How to have a sensible conversation about immigration

The poor of the world are on the move, eager to live and work in rich nations. What are the consequences? Talking about them cannot be a taboo.

Hell and high water: refugees wade through shoulder-deep water in Somalia as they prepare to board a boat for Yemen. Image: Alixandra Fazzina/Noor/Eyevine

Everyone has an opinion on migration but very few can justify it. I have reached this dismal conclusion in writing a book on the subject. I started the book out of concern about how migration was affecting those left behind in countries of emigration – I have spent my working life studying poor societies – but its scope gradually widened to include the effects on migrants and on the indigenous populations in host societies.

There is a serious technical literature that has studied various aspects of migration but very little of it has found its way into the media; nor has it been fitted together into an analytic whole. Instead, the media have been drowning in advocacy, supported by anecdote, assertion and moralising. As I read and listened, I was struck by the gulf between the strength with which opinions were held and the depth of ignorance on which they managed to remain afloat. I recognised this condition: it was the path to policy-based evidence.

The passion underpinning opinions on migration is fuelled by identities and fears. This is true on both sides of the debate but I will focus on the likely readership of this article – those who think of themselves as liberal intellectuals, my own circle. Among this group, distaste and disdain for opponents of immigration have become differentiating tests of identity. Beneath the vitriol is the fear that any concession to popular prejudice risks unleashing anti-immigrant violence.

Ever since Enoch Powell’s “rivers of blood” speech in 1968, serious discussion of migration has been taboo in British social science. I lost count of the number of times I was cautioned while writing my book Exodus not to include anything that could be ammunition for Ukip. In other words, I was told to write yet more policy-based evidence. British migration policy is too important and in too much disarray for this to be defensible.

In defying the advice not to give comfort to Ukip, I am confident that I am not unwittingly unleashing dark forces of latent violence. If indigenous mass violence against immigrants were a serious prospect, I would accept that caution was in order: the academic pulpit should be used responsibly.

Powell’s forecast of immigrant numbers was remarkably accurate but his forecast of their social consequences was grotesquely wrong. This is not because of any fortuitous management of British migration (which has been inept beyond the bounds of parody), nor of any unusual strengths in the English character. All high-income societies have developed robust conventions against intergroup violence. This is one of the defining and distinctive characteristics of high-income societies and it is a relatively recent achievement. Many poorer societies have not developed such conventions, as they depressingly demonstrate day by day. This is one reason why these societies remain poor.

As part of my research, I have come up with ten building blocks needed for reasoned analysis of migration. Some are straightforward; others are analytically tricky and you will need to chew on them. Indeed – with apologies for a self-serving remark – you will need to read the book.

Block 1 Around 40 per cent of the population of poor countries say that they would emigrate if they could. There is evidence that suggests this figure is not a wild exaggeration of how people would behave. If migration happened on anything approaching this scale, the host societies would suffer substantial reductions in living standards. Hence, in attractive countries, immigration controls are essential.

Block 2 Diasporas accelerate migration. By “diasporas”, I mean those immigrants and their descendants who have retained strong links with their home societies, rather than cutting loose and integrating into their host societies. These links cut the costs of migration and so fuel it. As a result, while diasporas are growing, migration is accelerating. Diasporas continue to increase until immigration is matched by the rate at which immigrants and their offspring are absorbed into the general population. A crucial implication of this interconnection is that the policies for migration and diasporas must be compatible.

Block 3 Most immigrants prefer to retain their own culture and hence to cluster together. This reduces the speed at which diasporas are absorbed into the general population. The slower the rate at which they are absorbed, the lower the rate of immigration that is compatible with stable diasporas and migration. By design, absorption is slower with multicultural policies than with assimilative policies.

Block 4 Migration from poor countries to rich ones is driven by the wide gap in income between them. This gap is the moral horror story of our times. The difference in incomes is ultimately due to differences in political and social structures: poor countries have political and social systems that are less functional than those in rich ones. Their dysfunctional systems persist in part because they are embedded in the identities and narratives of local cultures. Migrants are escaping the consequences of their systems but usually bring their culture with them.

