In a recent opinion piece in the Guardian, historian David Olusoga makes this shocking claim about slavery in the 17th Century British Empire:

The system [Hans Sloane] witnessed and wrote about was one in which human beings were worked to death. One in which enslaved people suffered and even died from malnutrition, as the economics of the slave trade meant that it was cheaper, at times, to starve people and then replace them than it was to provide them with food.

It seems not only shocking but somewhat counter-intuitive that replacing a slave could be cheaper than feeding them, but trade is not always rational. Were there circumstances during this period where this was true?

  • 4
    @IMSoP The claim definitely has some weasel words while flat out calling it a widespread practice of "the system". Then you ask for "sometimes". Either way, answers should be thorough as well as accurate.
    – user11643
    Sep 1, 2020 at 14:40
  • 11
    @fredsbend The quote makes a blanket statement about the system working people to death, but that's not the claim I'm asking about. I read the second quoted sentence as a separate, more specific assertion qualified by "and even ... at times". It wouldn't be particularly enlightening to ask if that was always the case, and receive an answer demonstrating a single counter-example.
    – IMSoP
    Sep 1, 2020 at 14:49
  • 18
    Not a full answer, but food prices are not constant but are subject to boom and bust cycles, particularly in places like an island that raises inedible cash crops. (See also the Potato Famine, when Ireland was exporting food in the midst of widespread starvation.) It's quite possible that starving the enslaved people wasn't a routine plan, but that there were periods of local food shortage in which slaveholders found it more profitable not to feed the humans they enslaved. A "yes" answer doesn't necessarily imply a continuous approach throughout the 300-year history of slavery.
    – Tiercelet
    Sep 1, 2020 at 19:51
  • 4
    "Starving" doesn't necessarily mean completely withholding food. It can also mean severely reduced portions. A person can survive quite a long time, even doing hard labor, on minimal rations. And because they aren't eating enough food to maintain their energy, will die of starvation, just slower. Sep 1, 2020 at 20:43
  • 6
    WRT malnutrition, it needs to be remembered that one can eat plenty of "food", yet suffer from malnutrition if the diet doesn't contain the required amounts of vitamins, minerals, and so on. This was about the time that the British Navy discovered that citrus fruit prevented scurvy. Even today, in some western countries we have people who are simultaneously obese and malnourished, due to a fast food diet.
    – jamesqf
    Sep 2, 2020 at 3:54

3 Answers 3


Was it cheaper to starve a slave to death and to replace him than it was to provide him with food? It depends on what you mean with "starve".

Starve has two meanings according to Merriam-Webster:

1a : to perish from lack of food - b : to suffer extreme hunger

Definition #1

If we were to take the first definition, the claim would come down to:

A slave was at some point in time cheaper than the food required to keep him alive.

How expensive was a slave? Seeing as the context of the claim referred to the slave trade in Jamaica, we will use the slave trade in Barbados as reference. Galenson (1982) calculated that the cheapest point in time to buy a slave was in 1681, where £6,51 could get you a slave girl (see image).

enter image description here

Using this currency converter I found online, £6,51 in 1681 was worth almost $1.600 in today's money. Assuming this currency converter is half decent, I consider it to be highly improbable that there was a point in time in Barbados where a slave was cheaper than the food required to keep him alive.

EDIT: I did a little more digging and found the following table by Eltis et al.

enter image description here

This table shows that at the lowest point a slave was worth about 420KG* of sugar. I assume sugar was a relative expensive product in those days, meaning that they were probably worth a lot more than 420kg of "normal" food.

*1710-1714 sugar prices were 54,64s/cwt = £2,73. Slave prices in 1710-1714 were £22,55, so 1 slave was worth 8,25cwt = 420KG of sugar.

What about the second definition by Merriam-Webster?

Definition #2

If we were to take the second definition, the claim would come down to:

It was cheaper to keep a slave malnourished and to replace him when he prematurely dies, than to keep the same slave well fed during the same period of time.

