Paul Collier
Page 14
exceptionalism—a higher rate of innovation—but the evidence for this is not general and may depend upon the specifics of host countries and countries of origin. The potential social costs—the decline in cooperation and generosity consequent upon increased diversity and diasporas attached to dysfunctional social models—are medium run. The habits of sociability are robust in the short run to an increase in diversity. In the long run the population blends and so the initial sociability can be reestablished. Does this pattern enable us to resolve the conflict between economic benefits and social costs? For the sparsely populated countries I think it does: since long-term effects are liable to predominate, for such countries far-sightedness favors migration. But for other countries, the open door may be the short-sighted option: an unsustainable economic boom, followed by complex and prolonged social problems.
138 HOST SOCIETIES: WELCOME OR RESENTMENT?
The other way of resolving how economic gains compare to
social costs is to find a way of combining them by means of a common metric. One of the most promising recent developments in
social science has been to recognize that income is not a good metric of the quality of life. Led by scholars such as Richard Layard, some economists are reformulating the objective of public policy as the maximization of happiness. Layard has been appointed an
official adviser to British prime minister David Cameron, and the government has introduced an official measure to track changes in happiness as distinct from income. Happiness is not the only objective in life, but it is pretty fundamental. Many of the other objectives you might want to replace it with—dignity, achievement, serenity, respect—are all not so much alternatives to happiness as routes toward it.
Potentially, a measure of happiness can synthesize economic and social effects into a common metric that is meaningful for policy purposes. Fortunately, there is one such study that measures the net impact of migration on the happiness of the indigenous population of host societies, namely that of Robert Putnam. Although the focus of his work was on the effects of migration on trust and social capital, he also measured its effect on happiness.
Although Putnam did not measure the economic effects, we can
reasonably infer that they were positive. There is no reason to think that the localities that Putnam chose experienced economic effects of migration other than the usual pattern of large gains for migrants and small gains for most of the indigenous population. But he found that the effects on happiness were dominated by social costs: the higher the concentration of immigrants in a community, the less happy was the indigenous population, controlling for other characteristics. That the adverse social effects should dominate the favorable income effects of immigration would not surprise the
GETTING MIGRATION POLICY WRONG 139
scholars of happiness. They find that above a relatively low income threshold, increases in income do not generate significant sustained increases in happiness. Further, the income gains from migration for the indigenous are likely to have been modest. Happiness
studies find that social relations are far more important than small changes in income, and “hunkering down” is essentially an erosion of these relations.
Too much should not be made of a single study. Unfortunately, my trawl of the literature has not revealed other rigorous studies that measure the effect of migration on the happiness of the indigenous population. There is a vacuum that research needs to fill.
Given the present inadequate state of knowledge, all that is warranted is a note of caution to be sounded against the overwhelming enthusiasm of social scientists for open-door migration policies.
The effect of migration to date on the overall well-being of indigenous host populations appears to have been modest and ambigu-
ous. Both the economic and the social effects of migration are positive at moderate rates, but beyond that they are likely to become negative. Quite why economists in particular are such strong sup-porters of increased migration is at this stage in the analysis a mys-tery: it cannot be coming from the effects on host populations. We will see the likely basis for it in the next chapter.
The Political Economy of Panic
What migration policies do the governments of host countries
adopt, and what are they likely to adopt? Among the policies available to the governments of host countries, the one that is most hotly debated is quantitative limits on the rate of migration. But other policies are potentially more important. One range of policies can affect the composition of migrants in various respects: skill levels,
140 HOST SOCIETIES: WELCOME OR RESENTMENT?
the balance between workers and dependents, and the weightings of the social models to which migrants are accustomed. Policies can also affect the rate of absorption of diasporas into the general population. These, rather than quantitative limits, are the important policies. To see this, I am going to use the model to tell an unfortunate story of how migration and policy are likely to evolve in the absence of good analysis.
The story has four phases and is illustrated in Figure 5.1. In the first phase there are no migration restrictions, so migration increases at its natural rate along the migration function as shown by the arrows. The desire to migrate is so strong that the function does not cross the diaspora schedule, and so there is no natural equilibrium.
The continual acceleration in migration becomes a salient political M
*
M
ion
e of Migrat
Rat
DS
**
M
Size of Diaspora
Figure 5.1 The Political Economy of Panic: Quantitative Migration Restrictions
GETTING MIGRATION POLICY WRONG 141
issue, and so I will term this initial period the anxiety phase. Eventually, the government imposes a quantitative restriction, freezing the rate at M* to prevent further increases.
