Bryan Caplan  

Why Do Slaves Cost Money?

Conard on Trade and the Trade ... "You Didn't Build That"...
I'm currently revising my notes for labor economics.  Main change: I'm cutting the week on slavery to add a full week on immigration.  It's a tough choice because I'm so fond of my slavery lectures.  But on reflection, the topic of immigration is simply more important in today's world.  Indeed, I have come to see immigration as the most important issue on earth.

Parts of my slavery notes are being moved to my lectures on human capital.  But cuts must be made.  My favorite section that I'm leaving on the cutting room floor: "why do slaves have a positive price?"  Full treatment:

IV. Why Do Slaves Have a Positive Price?

A. In many slave systems, free workers and slaves sometimes did the same kind of work, and slaves could be "rented."
B. The rental rates for slaves and the day rates for free laborers were comparable.  On the plus side, slaves could be worked harder, but on the minus side, they had to be monitored more.
C. In other words, a free worker and a slave-owner earned about the same income for one worker's labor.  Free and slave labor competed in the same market.
D. What does this show?  That free workers have always earned more than subsistence!
E. Slaves have a positive price because the owner gets to keep the difference between the competitive market wage and the cost of upkeep.

P.S. While the slavery notes are leaving my labor econ syllabus, I'll keep them at the bottom of the labor econ webpage as a bonus lecture.

Comments and Sharing

COMMENTS (15 to date)
Floccina writes:

I have always thought that slaves would be less productive than free laborers because they would be less motivated to make small innovations doing a task. Am I correct?

gwern writes:

> D. What does this show? That free workers have always earned more than subsistence!
> E. Slaves have a positive price because the owner gets to keep the difference between the competitive market wage and the cost of upkeep.

Hm, what more could we say about this from a Malthusian perspective...

If competitive market wages are greater than the cost of upkeep (the cost of living), then this says that the population could grow further: a worker could take the difference between his salary & cost of living and use it to raise more kids, and these kids could likewise have more kids.

If the population is growing, then this difference will sooner or later be erased because there will be more workers on the market driving down the competitive market wage. If this happens and the wage = cost of living, growth will stop, but there will also be no reason to keep slaves because there will be no profit from paying the same amount as cost of living rather than regular wages.

So in a large population in which per capita wealth has been driven as low as possible (everyone is poor), there's no reason to keep slaves. That's one way to get social justice, I suppose!

Conversely, the larger the difference between the cost of living and market wages, the more profit one could reap from having slaves. The difference is largest in wealthiest countries, of course.

So by this line of reasoning, we conclude: slaves are most profitable in America and least valuable in Africa.

dave smith writes:

When I was an undergrad, my history professor taught us that slavery was a way for employers to hold labor costs below equilibrium. I told him he was wrong using a similar argument you outline above and added that slaves could be thought of as capital goods with a competitive rental price.

I did not convince him.

Matthew Graves writes:

Bryan: I agree for a steady-state economy.

But what about seasonal effects for agricultural econ? I was under the impression that slavery was attractive to agricultural owners because they didn't have to pay higher wage rates during harvest time (when labor demand shot up and the labor supply curve stayed roughly constant, with the quantity supplied increasing thanks to migrant laborers chasing high wages).

That might just feed back to the capital cost of slaves, but it seems to me that you could shift the labor supply curve of location X by importing slaves to it, where the slaves are basically migrant workers being forced to move when the wage differential is not enough to justify voluntary movement, but is high enough to pay for the costs of capture.

Similarly, look at places where the cash crop export economy collapsed after slavery was abolished, because the slaves shifted to the less profitable but less work-intensive subsistence farming. Then again seems to be an effect- what was profitable with unfree labor became unprofitable with free labor.

I think the argument you outlined underestimates the value of slavery, because A is the weak point. A slave can be forced to do any task at just the cost of upkeep and monitoring- a freeman requires the market wage for that task, which includes other costs (based on unpleasantness and health risk, say).

dave smith writes:


"That might just feed back to the capital cost of slaves..."

Since slaves are bought in markets and those who bought them had to compete with others who wanted to buy them, there is every reason that your statement is completely true.

Hunter writes:

C. In other words, a free worker and a slave-owner earned about the same income for one worker's labor. Free and slave labor competed in the same market.

So the term wage slave is accurate?

dave smith writes:

Hunter: I'd say yes, because the slave did not get any money.

