About us

Leading international thinktank and political network


Register for all the latest updates in our regular newsletter

Home Opinion Immigration's 'dark side': a challenge for the left
Immigration • Social cohesion • Welfare

Immigration's 'dark side': a challenge for the left

Paul Collier - 05 December 2014

Discussing immigration purely in economic terms neglects the need to talk about social, political and electoral realities. Labour must be conscious that the erosion of social cohesion caused by immigration can create significant challenges for the successful delivery of a left-wing programme

The spectacular rise in support for the UK Independence party has, at its root, widespread concerns among ordinary people that immigration has become excessive. The instinctive response of the intellectual left has been to dismiss these concerns as a mixture of proxies for racism and misunderstandings of economic realities. The response from the radical chic has been social condescension: Englishness is naff and the St George’s flag risible. But these responses have in turn induced a dangerously debilitating reaction from ordinary people. They infer that political elites are disconnected from the realities that they themselves face. Roma arrive on the housing estates of Sheffield, not the lovely homes of Hampstead.

Ukip’s original purpose was to advocate the withdrawal of Britain from the EU – ‘Brexit’. Its recent focus on immigration has been somewhat opportunistic. It wanted Brexit; voters wanted control of immigration. In a political masterstroke worthy of Alex Salmond it has linked the two: we cannot control immigration without Brexit. In order to make this case, Ukip has had to focus upon immigration from the accession countries of the EU. This has had the further advantage of detoxifying its message: it is not about race. Judging by the email responses I received to a recent Guardian article where I suggested that Ukip may still be tainted by racism, this has certainly worked. Immigrant supporters of Ukip told me in no uncertain terms that I was wrong.

The place to start is to strip advocacy out of analysis. Some concerns about immigration can then be shown by evidence to be groundless, while others look to be legitimate. The groundless include all the short-term economic concerns about competition for jobs and claims on benefits. There is now rather solid evidence about the effects of immigration on the British labour market. It shows that the effects are positive but very small: the most recent and convincing study finds that a decade of rapid immigration raised British wages by 0.5%  (Docquier, Ozden and Peri, Economic Journal, 2014). This has implications for both sides of the debate. Ukip’s assertions that immigrants from ‘new’ Europe are stealing jobs and lowering wages are false. But so are the self-serving counter-claims of British business, and the narratives of immigration-driven prosperity peddled by the pro-migration lobby.

Of course it is convenient for British businesses to recruit sandwich makers from Hungary. Would their business collapse without it? No, though they might have to make the jobs a little more attractive. Moving up the scale, instead of recruiting ready-trained workers from Europe, firms would have to invest more in training British youth. Would this reduce their profits? Yes, individually, but collectively it might help to reverse the labour productivity disaster. Productivity is around a fifth lower than in Germany and France and has been falling. Left to themselves, businesses have been free-riding on the training of other firms in other countries. At the top end, would the City collapse if its foreign workforce did not have the non-dom subsidy? More likely, it would recruit rather fewer foreigners and the government would get higher tax receipts.

Ukip’s concerns about tax and benefit payments have no substance in respect of immigrants from Europe, but they are correct for non-EU immigration. The recent estimate of a net cost of £118bn over 16 years works out at around £120 cost per year per British citizen. This is not negligible, but it is almost cancelled by the small gain from the higher wages that immigration has generated. In other words, the net economic effect of immigration on the rest of us has effectively been zero. Within this there have, of course, been modest winners and modest losers. Probably the group that has lost most in economic terms from the upsurge in immigration of recent years has been the immigrants who were here already. The labour market evidence suggests that immigrants compete primarily with each other. Perhaps that is why Ukip is indeed gaining some support from established immigrants. To quote one of my emails from a Muslim immigrant: ‘there is significant opposition to immigration across all religious and ethnic minorities – a fact that seems to get forgotten – and some of these are also supporting Ukip.’

If the debate was about the economic effects of the surge in migration the only conclusion would be that the passion on both sides has been misplaced. But obviously, these are not the real issues. Immigration is a social, not an economic, phenomenon. The pro-immigration lobby does not really think that it is vital for the economy; it is terrified that any other narrative would unleash hostility to immigrants. The anti-immigration lobby does not really think that immigration has caused an economic shipwreck; it is fearful that social cohesion is being destroyed.  

