Tuesday, April 13, 2010

What is wrong with (voluntary) slavery? Why no property in people or in their labour?

In a post, David Ellerman asks the question “why was slavery wrong?” and argues that being involuntary is not the crucial problem, since there were and are arguments for slavery being a quasi-contractual (and so consensual) arrangement and for voluntary slavery. The crucial problem was rather that slavery treated people as things, and so alienated what cannot be alienated.

On the essentials, he is surely correct. If the only thing wrong with slavery was that it was involuntary, then voluntary slavery should be entirely fine and banning it is an infringement on people’s liberties to contract.

Some folk have argued that voluntary slavery is fine, notably Robert Nozick. A particularly trenchant argument in favour of voluntary slavery was put by J. Philmore (apparently a pseudonym for David Ellerman) in The Libertarian Case for Slavery (pdf) published in 1982.

In any discussion of human bondage, it is important to be clear about what particular form of bondage one is talking about, because the historical range (especially in what can be broadly called “serfdom”) is very large. It is clear that Philmore is talking about something quite different in his “voluntary slavery” than the historical conceptions of slavery. One historical version of slavery is what might be called “absolute slavery”, where a person is property completely. This was slavery in the Roman form, where a slave had the status of an animal and had no independent legal standing whatsoever. They were not even recognised by the criminal law, since their master was responsible for any criminal act they might perform.

Then there was what John Locke (Second Treatise on Government, section 24) called ‘Drudgery’, which is slavery in the Jewish, Christian (and Islamic) tradition, where slaves were owned but limits were put on what masters could do to them. So much so, that in antebellum Alabama, for example, slaves were recognised as persons for the criminal law (though for nothing else: they were otherwise property, to be bought and sold).

This second form of slavery has a certain level of tension, even contradiction, built into it. But, in either form of slavery, your children were also slaves. Roman law as codified by Justinian recognised three ways of being a slave: capture, birth or self-sale. If you were property, then your body was property and the products of your body were property. Slaves had no family rights, a major reason why slave populations typically had fertility rates well below replacement and had to be “replenished”. The only slave population that apparently did have a replacement rate fertility rate was in the antebellum South and there is reason to think even that has been exaggerated.

If your body was property, so also were the services of your body. Including, of course, sex: slave masters famously enjoyed the sexual services of their slaves. It was something of a standard “perk” of being a slave manager, for example.

To argue that people have a right to choose to be property is a big ask: particularly since it would also involve any children they may have being slaves as well. But that is not what Philmore is supporting: he is supporting voluntary slavery that is really “warranteeism”. That is, people selling their lifetime labour services (but only their labour services) while retaining family rights and, effectively, “pension” rights—support in old age. As Philmore puts it:
The quid pro quo in the warrantee contract is a lifetime guarantee of food, clothing, and shelter (or equivalent money income) in return for the lifetime right to one’s labour services. Moreover, there is no more need for a warrantee to give up his personal property and political rights in the lifetime labor contract than there is for an employee to do the same in the short term labor contract.
Which is actually quite a different thing from historical slavery. It is more like historical serfdom or debt bondage. Which people did sell themselves into, since there was a genuine exchange involved.
Read More...
This is distinct from selling yourself into slavery, as turning yourself into property then abolishes your ability to legally own property: not much of an exchange in the ordinary sense there—complicated by the fact that slaves, in practice, often did have effective property rights. Indeed, people typically put themselves into slavery to avoid execution or starvation: the quid quo pro of continuing to live was a benefit that did not require property ownership to enjoy. (Avoiding the the latter was also a reason for people to put themselves into serfdom or debt-bondage.) Parents sometimes sold their children into slavery, but that is clearly also an exchange (however outrageous or desperate).

