Evaluating the GDP–energy consumption nexus for the ASEAN‐5 countries using nonlinear ARDL model

• Date: 01 December 2017
• DOI: http://doi.org/10.1111/opec.12113
• Published date: 01 December 2017
• Author: Khalid M. Kisswani

Evaluating the GDP–energy consumption

nexus for the ASEAN-5 countries using

nonlinear ARDL model

Khalid M. Kisswani

Associate Professor, Department of Economics and Finance, Gulf University for Science and Technology,

P.O. Box 7207, Hawally32093, Kuwait. Email: kisswani.k@gust.edu.kw

Abstract

This study investigates the GDP–energy consumption nexus for five Asian countries (ASEAN-5)

during the 1971–2013 period by estimating the asymmetric long-run as well as short-run effects in

a cointegration framework. We apply the recently developed nonlinear autoregressive distributed

lags (NARDL) model of Shin et al. (2011) for individual country-by-country analysis. Meanwhile,

we employ the pooled mean group (PMG) estimator of Pesaran and Smith (1995) and Pesaran

et al. (1999) for the ASEAN-5 countries pooled as a panel set. The empirical results, in case of

country-by-country, show long-run asymmetry for Singapore and Thailand only, when considering

structural breaks. However, the panel analysis confirms asymmetry for the sample data. Finally, the

causality tests employed show mixed results for both set-ups: country-by-country and panel

sample.

1. Introduction

The 1970s witnessed two major energy shocks: an oil embargo and regime change in

Iran. These shocks hammered economic growth in many countries and generated

concerns in many developed countries regarding the supply of energy, triggering major

conservative energy policies. Furthermore, climate change and pollution worries in

recent decades enhanced the debate over the relationship between energy consumption

and economic growth. The pioneering work of Kraft and Kraft (1978) initiated a major

body of literature that has focused on the relationship between energy and output,

making it the focal point of much research for much of the past 30 years (Yu and

Choi, 1985; Yang, 2000; Soytas and Sari, 2003; Oh and Lee, 2004; Wolde -Rufael,

2004; Narayan and Popp, 2012; Apergis and Payne, 2012; among others). A survey of

the literature shows growing work has been devoted to examine the effect of energy

JEL classification: O13, O44, O53, Q43.

©2017 Organization of the Petroleum Exporting Countries. Published by John Wiley & Sons Ltd, 9600 Garsington

Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

318

consumption on economic growth along with the causal relationship. This was

performed by employing a range of different time series econometric techniques.

The renowned survey of Payne (2010) about the international evidence on the causal

relationship between energy consumption and growth reveals an abundant and growing

literature that employed various methodologies of time series analysis. Payne (2010)

implies that the empirical findings on the GDP–energy consumption nexus have

provided mixed outcomes, as the results show ambiguity. He summarises several reasons

for such ambiguity like differences in time horizons, variations in energy usage schemes,

heterogeneity in climate conditions and adopting different econometric approaches,

among many other reasons. Moreover, Payne (2010) affirms the fact that the bulk of the

work performed was focused on the developed and industrialised countries with less

attention paid to the rest of the world. Likewise, Ozturk (2010) in his literature survey

concludes that there is no consensus between energy consumption and economic growth

neither on the existence of the relationship nor on the direction of causality. He adds, to

the reasons of Payne (2010), the differences in countries’characteristics as part of not

finding the consensus.

Bozoklu and Yilanci (2013) summarise (using the work of Squalli, 2007) four

possible alternative hypotheses that organise the energy–growth nexus which give

justification for the energy conservation policies. First is the growth hypothesis which is

based on the production approach, wherein energy consumption acts as a supplement

factor for labour and capital in the production process. It assumes unidirectional

causality from energy consumption to economic growth. This means that energy

conservation policies will cause real GDP to decrease. Second is the conservation

hypothesis which is based on the demand-side approach. It assumes unidirectional

causality from economic growth to energy consumption and argues that energy

conservation policies do not affect real GDP. Third is the feedback hypothesis that

implies an interdependent causal association between real GDP and energy consumption

when there is a bidirectional causality, and energy conservation policies will not harm

real GDP. The final one is the neutrality hypothesis which suggests that conservative

energy policies do not decrease real GDP, as energy consumption has a little or no effect

on real GDP, and therefore, it suggests no causal relations between the two variables.

To this point and as indicated above, the literature suggests ambiguous conclusions

on the association between real GDP and energy consumption (for more details see:

Karanfil, 2009; Payne, 2010 for such discussion) despite the numerous studies that have

strived to examine the causality between energy and economic growth. In recent work,

Lee and Chang (2008) as well as Chontanawat et al. (2008) report long-run causal

relationship from energy consumption to GDP, whereas Lee et al. (2008) found a

bidirectional relationship. Narayan and Popp (2012) document significant evidence that

energy consumption negatively Granger causes real GDP for the G-6 panel of countries,

©2017 Organization of the Petroleum Exporting Countries OPEC Energy Review December 2017

Energy consumption & GDP by nonlinear ARDL model 319