The way out of the fly bottle is to recognise that the proposition that past immigration has been modestly beneficial is entirely compatible with the proposition that further immigration should be curtailed. Existing immigrants have made a useful contribution, but future rapid immigration should be discouraged because of its longer-term social effects. These effects work partly through the size of the population and partly through its composition.

Immigrants overwhelmingly come to England rather than the other, less crowded, parts of Britain. By European and North American standards, England is already very heavily populated. It shows up in the severe congestion of the major cities. Rapid population growth requires continuous encroachment on rural areas, something that is already deeply resisted by English environmentalists and also has direct practical implications for flooding (see, for example, ‘Effects of Urban Development on Floods’, Konrad, US Geological Survey). England’s transport infrastructure is overstretched, and its housing stock overpriced. If rapid immigration implies substantial population growth, it is not hard to see why many ordinary people would quite reasonably be opposed to it. The case for a stable population is an entirely sensible social goal.

The important implications of immigration for social composition are not about who we should let it. The notion that Britain should win some global scramble for talent by denuding poorer societies I find troubling, and it implicitly denigrates our own population. Rather, it is about the appropriate degree of diversity in our society. Social heterogeneity is a matter of degree. It is meaningful to think that Britain is a better place than if it were as homogenous as Japan, but that it would function much less well if it came to look like a miniature United Nations. Diversity brings variety and stimulus, but it also weakens cooperation and generosity towards the less fortunate.

Of particular concern to the left is that as diversity increases, cohesion erodes, and voters become less willing to support generous welfare programmes. I describe some of the supporting global research in my book Exodus. But the research available while I was writing it investigated the effects of social heterogeneity in general and only incidentally on that generated by immigration. New, highly rigorous experimental research has focused specifically upon the effect of immigration. The Oxford political scientist Sergi Pardos and his colleague Jordi Munos find that immigration reduces the willingness of non-immigrants to finance welfare benefits out of taxation. They further investigate which type of benefit is most undermined by immigration. Worryingly, their answer is that targeted benefits are considerably more vulnerable than universal benefits.

This matters because in the difficult fiscal circumstances of the coming decade, overall expenditure on benefits will clearly be squeezed. Ed Balls has already announced a reduction in child benefits, which will inevitably be the first of many. In such a fiscal climate, the most straightforward way to protect the needy would be for benefits to be converted from being universal to being targeted. The new research tells us that immigration makes that option substantially less politically acceptable. We can have a more generous society or more immigrants, but perhaps not both.

The most distinctive feature of Ukip supporters is that they are atypically likely to be dependent upon benefits. Their fear of declining social cohesion need not be based on an irrational craving for a vanished past. It could instead be coming from an entirely rational fear of a meaner future. Ukip supporters are not alone in this desire for cohesion. Indeed, quite explicitly it is currently the core appeal of a party that has been far more successful. Yes, this was the message that the Scottish National party pitched to voters in the Labour heartlands of west Scotland. The SNP is proposing to shrink the vote base on which social policies in Scotland are based, from a socially fragmenting Britain to a cohesive Scotland. Their explicit thesis is that the political basis for generosity would thereby be increased. It has become a commonplace that Ukip is essentially the English SNP, but this is usually meant merely as an insult that both are playing the nationalist card. The deeper truth is that both are offering a reduction in social diversity. There should be no surprise that this is appealing to those who depend upon cohesion. Alex Salmond can safely posture in praise of diversity. With a straight face he can welcome immigration because both he and Scottish voters know that it will not happen: immigrants will continue to go overwhelmingly to England. The politically correct nod to diversity is a cheery conceit.

Consistent with the research on the adverse effect of immigration on generosity, since the post-1997 surge in immigration there has been a huge decline in voter willingness to accept higher taxes in return for higher social spending. Other research shows an adverse effect of increased diversity on cooperation. Modern societies depend upon a myriad of cooperation games which we know from behavioural psychology typically crack at some threshold of non-compliance. A well-researched British example unrelated to immigration that will be congenial to readers is the disastrous legacy of the poll tax, which, by breaking the cooperative consensus on tax compliance, has permanently increased evasion ( Besley, Jensen and Persson, ‘Norm Enforcement and Tax Evasion’, 2014). Here are some uncongenial examples for which immigration may be pertinent:

