Observing the widening pay gap in the United States, economists grew worried because existing theories seemed to indicate that growing trade with poor countries might be at fault. An IMF report explained, “With exports from emerging markets and developing countries being intensive in labor, especially unskilled labor, traditional trade theory would predict that the integration of these countries into the world labor market would exert downward pressure on wages (corrected for productivity) of workers in advanced economies.”24 News coverage of companies closing factories in the United States and opening new ones in lower-wage countries further stoked fear and spread it among the general public. But while the media might have portrayed globalization as the main suspect, economists saw it as just one of many factors. The economy was shifting from manufacturing toward services, unions were declining, labor-saving technology was advancing rapidly, changes were taking place in unemployment benefits and in tax policies, and so on.
By the late 1990s, vigorous debate had produced something of a consensus among economists: technological change was the main cause of rising inequality, not globalization. Maybe globalization caused 10–20 percent of the increase in the wage gap—and some economists still maintained it didn't factor at all. As they saw it, trade with poorer countries was just too small in proportion to the rich economies to have such a big impact. Nonetheless, as trade with developing countries and inequality both grew in the 2000s, some economists reconsidered their earlier conclusions. Paul Krugman, Ben Bernanke, and Lawrence Summers all expressed the possibility that trade might play a bigger role than previously estimated.25
Robert Lawrence cast doubt on globalization's role after scrutinizing changes in the U.S. income distribution. The pay gap between skilled and unskilled workers had indeed risen in the 1980s. However, in the 1990s, he reported, “wages near the top of the income distribution (90th percentile) continued to rise more rapidly than wages at the median, but wages at the bottom of the distribution kept up or actually increased faster than wages in the middle. Since 2000, with the exception of the very top, wages have generally moved in tandem.”26
The United States became more integrated into international markets in the 1990s and 2000s, without widening the pay gap between skilled and unskilled workers. How can we square that with our (admittedly limited) theoretical understanding? Lawrence suggests three possible explanations for why the rest of the income distribution didn't become more unequal:
One is that the goods that the United States imports are actually very sophisticated and produced in the United States by relatively skilled workers. While it may cause displacement and could put downward pressure on wages generally, competition from low-wage countries does not increase wage inequality. A second more benign view is that a significant amount of what America imports today is no longer produced domestically. Thus, declining import prices simply yield consumer benefits and do not exert downward pressure on US wages or cause dislocation of US workers. A third view is that while US imports may be produced using labor-intensive methods abroad, when produced in the United States, capital- and skilled-labor-intensive methods are used.
Among the more intriguing observations that Lawrence makes while discussing these explanations is that “manufactured imports overall, even those from developing countries such as China and Mexico, are concentrated in US manufacturing sectors that pay significantly higher than average US wages. This means that import displacement does not fall disproportionately on less skilled workers.”27
Evidence indicates that, at least recently, trade hasn't increased inequality among the vast majority of American workers. However, as Lawrence suggested, American workers—skilled and unskilled—might all be losing out due to globalization. Is globalization letting more mobile capital claim an edge over less mobile labor? Here we can review evidence across advanced economies. The proportion of national income going to labor fell 3–4 percent in the United States and other Anglo-Saxon economies and about 10 percent in Europe and Japan from 1980 to 2006.28 But before we get too panicked about this, we should note that in the United States, Europe, and Japan, labor's share of the pie was in part just reversing earlier gains (see top panel of figure 9-4 for the U.S. data).
Figure 9-4: Labor share and compensation trends in developed countries
Sources: Top panel: Robert Z. Lawrence, Blue-Collar Blues: Is Trade to Blame for Rising US Income Inequality? Peterson Institute for International Economics, January 2008. Bottom panel: Florence Jaumotte and Irina Tytell, “The Globalization of Labor,” chapter 5 of IMF World Economic Outlook, April 2007.
What explains labor's declining share of national income across developed countries since the 1980s? Economists at the IMF did attempt to quantify the impacts of various factors: technology, cheaper traded goods, offshoring (trade in intermediate goods and services), immigration, and changes in labor market policies. Their results reaffirmed that technological change was again the most important factor reducing labor's share of national income. Falling prices of traded final goods, in contrast, appeared to increase the labor share across most developed countries, leading to the conclusion that “cheaper imports have increased the size of real total labor market compensation, implying that workers have participated in the benefits of the bigger economic ‘pie,’ although their share of it has declined.”29 Note that this is exactly what the bottom panel of figure 9-4 indicates.
