IMMIGRATION AFTER BREXIT.

Portes, Jonathan

Introduction

If the UK's vote to leave the European Union was a vote for, or more specifically against, anything, it was a vote against free movement of workers within the EU. The slogan, 'Vote Leave, Take Control', summed up the entire Leave campaign, whether referring to the fictional 350 million [pounds sterling] per week that the UK 'sends to Brussels', or to our supposed ability speedily to conclude advantageous free trade deals with third countries once freed from the dead hand of EU control of UK trade policy. However, it was particularly effective and resonant with respect to immigration policy and border control, because, of course, it contained a very large element of truth; free movement of workers is one of the foundational 'four freedoms' of the EU and, as long as it remains a member, the UK is obliged to respect this central obligation (Portes, 2015).

So the Remain campaign found it extremely difficult to counter the simple argument that the only way for the UK to 'control' immigration was to leave. Negative attitudes to immigration, and in particular free movement within the EU, are very strongly associated with opposition to UK membership. Ashcroft (2016) found that approximately 80 per cent of those who thought that immigration was mostly a force for good voted to Remain, while a similar proportion of those who thought of it as a force for ill voted to Leave. This correlation remains strong when controlling for socio-demographic factors (Vasilopoulou, 2016).

The evidence is less clear as to whether immigration, either from within the EU or more generally, was the key driving factor in the vote to leave. There is already a significant literature on this topic. A number of analyses (e.g. Clarke, 2016; Goodwin and Heath, 2016; Carozzi, 2016) find that areas with higher levels of immigration were, if anything, somewhat more likely to vote to Remain, but areas which had experienced large recent migrant flows (at least on a proportional basis) were more likely to vote Leave. But alternative explanations also exist--for example Colantone and Stanig (2016) show that the Leave vote was correlated with exposure to competition to trade with China, while at an individual level Kaufmann (2016) shows that social attitudes were more important than economic self-interest (as measured by socio-economic status). Clearly monocausal explanations are insufficient, and simple univariate or multivariate regression analysis is insufficient to establish causality. But to any casual observer of the national-level campaign, it is clear that--outside London at least--perceptions of the impact of migration were indeed a key factor in driving the Leave vote.

It is therefore not surprising that immigration--and what Brexit will mean for UK immigration policy--is equally central to the post-referendum debate. This paper is a follow-up to Portes (2016a), which examined the historical record of EU migration to the UK; here I look forward, and examine the short and long-term implications of the vote. Inevitably, this is more speculative in places, and the political and economic landscape is constantly changing. However, even before the UK leaves the EU--that is, while free movement remains in place much as now--there will be significant impacts. First, the vote itself will impact migration flows directly and indirectly. Second, policymakers need to address urgently the question of EEA nationals already resident in the UK. Post-Brexit, policy choices will need to be made on two key dimensions. There is a spectrum of options, ranging from only minor modifications to the principle of free movement through to a move to a system in which EEA nationals are treated similarly to those of third countries. At the same time, decisions will need to be taken as to just how restrictive any new system will be.

Short-term labour market impacts

EU migration is both driven by, and plays an important role in, UK labour market developments. The most recent data show that in the last quarter, essentially all the net increase in employment was accounted for by foreign born workers; the number of UK-born workers was flat (this does not, of course, mean that all 'new jobs' or 'new hires' were foreign born). Non-UK born workers now make up just over 17 per cent--more than 1 in 6--of those in work.

Over the past year (since these statistics are not seasonally adjusted, it is preferable to look at the annual change) the total net increase in employment has been just over 610,000; of this, the UK born account for about 250,000, EU (non-UK) born for 280,000 and non-EU born for 80,000. (ONS, 2016). (1)

Arguably these figures may give us the 'high water mark' for the impact of EU migration on the UK labour market. Over the next two years--that is, before any changes in law or policy have any impact, a sharp fall in net migration from the EU appears likely, for several reasons:

* Even before the referendum, employment growth in the UK had slowed (whether as a result of Brexit-related uncertainty, or, perhaps more likely, of other factors). Meanwhile unemployment is falling both in the EU as a whole, and in the Eurozone.

