Grinding the poor
Sep 27th 2001
From The Economist print edition
Sceptics charge that globalisation especially hurts poor workers in the developing countries. It does not
FOR the most part, it seems, workers in rich countries have little to fear from globalisation, and a lot to gain. But is the same thing true for workers in poor countries? The answer is that they are even more likely than their rich-country counterparts to benefit, because they have less to lose and more to gain.
Orthodox economics takes an optimistic line on integration and the developing countries. Openness to foreign trade and investment should encourage capital to flow to poor economies. In the developing world, capital is scarce, so the returns on investment there should be higher than in the industrialised countries, where the best opportunities to make money by adding capital to labour have already been used up. If poor countries lower their barriers to trade and investment, the theory goes, rich foreigners will want to send over some of their capital.
If this inflow of resources arrives in the form of loans or portfolio investment, it will supplement domestic savings and loosen the financial constraint on additional investment by local companies. If it arrives in the form of new foreign-controlled operations, FDI, so much the better: this kind of capital brings technology and skills from abroad packaged along with it, with less financial risk as well. In either case, the addition to investment ought to push incomes up, partly by raising the demand for labour and partly by making labour more productive.
This is why workers in FDI-receiving countries should be in an even better position to profit from integration than workers in FDI-sending countries. Also, with or without inflows of foreign capital, the same static and dynamic gains from trade should apply in developing countries as in rich ones. This gains-from-trade logic often arouses suspicion, because the benefits seem to come from nowhere. Surely one side or the other must lose. Not so. The benefits that a rich country gets through trade do not come at the expense of its poor-country trading partners, or vice versa. Recall that according to the theory, trade is a positive-sum game. In all these transactions, both sides—exporters and importers, borrowers and lenders, shareholders and workers—can gain.
What, if anything, might spoil the simple theory and make things go awry? Plenty, say the sceptics.
First, they argue, telling developing countries to grow through trade, rather than through building industries to serve domestic markets, involves a fallacy of composition. If all poor countries tried to do this simultaneously, the price of their exports would be driven down on world markets. The success of the East Asian tigers, the argument continues, owed much to the fact that so many other developing countries chose to discourage trade rather than promote it. This theory of “export pessimism” was influential with many developing-country governments up until the 1980s, and seems to lie behind the thinking of many sceptics today.
A second objection to the openness-is-good orthodoxy concerns not trade but FDI. The standard thinking assumes that foreign capital pays for investment that makes economic sense—the kind that will foster development. Experience shows that this is often not so. For one reason or another, the inflow of capital may produce little or nothing of value, sometimes less than nothing. The money may be wasted or stolen. If it was borrowed, all there will be to show for it is an insupportable debt to foreigners. Far from merely failing to advance development, this kind of financial integration sets it back.
Third, the sceptics point out, workers in developing countries lack the rights, legal protections and union representation enjoyed by their counterparts in rich countries. This is why, in the eyes of the multinationals, hiring them makes such good sense. Lacking in bargaining power, workers do not benefit as they should from an increase in the demand for labour. Their wages do not go up. They may have no choice but to work in sweatshops, suffering unhealthy or dangerous conditions, excessive hours or even physical abuse. In the worst cases, children as well as adults are the victims.
All this seems very complicated. Can the doubters be answered simply by measuring the overall effect of openness on economic growth? Some economists think so, and have produced a variety of much-quoted econometric studies apparently confirming that trade promotes development. Studies by Jeffrey Sachs and Andrew Warner at Harvard, by David Dollar and Aart Kraay of the World Bank, and by Jeffrey Frankel of Harvard and David Romer of Berkeley, are among the most frequently cited. Studies such as these are enough to convince most economists that trade does indeed promote growth. But they cannot be said to settle the matter. If the application of econometrics to other big, complicated questions in economics is any guide, they probably never will: the precise economic linkages that underlie the correlations may always be too difficult to uncover.
This is why a good number of economists, including some of the most distinguished advocates of liberal trade, are unpersuaded by this kind of work. For every regression “proving” that trade promotes growth, it is too easy to tweak a choice of variable here and a period of analysis there to “prove” that it does not. Among the sceptics, Dani Rodrik has led the assault on the pro-trade regression studies. But economists such as Jagdish Bhagwati and T.N. Srinivasan, both celebrated advocates of trade liberalisation, are also pretty scathing about the regression evidence.
Look elsewhere, though, and there is no lack of additional evidence, albeit of a more variegated and less easily summarised sort, that trade promotes development. Of the three criticisms just stated of the orthodox preference for liberal trade, the first and most influential down the years has been the “export pessimism” argument—the idea that liberalising trade will be self-defeating if too many developing countries try to do it simultaneously. What does the evidence say about that?
