Advertisement

Literature on Energy Demand

  • Nabaz T. KhayyatEmail author
Chapter
  • 552 Downloads
Part of the Green Energy and Technology book series (GREEN)

Abstract

The substitutability and complementarity of energy input have been widely studied during the last four decades. The empirical results were mixed between energy-capital complementarity and energy-capital substitutability. From the previous literature a flexible functional form (Translog) was generally used to model production, cost, energy demand or a combination of them depending on the objective of cost minimization or output maximization. For their empirical analyses the different studies utilized data covering different countries, regions, industrial sector, and in few cases firm levels. The results in general indicated substitution between capital and energy, while complementarity between capital and energy was also frequently observed. The degree of substitutability and complementarity differ significantly by different dimensions of the data and the unit’s characteristics. Energy efficiency is hard to conceptualize and there is no single commonly accepted definition. From the literature, energy intensity at the national level is calculated as the ratio of energy use to GDP. This variable is often taken as a proxy for general energy efficiency in production. However, this aggregate energy consumption to GDP ratio is too simple to explain an economy’s energy use pattern, and may lead to difficulties and misunderstandings in interpreting the energy intensity indicators. The energy/GDP ratio includes a number of other structural factors that can significantly affect those indicators. Hence, it is necessary to fix the structural change effect in measuring energy intensity at the aggregated level in the industrial sector. The demand for energy is defined as a derived demand that arises for satisfying some needs which are met through the use of appliances. The response to change in the energy demand is partially characterized and explained by changes in the behavior of the decision maker. Thus, the elasticity of energy that respond to changes in the short run is incomplete, while in the long run it will be accumulated over time and fully captured. A key hypothesis required for determining demand for input factors of production is the profit maximization , which depends on the level of output and a limited combination of input factors that give a highest production output. This is called a production function, in which it explains the maximum level of production given a number of possible combinations of input factors used in the process.

Keywords

Data Envelopment Analysis Stochastic Frontier Analysis Translog Cost Function Pool Time Series Energy Efficiency Index 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The substitutability and complementarity of energy input have been widely studied during the last four decades. The empirical results were mixed between energy-capital complementarity and energy-capital substitutability. From the previous literature a flexible functional form (Translog) was generally used to model production, cost, energy demand or a combination of them depending on the objective of cost minimization or output maximization. For their empirical analyses the different studies utilized data covering different countries, regions, industrial sector, and in few cases firm levels. The results in general indicated substitution between capital and energy, while complementarity between capital and energy was also frequently observed. The degree of substitutability and complementarity differ significantly by different dimensions of the data and the unit’s characteristics. Energy efficiency is hard to conceptualize and there is no single commonly accepted definition. From the literature, energy intensity at the national level is calculated as the ratio of energy use to GDP. This variable is often taken as a proxy for general energy efficiency in production. However, this aggregate energy consumption to GDP ratio is too simple to explain an economy’s energy use pattern, and may lead to difficulties and misunderstandings in interpreting the energy intensity indicators. The energy/GDP ratio includes a number of other structural factors that can significantly affect those indicators. Hence, it is necessary to fix the structural change effect in measuring energy intensity at the aggregated level in the industrial sector. The demand for energy is defined as aderived demand that arises for satisfying some needs which are met through the use of appliances. The response to change in the energy demand is partially characterized and explained by changes in the behavior of the decision maker. Thus, the elasticity of energy that respond to changes in the short run is incomplete, while in the long run it will be accumulated over time and fully captured. A key hypothesis required for determining demand for input factors of production is the profit maximization , which depends on the level of output and a limited combination of input factors that give a highest production output. This is called a production function, in which it explains the maximum level of production given a number of possible combinations of input factors used in the process.

3.1 Introduction

This chapter and the next one will present the theoretical foundation of this study. The current chapter will clearly outline the background of the problem, along with presenting the relevant theories and existing research related to energy demand. The short falling of the previous empirical studies regarding the energy demand will also be discussed in this chapter.

The theoretical foundations of this study amount to more than 300 reviews of books, peer reviewed journal articles, institutional and annual reports, dissertations, and several websites. The three subtopic areas: Stochastic production functions, factor requirement , and production risk on energy demand were researched to conduct a comprehensive literature search.