  • Gun culture: in the 1950s Jamaican criminals and police used guns, whereas British criminals and police did not. Unsurprisingly, when Jamaican criminals came to Britain they brought their gun culture with them. During subsequent decades British police and indigenous criminals have also started to use guns.
  • Loyalty: social scientists now think of frequency distributions of attitudes. Salmon Rushdie has recently drawn attention to the contrasting proportions of Muslim youth joining Islamic State and the British army. These are the observable tails of an unobservable distribution of loyalties that also affects less dramatic choices.
  • Otherness: as identities fragment, people start to see others as alien and so not subject to the normal restraints on behaviour. The ethnic prostitution gangs of Rotherham, Oxford and Bristol evidently regarded ethnic English girls as sub-human; the emergence of caste discrimination and slavery as common British phenomena are analogous.
  • Prison gangs: during the 1950s, rapid changes in the Californian prison population led to the formation of ethnic gangs. Despite their grim consequences they have proved impossible to eradicate (The Social Order of the Underworld, Skarbeck, 2014). With the rise of the Islamic extremist population in British prisons, those convicts who are ex-soldiers, scarcely timid individuals, have become fearful of revealing their military background. Our prisons may be in the first stage of ethnic gang formation.

At some point, fragmented identities become oppositional. With isolated exceptions English society is not yet so characterised and hopefully never will be. But more identity-divided societies such as Kenya are now at this stage. People are excessively loyal within their group: school boards cease to function properly as people refuse to sanction people of their own identity (Miguel and Gugerty, Journal of Public Economics, 2005). People are unable to cooperate across identity divides for local public goods (Miguel, World Politics, 2003). Voters support co-ethnics regardless of how they judge performance (Kenya: Policies for Prosperity, Adam, Collier and Ndungu (eds.), chapter 17, 2010). On production lines rival identity groups actively sabotage each other’s work even at cost to themselves (Hjort, Quarterly Journal of Economics, forthcoming).

Labour cannot match the SNP and Ukip in their proposals to reduce diversity. Influential Labour-supporting groups, such as the educated young, want further increases. But other, numerically more important groups, such as those on benefits, want less. A reasonable compromise within the Labour movement is for diversity to be stabilised at its present level. I suspect that the fear of the future, which is palpable among traditional Labour voters and can be seen in the polls, is due less to the present level of diversity than to an extrapolation of the recent trend.

Immigration targets and ‘Brexit’

So what would stabilising the population and stabilising social diversity imply for immigration targets and would any such targets be compatible with continued membership of the EU?

The implication of a stable population is that we indeed need some target for net immigration, although what it should be depends upon the fertility decisions of residents. That in turn depends upon child benefits. Labour increased them, inducing an increase in the birth rate to above replacement levels. The link from benefits to birth rates has been standard in economics for decades, but sceptics could be disabused by the recent work of Cohen, Dehejia and Romanov (Do Financial Incentives Affect Fertility?, Review of Economics and Statistics, 2013). They showed rigorously that fertility rates are sensitive to changes in government child subsidies, with the effect most pronounced among poorer families.  

The current high level of fertility implies there should be a target of net ‘out’-migration. Although parts of the UK, including Scotland, have had net outmigration over long periods, this would be challenging. Instead, as benefits become more targeted, it might be worth considering reining in those for larger families, subject to safeguards through grandfathering. Government-provided benefits help to set social norms and so generous provision for large families can delay the demographic transition to smaller family size that many immigrant communities will gradually go through.

The implication of stabilising diversity is that the rate of immigration should match the rate at which past immigrants absorb into mainstream society. Stabilising diversity need not be achieved entirely by reducing immigration; it can also be achieved by promoting integration. The official allergy to Englishness may, inadvertently, have inhibited the forging of a shared identity. Just as the SNP has rightly promoted the narrative that everyone who lives in Scotland can think of themselves as ‘Scottish’, so everyone who lives in England could expect to be ‘English’. ‘British’, as should have been evident from introspection, is too weak an identity to cut the mustard: something that has just been confirmed by the referendum. Instead, ‘English’ has been allowed to drift into designating a subversive ethnic identity. On the ubiquitous official tick-box identity forms, used in the NHS, museums and schools, ‘English’ is not even a permitted category, though ‘Irish’ is welcome. Even at its best, the multiculturalism agenda can produce only mutual ‘respect’ between distinct clusters of identity.

