Exporting and environmental performance: Where you export matters

Abstract

Empirical analyses that rely on micro-level panel data have found that exporters are generally less pollutant than non-exporters. While alternative explanations have been proposed, firm-level data has not been used to examine the role of destination markets behind the relationship between exports and pollution. In this paper, we argue that because consumers in high-income countries have higher valuations for clean environments than consumers in developing countries, exporters targeting high-income countries are more likely to improve their environmental outcomes than exporters targeting destinations where valuations for the environment are not high. Using a panel of firm-level data from Chile we find support to this hypothesis. A 10 percentage point increase in the share of exports to high-income countries is associated with a reduction in CO₂ pollution intensity of about 16%. The results have important implications for firms in developing countries aiming to target high-income markets.

Keywords:

• Exports
• environment
• climate change
• emissions

JEL Classifications:

• F14
• F18
• Q56

Disclosure statement

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

Notes

1 See, for example, Cherniwchan, Copeland, and Taylor (2017) for a review.

2 Cole (2006) argues that firms that compete in the global marketplace are likely to come under closer international monitoring from their buyers which may lead to better environmental outcomes. Distelhorst and Locke (2018) present insightful evidence that sellers that comply with labor and environmental standards export more than non-complying firms. Neither of these papers, however, examine whether the environmental outcomes of the exporters change depending on the destination market. Cui (2017) presents a theoretical trade model with heterogenous firms, technology adoption and environmental regulation. The analysis is focused on changes in environmental regulation at home, but do not model the role of environmental regulations in the destination countries.

3 Firm size, measured in terms of total employment, is defined by INE according to the following groups: small firms (less than 50 employees), medium size firms (between 50 and 250 employees), and large firms (more than 250 employees).

4 We use alternative lists of high-income countries, based on groupings made by the OECD and the World Bank. The results remain very similar regardless of the list that we use. Table A2 shows the list of high-income countries from the OECD. The results in the rest of the paper are based on this group.

5 In the empirical estimation, we control for sector differences across establishments at a more disaggregated level (ISIC 4-digit level).

6 We do the same for the instrument of this variable.

7 We prefer this strategy than performing two separate regressions for two sub-periods. The latter approach reduces the sample size for each sub-period, which tends to lead to less precise estimates.

8 To identify the PTAs with environmental provisions that Chile has signed, we use Martínez-Zarzoso (2018). Table A3 shows the list of FTAs.

9 We adjust the instrument accordingly to reflect the exchange rate with respect to these countries.

10 We adjust the instrument accordingly to reflect the exchange rate with respect to these countries.

11 We also run additional regressions (not shown) using the upper tercile of the EPI-2010 index to calculate our destination-based export variable and the results fall in line with those with the ESI.