Applying the dynamic DEA model to evaluate the energy efficiency of OECD countries and China
Introduction
The 1996 Intergovernmental Panel on Climate Change (IPCC) of the United Nations Climate Change estimated that in order to stabilize carbon dioxide concentrations by double the pre-industrial (550 ppm) levels, then the world’s carbon emissions must be reduced by half in the 21st century. Thus, in December 1997 in Kyoto, Japan the Third Assembly of States Parties signed the Kyoto Protocol to control the emissions of greenhouse gases by 38 countries and the European Union so as to reduce the greenhouse effect on the global environment. The Protocol is an international agreement linked to the United Nations Framework Convention on Climate Change (UNFCCC), which commits its Parties by setting internationally binding CO2 emission reduction targets. This has placed a heavier burden on developed nations under the principle of “common but differentiated responsibilities”. The detailed rules for the implementation of the Protocol were adopted at COP 7 in Marrakesh, Morocco in 2001 and are referred to as the “Marrakesh Accords”. Its first commitment period started in 2008 and ended in 2012.
The overall perception of environment quality has stimulated policy makers to take accurate measures of the environmental impact of their policy decisions. As environmental topics become more noticeable and regarded as the world’s affairs, many countries are being required to carefully study, survey, adjust, and publish appropriate information of their policies’ impact on a set of economic targets, ranging from national accounts to social targets. An elementary step is an assessment that internalizes negative externalities as the basic part of the production course. However, traditional measures of managing efficiency focus only on the production of desirable outputs, while failing to consider environmentally undesirable by-products of production.
For the last two decades, the DEA (data envelopment analysis) analytical framework has been widely applied in the environmental and energy fields (see Färe et al. [1], Pasurka [2], Wang [3], Zaim [4], Zhou et al. [5], [6], Zhou and Ang [7]). There are two ways to deal with undesirable outputs: 1) treat undesirable outputs as weakly disposable (WD) variables in their original forms; see, for example, Färe and Grosskopf [8], [9], [10]; 2) consider the undesirable outputs as strongly (freely) disposable (SD) in various translated forms, such as in the form of their reciprocals or in the form of their additive inverses; see Lovell [11],Athanassopoulos and Thanassoulis [12] and Seiford and Zhu [13]. However, the above research studies do not consider the factor of carry-over change over time.
The purpose of this research is to encompass the results in Seiford and Zhu [13] to promote an undesirable input and to consider the variable of carry-over in the dynamic models [14]. The method in the literature for calculating system efficiency generates over-estimated scores when the dynamic essence is ignored. This makes it necessary to carry out a dynamic analysis whenever data are available. The data can hence be decomposed into efficiency change components and at the same time one can solve the change over time.
This study selects OECD countries and China during the period 2000 to 2010 and uses the dynamic DEA model to calculate and analyze these 27 countries’ executive efficiency based on fossil-fuel CO2 emissions. We also study the effects of the undesirable output and carry-over variable in order to rank the OECD countries and China. Finally, we apply environmental externalities and the overall efficiency change over time by the dynamic DEA model.
Traditional DEA models are usually devised to measure the technical and total efficiencies of decision making units (DMUs) over a certain period of time in a static manner. However, biased measurements of inefficiency may exist if we adopt an assumption of static optimization, because quasi-fixed inputs during a long-run period might not be allocated efficiently or adjusted to the optimal levels (Nemoto and Goto [15]). Moreover, when we incorporate several periods with inter-temporal effects to investigate the overall efficiency, we must take into account the inter-relationship during consecutive periods in a dynamic manner (Kao [16]).
