THE liberty that makes economic integration possible is desirable in itself. In addition, advocates of globalisation argue, integration is good for people in material terms—that is why free people choose it. Sceptics disagree on both points: globalisation militates against liberty and democracy, they say, and while it makes some people who are already rich even richer, it does this by keeping the poor in poverty. After all, globalisation is merely capitalism writ large. A later chapter of this survey will deal with the implications of globalisation for democracy. But first, is it true that globalisation harms the poor?

In a narrow sense, the answer is yes: it does harm some of the poor. Free trade and foreign direct investment may take jobs from workers (including low-paid workers) in the advanced industrial economies and give them to cheaper workers in poor countries. Thanks to the North American Free-Trade Agreement (NAFTA), for instance, there are no tariffs or investment restrictions to stop an American manufacturer closing an old factory in the United States and opening a new one in Mexico.

Sceptics score this strategy as a double crime. The rich-country workers, who were probably on low wages by local standards to begin with, are out of work. That increase in the local supply of labour drives down other wages. Meanwhile, the poor-country workers are drawn into jobs that exploit them. How do you know that the poor-country workers are being exploited? Because they are being paid less, often much less, than their rich-country counterparts got before trade opened up—and in all likelihood they are working longer hours in shabbier premises as well. The only gain from this kind of trade, the indictment continues, accrues to the owners of the companies who have shifted their operations from low-wage factories in industrialised countries to poverty-wage factories in the south.

Some of this is true. Trade displaces workers in the industrialised countries; other things equal, this will have some depressing effect on the wages of other workers; and pay and conditions in developing-country factories are likely to be worse than in their rich-country counterparts. But whereas the displaced rich-country workers are plainly worse off than they were before, the newly employed poor-country workers are plainly better off. They must be, because they have chosen to take those jobs.

Altogether, given freer trade, both rich-country and poor-country living standards will rise

As for profits, yes, that is the spur for moving production to a lower-wage area. But no company can expect to hang on to this windfall for long, because it will be competed away as other companies do the same thing and cut their prices. That lowering of prices is crucial in understanding the broader benefits of the change. It is what makes consumers at large—including poor consumers—better off, raising real incomes in the aggregate.

What about the rich-country workers who are not displaced, but whose wages may nonetheless come under downward pressure? It is hard to generalise. On the one hand, their wages may fall, or fail to rise as quickly as they would have done otherwise; on the other, they benefit from lower prices along with everybody else. On balance, you would expect that some will lose, some will gain, and some will be about as well off as they were before. In developing countries, the labour-market side of this process will tend to work in the other direction. The increase in demand for poor-country labour ought to push up wages even for workers who are not employed in the new trade-related jobs.

So capitalism-globalisation is not mainly concerned with shifting income from workers to investors, as the sceptics maintain. Rather, it makes some workers worse off while making others (including the poorest ones of all, to begin with) better off. And in the aggregate it makes consumers (that is, people with or without a job) better off as well. Altogether, given freer trade, both rich-country and poor-country living standards rise. That gives governments more to spend on welfare, education and other public services.

Note that all this counts only the so-called static gains from trade: the effects of a once-and-for-all shift in the pattern of production and consumption. Modern economics also emphasises the importance of dynamic gains, arising especially from the economies of scale that freer trade makes possible. The aggregate long-term gain for rich and poor countries alike is likely to be far bigger than the simple arithmetic would suggest.

Moreover, few displaced rich-country workers are likely to be permanently out of work. Most will move to other jobs. Also, new jobs will be created by the economic opportunities that trade opens up. Overall, trade neither reduces the number of jobs in the economy nor increases them. In principle, there is no reason to expect employment or unemployment to be any higher or lower in an open economy than in a closed economy—or, for that matter, in a rich economy as compared to a poor economy. Still, none of this is to deny that the displaced rich-country workers lose out: many, perhaps most, of those who find alternative work will be paid less than they were before.

In thinking through the economic theory of liberal trade, it is helpful to draw a parallel with technological progress. Trade allows a country to shift its pattern of production in such a way that, after exporting those goods it does not want and importing those it does, it can consume more without there having been any increase in its available resources. Advancing technology allows a country to do something very similar: to make more with less. You can think of trade as a machine (with no running costs or depreciation): goods you can make cheaply go in at one end, and goods that would cost you a lot more to make come out at the other. The logic of protectionism would demand that such a miraculous machine be dismantled and the blueprint destroyed, in order to save jobs.

No question, technological progress, just like trade, creates losers as well as winners. The Industrial Revolution involved hugely painful economic and social dislocations—though nearly everybody would now agree that the gains in human welfare were worth the cost. Even in far milder periods of economic transformation, such as today's, new machines and new methods make old skills obsolete. The Luddites understood that, which made them more coherent on the subject than some of today's sceptics, who oppose integration but not technological progress. Logically, they should oppose both or neither.

Politically, of course, it is essential to keep the two separate. Sceptics can expect to win popular support for the view that freer trade is harmful, but could never hope to gain broad backing for the idea that, so far as possible, technological progress should be brought to a halt. Still, it might be better if the sceptics concentrated not on attacking trade as such, but on demanding help for the workers who suffer as a result of economic progress, whether the cause is trade or technology.