It is “warranteeism” rather than full slavery that Paul Samuelson is invoking when he writes, as Philmore quotes,:

Since slavery was abolished, human earning power is forbidden by law to be capitalized. A man is not even free to sell himself: he must rent himself at a wage.
This a fairly standard observation in the economics literature. Philmore quotes Carl Christ who wrote, as part of the working assumptions for complete allocative efficiency, that:
The institution of private property and free contract as we know it is modified to permit individuals to sell or mortgage their person in return for present and/or future benefits.
But there is an unremarked shift within what both Samuelson and Christ write where what is being considered talks about (not) being able to capitalise your own labour by self-sale yet invokes a notion of turning yourself into property, which does rather more than merely capitalise your labour. Full slavery capitalises your body—including sexual services and your progeny—while denying yourself legal rights to own property. In other words, capitalising your labour, turning your labour into property, is not the same as genuinely selling yourself, as turning yourself into property.

From this, we can see that there is a problem with the expression “rent yourself”. There is a difference between renting your labour and renting yourself: renting your labour involves only providing labour services of specified type for specified duration. Renting you involves a much more complete offer. The notion of people-as-property has become so foreign that we glide over the distinction. But it is a very real difference.

So, care has to be taken with use of terms. For example, Philmore’s use of the term ‘human capital’ is incorrect. Human capital is not the ability to produce labour, it is the investment in improving the quality of labour through learning skills and gaining experience. As such, it is tied to a particular person, and is not transferable in the way land, labour or fixed capital is. (It can be transmitted, but not transferred.) The production of a person (who may or may not be engaged in economic activity and who has a much broader ambit than their economic role) is not the same as the production of saleable skills. Misusing the term ‘human capital’ in this way is actually a sign of not being alert to this, rather crucial, distinction.

Philmore understands that what he calls “voluntary slavery” is really “warranteeism”, but ‘slavery’ is the more familiar (and more shocking) term, so he uses it. This is unfortunate, because it encourages the confusion between selling your labour and selling yourself: between offering lifetime labour service and becoming property.

Now, we ban both turning yourself into property and turning your labour services into property. That is, you can neither sell yourself into slavery nor enter into any form of serfdom, debt bondage or other manifestation of “warranteeism”.

People barred from being property
By banning slavery, what we are saying is that people cannot be property. For that was the crucial feature of slavery: it made people property. They were owned. Modern law says that people are not the sort of thing that can be property: they cannot be treated as things in that way.

A good question is why? It is not because they are living things: we permit ownership of plants and animals. Indeed, the rise of humans from hunter-gathering has been based on ownership of living things. Such ownership even preceded agriculture, since it is likely dogs were domesticated long before agriculture.

It is because humans are a certain sort of living thing: a living thing that is a moral agent.

Not that all humans are full moral agents. Children and mental incompetents are placed under the care of others. We do this because we hold them not to have developed the full functions of moral agents.

So, in effect, we are saying that moral agents cannot be property. And we are saying that whether or not they are full moral agents, because you are not allowed to own children or mental incompetents either. Guardianship (parental or otherwise) is not ownership.

Why do we have guardianship? Because humans are worthy of moral concern: we have guardianship because, even if children and mental incompetents are not full moral agents, they need someone to be concerned for their welfare, someone able to decide on their behalf.

Slave-masters had some concern for the welfare of their slaves: they wanted them to be productive. People tend to take care of their property: so much so that systems of private property have repeatedly shown themselves to be more productive than command economies. But such concern for one’s property is entirely an instrumental concern. In the case of the slave-master, it is still predicated on the person being a thing: a useful thing, but still a thing, not a person.

So, for example, slavery in the antebellum South was less harsh than in the Caribbean and Brazil for purely instrumental reasons. In the American South, death rates were comparatively low and breeding rates comparatively high, so slaves functioned more as assets: things producing income beyond a particular production cycle. In the Caribbean and Brazil, death rates were high and breeding rates low, so slaves functioned more as intermediate goods used up in the process of production.

