Chip Taylor

Privatization, public goods, lighthouses, and parks

In a comment to my earlier post about private parks or at least park-like areas, Rey Hernandez (who I ought to go meet in person, seeing as how we are on the same campus) helpfully reminds me of Coase's lighthouse article (Link for those with JSTOR access).

I read this article for a political economy course last semester, but we didn't spend a lot of time on it in class and I went away with the impression that I hadn't quite gotten the point that Coase was trying to make. Rey's reminding me about the article prompted me to read it again.

The first time I read the article, last spring, it seemed to me that Coase was offering the history of the British lighthouse system mainly as an example of privately provided lighthouses, since private firms were involved in constructing and operating many of the lighthouses. That appeared to me to be wrong, since in the British system even private owner-operators of lighthouses depended on government grants of authority (and in some cases government customs agents) to collect the fees that paid for the lighthouses. Coase saw the government role as limited to the establishment of property rights in the lighthouse and its output. I think the government role amounts to something a bit more than that, since I see something of a regulatory or allocational role also. By that role, I mean that the government decides if my proposed lighthouse is needed or not; they probably wouldn't let let me levy tolls for a lighthouse built right next to one already in operation. This sort of thing comes up in utility regulation in the form of certificates of "necessity and convenience."

So, if you think of privatization as a continuum, the British lighthouse system is more private than many public services (say the postal service) but not exactly an example of full-blown private enterprise.

But anyway...

On re-reading the article I get a broader, though I suspect still incomplete, understanding of the article. Now, it seems to me that what Coase is also doing is disputing that the lighthouse is an example of a pure public good and disputing Samuelson's contention that lighthouses should be funded from general revenue rather than user fees.

Coase cites Mill, Sidgwick, and Pigou, who all three present the lighthouse as something to be provided by the government because it is non-excludable. In other words, as a private entrepreneur, it is exceedingly difficult for me to go into the lighthouse business, because I have no ready way (on my own) of charging passing ships for the use of my lighthouse. If there were some way for me to do so, private lighthouses would be feasible.

If you imagine what's possible with modern technology (satellites, GPS, digital radios), I imagine private navigation aids (if not lighthouses, per se) are, or could be, available by subscription. The point is, if you can exclude non-payers from using your system you can go private. But with 19th century (and earlier) technology, they had to rely on the state's customs authority to levy the fees for the lighthouses.

Anyway, Coase's beef isn't with Mill, Sidgwick, and Pigou. They all seem to agree that if you could make the lighthouse excludable, it could be privatized. (Actually, he does seem to have a minor beef with Mill, but I can't quite figure out what it is.) Coase's beef is with Samuelson. I'll get to Samuelson in a minute.

A pure public good is not only non-excludable, but also non-rival -- my consumption of the good doesn't interfere with your consumption. A lighthouse fits the non-rival description. Once it is lit up, my taking a bearing off it doesn't prevent you from doing so, too.

Samuelson pointed out the non-rival nature of the lighthouse as well as its non-excludability. He used the non-rivalry as the main jusitifcation for government provision of lighthouses. He claimed that since the lighthouse was non-rival, that even if you could exclude a non-payer you shouldn't. Allowing an additional ship to pass by would impose no additional cost; excluding it and denying it the use of the waters in the vicinity of the light would impose a social cost in the form of reduced trade. In Coase's view, Samuelson is arguing against even government-imposed user fees to fund lighthouses; the lighthouses should, instead, be paid for out of general tax revenues. Coase thinks Samuelson is wrong.

Coase describes the operation of the British lighthouse system and illustrates that a) the system was funded by user fees; b) the fee system had exemptions for low-volume users and caps for high-volume users that were, perhaps, an effort the minimize the social costs of exclusion alluded to by Samuelson; and c) the system involved the main stakeholders (as we would call them today) of the lighthouse system in making decisions abouts its funding and operation.

So, it turns out that the lighthouse system is partially excludable if you bundle it with other services that the ships couldn't avoid using -- the ports. But the system isn't totally private since you've bundled it with a government service. But this seems enough to remove the lighthouse from the list of pure public goods. If the ports themselves were completely privatized, the lighthouse system could be also.

But Samuelson was saying that even in this case, you shouldn't be funding the lighthouses with user fees, you should use general tax revenues. Coase thought that idea a total crock. He argues that the existing system is more efficient since those who are using it are paying for it and are involved in making decisions about how to run it. If you scrap that system and go to one that funds lighthouses from general revenues, then you end up with the government running a lighthouse system that is less efficient and less responsive to the users. On this point, I agree with Coase completely.

So, what has all this got to do with parks? Well, briefly (because this is too long already and also because I need to give this some more thought) it has to do with bundling public and private goods. The examples of private park-like areas that I mentioned in the other post were made possible by bundling them with private property. Geoffrey Heal has written about this in terms of environmental amenities. So, thinking back to my previous post, the question is how to do the bundling in post-development situations and how to do it so that more people have access to such areas (I originally said "open-access" but the bundling is based on exclusion, so that's not really the right term). Rey's suggestion to look for articles that cite Coase seems like one good way to start looking.