Elsevier

Energy Economics

Volume 33, Issue 5, September 2011, Pages 896-902
Energy Economics

Electricity demand elasticities and temperature: Evidence from panel smooth transition regression with instrumental variable approach

https://doi.org/10.1016/j.eneco.2011.05.009Get rights and content

Abstract

This study applies a non-linear model, i.e. the recently developed panel smooth transition regression (PSTR) model, and takes into account the potential endogeneity biases to investigate the demand function of electricity for 24 OECD countries from the period 1978–2004. Our empirical results demonstrate that there is a strongly non-linear link among electricity consumption, real income, electricity price, and temperature, a result that is new to the literature. As real income rises, electricity consumption rapidly increases first, and after the level of real income exceeds approximately US$2500, its increasing rate turns slow down. An increase in electricity price has a negative or no influence on electricity consumption. Evidence of a U-shaped relationship between electricity consumption and temperature is supported, and the threshold value of temperature is approximately 53 °F, which is endogenously determined. Furthermore, the estimated elasticities of time dynamic indicate that electricity demand is income inelastic, price inelastic, and temperature inelastic. As time goes on, the absolute elasticities of electricity demand gradually decrease with respect to real GDP and electricity price, whereas they gradually increase with respect to temperature, suggesting that the impact of temperature on electricity demand is becoming more important in recent years.

Highlights

► This study applies a non-linear method to examine the electricity demand model. ► Evidence of non-linear effects on the electricity demands. ► Evidence of a U-shaped relationship between electricity consumption and temperature. ► The impact of temperature on electricity demand is becoming increasingly important.

Introduction

Electricity plays an important role in economic development and technological progress in many countries. “No country in the world has succeeded in shaking loose from subsistence economy without access to the services that modern energy provides” (World Bank, 1996). For developed countries, Ferguson et al. (2000) indicate that there is a strongly positive correlation between wealth and energy (or electricity) consumption, and the correlation between electricity and wealth is stronger than the correlation between total energy and wealth. However, the process of electricity production and consumption may emit air pollution and greenhouse gasses. The long-run accumulated greenhouse gas emissions are an important factor for global warming, which accelerates unusual climate change in the world.

Many countries have paid attention to greenhouse gas emissions and problems of global warming. In 2005 the Kyoto Protocol was drawn up and co-signing countries agreed to reduce greenhouse gas emissions by 5.2% from the level in 1990. Since the greenhouse effect and the reduction of pollution emissions are global concerns, one needs to clarify the determinants of electricity demand, which include real income, own price, climate change, and so on. Accurately estimating and analyzing the determinants of electricity consumption can provide some information for governments to discuss and anticipate the supply and demand of electricity, and then provide the basis of setting up appropriate environmental policies, i.e. pollution and energy taxes. Thus, in the framework of global data, it is more important to investigate electricity demand.

The paper aims to make the following contributions to the electricity demand literature. First, we apply the panel smooth transition regression (PSTR) model of González et al. (2005) to investigate the relationship among electricity consumption, real income, electricity price, and temperature for 24 Organization for Economic Cooperation and Development (OECD) countries from the period 1978–2004. Which economic variables could possibly explain the transition from one regime to another? In order to find out the optimal threshold variable of the electricity demand model, this study carries out non-linear tests by way of the potential threshold variables (Fouquau et al., 2008, Huang et al., 2008), which are real GDP per capita (Model (1)), electricity price (Model (2)), and temperature (Model (3)).

Second, most studies in the literature focus on analyzing the demand elasticities of electricity with respect to electricity price and income, but they seldom consider the impact of climate change on electricity consumption. How significant is temperature for the rising electricity consumption? We enrich the existing literature by simultaneously examining the impacts of real income, electricity price and temperature on electricity consumption and take into account endogenous determination of the types of our PSTR models for electricity demand.

