Paul Collier
Page 25
migrants and any policies set by these governments. Left to the decisions of migrants, migration is liable to accelerate beyond the happy medium at which those left behind gain most from it. It would also accelerate beyond the point at which host populations gain from further migration. Migration cannot be left to the decisions of individual migrants; it must be managed by governments.
But migration policy is unavoidably complicated. To be fit for purpose, policy must get to grips with these complexities. On many of the issues, research is not yet at the stage where it can provide reliable answers. Meanwhile, official pronouncements have forfeited the trust of ordinary citizens by a continuous litany of complacent reassurance: recall that spectacularly erroneous forecast by the British Home Office as to likely migration from eastern Europe. But until the taboos are broken and the parameters of future policies are widely understood, such research will not even start. In chapter 5 I set out a schematic prediction of how migration policy might blunder into mistakes in the typical high-income society. I termed it the political economy of panic. I now return to precisely the initial
256 RETHINKING MIGRATION POLICIES
conditions that produced that disturbing policy sequence and propose a different one.
As in the political economy of panic, the initial configuration of the migration function and the diaspora schedule implies that there is no equilibrium. In the absence of controls, migration and the diaspora will expand without limit. However, instead of leaving migration to accelerate until the point of policy panic, the government of the host country now adopts a package of policies designed around ceilings, the selection of migrants, the integration of diasporas, and the legalization of illegal immigrants.
Ceilings
At minimum, the task for migration policy is to prevent its acceleration to rates that would become damaging, both for those left behind in poor countries of origin and for the indigenous people of host countries. Migration has not yet generated such damage, so there is no need for policies of panic. But we should recognize that fundamental forces will lead migration to accelerate and that preventative policies are greatly superior to reactive ones. Indeed, I suspect that by putting effective preventative policies in place, mainstream politicians would stymie the current appeal of extremist parties to ordinary citizens and avert the conditions under which that appeal might spread. What is the rationale for ceilings? It unites enlightened self-interest and compassion.
The argument from enlightened self-interest is preventative: it does not suggest that migration has already caused net damage to high-income societies. The economic rationale is that continued accelerating migration would drive wages down for indigenous
workers and seriously dilute public goods. There are practical limits to how rapidly jobs markets in high-income countries are able to
MAKING MIGRATION POLICIES FIT FOR PURPOSE 257
generate high-productivity employment: they are already struggling. At the moderate rates of migration experienced for most of the past half-century, which happened to be coincident with prolonged boom conditions, favorable offsetting effects sustain and indeed modestly enhance wages. But these effects cannot be extrapolated to what would happen in the absence of migration controls.
The social rationale is that continued acceleration would increase diversity to the point at which it undermined mutual regard.
The case from compassion is that the neediest people in the world are not the migrants from poor countries. Migrants are usually drawn from the better-off in their own countries because the poorest cannot afford the costs of migration. The neediest are the people who are left behind. This is the great moral challenge of our age, and softheaded-ness about migration is not the remedy. China would continue to gain from accelerating migration, but Haiti would not, and it is Haiti that we should be concerned about, not China. While migration at moderate rates helps these people, even present rates of migration are most likely beyond the happy medium at which it is most beneficial to them. At the margin, migration is already handicapping their struggle out of poverty. The argument from compassion thus implies both more
urgent and more restrictive policies than the argument from enlightened self-interest.
So there is a sound case from both self-interest and compassion for ceilings on migration. Such policies are not a vestige of a bygone age: accelerating mass migration from poor societies to rich ones is a new, and indeed prospective, phenomenon analogous to global warming. As with global warming, we do not yet have an adequate research base on which to model it in the necessary detail, but it is already evident that controls will become increasingly necessary in the next few decades. Growing awareness of climate change is teaching the high-income societies to think long term and to consider the
258 RETHINKING MIGRATION POLICIES
potential risks of carbon emissions. Migration policy is analogous: indeed, the two processes share the essential feature that flows in excess of a threshold accumulate into stocks. In respect of climate change, analysts have realized that the safe rate of carbon emissions is derived from the safe stock of carbon dioxide in the atmosphere.
In respect of migration, the equivalent concept is the safe size of the unabsorbed diaspora. The diaspora is the accumulated stock of unabsorbed migrants, so it is the diaspora that measures the impact of migration on diversity. It is the degree of diversity that should be the ultimate objective of migration policy, not the rate of migration itself. Analogous to climate change, we do not know how large an unabsorbed diaspora would need to be before it significantly weakened the mutual regard on which the high-income societies depend.