Handler and Corruccini (1983) wrote about the plantation live in Barbados (emphasize mine):

Dirks calculates caloric levels and protein intake. He concludes that plantation food allowances were clearly inadequate "to the total energy required by the average field laborer," and that protein rations were also "marginal at best and more likely inadequate to the extraordinary demands of life and labor on a West Indian estate." Moreover, the foods that the slaves provided for themselves did not augment plantation food allowances sufficiently to produce "an overall level of nutritional adequacy." Dirks's findings can be extended to Barbados, where the historical and physical anthropological evidence support a view of a malnourished slave population.16

Regarding the reason for the malnutrition, they write (emphasize mine):

Plantation allocations varied as a result of a variety of factors within the control of individual managements: for example, what they were willing to spend on food in their efforts to maximize profits and reduce costs and how much acreage they were prepared to plant in food crops. Factors beyond their control also affected food allocations as, for example, when a disruption in trade patterns caused an increase in imported food prices with a concomitant skimping on slave rations, and when droughts, storms, and hurricanes affected the supply of locally grown foods and severely reduced the slaves' diet, sometimes to the point of producing famine conditions.

I couldn't find a source that actually did the cost-benefit analysis of slave malnutrition, but Handler and Corruccini write there were plantation owners who at least acted like it was good economics.


Based on what I could find, I consider it highly unlikely (but your mileage may very) that there was a culmination of circumstances that led to slaves being cheaper than the food required to keep them alive. However, according to Handler and Corruccini (1983) slaves were routinely malnourished in order to "maximize profits and reduce costs". So the claim is true from at least one perspective. Seeing as the first definition is a very extreme interpretation (IMHO), and considering the Principle of Charity, I am comfortable with saying that the claim is true.

Edit: some final notes. The answers to this question depend on how the claim is interpreted. During my research for this answer I found an anecdote, I can't find it anymore but I believe it was by Frederick Douglass, where a lame female slave was flogged and send away because she was useless to her master. The anecdote concluded that she likely starved to death. Does that mean that the claim is true? In this specific case her owner decided that she was worth absolutely nothing, does that verify the claim? Or is the claim about the widespread use of this practice during economic hard times? The ambiguity makes the question hard to answer, but I hope this answer brought some perspectives that are helpful.


Galenson, D. W. (1982). The atlantic slave trade and the Barbados market, 1673-1723. *Journal of Economic History*, 491-511.

Eltis, D., Lewis, F. D., & Richardson, D. (2005). Slave prices, the African slave trade, and productivity in the Caribbean, 1674–1807 1. The Economic History Review, 58(4), 673-700.

Handler, J. S., & Corruccini, R. S. (1983). Plantation slave life in Barbados: A physical anthropological analysis. The Journal of interdisciplinary history, 14(1), 65-90.

Further Reading

Aworawo, D. (2010). Bitterness on a Sugar Island: British colonialism and the socio-economic development of Jamaica (1655-1750). *Lagos Notes and Records, 16(1)*, 189-214.
  • @IMSoP, for most of history, the price of feeding someone for a day was a few pence. If the cheapest available slave goes for an average of 1560 pence, you don't need to look up exact prices -- it's self-evident that starving slaves and replacing them costs at least two orders of magnitude more than feeding them.
    – Mark
    Sep 4, 2020 at 2:14
  • @Jordy That Table 2 from Galenson says underneath: "The unit of the price estimates is local Barbados currency", so the conversion you give (for pounds Sterling) is probably not relevant.
    – Brian Z
    Sep 4, 2020 at 15:54
  • 1
    @BrianZ, nice find, I spotted that too. But after that I found Eltis, et al. (added to further reading for your convenience) which has the same info with sterling pound (see table 2). The numbers are very similar. Sep 4, 2020 at 16:51
  • 1
    My figures are off because this would have been a "long" hundredweight of 112 pounds (8 stone), so it's actually more like 550 times.
    – phoog
    Sep 5, 2020 at 17:54
  • 1
    I've deleted my previous comments and accepted this answer. While I still think it would be interesting to have a source discussing the costs of food for the slaves, I agree that would be difficult to pin down, and this answer makes a clear argument for its position.
    – IMSoP
    Sep 7, 2020 at 8:11

