ascent to prosperity. Even the Glorious Revolution was not undertaken with the objective of unleashing prosperity: it was triggered by a mixture of religious prejudice and political opportunism. The English social model that emerged in the eighteenth century was replicated and improved in America. This in turn influenced social revolution in France, which exported its new institutions by force of arms across western Europe. The key point I wish to convey is that the present prosperity enjoyed in the Western world, and which is belatedly spreading more widely, is not the outcome of some inevitable march of progress. For thousands of years until the twentieth century ordinary people were poor, everywhere.
A high living standard was the privilege of extractive elites rather than the normal reward for productive work. Had it not been for a fortuitous combination of circumstances that relatively recently produced a social model conducive to growth, this dreary state of affairs would most likely have continued. In poor countries it continues still.
If the prosperity of the high-income world rests on this platform, it has crucial implications for migration. Migrants are essentially escaping from countries with dysfunctional social models. It may be well to reread that last sentence and ponder its implications.
For example, it might make you a little more wary of the well-intentioned mantra of the need to have “respect for other cultures.”
The cultures—or norms and narratives—of poor societies, along with their institutions and organizations, stand suspected of being the primary cause of their poverty. Of course, on criteria other than whether they are conducive to prosperity these cultures may be the
WHY MIGRATION ACCELERATES 35
equal of, or superior to, the social models of high-income societies.
They may be preferable in terms of dignity, humanity, artistic creativity, humor, honor, and virtue. But migrants themselves are voting with their feet in favor of the high-income social model.
Recognizing that poor societies are economically dysfunctional is not a license for condescension toward their people: people can as readily earn the right to respect while struggling against a hostile environment as while succeeding in a benign one. But it should put us on our guard against the lazier assertions of multiculturalism: if a decent living standard is something to be valued, then on this criterion not all cultures are equal.
Workers who migrate from poor countries to rich ones are
switching social models. As a result their productivity rockets upward. It is possible to get the same productivity gain if functional social models spread to the low-productivity societies, instead of transferring their people to the high-productivity societies. Ideas are ultimately decisive, and they can flow through many channels.
Societies do indeed pick up ideas and thereby transform them-
selves: in my lifetime I have seen several such episodes. Within western Europe in the 1970s Spain, Greece, and Portugal cast off dictatorships and embraced democracy. In 1989 the Soviet empire cast off communism, a transformation that reverberated around other regions as military regimes in Latin America and Africa were toppled. A remarkable wave of transformation is right before our eyes: the Arab Spring, which has transformed Tunisia, Egypt, Libya, and, as I write, shortly Syria. These transformations each demonstrate the potency of the idea of democratic institutions. At the onset of the Cold War the Soviet leader, Stalin, reportedly posed the rhetorical question, “How many divisions has the pope?” His implication that Soviet power trumped religious belief has since been revealed as precisely wrong: ideas beat guns. The question
36 THE QUESTIONS AND THE PROCESS
that should have been troubling him was, “Is the communist social model viable?” The transfer of ideas is enabling the rapid convergence of many once-poor countries with high-wage economies.
This will reduce the need for migration and may reverse it. But there is no simple institutional blueprint that just needs to be copied. Institutions, narratives, norms, and organizations do not need to look the same everywhere, but they do need to cohere.
The movement of goods can also be a substitute for the move-
ment of people. Indeed, an initial impetus to the migration of workers to high-wage countries was the need to get around the trade restrictions that rich countries imposed on imports from poor countries. In Britain the major clusters of Asian migrants in Bradford and Leicester were initially recruited to shore up the textile factories based there. The factories were no longer able to attract British workers because of rising wages in the rest of the economy.
It would have been more efficient to relocate the textile factories to Asia, as indeed happened a decade or so later. But British trade barriers on the import of textiles closed off this option. As a result, the trade protection that temporarily preserved the factories bequeathed a permanent legacy of clusters of Asian immigration. Restricting the movement of goods, as Britain did, thereby inducing an offsetting movement of people, offers no overall economic gains. But it does generate a range of social costs. It is often asserted that increasing migration is an inevitable facet of globalization. But in fact this is just lazy rhetoric. Far from the movement of people being all of a piece with other aspects of globalization, movements in goods, capital, and ideas are all alternatives to moving people.
Where it is possible to reap the productivity gain by moving ideas, goods, or money rather than people, it is surely sensible to do so.
Over the time frame of the next century, this is indeed what is likely to happen. But, as I now elaborate, these alternatives to migration
WHY MIGRATION ACCELERATES 37
are all too slow to close the massive gap in income between the poorest societies and the rich within our own lifetimes.