* Moreover, for some countries at least (in particular Romania and Bulgaria), the very high levels of recent inflows is likely to reflect the impact of the lifting of transitional controls in 2014; this seems likely to run its course. So even if there had been no referendum, it is plausible that immigration would have fallen back somewhat from its peak earlier this year.

* The referendum could make this fall much sharper. This is not just because of the overall economic impact of Brexit on growth, output and employment, discussed elsewhere in this issue of the Review. A Brexit-related slowdown is likely to impact some sectors and regions--such as the finance sector in London--that employ large numbers of EU migrants. Moreover, migration from some EU countries--Poland for example--appears to respond quite quickly and substantially to exchange rate changes, presumably because migrants compare the salaries that they could earn at home to what they can earn here (and, in part, remit back to family).The value of the UK minimum wage, expressed in zlotys, has fallen by almost 15 per cent already.

* To these economic reasons must be added legal and psychological ones. As discussed below, EEA citizens already resident here may have a legitimate expectation, supported by most if not all politicians, that they will be allowed to remain legally indefinitely. But there will inevitably be a prolonged period of uncertainty before we know exactly what that means. If people cannot plan with any confidence, not just about themselves but their families, they are less likely to come and less likely to stay. Moreover, not only have we seen isolated but very unpleasant outbreaks of racism (National Police Chiefs Council, 2016), with calls for EU citizens resident here to leave, but there is a much more widespread and more general sense that they are no longer welcome. There is already some anecdotal evidence that this is leading some to contemplate returning to their countries of origin or (particularly in the case of skilled workers, who may have a variety of options) moving elsewhere in the EU.

It would, of course, be hugely ironic if it was the referendum result--rather than any change in policy--that led to a significant reduction in net migration from the EU. Equally ironic, but much more serious, will be the economic consequences. The fiscal impact of a large fall in EU migration for work purposes will make the post-referendum hole in the public finances even bigger (Nardelli, 2015).

Transitional arrangements

Prior to the referendum, one concern that was the reverse of what the discussion above suggests--that a Brexit vote might lead to a sudden surge in migration to the UK, as migrants sought to establish legal residence in the UK before any restrictions were introduced. While such concerns were not a priori absurd, there does not appear to be any evidence, even anecdotal, to date to support them. Nevertheless, the question of the status of EEA nationals resident here prior to the referendum, and those moving here between the referendum date and that of any eventual Brexit, remains very much an issue. There have been vocal calls from all sides of the political spectrum for the government to guarantee the rights of EEA citizens currently living in the UK (see Daily Telegraph, 2016).

It is estimated that there are approximately 3 million EU nationals currently resident (ONS, 2016), although even this may be an underestimate (Portes, 2016b). Of these, a substantial majority have been living here more than five years, and are thus legally entitled to permanent residence (although, for the reasons described below, that does not mean they will be able to prove continuous residence for five years); however, a large number, probably close to a million, are more recent migrants. There are also, of course, far more complicated cases, involving people who have moved back and forth, or those who have family ties.

The official position remains that the status of EEA nationals post-Brexit will depend on the withdrawal negotiations, and in particular agreement on the status of UK nationals resident in other EU countries (unless of course we remain in the EEA with free movement as now, which, as discusssed below, does not seem politically feasible at present). It appears highly probable that, one way or another, a political agreement in principle will be reached rapidly. It is not in anyone's interests to make this particular issue central to the negotiations, and there is no reason to believe that any significant UK politician--and certainly no-one in any key government role wants to be in a position where they are threatening to deport large numbers of law-abiding EU nationals.

But that does not mean that achieving this outcome will be easy. First, there will be have to be some sort of cutoff date. A number have been suggested--the date of the referendum; the date that we formally notify the EU of our intention to leave under Article 50; or the date of actual legal exit. The referendum date would avert some of the obvious issues of announcing and then have people coming to establish a permanent right of residence before Brexit. But making any cut-off retrospective will make the administration even more difficult, for the reasons described below.

Whatever cut-off date is chosen, the genuinely problematic issue is to identify who was resident here on that date. This cannot be based on simple physical presence: EEA nationals here for a week for tourism would qualify, while long-established residents who happened to be out of the country would not. Obviously, any system will need to grant residence to the latter but not the former. But it is unclear what this means. Under the current freedom of movement rules, the right of EEA nationals to residence in the UK is automatic; it does not depend on any form of registration, which is purely voluntary (Cabinet Office 2016).