Greg Jaxon writes:

You'll have to teach the IRS a new trick then.
They don't believe that my income = my wage - my upkeep. They keep trying to translate "upkeep" into "deductions", which they limit and nitpick over until it comes to practically zero.

Yet, I'm quite sure the tax is only measured by one's "income", defined as you describe it.

Silas Barta writes:

@gwern: Good points, but people generally don't redirect the excess wealth into producing more kids, and after about 1800 AD, the gains in productivity stated (and haven't stopped) outpacing increases in population.

Bryan Willman writes:

That free people are worth more to society and the economy as a whole, than the same people as slaves, seems basically proven by history.

Slaves can only have a "positive price" when the slave owner is "collecting a rent". That is, only when an economy based on "arrogance and cotton" is viable. It would appear that such an economy ceased to be viable in the US in middle of the 19th century.

Of course, Robots (universal "slaves") have a positive price today - but a large part of that is because they can be worked very very hard for very low upkeep. ("Robot" was a Czech word that translates roughly as slave...)

So I suppose how much of a positive price human slaves can have is a function of what level of automation is available in the economy at the time. Human slaves would probably not have a positive price today, if we ignore sexual slavery of various sorts. (And one can imagine that very advanced sexbots would/will drive pimps and prostitutes out of business.)

Jim Glass writes:
"I have always thought that slaves would be less productive than free laborers because they would be less motivated to make small innovations...

"...the slave did not get any money."

Don't paint slavery with one brush. Slavery has been one of the most endemic institutions throughout human history, and across time and place there has been slavery and slavery and slavery, with huge variations.

Prisoners of war and criminals and undesirables of all sorts have often been worked straight to death, for instance in the Athenian silver mines. (And by the Nazis.)

At the other extreme in some cases slaves could have significant social standing, legal rights, power and wealth of the their own. In Rome some of the most powerful people in the empire were slaves of the Emperor running the state ... slaves could earn their own income and buy their freedom (the play "A Funny Thing Happened on the Way to the Forum" is based on a popular Roman play, showing that the highly "motivated and innovative" slave was well-known figure in Roman culture) ... and free people voluntarily selling themselves into slavery was not uncommon. In a harsh world, consider that trading freedom for security -- the slave owner would take care of you to protect his investment in you, when you might not be able to take care of yourself. And you might be able to buy your freedom back later. An upscale variation on this was the gladiator -- the star athlete "rock star" celebrity of the day, getting all the women and perks that went with being such. But he couldn't be a gladiator on his own, so gladiators often voluntarily became slaves of those who put up the resources to run the business ... There have also been slaves who have worked and lived entirely on their own, with their own income and property, paying a sum of some sort to their owners, which to our eyes looks much more like some sort of tax or fee than servitude.

So institutional slavery has ranged from the most murderously brutal pure exploitation to voluntary slavery with a legal right to reacquire freedom {which might even be considered a form of employment contract} to a tax. That's quite a range.

The economics of it would vary just as much, and I'd think could only be considered in detail in light of the particular arrangement.

BTW, at EconTalk, there's an interesting podcast on slavery with Stanley Engerman, who co-authored the "breakthrough analysis" of US slavery with Robert Fogel some years back.

Ghost of Christmas Past writes:
free workers have always earned more than subsistence!

It may be a minor point but that's not quite right. Adam Smith himself introduced the notion of the subsistence wage in Wealth of Nations and noted that it varied not only with the price of food but also by some social/customary considerations. In Smith's time, he explained, "subsistence" for an English worker included the cost of shoes, but French workers were content to go barefoot. (If I've misremembered that in some small detail, don't be too unkind... my books are in boxes just now.)

Anyway, Smith explained that many workers earned only subsistence, so if the absolute quantity of a free worker's subsistence was slighly more than a slave's that probably reflected a lower dignitary component in the slave's quantum, or perhaps just an adjustment for the negative value of the extra coercion to which the slave was subject.

In times of scarcity free folk of the laboring classes have sometimes been unable to command even enough of a wage to feed themselves, hence the parish poorhouse in Old England, or the more remarkable fact of people selling themselves into slavery in medieval Russia. Obviously in such circumstances free labor does not earn more than slave, probably because the extra coercion to which slaves may be subjected makes them more productive (for the slaveholder) than free workers.

chris writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

William Barghest writes:

Wasn't life expectancy for slaves in the caribbean quite short?
Is a 5 year half life considered subsistence?

Chris writes:

Wow, those notes are pretty racist. I'd hate to be the black kid in your class, Bryan. Shame on you.

Comments for this entry have been closed
Return to top