Our recent discussions with Glen Whitman about slavery, Haitian zombies, and the Coase Theorem has led us to think deeper about the relation between the Coase Theorem and other “legal failures.” (We consider the institution of slavery a paradigm case of a legal failure because it was the law, not free markets, that made slavery possible prior to abolition. By the way, you can find other examples (some fictional) of legal failures in our short essay “Buy or Bite?,” which also appears as Chapter 12 in the new book Economics of the Undead, which we highly recommend.)
Consider the U.S. Civil War — or the War between the States, as this terrible conflict is sometimes referred to in the South. As Paul Samuelson once noted in his short paper Some uneasiness with the Coase theorem: “the Civil War was not aborted by purchase of the slaves and setting them free.” In other words, although it would have been in the mutual interest of both abolitionists and slave-owners to negotiate a deal rather than go to war, such a Panglossian outcome or “Coasian bargain” did not occur. Dogmatic defenders of Coase’s theorem, of course, will resort to postulating the existence of “transaction costs” or market frictions preventing the parties from negotiating a mutually-beneficial deal, but doesn’t this vague and all-purpose excuse make the Coase theorem unfalsifiable in principle?
But aside from the important issue of falsification (an issue that is essential for us academics who care about intellectual honesty), we wish to make an even more important point, especially in the context of slavery. In short, what about the interests of the slaves themselves? In fairness to Glen, he actually takes the time to address this concern in one of his replies to our initial critique of his post on Haitian zombies. Our point here is that most slaves were simply unable to purchase their own freedom as a matter of common law (i.e. slaves lacked the “legal capacity” to enter into contracts), so how can we talk about the Coase theorem when one side to a possible transaction is prevented from bargaining at all? More importantly, notice that our focus is not on slave-owners or the Civil War; our focus is on the slaves themselves and their unjust plight created by the law.
Next time, let’s negotiate.
prior probability 
Falsifiability and the methodology of science are deeper waters than I have time to wade into right now. But I do want to address an issue of historical accuracy.
Although it’s true that slaves had no standing before the law, that didn’t prevent slave owners from making transactions of various kinds with them. Sometimes they would pay them wage-like bonuses for performing certain kinds of work. And in some cases, the slaves would save up enough money from these wages to buy their freedom from their masters. Legally, it would have been possible for the master to just take back the wages, and probably some did. But for whatever reason — perhaps a sense of honor? — some masters would stand by the agreements they made with their slaves and respect their (minimal) property claims. Also, once papers of manumission had been signed, the slave became a free person and had rights before the law.
To the extent that masters allowed and honored agreements like these, I would argue that the slavery was less the complete — i.e., that the slaves were partial self-owners, in a de facto if not de jure sense. There’s an interesting question as to why owners would have any incentive to allow that. I believe the answer is that some aspects of a person simply cannot be owned by anyone else, because they cannot easily be monitored and controlled. I alluded to this in the original blog post when I talked about how it’s impossible to know whether a slave is really “doing his best” when performing a skilled job. So in order to motivate the best possible effort, it’s necessary for the master to give the slave a share in the proceeds. This is a special case of the principal-agent problem.
In one particularly extreme case, there was a slave in Louisiana who was also a skilled entrepreneur and businessman. He actually owned and operated a riverboat on the Mississippi. In every visible respect he was a free man, except one: he had to pay his “master” a flat fee every year for the right to act in this way. It’s significant that the fee was flat, because that meant there was no marginal tax on the riverboat’s revenues, giving the slave the strongest possible incentive to maximize profits. In the principal-agent literature, this kind of outcome is referred to as the “selling-the-store solution.” (Unfortunately, I can’t recall where I read about this example. I’ll post again if I remember.)
Thanks for these historical insights … Another aspect of the principal-agent problem is that slave owners would be legally liable for the torts committed by their slaves, but I appreciate your larger point about the possibility of informal bargaining, even under such legally unjust circumstances.