3.2 Inter-Factor Substitutability and Complementarity

In this section, previous relevant literature on inter-factor substitutability is introduced. The main focus is particularly dealing with substitutability between energy and other input factors of production such as non-ICT capital and labor. The issue of energy substitutability and complementarity has been widely studied during the last four decades. The empirical results were mixed between energy-capital complementarity and energy-capital substitutability. In the following, the literature and its main findings are presented in chronological order.

Hudson and Jorgenson (1974) constructed an inter-industry production model aimed at energy policy analysis. They divided the US business sector into nine industries, namely agriculture, non-fuel mining and construction, manufacturing excluding petroleum refining, transportation, communications, trade and services, coal mining, crude petroleum and natural gas, petroleum refining, electric utilities, and finally gas utilities. Using time series data covering the period 1947–1971, they aggregated the input factors into four main commodity groups, namely capital, labor, materials, and energy. Hudson and Jorgenson (1974) thus, concluded that energy, capital, and materials are complements in the US industrial sector.

In a first attempt Berndt and Wood (1975) have empirically tested the substitutability between energy and non-energy input factors. They assumed a Translog functional form in modeling the production structure for the US manufacturing. For their analysis they consigned an empirical value on the elasticity of substitution and found that energy demand is price elastic, while energy and capital are having a complimentary relationship.

By using pooled panel data set of manufacturing for nine countries, namely Belgium, Denmark, France, Italy, Netherlands, Norway, UK, US, and West Germany, Griffin and Gregory (1976) studied the intersubstitutability between energy and capital. They applied the Translog production function to represent the production technology . In their research, the authors identified the long run substitutability between energy and capital.

The energy demand for Canadian manufacturing sector is estimated by Denny et al. (1978) during the period 1949–1970. The authors applied a non-homothetic generalized Leontief cost function. Their findings revealed that energy and capital are complement. Magnus (1979) applied a generalized Cobb-Douglas cost function using annual aggregate time series data for the Netherlands’ economy covering the periods 1950–1976. According to his results, energy and labor were substitutes , whereas, energy and capital were complement. A pooled, cross sectional and time series data of manufacturing sector for the US, Canada, West Germany, Japan, the Netherlands, Norway, and Sweden covering the period 1963–1974 is used by Ozatalay et al. (1979). They estimated a Translog cost function and found that energy and capital are substituting each other.

In a ground breaking paper Pindyck (1979) has introduced an econometric model to analyze industrial demand for energy. The model was applied to ten industrial countries, namely Canada, France, Italy, Japan, Netherlands, Norway, Sweden, UK, US, and West Germany, covered the period 1963–1973. His analysis was aiming at determining the level of substitution effects among capital, labor, and energy inputs. Subsequently, comprehensive literatures have been developed based on Pindyck’s original model.

By constructing a pooled dataset of ten industries in the US manufacturing sector, Field and Grebenstein (1980) disaggregated the capital stock into physical capital and working capital in their study. The disaggregation was an attempt to reveal the argument about the role of energy and its relationship’s change by capital type. They found a large complementarity relationship between physical capital and energy, while substitutability was observed between working capital and energy.

The Cobb-Douglas production function is applied by Suzuki and Takenaka (1981) incorporating energy and capital investment factors as input substitutions. They found that the Japanese economy will achieve higher growth rate if actively substitutes capital for energy. In a similar study, Hazilla and Kopp (1982) by dividing the physical capital into structure and equipment, they found complementarity between energy and one component of physical capital, and substitutability between energy and other components of physical capital.

The inter-factor substitutability is investigated by Turnovsky et al. (1982) using time series data for the Australian manufacturing sector during two periods 1946–1947 and 1974–1975 focusing on energy inputs. They estimated elasticity of substitution for capital, labor, materials, and energy, and found that energy and capital have a substitutability relationship. Harper and Field (1983) estimated the elasticity of substitution for capital, labor, materials, and energy for the US manufacturing sector during the period 1971–1973, using regional cross sectional data and utilizing a Translog approximation approach. They found that capital and energy are substitutes and the degree of substitution differs by regional location.

Chichilnisky and Heal (1993) came up with a different result about the substitutability and complementarity of energy with non-energy inputs. They developed a total cross price elasticity of demand for energy and capital, in which it considers full adjustments in the long run in multi-sector economy once the energy price changes in the long run. Their finding illustrates that the capital and energy’s substitutability relationship tends to change into complementarity once the energy price rises in the long run. Hunt (1984) extended the results obtained by Berndt and Wood (1979) through investigating the role of technological progress in production with the presence of factor enhancing technological progress. Hunt’s study was conducted through accounting for linear trend as a determinant factor, while Iqbal (1986) applied the Translog cost function to estimate the inter-factor substitutability of labor, capital, energy, and fuel types for five manufacturing industries in Pakistan. She found that labor, capital, and energy are substitutes. Saicheua (1987) through the use of pooled cross section and time series data of manufacturing sector in Thailand for the periods 1974–1977, found the substitutability between input demand factors (capital, labor, and energy). In addition, Saicheua found that in all industries capital and energy were substitutes.

The elasticities of demand for energy and non-energy inputs are measured by Siddayao et al. (1987) for two industries in three Asian countries, namely Bangladesh for the period 1970–1978, the Philippines 1970–1980, and Thailand 1974–1977. They found labor and energy are substitutes , and the elasticity is higher than in the developed countries’ industrial sector . A study conducted by Kim and Labys (1988) to investigate the long run elasticity between energy demand and price of energy and the level of inter-factor substitutability. They analyzed the production structure of Korean industrial sector using pooled time series data, covering the period 1960–1980. Their finding reveled that energy and capital have substitutability relationship in the total manufacturing and total industry level, while complementarity in some others sub-industrial sectors. The factor demands of manufacturing sectors in Japan and US is investigated by Morrison (1988) to characterize the short run and long run price elasticities of demand, she found that in both countries the energy and capital are complement , while other inputs were substitutes.

Apostolakis (1990) conducted a literature survey on energy and capital relationship. He found that studies used time series data and methodology to capture the short run effects are mainly implied complementarity between capital and energy, whereas studies that used cross sectional data captured the long run effects implied substitutability between the two factors. McNown et al. (1991) studied the substitution elasticities of capital, labor, and energy for manufacturing sector in India, Pakistan, and Bangladesh. They proved the substitutability of capital and energy using Translog cost function although the substitutability was differed in elasticity measures for the three countries.

The relationship between economic growth and elasticity of substitution is investigated by Yuhn (1991) through analyzing the inter-factor substitutability between input factors (capital, materials, labor, and energy) comparing South Korea with the US manufacturing sector. The study found substitutability between capital and energy in both countries. Watanabe (1992) through investigating the substitutability of energy and capital for Japanese manufacturing sector during the period 1970–1987 argued that the energy and capital substitution is resulted from the technological innovation and R&D investment effort that led to faster growth of Japanese industrial technology.

Atkson and Kehoe (1995) derived a model called putty-clay model and applied it to study the equilibrium dynamic of investment capital, wages, and energy. They found that energy and capital are negatively correlated and thereby are considered substitutes. Christopoulos (2000) used a Translog cost function to model a dynamic structure of production and to measure the substitutability degree between three types of energy Crude oil, electricity, and diesel, and capital and labor. He used the Greek’s manufacturing sector time series data covering the period 1970–1990 and found energy and capital are substitutes.

In an attempt to study the substitution relationships in the German economy, Koschel (2000) argued that energy, materials, and capital inputs are substitutes. He applied the Translog function and used a pooled time series and cross sectional data for the period 1978–1990 to estimate price and substitutional elasticities between capital, labor, materials, and energy for fifty sectors aggregated into four sectors energy-supply, energy- intensive manufacturing, non-energy intensive manufacturing, and service sectors. The results showed variations in the degree of substitutability between capital, materials, labor, and energy for the different sectors. Kemfert and Welsch (2000) estimated the nested constant elasticity of substitution (CES) production function and the elasticity substitution using two different datasets for German economy. The datasets included aggregate time series data covering entire German industrial sector for the period 1970–1988 and a time series data which covered the same period for seven industries in Germany. The industries involved were chemical industry, stone and earth, iron, non-ferrous metal, vehicles, food and paper. They found energy and capital were substitutes, based on the aggregated time series data and the degree of substitutability were differing across the sectors under study based on the second time series dataset.

