Comprehensive Economic and Trade Agreement (CETA) and air pollution

ABSTRACT

The study empirically investigates and shows that on average, the implementation of the Comprehensive Economic and Trade Agreement (CETA) may contribute in the fight against global warming. This study finds that on average, a 1 percent increase of a percentage point in the bilateral volume of trade as a portion of GDP between Canada and a typical EU member could help reduce annual per capita emissions of GHGs in an average CETA member by about .57%. The results also show that the presence of CETA may decrease annual per capita emissions of GHGs in almost all CETA members. There is no statistically significant evidence suggesting an increase of GHGs per capita emissions in any CETA member, regardless of the model or statistical method employed in the paper. These results stand because of the combinations of the factor endowment hypothesis (FEH), the pollution haven hypothesis based on population density variations (PHH2) and the pollution haven hypothesis based on national income differences (PHH1) between each EU member and Canada.

KEYWORDS:

• Free trade
• environmental economics
• CETA

JEL CLASSIFICATION:

• F18
• F53
• F64

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 During the first year of CETA's life (September 2017–October 2108), there has been a 7.7% increase on bilateral trade volume between the EU and Canada relative to the pre-CETA period (there was a 3.9% increase on Canadian exports into the EU and a 10.8% increase of the EU exports into Canada). During this year, the total bilateral volume of trade between the trade partners reached $155.2 billion Canadian Dollars. For more details and news on topics related to CETA see its EU official website at http://ec.europa.eu/trade/policy/in-focus/ceta/ and its Canadian counterpart at https://www.international.gc.ca/gac-amc/campaign-campagne/ceta-aecg/year_one-premiere_annee.aspx?lang=eng.

2 See Copeland and Taylor (1994, 2013), and Antweiler, Copeland, and Taylor (2001) for theoretical explorations over the existence of FEH and PHH1. See also Grossman and Krueger (1993), Antweiler, Copeland, and Taylor (2001), Cole and Elliott (2003) and Davis and Caldeira (2010) for empirical evidence consistent to FEH. See Scott (2005), Levinson and Taylor (2008), Cole and Fredriksson (2009), and López, Arce, and Zafrilla (2013) for empirical evidence in support of PHH1. In regards to PHH2, Frankel and Rose (2005) was the first study to investigate the existence of a pollution haven motive based on national population density variations when evaluating the effects of trade openness on air pollution. However, they find no empirical evidence in support of PHH2.

3 For theoretical models on trade and environment, see Markusen, Morey, and Olewiler (1993), Copeland (1994); Copeland and Taylor (2005), Copeland and Taylor (1995), Antweiler, Copeland, and Taylor (2001), Benarroch and Weder (2006), and Shapiro and Walker (2018) among others. See also Copeland and Taylor (2004), Copeland (2011), and Cherniwchan, Copeland, and Taylor (2017) for a comprehensive review of theory and evidence of the role of trade on pollution. See also Copeland and Taylor (2017) for a review of trade and environment literature in a Canadian retrospective.

4 See also Qirjo and Pascalau (2019a) and Qirjo, Pascalau, and Christopherson (2020), where they analyse the impact of TTIP on other eight pollutants.

5 This is an interesting result because it contradicts other recent empirical findings that suggest that trade openness has led to lower air pollution in the USA. See for example Cherniwchan (2017) that provide empirical evidence of this argument in the case of NAFTA. Levinson (2009, 2015) also empirically validate the latter claim but for trade intensity between the USA and the rest of the world.

6 Since our data of the GHGs are in per capita emission levels, we cannot separate the scale from the technique effects of growth. Therefore, the income per capita variable measures both the scale and technique effects. In the specialized literature, this is known as the scale-technique effect.

7 We analyse the impacts of CETA in each of the above main four groups of GHGs, respectively, in a follow-up project. We find that the implementation of CETA may help reduce per capita emissions of CO${}_2$, CH${}_4$, and N${}_2$O, respectively. For more details see, Qirjo, Pascalau, and Krichevkiy (2020).

8 Hence, in each EU member i, $T_i=\left(\frac{X_i+M_i}{GD{P}_i}\right)$, where $X_i$ and $M_i$ denote each EU member's exports and imports with Canada, respectively. For Canada, $X_{Canada.}=\sum _i{M}_i$ and $M_{Canada.}=\sum _i{X}_i$, respectively.

9 In the dataset, most of the volume of trade data are complete. However, there are missing observations for some of the Ex-Communist EU members in the early 90s (1990–1993) because these countries gained their independence during these years. For example, Croatia and Slovenia gained their independence from Yugoslavia in early 90, but we have complete data for these two countries from 1994–2016. Analogously, there are missing observations from 1990 to 1992 for the Baltic countries (Estonia, Latvia and Lithuania), but we have complete data for these Baltic EU members from 1993 to 2016. The latter countries were part of the Soviet Union and gained their Independence in early 90s. However, since the missing data are in the very beginning of the sample and the series exhibits a clear monotonic trend, we use a simple trend regression to fill in the missing observations. Furthermore, for the countries like the Czech Republic and Slovakia that did not become independent until 1993, we impute the data (1990–1993) by using the information for Czechoslovakia and using a ‘proportional’ approach based on a counterfactual analysis. The same strategy is used for Luxembourg during 1990–1996 time period, using the available data for Luxembourg and Belgium (reported as one country in this time period) and Belgium (that has also complete observations as a single country during this time period).

