See the complete article in the book Eco-Integration in Latin America.
An article by Inmaculada Martínez-Zarzoso in Eco-Integration in Latin America analyzes the environmental provisions (EPs) included in free trade agreements (FTAs), especially those signed by countries in the Americas over the last few decades.
From the mid-1990s and early 2000s, a growing number of FTAs and economic integration agreements (EIAs) have included environmental aspects in the agreement text, or have been accompanied by a separate environmental agreement, as was the case in the North American Free Trade Agreement (NAFTA). The main objective for including EPs tends to be preventing the elimination of barriers to trade and the resulting increase in trade from having a negative effect on environmental quality in signatory countries.
From the start of the 1990s, the interaction between global trade and environmental quality has been widely recognized within studies on international economics and international relations and has been taken into account in a broad sense during the negotiation of both FTAs and EIAs. As early as 1992, in the Rio+20 agreement, environmental protection was considered necessary for guaranteeing the sustainability of countries’ economic growth. Likewise, there have been regional environmental cooperation agreements between the US and Central America since the mid-1990s that are not necessarily linked to trade.
At the same time, the number of trade agreements that have entered into force in the last two decades has grown considerably, reaching a total of over 250 FTAs by 2016, a growing number of which include devices that extend into other areas, such as the environment and labor protection. Figure 1 shows the cumulative number of FTAs that have been signed since the mid-1990s, and the number of FTAs that include EPs or parallel environmental agreements. The cumulative figures for 2014 indicate that approximately 25% of agreements include EPs. Figure 1 also shows the number of agreements that include labor provisions, so as to compare the relative importance of environmental and labor-related content. Approximately 21% of FTAs touch on issues related to labor rights or social protection. The number of more inclusive agreements has grown, especially since 2005. Since 2007, the Organisation for Economic Co-operation and Development (OECD) has periodically reviewed how environmental issues have been handled in FTAs (OECD, 2007) and has compiled an inventory of FTAs with EPs. In the studies it is currently carrying out, the OECD is promoting econometric analysis to evaluate whether the inclusion of EPs in FTAs reduces emissions and improves environmental quality in signatory countries.
Figure 1. Cumulative Number of FTAs by Date of Entry into Force
It is worth noting that, to date, very few quantitative studies have attempted an ex-post assessment of the effectiveness of these EPs. In the academic sphere, there have been two empirical studies based on econometric models (Ghosh and Yamarik, 2006; Baghdadi, Martínez-Zarzoso, and Zitouna, 2013), but only the latter distinguishes between FTAs with and without EPs, while the former evaluates the general effects of FTAs on the environment and uses emissions data from 1990. Ghosh and Yamarik’s (2006) main findings show that signing FTAs reduces pollution, but this effect is indirect and derives from the positive effect of increased trade on per capita income, which in turn effects environmental quality. In contrast, they found no evidence of a direct effect between signing FTAs and emission reductions. Baghdadi et al. (2013) do distinguish between agreements with and without EPs between 1980 and 2008 and obtained evidence for the existence of a direct effect of signing FTAs on the reduction of CO2 emissions, but only for FTAs with EPs and not for those that do not include environmental aspects.
This study was based on a methodology that is widely accepted in economics to identify whether an effect is causal rather than merely indicating a positive correlation without pinpointing causality. It uses instrumental variables and a fixed effects estimator (panel data). At the initial stage, trade was estimated using a gravitational equation in which bilateral trade was explained using geographical determinants, and a growth model was used to estimate per capita income, in line with Baghdadi et al. (2013). During the second stage, the equation for determinants of emissions was estimated using panel data and fixed effects, and the predictions obtained during the first stage for trade and per capita income were incorporated. This combined use of instrumental techniques and variables made it easier to identify causality.
 An early version of this paper titled “Is the Road to Regional Integration Paved with Pollution Convergence?” was presented at the 10th Annual Conference of the Euro-Latin Study Network on Integration and Trade (ELSNIT), which was sponsored by the IDB. http://events.iadb.org/calendar/eventDetail.aspx?lang=es&id=3735. This study focuses exclusively on the effects of FTAs with EPs on the convergence of CO2 emissions (and not the effect on emissions levels by country) between 1980 and 2008 and does not consider other pollutants.
 The methodological use of these techniques is explained in detail in Baghdadi et al. (2013).