Yet the requirements of a compassionate society are far more demanding: not merely mutual respect, but mutual ‘regard’. For regard to be more than the patronising compassion of de haut en bas, it has to come from a sense of common identity. I have spent my life trying to encourage a sense of common humanity between rich and poor societies. But I have to face the reality that by far the most effective expression of mass mutual regard that mankind has achieved has been through a few inclusive national identities that embrace large populations. A common English identity is not an embarrassing anachronism, but a marvellous inheritance of immense social value which we should be protecting.

Both stabilising the population and stabilising diversity imply major reductions in the rate of immigration from the levels of the recent past. Such reductions need not be permanent: since 1945 the rate of immigration has fluctuated considerably as a result of changes in policies. A pause would, however, enable us to learn how the demographic transition among immigrant communities is proceeding, and how policies could realistically increase the pace of integration.

The more substantive issue is not what the immigration target should be, but whether it is deliverable. Commitment to a target is not politically astute if there is no way of achieving it. In particular, is Ukip correct in its claim that immigration control is fundamentally incompatible with European commission rules so that withdrawal from the EU is a necessary precursor?

This allegation creates a further choice for the established parties. They can either demonstrate that Ukip’s premise is wrong by implementing reductions while remaining in the EU. Or, they can agree with Ukip that without withdrawal immigration controls are infeasible, but argue that the consequences of withdrawal would be far worse than the consequences of continued immigration. All three parties have flirted with this second line, but it has become extremely dangerous. Rightly or wrongly, the institutions of the EU are not well-regarded by ordinary people. President of the European commission Jean-Claude Juncker was elected without the support of a single British vote and is already embroiled in embarrassment, squirming over his evident responsibility as the long-serving prime minister of Luxembourg for his country’s beggar-thy-neighbour tax haven policies. The message from the established parties to voters would be that their top concern cannot be met because an institution for which they have little respect will not permit it. Yet they must nevertheless obey because, after Brexit, President Juncker and our erstwhile partners would kick us in the teeth. Faced with this message voters might buckle under with a resigned whimper. But they might not: the ranks of the established parties are thinning out. Since Brexit would indeed be catastrophic – our erstwhile partners really would kick us in the teeth – the strategy would be irresponsible. The only alternative is to argue that the premise is false: the control of immigration is potentially compatible with continued membership of the European Union. Fortunately, that is indeed the case.

First, to state the obvious, immigration from non-EU countries is unaffected by EU rules. We are not in Schengen and so can do whatever we think best. Recall that the link between EU membership and immigration is primarily a Ukip artifice. Since no target for immigration would be country-specific, we can, if necessary, achieve an aggregate target for net immigration simply by subtracting actual net EU immigration from it and making this the target for net non-EU migration. Because official policy has lost much of its credibility with the electorate, it is vital that any such target should not be ‘met’ by a sleight-of-hand. In particular, squeezing student entry would deliver short-term reductions in numbers, but not long-term substance. Proper safeguards on abuse of student stay are indeed important, but training them is good for their home societies, and is Britain’s future road to soft power.  

Further, there is considerable scope for curtailing EU immigration within EU rules. To some extent immigration from the accession countries will gradually diminish even without policy intervention. Despite the present appalling continental European recession, the accession countries are growing: it is the old Europe of France, Italy, Spain and Portugal that is falling apart. As the accession countries converge, pressure to emigrate will diminish, just as Turkish immigration to Germany is no longer significant.

We also have considerable scope to reduce EU immigration through new policies that are compatible with ‘free movement of labour’; or at least not clearly incompatible with it. We should realise that ‘free movement’ was introduced as a piece of symbolism for the dream of a proto-state, rather than as a practical, negotiated policy designed to generate mutual benefits. As with that other piece of symbolism, the euro, it has had some damaging unintended consequences. But, unlike the euro, there is room for reinterpretation, as demonstrated by the recent decision of the European court of justice on benefits entitlement for immigrants to Germany.  

Evidently, the original surge occurred because, unlike other member countries, we chose not to avail ourselves of transition arrangements. This was clearly a serious mistake that created its own momentum: an important determinant of migrant inflows from a country is the existing stock of immigrants from that country. But we should now be asking ourselves why immigrants are queuing at Calais when they could go anywhere in the vast Schengen Area without the least impediment. Some parts of the Schengen area, such as Germany and Norway, have low unemployment and higher wages than Britain, so it is not just the attraction of jobs.