Of course, accelerating migration would also at some stage reduce wages, but the weakening of mutual regard is the more important danger on which to focus because it is less obvious and probably has long lags. This makes it more susceptible to serious policy mistakes: if a society stumbles into it, it is difficult to correct. People will disagree about the risks of growing diversity, just as they disagree as to whether a risk of three, four, or five degrees of global warming is acceptable. But at least in respect of climate people are now having that discussion. The same is needed in respect of diasporas: should the ceiling on diasporas as a percentage of a population be 10 percent, 30 percent, or 50 percent, bearing in mind that left to themselves diasporas will cluster heavily in some cities? For climate change we not only have the right concepts, we are increasingly measuring them. For migration policy we have neither.
Given some ceiling to the safe size of the diaspora, whatever it might be, the next key number on which policy should be built is not the rate of migration but the rate at which the diaspora is absorbed. The core insight of our workhorse was that the sustainable
MAKING MIGRATION POLICIES FIT FOR PURPOSE 259
rate of migration that corresponds to any particular ceiling on the diaspora depends upon how rapidly the diaspora is absorbed. This rate differs massively among immigrant groups and between host societies: for example, Tongans in New Zealand have a far higher absorption rate than Turks in Germany. In most societies this key information is not even measured properly, so initially it would need to be approximated and gradually refined.
Between them, the safe ceiling on the diaspora and the rate of its absorption lead us to the sustainable ceiling on the rate of migration. A high rate of migration is only consistent with a stable diaspora if combined with a high rate of absorption. Conversely, a low rate of absorption is only consistent with a stable diaspora if the rate of migration is kept low. This ceiling on the rate of migration evidently relates to the gross flow of immigration. There is nothing outrageous about specifying a ceiling in gross terms: for example, the various lottery systems adopted in some high-income countries for controlling migration automatically specify a ceiling in terms of gross inflows. Yet the ceiling currently being debated by British politicians is for the net flow of immigration minus emigration.
This bears little relation to the concept that real
ly matters, which is the size of the diaspora. It would be pertinent only for concerns about overpopulation. I doubt that the current majority opinion in Britain that “migration is excessive” reflects anxieties about overpopulation. More likely, it reflects a vague unease that unabsorbed diasporas are getting too large. Accelerating emigration might warrant being an objective of policy in its own right: for high-income countries it is damaging to the remaining population due to the loss of skills.
Once we are able to distinguish between gross immigration and gross emigration, other important distinctions follow. Faster migration for the purpose of settlement augments the diaspora, while
260 RETHINKING MIGRATION POLICIES
draining the poorest countries of talent. In contrast, faster temporary migration for the purpose of higher education does not increase the diaspora, augments vital skills in poor countries, transfers values, and trains future leaders. A parody of Soviet central planning in the old USSR recounts how a target specified in terms of heads of cattle had been met by breeding the two-headed cow. Meeting a migration target by reducing the inflow of foreign students nests in the same category of policy design. 3
Selectivity
Having established an overall ceiling for gross migration, the next component of a fit-for-purpose public policy would be to shape its composition. The salient dimensions are household status, education, employability, cultural origins, and vulnerability.
If the right to migrate is conferred simply by a relationship or prospective relationship to an existing immigrant, all other criteria are of little moment. Dependent relatives of the diaspora will increasingly crowd out other would-be migrants as diaspora-fueled migration accelerates, and that is the end of the story. Further, generous rights to bring in relatives reduce the incentives to make remittances, the lifeline that migration provides to the poorest countries. It is therefore a crucial, albeit sensitive issue, as to how these rights are defined. I have argued that these rights only exist because the indigenous population rarely uses them. As rights, they do not meet Kant’s categorical imperative test of whether something is ethical: what if everyone did that? They are only viable because, in respect of the indigenous population, the answer to Kant is “Fortunately, they don’t.” So the reasonable extension of these little-used indigenous rights to migrants is to confer them with the same proviso: that they should be little used. As a practical
MAKING MIGRATION POLICIES FIT FOR PURPOSE 261
matter, this implies a lottery system in which migrants as a group receive the same proportion of immigration slots for their relatives as do the indigenous. Restricting the migration of dependents in this way opens up room for the immigration of workers. How
should workers be selected?
The most obviously desirable characteristic of immigrant workers is that they should be educated or equivalently skilled. If immigrants are more educated than the indigenous population, they tend to raise the wages of the indigenous; if they are less educated, they tend to lower them, at least toward the bottom of the wage spectrum. So, based on the self-interest of host societies, policy should select potential migrants based on a threshold level of education. This is becoming increasingly common in high-income
societies, although there are currently wide variations between them. As education levels continue to rise, this threshold will also need to rise. As I discussed in part 4, from the perspective of those left behind in the poorest countries, this is not ideal. The poorest countries are already suffering a brain drain, and this weakens their capacity to catch up with modernity through adopting and adapting global technologies. Further, there is some evidence that beyond a point, highly educated migrants send less money back home than those who are not so highly educated.
Beyond education comes employability. While educational criteria lend themselves to the checklist regulation of applications for immigration, they miss enormous amounts of other information that is pertinent for a working environment. Anyone familiar with universities will recognize that some of their students, and indeed some of their staff, are virtually unemployable despite being highly educated.