Third, based on the characteristics of the PSTR model, we can consider the elasticity of electricity demand changes with country and time to analyze the elasticities of heterogeneous countries and the potential impacts of structural breaks (parameter instability) on the electricity demand's elasticities in the panel framework. The structural breaks are a common problem in macroeconomic series when they are usually affected by exogenous shocks or regime changes in environmental or economic events, i.e. economic development, energy crisis, global warming, the Kyoto Protocol, renewable energy technology, and so on (Lee and Chang, 2007, Lee and Lee, 2009).

Fourth and finally, many existing studies have found unidirectional causality from electricity consumption to real income and/or from real income to electricity consumption (Jumbe, 2004, Lee and Lee, 2010, Mozumder and Marathe, 2007, Ouédraogo, 2010). On the other hand, greenhouse gas emissions may increase with electricity consumption and then lead to rising temperatures (global warming), whereas an increase in temperature also influences electricity consumption. Thus, the problem of potential endogeneity exists in the electricity demand model.1 To the best of our knowledge, none of the studies on electricity demand in the existing literature to date notice this problem. To consider the potential endogeneity biases, we apply the PSTR model with instrumental variables developed by Fouquau et al. (2008).

The remainder of this study is organized as follows: In the next section, we discuss the reasons why it is important to test for nonlinearity in the energy demand model. Section 3 introduces the PSTR model with instrumental variables and illustrates the variables' definitions and data sources. Section 4 describes the data specification. Section 5 provides the empirical results, and a conclusion is offered in Section 6.

Section snippets

Importance of the non-linear analysis of the energy demand model

Looking at history, the two energy crises were a clear sign of their very strong shocks to the world's energy markets, which undoubtedly impacted the economic activities of almost every country, forcing them into a recession and causing them to adopt severe energy-cutting measures. However, a change in an administration's energy policy results in a heavy impact on people's energy consumption habits, and this surely brings about structural change in the relationship between energy consumption

Methodology

Following González et al., 2005, Fouquau et al., 2008, the two-regime PSTR model with fixed effects is defined as follows3:LELEit=ai+b1LRYit+c1LELEPit+d1LTEMPit+b2LRYit+c2LELEPit+d2LT

Data

The panel dataset is yearly and covers the period from 1978 to 2004 for 24 OECD countries.5 The 24 OECD countries include Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, South Korea, Luxembourg, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States. Electricity consumption per capita and real GDP

Empirical results

Before estimating the PSTR model with instrumental variable approaches, this study utilizes the MW-ADF and MW-PP unit-root tests (Maddala and Wu, 1999) to examine whether all variables are stationary, i.e. integrated of order zero (I(0)).7 The results of Table A1 show that all variables are stationary. We next examine whether there is a

Concluding remarks and implications

Electricity plays an important role in economic development and technological progress in many countries. Accurately estimating the determinants of electricity consumption can provide some information to discuss and anticipate the supply and demand of electricity and to set up appropriate environmental policies. Thus, it is important to examine the determinants of electricity consumption in the framework of global data.

The empirical results of this study are summarized as follows. First, the

Acknowledgments

The authors are grateful to Professor Fouquau and Professor Hurlin for kindly making available the MATLAB computer codes used in this paper. We are also grateful to the National Science Council of Taiwan for financial support through grants NSC 99-2623-E-110-001-NU and NSC 98-2410-H-110-069-MY2.

References (44)

Cited by (82)

  • Residential and industrial electricity consumption in Taiwan: Weather or macroeconomic condition (or both)

    2022, Energy Strategy Reviews
    Citation Excerpt :

    This study also contributes to the literature at large on the links between electricity consumption and weather or\and macroeconomic conditions. This literature provides fairly clear evidence of a significant relationship between residential electricity consumption and temperature change (e.g. Refs. [27–30,32,38]. However, little is known about the links between weather or\and macroeconomic conditions and collective electricity consumption.

View all citing articles on Scopus
View full text