Block 5 In economic terms, migrants are the principal beneficiaries of migration but many suffer a wrenching psychological shock. As far as can be judged from the net effect on happiness, the economic gains and psychological costs broadly offset each other, although the evidence on this is currently sketchy.

Block 6 Because migration is costly, migrants are not among the poorest people in their home countries. The effect on those left behind depends ultimately on whether emigrants speed political and social change back home or slow it down. A modest rate of emigration, as experienced by China and India, helps, especially if many migrants return home. However, an exodus of the young and skilled – as suffered by Haiti, for example – causes a haemorrhage that traps the society in poverty.

Block 7 In high-income societies, the effect of immigration on the average incomes of the indigenous population is trivial. Economies are not damaged by immigration; nor do they need it. The distributional effects can be more substantial but they depend on the composition of immigration.

In Australia, which permits only the immigration of the skilled, the working classes probably gain from having more skilled people to work with. In Europe, which attracts many low-skilled migrants, the indigenous poor probably lose out through competition for social housing, welfare, training and work. The clearest effect on the jobs market is that new migrants compete with existing migrants, who would consequently be substantial beneficiaries of tighter controls.

Block 8 The social effects of immigration outweigh the economic, so they should be the main criteria for policy. These effects come from diversity. Diversity increases variety and this widening of choices and horizons is a social gain.

Yet diversity also potentially jeopardises co-operation and generosity. Co-operation rests on co-ordination games that support both the provision of public goods and myriad socially enforced conventions. Generosity rests on a widespread sense of mutual regard that supports welfare systems. Both public goods and welfare systems benefit the indigenous poor, which means they are the group most at risk of loss. As diversity increases, the additional benefits of variety get smaller, whereas the risks to co-operation and generosity get greater. Each host society has an ideal level of diversity and hence an ideal size of diasporas.

Block 9 The control of immigration is a human right. The group instinct to defend territory is common throughout the animal kingdom; it is likely to be even more fundamental than the individual right to property. The right to control immigration is asserted by all societies. You do not have the automatic right to move to Kuwait; nor do the Chinese have the automatic right to move to Angola, although millions would if they could. Nor do Bangladeshis have the automatic right to move to Britain and claim a share of its social and economic capital.

It sometimes makes sense to grant the right to migrate on a reciprocal basis. Thousands of French people want to live in Britain, while thousands of Britons want to live in France. Yet if flows become too unbalanced, rights derived from mutual advantage can be withdrawn: Australia, for instance, withdrew them from Britain. The expansion of the EU has created these unbalanced situations and the original reciprocal right may therefore need modification.

Block 10 Migration is not an inevitable consequence of globalisation. The vast expansion in trade and capital flows among developed countries has coincided with a decline in migration between them.

These ten building blocks are not incontrovertible truths but the weight of evidence favours them to varying degrees. If your views on migration are incompatible with them, they rest on a base too fragile for passionate conviction.

Once accepted, the building blocks still leave plenty of room for disagreement as to their application to particular societies at particular times. Yet they do not leave room for existential confrontations between polarised positions. Liberal intellectuals want to combine rapid immigration, the multiculturalism that entitles migrants to remain within a distinct cultural community, and an egalitarian society. This is a noble vision but unfortunately the desirability of a policy combination does not ensure that it is technically possible.

In economics, there is a well-known “impossible trinity”: the attempt to combine open capital accounts with targets on exchange rates and interest rates is doomed to fail. I fear that the open door, multiculturalism and generous provision of public services may be the social policy equivalent – another impossible trinity. This is because the major social risk posed by rapid migration combined with multiculturalism is not that the society polarises but that it atomises. It’s not that England would descend into violence but that our tacit norms of co-operationand generosity would gradually be undermined.