Guardianship respects someone’s personhood, slavery negates it: slavery takes people outside the web of morality and legal personality in a way we now find unacceptable regardless of whether they may offer themselves up for it or not.

Not that this was always true. Romans happily regarded slaves as “talking instruments” (instrumentum vocale), legally treated them as animals put to work by their masters, and held their masters responsible for any criminal act by them (under the Lex Aquilia of 286BC) in the same way that we hold animal owners responsible for any criminal damage done by them.

Christians had systems of slavery for centuries. But they had a problem, in that it was incompatible with Christianity to class slaves as animals, so Christian slave law was “polluted” by Christian qualms on this issue, as Jewish law was "polluted" by similar qualms. Hence, for example, the law in the antebellum Alabama held slaves to be criminally liable for their own actions. Hence also the elaboration of contractual arguments for slavery by various theorists within Christendom that Ellerman documents so usefully (pdf).

In the end, the contradiction between demoting humans to the status of animals—and thus things that could be owned—and humans as moral-beings-with-souls “equal in the sight of God” was resolved in the favour of the latter. But it took a long time and a brutal civil war in the US.

So, we now say that people cannot demote themselves to the legal status of animals. They cannot demote themselves to things, to that which can be owned. You cannot choose to be what you are intrinsically are not. And slavers are the enemy of all mankind because they force people into such a degraded state: they attack the web of morality in a quite fundamental way.

But intrinsically only under a certain conception of morality: a conception of everyone as belonging to the moral community, of the web of morality as a universal condition which puts limits on the web of property. Once one has such a universal conception of morality, then slavery is a universal affront. It turns “someone like us” into a thing, into the legal status like that of an animal in flat contradiction to their moral status. Hence, for example, the famous anti-slavery slogan of “Am I not a man and a Brother?”. As such, slavery is an affront regardless of whether anyone chose the state or not for it is about the status of being human, not the particular choices of particular humans.

People barred from being bound to property
Hence serfdom is likewise not acceptable. Serfdom is also a form of bondage but, unlike a slave, a serf could own property and had family rights. They always remained a legal person. Just one under the direction of another who could not leave the land they were bound to without the permission of the landowner. This lack of legal ability to exit meant that their payment could be kept down to subsistence level and a larger surplus extracted. It was a way of funding local provision of public goods (primarily protection and law enforcement) and the provision of military service to the crown.

Serfdom decayed largely for instrumental reasons: as a form of exploitation it relied on free wages being higher than subsistence plus the cost of control (including any loss of productivity) and on enforcement by the crown. As populations rose and free wages fell, landowners in Latin Christendom became more and more willing to allow serfs to purchase their freedom. (Not so much the Church, since canon law forbade alienation of Church property.) But serfdom was so clearly a system of exploitation, that its survival into post- Enlightenment times became more and more problematic and eventually rulers withdrew legal enforcement⎯without which it could not persist (as a mass condition)⎯by abolishing it in law.

While serfs were not property in the full sense that slaves were, control over them was a form of property and it also fell afoul of the notion that the web of morality put limits on the web of property. The last hurrah of serfdom (though it was, of course, never labelled in that way) was in Stalinist Soviet Union, and it was abolished soon after the death of Stalin.

One sign of the degraded status of the serf and the slave is that the power of direction extended beyond the process of production. An employee is only subject to direction during work hours, only on matters pertaining to the supply of labour to the productive process and, even there, there are legal limits on what direction is permitted. The slave and the serf were subject to legal direction extending into the rest of their life: in the case of the slave, to all aspects of their life. This was another way that being property, or being tied to property, diminished them.

To put all this another way, requiring consent is a manifestation of moral agency. We now take the view that a manifestation of moral agency cannot be used to extinguish or seriously diminish moral agency:in particular, not undermine the status of being a moral agent—even that of the person offering to consent to it. Slavery is wrong because it attacks human moral agency in a way we now find unacceptable: an unacceptable diminution of the status of human moral agency. Involuntary slavery does in both the condition itself and how it was entered into, voluntary slavery in the condition itself regardless of how it was entered into.