Government visa offices are ill-equipped to elicit such information, and the degree of discretionary power that would be handed to immigration officials were they tasked with doing so would invite
262 RETHINKING MIGRATION POLICIES
increased corruption. The sensible way for a society to use this information is to add a layer to the migration decision that is administered by firms. Having satisfied the criteria set by government, would-be migrants must also satisfy a firm that it wants to employ them. New Zealand and Germany both operate such a system. Employers have the incentive to vet the applicant, thereby taking into account a more balanced array of characteristics. Countries that select migrants only by means of mechanically applied points systems are liable to lose out to those that also vet, because they will attract people who meet the letter of the requirements but are otherwise unsuitable. 4
Beyond these work-based attributes is culture: a message of this book has been that cultures matter. Culture is what separates diasporas from the indigenous, and some cultures are more distant from the culture of the indigenous population than others. The more distant the culture is, the slower will be the rate of absorption of its diaspora, and also slower will be the sustainable rate of migration.
Yet, in one of the paradoxes of migration, in the absence of culturally differentiated controls, the culturally distant will be advantaged in migration decisions. Precisely because their diasporas take longer to be absorbed than the culturally proximate, these large diasporas facilitate further migration. So to the extent possible without transgression into racism, a fit-for-purpose migration policy sets the rights to migration from particular countries so as to offset these perverse effects of cultural distance. As an example of culturally targeted but politically acceptable differential controls, in both Sweden and Britain there is currently no restriction placed upon immigration from Poland, but immigration from Turkey is restricted because
Turkey has not been admitted to the European Union. 5
The last, though not the least, criterion is vulnerability. Although the status of asylum is abused, as a category it is extremely important. Helping the vulnerable is unlikely to confer economic benefits
MAKING MIGRATION POLICIES FIT FOR PURPOSE 263
on the indigenous population. That is not its rationale. By helping the most stressed societies, the high-income societies retain their self-respect. However, there is scope for reforming the asylum process. A fit-for-purpose migration policy would target asylum on those few countries in the throes of civil war, brutal dictatorship, minority persecution, or equivalent severe social disturbance. For the citizens of such countries asylum would be granted swiftly and generously. But this liberality would be combined with time-bound rights of residence: when peace is restored, people would be
required to return. The rationale for this rider is that postconflict countries face an acute coordination problem. Though they are desperately short of skilled people, individual members of the diaspora are reluctant to return. Only if many people return together are the prospects of the country sufficiently promising for return to be other than quixotic. Analytically, we are back to the discussion of chapter 3: the difficulties of coordinating cooperation.
But whereas there we were concerned with the fragility of existing cooperation in the high-income societies, now we are concerned with how to get coordination started in some of the poorest. The governments of postconflict states usually try despairingly to attract their diasporas back to the country, but they lack the means to engineer a coordinated return. Only the host governments of asylum-seeking migrants have this power. In the interests of these societies at the bottom of the global heap, they should use it. The purpose of asylum in conflict situations is not to confer a permanently transformed life onto the fortunate minority who are able to get out but to preserve the country’s critically i
mportant skilled and politically engaged people until it is safe for them to return to rebuild their society. The duty of rescue does not absolve the high-income societies from the duty to think through the implications of their policies.
264 RETHINKING MIGRATION POLICIES
Integration
Controlling the size and composition of migration is not the only means of containing diversity and stabilizing the size of the diaspora. The other means is to increase the rate of absorption. This opens slots in the diaspora, enabling migration to fill them up. The rate at which diasporas are absorbed depends in part upon the choice between multiculturalism and assimilation.
Absorption has turned out to be more difficult than social scientists and policymakers initially imagined. In part the switch to multiculturalism was probably a psychological response to this failure:
“What cannot be eschewed must be embraced.” But for any ceiling on diversity, the lower the rate of absorption the lower must be migration, so multiculturalism has a clear cost. It is premature to give up on integration. A fit-for-purpose migration policy therefore adopts a range of strategies designed to increase the absorption of diasporas.
The government cracks down hard on racism and discrimination on the part of the indigenous population. It adopts Canadian-style policies of requiring geographic dispersion of migrants. It adopts America-in-the-1970s-style policies of integrating schools, imposing a ceiling on the percentage of pupils from diasporas. It requires migrants to learn the indigenous language and provides the resources that make this feasible. It also promotes the symbols and ceremonies of common citizenship.
Most people who consider themselves progressive want multiculturalism combined with rapid migration and generous social welfare programs. But some combinations of policy choices may be unsustainable. Electorates have gradually learned to be skeptical of the alluring policy combination of low taxes, high spending, and stable debt offered by rogue politicians. One level up in economic sophistication, an important insight of modern international economics is