Saturday, July 25, 2009

Serfdom and slavery

Having previously posted on these matters, have also done some useful reading on serfdom and unfree labour. Notably, Evsey Domar’s classic article (scroll down), summarised here by Paul Krugman. Also found an updated version of Domar’s model. (And further.)

Basically, if labour is scarce compared to land – as it was in the New World, in Europe from the late Roman Empire to high medieval period, and in post-gunpowder Russia as the former nomad-dominated lands opened up – then if labour can be controlled suitably cheaply, it pays to use unfree labour (i.e. reducing labour costs to subsistence levels), even after any loss of productivity.

As population increases (i.e. labour becomes less scarce), the cost of free competitive labour falls (in the absence of any countervailing technological improvement or other capital accumulation) and the cost of control increasingly consumes any benefit in having unfree labour, leading to the replacement of unfree by free labour. So, as the population of medieval Europe rose, the "control premium" fell, resulting in the phasing out of (that vast array of bound labour-service arrangements we call) serfdom.

The question then becomes: why didn’t the Black Death (which made labour scarce again and thus free labour expensive) result in the re-imposition of serfdom in Western Europe? There were some attempts in that direction, but they were defeated.
There doesn’t seem to be any scholarly consensus. It is clear enough from the economic models that large magnates have other ways to extract rents from land than imposing unfreedom – it is harder to leave their territories, they can provide a wider range of services, they have more bidding resources, etc. Hence the boyar magnates in Russia fought against imposing restrictions on the peasants, and the English magnates, for example, don’t seem to have been as keen to re-wind peasant conditions as the smaller landlords. The Black Death was also highly variable in local occurrence, which helped break down landlord collusion further.

In the case of Russia (and the Polish Commonwealth, etc), the Crown need for the military services of the pomestia servitors made it rule in their favour when the defeat of the “Tatars” greatly increased the land available, reducing the value of land and increasing the cost of free labour. (Indeed, restrictions on sale of land to ensure military servitors preceded those on labour).

In Western Europe, the C14th English Crown – which had already largely monetised military service, which also used yeoman longbowmen and received judicial rents from court cases involving free peasants – lacked such incentives. These factors also operated to some extent on the continent (substituting crossbows for longbows: either way the coercive balance was not as much in favour as landlords as it had been). The vigorous commercial life of the cities, disproportionately important for royal tax revenues, provided complicating competition for labour services. The increased ability of the late-medieval Crown to raise taxes also provided it with a further interest in expanded economic activity. The comparative technological dynamism of medieval Europe may have provided some opportunity for capital substitution (in the case of England, particularly the sheep trade) and certainly provided other ways of extracting incomes from peasants. Representative legislative bodies provided avenues for landlords to restrain Crown demands and gain other trade-offs. Tenant contracts were more enforceable in Latin Europe, particularly England, than in Russia.

Again, the models suggest relatively small price changes can tip the balance of action one way or another. Jonathan Conning’s aforementioned update of Domar’s model is extraordinarily fruitful. It makes the process of development of latifundia much clearer. In particular, aspects of development in Roman agrarian society which had always puzzled me – notably the tendency for small holdings to fall below viability – were greatly clarified.

The rise of demand for a staple crop – wheat in the Classical era, feeding Rome and other cities; wheat in Eastern Europe feeding the cities of Flanders; sugar, rubber and cotton in the New World – can also provide the necessary income to make imposing unfreedom worthwhile.

There is also the issue whether free labour is available at all. In Classical Greece, for example, it is not clear that much alternative was available than slaves for labour to feed the demands of a growing commercial society, given the rules were made by citizens proud of their status, able to provide coercive power and unwilling to work for fellow citizens.

In the case of the coloni of Late Imperial Rome and the serfs of medieval Europe, their landlords also provided forms of public goods, notably protection (in the former case, against the increasing depredations of the Roman state; in the latter against all and sundry). To a significant degree, these were in fact what we would call protection rackets, but they were not entirely such. Presumably, the existence of a genuine trade-off made unfree arrangements more bearable and gave peasants a collective interest in being forced to pay. (In Russia, collective tax obligations had similar effect, since if peasants left their tax liability fell on the remaining peasants until the next census). As the late-Medieval state became more able to provide public goods such as protection, this factor also dwindled.

I am particularly interested in the difference between slavery and serfdom (by which I mean any variety of bound labour-service tenancy). Slavery has, with some exceptions for state slaves, standard characteristics. Serfdom is much more variable and much more common as a mass condition. While slavery as a condition is very historically widespread (and still exists), only five mass slave societies have occurred in history – Classical Greece and Rome, colonial Caribbean, Brazil, the antebellum American South (though significant levels of state slavery were used by various Islamic states, Nazi Germany, the Soviet Union and other Leninist states).

Slaves are property, fully and completely. They have no family rights, they cannot legally own property (though there can be practical limitations on that) and, with the sole exception of the antebellum American South (which took in only about 7% of the 9.9m slaves transported across the Atlantic, but note this caveat), no mass slave population has ever fully reproduced itself, so they need regular replenishing. The antebellum American South was thus the only slave society to make slaves reliable long-term investment assets (hence its politically powerful slaveowners put up by far the strongest resistance to emancipation, as slaves involved a significant amount of the South’s capital).

Generally, the products of serfs are only partly monetised (or may not be at all). With the interesting exception of Russia, serfs were not directly saleable, being attached to the land, so the market did not extend to serf-labourers. Unlike slavery, which always entails a market in labourers (unless the state is the only slaveowner: though Nazi Germany engaged in contracting-out arrangements).

Serfs can own property, they do have family rights and they do reproduce themselves. They also entail lower costs of control, requiring both less supervision and having better incentives than slaves.

Serfdom and slavery are, to a significant degree, mutually exclusory. If serfdom rises, mass slavery declines, a result that the updated Domar model predicts in cases of concentrated land ownership. The only Greek polity without a significant private-slave population was Sparta – which had the helots. As the coloni increased in number in late Imperial Rome, slavery declined. Once the Roman State had sufficiently devalued citizenship that non-slave bound labour became significant, slaves tended to be replaced by a more cost-effective alternative, and had never been as widespread in the eastern sections of the Empire, where ancient forms of bound-peasantry persisted. New World slavery got its value from the lack of a suitable serf population and the economic efficiency of the gang-system. Differential skin colour made enforcement of slavery cheaper and provided a basis for rationalising obvious contradiction against the principles of the American Revolution.

Even though mass slavery is historically unusual, so is mass free labour (though less so) and a mass free peasantry (though less so again). It is very likely that the most common social form in agrarian societies, below a given population density, was a peasant owing some form of labour service. In such societies, labour was scarce compared to land and elites had both the power and the incentive to extract a surplus by some form of service-binding.

The only attempt to revive serfdom in a modern industrial economy – in Stalin's Soviet Union – was abandoned. Though various forms of bonded labour continue to exist. Slavery also persists in the modern world, both in the forced labour camps of Leninist states (most infamously in North Korea), but in various private forms. While people can be directly controlled, and a surplus extracted large enough to exceed any loss of productivity plus the cost of control, it will persist.

No comments:

Post a Comment