many thousands of people - where employees have to do as they're told by the boss and the rules of the market don't apply?

His answer was that buying things from others in the marketplace involved hidden costs, which he dubbed "transaction costs" - the cost of finding the best deal, checking quality, negotiating a price, writing a watertight contract and then, if necessary, enforcing that contract.

Business people would do things "in house" whenever this was cheaper than incurring all the transaction costs involved in buying from the market. (But once you start thinking like that, it eventually occurs to you that there will be times when it's cheaper to "outsource" the provision of services you formerly provided in-house.)

In a paper on the US Federal Communications Commission, written in 1959, Coase argued that the transaction costs faced by the commission in deciding which of the many applicants for a broadcast licence would make the greatest contribution to the economy were impossibly high.

But this did not justify the commission continuing to give away licences to whomever it saw fit. It would be better to replicate market conditions by auctioning the licence to the highest bidder. This way, the licence would go to the firm most likely to put the licence to its "highest-valued use".

Do you see how this led to the invention of the tradeable permit? Say the government is trying to limit to a certain level the catching of a particular type of fish, or limit emissions that cause acid rain, or those that cause climate change.

It issues permits for firms to catch or emit up to that level. Because this level is lower than the market would otherwise produce, it has thereby increased the item's "scarcity value", allowing firms with permits to get away with charging a higher price.

If it gives the permits away to firms, it's effectively allowing them to levy a tax on their customers. If it auctions the permits, it's ensuring the proceeds of the disguised tax are collected by the taxman.

The firms that get the licences by bidding highest can be expected to pay no more than allows them to continue profitably producing whatever it is. They'll also have a monetary incentive to find ways to continue producing their product while generating fewer emissions.

And by allowing firms to trade their permits - say, to sell any they discover they don't need - you increase their incentive to find ways to reduce their emissions, as well as ensuring the burden of reducing emissions is shifted to those firms that can do so at the lowest cost.

But Coase's greatest claim to fame came from a paper he wrote in 1960, The Problem of Social Cost, which became the all-time most cited paper by other academic economists and made him the darling of libertarians and free-market conservatives.

Social costs - also known as "negative externalities" - are costs imposed on third parties by transactions between people in the marketplace.

Say I run a factory that imposes a lot of noise on my neighbours, emits fumes and puts gunk into the local river. Since this polluting costs me nothing it represents costs borne by the community, not by me and my customers. It's a cost that's "external" to the market.

What should governments do about this problem? The traditional answer was for them to protect the victims of this action by imposing restrictions or obligations on the perpetrator.

But Coase argued that, simply by clarifying the property rights involved, governments could leave it to the affected parties to negotiate a satisfactory solution. Again, the solution could be left to the market.

What's more, this ability to reach a privately negotiated solution meant it didn't matter to which side the government awarded the property rights. The libertarians loved this so much they called it the "Coase theorem".

What they liked was that it appeared to justify a greatly reduced role for governments in solving environmental problems. That it would also favour the rich and powerful was, of course, purely coincidental.

Over the years, however, Coase made it clear the libertarians had taken him out of context. For one thing, he'd argued that to whom you awarded the property rights made no difference from the perspective of economic efficiency. Obviously, it made a big difference from an equity or fairness perspective.

And his theorem had been based on the explicit assumption that the transaction costs involved in negotiating a solution were negligible. Not surprisingly, the man who had discovered transaction costs thought that, in the real world, transaction costs would be significant and often prohibitive.

Is it easy for all the people affected by a factory's pollution to get together and negotiate a satisfactory solution with a rich factory owner? Sounds to me like a case for government intervention.

Ross Gittins is the economics editor.