So there is no legal or practical obligation on EEA nationals to register their residence--not before the referendum, and not now. Nor do we systematically record UK or other EEA citizens entering or leaving the country. Passports are not stamped (indeed, for those crossing the border with Ireland, they are not even scanned). There is no population register. So it is literally impossible to determine who was physically present on a retrospective basis.

It therefore seems likely that the Home Office will have to adopt a modified version of the approach it normally uses in establishing continuous residence (as opposed to presence, or residence, on one particular date). People could submit payslips, tax returns, electricity bills, and other forms of evidence to show that they were here and for more than just a tourist visit--at some point in some particular period, or for some length of time.

Crucially, as I and others have proposed (Portes, 2016c), this could be supplemented by use of HMRC and DWP systems recording the issuance and usage of National Insurance numbers.

However, for any such system to be remotely workable, it will--in contrast to the present system, which requires the completion of a lengthy form and the provision of extensive documentary evidence, which is resourceintensive to process--have to be applied in a light way, with relatively loose criteria, and based on a significant element of self-certification. There is simply no way in which the Home Office--understaffed for its current tasks--could cope with having to examine and verify, on a case by case basis, the documentation of up to 3 million people. The result would be lengthy delays and a very high volume of appeals and legal challenges, placing further strains on the system.

Nevertheless, whatever system is adopted, there will have to be definitions, criteria, rules, guidance and a bureaucratic process, covering not just simple cases but much more complex ones: for example, the status of non-resident spouses who would not qualify for residence under the normal rules; parents of UK-born children who themselves do not qualify, etc. All this will have to be set out in laws and regulations (which will, no doubt, be subject to challenge on human rights and other grounds). This in turn means that, by definition, some people will be excluded: people who were here for several years but not quite at the right time; people who don't have or never had documents. Such cases are in turn disproportionally likely to be among relatively disadvantaged groups, such as the homeless, destitute and the mentally ill (Fitzpatrick, Johnsen and Bramley, 2012). Of course, all these issues arise under the current immigration system for non-EEA nationals, but the number of such cases could easily be in the tens, if not hundreds of thousands.

Given this inevitable complexity, it would be sensible for the government--without prejudice to its negotiating position--to try to scope out the size of the problem, so as to ensure that whatever system eventually adopted is administratively feasible. Once again, HMRC and DWP data present an opportunity here. It would be feasible to conduct a random sample of EEA nationals who have registered for NINos in the past ten years and to match that with the data it already holds on their tax and benefit records). How many are still in the UK? How many have already acquired UK citizenship? How many have married a UK citizen and/or have UK-born children? How many would actually want to take up UK citizenship and/or leave/remain? At the moment neither the government nor external researchers have more than a vague idea of the answer to any of these questions.

After Brexit

The issues described above arise as a consequence of the referendum. But they do not impact directly on the longer-term issue of what UK immigration policy will actually look like after Brexit. As I will discuss below, there are a wide spectrum of possibilities for the nature of any future system. However, some overarching observations are likely to be applicable to all possible eventualities.

* Immigration control (of EEA nationals) does not (mostly) mean border control. It does not seem likely or feasible that we would restrict EEA nationals' right to enter the UK without a visa, given that we do not do so for most other developed countries (although it is possible that some sort of electronic pre-entry system, like ESTA in the US, will eventually be introduced for movements to and from the remaining EU. This does not, however, constitute a visa system). A fully fledged visa regime for EEA nationals would be hugely disruptive to trade, travel and tourism (Neumayer, 2011), even leaving aside the obvious point that this would mean UK nationals would require visas to travel to continental Europe. And it would mean that they were treated materially worse than, for example, Americans or Australians, who do not need a visa to enter the UK. So EEA nationals (like UK nationals) already have their passports checked at entry and this will continue; equally, they are only denied entry in relatively rare circumstances. The range of possible circumstances will expand post-Brexit, but the proportion denied entry will (as with Americans and Australians) remain insignificant in the context of the overall immigration figures. This does not in itself directly affect immigration policy: Americans or Australians have no automatic right to work or settle in the UK. But it does mean that control over how many and which EEA nationals are allowed to work in the UK will not, in practice, be applied at the border in the vast majority of cases. As with other non-visa nationals, like Americans, it will be applied in the workplace; employers will have to verify that EEA nationals are entitled to work in the UK, just as they currently do for non-EEA nationals. Talking about 'border controls' for EEA nationals or reintroducing 'controls' over 'who enters the country' misses the point (from an immigration policy perspective) almost entirely.

