for example, according to a recent summary by the economists michael clemens, claudio montenegro and lant pritchett, white men earn 27 per cent more in the us than white women. that figure compares the hourly wage of full-time workers with similar qualifications and experience. again making best efforts to compare like with like, the economists found that white men earn 7 per cent more than black men in the us. look back to 1939, and the like-with-like wage premium for whites in the us was 60 per cent. in modern pakistan, meanwhile, men earn three times as much as equally qualified women. none of these numbers is trivial: most are appalling.

it is even possible – although perhaps only an economist would think it pertinent – to calculate the implicit wage loss suffered by us slaves. several economists have attempted to do this by comparing the “compensation” – food, clothes, shelter and perhaps some medical care – received by slaves with how much one slave-owner would pay another to rent a slave. of course, low wages were hardly the chief reason that slavery was an atrocity. yet had slaves earned for their labour what slave-owners paid each other for it, the wage would have been three or four times higher than the basic subsistence owners saw fit to provide.

there is a huge gap between what slaves would have earned in a free labour market and what in fact they were forced to accept. but the gap is dwarfed by the difference between what a nigerian-born, nigerian-educated man could earn in the formal sector in nigeria, and what he could earn if allowed to work in a rich country – more than eight times as much. nigeria is an extreme example, but there are many other countries in which all that would be needed to quadruple or quintuple a person''s income would be permission to work in a rich country. restrictions on immigration cause a greater loss of wages than racial and sexual discrimination – and perhaps greater even than slavery. this is what clemens and his colleagues call “the great discrimination”.

this is unquestionably a research paper with an agenda: lant pritchett is a vocal advocate of more liberal immigration rules. despite the agenda, i see no reason to doubt the numbers. migrants from very poor countries see huge leaps in wages if allowed to move to wealthy countries – that much is obvious. the question is whether voters in wealthy countries feel morally obliged to take those gains into account. so far, they don''t.

economists have a habit of poking these sore points. steven landsburg, author of the armchair economist, secured notoriety four years ago by labelling the vice-presidential candidate john edwards a “xenophobe”, arguing that his protectionism arbitrarily privileged americans over foreigners and was no better than arbitrarily privileging whites over blacks. few non-economists see things that way.

i have no idea how immigration barriers will be viewed by our descendants. but it is worth reflecting, if only for a moment, on the costs they impose on those trapped on the other side.