Will changes in the student visa system really cost the economy £3.6 billion over four years, as The Guardian, The Times, The Independent and the BBC report?

The Home Affairs Select Committee seems to think so, but Home Secretary Theresa May doesn’t, even though her own department came up with the figures.

The first point is that both the media and the select committee have taken the worst-case scenario from the impact assessment Ms May so disdains. The “central case” made by the assessment is a net cost of £2.43 billion, the difference between losses to the economy of £3.55 billion and savings of £1.2 billion. (The best case is a net cost of £1.3 billion, but nobody ever mentions best cases in an argument like this.)

The select committee is angry that Ms May rejects the finding, saying she is departing from evidence-based policy. And it’s plain from the delay in publishing the assessment - it appeared 12 weeks after the policy was announced - that it carried relatively little weight in the decision. But close examination of the figures suggests the Home Secretary may have a point.

The policy aims to reduce abuses in the points-based system that allowed some students to enter the country with no intention of studying after being provided with visas by “colleges” with no premises or teachers; to reduce overall numbers; to ensure that those who come to study, and then may wish to seek jobs, are the best available; and to restore confidence in the immigration system.

The means of achieving this include tougher accreditation of colleges issuing visas; limiting the right of some students to work for 20 hours a week; removing the right of graduating students to work for two years (the post-study work route) but opening up an alternative route into work; seeking better evidence that would-be students can speak English; and limiting the dependents they can bring in.

The assessment calculates, in Annex 7, that the effect of these changes in policy will be to reduce student numbers by 81,000 a year and their dependents by 32,000 a year, while post-study work route applicants will fall by 26,000 and their dependents by 6,000.

The economic consequences are many, including (obviously) reduced income for schools and colleges. But the calculations of cost are dominated by three items: the loss to the economy of the money that students and dependents spend; the loss accounted for by the students and dependents who will no longer come; and the loss of output of those denied the post-study route into work, together with their dependents. Together, these three items account for around £2.95 billion over four years, out of a total loss of £3.6 billion (83 per cent).

The underlying assumption is that the work the immigrants would have done will not be done by anybody else: the jobs vacated not taken up by those currently unemployed. That appears unlikely, and very much a worst-case scenario, especially as the post-study work route is being closed because it became clear that many of the jobs taken were low-skilled ones. The assessment assumes an average salary of £15,000 pa for the lost value of these jobs, so it is clear that many or all of them would be well within the capabilities of indigenous labour.

The results are quite sensitive to this assumption. The assessment’s central case calculates costs at £3.558 billion, and benefits (reduced health and education costs are the most important) of £1.119 billion, with a net cost of £2.439 billion. This assumes that none of the jobs that are done by students migrating to the UK are at the cost of those already here. If you assume that just 10 per cent are, the net cost falls to £2.1 billion. If 20 per cent are, it falls to £1.8 billion.

So for every 10 per cent of jobs displaced, the net cost falls by £300 million over four years. Given the recent Labour Market Survey showing that of the increase in employment last year of 416,000, those born outside the UK filled 334,000 of the jobs, it appears implausible that migrants here on education courses are not displacing some UK-born workers. Yet even in its best case, the assessment assumes that only 10 per cent of jobs done by migrant students result in UK-born workers being displaced.

This appears to be the Home Secretary’s principal objection to the assessment, and she has asked the migration advisory committee to look again “because I do not believe that the impact assessment gives a full and true picture at the moment.”

Clearly any change in immigration policy that results in a smaller workforce will have economic consequences. But the immigration policy was not driven by economics but by political judgement that the existing policy was out of control, a claim now acknowledged by those who devised it.

To focus on the worst case scenario and then accuse the Home Secretary of disregarding evidence is not calculated to improve the reputation of the select committee or the media who reported it largely uncritically. The committee is on stronger grounds when it complains of the late publication of the assessment, and more generally, of the poor state of migration statistics. But that’s hardly a new complaint.