How the Income Gap Affects Migration
The growth of rich countries during that Golden Thirty Years and the stagnation of the poor are fundamental to understanding the origins of modern migration. The unprecedented prosperity of
that period created pressures to reopen the doors. Full employment made employers desperate for workers. It also removed the fear that had inhibited workers from collective action, and so unions expanded and became more militant. Governments were
themselves the main employer in the country, and so directly
faced a labor shortage, but they also suffered the backlash of strikes and wage inflation that accompanied union militancy. In the race for growth, bringing in workers from countries with much lower living standards looked to be an astute move. The political Left needed to recruit for the expansion of public services and infrastructure; the political Right needed immigrants who would go precisely to the bottleneck areas, thereby accelerating growth and curbing militancy. So governments loosened immigration
restrictions and indeed actively sought to attract foreign workers.
Germany targeted Turks, France North Africans, Britain those
from the Caribbean, and the United States Latin Americans; for example, America radically eased immigration through the 1965
Immigration Act.
In opening the door they could be confident that people would wish to come through it. The wide income gap gave people in poor countries a powerful economic incentive to move to rich ones.
But despite the wide gap, the initial flow of migrants was a trickle, not a flood. As I will discuss in chapter 6, there are many daunting
38 THE QUESTIONS AND THE PROCESS
impediments to international migration over and above any legal restrictions.
Economists have only recently been able to model migration
with the full array of techniques available to the discipline. The obstacle has always been that data on international migration has been hopelessly inadequate: economists could build theories, but we could not test them. Big data sets are the public capital goods of applied economics: the prolonged effort required to put them
together discourages individual researchers from doing so, and so the task falls to the internati
onal economic organizations that have sustained resources and public interest mandates. In the last few years such data sets have been trickling out, but only in 2012 did the World Bank release a major one that will likely prove to be a vital resource for analysis. Our factual knowledge has advanced more in the past five years than in the previous fifty, but even so much of our data still stops at 2000.
With this caveat, we now know three big things about what
drives international migration. One is that migration is an economic response to the gap in income: other things being equal, the wider the gap in income, the stronger the pressure to migrate. The second is that there are a myriad of impediments to migration, economic, legal, and social, that are cumulatively important, so that migration is an investment: costs must be borne before benefits can be reaped. Since poor people are least able to meet the costs of investment, this generates an offset to the pressure coming from a wide gap in income. If the gap is wide because people in the country of origin are desperately poor, their desire to migrate is likely to be frustrated. The third big thing we know is that the costs of migration are greatly eased by the presence in the host country of a diaspora
from the country of origin. 9 The costs of migration fall as the size of
the network of immigrants who are already settled increases. 10 So
WHY MIGRATION ACCELERATES 39
the rate of migration is determined by the width of the gap, the level of income in countries of origin, and the size of the diaspora. The relationship is not additive but multiplicative: a wide gap but a small diaspora, and a small gap and a large diaspora, will both only generate a trickle of migration. Big flows depend upon a wide gap interacting with a large diaspora and an adequate level of income in countries of origin.
By the 1970s the gap between the rich world and the poor was
horrendous, but then the Golden Thirty Years came to an end and growth rates in the rich world slowed. Gradually, the baton of fast growth was picked up by developing countries, starting with East Asia. By the 1980s China and India, home to a third of mankind, were accelerating, in the 1990s Latin America started growing, and since the millennium Africa has been growing. But if the gap in income is initially wide enough, even if poorer countries grow more rapidly than rich ones there is a prolonged period during which the absolute gap widens. Suppose that per capita income is $30,000 in a rich country and $2,000 in a poor one, but that the poor country grows at 10 percent and the rich one at only 2 percent. Measured in ratios the two countries are converging at a rapid rate, but the absolute gap in income increases from $28,000 to $28,400 in a year.
Measured in the dollar return on an investment in migration, migration is becoming more attractive, not less. Further, rising incomes in the countries of origin imply that the initial investment costs of migration are more affordable. Compound growth rates will eventually work their magic. If the poor country keeps on growing more rapidly than the rich one, at some stage the absolute gap in incomes will start to narrow again and the additional income will make little further difference to whether migration is affordable. But starting from a wide gap, the lag between growth rates and the narrowing of the income gap is very long. China is at last reaching the stage at
40 THE QUESTIONS AND THE PROCESS
which its absolute income gap with the rich countries is likely to narrow. But the absolute gap between the low-income countries and the rich ones will continue to widen for decades. Further, in the low-income countries, income remains so low that the costs of migration still matter: rising incomes will finance investment in migration. So although there are good prospects that poor countries will gradually catch up, for several decades the income gap will be wide enough to constitute a strong incentive to migration, and one that is actually increasing.
