If it were up to me, I would keep freedom of movement after Brexit. Not just for a transition period, but indefinitely. Migration from the European Economic Area (EEA) has been unequivocally beneficial for the UK. The average Eastern European migrant contributes about £1,000 more to the UK’s public finances per annum than the average UK-born adult, net of the benefits they receive, and of the value of the public services they use. Their earnings are, on average, lower than those of UK-born workers, but their employment rate is about 8 percentage points higher, and they probably also work longer hours.

The situation for Western Europeans is even more clear-cut. The average Western European migrant’s annual fiscal net contribution is about £3,700 higher than that of British-born adults. Western Europeans earn, on average, more than UK-born workers, they are more likely to be in work, and only half as likely to live in social housing. Two thirds of them are highly skilled, meaning a university degree or a vocational equivalent.

Best of all, free movement achieves all this at a remarkably low administrative cost. The system is virtually bureaucracy-free, hassle-free and paperwork-free, not just for EEA migrants themselves, but also for their employers, their landlords, their local councils etc. Free movement certainly delivers “bang for the buck”.

Claims that European migrants are “taking away” the jobs of British workers, depressing their wages, or reducing training opportunities, are all demonstrably wrong. The impact of migration on employment rates, wages and training among the native workforce is around zero. This is true in the aggregate, and it does not change fundamentally if we focus specifically on low-skilled Britons. The perception that migration benefits an aloof “liberal metropolitan elite”, at the expense of the working classes, is popular, and intuitively plausible. But like most ideas that are popular and intuitively plausible, it is also completely wrong.

And yet: I accept that supporters of free movement, like myself, have lost this argument. Keeping free movement after Brexit would be widely seen as incompatible with the spirit of the 2016 EU Referendum result, even if it complies with the letter. It would have been a different story if ongoing free movement had been announced in the immediate aftermath of the Referendum. But the intervening two years, and the government’s repeated, explicit commitments to ending free movement, have hardened expectations. Keeping free movement would now require a major U-turn, which would come at a substantial political cost.

With this in mind – is there a way to salvage at least a part of it? Are there any aspects of free movement that could be politically palatable even to most Leave voters?

There are. So far, our entire debate about post-Brexit immigration has been predicated on the assumption that after Brexit, the UK would have one single immigration regime vis-à-vis all EEA countries. Why?

This does not have to be the case. The EU is a trading bloc, but it is not an ‘immigration bloc’. We could not have one trading arrangement with France, and a different one with Poland – but there is no reason why we could not have one set of immigration policies for French citizens, and another one for Polish citizens.

YouGov runs a repeated survey which asks people which countries they would accept more, or at least the current levels of immigration from, and which countries they would like to reduce immigration from. It shows that there are enormous differences in popularity between different migrant groups, with respect to country of origin. For Europe specifically, there is a huge East-West gap. The majority of respondents do not seem to have a problem with ‘uncontrolled’ migration from Western Europe. For Poland, the responses are much more polarised, and for Romania, they are overwhelmingly negative.

This suggests that it is not “free movement” as such that was rejected in 2016. It was free movement for people from specific countries. Responding to this by ending free movement with all EEA countries would mean taking a sledgehammer to crack a nut.

We can also see this from the fact that free movement was never politically controversial until the 2004 EU enlargement, and even then, it did not immediately become an issue. In the 2005 General Election, even UKIP did not yet have a lot to say on free movement. They opposed it, but treated it as a relatively minor issue.

So here’s what should happen. After the end of the post-Brexit transition period in 2021, the UK should, in terms of immigration policy, go back to the pre-EU-enlargement status quo ante. It should keep free movement with the old member states (the EU-14) and with the EFTA countries. Ideally, this should be done on a reciprocal basis, but if that cannot be attained, it should be done unilaterally.

The newer EU member states could, for now, be treated like non-European countries, meaning that Eastern Europeans wishing to move to Britain would have to go through the tier system. (That said: the tier system itself could do with a thorough tidying up, simplification and liberalisation, which should start with an abolition of the visa cap.)

Later on, however, the ‘free movement zone’ could be extended eastwards again. The East-West gap in living standards, which has so far been the main driver of East-West migration, will not be there forever. “Uncontrolled” Eastern European migration is controversial now, but it need not remain controversial forever.

Take Poland. In the years after the fall of socialism, Poland’s national income per capita was not much more than one third of the UK level. By the time Poland joined the EU, it was still less than half. It now stands at two thirds of the UK level. Given that “uncontrolled” migration from rich countries is generally uncontroversial, we could, in fact, have a rule of thumb under which countries qualify for free movement once their national income per capita reaches, say, three quarters of the UK level.

This need not be limited to Europe. The EU never stopped the UK from adopting free movement arrangements, similar or identical to free movement within Europe, for countries with which the UK has close historical and personal ties. Canada, New Zealand, Australia and the US would have been obvious candidates. This could still happen now.

Free movement in its current boundaries may be a lost cause for now, but the idea of bureaucracy-free, hassle-free migration is not. To misquote some of my socialist detractors: REAL free movement has never been tried.