“LET me tell you the problem,” says Victor Trevino, Mexico's deputy consul in the border town of El Paso. “The minimum wage in the United States is $5.15 an hour. In Mexico, many people earn $5 a day. In a supermarket here, a gallon of milk costs $3. Across the border, in Juarez, it often costs 10 or 15 cents more.” No wonder they come. But what do the people in the host country gain?

Magnum

The locals didn't want the job

For most people, this question revolves around tax and welfare benefits. Do immigrants come to scrounge? Or will they be the salvation of rich countries that have promised their citizens pensions that the public purse cannot afford? There are wider questions too. Do immigrants depress the pay of native workers? Do they retard the growth of productivity? Or do they bolster global economic growth? To such questions, the answer is mostly an unsatisfying “it all depends”.

Take tax and welfare first. Asked whether he thought that the United States should open its borders to all comers, Milton Friedman, the grand old man of liberal economics, replied: “Unfortunately not. You cannot simultaneously have free immigration and a welfare state.”

At a given moment, migrants are generally net contributors to the public purse: they are disproportionately of working age, and the receiving country has not had to pay for their education. A study by Britain's Home Office estimated that the foreign-born population paid about 10% more to the government than it received in expenditure. However, a magisterial study in 1997 of the economic impacts of immigration, by America's National Research Council, found that the picture changes if one looks across time instead of taking a snapshot. In that case, the NRC found, first-generation migrants imposed an average net fiscal cost of $3,000 at present discounted value; but the second generation yielded a $80,000 fiscal gain.

Some immigrants contribute more than others. Those who come as students or relatives or asylum-seekers (as most do) may not work at all—or may not be allowed to work. In addition, immigrants tend to have larger families, to be poorer and to be more often unemployed than the native-born. The report for the Fondazione Rodolfo Debenedetti found that, in some European countries, such as Denmark and the Netherlands, welfare benefits were so generous that they probably distorted the inflow of immigrants and encouraged them to draw welfare.

As with so many other aspects of immigration, skill levels affect costs. Migrants from poor countries are much more likely to claim benefits than migrants from rich ones; unskilled migrants are much more likely than natives or skilled migrants to lose their jobs in a recession. And countries treat different classes of immigrants in different ways for welfare-benefit purposes. Asylum-seekers in the United States cannot claim welfare during the six months in which their claim is normally determined. In Germany, by contrast, asylum-seekers can claim welfare benefits while their asylum claim is processed. That can easily take years, during which time they may not be allowed to work if they have been sent to live in an area of high unemployment.

Even if immigrants pay more tax than they get back in public spending, they may create some imbalances on the way. The tax they pay may go to the national government but the spending may burden a city or state, which houses and educates them and their children. The NRC study noted that the fiscal gain from immigrants was spread fairly evenly across the United States, but that the burden on states varied, depending on the type of immigrants they attracted. In 1989-90, in New Jersey, where almost half of new immigrants are from Europe or Canada, native households paid a net $232 a year more because of immigration; in California, where more than half of all immigrants come from Latin America, they paid $1,178 a year more.

Immigrants are on average younger than the native-born, and they have more children—partly because of their ages, but also because they usually come from countries with higher fertility rates. Fertility in most rich countries, and especially in Europe, is below replacement level, so their populations will age and shrink over the next half-century. For instance, the median age in Italy may rise from 41 today to 53 in 2050. By mid-century, the population of Spain may be 22% smaller than it is now, and that of the Netherlands 10% smaller. All told, by 2050 no fewer than 34 countries will face a decline in population.

Two years ago, the United Nations Population Division tried to establish what levels of immigration might be needed to prevent such a population decline, and what might be required to maintain the existing ratio of workers to those needing support. Its findings produced international uproar. “Quite a lot of my colleagues had to cancel their holidays when the report came out,” says Joseph Chamie, head of the division. Not only did the levels of migration needed to stabilise the working-age population turn out to be large, but the flows needed to stabilise the present support ratio proved to be immense, at least in Europe. The EU would require an annual inflow of nearly 3m migrants a year, or roughly twice the present legal and illegal flow from outside the EU, to prevent the future support ratio (of those aged 15-64 to those aged 65 and over) dropping from about four at present to below three. In the United States, where the support level is currently above five, it would take just under 1m immigrants a year (or about two-thirds the present inflow) to stabilise at three.

Would it help to increase the proportion of the working-age population who actually work? Yes, said the report, but only up to a point: even if everyone of working age were in work, support ratios in 2050 would still drop below three even in the United States. Only by raising the retirement age to 75 could current support ratios be maintained.

For Europe, these calculations offer an uncomfortable reminder that its population may have peaked around 1997 and may now be declining. America's population, by contrast, will grow over the next half-century, perhaps by 40%—and four-fifths of that growth will be due to today's immigrants and their descendants. Numbers count, and Europe's resistance to immigration may count against it.

Certainly, ageing countries will attract immigrants. Demand from older people for labour-intensive services will drive up the wages of the unskilled. The choice for Europe's old may be between being cared for by legal migrants or illegal ones. However, immigration is not a solution to the strains that ageing will bring. For one thing, by 2050 fertility rates will have dropped below replacement levels even in many traditional emigration countries, including Mexico, Egypt, Brazil, the Philippines and Indonesia. If young people continue to leave these countries in large numbers, supporting the elderly there will one day become an even bigger problem than in the rich world.

Besides, young and fertile migrants grow old and their fertility rates rapidly decline. “There are no feasible migration solutions to the age-structure change and its effects on social security,” insists David Coleman, a demographer at Oxford University, who argues that integrating the existing foreign-born and their children should come first.

