World 3.0

by Pankaj Ghemawat

  The Direct Effects of Globalization

  Globalization's most obvious direct environmental impact involves pollution caused by transporting goods and people between countries. Transportation accounted for 23 percent of energy-related CO2 emissions in 2004,1 or more broadly, 13 percent of all human-caused greenhouse gas emissions in the same year.2 Yet most of those emissions reflected domestic transportation. A full 74 percent of transport-related emissions were caused by ground transportation,3 mainly from trips within the borders of a given country. International aviation and marine transportation accounted for 6 and 9 percent of transport-related CO2 emissions respectively in 2007.4 Multiplying all this out, international air and sea transport emissions, those most directly traceable to globalization, were 1.4 and 2.1 percent of total energy-related CO2 emissions respectively (and an even smaller proportion of human-caused greenhouse gas emissions).

  Given all the concern about the environmental impact of moving products and people over long distances, let's dig a little bit deeper into international shipping and air transport, starting with international shipping of products, which embodies the most direct environmental impact of merchandise trade. According to the WTO, “90 percent of global merchandise trade by volume is transported by sea, and the bulk of CO2 emissions in the transport sector comes from road transport, [so] international trade does not seem to play a major role in the generation of emissions from the transport sector.”5 Other sources put maritime transport's impact at a level similar to the one cited above. The International Maritime Organization estimates that international shipping produced 2.7 percent of world CO2 emissions in 2007 (and maritime shipping 3.3 percent overall).6 On a per-ton-kilometer basis, a cargo ship emits only 15–21 grams of CO2, compared to a heavy truck and trailer's 50 grams and a jumbo jet's 550 grams on a 1,200-kilometer flight.7 Environmental damage therefore varies greatly based on mode of transportation, not just the distance over which goods are transported. Something carried a very long distance across the ocean may actually cause less harm than something transported a shorter distance over land.

  Given marine transport's importance to globalization, it's also worth noting other environmental harms stemming from shipping cargo around the world. One particular focus in the shipping industry has been oil spills, whose incidence has come down significantly since the 1970s. Regulations have also been put in place to limit releases of other pollutants and to address the unintended transport of marine organisms into nonnative environments in ships' ballast water.8 Thus, while marine transport does create environmental harms, in relative terms this is a very energy efficient and clean mode of transport.

  Aviation, both passenger and cargo, is the other form of transportation particularly associated with long-distance international flows. According to a group representing the air transport industry, “Aviation's overall contribution to human climate change is currently thought to be about 3.5 percent. This share can be expected to grow to 5 percent in 2050 due to the high level of demand for aviation services.”9 The same source indicates that 80 percent of aviation greenhouse gas emissions come from flights of 1,500 kilometers or longer, so we can assume that a large portion of aviation emissions are due to international flights.10 Proportionally, though, long-distance flights cause less noise and air pollution near airports than short flights.

  In sum, the two modes of transport that are most critical to cross-border flows of products and people do indeed contribute to global warming and other environmental harms, but in a modest way compared to domestic causes. You could have predicted this based on chapter 2. Since most kinds of human activity are mostly domestic, it shouldn't surprise us that domestic transportation is a much bigger polluter than international transportation. Of course, that isn't to say that that should let shipping and airline industries off the hook internationally or domestically for the damage that they do.

  Another environmental harm that arises directly from globalization is the introduction, intentional or otherwise, of alien species (and diseases) into local environments, which in some cases wreaks havoc on local ecosystems. Examples of the former include the introduction of rabbits into Australia and kudzu and Asian carp into the United States, and of the latter, the Mediterranean fruit fly (in infested imported fruit) and the Asian longhorn beetle (in wooden packing material). A review of this literature cites a report from the Office of Technology Assessment (OTA) estimating the direct monetary costs of such phenomena in the United States at $5 billion annually.11

  We could go farther and analyze the spread of infectious disease—most often, through international travel—in similar terms. This does, however, raise a practical issue. Even the sub–1 percent globalization of the Middle Ages was enough for the Black Death, transmitted by rats from Asia, to ravage Europe—and reshuffle the ranks of trading cities. Unless what is being contemplated is an extreme World 0.0/1.0 clampdown on cross-border interactions, some of the risks of global pandemics are going to be with us no matter how we choose among the globalization-related choices that we face. So while pandemics and other risks do need to be managed—the next chapter discusses the principles of risk management in World 3.0—they are, to a first approximation, neutral with respect to the choices that we must make.