Strictly speaking, the quoted claim is most immediately one about Hans Sloane's writings. So if this is an accurate description of what Sloane wrote, Olusoga is telling the truth. Olusoga is not specific about which writings he is referring to. At least some of Sloane's writings are relatively easy to find online, but so far, I'm not able to pinpoint quotes that fit what Olusoga attributes to him. In this sense, I would say the claim is unconfirmed, until we can identify the sources of specific quotes consistent with Olusoga's representation.

Ultimately the more germane question though is about British 17th-century slave plantations in general. It is easy to find systematic studies of mortality patterns on slave plantations such as this one about the US South and this one on Trinidad. These generally apply to the 18th and 19th centuries, rather than the 17th. They do show periods of high mortality, and some evidence to suggest malnutrition was at least a contributing cause.

An article about Sloane's "Natural History of Jamaica" has an interesting footnote:

Dunn states that while an estimated 12,000 Englishmen came to Jamaica in the first 6 years (1655-1661) of its settlement, by the end of that time the colony's population was only 3,470. Tropical fevers and starvation accounted for most of the deaths. Although the mortality rate had improved by the time Sloane came to Jamaica, the island still had the reputation of being a tropical hell.

If even a small number of white British settlers were starving, it would hardly be surprising if their slaves were too. In any event, here is a more directly relevant quote from Dunn's book, describing Barbados c. 1680:

In only one generation these planters had turned their small island into an amazingly effective sugar-production machine and had built a social structure to rival the tradition-encrusted hierarchy of old England. But the irony is that in accomplishing all this they had made their tropical paradise almost uninhabitable. By crowding so many black and white laborers on to a few square miles they had aggravated health hazards and over-taxed the food supply, condemning most inhabitants of the island to a semistarvation diet. Those who had money squandered it by overdressing, overeating, and overdrinking and by living in ornate English-style houses unsuited to the climate. Even the rich were unhappy in Barbados, for they suffered from claustrophobia, heat, and tropical fevers and longed for the dank, chill weather they were used to at home. Most of all they hated and feared the hordes of restive black captives they had surrounded themselves with. The mark of a successful Barbados planter was his ability to escape from the island and retire grandly to England. (p. 116)

Dunn gives various pieces of historical evidence to support his description. He cites a contemporary witness who described a particular planter (Edward Atcherley) as "an irresponsible drunkard who idled away his time at Port Royal while the neglected slaves and servants at Bybrook starved, stole, plotted, or ran away (p. 217)." Dunn points out that the a 1688 update to the Barbados Slave Act, "admitted that some Negroes stole food because they were starving." (p. 242) He clearly shows that the plantation called Bybrook had periods of high mortality and that an underfeeding of the slaves was a contributing factor.

Based on this I would say that yes, there is considerable historical basis for the statement, if we do not insist on taking it too literally. We don't know based on the above that it was cheaper to replace slaves then feed them because planters were clearly cruel and incompetent and may not have been making economically rational decisions. There may also be room to argue that death due to malnutrition in conjunction with other causes may not literally equal "starvation" but I think that point is rather trivial in this context.

EDIT: Several comments have pushed for direct data on the question of relative prices. According to Galenson in late 17th century Barbados, an adult male slave cost around £20 while a young girl might be closer to £10, sometimes less. Meanwhile, Eltis discusses some relevant estimates of yearly provisions for slaves and servants for around the same time, mostly in the range of £1 to £5 per year. Whether this data shows that underfeeding slaves fit some kind of rational cost-benefit analysis for some planters or not, I have absolutely no idea. Particularly on the lower end of food cost estimates, this is assuming slaves grow some significant portion of their own food. So to try to make such a cost-benefit analysis would require an understanding of the opportunity costs involved in taking slaves away from sugar production. I'm not sure if anyone with relevant expertise has looked at this in any detail. Unless someone can find such an analysis, I think the best resolution to this question would be to simply prove whether or not Olusoga accurately represents something Sloane wrote or not.