It does not say that the claim is nonsense. History shows that the prediction of persistently falling export prices has proved correct for some commodity exporters: demand for some commodities has failed to keep pace with growth in global incomes. And nobody will ever know what would have happened over the past few decades if all the developing countries had promoted trade more vigorously, because they didn't. But there are good practical reasons to regard the pessimism argument, as applied to poor-country exports in general, as wrong.
The developing countries as a group may be enormous in terms of geography and population, but in economic terms they are small. Taken together, the exports of all the world's poor and middle-income countries (including comparative giants such as China, India, Brazil and Mexico, big oil exporters such as Saudi Arabia, and large-scale manufacturers such as South Korea, Taiwan and Malaysia) represent only about 5% of global output. This is an amount roughly equivalent to the GDP of Britain. Even if growth in the global demand for imports were somehow capped, a concerted export drive by those parts of the developing world not already engaged in the effort would put no great strain on the global trading system.
In any event, though, the demand for imports is not capped. In effect, export pessimism involves a fallacy of its own—a “lump-of-trade” fallacy, akin to the idea of a “lump of labour” (whereby a growing population is taken to imply an ever-rising rate of unemployment, there being only so many jobs to go round). The overall growth of trade, and the kinds of product that any particular country may buy or sell, are not pre-ordained. As Mr Bhagwati and Mr Srinivasan argued in a recent review of the connections between trade and development, forecasts of the poor countries' potential to expand their exports have usually been too low, partly because forecasters concentrate on existing exports and neglect new ones, some of which may be completely unforeseen. Unexpected shifts in the pattern of output have often proved very important.
Pessimists also make too little of the scope for intra-industry specialisation in trade, which gives developing countries a further set of new opportunities. The same goes for new trade among developing countries, as opposed to trade with the rich world. Often, as developing countries grow, they move away from labour-intensive manufactures to more sophisticated kinds of production: this makes room in the markets they previously served for goods from countries that are not yet so advanced. For example, in the 1970s, Japan withdrew from labour-intensive manufacturing, making way for exports from the East Asian tigers. In the 1980s and 1990s, the tigers did the same, as China began moving into those markets. And as developing countries grow by exporting, their own demand for imports rises.
It is one thing to argue that relying on trade is likely to be self-defeating, as the export pessimists claim; it is another to say that trade actually succeeds in promoting growth. The most persuasive evidence that it does lies in the contrasting experiences from the 1950s onwards of the East Asian tigers, on one side, and the countries that chose to discourage trade and pursue “import-substituting industrialisation” (ISI) on the other, such as India, much of Latin America and much of Africa.
Years ago, in an overlapping series of research projects, great effort went into examining the developing countries' experience with trade policy during the 1950s, 60s and early 70s. This period saw lasting surges of growth without precedent in history. At the outset, South Korea, for instance, was a poor country, with an income per head in 1955 of around $400 (in today's prices), and such poor economic prospects that American officials predicted abject and indefinite dependence on aid. Within a single generation it became a mighty exporter and world-ranking industrial power.
Examining the record up to the 1970s, and the experience of development elsewhere in East Asia and other poor regions of the world, economists at the OECD, the World Bank and America's National Bureau of Economic Research came to see the crucial importance of “outward orientation”—that is, of the link between trade and growth. The finding held across a range of countries, regardless of differences in particular policies, institutions and political conditions, all of which varied widely. An unusually impressive body of evidence and analysis discredited the ISI orthodoxy and replaced it with a new one, emphasising trade.
What was wrong with ISI, according to these researchers? In principle, nothing much; the problems arose over how it worked in practice. The whole idea of ISI was to drive a wedge between world prices and domestic prices, so as to create a bias in favour of producing for the home market and therefore a bias against producing for the export market. In principle, this bias could be modest and uniform; in practice, ISI often produced an anti-export bias both severe and wildly variable between industries. Managing the price-rigging apparatus proved too much for the governments that were attempting it: the policy produced inadvertently large and complex distortions in the pattern of production that often became self-perpetuating and even self-reinforcing. Once investment had been sunk in activities that were profitable only because of tariffs and quotas, any attempt to remove those restrictions was strongly resisted.
ISI also often had an even more pernicious consequence: corruption. The more protected the economy, the greater the gains to be had from illicit activity such as smuggling. The bigger the economic distortions, the bigger the incentive to bribe the government to tweak the rules and tilt the corresponding pattern of surpluses and shortages. Corruption and controls go hand in hand. ISI is not the only instance of this rule in the developing countries, but it has proved especially susceptible to shady practices.