Sengupta [17] and Färe and Grosskopf [18] made an important contribution to the development of dynamic DEA. Sengupta presented a dynamic DEA model by introducing the adjustment cost approach to analyze the risk and output fluctuations on the dynamic production frontier when incorporating shadow values of quasi-fixed inputs and their optimal paths into an analytic linear programming problem [17]. Färe and Grosskopf formulated several kinds of inter-temporal variables into realistic multi-output production processes across periods [18]. Later studies on dynamic DEA followed up and were developed based upon their model (see Sengupta [19], Nemoto and Goto [15], Jaenicke [20], Sueyoshi and Sekitani [21], Emrouznejad and Thanassoulis [22], Silva and Stefanou [23], Ouellette and Yan [24], and Chen and van Dalen [25]). Tone and Tsutsui [14] incorporated carry-overs as the connection of two consecutive periods to construct a slacks-based model to measure overall efficiency and period efficiency. Jafarian-Moghaddam and Ghoseiri [26] developed a fuzzy dynamic multi-objective DEA model to assess the performance of railways. Soleimani-damaneh [27] provided a new technique to gain an algorithm with computational advantages, using dynamic DEA models for estimating returns to scale. Tihana [28] applied a dynamic slacks-based measure approach to evaluate the relative efficiency of stocks in each quarter, gathering quarterly data of the Zagreb Stock Exchange during the period April 2009 to June 2012. Sueyoshi et al. [29] dealt with dynamic DEA window analysis in a time-shift frontier to assess the environmental performance of U.S. coal-fired power plants during 1995–2007 and concluded that it is necessary for the United States to extend the scope of the Clean Air Act (CAA) for controlling the amount of CO2 emissions.
With the increasingly serious problem of global warming, concerns on E&E (energy and environment) have attracted many researchers to find solutions for the trade-off point between environment protection and economic growth, or CO2 emission and energy consumption. Färe et al. [30] incorporated a DEA framework from the Malmquist index to calculate the relative efficiency change and technical progress of production technology. For the last two decades, DEA analytical frameworks have been widely applied in the environmental and energy fields (see Färe et al. [1], Pasurka [2], Wang [3], Zaim [4], Zhou et al. [5], [6], Zhou and Ang [7]. In energy efficiency studies, undesirable outputs are usually discussed in DEA models (Scheel [31]), with two ways to deal with them: 1) treat undesirable outputs as weakly disposable (WD) variables in their original forms; see, for example, Färe and Grosskopf [8], [9], [10], [32] who measured energy efficiency using the assumption that WD variables are consistent with the physical laws and the standard axioms of production theory; 2) consider the undesirable outputs as strongly (freely) disposable (SD) in various translated forms, such as in the form of their reciprocals or in the form of their additive inverses [11], [12], [13].
There is a large number of application studies on energy efficiency that use DEA within many countries. Hu and Wang [33] employed DEA to analyze the energy efficiencies of 29 administrative regions in China for the period 1995–2002 with the total-factor energy efficiency (TFEE) index. Mukherjee [34] applied DEA to measure the energy efficiency of manufacturing industries in India’s major states based on technical efficiency and cost efficiency models. Tsutsui and Goto [35] provided a weighted SBM (WSBM) model to measure the overall management efficiency of 90 electric power companies in the U.S. during the 1990s. Zhang et al. [36] used DEA window analysis to investigate the total-factor energy efficiency of 23 developing countries during the period 1980–2005. Biresh et al. [37] applied output-oriented environmental DEA models using a confected dataset of ten firms and a real-life dataset of 22 OECD countries. Honma and Hu [38] utilized DEA to calculate the total-factor energy efficiency at the industry level to compare Japan’s energy efficiency with 14 other developed countries during 1995–2005.
Although there are many above papers that focused on energy and environment efficiencies - e.g., CO2 emissions and hazardous wastes (Arcelus and Arocena [39], Ramanathan [40], Zhou et al. [6], [7], Yang and Pollitt [41], Emrouznejad et al. [42], Sueyoshi and Goto [43], [44], Wang et al. [45]) - one limitation of them is that they usually investigate undesirable output (e.g., CO2) efficiency within cross-sectional data and not over time. It is therefore worthwhile to build a dynamic DEA model to analyze energy efficiency.
The rest of this study is organized as follows. Section 2 shows the dynamic DEA model. Section 3 presents the results of empirical evaluation. Section 4 reveals a discussion of the results. Section 5 offers conclusions and policy implications.