  • 26
    I have little doubt that slave owners mistreated them, and preferred to let them die of diseases rather than invest in caring for them. But I doubt it was "cheaper" to buy another slave than feeding the ones who were already young, strong, and healthy in their possession.
    – Mari-Lou A
    Sep 1, 2020 at 17:19
  • 7
    "yes, there is considerable historical basis for the statement, if we [interpret the claim to mean something completely different]" Sep 1, 2020 at 18:59
  • 14
    @ChrisH (and BrianZ) To spell out the horrific economics beyond "I have absolutely no idea": An unfed person cannot work more than a month at the very most. The highest food cost given is less than £0.5 per month. So, to be very conservative, it would typically cost >20 times as much to starve and replace enslaved people as to feed them. And so it would take a quite large anomalous surplus of enslaved people and shortage of food to reverse this.
    – nanoman
    Sep 2, 2020 at 9:36
  • 27
    I read the claim as: "it was cheaper to underfeed slaves over a longer period of time and to buy a new one if they eventually died, than to keep slaves well fed during the same time." Many of the comments go for the most extreme interpretation: that the food required to sustain a person for a week or two was cheaper than the cost of buying a new slave. Sep 2, 2020 at 10:04
  • 5
    @nanoman As Jordy points out, that's taking the extreme of providing no food at all. If you can save £4 per year by underfeeding your slaves, and a new slave costs £20, they only need to survive for 5 years.
    – IMSoP
    Sep 2, 2020 at 15:07

Your starting point is the 17th century slave trade, but your actual question appears to be more general - was this ever true for slavery? In that case we have to say definitely yes.

In WWII, prisoners of war were used fairly extensively as slave labour by both Germany and Japan - and a key feature of prisoners of war is that they cost nothing to acquire. The Germans mostly treated West European prisoners of war reasonably but mostly worked Soviet prisoners to death. The Germans of course also used Jews as slave labour too, in companies such as IG Farben. And of course the Japanese treatment of prisoners of war on the railways was infamous; as the link says, prisoners were fed less than a cup of rice a day.

Further back, it was not uncommon for convicts to be worked on limited rations until they died, and again there are always more convicts. This was used in recent history by the USSR, but penal labour was historically used by most European countries. The intent was not always to literally work the convicts to death, but it was certainly to work them in excess of the rations supplied.

  • 13
    I appreciate that this meets the literal wording of the question, but I think it takes us too far away from the original claim. You also haven't actually provided any references, as is required for a high-quality answer on this site.
    – IMSoP
    Sep 1, 2020 at 20:42
  • 1
    @IMSoP Fair point on the references. I'm on a phone and it's a little harder. I don't believe they should be obscure to anyone, but it would improve the answer, so I'll backfill links shortly. Re the claim, I do appreciate it's some distance from the OP's example, so I don't expect it to get high votes. :) But as a sceptic, I think it does put useful bounds on the result.
    – Graham
    Sep 1, 2020 at 21:02
  • 7
    -1 As pointed out, answers here need to be sourced, and this isn't really addressing the claim (which is about slavery in the 17th century). To prevent the answer from deletion, I'd suggest to add sources. Note also that "Germans mostly treated prisoners of war reasonably" is false in general. 66% of sowjet POW died in German camps (compared to 35% of German prisoners in the UdSSR). In the case of Germany, this was very much by design (see Hungerplan & Vernichtung durch Arbeit).
    – tim
    Sep 2, 2020 at 8:02
  • 6
    Did either the USSR and Germany purchase their prisoners? Linking POW to enslavement where a purchased human being becomes your property, and is listed as such, is not the same.
    – Mari-Lou A
    Sep 2, 2020 at 9:38
  • 2
    @Mari-LouA slaves do no require to be purchased to be counted as ones. If we go back further back in time, the practice of enslaving people on a captured territory is quite common.
    – Dan M.
    Sep 2, 2020 at 11:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .