Environmental Kuznets Curve: Conclusive Econometric Evidence for CO2

• Published date: 01 February 2014
• Date: 01 February 2014
• Author: Jie Li,Gregory C. Chow
• DOI: http://doi.org/10.1111/1468-0106.12048

ENVIRONMENTAL KUZNETS CURVE: CONCLUSIVE

ECONOMETRIC EVIDENCE FOR CO2

GREGORY C. CHOW*Princeton University

JIE LIThe World Bank

Abstract. This paper begins with a statistical formulation of the Environmental Kuznets Curve

(EKC) and discusses the major econometric problems raised in the literature in testing this hypoth-

esis using panel data. A simple t-test is then presented that avoids these econometric problems. The

result from applying the t-test conf‌irms conclusively the EKC for CO2as formulated.

1. INTRODUCTION

There has been a vast literature on the Environmental Kuznets Curve (EKC)

since the hypothesis was presented in Grossman and Krueger (1995). Much of it

raises skepticism of the econometric methods used to test the hypothesis of the

EKC, as contained in the survey articles of Stern (2004), Dinda (2004) and

especially Romero-Avila (2008). There is also much less consensus on the exist-

ence of the EKC for CO2than for the other traditional pollutants such as SO2

and NOx. The reason is that compared with those of other pollutants the

impacts of CO2are indirect, global and long-term. In this paper we f‌irst present

a statistical formulation of the EKC in Section 2. In Section 3 we discuss the

econometric problems in testing the stated hypothesis using panel data. In

Section 4, a simple t-test is provided to test the stated hypothesis that avoids all

the abovementioned statistical pitfalls. The t-test is applied to test the EKC in

Section 5 using data for 132 countries for the period 1992 to 2004. The test result

supports the EKC conclusively. Section 6 concludes. It is hoped that this paper

will dispel any doubt that may exist regarding the existence of the EKC for CO2

during the period observed.

2. STATISTICAL FORMULATION OF THE ENVIRONMENTAL KUZNETS CURVE

For the purpose of our discussion, an EKC is def‌ined as an empirical relation-

ship during the course of economic development where per capita CO2emissions

f‌irst increases with per capita real GDP and later decreases with per capita real

GDP. This def‌inition is accepted in much of the literature (e.g. Stern, 2004;

Dinda, 2004; Romero-Avila (2008), although disagreement on def‌inition exists.

We will discuss one simple formulation in this paper and point out that our

method of testing applies also to other formulations using different dependent

variables and explanatory variables.

*Address for Correspondence: 205 Fisher Hall, Princeton University, Princeton, New Jersey 08544

USA. E-mail: gchow@princeton.edu. The f‌irst author would like to acknowledge the support of the

Gregory C Chow Econometric Research Program of Princeton University in the preparation of this

paper.

bs_bs_banner

Pacif‌ic Economic Review, 19: 1 (2014) pp. 1–7

doi: 10.1111/1468-0106.12048

© 2014 Wiley Publishing Asia Pty Ltd