Mahmud (2000) studied the role of energy in Pakistan’s manufacturing sector applying the Generalized Leontief restricted cost function on the manufacturing sector’s time series data for the period 1972–1993. He found the inter-factor substitutability between energy and capital and inter-fuel substitutability between electricity and gas. Frondel and Schmidt (2002) claimed that the issue of substitutability and complementarity of energy and capital is not about the econometric methodology as discussed in previous literature such as Apostolakis (1990). Instead, they argued that the estimated Translog cost function for cost share is more appropriate for this issue. Their implication was based on the review of previous empirical works and showed that there is a correlation between cross price elasticity and the cost share of capital and energy due to technological change. In addition, they found evidence of complementarities occurring only when the cost share of both inputs are small; otherwise, the two inputs are always substitutes.

Thompson (2006) in addition to his finding about energy-capital substitutability, emphasized on the degree and direction of this substitutability. The study described the substitution of capital and energy input through the derivation of cross price elasticity, using Cobb-Douglas and the Translog production and cost functions . In contrast a high degree of complementarity between energy and capital is found in a recent study conducted by Kander and Schön (2007) on Sweden industrial and manufacturing sectors for the period 1870–2000. They used a direct measure of technical efficiency and investigated the short and long-run energy and capital relationships to identify the type of relation between capital and energy.

Arnberg and Bjorner (2007) applied Translog and linear logit approximation to estimate factor demand models for capital, labor, and energy inputs using micro panel data of Danish industrial companies for the years 1993, 1995, 1996, and 1997. The authors found labor to be substitutable with energy and capital inputs. Ma et al. (2008) applied a two-stage Translog cost function on a panel data of 31 autonomous regions in China covering the periods 1995–2004. The objective was to measure the elasticities of substitution. They found inter-factor substitutability, i.e. capital and labor are substitutes for energy. In addition, they found the inter-fuel complementarity between coal and electricity and inter-fuel substitutability between electricity and diesel. Koetse et al. (2008) through their literature survey about elasticity of substitution applied the Meta regression analysis of previous literature’s results and found energy and capital are substitutes, and the degree of the substitutability differs across regions and time periods.

A different approach is taken to model the structure of Korean industries using a dynamic factor demand model by Khayyat et al. (2014), they examined the changes in the South Korean industrial productivity between 1980 and 2009. Their finding reveled that ICT and non-ICT capital are substitutes for labor and energy use.

In sum, the review of the comprehensive literature presented above suggest that a flexible functional form (Translog) is used to model production, cost, energy demand or a combination of them depending on the objectives of cost minimization or output maximization. For their empirical analysis the different studies utilized data covering different countries, regions, industrial sector and in few cases firm levels. The results in general indicate substitution between capital and energy, while complementarity between energy and capital is also frequently observed. The degree of substitutability and complementarity differ significantly by different dimensions of the data and the unit’s characteristics.

3.3 Energy Efficiency

Energy efficiency is hard to conceptualize, and there is no single commonly accepted definition. A frequently occurring question concerns the level of detail necessary to carry out a cross-country or cross-industry comparison without distortions due to structural differences.

From the literature, energy intensity at the national level is calculated as the ratio of energy use to GDP, and this variable is often taken as a proxy of general energy efficiency in production (Ang 2006). A lower rate of use per unit of output indicates a higher level of efficiency. At the industry level, it is measured as the ratio of energy use to value of production for a given period of time.

However, this approach has several limitations for example the aggregate energy consumption to GDP ratio is too simple to explain an economy’s energy use patterns. Furthermore, this could lead to difficulties and misunderstandings in interpreting the energy intensity indicators, because energy/GDP ratio includes a number of other structural factors that can significantly affect these indicators. Hence, it is necessary to fix the structural change effect in measuring energy intensity at the aggregated level in the industrial sector (Ang 2004; Boyd et al. 1988).

There are several studies elaborate with the structural change challenge. A look at the case of South Korea, Choi et al. (1995) proposed a method to decompose the aggregate energy demand applying the Divisia approach and using the data of the South Korean manufacturing industry. Three components are distinguished: Structural change, inter-fuel substitution, and real energy intensity. The results showed that the increase in the aggregated energy intensity since 1988 was mainly due to increase in the real energy intensity, and the contribution from the effect of structural change and fuel substitution is small. Jung and Park (2000) applied the method of real energy intensity to analyze the industrial structural change effect from energy intensity. The conventional aggregated energy intensity in South Korean manufacturing sector had improved by almost three times than the real energy intensity. It is found that the conventional energy intensity could be overestimated, because it contains the effect of structural change.

