Every undergraduate who has taken a course in microeconomics has met the lighthouse. It appears, almost without exception, in the chapter on market failure. The argument runs like this: a lighthouse provides a service — a warning signal — that no ship can be prevented from using. If you cannot exclude non-payers, you cannot collect from them. If you cannot collect, no private firm will build the lighthouse, because no private firm can make money from it. Therefore, the textbook concludes, lighthouses must be provided by the state.
This is one of the most elegant arguments in the discipline. It is also, as Ronald Coase pointed out in a now-famous 1974 paper, entirely wrong.
What actually happened
For most of British maritime history — roughly 1610 to 1842 — lighthouses on the English coast were built and operated by private firms. They were profitable. The mechanism was simple: ports collected a small per-ton "light duty" from every ship that docked, and remitted the proceeds to whichever lighthouse-keeper was nearest. Ships that did not use a port that year did not pay. Ships that did use a port paid in proportion to their size, which was a reasonable proxy for benefit. The market cleared.
"The lighthouses were built by private men. They were operated by private men. They were lit by private men. And they made money."
The first such lighthouse was at Caister, on the Norfolk coast, in 1610. Over the next two centuries, dozens more were built — at Eddystone, the Skerries, the Smalls, Beachy Head, the Casquets. By 1820 there were forty-six lighthouses on the English coast. About a third were operated by the public chartered trust Trinity House. The rest were private.
Why the textbook got it wrong
The standard story makes a category error. It conflates two questions. The first is: can a lighthouse exclude beneficiaries? The answer is no — any ship within sight can use the light. The second is: can a system be designed in which beneficiaries can be charged? The answer, demonstrably, is yes. The port did the charging. The lighthouse did the lighting. The two were institutionally linked, and the link was sufficient.
This distinction — between the technical features of a good and the institutional arrangements that govern it — is the substance of Coase's career. His earlier work on the firm, on social cost, and on broadcast frequencies all turn on the same point. The textbook category of "public good" is descriptively useful but normatively empty. The question is not whether a good has certain features. The question is what arrangements people have managed to build around it.
The nationalization
The interesting historical question is not why private lighthouses worked. They worked. The interesting question is why they stopped. The answer, traced out in the parliamentary record, has almost nothing to do with economics and almost everything to do with politics.
By the 1830s, the lighthouse system had become a vestigial patronage network. The private patents were inherited assets, sometimes held by families that had not built or operated a lighthouse in three generations. The dues were collected and the lights were kept, but the surplus accrued to absentee owners. The shipping industry — by then a politically organized force in its own right — wanted the system reformed.
The 1836 Lighthouses Act transferred the remaining private lighthouses to Trinity House at compulsory purchase. The amounts paid out were enormous: about £1.2 million in 1836 pounds, more than seventy million today. The shipping industry got what it wanted, which was lower dues. The lighthouse-owners got what they wanted, which was an exit at a generous valuation. The arrangement was not, in any sense, a response to market failure. It was a response to political pressure.
Why this matters
The story of the British lighthouses is, in some sense, a small one. There are only ever a few hundred such structures, and their economics is sui generis. But the larger point — that the textbook category of "public good" tells you very little about whether a market can actually provide a service — is one of the most important and least absorbed lessons of twentieth-century economics.
The list of services that economists have, at various points, declared impossible to provide privately is long and embarrassing. It includes mail delivery, road building, education, broadcast media, weather forecasting, and certain forms of scientific research. In every case, on close examination, the impossibility turns out to be conditional — possible in some institutional arrangements and not others. The argument is rarely about the good. It is almost always about the institutions.
This is the conservative reading of Coase. The progressive reading is the same. What both share is a recognition that the question of whether to do something publicly or privately is, in the deepest sense, prior to the question of whether the thing in question is a "public good." We choose our institutions. We do not have to discover them by classification.