* It follows from (i) that any plausible system, even one that preserves relatively free access for EEA nationals to the UK labour market, is likely to imply a significant impact in the regulatory burden on business. Employers will be obliged to check, for all EEA nationals, that they have the right to work in the UK, which will no longer be simply a matter of showing a passport. Depending on the system, these rights might be time-limited or dependent on the possession of a valid NI number or work visa. Experience with non-EEA nationals suggests that such a system is far from straightforward. Indeed, it would appear reasonably safe to assume that the consequent extra regulation would, in itself, more than outweigh any remotely plausible gains from reducing 'EU red tape' post-Brexit.

* A further consequence is a likely increase in illegal working. Just as previous extensions of free movement rights to the citizens of new Member States reduced illegal working (both directly, by regularising the position of those already here, and indirectly, by allowing employers to recruit EEA citizens who could work legally rather than non-EEA ones who could not), any controls will have the reverse effect. The fact that EEA citizens will have visa-free access means that this will be considerably more difficult to control than it is for non-visa nationals.

Finally, there is a broader point, often overlooked in this debate. It is often assumed that if the UK could impose restrictions on migration from the EU, the impact would be to reduce migration for unskilled and/ or low paid work, while having no impact on skilled migration; possibly even resulting in an increase. This ignores the nature both of migration systems and outcomes. First, no system can select perfectly, or even close to it. The view that we can devise an immigration system that allows in those, and only those 'immigrants that have the skills we need' implicitly assumes both the feasibility and the desirability of a centrally planned labour market. It is particularly ironic that such assertions are frequently made by those who would reject central planning as a mode of economic organisation in any other context.

Second, this also ignores the fact that migration is not just a matter of the UK choosing migrants; migrants have to choose us. Even if we wish to remain open to skilled migrants from elsewhere in the EU post-Brexit, they may not choose to come here (or remain here). Almost any system is likely to involve some increase in bureaucracy, as well as some reduction in the rights of those who come. When the Chancellor says that "We would use it [new controls] to facilitate movement of highly skilled people between financial institutions and business" that is all well and good; but it does not deal with the rights of their spouses to work and study, of their children to citizenship, access to the NHS and so on. And beyond this are the wider psychological and attitudinal consequences described above. Already, there is some anecdotal evidence that it is more difficult for UK-based businesses and institutions to recruit skilled EU workers (or retain the ones they already have) (The Guardian, 2016).

The parameters of a new system

Before the referendum, it appeared that a Brexit vote would mean that the UK faced a clear choice on immigration policy (Portes, 2016b). If we wanted as far as possible to retain access to the Single Market--either by maintaining membership of the European Economic Area (like Norway) or via a series of bilateral agreements (like Switzerland)--then we would need to accept that freedom of movement would continue much as now. At the same time, as noted above, it is impossible to view the referendum result as anything other than a rejection of free movement in its current form. Some degree of control of EU migration for work purposes would appear to be a political necessity. As the Prime Minister put it: "we are not leaving the European Union only to give up control of immigration again".

However, this in itself tells us little about what a new system would look like. So far, political debate has been dominated by discussion about whether the UK should adopt an 'Australian-style points system', as advocated by Vote Leave during the campaign, or some variety of the current 'work permit' system that currently applies to non-EEA migrants coming to the UK to work.

This debate is almost entirely vacuous. It appears certain that we will move to a system where the UK government sets out some criteria (skills, occupations, salaries, numerical limits etc) for work-related migration from the EEA, as it does at present for non-EEA migration. Such a system could be described as an 'Australian-style' system, since this is what Australia (like many other countries) does; it could also be described as a work permit system, since those who qualify receive permission to work here. So this debate tells us essentially nothing about what the overall system will look like post-Brexit.