The energy efficiency is a critical issue of many national energy policies, but little attention has been paid to define and measure the efficiency index. However, there has been continuous efforts to calculate the energy efficiency index by using the concept of Stochastic Frontier Analysis (SFA) and Data Envelopment Analysis (DEA). Below are some key literatures that evaluated both SFA and DEA:

Boyd et al. (1988) used SFA to develop an energy performance index (EPI), which is a statistical benchmarking tool of the US EPA Energy Star Program to assess industrial plant energy efficiency. Hjalmarsson et al. (1996) provided a comparison of SFA and DEA, and Heshmati (2003) provided a review of the literature on performance measurement in manufacturing and service industries.

Reinhard et al. (2000) estimated environmental efficiency measures for Dutch dairy farms. They defined environmental efficiency as the ratio of minimum feasible to observed use of environmentally detrimental inputs such as nitrogen surplus, phosphate surplus, and the total energy use, they compared two methods for calculating efficiency namely SFA and DEA. The result suggested that the environmentally detrimental input is used most inefficiently, both at individual farms and at the aggregate levels.

Hu and Wang (2006) analyzed energy efficiency of 29 administrative regions in China for the period 1995–2002. Unlike several other studies for regional productivity and efficiency in China, where energy input is neglected, this study included the energy use to find the target energy input by using DEA. The index of total factor energy efficiency (TFEE) is defined as the ratio of the target energy input to the actual energy input. The developed area (East) in China has the highest TFEE, the least developed area (West) has the second best rank, while the developing area (Central) has the worst rank, even though this area shows second highest level of GDP output in China. This “U-shaped” relationship between the area’s TFEE and per capita income confirms that energy use efficiency eventually improves the economic growth.

In a recent study, Filippini and Hunt (2011) estimated aggregate energy demand frontier by using SFA for 29 countries over the period 1978–2006. Energy intensity might give a reasonable indication of energy efficiency improvements but this is not always the case. Hence, they suggested an alternative way to estimate the economy-wide level of energy efficiency, in particular through frontier estimation and energy demand modeling. Zhou et al. (2012) proposed a parametric frontier approach to estimate economy-wide energy efficiency. They used the Shephard energy distance function (Shephard 1953) to define energy efficiency index, they adopted the stochastic frontier analysis (SFA) to estimate the index by using a sample of 21 OECD countries. It is found that the proposed parametric frontier approach has a higher explanation power in energy efficiency index compared to its non-parametric DEA frontier counterpart.

The stochastic frontier function has generally been used in production theory to measure economic performance of production units (See for example: Aigner et al. 1977; Battese and Coelli 1995; Jondrow et al. 1982). The main concept of frontier approach is that the function presents maximum output or minimum level of economic input indicators. Kumbhakar and Lovell (2000) discussed the interpretation of the efficiency in an input requirement function. An input requirement function gives a minimum level of input used by an industry for production of any given level of output. Literatures on input requirement function were mainly focused on labor use efficiency because labor is an important part of input factors in production, e.g. (See for example: Battese et al. 2000; Kumbhakar et al. 2002; Masso and Heshmati 2004). However, the energy use is the main focus of this study. Therefore, energy use efficiency is estimated by means of stochastic energy requirement function.

Attempts have been also made to analyze the dynamic factor demand and its adjustment process. Pindyck and Rotemberg (1983) examined how input factors respond over time when changes in the price of energy or output level can be anticipated. This study focused on the importance of adjustment cost and the role of energy as a production factor. Urga and Walters (2003) compared dynamic flexible cost functions to analyze inter-fuel substitution in the US industrial energy demand, while Yi (2000) compared dynamic energy demand models using Swedish manufacturing industries.

The industrial demand for energy has been frequently studied but these studies solely investigated the relationships between energy and non-energy factors. A complementary relation between energy, capital, and labor were investigated based on the US manufacturing time series data. The models have different views of production technology , yet can distinguish the relationships between any two factors in forms of complementarity or substitutability.