British society depends on these norms that, for the most part, we take so much for granted that we are barely aware of them. The weight of the evidence suggests overwhelmingly that if a society fragments between an indigenous population and a variety of diaspora communities, co-operation will weaken. More surprisingly, diversity even appears to weaken co-operation within the indigenous population: as indigenous networks are disrupted, people withdraw into more isolated lives.

The evidence for these adverse effects of diversity is partly analytical, partly statistical and partly the results of experimental games. In a magazine article such as this, however, there is not enough space to present it all, so you will have to make do with a practical example of a social convention that could be threatened by immigration.

A word of caution is in order. Examples are merely illustrative: anecdotes are not analysis. They can be stacked up to illustrate anything and so, while they are handy to understand an argument, they cannot be used to assess its merit.

Britain has an unarmed police force. The rarity of this practice should alert us that it is potentially fragile. It rests on a convention among criminals that faced with unarmed police they, too, should be unarmed. Individual criminals would gain an advantage by arming themselves, so for the convention to remain in equilibrium depends on an unlikely pattern of collective behaviour among criminals – a bizarre code of decency among crooks.

This highly attractive convention is under threat. One reason is that criminals in some diasporas have brought with them entirely different conventions. Jamaica has a murder rate that is 50 times that of Britain and so its criminals have an ingrained gun culture. Somalia has had inter-clan warfare for a generation, so its criminals are socialised into extreme violence. A sufficiently high frequency of Jamaican and Somali criminals in Britain would be liable to change the behaviour of indigenous criminals: why follow a personally disadvantageous convention if many others are breaking it?

At some threshold of criminal violence, an unarmed police force would become quixotic and our society would lose something we cherish. Judging by the difficulty that other societies have had in establishing the convention, it is likely that, once lost, it could not be re-established.

Race and culture are distinct. Jamaicans and Somalis are not genetically prone to crime and violence. Anyone of any race can absorb any culture: ethnic Jamaicans and Somalis can become culturally indistinguishable from the indigenous English and many have done so; just as ethnic English criminals might adopt the norms of Jamaica and Somalia.

A diaspora is not a race; it is a cultural network. It only becomes synonymous with race if there is no cultural absorption. Multiculturalism, as its name implies, is about culture, not race. Its objective is to slow or prevent the absorption of diasporas into mainstream indigenous culture. In so doing, it reduces the rate of migration compatible with a stable diaspora and hence with stable migration.

With an open door, migration keeps accelerating beyond this level so that diasporas keep expanding, thereby increasing diversity. In turn, beyond a point diversity starts to undermine the co-operation and generosity on which egalitarian policies rest. This is why the trinity of policies judged desirable by liberal intellectuals may be impossible.

With the current state of knowledge, such questions remain open. The analytics suggest that the net effects of migration follow an inverse-U pattern: moderate migration is modestly beneficial, whereas rapid migration carries potentially large risks. We lack the research to determine where our society is along this inverse U.

My guess is that, to date, Britain has had net benefits. We do, however, know that uncontrolled migration accelerates. Consequently, at some point the costs of additional diversity would outweigh the benefits. We do not know at what rate diasporas are being absorbed. Hence, we do not know what rate of migration would be compatible with stable diasporas. We do not know at what point particular social conventions would start to crack in the face of rising diversity. We do not know what the costs of such cracks would be.

In these circumstances, liberal intellectuals who dismiss concerns about future migration, as distinct from the complaints about its past effects, are being cavalier at other people’s expense. It is the indigenous poor, existing immigrants and people left behind in the countries of origin who are potentially at risk, not the middle classes.

If future migration is to be controlled and the permitted rate related to stabilising the size of the diaspora, migration policy and the supporting data will need revision. We do not even know the rate of illegal migration in Britain. Nor do we know the size of the diaspora: it needs to be defined and measured not by place of birth but by degree of cultural separation.

The core objective of migration policy should probably be to stabilise the size of the diaspora, culturally defined. There are evident legal, political and ethical limits that will need to be respected – not least arising from the EU – but a commitment to stabilising diversity, reasonably defined and measured, remains a meaningful and implementable objective.