However, in one respect immigrants may make a substantial difference. The ageing countries of Europe face an unsustainable gap between future tax revenues and commitments to spend and to service government debt. In Germany, for example, that prospective gap is of the order of an annual 6% of GDP. If migrants make a net contribution to taxes over their lives, they reduce that debt. But even if they do not, argues DIW's Mr Brücker, they increase the number of future taxpayers. The same debt spread over more payers automatically reduces the individual burden of future taxpayers. In short, migration cannot prevent ageing, but it can significantly reduce its fiscal consequences.

To sum up, immigration changes both the size of a country's economy and the way the gains are shared out. Measurement is difficult, and results often disagree. But a number of broad conclusions emerge from all this.

First, migration probably raises the living standards of the rich (think of all those foreign nannies and waiters) and the returns to capital (hence the enthusiasm of employers for more flexible policies). It does not seem to increase unemployment among the native-born, although it may reduce their pay.

On balance, argued the report by America's NRC, the people whose wages immigration harms are mainly previous immigrants, for whom new immigrants are close substitutes in the job market. In addition, immigration seems to account for almost half the fall in wages of high-school drop-outs in the 1980s and early 1990s. But this is a smallish group, now less than 10% of the American workforce. Trade unions in the United States no longer argue for a ban on immigration, realising that this is a lost cause: instead, they want to legalise the undocumented, who are much more likely to undercut their less-skilled members than are unionised legal immigrants.

The effect on jobs depends partly on whether immigrants are complements to or substitutes for native labour. Are the immigrants doing jobs that natives might have done, or would those jobs simply not exist if immigrants were not there to do them? Advertise for a cleaner in London at twice the minimum hourly wage, and you will get no response from local school drop-outs or Liverpool's unemployed. More probably, the applicants will be from Ukraine, Colombia or Poland.

Some labour economists are puzzled that immigration does not appear to have made much impact on wages or jobs. Trade economists such as Berlin's Mr Brücker are not. They point out that an open economy may change its mix of output, leaving wages and unemployment unaltered. Indeed, in many countries, immigrants are densely concentrated in export industries such as textiles, car making and agriculture.

Immigrants also cluster in areas where the job market is tight. In Canada, half of all immigrants go to Toronto; in Britain, an even higher proportion settles in London. Harvard's Professor Borjas points out that immigrants incur much lower costs than natives in choosing to move to a particular place, because they have already decided to uproot. They gain the greatest benefit by moving to those places where their skills are in greatest demand. Not surprisingly, he finds that new immigrants are disproportionately clustered in America's high-wage states, where workers are scarcest.

That clustering helps to stop wages rising even further, and allows the entire economy to run at higher speed than might otherwise be possible. The finding has important consequences for Europe, with its lower geographical mobility and more inflexible job markets. There, the importance of immigrants as a flexible workforce is potentially even greater than in the United States.

Immigrants may boost economic growth in other ways that are harder to measure, and that depend on their skills and experience. They may bring entrepreneurial skills: almost 30% of new companies in Silicon Valley in 1995-98 were started by Chinese and Indian immigrants. Some depressed regions actively court immigrants: the chamber of commerce at Nashville, Tennessee, sees them as a source of dynamism, as does Tom Vilsack, the governor of Iowa, whose state has a meat-packing industry relying largely on Bosnian refugees.

Against that, unskilled immigrants may discourage investment. Mr Krikorian, of the Centre for Immigration Studies, argues that productivity in California's raisin industry is far lower than it could be, because of the ready supply of cheap illegal Mexicans to pick the fruit by hand. Australia, with far fewer illegals, grows grapes over trellises, which allows automated harvesting.

As far as the host country's population is concerned, then, the benefits of immigration may be modest and unevenly distributed. The NRC study estimated them at up to $10 billion a year—chickenfeed in an economy of $10 trillion. “The economic pluses and minuses are much smaller than the political and emotional salience,” says Rand's Mr Smith. In Europe, where they have been less carefully measured, they may be larger: Britain's government recently increased its underlying forecast for economic growth by a quarter of a percentage point because it now expects higher net inward migration.

For the individual immigrants, on the other hand, the potential gains are very large. This explains why many trade economists argue that humanity as a whole benefits enormously from migration. Alan Winters of Britain's Sussex University, in a study for the Commonwealth Secretariat, has tried to quantify these gains. He concludes that, if the rich countries raised the number of foreign workers that they allowed in temporarily by the equivalent of 3% of their existing workforce, world welfare would improve by more than $150 billion a year. That is bigger, he points out, that the gains from any imaginable liberalisation of trade in goods.

Dani Rodrik, a trade economist at Harvard University, has independently reached similar conclusions. Gaps in the prices of traded goods have become much smaller after many years of liberalisation. Not so gaps in the wages of similarly qualified individuals in different parts of the world. So the gains from liberalising immigration restrictions are vastly greater than those from further freeing the movement of goods or capital.

Both Mr Winters and Mr Rodrik agree that the biggest gains are from freeing the movement of unskilled rather than skilled labour, for there the wage gaps among countries are greatest. In addition, the loss of skilled people may do greater damage to the developing countries that have trained them. So both economists independently reach another conclusion: to minimise harm to sending countries, migration should be temporary.

More migration will not be popular. But neither, says Mr Rodrik, is liberalising trade. And indeed, the plight of the miner in Wales or the farmer in Japan whose way of life is destroyed by cheaper imports is not so different from that of the old lady in downtown Los Angeles whose neighbourhood has become Hispanic. All, in different ways, are victims of the march of globalisation. The best remedy is to redistribute some of globalisation's gains to its victims.

Copyright © 2002 The Economist Newspaper and The Economist Group. All rights reserved.