When I wrote my chapter on Adam Smith in The Visions of inequality (to be published October 10) I considered including a section discussing Smith’s views on slavery. I eventually decided against it because slavery plays only an incidental role in Smith; most of his analysis in The Wealth of Nations assumes the three standard social classes that are all formally free and the interplay between them generates their incomes and social positions. In that context slaves who no longer existed in countries with which Smith was mostly concerned, namely England, Scotland, France, Netherlands, and the rest of Western Europe, were somewhat outside of the main scope of his interest.
However Smith's views on slavery are, as is well-known, fairly complex. I wrote about his argument that the position of slaves was better in autocratic than in oligarchic democratic governments. The reason is that in an oligarchic and representative government the power is held by the people who themselves are the largest slave-holders (US comes to mind) and they are very unlikely to emancipate slaves because they would thus deprive themselves and their friends of a large amount of capital. They are wary to treat them well because they are permanently afraid of slaves’ insurrection. In richer societies (Smith explains this at some length), the income and social gap between the master and the slave is much greater. Slave has also much more to covet. Hence, in more democratic and richer societies the treatment of slaves, in order to forestall their insurrection, must be particularly harsh. In one of perhaps most striking passages in The Lectures on Jurisprudence Smith says:
Opulence and freedom, the two greatest blessings the men can possess, tend greatly to the misery of this body of men, which in most countries where slavery is allowed makes by far the greatest part. A humane man would wish therefore, if slavery has to be generally established, that these greatest blessings, being incompatible with the happiness of the greatest part of mankind, were never to take place. LoJ (A), February 16, 1763.
(The passage is, I think, so seldom cited because it questions the two most cherished beliefs: popular government and opulence. If both can exist only if slaves are terrorized, what good are they? It is a striking statement.)
In more autocratic forms of government, the kings are not necessarily dependent on slaves for their power and, as Adam Smith explains regarding the abolition of serfdom in Western Europe, they might even like to get rid of forced labor if it reduces the power of local lords that could use slave armies to challenge the king.
But the most important part which attracted attention of the economists, and I want to single out a recent paper by Barry Weingast, is the following. Smith very clearly in his Lectures on Jurisprudence states that slavery, as an economic institution is, inefficient. That means that a slave-toiled land would yield to the owner much less than if the same land were leased for a reasonable rent to the tenant. The reason is obvious: the tenant farmer has an incentive to increase production even if he has to give one-third (Smith’s estimate, but we can substitute any number) of the crop to the owner. The owner can easily calculate that over a period of x years, the Net Present Value (NPV) of the rent he gets would exceed the amount of the surplus product that he receives from slaves, to whom he pays only the subsistence wage. We are not talking here about high mathematics. Things are pretty simple. We can introduce different discount rates, and different amounts for rent, different productivity differentials for the land worked by free vs. slave labor etc., as Weingast does, but whatever set of (reasonable) assumptions we choose the final result is the same. The landlord should just dismiss his slaves, offer to some of them to continue working as free tenant-farmers, find new tenant famers if needed, and spend the rest of his time at leisure enjoying the rent.
So why did not slave owners do precisely that? Why we never observe in history endogenous scrapping of slave-holding when it is such an inefficient system and when the owners could actually make more money by hiring free labor than by using slave?
It is a fair question and Weingast solves it by using a game theoretic framework showing that neither side can credibly commit to their part of the bargain. (He does this because he introduces, in my opinion, totally unnecessarily further assumption of slaves buying over time their freedom. Yes, slave owners can try to extract that too, but in a simple model, it is superfluous. For slave owners are better off even if they emancipate slaves for free.)
The reason why this spontaneous emancipation has never happened in the real world is the following. What would be the reaction that our emancipatory landlord would meet from his peers at the news that he has just released all his slaves? How would such an act be viewed by them? Certainly not with pleasure and warmth. As Smith writes, slave-owners like all property owners, the richer they are the more they live in a perpetual fear of expropriation by the poor. Every slave-owner forms part of that implicit compact. Everyone who, as a slave-owner, does not observe the compact is as much of a threat, if not more, to other slave-owners as slaves themselves.
