Everything Mike O’Hare says about sunk v. recoverable costs is true and important. So are his reflections on real opportunity or resource costs vs. mere transfers.
But one of his examples — a standard one in the literature — is, in my view, profoundly wrong. When a thief steals a $250 camera, Mike says, there’s no physical destruction; the camera is just transferred from the owner to the thief. The transaction may be undesirable, he says, but it’s not a social loss of $250, because the owner’s loss is balanced by the thief’s gain.
This needs to be unpacked on two levels: the actual value of the stolen camera, and the deadweight losses associated with the practice of theft.
The deadweight-loss issue is obvious. The effort by the thief to steal the camera, and the effort by the owner to prevent its theft (which might range from locking a suitcase to moving to the suburbs) produce no social value. The costs of hiring police to chase the thief, prosecutors to charge him, judges to try him, and prison cells to hold him, and the costs to the thief and his intimates of arrest, trial, incarceration, and post-release loss of economic opportunity are still more deadweight. The anxiety experienced by the owner about the possible loss of his property, and the anxiety experienced by the thief (and his mother, girlfriend, and son) about losing his liberty, are also pure costs without offsetting benefits. Insofar (as Mike hypothesizes) the game of “who has the camera” is zero-sum, the investment of scarce resources make the activity as a whole negative-sum. By my calculations, the social costs of theft are a very large multiple of the value of stolen property.
But put all that aside, and just look at the camera itself. If it hasn’t been damaged in the process of theft, you might think, then it’s the same camera it was, and must therefore be just as valuable in the thief’s hands as it was in the owner’s. Not so.
For one thing, the owner wanted it enough to buy it. The thief only wanted it enough to steal it. In a burglary or an auto break-in, the thief might not even have known that a camera was going to be part of the loot. While the owner had a chance to pick a particular camera, the thief had no such choice, and could only take what was available. Perhaps the thief already has a camera, or doesn’t know how to use a camera and isn’t interested in learning. So even the physically identical camera is worth less to the thief than it was to the owner.
And of course in general thieves don’t keep the stuff they steal; instead they sell it either to professional fences or to legitimate dealers in second-hand goods or pawnshops; perhaps nowadays some of them put the stuff on eBay. Prices in those second-hand markets are much lower than prices in stores; a fence might pay $25 or $50 for a $250 camera. That’s partly because the stolen item isn’t identical to a purchased item; it may lack some accessories, probably lacks an instruction manual, and certainly lacks a warranty.
The fence or other second-hand dealer is likely to sell the stolen item for more than he gave the thief for it. But the difference isn’t pure gain; it’s the payment for the receiver’s time, effort, expertise, and risk. And the eventual buyer may well be getting a bargain, but to get that bargain he’s likely to have to expend more effort and incur more risk (at least risk as to the quality of the item than would be required to buy it at Circuit City; that’s why second-hand prices are lower. And to most people owning a stolen item is less desirable than owning a physically identical item acquired honestly.
So the conclusion of all this reasoning is that the value of the stolen $250 camera is roughly what the fence is willing to pay for it: typically a fifth to a tenth of its value to the owner.
Socially, we would be better off if we could program all of our possessions to melt down if stolen. Transfer is different from destruction, but theft is much more than transfer.