So much for the basic theory. What does the evidence say? For the moment, concentrate on the prospects for workers in rich countries such as the United States (the next section will look in more detail at workers in poor countries). By and large, the evidence agrees with the theory—though things, as always, get more complicated the closer you look.

A first qualification is that most outward foreign direct investment (FDI) from rich countries goes not to poor countries at all, but to other rich countries. In the late 1990s, roughly 80% of the stock of America's outward FDI was in Canada, Japan and Western Europe, and nearly all of the rest was in middle-income developing countries such as Brazil, Mexico, Indonesia and Thailand. The poorest developing countries accounted for 1% of America's outward FDI (see table 1). Capital is hardly flooding to the world's poorest countries—more's the pity, from their point of view.

The notion that outward FDI reduces the demand for labour in the sending country and increases it in the receiving one needs to be revised as well. It was based on the assumption that when rich-country firms invest in poor countries, rich-country exports (and jobs) are replaced by poor-country domestic production. In fact, evidence from the United States and other countries suggests that outward FDI does not displace exports, it creates them: FDI and exports are, in the jargon, net complements. This is because the affiliates of multinationals trade with each other. Figures for 1995 show that America's exports to its foreign-owned affiliates actually exceeded its imports from them (see table 2).

Before FDI, the companies exported finished goods. After FDI, they ship, let us suppose, a mixture of finished goods and intermediate goods. The intermediate goods will be used to make finished goods in the FDI-receiving country. The corresponding increase in exports of intermediate goods outweighs the fall, if any, in exports of finished goods. Overall, then, exports from the FDI-sending country rise. At the same time, the sending country's imports rise as well, partly because the affiliate sells goods back to the sending country. Exports rise, which increases the demand for labour; and imports rise, which decreases the demand for labour.

What does all this mean for the labour markets of the rich, FDI-sending countries? Jobs are created in exporting industries which will tend to be relatively high-paying, but overall employment will not rise, for reasons explained earlier. For every job created, another one somewhere else will be destroyed. The jobs that go will tend to be in industries that compete with imports. On average, studies suggest, those jobs pay lower wages.

On balance, then, you could say that the economy has gained: it now has more higher-paying jobs and fewer lower-paying jobs. A policy which attempted to resist a shift like that would be difficult to defend on its merits. Unfortunately, though, the people getting the higher-paying jobs are not necessarily the ones who have lost the lower-paying jobs. Because of the boost to exports, the overall effect of outward FDI on jobs and wages in the sending country is more benign than the simple theory suggests—but some people still lose.

Another implication of the shift in the demand for labour in the rich, FDI-sending countries is a possible widening of income inequality. In a country such as the United States, the combined action of trade and capital flows is likely to raise the demand for relatively skilled labour and lower the demand for relatively unskilled labour. Some hitherto low-wage workers may succeed in trading up to higher-paid jobs, but many others will be left behind in industries where wages are falling. In this scenario, high and average wages may be rising, but wages at the bottom may be falling—and that means greater inequality.

You would expect to see a similar pattern in an economy that was undergoing rapid technological change. So in the United States, which fits that description better than most in the 1990s, you could say that economic integration may have added to the already powerful pressures that were acting to increase inequality. Since those same pressures were raising living standards in the aggregate—not just for the very rich—it would be a misleading summary, but not a false one.

Of these two unequalising forces, economic integration and technological progress, which is likely to be more powerful? If it were the latter, that would raise doubts over the sceptics' focus on globalisation as the primary cause of social friction. The evidence suggests that technology is indeed much the more powerful driver of inequality. One study, by William Cline, estimated that technological change was perhaps five times more powerful in widening inequality in America between 1973 and 1993 than trade (including trade due to FDI), and that trade accounted for only around six percentage points of all the unequalising forces at work during that period. That is just one study, but it is not unrepresentative. The consensus is that integration has exerted a far milder influence on wage inequality than technology.

Mr Cline's study in fact deserves a closer look. It found to begin with that the total increase in the ratio of skilled to unskilled wages in the two decades to the early 1990s was 18%. This was the net result of opposing influences. An increase in the supply of skilled labour relative to the supply of unskilled labour acted to equalise wages, by making unskilled labour relatively scarce. By itself, this would have driven the wage ratio down by 40% (see table 3). But at the same time a variety of unequalising forces pushed the ratio up by 97%, resulting in the net increase of 18%. These unequalising forces included not just trade and technology, but also immigration, reductions in the real value of the minimum wage, and de-unionisation.

Two things strike you about the numbers. First, trade has been relatively unimportant in widening income inequality. Second, this effect is overwhelmed not just by technology but also by the main force operating in the opposite, equalising, direction: education and training.

This means that globalisation sceptics are missing the point if they are worried mainly about the effect of integration on rich-country losers: trade is a much smaller factor than technology. Some people in rich countries do lose out from the combination of trade and technology. The remedy lies with education and training, and with help in changing jobs. Spending in those areas, together perhaps with more generous and effective help for people forced to change jobs by economic growth, addresses the problem directly—and in a way that adds to society's economic resources rather than subtracting from them, as efforts to hold back either technological progress or trade would do.

Copyright © 2001 The Economist Newspaper and The Economist Group. All rights reserved.

Websites