In one example, Clifton (1995) analyzed the inter-fuel substitution of the US industrial sector for the period 1960–1992 and found that dynamic linear logit model is providing global properties that are superior to those of a comparable dynamic Translog models. Ang and Lee (1994) developed an energy consumption decomposition model, using data from Singapore and Taiwan. The authors attempted to identify the effects of structural change on energy efficiency based on energy coefficient and measures of elasticity of demand. An analysis of the relationship between energy intensity and total factor productivity is conducted recently by Sahu and Narayanan (2011). Their finding indicated that energy intensity is negatively related to the total factor productivity, and hence energy use efficiency is required by the industry to operate efficiently.

3.4 Energy Demand

The demand for energy is defined by Bhattacharyya and Timilsina (2009, p. 16) as follows:

…a derived demand that arises for satisfying some needs which are met through use of appliances.

According to this definition, the energy demand depends on the type of energy chosen to be used in a device for a process or activity, in which it will be influenced by the price of the chosen energy type, the price of the device used by the energy type, the availability of the devise used, and other factors such as environmental conditions, decision maker’s preferences, income, demand for energy substitutes, etc. Accordingly, changes in the demand for energy depend mainly on the supply of the device used. Thus, response to the change will lead to inflexible results, as changes in response to the changes in supply of the device might be influenced by factors other than energy demand. The supply of the device used depends mainly on a set of characteristics such as device cost, availability, and efficiency (Bhattacharyya and Timilsina 2009).

The response to change in the energy demand is partially characterized and explained by changes in the behavior of the decision maker. Thus, the elasticity of energy that respond to changes in the short run is incomplete, while in the long run will be accumulated over time and fully captured. The short run elasticity will depend on the output level, while in the long run other factors in addition to the level of output will determine the size of the elasticity such as taxes, prices, technical progress, changes in the industry structure, and policies toward more efficient use of energy (Schön 2000).

Factors that derive the demand for energy by industries are determined based on the production theory with a priori expected outcome. These factors are different by industries as well as over time. Energy is considered as an input in the production, and hence, the cost minimization approach is applied when the firm is maximizing the profit (Uri 1982).

The cost minimization and profit maximization goals of the producer in the industrial sector are subjected to a number of restrictions such as the production process and its capacity in producing maximum quantity of output given the level of inputs available and used, the fixed capacity of the firm during a certain time period, price and availability of different inputs used in the production process, and the price of their substitutes . The factor demand functions can be derived from the cost minimization approach, which aims at producing units of outputs up to the level that the rate of technical substitution will be equal to the price of the inputs used (Bhattacharyya and Timilsina 2009).

A key hypothesis required for determining the demand for input factors of production is the profit maximization, which depends on the level of output and a limited combination of input factors that give a highest production output. This is called a production function, in which it explains the maximum level of production given a number of possible combinations of input factors used in the process (Dougherty 2007).

In order to illustrate the discussion above in context of production of output and use of inputs, let Y it be an amount of output which can be produced by an industry i at time t. Y will use different combinations of non-ICT capital K, labor L, materials M, value added service S, ICT-capital I, and energy E. In addition to that, exogenous technical changes represented by time trend T will have positive influence on the production (Heshmati 2003). In similar with the output, all these inputs are varying by industry and time. Given initial conditions described above, the production function will be specified as follows:
$$Y_{it} = F\left( {K,L,M,S,I,E,T} \right),$$
(3.1)
Here the demand for energy can be derived using the Shephard’s lemma approach (Shephard 1953), and based on Diewert (1974). It is labeled as inverted factor demand (or factor requirement function) as follows:
$$E = f\left( {Y,P,K,L,M,S,I,T} \right),$$
(3.2)

The price of energy P is included due to the cost minimization requirement. This is a demand function for energy, it depends on output, the own price, other inputs, and time trend representing the state of technology. The price of alternative energy, if available, can be included to capture prevalence of substitution and complementarity in the demand for energy.

3.5 The Elasticity of Demand

The elasticity can be defined as responsiveness of the dependent variable to changes in the explanatory variables. It is a measure of changes in explanatory variables that affect the dependent variable. If the left and right-hand side variables are expressed in logarithmic form, all the variables then can be in different units of measurement, and yet the changes will express percentage changes, or elasticities. The elasticity is defined as if an explanatory variable such as materials use increases with one percent, how many percent the demand for energy will be changed, ceteris paribus (meaning everything else unchanged) (Allen et al. 2009).