  The Indirect Effects of Globalization

  We saw in the preceding section that cross-border flows' direct environmental effects are nonnegligible but much smaller than other sources of environmental harm, like domestic land transport, domestic industry, and deforestation. But does globalization cause more problems indirectly, through changes it induces in the broader volume and pattern of economic output? Environmental economists have distinguished among three types of indirect environmental effects. First, there are scale effects: globalization expands output and therefore pollution. More broadly, while economic growth is good, it does have this environmental downside that must be accounted for. Second, there are technique effects related to the degree of cleanliness or dirtiness of the production and disposal techniques used. Despite fears of a “race to the bottom,” globalization is generally viewed as having a positive influence on the environment through this channel by promoting the adoption of better techniques. And third, there are composition effects that derive from shifts in the mix of output across countries. These can have either positive or negative effects, and one particular concern is that dirty industries might migrate to poor countries with laxer environmental standards or enforcement, leading to the creation of “pollution havens” even if aggregate world pollution does not increase.12

  All three effects seem to be real, so globalization's indirect environmental impact depends on their relative strengths. If the negative effects exceed the positive ones on balance, we can expect globalization to lead to economic growth but a dirtier environment. If this were the case, we would have to address some really hard trade-offs and distributional questions, particularly in less developed countries. We certainly don't want our rallying cry to be, “protecting the planet through poverty.” On the other hand, if the positive effects are greater, we arrive at the surprising but hopeful conclusion that globalization and economic development can help make the environment cleaner.

  Research suggests that it is impossible to conclude that globalization (or economic growth) is always bad (or good) for the environment; there are simply too many different types of pollutants and country situations. For sulfur dioxide, to take one example, the evidence generally indicates a positive effect. Multiple studies have linked openness to international trade and lower sulfur dioxide concentrations, with one finding that “a 1-percent increase in the scale of economic activity raises pollution concentrations by 0.25 to 0.5 percent for an average country in our sample, but the accompanying increase in income drives concentrations down by 1.25–1.5 percent via a technique effect.”13 While the evidence isn't quite as clear for nitrogen dioxide and particulate matter, it seems that openness is associated with lower concentrations of these pollutants as well.14


  The idea that trade may not increase many types of air pollution and may even help decrease some of them is consistent with broader research on the ecological impacts of economic growth. For most pollutants (particularly those with localized impacts), economic growth in relatively poor countries leads to more pollution: the scale effect dominates. But as countries get richer, more output leads to less pollution; governments and polluters come under pressure to improve the environment and they have the wherewithal to do so, so they clean things up (the technique effect). Also, dirty industries generally form a smaller proportion of economic output as countries shift from primary production to more sophisticated manufacturing and eventually to services (the composition effect). This pattern has been observed for a broad range of pollutants, including sulfur dioxide, particulates, nitrous oxide, automotive lead emissions, and arsenic.15

  Trade and foreign investment can also enhance positive technique and composition effects. To maintain consistency across their plants and avoid negative publicity, foreign companies bring in new technologies and sometimes apply higher environmental standards than local firms. A related positive effect, particularly in agriculture, can come from leveraging cross-country geographic differences pertaining to climate and natural resources. Sometimes it is beneficial to produce products where their environmental impact is lowest (e.g., agricultural products where the climate is most favorable and industrial products closest to required natural resource inputs) and then ship them across borders for additional processing or final sale.

  Still, there are exceptions to this pattern and to the generally good- to-neutral tone of the news about trade, the most important one being carbon dioxide. Most research indicates that CO2 emissions have not fallen as countries get richer. Instead they have kept on growing, although as I write this, there are some tentative indications that this might be starting to change, perhaps due to the Kyoto Protocol and continuing multilateral efforts to regulate these emissions. But trade still seems to exacerbate CO2.16 So clearly CO2 is more vexing than other pollutants. The next section looks at some of its distinctive properties, and the problem of climate change is considered in a bit more detail in the section after that.

  But let's first address two other, related concerns about globalization's general environmental impact, the potential for a “race to the bottom” in environmental standards and for “pollution havens” to develop in poor countries. The race to the bottom simply doesn't seem to have happened. We already saw some evidence of this in the positive relationship between openness to trade and some kinds of environmental quality, the opposite of what would be predicted if there were such a race. Furthermore, developed countries have consistently raised—not lowered—environmental standards over the past thirty to forty years, and developing countries have followed suit.

  Germany is a prime example. Nearly a third of its GDP comes from exports, so it's exactly the kind of country that would be expected, in a race to the bottom, to experience pressure to lower its environmental standards. Instead, the opposite happened. In the 1970s and 1980s, Germany was reportedly considered one of Europe's “environmental laggards,” according to Newsweek, and its “river Rhine was a stinking cesspool, poisoned by heavy industry.” But by 2008, “Germany may be the world's greenest country—and not just because salmon once again return to spawn in the Rhine.”17 Germany now is cleaner on a host of environmental metrics.

  Tighter German standards have also, in an instance of positive rather than negative externalities, spilled over to other countries, including developing ones. Thus, some Chinese companies have matched German requirements even for their own domestic products, standardizing output in support of exports to Europe. Transfer of clean production processes via foreign direct investment, exports of clean technology products, and environmental provisions in trade agreements have also supported environmental protection in developing countries. And governments and firms in developing as well as developed countries are excited about global markets for green products and the jobs they might generate.