People are barred from their labour being property
“Warranteeism” and debt-bondage would seem to get around this because only the labour is being sold. They are bound to the owner of their labour services, but otherwise remain legal personalities and get a clear return on their exchange.

Remember how Philmore presented it:
The quid pro quo in the warrantee contract is a lifetime guarantee of food, clothing, and shelter (or equivalent money income) in return for the lifetime right to one’s labour services. Moreover, there is no more need for a warrantee to give up his personal property and political rights in the lifetime labor contract than there is for an employee to do the same in the short term labor contract.
Indeed, such an arrangement is not only economically efficient, it is required for complete allocative efficiency. As economist Carl Christ wrote, in developing the implications of the Arrow-Debreu theorem, to get full efficiency he assumed that:
The institution of private property and free contract as we know it is modified to permit individuals to sell or mortgage their person in return for present and/or future benefits.
Now, as already discussed, it is not their person being sold, but their lifetime labour services but, with that adjustment, the thing goes through. (Philmore’s parallel example of Japanese lifetime employment is not really a case of “warranteeism” as their services are not turned into property.)

Lifetime commitment would seem no more objectionable in itself than it is for marriage. Moreover, one can exit from a “warrantee” contract by “paying out” the contract. Nor is future uncertainty an issue. In a society where capital grows faster than the workforce, labour income will tend to rise but the issue of uncertainty about future income applies equally to any purchase of land or other continuing assets. Moreover, such risk applies in both directions—to the purchaser of the “warrant” as well as the seller.

It is normal with assets that they can be sold at any time. If labour services are turned into a saleable asset, then the ability to on-sell would appear to be a necessary element. Which would, of course, raise the value of the asset and improve the security for the warrantee, as they would not be entirely beholden to the ability of the original purchaser to fulfil their element of the contract.

There is nothing inherently coercive about such warranteeism and it is only labour services that are being sold. The person does not become property, nor are they bound to property, only their labour becomes property.

So, if it is objectionable, it is certainly is not objectionable on grounds of lack of consent, lack of contractual exchange, turning people into property or subordinating people to property.

Or is it? To turn someone’s lifetime services into property—with all the implications (including resale) that being property implies—is subordinating people to property in a quite direct way: property, moreover that cannot be separated from them. Surely, that is the objection: on the grounds that no part, particularly no non-separable party, of a person can be the property of another.

Which is why there is angst about selling kidneys, selling use of the womb even selling blood. We feel that it violates the principle that the web of property does not extend to the human body—as the carrier of our personhood—even in part. So selling blood seems the least objectionable because blood is replenishable. And, while the use of the womb is replenishable, it is too close to “selling babies” for people to be entirely comfortable with it.

If we are uncomfortable with selling separable bits of the body, with turning body parts into property, how much more uncomfortable are we going to be with turning something as irretrievably tied to you as your labour services into property. The point is not whether such an arrangement is consented to, the point are boundary issues:
1. whether consent is the only issue for the preservation of human autonomy;
2. what is required to defend the status of humans as moral agents; and
3. what are the boundaries between people and property.
It is entirely reasonable to say there is more to human autonomy than any act of consent, that respect for people as moral agents also means respect for their status as moral agents, for the moral dignity of being human. It is also entirely reasonable to say the web of property cannot be used to attack the very basis of human autonomy, that people—particularly non-separable aspects of people—cannot be property.

First, because the essence of property is control and we are not prepared to give any individual that level of control over another regardless of whether it is consented to. Second, because property is control of some attribute and we are not prepared to allow people—or integral parts thereof—to be mere attributes in the way animals or things can be.

The problem with the way things are expressed in the economic literature—the failure to distinguish between selling labour services and selling yourself—is not only a problem with a lack of grip on the notion of people-as-property. It is also a sign within the wider tendency of modern economics to overlook the central significance of property rights.