Instead, it is helpful to consider the options along two dimensions:

* will the new system give a considerable degree of preference to EEA citizens, even if not full free movement, compared to those outside the EEA, or will it treat all non-UK citizens equally (with the possible exception of Irish citizens, not discussed here);

* will the new system be relatively liberal--accepting perhaps an increase in skilled migration from outside the EEA at the same time as reducing EU migration; or will it be restrictive, with the overarching objective still being to hit the government's target to reduce net migration to the tens of thousands?

EEA preference

Should the UK wish to preserve, as far as possible, the current levels of economic integration with the rest of the EU, at the same time as ending political integration and regaining control over migration from the EU, it is likely to have to accept that labour mobility between the UK and the EEA will continue with considerably fewer restrictions than currently apply to non-EEA nationals. This implies that while free movement would not continue as now, we would not move to an immigration system that gives effectively equal treatment to EEA and non-EEA nationals; there would still be a considerable degree of preference for the former. The objective would be to be able credibly to claim to other EU countries that we were retaining a system that preserved the essential principles of free movement, albeit with some restrictions and modifications, while also being able to maintain credibly to a domestic audience that the UK had regained control over migration from elsewhere in the EU. Is it feasible to square this circle, and what might this entail, within the limits of administrative and political feasibility?

There are a number of ways that a system that imposes restrictions on EEA nationals that are more restrictive than the current system, but less so than that applying to non-EEA nationals, could work:

* To oblige EEA nationals who want to work legally to apply for a work visa, as non-EEA nationals, but with less restrictive rules. This could mean lower qualification thresholds, a wider variety of occupations for which work visas were automatically issued, a separate and higher quota for Tier 2 visas, fewer or no restrictions on intra-company transfers, and so on. EEA nationals would presumably, unlike most non-EEA nationals, be permitted to apply for work visas from within the UK as well as from their home countries. It should be noted that the suggestion made by some that EEA nationals should need a job offer prior to arrival, but with no restrictions on the nature of that job (occupation or skill level), appears pointless; it would add an extra layer of bureaucratic regulation, but without additional restrictions would add almost nothing to the degree of control exercised by UK authorities in practice.

* To impose no specific restrictions with respect to occupation or skill level, but simply an overall ceiling or quota. This could be on the issuance of work visas (as with the current Tier 2 system), or it could operate through the issuance of new National Insurance numbers to EEA nationals. In either case, once the ceiling were hit, any further EEA nationals seeking to work in the UK would have to apply through the system that currently applies to non-EEA nationals. This would not stop them travelling to or living in the UK, but they would not be able to work legally.

* A further variant would be to (re)-introduce sectorspecific migration schemes, like the Seasonal Agricultural Workers Scheme, which allowed agents to arrange the entry of a limited number of workers (in practice, mostly students) from specific countries. The agents were required to ensure that the workers had employment, accommodation and that they returned to their countries of origin after the specified period.

Each system (or a combination) would in principle be feasible, albeit complex. The advantage of the former would be that it would at least partially address the concerns of those who complain that, unlike non-EEA nationals migrating for work purposes, we do not 'select' EEA nationals by occupation or skill level, and as a consequence a very large proportion work in low-skilled or low-paid jobs. The disadvantage, however, is that it would replicate the bureaucratic and inflexible Home Office work visa system, albeit at a different level. This would require significant extra resources, which are unlikely to be forthcoming, and even if properly resourced would result--if the current system for non-EEA nationals is anything to go by--in large costs to business and a significant reduction in labour market flexibility.

The second option also has disadvantages--it would mean that much continued EEA migration would be for relatively unskilled or low-paid jobs. Another possible downside might be an increase in irregular work. But against that, it could be administered relatively cheaply and in a light-touch way--the only obligation on employers would be to verify that an EEA national had a valid visa or National Insurance number. And, although it is difficult to judge at present, since it has at least some resemblance to the type of 'emergency brake' currently available (though never used) by EEA members, it is likely to be considerably easier to negotiate with the remaining EU member states than something which looks like a watered-down version of the system applying to non-EEA nationals. Finally, it also has the potential advantage that if EEA migration does indeed fall sharply over the next two years, as the UK economy weakens and EEA nationals feel less welcome here, any quota may in practice not have much impact.