The IMF study also estimated that immigration and offshoring reduced the labor share to varying degrees across countries, with immigration having a larger impact than offshoring.30 But immigration's estimated impact on natives' wages was very small and often temporary.31 And offshoring was estimated to contribute less than 0.1 percent annually to labor's declining share of income.32
Let's move to our third concern—does globalization make jobs less secure? In the United States, the Great Moderation of the last few decades did not translate into more stable incomes or enhanced job security. Household income volatility—particularly men's labor income volatility—skyrocketed. The standard deviation of changes in U.S. household income “increased by one-third between the early 1970s and early 2000s,” with most of the growth prior to the mid-1990s.33 Average job tenures and the incidence of long-term employment also declined.34 And worries about job security weren't exclusively an American phenomenon. Much of Europe experienced persistent high unemployment. Strong labor protections curtailed job losses but also stifled job creation.
Theoretically, there are a number of ways in which globalization could make labor markets more volatile. As markets open up and comparative advantage shifts among countries, workers in industries where imports are growing face a heightened risk of job loss, which could also lead to more frictional unemployment.35 Another proposed link between globalization and employment volatility is the concern that globalization could make it easier for employers to substitute foreign for domestic workers. In this case, workers might “have to incur greater instability in earnings and hours worked in response to shocks to labor demand or labor productivity.”36 However, econometric studies have cast doubt on the presumed increase in substitution of foreign workers for domestic ones.37
So much for theory—what about real jobs? Only 2–3 percent of U.S. job losses in the 1990s and early 2000s were related to competition from imports or firms relocating positions overseas.38 And that counts only losses. It's harder to track the jobs created because of trade, but research suggests that a given amount of demand growth from exports creates more (and better-paying) jobs than a similar increase in domestic demand; and more jobs than would be lost if imports grew by the same amount.39 Earlier studies mainly reflected the loss of manufacturing jobs, and today there's more fear about offshore services. But the phenomenon still is not big enough to affect national employment patterns.40 Furthermore, we can take comfort in knowing that rich countries more open to trade and FDI don't systematically suffer from higher unemployment. Some of the most open countries have among the highest employm
ent rates.41
While some workers in rich countries really do lose jobs to foreign competition and offshoring—and theories can explain how globalization could help lower wages—labor markets remain overwhelmingly domestic. We need to do better at assisting workers who get hurt by globalization, but if we really care about labor markets in advanced countries, we should expend most of our energy addressing factors with bigger effects than globalization.
Developing Concerns
Free trade enjoys stronger public support in poorer countries than in rich ones, yet that hasn't stopped intellectuals on the left from arguing that “neoliberal” free-trade agreements pursued during the last twenty or so years systematically exploit the poor.42 Why so much skepticism about globalization's benefits for poorer countries when the prevailing economic theories imply that integration promotes growth and convergence, which are obviously good for the poor? Much of the worry about globalization that exists in the developing world reflects the real exploitation that many countries suffered under colonial rule. Recall that economists believe all freely chosen transactions should be mutually beneficial. Under previous episodes of globalization, poorer countries engaged in a host of transactions that were far from freely chosen: exporting mineral resources, purchasing manufactured goods from colonial masters, and so on. With that kind of history, it's no surprise that foreign ownership of major industries might be met with some suspicion.
Let's look at what actually has happened under globalization in developing countries. Most importantly, poverty has declined markedly, as shown in figure 9-5 (top panel). The number of human beings living on less than $2 a day declined from 1.6 billion in 1970 to 850 million in 2006, even though the world population almost doubled over the same period.43 That 850 million people live on under $2 a day is a tragedy, but claims that the integration of the last few decades harmed the poor just don't fit the facts. Skeptics argue that graphics like figure 9-5 focus on gains in big countries like China and overlook smaller countries that have done poorly. Actually, many (though not all) smaller economies are also benefiting. Think of the growth of commodity producers, particularly in Africa, fueled by China's rise.
Rising inequality within many developing countries is also a concern. This trend might seem surprising based on trade theory, but scholars have tied it to shifts toward more technologically advanced exports along with domestic developments like urbanization.44 Back in 1955, Simon Kuznets famously linked early stages of economic development to rising inequality,45 though subsequent research has cast doubt on that relationship.46
Consider China's changing income distribution (bottom of figure 9-5). Before what Chinese leaders call the “period of reform and opening up” began in 1978, China had very little inequality and a lot of poverty (the vertical lines on the graph show the poverty line at only $1 per day). But as China's leader at the time, Deng Xiaoping, explained, “Some people will become prosperous first, and then others will become prosperous later.”47 China is now much less equal than thirty years ago, but the country has nearly eradicated poverty at the $1 a day level. And recent data indicate that Chinese inequality may be starting to decline with the government's emphasis on building a “harmonious society.”48
Figure 9-5: World poverty rate at $1/2/3/5 per day versus exports as percent of GDP (top) and China's income distribution (1970–2006) (bottom)
Sources: Poverty rates and Chinese income distribution from Maxim Pinkovskiy and Xavier Sala-i-Martin, “Parametric Estimations of the World Distribution of Income,” NBER working paper 15433, October 2009; exports as percent of GDP from World Bank; World Development Indicators.