Section snippets
Methods
DEA dynamic development, first proposed from Kloop [46], window analysis, the subsequent development, Malmquist index, proposed from Malmquist [47], Färe, Grosskoft, Norris and Zhang [30]. (Divided into two different categories, catch up and innovation effect), but these analyzes did not analyze the impact of “the effect of carry-over activities” during these two periods. Färe and Grosskopf [48] proposed the first analyze of dynamic which put the “inter-connecting activities” into it, the
Results
The research sample covers 26 OECD countries and China during 2000–2010 according to the United Nations data (UNdata) and the World Bank data in 2015. One problem in data collection is that we did not find data for Estonia, Ireland, Luxembourg, New Zealand, Norway, Slovak Republic, South Korea, and Slovenia. We employ six variables: three are input variables, two output variables, and one carry-over variable. The input variables are land area, population, and energy use, the output variables
Discussion
The 1996 Intergovernmental Panel on Climate Change (IPCC) of the United Nations Climate Change estimated that in order to stabilize carbon dioxide concentrations by double the pre-industrial (550 ppm) levels, then the world’s carbons must be reduced by half in the 21st century. Thus, in December 1997 in Kyoto, Japan the Third Assembly of States Parties signed the Kyoto Protocol to regulate 38 countries and the European Union to control emissions of greenhouse gases in order to reduce the
Conclusions and policy implications
Environmental topics in recent years have become very popular, as many scholars and countries have talked about energy efficiency and the greenhouse effect. Some countries are putting a lot of resources into controlling CO2 emissions after the Kyoto Protocol, but some countries have even cut their productivity efficiency during this period. In other words, some countries’ productivity efficiency fell in order to control CO2 emissions. This study employs the dynamic DEA model to evaluate the
Acknowledgements
This thesis is supported by the Major project of Ministry of Education in China (14JZD020).
References (54)
- et al.
An activity analysis model of the environmental performance of firms application to fossil-fuel-fired electric utilities
Ecol Econ
(1996) Decomposing electric power plant emissions within a joint production framework
Energy Econ
(2006)Decomposing energy productivity change: a distance function approach
Energy
(2007)Measuring environmental performance of state manufacturing through changes in pollution intensities: a DEA framework
Ecol Econ
(2004)- et al.
Slacks-based efficiency measures for modeling environmental performance
Ecol Econ
(2006) - et al.
A non-radial DEA approach to measuring environmental performance
Eur J Operational Res
(2007) - et al.
Decomposition of aggregate CO2 emissions: a production theoretical approach
Energy Econ
(2008) - et al.
Modeling undesirable factors in efficiency evaluation: Comment
Eur J Operational Res
(2004) - et al.
Modeling undesirable factors in efficiency evaluation
Eur J Operational Res
(2002) - et al.
Dynamic DEA: a slacks-based measure approach
Omega
(2010)
Dynamic data envelopment analysis modeling intertemporal behavior of a firm in the presence of productive inefficiencies
Econ Lett
Dynamic data envelopment analysis: a relational analysis
Eur J Operational Res
Measuring dynamic efficiency under risk aversion
Eur J Operational Res
A dynamic efficiency model using data envelopment analysis
Int J Prod Econ
Returns to scale in dynamic DEA
Eur J Operational Res
A mathematical model for dynamic efficiency using data envelopment analysis
Appl Math Comput
Measuring dynamic efficiency: theories and an integrated methodology
Eur J Operational Res
Fuzzy dynamic multi-objective data envelopment analysis model
Expert Syst Appl
DEA window analysis for environmental assessment in a dynamic time shift: performance assessment of U.S. coal-fired power plants
Energy Econ
Undesirable outputs in efficiency evaluations
Eur J Operational Res
Pollution abatement activities and traditional productivity
Ecol Econ
Total-factor energy efficiency of regions in China
Energy Policy
Energy use efficiency in the Indian manufacturing sector: an interstate analysis
Energy Policy
A multi-division efficiency evaluation of U.S. electric power companies using a weighted slacks-based measure
Socio-Economic Plan Sci
Total-factor energy efficiency in developing countries
Energy Policy
Industry-level total-factor energy efficiency in developed countries: a Japan-centered analysis
Appl Energy
Incorporating both undesirable outputs and uncontrollable variables into DEA: the performance of Chinese coal-fired power plants
Eur J Operational Res
Cited by (142)
Impact of new energy demonstration city policy on energy efficiency: Evidence from China
2023, Journal of Cleaner ProductionThe effect of electricity market reform on energy efficiency in China
2023, Energy Policy