Coase and Lighthouses

I read a series of lectures Coase wrote in the 50s and was struck by how lucid and insightful his thinking was. I've been reading D-Squared Digest (Daniel Davies) for a few months and while some of it is quite fun, it's not Coase (no offense intended). Imagine my surprise when I see " Score: Daniel Davies, 393; Ronald Coase, 0" on J Brad DeLong's site.

It turns out to center on an argument between Coase and Samuelson on lighthouses. Samuelson posited that lighthouses are a "public good" in that, since it is very hard to charge for the services lighthouses render, there will be no lighthouses unless the government steps in and provides them. Coase looked at the history books and found that there were lighthouses before the government stepped in to provide them (1836), and they were paid by charging ships when they came into port (so much so for "public goods" then). D-Squared Digest then looks at the history books more closely, and notes that Henry III created a a Royal Patent dated 1261, where some private individuals can collect "light duties" from shipping to pay for the upkeep of a light house. He goes on to argue that this sort of "mandatory duty" does not jibe well with the free-market / free-participant negotiations Coase implied sorted out the lighthouse problem long before the government got involved. OK.

D-Squared Digest generally paints a false caricature of Chicago style economics, and this most recent post is no different. While Daniel claims a "mandatory duty" does not seem like a typical free-market, to me it sounds like backward integration by consumers to set up a two-part tariff regime to provide a necessary service where a more classical free market is not possible.

Here's what I mean. When you have a good with a marginal cost of zero (like light from a lighthouse, or an mp3) you face the problem of covering your upfront, fixed costs (like building a lighthouse, or recording a song for the first time). This is a very tricky area, and no solution works very well, but one common system is the "two part tariff" where a consumer pays some upfront cost to "enter" the market, and then pays the marginal cost after that. The upfront "entrance fee" covers fixed costs, while the marginal price after that covers marginal costs. In the lighthouse example, ships had to pay an upfront fee (in the form of a port tax) to sail around, where they consumed light from lighthouses for their marginal cost (zero).

To me this seems to fit perfectly well within the neo-classical economics systems. Chicago school economics is not against government regulation per se, it just feels that governments should support markets and let individuals sort it out from there, and that governments should not engage in social engineering. In the lighthouse example, private parties (the sailors) needed a two-part tariff system to create a market for a good they needed but would not otherwise be provided (lighthouses). They contracted with other private parties (ports) to arrange this, and the government supplied the legal infrastructure needed to support that market (making paying the tax mandatory).

Samuelson's public good argument had the entire population paying taxes to support lighthouses only used by sailors (which seems strange) and after 1836, the government itself started to operate lighthouses (which seems even stranger since private individuals were already doing that just fine). And perhaps most bizarrely, Daniel notes "...there are numerous accounts of "rent-seeking" behavior in lighthouses, whereby lighthouse entrepreneurs with good political connections sought to build unnecessary lighthouses in anticipation of the stream of light duties they would be allowed to extract" as if this refutes the neo-classical interpretation of what would happen in the market and seems to deny that exactly the same sort of rent-seeking would go on if the government ran lighthouses itself. Note how, in the US, states lobby for all kinds of unnecessary pork-barrel public works projects that seek only to transfer wealth from other states to their own.

I'll will also note that in 1995 in Britain, there were 56 automated lighthouses, 11 staffed lighthouses, 10 automated lightvessels, 2 automated lightfloats, 2 large automatic navigation buoys (lanbys), 414 buoys (of which 94 were unlit), 26 beacons, 62 radar or radio beacons, and 11 Decca Navigator stations. All in an age where a GPS system costs about $100. I'm not saying there would be more or less of these things if the government was not in the business of lighthouse provision and therefore had a bureaucracy invested in lighthouses, I'm just sayin'.