Smith holds that in richer countries, where the slave-owners are much richer than slaves, they are especially fearful of rebellions. The only way to hold slaves down is to apply brutal control that Smith details, with gory detail: from whipping them on a daily basis, keeping people in chains during most of the day, or even in the most gruesome example crucifying them for trivial offenses, hacking their bodies into pieces and then feeding them to the fish. In such an atmosphere of fear of constant rebellion a landlord who would follow the economic logic described above would find no sympathy from other landlords. Moreover, other landlords would do everything to stop the manumission: they might claim that our economist-landlord has lost his mind, has been bewitched by slave wizards, is mad, a traitor, foreign agent, anything. They would do everything to make his life impossible, to make him change his decision and short of it, to punish him by seizing the land which in their view he is mismanaging. So, in calculating gains and losses from slavery on a piece of paper is a very different proposition from acting on what the calculation reveals.
But let us go further and assume that for some reason the rationality prevails and that his decision to emancipate the slaves gets imitated by many, and enjoys a sufficient momentum so that at certain time a large number of slaves become free, as it were, spontaneously. Make even further a very unrealistic assumption that the powers-that-be which Smith very clearly associate with people who own slaves and who, as he says, “hate them” overcome the hatred because of material interest and agree to the emancipation.
We have to ask the new question then: what will happen next? If the land was originally toiled by 5 slaves and can now be worked equally productively (yielding the same output) by 1 tenant this means that all the released slaves will not be able to find a job on the land on which they worked before. Perhaps only a fraction would. Others would congregate in the urban areas looking for jobs there.
And that's where the problems begin. There will be a massive surge of brutalized, poor and unskilled people who would get into a cut-throat competition with the urban proletariat, driving the wages down and getting into conflicts, perhaps riots and fights, with the working class. Moreover in the urban areas there would be now two large classes of malcontents with unstable jobs, subsistence wages, deep hatred of each other and even deeper hatred for the rich. This is something the top politicians among slave-owning elite are unlikely to overlook.
This is not a very different situation, just more dramatic, than what many contemporary developing countries have faced: immense inflow of labor from the countryside into the urban areas which has created political instability, violence, crime, and conflict between different parts of the lower classes and then ultimately the conflict between the lower classes and the upper classes.
Only a little dose of reality shows that the Panglossian idea, shared by some economists (but not by Adam Smith) that by proving that slave holding was not profitable for the individual landlord should have led to the spontaneous emancipation of slaves. Such economists make several logical mistakes: they make a composition error by assuming that what is true for one might also be true when it is extended to everybody, and they ignore political externality by assuming that other slave-owners are not affected by the decision of some of them.
Most importantly they make a mistake that Smith did not make: they disregard political forces and political implications of such a big move. After all we have by now sufficient historical evidence that no society has ceased being slave-holding spontaneously through the working of economic factors alone: not the slave-holders of Egypt, nor Persians, nor the Arabic slave-holders, nor the slaveholders of Justinian’s Eastern Roman empire, nor the slave-holders of the Western Roman Empire, nor the slaveholders among the Visigoths and Ostrogoths, not was forced labor abolished spontaneously in France, nor in the Habsburg empire, nor in Russia, and obviously not in the West Indies and the American South. And it is not because slave-owners were not numerate enough to calculate Net Present Value that economists have discovered.
The way slavery was crushed and abolished was almost always through the use of violence: riots, revolution, threats to government, and civil wars. Adam Smith, who believed that anti-slavery forces will never be strong enough to abolish slavery: “Notwithstanding of those superior [economic] advantages it is not likely that slavery should be ever abolished, and it was owing to some peculiar circumstances that has been abolished in the small corner of the world in which it now is” was wrong on that score, but he offers us a cautionary tale showing that looking only at economic advantages without considering them in their political context is insufficient. We can draw game theoretic boxes but they are a paltry substitute for what we observe in reality. They may be a starting, but never an ending, point.