Trade liberalization, consumption shifting and pollution: Evidence from Mexico's used vehicle imports

• Author: Rui Wan,B. Cecilia Garcia‐Medina,Liang Chen
• Date: 01 November 2019
• DOI: http://doi.org/10.1111/roie.12436
• Published date: 01 November 2019

Rev Int Econ. 2019;27:1591–1608. wileyonlinelibrary.com/journal/roie

|

1591

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

The linkage between trade and the environment has been analyzed since the mid‐1970s and is mainly

focused on pollution generated from production, particularly in the manufacturing sector.1

However,

pollution generated from the consumption side has been less studied, even though it is a significant

contributor to global pollution. A good example of consumption‐generated pollution is vehicle driv-

ing, which accounts for a large fraction of air pollution emissions.2

Moreover, the global trade in

vehicles is large, a significant fraction of which involves used vehicles. For example, Japan exported

nearly 4million vehicles to the world in 2015, 32% of which were used vehicles, while the United

States exported 2.4million vehicles to other countries in 2015, 22% of which were used vehicles.3

Received: 3 October 2018

|

Revised: 4 July 2019

|

Accepted: 8 July 2019

DOI: 10.1111/roie.12436

ORIGINAL ARTICLE

Trade liberalization, consumption shifting and

pollution: Evidence from Mexico's used vehicle

imports

LiangChen1

|

B. CeciliaGarcia‐Medina2

|

RuiWan3

1Department of Economics, Economics and

Management School,Wuhan University,

Wuhan, China

2Dirección de Investigación

Económica,Banco de México, Ciudad de

México, Mexico

3Department of International Economics

& Trade, School of Economics,Nanjing

University, Nanjing, China

Correspondence

Rui Wan, Department of International

Economics & Trade, School of Economics,

Nanjing University, 22 Hankou Road,

210093, Nanjing, China.

Email: wanrui@nju.edu.cn

Funding information

Chen acknowledges support from the

Humanities Research Council of Wuhan

University under Grant No. 413000048.

Abstract

This paper develops a model of used vehicle trade between

the United States and Mexico, countries with different en-

vironmental regulations regarding vehicle emissions. We

demonstrate that the United States, with its strict environ-

mental regulations, has a motive to export used vehicles to

Mexico, which impacts air pollution emissions from vehi-

cle driving in Mexico. Using unique data on the import and

registration of vehicles in Mexico after NAFTA, we find

that Mexico's used vehicle imports reduce pollution emis-

sions generated from vehicle driving mainly because of the

technique effect, that is, model–age‐comparable vehicles

imported from the United States emit less than those origi-

nally operated in Mexico.

JEL CLASSIFICATION

F13; L62; Q56; Q51

1592

|

CHEN Et al.

Therefore, in this article, we are interested in examining the effects of trade in used vehicles on pollu-

tion emissions generated by vehicle driving.

First, we develop a theoretical model of trade between two countries, the North and the South,

differing entirely in their environmental regulations on emissions. Vehicles are assumed to exist for

three periods: new, used, and clunkers (i.e., an old car in poor repair). The North has a monopoly on

producing new vehicles and selling them in both countries. Used vehicles and clunkers can be sold

in perfect second‐hand markets without transaction costs. Consumers purchase at most one vehicle in

each period, and consumers' preferences for a vehicle's quality are uniformly distributed in both the

North and South to generate smooth aggregate demand. To simplify the dynamics, we solve the model

in stationary equilibrium in which the monopolist chooses a constant price path for the new vehicles

in each period.

Within this context, we examine the effects of trade in used vehicles by comparing the equilibrium

outcomes in three trade regimes: (a) no used vehicles are allowed to be traded (“Autarky” regime);

(b) only used vehicles are allowed be traded (“UC” regime); and (c) both used vehicles and clunkers

are allowed to be traded (“UCC” regime). We find that, in autarky the North, the country with stricter

emissions regulations, has a comparative advantage in used vehicles and clunkers, which creates a

motive for trading in used vehicles. As a result, the North will export used vehicles and/or clunkers to

the South in UC and/or UCC regimes when trade in used vehicles is open between the two countries,

which shifts some of the consumption of new vehicles to older used vehicles in the South. The con-

sumption shifting changes the size and composition of the new vehicles, used vehicles, and clunkers

on the road in the South. These results imply that trade in used vehicles affects pollution emissions

generated from vehicle driving in the South through three potential channels: scale effect, composition

effect, and technique effect.

We then focus our analysis on empirically quantifying the effects of trade in used vehicles on pol-

lution emissions. In particular, we focus on used vehicle trade between Mexico and the United States

after NAFTA, which is perhaps the best‐documented example of North–South trade in used vehicles.

Although NAFTA came into effect in 1994, Mexico was allowed to restrict its used vehicle imports

until January 1, 2019. Imports of used vehicles into Mexico were strictly prohibited before 2005. After

2005, there were restrictions on the age of used vehicles allowed for import into Mexico, and these

restrictions were gradually relaxed until 2019. Therefore, we collect a unique dataset on used vehicle

imports, new vehicle sales, vehicle registration by state and age in Mexico from 1995 to 2015 for

implementing our empirical analysis.

Using the data, we first run regressions to investigate the effects of used vehicle imports on

vehicle fleet size and composition in Mexico. Our estimation results demonstrate that when mov-

ing from the Autarky to the UC regime between 2005 and 2008 in Mexico (when only vehicles

between 10 and 15years old were allowed to be imported), imports of used vehicles decrease new

vehicle sales, while when moving from the UC regime to the UCC regime from 2009 to 2015

(when restrictions on the age of used vehicles allowed for import was gradually relaxed beyond

10 to 15years old), imports of used vehicles increase new vehicle sales. The regressions on the

quantities of registered vehicles in the full sample for vehicles at all age ranges and subsamples for

vehicles at different ages indicate that used vehicle imports in Mexico slightly increase the total

fleet size of registered vehicles. Meanwhile, the estimated effects of used vehicle imports differ

across subsamples for vehicles at different age ranges. These results suggest that used vehicle im-

ports in Mexico affect the fractions of registered vehicles belonging to different age ranges, which

suggests a composition effect.

Finally, we use the above estimates to analyze the effect of used vehicle imports on pollution

emissions generated by vehicle driving in Mexico. We focus our study on four major pollutants: CH4,