Today, developing-country governments are constantly, and rightly, urged to battle corruption and establish the rule of law. This has become a cliché that all sides in the development debate can agree on. But defeating corruption in an economy with pervasive market-suppressing controls, where the rewards to illegality are so high, is extraordinarily hard. This is a connection that people who favour closed or restricted markets prefer to ignore. Limited government, to be sure, is not necessarily clean; but unlimited government, history suggests, never is.
On the whole, ISI failed; almost everywhere, trade has been good for growth. The trouble is, this verdict was handed down too long ago. Economists are notoriously ignorant of even recent economic history. The lessons about what world markets did for the tigers in the space of few decades, and the missed opportunities of, say, India (which was well placed to achieve as much), have already been forgotten by many. The East Asian financial crisis of 1997-98 also helped to erase whatever lessons had been learned. And yet the prosperity of East Asia today, crisis and continuing difficulties notwithstanding, bears no comparison with the economic position of India, or Pakistan, or any of the other countries that separated themselves for so much longer from the international economy.
By and large, though, the governments of many developing countries continue to be guided by the open-market orthodoxy that has prevailed since the 1980s. Many want to promote trade in particular and engagement with the world economy in general. Even some sceptics might agree that trade is good for growth—but they would add that growth is not necessarily good for poor workers. In fact, it is likely to be bad for the poor, they argue, if the growth in question has been promoted by trade or foreign capital.
Capital inflows, they say, make economies less stable, exposing workers to the risk of financial crisis and to the attentions of western banks and the International Monetary Fund. Also, they argue, growth that is driven by trade or by FDI gives western multinationals a leading role in third-world development. That is bad, because western multinationals are not interested in development at all, only in making bigger profits by ensuring that the poor stay poor. The proof of this, say sceptics, lies in the evidence that economic inequality increases even as developing countries (and rich countries, for that matter) increase their national income, and in the multinationals' direct or indirect use of third-world sweatshops. So if workers' welfare is your main concern, the fact that trade promotes growth, even if true, is beside the point.
Yet there is solid evidence that growth helps the poor. Developing countries that have achieved sustained and rapid growth, as in East Asia, have made remarkable progress in reducing poverty. And the countries where widespread poverty persists, or is worsening, are those where growth is weakest, notably in Africa. Although economic policy can make a big difference to the extent of poverty, in the long run growth is much more important.
It is sometimes claimed that growth is less effective in raising the incomes of the poor in developing countries than in rich countries. This is a fallacy. A recent study confirms that, in 80 countries across the world over the past 40 years, the incomes of the poor have risen one for one with overall growth (see chart 4).
If all this is true, why does global income inequality seem to be widening? First, the evidence is not at all clear-cut. Much depends on how you make your comparisons. An overall comparison of country aggregates—comparing rich countries with poor countries—is generally more encouraging than a comparison of the richest 10% of people in the world with the poorest 10%. In 1975, America's income per head was 19 times bigger than China's ($16,000 against $850); by 1995, the ratio had fallen to six ($23,000 against $3,700). On the other hand it is true that Africa's income per head is rising more slowly than America's: as a result, their income-gap ratio has increased, from 12 in 1975 to 19 in 1995. But it would be odd to blame globalisation for holding Africa back. Africa has been left out of the global economy, partly because its governments used to prefer it that way. China has embraced the global economy with a vengeance—and see how well it has done.
Statistical difficulties aside, suppose it were true that global inequality is increasing. Would that be a terrible indictment of globalisation, as sceptics seem to suppose? Perhaps not. It would be disturbing, and extremely surprising, if poor countries engaged in globalisation were failing to catch up—but they aren't, as China and many other avid globalisers show. It would also be disturbing if inequality across the world as a whole were rising because the incomes of the poorest were falling in absolute terms, rather than merely in relative terms—but this is extremely rare. Even in Africa, which is doing so badly in relative terms, incomes have been rising and broader measures of development have been getting better. It may be too little, but it is not nothing, merely because other countries have been doing better.
The sceptics are right to be disturbed by sweatshops, child labour, bonded labour and the other gross abuses that go on in many poor countries (and in the darkest corners of rich ones, too). But what makes people vulnerable to these practices is poverty. It is essential to ask if remedial measures proposed will reduce poverty: otherwise, in attacking the symptoms of the problem, you may be strengthening their underlying cause. It is one thing for the sceptics to insist, for instance, that child labour be prohibited; it is quite another to ensure that the children concerned go to school instead, rather than being driven to scrape a living in even crueller conditions.