Migration produces diasporas, and diasporas produce migration: which is the chicken and which is the egg? For once, there is no conundrum. The prolonged period during the twentieth century in which the borders of rich countries were closed to migrants from poor countries implied that as of around 1960 there were no
significant diasporas. Starting from 1960, migration preceded the buildup of diasporas. Because diasporas were initially negligible, despite the wide income gap, even once borders were opened, there was only modest migration. In the absence of a diaspora to receive migrants, the costs of migration were too high.
The interaction of the income gap and diaspora creates a striking and straightforward dynamic: the flow of migration depends upon the gap and the previous stock of migrants. As the stock accumulates, the flow increases, so that for a given gap, migration accelerates. Economists always search for equilibrium: a point at which opposing forces balance so that the system is at rest. The system of migration could be at rest in two distinct ways. The rate of migration could stay the same instead of accelerating, or, in a more profound sense of being at rest, the net flow of people
between countries might cease. Might this simple process of interaction between an income gap and the diaspora produce either of these equilibria?
WHY MIGRATION ACCELERATES 41
Why There Need Be No Equilibrium
For a given income gap, migration would only cease to accelerate if the diaspora stopped growing. Since migration is constantly adding to the diaspora, it will only cease to grow if there is some offsetting process reducing the size of the diaspora. The diaspora is a simple concept to understand, but a tricky concept to measure. Typically, measurements use proxies such as the number of people resident in the country who were not born in it. But the pertinent concept of the diaspora is defined not by birth but by behavior. What matters for the rate of migration is the number of people who are related to new migrants and who are prepared to help them. In that sense, the rate of exit from the diaspora depends not upon rates of mortality among immigrants, but upon the transmission of culture and obligations. I am the grandson of an immigrant, but completely useless for any aspiring immigrants to Britain from Ernsbach.
Although I once went back to the beautiful village that my grandfather left, I have no connection either with its people or with other descendants of Germans in Britain: I am not part of a diaspora. But some other grandchildren of immigrants do belong to a diaspora so defined.
In most societies the boundaries of the diaspora are blurred: many people have one foot in their migrant past and the other in a mainstream future. But for purposes of analysis it is often useful to create clear categories and stylized processes that approximate to reality. We sacrifice the accuracy of a complete portrayal in return for simplifications that enable us to work out the likely implications of interrelationships. So I will consider a stylized society in which an unabsorbed diaspora gradually merges into mainstream society by a process in which each year a certain percentage of the diaspora switches into the mainstream. The process of switching may take
42 THE QUESTIONS AND THE PROCESS
many different forms. An immigrant might simply lose touch with, and interest in, the society that she has left. A child of immigrants might redefine itself as a member of the host society, as did my father. Or, over time, each successive generation of descendants from immigrant families may become more psychologically distant from their country of origin. The proportion of the diaspora that switches each year may be high or low, and I will refer to it as the absorption rate. So, for example, if each year two out of every hundred members of the diaspora were absorbed into mainstream
society, the absorption rate would be 2 percent.
Absorption rates vary depending on where migrants are from
and where they have arrived. They vary according to government policies. I discuss such influences more thoroughly in chapter 3.
But at this stage I will introduce only one straightforward influence on the absorption rate: it depends directly upon the size of the diaspora itself.
Size matters because the more interactions that a member of the diaspora has with the indigenous population, t
he more rapidly she is likely to merge into it. But while some of her interactions will be with the indigenous population, some will be with other members of the diaspora. The larger is the size of the diaspora relative to the indigenous population, the smaller will be the proportion of her interactions with the indigenous. This is because there is a practical limit to the total number of interactions that people can manage.
Typically, the total number of genuine person-to-person interac-
tions is limited to around 150. 11 So the larger the diaspora, the
fewer social interactions there will be with the indigenous, and therefore the slower the absorption rate will be. I should note that in principle, there is an offsetting effect. The larger the diaspora, the more social interactions the indigenous population has with it, and so indigenes absorb diaspora culture more rapidly. But while ever
WHY MIGRATION ACCELERATES 43
the diaspora remains a minority, the typical diaspora member will have far more contacts with indigenes than the typical indigene has with the diaspora. So if contact carries the same absorptive punch in each direction, the absorption process will be predominantly through the adaptation of migrants. Although a larger diaspora increases adaptation by indigenes, this is unlikely to offset the reduced rate of
adaptation of migrants. 12 The important implication is that the larger
the diaspora is, the slower its rate of absorption will be.
Introducing a Workhorse
We now have all three of the building blocks we need to understand the dynamics of migration. The first block is that migration depends upon the size of the diaspora: the larger the diaspora is, the easier migration will be. The second is that migration adds to the diaspora, whereas absorption into mainstream society reduces it. The third is that the rate of absorption depends upon the size of the diaspora: the larger the diaspora is, the slower its absorption. It is time to fit the three blocks together. If you possess intuitive genius you will be able to do this unaided. But most of us need some help, and this is what models are for.
Paul Collier Page 4