  What about evidence for pollution havens in poor countries? If globalization isn't leading to more pollution overall, is it saddling the poor with more pollution while allowing rich countries to dispatch their polluting industries far away? Here, the evidence is mixed. A recent study found that in low-income countries, more trade is associated with higher per capita energy consumption—and more emissions—while the opposite holds true in high-income countries.18 This seems to fit with findings that imports into rich countries are more pollution-intensive than rich countries' exports.19 On the other hand, for SO2 emissions, “trade apparently leads to a reallocation of pollution from the poor country to the rich country, rather than the other way around.”20 It turns out that industries that cause a lot of SO2 emissions also tend to have high capital-to-labor ratios. Since capital costs are normally lower and labor costs higher in wealthier countries, this consideration can offset higher environmental compliance costs.

  Those, of course, are the most broad-brush of characterizations. More detailed studies of other aspects of the pollution haven hypothesis—for example, the links with foreign direct investment rather than trade—yield more equivocal results. Part of the difficulty of making sense of the results seems to me to be intrinsic to the pollution haven hypothesis, in terms of what it assumes and what it leaves out. It assumes that all redeployment of dirty industries from countries with tough environmental standards to countries with looser ones is potentially problematic. But surely, some responsiveness to differences in standards is to be expected: the question is whether it reaches problematic levels. And it glosses over the fact that environmental impact depends on the standards actually applied to the production transferred and how they develop over time. While India has severe environmental problems, when a company with high standards such as Bharat Forge shuts an antiquated foundry in Europe and transfers production to India, I'm not very concerned about the environmental impact. And while China became a haven for dirty rare earth extraction in the 1980s and 1990s, by the late 2000s, China began sending police to shut down illegal mines and blowtorch their equipment,21 indicating that such problems can also be self-correcting.

  More broadly, a focus on pollution havens implies looking at pollution based on its putative source. But since there are many sources of pollution and globalization's contribution to the total is generally limited, a focus on whether globalization has or hasn't led to the emergence of pollution havens may miss out on the very real problems of areas that are—for whatever reason—pollution hotspots. So for purposes of thinking about how to regulate environmental problems, we shift to looking at overall patterns of pollution and what to do about them instead of trying to refine estimates of how much pollution to attribute to globalization.

  Pollution and Distance

  To figure out what to do about environmental externalities, it is useful to begin by looking at their actual incidence, that is, their actual geographic scope. Focusing on pollution as a singularly important case is helpful because most forms of pollution are subject to the law of distance: the greater the geographic distance between polluter and pollutee, the lower the pollutant concentrations experienced by the pollutee.

  A general sense of the distance sensitivity of pollutants is conveyed by maps such as the one in figure 6-1, which depicts air quality in the United States at the county level. As you can see from the map, air quality exhibits significant regional variation, broadly reflecting patterns of human habitation and industrial production. And variation is also apparent at smaller scales. At the metropolitan level, maps of the intensity of many air pollutants are mostly tracings of a city's major traffic arteries.

  As subnational differences in air quality imply, environmental harms still have a significant local dimension. Thus, Jeffrey Frankel concludes in his excellent review of the literature that for “most kinds of air and water pollution, the latter a particularly great health hazard in the third world … most of t
he damage is felt within the country in question.”22 Nobel Prize winner Joseph Stiglitz takes a similar perspective, noting that “the quality of ground water, lakes, or air usually affects only those nearby.”23

  Those average effects do, however, mask significant variation across pollutants, especially air pollutants. Figure 6-2 summarizes the “effective distances” that various air pollutants travel and indicates a significant difference between short-lived air pollutants with very localized effects, moderately long-lived pollutants with national/regional effects, and really long-lived pollutants with potentially global effects. These distance effects are measured a bit differently than the ones discussed previously, but for SO2 (sulfur dioxide), to take one example, the distance sensitivity, calculated as in chapter 3, is about -1, implying a regional footprint, as corroborated by other sources. For some of the really short-lived air pollutants in the figure, the distance effects might be as much as -3 or -4. CO2 (carbon dioxide), in contrast, is literally off the charts: its distance sensitivity is close to 0 in the sense that a ton of CO2 will make the same contribution to global warming irrespective of where it is emitted.

  As in the previous chapter, the concept of distance sensitivity helps scope the problem. For highly distance-sensitive pollutants, pollution is likely to happen within a single country (or province or city). In such situations, people expect the relevant governmental authority to take care of the matter. It might take some pressure from environmental activists, but in countries with democratic governance and sufficient resources to tackle the problem, such matters tend to get addressed.24 Special safeguards may be required, however, to prevent countries without much money or representative democracy from turning themselves into trash dumps to the world. And if such pollution happens close to national borders, cross-border agreements may also be necessary, as in the acrimonious case of the paper mill in Fray Bentos, Uruguay, just across the border from Argentina. But apart from such complexities, these are the simplest kinds of externalities to deal with given their limited geographic scope.