The barriers to trade that many sceptics call for seem calculated to make these problems worse. Some sceptics want, in effect, to punish every export worker in India for the persistence of child labour in parts of the Indian economy. This seems morally indefensible as well as counter-productive in economic terms. The same goes for the campaign to hobble the multinationals. The more thoroughly these companies penetrate the markets of the third world, the faster they introduce their capital and working practices, the sooner poverty will retreat and the harder it will be for such abuses to persist.
This is not to deny that the multinationals are in it for the money—and will strive to hire labour as cheaply as they can. But this does not appear to be a problem for the workers who compete to take those jobs. People who go to work for a foreign-owned company do so because they prefer it to the alternative, whatever that may be. In their own judgment, the new jobs make them better off.
But suppose for the moment that the sceptics are right, and that these workers, notwithstanding their own preferences, are victims of exploitation. One possibility would be to encourage foreign firms to pay higher wages in the third world. Another course, favoured by many sceptics, is to discourage multinationals from operating in the third world at all. But if the aim is to help the developing-country workers, this second strategy is surely wrong. If multinationals stopped hiring in the third world, the workers concerned would, on their own estimation, become worse off.
Compared with demands that the multinationals stay out of the third world altogether, the idea of merely shaming them into paying their workers higher wages seems a model of logic and compassion. Still, even this apparently harmless plan needs to be handled cautiously.
The question is, how much more is enough? At one extreme, you could argue that if a multinational company hires workers in developing countries for less than it pays their rich-country counterparts, it is guilty of exploitation. But to insist on parity would be tantamount to putting a stop to direct investment in the third world. By and large, workers in developing countries are paid less than workers in rich countries because they are less productive: those workers are attractive to rich-country firms, despite their lower productivity, because they are cheap. If you were to eliminate that offsetting advantage, you would make them unemployable.
Of course you could argue that decency merely requires multinationals to pay wages that are “fair”, even if not on a par with wages in the industrial countries. Any mandatory increase in wages runs the risk of reducing the number of jobs created, but you could reply that the improvement in welfare for those who get the higher pay, so long as the mandated increase was moderate and feasible, would outweigh that drawback. Even then, however, two difficult questions would still need to be answered. What is a “fair” wage, and who is to decide?
A “fair” wage can be deduced, you might argue, from economic principles: if workers are paid a wage that is less than their marginal productivity, you could say they are being exploited. Some sceptics regard it as obvious that third-world workers are being paid less than this. Their reasoning is that such workers are about as productive as their rich-country counterparts, and yet are paid only a small fraction of what rich-country workers receive. Yet there is clear evidence that third-world workers are not as productive as rich-country workers. Often they are working with less advanced machinery; and their productivity also depends on the surrounding economic infrastructure. More tellingly, though, if poor-country workers were being paid less than their marginal productivity, firms could raise their profits by hiring more of them in order to increase output. Sceptics should not need reminding that companies always prefer more profit to less.
Productivity aside, should “good practice” require, at least, that multinationals pay their poor-country employees more than other local workers? Not necessarily. To hire the workers they need, they may not have to offer a premium over local wages if they can provide other advantages. In any case, lack of a premium need not imply that they are failing to raise living standards. By entering the local labour market and adding to the total demand for labour, the multinationals would most likely be raising wages for all workers, not just those they hire.
In fact, though, the evidence suggests that multinationals do pay a wage premium—a reflection, presumably, of efforts to recruit relatively skilled workers. Table 5 shows that the wages paid by foreign affiliates to poor-country workers are about double the local manufacturing wage; wages paid by affiliates to workers in middle-income countries are about 1.8 times the local manufacturing wage (both calculations exclude wages paid to the firms' expatriate employees). The numbers come from calculations by Edward Graham at the Institute for International Economics. Mr Graham cites other research which shows that wages in Mexico are highest near the border with the United States, where the operations of American-controlled firms are concentrated. Separate studies on Mexico, Venezuela, China and Indonesia have all found that foreign investors pay their local workers significantly better than other local employers.
Despite all this, you might still claim that the workers are not being paid a “fair” wage. But in the end, who is to make this judgment? The sceptics distrust governments, politicians, international bureaucrats and markets alike. So they end up appointing themselves as judges, overruling not just governments and markets but also the voluntary preferences of the workers most directly concerned. That seems a great deal to take on.
|OPINION | WORLD | BUSINESS | FINANCE & ECONOMICS | SCIENCE &
PEOPLE | BOOKS & ARTS | MARKETS & DATA | DIVERSIONS | PRINT EDITION
Copyright © The Economist Newspaper Limited 2005. All rights reserved.