Employment contracts
What is different, therefore, about employment contracts? For we obviously allow people to rent their labour services. To rent them, what’s more, not as a simple commercial exchange, but as an arrangement where someone directs another. In the words of a legal text quoted by Coase and then Philmore:
The master must have the right to control the servant’s work, either personally or by another servant or agent. It is the right of control or interference, of being entitled to tell the servant when to work (within the hours of service) or when not to work, and what work to do and how to do it (within the terms of such service), which is the dominant characteristic in this relation and marks off the servant from an independent contractor, or from one employed merely to give to his employer the fruits or results of his labour.
A warrantee would surely be directed, within the hours and terms of service, like any employee. So, it is not the issue of being directed that is different. Certainly, the direction must be consented to by the warrantee to be effected, but that applies to any employee as well.

Nor is the difference that it happens under contract. In both cases, the ability to direct flows from the contract.

The problem is with the nature of the contract. An employment contract is a contract between people who remain full legal individuals: using the web of property but not part of it. A contract that is transferable only by the supplier of the labour and is transferable without any limit on their part (apart from the normal notice period). Indeed, the purchaser of the labour services often has, or takes on, various obligations of notice and/or payment if they terminate the contract.

A “warrantee” contract places an attribute of a person—their labour services—within the web of property and makes it transferable entirely by the holder of the warrant (as it must be to be the sale of lifetime service). Even if it was made non-transferrable (which would significantly reduce its value), it has still pushed the boundary of property into someone’s personhood, which the employment contract does not.

A revealing comparison is with coverture marriage, where the legal rights of the wife were subsumed into that of the husband. In the Blackstone’s words:
By marriage, the husband and wife are one person in law: that is, the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated and consolidated into that of the husband: under whose wing, protection, and cover, she performs every thing; and is therefore called in our law-French a feme-covert; is said to be covert-baron, or under the protection and influence of her husband, her baron, or lord; and her condition during her marriage is called her coverture. Upon this principle, of a union of person in husband and wife, depend almost all the legal rights, duties, and disabilities, that either of them acquire by the marriage. I speak not at present of the rights of property, but of such as are merely personal. For this reason, a man cannot grant any thing to his wife, or enter into covenant with her: for the grant would be to suppose her separate existence; and to covenant with her, would be only to covenant with himself: and therefore it is also generally true, that all compacts made between husband and wife, when single, are voided by the intermarriage.
Coverture marriage only began to be abolished with the passage of the married women’s property acts, a process that continued for over a century until the recognition of rape within marriage. (A useful comment on coverture marriage is provided by skepticlawyer here: [scroll down] [archived here].)

Coverture marriage was not mere placement of a wife under the husband’s direction, it was the subsuming of the legal personality of the wife within the legal personality of the husband. It was, therefore, a far more complete subordination than that of the employment contract, whose only restriction of the legal personality of the employee is to protect the employee from legal liability for following lawful directions. Such direction does not, otherwise, stop the employee having a legal personality: in particular, being under direction does not stop the employment contract being a contract, breaches of which can be actionable.

People and property
Under modern legal systems, people cannot be slaves because they cannot be property, and people cannot be serfs because they cannot be bound to property, so they cannot be warrantees because non-separable aspects of themselves cannot be turned into property. The rejection of “warranteeism” is not about consent, nor about being under direction. It is about maintaining the boundary between people and property. That property is something that natural persons own: never, in any form, the reverse. Hence there is no legally recognised sense in which one person can own another, in part or in full. To allow that is an affront to the dignity of the human in a way simple service is not.

So, renting your services is fine. Selling your services is fine. Capitalising your services—turning your labour service into property that someone else owns—is not.

ADDENDA This post has been amended to clarify some points and add the link to skepticlawyer's comment on coverture marriage.

No comments:

Post a Comment