As a result we now enter the second phase, which I will term the panic phase. While the binding limit on the rate of migration prevents further acceleration in migration, it cannot in itself lead to an equilibrium size of the diaspora. As the size of the diaspora increases, the combination of reduced interaction with the indigenous population, the widening cultural distance in the composition of migrants, and the feedback from reduced trust all reduce the rate of absorption to the extent that beyond some size, the diaspora schedule twists back upon itself. In this policy scenario, the rate at which migration happens to be frozen, M*, is not compatible with a stable diaspora. Hence, in the panic phase, although migration is frozen, the unabsorbed diaspora continues to increase: this is shown by the arrows that march along the horizontal line marking the binding migration control. As the unabsorbed diaspora keeps growing, at some point rising social costs, such as the decline in trust within the indigenous population, and competition for social services between the diaspora and the indigenous population, generate renewed political pressure. In this scenario the only policy that the government uses is the quantitative limit on the rate of migration. So the government imposes ever-tightening restrictions.
This takes us into the third phase, which I will call the ugly phase.
This phase is ugly because until migration is reduced to a level below the diaspora schedule, no matter how much it is reduced, the unabsorbed diaspora keeps increasing, and so social costs and political pressure keep mounting. This is depicted by the arrows that trace a path from M* to M**, at which the rate of migration has been so reduced that the diaspora begins to decline.
142 HOST SOCIETIES: WELCOME OR RESENTMENT?
This takes us to the final phase of diaspora absorption. During this phase, which might take many decades, migration is severely restricted while the diaspora is gradually absorbed into the general population and social trust is rebuilt, enabling the fragile equilibria of cooperation to be reestablished.
This story of migration is not especially encouraging. There are ver
y large changes in the rate of migration: it swings from being very high to very low. This is unlikely to be optimal from any perspective. There are also very large changes in the size of the diaspora, with a prolonged period in which it may be so high as to inflict significant social costs on the indigenous population.
While it is not encouraging, it is also not inevitable. In the final chapter I will return to this scenario, starting from precisely the same objective circumstances—the same migration function and the same diaspora schedule—and show how different policies could
produce a far superior outcome.
But first, I turn from the interests of the indigenous population of high-income host countries to those of the migrants who come to join them. Other than in the guest-worker approach to migration, migrants become members of their new societies. How does this affect them?
PART 3
Migrants
Grievance or Gratitude?
This page intentionally left blank
CHAPTER 6
Migrants
The Winners from Migration
MIGRANTS ARE BOTH THE BIG ECONOMIC WINNERS and
the big economic losers from migration. If economic self-interest were the only influence on behavior, the people in poor countries should move heaven and earth to migrate to rich countries, but once there they would vote for the political parties that advocate tightening restrictions on immigration. This conclusion from the economic analysis of how migration affects migrants is sufficiently bizarre that we will take it in stages.
Why Migrants Are the Big Winners from Migration
The first stage is the unsurprising part: migrants are the big winners from migration. The big gains accrue as a result of moving from a country in which workers are paid little to one in which they are paid a lot. The magnitude of the wage differential between rich and poor countries is staggering: it is, indeed, a mirror of the overall gap
146 MIGRANTS: GRIEVANCE OR GRATITUDE?
in income between the rich world of the Organisation for Economic Co-operation and Development and the poor world of the bottom billion. We cannot infer directly from a differential in wages that by moving from a low-wage country to a high-wage country a worker would be able to earn a high wage. Something that every economist takes for granted, but noneconomists find uncomfortable, is that to a first approximation differentials in wages reflect differentials in productivity: people are more or less paid what they are worth.
Of course, we are all aware of glaring instances where this is not the case: some people are paid far more than they are worth and others too little. But if employers made large systematic mistakes in matching wages to productivity, they would go bankrupt. So the key issue becomes not whether but why workers in high-wage countries are so much more productive than workers in low-wage countries.
As a matter of logic there are only two possibilities: the productivity gap is either due to the characteristics of workers or to the characteristics of countries. Economists have studied this issue, and the bulk of the gap is due to differences in the characteristics of countries rather than the people in them. How have they reached this conclusion?
