In addition to the increase in emissions from freight transport, trade affects climate change by changing production techniques and specialisations. The overall effect depends on comparative advantages, environmental policies and the nature of trade agreements.
The issue of the impact of international trade on the environment is not a new one; it has been present in the economic literature since the 1970s. The work multiplied in the 1990s, during the debates surrounding the North American Free Trade Agreement negotiations and the Uruguay Round of the GATT, and as the volumes traded increased (multiplied by 9 between 1980 and 2014, with the share of trade in world gross domestic product now reaching 30%).
Beyond the direct effect on emissions from freight transport, theoretical work has identified three main mechanisms through which trade can have an impact on climate change (see Copeland and Taylor, 2004 for a summary).
An increase in trade can induce an increase in production and thus, other things being equal, in greenhouse gas emissions. This is known as the "scale effect".
Trade liberalization shifts the production of goods and services according to the comparative advantages of the countries concerned. Thus, taking prices into account, production can be relocated to countries with high unit emissions (or vice versa). This effect is known as the "composition effect".
Finally, trade can have a "technical effect": it can make available, or reduce the cost of certain technologies and thus modify production methods, and thus emission intensities. The discussions on environmental goods currently underway at the World Trade Organization thus aim to reduce tariff barriers for the least polluting technologies.
Scale effects tend to increase emissions while technology effects tend to decrease them. Composition effects are more ambiguous. They depend on comparative advantages, the latter being influenced by the country's endowment of production factors but also by the environmental policies in place (a low-carbon taxing country will have an advantage in the production of high emitting goods). It is on the balance between these two elements that an important part of the debate has taken place: while the effect of environmental policies is preponderant, it is expected that, during trade liberalization, the production of polluting goods will be concentrated where environmental policies are less strict, especially in developing countries, constituting a "pollution haven". If, on the contrary, the factor endowment effect dominates, polluting industries are expected to concentrate in developed countries. Indeed, these industries are often capital-intensive and the most capital-intensive countries are the developed countries. These considerations are further complicated by the fact that environmental policies are not fixed but evolve with income. Thus, trade, by generating growth, tends to harden environmental policies. The location of polluting industries is important when considering local pollution. But when we look at climate change, it becomes less preponderant (leaving aside ethical issues): greenhouse gases have the same global impact wherever they are emitted. If a high-emitting industry moves from one country to another (which is a form of carbon leakage), it does not affect global climate change as long as its emission intensity remains the same. However, if this industry, by moving to a country where environmental policies are less strict than in its home country, increases its emission intensity, it increases global emissions all the more.
If there is one thing to be retained from the theoretical literature, it is that the mechanisms by which trade affects climate change are complex and multiple. They depend on the characteristics and policies of each country and have ambiguous effects.
The empirical literature quantifies the significance of the effects involved and determines the conditions under which trade has a positive or negative impact on greenhouse gas emissions. The first empirical studies concerned sulphur dioxide emissions, for which data were more readily available than for carbon dioxide emissions. However, by the early 2000s it was clear that the impacts of trade on emissions depended on the gas under consideration. Articles using estimation techniques that take into account the links between trade, growth and the environment as much as possible find a wide range of results. Frankel and Rose (2005) find no significant effect of trade on CO2 emissions, while Managi (2009) finds a differentiated impact between countries: when trade is liberalized, CO2 emissions decrease in developed countries (the dominant technical effect on scale and composition effects), while they increase in developing countries. Finally, Baghdadi et al (2013) show that differences in CO2 emissions between countries that are signatories to the same free trade agreement are reduced, with the country with the highest emissions coming closer to the lowest emissions when environmental clauses are included in the agreements. Trade does not only have an impact on production-related emissions, but also on those due to freight transport. Such transport can be by land, air or sea, the latter having the lowest emission intensity. Considering that all maritime transport, heavy road vehicles and two thirds of air transport are dedicated to goods, international trade would be responsible for 43% of transport sector emissions, or 6% of global emissions in 2010 (IEA, 2012). However, a strictly local supply does not guarantee lower emissions. Some countries have modes of production that are sufficiently low-emission-intensive for shifting production to them to lead to an overall reduction in emissions, despite the additional emissions from transport. This is the case, for example, of lettuce exported from Spain to the United Kingdom between November and December: its carbon balance (including transport) is better than that of lettuce produced in the United Kingdom (Edwards-Jones et al., 2008). In this case, the issue is the consumption of seasonal rather than local produce.
Finally, trade should not only be considered from the perspective of climate change mitigation, but also from that of adaptation to its effects. Climate change is likely to change the places where certain goods, particularly agricultural goods, are produced. Trade could enable the most affected countries to continue to obtain supplies despite the decline in their production.
Faced with these complex environmental effects, the outcome of trade agreements is uncertain, but it depends largely on the modalities chosen and the accompanying policies. If trade agreements are to help combat climate change, it is necessary to ensure that they effectively enable the wide dissemination of low-emissions technologies by facilitating trade and technical assistance in the sectors concerned. It is also necessary to limit the risks of "leakage" due to differences in regulations between partner countries, which justifies binding environmental clauses. Examples abound of such clauses, but their binding nature remains to be proven. The Trans-Pacific Partnership (TPP), signed on 5 November last, is a particularly interesting example in this area, since it contains environmental commitments whose non-compliance can be managed by the agreement's own dispute settlement mechanism. As the Americans have demonstrated their political will to make these commitments effectively binding, the reality of their effects (if the agreement is ratified) will be a test of the ability of modern trade agreements to play a constructive role in international cooperation on environmental and climate issues. While the commitments made in the SPR relate mainly to fisheries resources, this type of mechanism could indeed be equally applicable to climate change in other agreements under negotiation.