10 For the GDP variable, there are also some missing observations for the same years and countries as those of the volume of trade. In the case of Czech Republic, Slovakia and Luxembourg, we employ the same strategy as in the volume of trade. For the other countries (Croatia, Slovenia, Estonia, Latvia and Lithuania), we use a square polynomial trend (since GDP follows an exponential trend) in order to fill in the missing observations for 1990–1993 time period.

11 There are also some missing observations for the FDI variable. Here, we employ a simple trend or square polynomial trend regression to fill in the missing observations. In particular, we fill out the data for Belgium and Luxembourg during 1990–2000, Croatia during 1990–1993, and Estonia, Latvia and Lithuania during 1990–1992 time periods.

12 Note that Euro came to life as a common currency in 1999 and it was an official currency only for 11 EU members (Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Portugal and Spain). Greece joined in 2001, Slovenia in 2007, Cyprus and Malta in 2008, Slovakia in 2009, Estonia in 2011, Latvia in 2014, Lithuania in 2015. In 2016, there were 19 EU members that have adopted Euro as their official currency.

13 Therefore, there is no empirical evidence that suggest that poor EU members act as pollution havens due to more trade with Canada. Note that all Ex-Communist EU members are poorer than Canada, and there is no statistically significant evidence that suggest that they act as pollution havens. This could be an important result since it is contradictory to earlier empirical evidence that suggests that trade openness between Central and Eastern European countries and the rest of the world may force these countries to act as pollution havens (e.g. Kheder and Zugravu 2008).

14 We investigate further the empirical validity of the EKC in a follow up project, where we focus on four specific groups of GHGs. We confirm the results found here in regard to the inconsistency of the EKC even when we add the cubic per capita income and break down GHGs into 4 main GHGs. For more details, see Qirjo, Pascalau, and Krichevkiy (2020). See also Pascalau and Qirjo (2017b) and Qirjo and Pascalau (2020) for a similar analysis applied in the case of TTIP. They confirm that empirical validity of the EKC for CO${}_2$, CH${}_4$, HFC/PFC/SF${}_6$, but they find a monotonically positive relationship between income per capita and GHGs per capita. They claim that the latter result is associated mainly to the positive relationship between per capita income and N${}_2$O per capita.

15 See Porter (1991) on details on the latter hypothesis. See also Ambec et al. (2013) and Cohen and Tubb (2018) for recent theoretical and empirical evidence on evaluating the Porter Hypothesis.

16 The Driscoll–Kraay Fixed Effects approach controls both for cross-section dependence and serial correlation in standard errors up to the second lag.

17 We also use the Arrelano-Bond two-step difference GMM estimator and a system GMM specification. When using both instrumental variable approaches, we run into an instrumental proliferation issue, where we get a perfect Hansen statistics of 1. This is also the case when we instrument both trade and income with lags using the Arrelano-Bond one-step difference GMM estimator. Under all these scenarios, instruments either fit the endogenous variables or (and) they outnumber the individual countries.

18 The Driscoll–Kraay Fixed Effects regression should be theoretically superior to the rest, since it controls both for cross-section dependence and serial correlation in the standard errors.

19 Copeland (1996) is the first theoretical study to introduce a pollution content tax on imports. Abrego et al. (2001) derive implications on linking global trade and environmental policy. Carbone, Helm, and Rutherford (2009) use a game-theoretical structure to analyse potential benefits of global trade in emissions permits.

20 For more details, see http://ec.europa.eu/trade/policy/in-focus/ceta/ceta-chapter-by-chapter/. In addition to chapter 24 on Trade and Environment, see also chapter 22 entitled ‘Trade and Sustainable Development (TSD)’. Note that the largest international free trade agreement that has included a chapter of environment and trade in its negotiations is TTIP. For more details on this chapter, see http://trade.ec.europa.eu/doclib/docs/2016/august/tradoc_154837.pdf. However, despite the earlier progress on trade negotiations between the EU and the US, the trade talks have been put into a halt from August of 2016. On the other hand, Canada and the EU not only have reached an agreement, but also 98% of it is ratified by each member and their respective regional governments. On 13 September 2018, EU and Canadian committee members of TSD, have met to discuss progress on the procedure and institutional structures for the effective implementation of TSD chapters and exchange views on priority areas of trade and environment. For more details, see http://trade.ec.europa.eu/doclib/docs/2018/september/tradoc_157409.pdf.

21 For more information in regard to the increase of the volume of trade between each EU member and Canada, or and to see the top 5 imports and exports categories for each EU member and Canada under the September 2017–October 2018 time period, see https://www.international.gc.ca/gac-amc/campaign-campagne/ceta-aecg/year_one-premiere_annee.aspx?lang=eng

22 See for example a short report, http://trade.ec.europa.eu/doclib/docs/2016/july/tradoc_154775.pdf. See also various interviews from the EU and Canadian politicians emphasizing the positive economic impacts of CETA in both trade partners. They have summarized the positive effects of CETA on innovation, productivity, employment, quality of products, etc. There is not even an interview in support on the environmental gains of CETA from the economics point of view. For more information, see the official EU website on the CETA related issues, http://ec.europa.eu/trade/policy/in-focus/ceta/index_en.htm, or the official Canadian website on CETA, https://www.international.gc.ca/gac-amc/campaign-campagne/ceta-aecg/index.aspx?lang=eng. There are news on the possible implementation and inclusion of Paris Climate Change Agreement in CETA, but there are no reports or news in the economics behind the reduction of GHGs per capita that the presence of CETA may bring on average. See for example https://ec.europa.eu/clima/events/ceta-taking-action-trade-and-climate_en.