Partly it is because of the design our welfare system (both benefits and tax credits), which could be changed so as to align better with the rest of Europe. Elsewhere, entitlements are commonly based on past contributions rather than residence. Rather than demanding exemptions from European rules, we could turn the penchant for Euro-symbolism to our advantage. A smart move would be to propose that the commission sets a pan-European floor rate for social benefits. Necessarily, in view of the fiscal crisis besetting its poorest member states, these would need to be set at very modest levels. We could then legitimately adopt these rates for payments to immigrants from the EU.

Partly it is because of Britain’s easy-hire, easy-fire, ‘don’t-bother-to-train’, employment culture. The ‘great jobs machine’ is not an unmitigated triumph. We could take action to enforce our employment laws more effectively. Labour could revive its policy of identity cards: as a consequence of the Conservative campaign against them we have become a paradise for illegality.

If we reach a stage where we really need temporary controls on the immigration of European workers, we should again propose it in a smart way that minimises affront to the symbolism by proposing a policy that has general applicability and is linked to clear economic criteria for which the commission has responsibility. For example, a brake could be warranted during periods of severe macroeconomic misalignment, such as wide differences in unemployment rates or growth rates. After all, beyond the symbolism, a core function of the commission is to promote economic convergence. On the occasions when instead there is divergence, the commission must take some responsibility, and consequently permit states to take mitigating actions to deal with the fallout.

Labour’s rethink

Ukip has ingeniously linked its core obsession with Europe to the electorate’s core obsession with immigration. The attempt to counter it by the message that continued immigration is economically necessary is seen by ordinary people as a self-serving elite narrative that conceals contempt for their concerns. This is now building to a dangerous political upheaval: no party will garner the legitimacy to govern if it secures less than a third of the vote, regardless of seats in parliament.

The left is best-placed to break out of this spiral: an analogy from history is the structural break in American policy towards China. For a generation, normalisation of American relations with China was hamstrung by the myth of the ‘yellow peril’. The Democrats dared not challenge it because they would be exposed to the charge of being weak on security. Richard Nixon’s finest hour, indeed his only fine hour, was his visit to China, which broke the myth. Britain now needs policy change on immigration, but for the legitimate right the subject remains toxic because of the long shadow of Enoch Powell. Only Labour can credibly justify effective controls by a rationale of social cohesion and environmental sustainability, untainted by accusations of racism.

A rethink does not imply a renunciation of core beliefs. That the rise in diversity in England has been a net benefit to the non-immigrant population is not in dispute. Rather, it involves recognising that diversity also has a dark side and that as it increases, so do the risks. Further increases in diversity might significantly erode generosity and cooperation, and might at some point tip the balance into the range of problems now evident from Kenya’s multiculturalism. Or they might not: we cannot tell. But we have reached the range in which there are no significant economic benefits to the current population from further rapid immigration, while we already have enormous variety of choice. By the nature of these risks, and the powerful psychological forces that discount inconvenient evidence, if the only process whereby Labour policy on immigration gets rethought is to wait until social damage becomes incontrovertible, it may be extremely difficult to reverse. On the balance of gains v risks, the prudent policy is to have an immigration pause while we assess how both integration and fertility evolve. Unfortunately, this class of argument requires judgment rather than dogmatism. Acting to pre-empt deterioration in social cohesion is analogous to acting to pre-empt global warming. ‘Wait and see’ may be convenient, but it would not be wise.

Paul Collier is professor of economics and public policy at the University of Oxford, a director of the International Growth Centre, and the author of Exodus: Immigration and Multiculturalism in the 21st Century

This essay continues Policy Network's debate on The Freedom of Movement of Labour in Europe.

This is a contribution to Policy Network's work on Understanding Populism .

Tags: Immigration , Brexit , Ukip , Welfare , Labour , Social cohesion , Racism , Metropolitan elites , Disillusionment , Populism

Add comment


Enter the code shown:

The Policy Network Observatory promotes critical debate and reflection on progressive politics. It is centre-left orientated but determinedly challenges social democracy. It is pro-European but restlessly questions EU institutions and practices.

Most read this month

Search Posts

search form
  • Keyword
  • Title
  • Author
  • Date posted