Sector-specific schemes, while they may appear attractive to Ministers faced with lobbying from industry or trade groups, are likely to be significantly more difficult to implement and enforce in sectors which are considerably more diverse (geographically and in terms of the number and nature of businesses operating in them) than agriculture (see Rolfe and Hudson Sharpe, 2016). It is not obvious how a sectoral scheme would work in catering and hospitality, for example. Nevertheless, they may have a role to play.

Whatever system is adopted would inevitably be bureaucratically and administratively complex, further complicated by the position of EEA nationals who have already exercised their free movement rights. However, given the alternative--UK exclusion from the Single Market and an end to free movement, both of which would be economically damaging, both to the UK and to the remaining EU--some option on these lines may in practice be the least bad option from an economic perspective.

But this outcome is far from guaranteed. It may not be negotiable: there are those who argue that, if the UK wishes to retain all or most of the privileges of membership of the Single Market, no meaningful restrictions on free movement, of the sort described above, are feasible. They may be right; negotiations that traded off such controls for concessions from the UK side will be complex legally, economically, and politically. The UK cannot have its cake and eat it. Nor is it the preferred option of many of those who campaigned for a Leave vote.

Overall policy--restrictive or liberal

A separate, but related issue is the extent to which any new system is primarily directed at reducing overall immigration levels, or whether it also takes into account the likely economic downsides of significant further restrictions. A restrictive policy--which would be consistent with the previously expressed views of Prime Minister May--would aim to reduce migration to the government's target of 'tens of thousands'. Migration Watch (2016), for example, argue that applying the same migration rules to EEA nationals as currently apply to non-EEA ones would reduce net migration by about 100,000. This would be 'upward harmonisation'--that is, the new system for work-related migrants would be at least as restrictive as that currently applying to non-EEA migrants. Combined with reductions in EU migration resulting from the other factors I set out above, as well as further downward pressure on non-EU migration, this could lead to quite large falls in overall migration levels (although it would be foolish to speculate on whether this would allow the government to hit its net migration target).

The economic consequences would be significant. As Nardelli (2015) points out, in the short term a reduction in migration on this scale would jeopardise the government's ability to achieve its fiscal targets. Looking at the longer term, Lisenkova and Sanchez-Martinez (2016), using an overlapping generations model to simulate the impacts of migration to 2060 for the UK, estimates that a reduction in EU net migration by two thirds requires an increase in the tax rate on labour income of about 1.7 per cent to preserve budget balance. Wadsworth et al. (2016) summarised: "We cannot be precise about the size of the losses from restricting immigration following a Brexit.... At the national level, falls in EU immigration are likely to lead to lower living standards for the UK-born. This is partly because immigrants help to reduce the deficit: they are more likely to work and pay tax and less likely to use public services as they are younger and better educated than the UK-born. It is also partly due to the positive effects of EU immigrants on productivity."

It is important to emphasise, as I explain above, that such falls would not come exclusively, or even necessarily mainly, from reductions in migration for low-skilled work; a substantial reduction in high-skilled migration would also be likely, even if not the policy objective.

At a sectoral level, the consequences of restrictions for some industries that rely on migration would be significant. At least in theory, it is possible to construct a plausible economic rationale; while restrictions would cut growth in the short term, it might over time incentivise firms into productivity-enhancing investment or training. However, there is little or no evidence that occupational or sectoral usage of migrant labour is associated with lower levels of training (MAC, 2016). Other research has focused on migration from outside the EU and in relation to higher-level skills, but has similar findings (George et al., 2012; CIPD, 2014). In low paid sectors the difficulty of attracting and retaining young people appears to act as a disincentive to training, rather than the availability of migrants (Rolfe and Hudson-Sharp, 2016).

The evidence on the impact on wages suggests that there a restrictive policy might have some (relatively small) positive direct impact on wages for low skilled workers, although little or none for medium and highly skilled workers (Nickell and Salaheen, 2015). However, the impact on incomes would be more than offset by the wider negative economic and fiscal impacts. Lisenkova and Sanchez-Martinez (2016) suggest that the small positive impact on unskilled wages from lower migration is more than offset by the tax rises (or public spending cuts) resulting from the negative fiscal impacts, so real post-tax incomes would fall.