We do not know whether the current size of the diaspora is right but the caution appropriate to scientific ignorance and the evidence of widespread concern among the indigenous population suggest that stabilisation would be a sensible medium-term objective until we know better. To stabilise the diaspora, the measure of migration that matters is gross, not net. Net migration, which is the current policy target, is not a completely pointless measure but the only sensible use for it is in conjunction with domestic demographics: a net migration target could presumably aim in the long term to stabilise the size of the population.

Having rethought targets, we will then need to rethink the mechanics of how migration should best be controlled. All forms of temporary migration are benign and should be excluded from targets. Students from poor countries who return home are highly beneficial to those left behind. There is good research evidence that over and above the skills they take back with them, students’ exposure to the political and social values of their host societies rubs off on them: once they return, they spread these functional values. It’s true that student visas are open to abuse – students refusing to go home is not just a problem for the host society; it is debilitating for poor countries. Yet the current policy response of restricting students from coming here is damaging for such countries and it chokes off one of our most promising economic opportunities.

There is ample scope for a deal between the government and universities: more students in return for effective controls. Universities have powerful control points – the award of a degree and prepayments – which they already use without compunction to ensure that students pay their bills. The same system could be used to ensure that they leave the country on completion of their studies.

One of the key insights of recent research is that the composition of migration matters more than its rate. Migrants are more beneficial if they are skilled and employable. Whereas skill can be assessed by a points system, employability can only be assessed by employers and so a sensible requirement is that migrants should already have a job offer. Existing migrants want to bring in their dependants. This occupies migration slots that could otherwise have been occupied by skilled workers; it is likely to slow the rate at which diasporas absorb into mainstream society and increases the burden on welfare systems.

A particularly sensitive issue is the right to bring in a prospective spouse. Indigenous citizens have the right to bring in foreign spouses but it is viable as a right only because few indigenous citizens wish to use it. It may be unreasonable to extrapolate from the largely unexercised rights of the indigenous to infer a right that would be used very frequently by immigrants. A similarly contestable extrapolation to the rights of immigrants concerns social housing and related benefits. Many migrants from poor countries will have needs that exceed the low-income indigenous population but should they preempt limited provision?

Granting the right to immigrate can be part of a package of rights and obligations designed to protect the rights of the indigenous. This is particularly important where entitlement to welfare systems is needsbased, as in Britain, rather than dependent on past contributions, as in continental Europe. Migrants to high-income countries from poor countries have won the lottery: a package that protects the rights of the indigenous poor scarcely encroaches on this good fortune while defusing one of the reasons for fear.

Parents naturally want to pass on their culture to their children. This is true both of immigrants and of the indigenous English population. It has become an article of faith in liberal and official discourse that cultural diversity should be respected and promoted. However, not all cultures are equally functional. The cultures that immigrants from poor societies bring with them have some attractive features but they are implicated in the social failures from which these migrants are escaping. For example, poor societies typically have far lower levels of trust and higher levels of violence and intolerance. There is therefore a solid, objective reason why we should want the children of immigrants to absorb English culture.

Left to their own inclinations, migrants cluster. In Britain, clustering is getting more pronounced. This reduces the rate of absorption of even basic attributes such as language. High concentrations of diaspora clusters in schools are so likely to slow absorption that it may be sensible to cap the proportion permitted. Other countries have had more active policies of dispersal: location is made part of the package of rights and obligations.

The consequences of uncontrolled future immigration are potentially serious. Designing controls that are effective, just and advantageous to citizens will be complex and contentious. Yet the task cannot be ducked. In saying so, I am aware that I have broken a taboo and that a fatwa awaits me. But Exodus will not unleash the forces of prejudice. It may, however, lance a boil.

Paul Collier is professor of economics and public policy and director of the Centre for the Study of African Economies at Oxford University. His latest book is “Exodus: Immigration and Multiculturalism in the 21st Century” (Allen Lane, £20)