How inequality is generated also influences its effects on growth: the concentration of wealth among entrepreneurs is very different from its concentration in entrenched oligarchies. Studies have shown that self-made billionaire wealth promotes faster economic growth, while high levels of heir-controlled wealth subtract as much as two percentage points a year. Entrenched wealth poses a particular challenge in some developing countries: billionaires' holdings average less than 3 percent of GDP in industrial countries but 13.3 percent in some East Asian countries.49 Globalization can help by broadening the playing field.
While labor market globalization generally looks attractive from the perspective of developing countries, it is worth remembering that poor people can also lose jobs due to foreign competition—with consequences that can, in their cases, literally include starvation. Consider Haiti's rice farmers. Three decades ago, Haiti exported rice. During the 1990s, the United States pushed Haiti to open its rice market. Now 80 percent of rice eaten in Haiti is imported. Six pounds of U.S. rice cost only $3.80, versus $5.12 for Haitian rice, a big benefit for Haiti's poor.50 But rice farming employed some 20 percent of the population in cultivation and many more in related sectors; cheaper American rice hardly makes up for their losses.51 In March 2010, former U.S. President Bill Clinton apologized, saying, “It was a mistake that I was a party to … I have to live every day with the consequences of the lost capacity to produce a rice crop in Haiti to feed those people.”52 Despite globalization's big benefits for the poor, we need to think twice before doing anything to jeopardize the livelihoods of those who are already so vulnerable.
In summary, globalization does seem to benefit most categories of workers in developing countries. And the demographic pattern we looked at in the previous chapter, rapid growth of the working age populations in most developing countries, implies that such countries should keep pushing for more openness to fuel more job creation. Demographic windows of opportunity don't automatically translate into growth. Smart policies and effective entrepreneurship are required to convert demographic potential into real prosperity.
Toward a Fairer World
As we have seen, globalization isn't fundamentally exploitative, but some people do lose out. The strength of the antiglobalization movement—however misguided its ideas—indicates that prevailing approaches to making trade fair for all have failed to pass muster with broad segments of the public. Part of the problem is rhetorical. World 2.0 believers get so enamored with integration that they fail to acknowledge any negative consequences, and their lack of sensitivity fuels the backlash.
Public demands for government to stop the “exporting of jobs” lead predictably to World 1.0 proposals to restrict trade. It's as if policy makers are using the simplistic bathtub model and think everything will turn out fine if you plug up the leaks and keep the “domestic jobs” at home. The trouble is that these policy makers seldom seem to pay attention to the actual costs. In 2009, the United States imposed a special tariff to protect the U.S. tire industry from a surge of Chinese imports. According to a very rough estimate by economist Brad DeLong, the tariffs will cost Americans $140,000 per job saved per year, and that doesn't include the cost of retaliation by China and dented perceptions of the U.S. business climate.53 Earlier and more systematic studies indicate it costs $170,000 to save an American job via trade protection,54 and even more in some industries like shipping ($415,000) and sugar ($600,000).55 Obviously the costs exceed the relevant workers' salaries.
Furthermore, if we're concerned about inequality, we should think twice about raising tariffs, since these tend to serve as a particularly regressive form of tax. In the United States, staple consumer products—especially low-priced ones—draw among the country's highest tariffs,56 putting the cost of protectionism squarely on lower-income Americans, as well as on people in poorer countries, who produce most of these goods.57 Tariffs on raw material inputs and intermediate goods can be even worse. The high U.S. sugar tariffs discussed in chapter 3 drove the production of Life Savers candy to Canada, costing the United States six hundred jobs.58
Another problematic proposal inspired by World 1.0 involves curbing the international expansion of domestic firms to make it harder for them to shift work abroad. Bad choice. German data indicate that domestic firms that expand overseas actually have a better track record of retaining home c
ountry jobs than those that stay at home.59 Protecting jobs is thus a losing proposition. Downstream producers and consumers both lose. More efficient producers overseas lose. The costs far exceed the benefits.
My mantra for creating a fairer and more prosperous World 3.0 is to promote productivity and protect people, rather than protecting jobs. Protecting jobs is outrageously expensive. And think about the jobs your grandparents had. Do we really want to lock ourselves into the past? Given the push for prosperity, we should seize on gains from trade and other opportunities to grow the pie. But we shouldn't let ourselves forget real people's yearning for job security and the pain caused by job losses. For most of us, a job is a lot more than a source of income.60 A recent study in Sweden showed that “overall mortality risk for men was increased by 44 percent during the first four years following job loss.”61
So how should we go about promoting productivity while protecting people? On the protection of people, the ideas are familiar ones. Start with investment in education to make workers more flexible. And cushion the blow when job losses come: many developed countries have trade adjustment assistance programs, but most seem inadequate. And in terms of innovation, experiments with “wage insurance” show some promise.62 Such programs supplement workers' income when they move into new jobs that pay lower wages, reversing the incentive to wait longer for a higher-paying job, which can be associated with long-term unemployment.
World 3.0 Page 21