There are a few smart ways of getting at the importance of worker characteristics. One is to compare matched jobs in host countries and countries of origin. Job for job, the wage differentials are massive: for example, in moving from Haiti to the same job in the United States a typical worker would increase his earnings tenfold. 1
Another way is to compare immigrants with indigenous workers in host countries. Immigrants turn out to be almost as productive as indigenous workers. Even this is not quite killer evidence: conceivably, the most productive cream of workers in the country of origin are the ones who choose to migrate. Getting around this possibility requires a degree of ingenuity. The trick has been to find natural
MIGRANTS: THE WINNERS 147
experiments in which migration has been randomized. Random
migration may sound unlikely, but the literally random allocation process for some visas comes very close to the statistical ideal. For example, the United States conducts an annual lottery for 50,000
visas that attracts around 14 million applicants. So the lucky few who are granted a visa are very likely to take it up and are unlikely to differ from the unlucky majority. New Zealand holds a similar lottery for would-be immigrants from Tonga. Studies that have investigated whether the lucky minority are atypically productive in their home environments have concluded that they are not. Personal characteristics do not account for most of the productivity
cliff that migrants climb. 2 A final approach is to explain differences in productivity between countries solely in terms of differences in country characteristics. It reaches the same conclusion: most, though not all, differences in income between rich and poor countries are because of differences in productivity that are due to characteristics of countries rather than of individual workers. This is consistent with the explanation of income gaps that I offered in chapter 2: differences in social models.
Some differences in productivity do remain, even when immi-
grants and indigenous workers have the same level of education.
Typically, immigrants end up taking jobs that are several notches below those ostensibly appropriate for their formal educational attainments. This may reflect sheer discrimination, but it may also reflect underlying differences in skills. However, even if it does, the discount is modest relative to the wage differential between poor countries and rich ones.
That the income gap between the rich world and the bottom billion is predominantly due to the characteristics of countries rather than the characteristics of workers has powerful implications. One tells us what it will take for the bottom billion to catch up with the
148 MIGRANTS: GRIEVANCE OR GRATITUDE?
developed world. Its message is that key characteristics of these societies will need to change. It is not simply a matter of training individuals in skills. As I discussed in chapter 2, the societies of countries of origin will need to change their social models. An uncomfortable corollary is that it is not desirable for migrants from them to bring their social model to their host country. Whether or not migrants realize it, the impetus for their emigration is to escape from those aspects of their countries of origin that have condemned people to low productivity. By the same reasoning, for emigration to have a fundamental impact on countries of origin, it would be through inducing changes in their social models.
Another implication of income differences being due to country characteristics is that migrants will nation-shop. In my research with Anke Hoeffler, we have arranged global migration flows into a large matrix showing all possible host countries for each country of origin. We find that the rate of migration from a country of origin to any particular host country is influenced by not just the income gap between them but the income of the host country relative to other possible host countries.
Nation-shopping is driven not just by differences in average
incomes among possible host countries but by where migrants
expect to slot in the distribution of income. Those migrants who expect to earn below-average incomes should prefer countries with high redistributive taxation, whereas those who expect to earn above-average incomes should prefer countries that accept greater inequality. Migrants should shop for their preferred tax-cum-welfare system as well as their preferred wage. This was first recognized just as a theoretical possibility: low-skilled migrants would prefer equal countries, and high-skill migrants would prefer unequal
ones.3 But more recently this expected pattern has found some empirical support: the skill profiles of migrants to Europe and
MIGRANTS: THE WINNERS 149
America differ in a way that matches the prediction. 4 Europe is more equitable, with the most generous welfare systems in the world, and it tends to attract migrants with a lower skill profile, although this composition of migration may also be explained by other factors
.
The final implication is that simply by shifting a worker from a dysfunctional society to a more functional one, her productivity can increase tenfold. This is an order of magnitude greater than any other process by which productivity can be increased. Globally, the vast edifice of technological research enables productivity to inch forward by a couple of percentage points per year. Over the past two decades China has indeed been an astounding exception:
cumulatively it has also achieved around a tenfold increase in productivity. But this is without historical precedent and has required a staggering willingness to postpone consumption: despite China’s initial poverty, for the past two decades, half of all income has been saved and invested. Yet simply by getting on a plane, workers can reproduce that hard-won Chinese productivity boost. This is why economists get so excited by enhanced migration: it is the closest the world economy comes to a free lunch.
Who Should Get the Gains from Migration?
Who should eat that free lunch? That is, to whom should the productivity gains from migration accrue? In a market economy the default option is that productivity accrues to the producer: workers are paid according to their productivity. So, in the absence of a policy override, the gains from migration will accrue to migrants.