Agriculture, trade openness and emissions: an empirical analysis and policy options

Shuddhasattwa Rafiq, Ruhul Salim, Nicholas Apergis

    Research output: Contribution to journalArticlepeer-review

    59 Citations (Scopus)


    This article investigates the impact of sectoral production allocation, energy usage patterns and trade openness on pollutant emissions in a panel consisting of high-, medium- and low-income countries. Extended STIRPAT (Stochastic Impact by Regression on Population, Affluence and Technology) and EKC (Environmental Kuznets Curve) models are conducted to systematically identify these factors driving CO2 emissions in these countries during the period 1980–2010. To this end, the study employs three different heterogeneous, dynamic mean group-type linear panel models and one nonlinear panel data estimation procedure that allows for cross-sectional dependence. While affluence, nonrenewable energy consumption and energy intensity variables are found to drive pollutant emissions in linear models, population is also found to be a significant driver in the nonlinear model. Both service sector and agricultural value-added levels play a significant role in reducing pollution levels, whereas industrialisation increases pollution levels. Although the linear model fails to track any significant impact of trade openness, the nonlinear model finds trade liberalisation to significantly affect emission reduction levels. All of these results suggest that economic development, and especially industrialisation strategies and environmental policies, need to be coordinated to play a greater role in emission reduction due to trade liberalisation.
    Original languageEnglish
    Pages (from-to)348-365
    JournalAustralian Journal of Agricultural and Resource Economics
    Issue number3
    Early online date5 Nov 2015
    Publication statusPublished - Jul 2016


    Dive into the research topics of 'Agriculture, trade openness and emissions: an empirical analysis and policy options'. Together they form a unique fingerprint.

    Cite this