A more liberal alternative would be 'downward harmonisation'; that is, the new system for work-related migration would be less restrictive than that currently applying to non-EEA nationals (although, obviously, more restrictive for EEA nationals to whom free movement currently applies). This would mean that immigration continued to run at historically fairly high levels; it could, however, support a rebalancing from unskilled jobs to skilled migration, as well as from EU to non-EU migration. What would be the likely impacts?

* If appropriately structured, a relaxation of controls on skilled migration could potentially relieve some of the barriers to growth imposed by current government policy. At the moment, the quota for Tier 2 skilled workers is preventing some companies from recruiting for skilled jobs, and the government intends to further tighten restrictions, despite the evidence that such restrictions are likely to further damage productivity and growth (MAC, 2016).

* A particular issue would be the impact on relatively low paid, but skilled, occupations in the public sector, such as nurses. Current restrictions on non-EU migrants have led to significant staff shortages. If a new system precluded EU migrants as well in these sectors, this would exacerbate the problem; if, on the other hand, it allowed more non-EU migrants as well, matters might improve.

* As noted above, the consequence for some sectors that rely heavily on migration from within the EU to fill relatively low skilled jobs could be significant.

* The impacts on wages would be stronger and, for the unskilled, more positive in this scenario. Rebalancing towards higher skilled migration could, in principle, both raise wages for the lower skilled and improve the fiscal impacts of migration, boosting post-tax incomes.

From an economic perspective, a relatively liberal approach would clearly be preferable, and would minimise the damage resulting from extra regulation, as well as any wider economic consequences of Brexit. However, there is little to suggest that this is the current direction of travel. The Prime Minister, in particular, has emphasised that reducing immigration takes precedence over economic considerations.

Conclusion

The spectrum of options for UK immigration policy post-Brexit remains wide--both as to whether the treatment of EEA and non-EEA nationals is harmonised, and the degree of restrictiveness of any new system. However, almost any plausible outcome will result in an increase in regulatory burdens on business; a reduction in the flows of both unskilled and skilled workers; and an increase in illegal working. The key question for policymakers will be how to minimise these negative impacts while at the same time addressing domestic political demands for increased control without antagonising our EU partners to the point of prejudicing other key aspects of the negotiations. This will not be an easy task.

NOTES

(1) It is important to emphasise that these figures represent net changes in employment, not 'new jobs' or 'new hires'. Labour market flows are far greater in magnitude, and the UK born account for the vast majority of both new hires and job separations. See Portes (2016 and 2016a).

REFERENCES

Ashcroft (2016), 'How the UK voted on Thursday, and why', http:// lordashcroftpolls.com/2016/06/how-the-united-kingdom-voted-and-why/.

Cabinet Office (2016), https://www.gov.uk/government/news/ statement-the-status-of-eu-nationals-in-the-uk.

Carozzi, F. (2016), 'Brexit and the location of migrants', Spatial Economics Research Centre, LSE, July. http://spatial-economics. blogspot.de/2016/07/brexit-and-location-of-migrants.html.

CIPD (2014), 'The growth of EU labour. Assessing the impact on the UK labour market', CIPD http://www.cipd.co.uk/binaries/the-growth-of-eu-labour-assessing-impact-uk-labour-market_2014. pdf.

Clarke, S. (2016), http://www.resolutionfoundation.org/media/blog/ why-did-we-vote-to-leave-what-an-analysis-of-place-can-tell-us-about-brexit/.

Colantone, I. and Stanig, P. (2016), 'The real reason the U.K. voted for Brexit? Jobs lost to Chinese competition', Washington Post, [https://www.wash ingtonpost.com/n ews/monkey-cage/ wp/2016/07/07/the-real-reason-the-u-k-voted-for-brexit-economics-not-identity/].

Daily Telegraph (2016), letter available at http://www.telegraph. co.uk/opinion/2016/07/02/letters-eu-citizens-who-have-already-settled-in-britain-must-be/.

Fitzpatrick, Johnsen Bramley (2012), http://www.feantsaresearch. org/IMG/pdf/article-2-4.pdf.

Goodwin, M.J. and Heath, O. (2016), 'The 2016 Referendum, Brexit and the Left Behind? An aggregate-level analysis of the result', forthcoming in Political Quarterly.

