Did It Pay, and What Follows

In 1944 Eric Williams, a Trinidadian historian who later became his country's first prime minister, published Capitalism and Slavery. It made two claims that have shaped the argument ever since, and they have fared very differently.

The first was that profits from the slave trade and the plantations helped finance British industrialisation. The second, the decline thesis, was that Britain abolished the trade and then slavery when the West Indian colonies had already become economically marginal: that abolition was a rationalisation of decline dressed as a moral achievement.

The decline thesis has taken heavy fire and has largely not survived. Seymour Drescher's Econocide assembled the figures: at the point of abolition the British Caribbean was expanding rather than contracting, output was rising, and slave prices were buoyant. Britain abolished the most profitable slave trade in the world while it was profitable, and then abolished slavery in colonies that were still making money.

That is the strongest available evidence that abolition was a genuine political achievement rather than an economic inevitability with a moral story attached. It does not make the abolitionists disinterested (nobody is) but it does mean their success is not explained by their interests.

The financing claim has gone the other way and has been substantially rehabilitated. Recent work has traced slave-derived capital into British banking, insurance, shipping, and industrial investment in considerable detail, and the Legacies of British Slavery project at UCL has done something similar from the other end.

That project's source is the compensation records, and the compensation is worth sitting with.

When Britain abolished slavery in 1833, it paid £20 million in compensation. Not to the people who had been enslaved. To the people who had owned them, for the loss of their property. The sum was around 40% of the government's annual expenditure, financed by a loan on which the Treasury was still paying interest until 2015.

Because claimants had to document their ownership in order to be paid, the resulting archive is the most detailed evidence anywhere of how widely slave wealth was distributed through a society, and the answer is very widely indeed. Not a handful of planters but thousands of individuals: widows with a few shares, clergymen, merchants, minor gentry, and the ancestors of a great many institutions. The money went into country houses, railways, art collections and banks.

Compensation, incidentally, is a good general indicator, and it does not measure merit. It measures the political strength of the people being bought out. West Indian interests were heavily represented in Parliament, and paying them converted a fight over principle into a transaction they could live with. The size of the cheque records how strong the losers were, not how right.

The separate question (whether slave labour was economically efficient) needs care, because two different things get run together.

Fogel and Engerman's Time on the Cross in 1974 argued that southern plantations were productive and profitable, using the gang system to extract output per worker that compared well with free agriculture. The book caused a furious controversy, much of it deserved for its tone and some of its methods, and the core finding has broadly held.

It is essential to be clear that this is not a defence of anything. Profitability and moral character are unrelated quantities, and an institution can be both highly efficient and monstrous. Indeed the efficiency was achieved by violence, which is what the gang system was for. The finding makes abolition harder to explain, not easier, which is precisely why it matters.

The long-run finding runs the other way and is at least as important. Economies built on slavery under-invested in almost everything that produces growth: schooling, because an educated enslaved population is dangerous and a poor white population is politically convenient; immigration, because free labour will not compete with slave labour; manufacturing, because capital is tied up in people and land; and infrastructure beyond what the export crop requires. Regions dependent on slavery lagged their neighbours for a century after abolition, and in several cases still do.

So: enormously profitable for owners, and corrosive to the development of the places that relied on it. Those two statements are both true and they are not in tension, because the people making the profits and the people bearing the long-run cost were not the same people. That gap is what the word "extractive" means when economists use it about institutions, and the Atlantic slave economy is where the concept was learned.