Guardian, The (2016), https://www.theguardian.com/education/2016/ sep/25/brexit-may-force-15-of-staff-at-uk-universities-to-leave-warns-group?CMP=Share_iOSApp_Other.

George A, Lalani M., Mason G., Rolfe H. and Rosazza Bondibene, C. (2012), 'Skilled immigration and strategically important skills in the UK economy', Migration Advisory Committee.

Her Majesty's Revenue and Customs (2016), 'Further statistics on the tax contributions of EEA nationals in 2013-14', https:// www.gov.uk/government/statistics/further-statistics-on-tax-contribution-of-eea-nationals-for-2013-to-2014.

Kaufman, E. (2016), 'It's not the economy, stupid', LSE Brexit blog, http://blogs.lse.ac.uk/politicsandpolicy/personal-values-brexit-vote/.

Lisenkova, K. and Sanchez-Martinez, M. (2016), 'The long-term macroeconomic effects of lower migration to the UK', NIESR Discussion Paper 460, http://www.niesr.ac.uk/sites/default/files/ publications/dp460.pdf.

Migration Advisory Committee (2016), Review of Tier 2: Balancing migrant selectivity, investment in skills and impacts on UK productivity and competitiveness, London: Migration Advisory Committee.

Migration Watch (2016), 'UK immigration policy outside the EU', available athttp://www.migrationwatchuk.org/briefing-paper/371.

Nardelli, A. (2015), 'Osborne reliant on rising immigration levels to achieve budget surplus', The Guardian, December, available at http://www.theguardian.com/uk-news/2015/dec/01/ osborne-reliant-on-rising-immigration-levels-to-achieve-budget-surplushttp://www.migrationpolicy.org/article/free-movement-europe-past-and-present.

National Police Chiefs Council (2016), http://news.npcc.police.uk/ releases/tackling-hate-crime-remains-a-priority.

Neumayer, E. (20II), 'On the detrimental impact of visa restrictions on bilateral trade and foreign direct investment', Applied Geography, 3I (3), pp. 90I-907, http://eprints.lse.ac.uk/37226/.

Nickell, S. and Salaheen, J. (2015), 'The impact of immigration on occupational wages: evidence from Britain', Staff Working Paper No. 574, available at http://www.bankofengland.co.uk/research/ Pages/workingpapers/2015/swp574.aspx.

Office of National Statistics (2016), Labour Market Statistics, August.

Portes, J. (2015), 'Labour mobility in the European Union', in Durlaf, S.N. and Blume, L.E. (eds), The New Palgrave Dictionary of Economics, online Edition, Basingstoke, UK: Palgrave Macmillan. http://www.dictionaryofeconomics.com/article?id=pde2015_ L000248> doi:I0.I057/9780230226203.3943.

--(2016a), 'Immigration, free movement and the UK Referendum', National Institute Economic Review, 236, May.

--(2016b), '(Under)counting European migrants in the UK', http://www.niesr.ac.uk/blog/under-counting-europeans-uk#. V9V7h5grJpkPortes.

--(2016c), 'Let them stay?' [blog].

--(2016d), http://www.niesr.ac.uk/blog/brexit-migration-andlabour-market#.V-rGY_ArJpk.

Rolfe H. and Hudson-Sharp N. (2016), 'The impact of free movement on the labour market: case studies of hospitality, food processing and construction', London: NIESR.

Vasilopoulou, S. (2016), 'Explaining UK euroscepticism: freedom of movement, employment, welfare and the UK referendum', University of York, Political Quarterly (forthcoming).

Wadsworth, J., Dhingra, S., Ottaviano, G. and Van Reenen, J. (2016), 'Brexit and the impact of immigration on the UK', Brexit Analysis No. 5, London: Centre for Economic Performance.

Wadsworth, J. and Vaitilingam, R. (2015), 'Immigration, the European Union and the UK labour market', LSE, CEP Policy Analysis, available at http://cep.lse.ac.uk/pubs/download/pa015.pdf.

Jonathan Portes, Senior Fellow, UK in a Changing Europe; Fellow, National Institute of Economic and Social Research. E-mail: j.portes@niesr.ac.uk.

Caption: Figure I. Non-UK born people working in the UK, not seasonally adjusted