3.6 Critique of Previous Literature

The data used to estimate energy demand in previous literature were mainly of two types: Cross sectional data within a country, in which it is considered inadequate due to the effects of location that exaggerate the elasticities such as price elasticity. The other data type used is the international cross sections, which also considered insufficient due to structural differences that direct the elasticities away from zero. Hence, the pooled time series cross sectional data is more desirable, as it addressed the shortcoming mentioned above by powerful econometric techniques such as flexible Translog production function (Hartman, 1979). The model also allows for capturing both dynamics and heterogeneity in production and energy demand.

An ideal model is required to combine theoretical and empirical tools of inter-factor substitution model often called (KLEMS) which refers to capital K, labor L, energy E, materials M, and value added services S. Further extensions of the inter-fuel substitution, dynamic partial adjustment, demand model for quasi-fixed factors, and econometric model that utilized Translog flexible functional form with production risk approach are incorporated. Furthermore, explicit treatment of elasticity demand is accounted for in this study in order to identify behavioral characteristics of individual industry and to derive relevant specific policy variables and recommendations.

3.7 Summary

From the study of inter-factor substitutability between energy and other factors of production, it is found that there are two directional approaches: One claims the substitutability and the other claims complementarity, and both are providing strong theoretical and empirical evidences. For their empirical analyses, these studies have utilized data of different countries, regions, industrial sector , and in a few cases, based on firm levels. The results in general indicate substitution between capital and energy, while complementarity between capital and energy is also frequently observed. The degree of substitutability and complementarity differ significantly by different dimensions of the data and the unit’s characteristics.

An ideal model is required to combine theoretical and empirical tools of inter-factor substitution model often called as (KLEMS) which refers to capital, labor, energy, materials, and value added services. A derivation for energy as an input factor demand function (or factor requirement function) is offered and the factors that derive the demand for energy by industries and over time are determined based on the production theory with a priori expected outcome. The cost minimization approach is applied for firm’s profit maximization, as the energy is considered an input factor of production.

Further extensions of the inter-fuel substitution, dynamic partial adjustment, demand model for quasi-fixed factors, and econometric model that utilized Translog flexible functional form with production risk approach are incorporated. Furthermore, explicit treatment of elasticity demand is accounted for in this study in order to identify behavioral characteristics of individual industry, and to derive relevant specific policy variables and recommendations.

References

  1. Aigner, D. J., Lovell, C. A. K., & Schmidt, P. (1977). Formulation and estimation of stochastic production function models. Journal of Econometrics, 6(1), 21–37.MathSciNetCrossRefzbMATHGoogle Scholar
  2. Allen, W. B., Weigelt, K., Doherty, N., & Mansfield, E. (2009). Managerial economics: Theory, applications, and cases. New York, USA: Norton.Google Scholar
  3. Ang, B. W. (2004). Decomposition analysis for policymaking in energy. Energy Policy, 32(9), 1131–1139. doi: 10.1016/s0301-4215(03)00076-4.CrossRefGoogle Scholar
  4. Ang, B. W. (2006). Monitoring changes in economy-wide energy efficiency: From energy–GDP ratio to composite efficiency index. Energy Policy, 34(5), 574–582. doi: 10.1016/j.enpol.2005.11.011.CrossRefGoogle Scholar
  5. Ang, B. W., & Lee, S. Y. (1994). Decomposition of industrial energy-consumption—some methodological and application issues. Energy Economics, 16(2), 83–92. doi: 10.1016/0140-9883(94)90001-9.CrossRefGoogle Scholar
  6. Apostolakis, B. E. (1990). Energy-capital substitutability/complementarity. Energy Economics, 12(1), 48–58. doi: 10.1016/0140-9883(90)90007-3.CrossRefGoogle Scholar
  7. Arnberg, S., & Bjorner, T. B. (2007). Substitution between energy, capital and labour within industrial companies: A micro panel data analysis. Resource and Energy Economics, 29(2), 122–136. doi: 10.1016/j.reseneeco.2006.01.001.CrossRefGoogle Scholar
  8. Atkson, A., & Kehoe, P. J. (1995). Putty-clay capital and energy (Working Paper No. 548). Retrieved from The Federal Reserve Bank of Minneapolis website: http://www.minneapolisfed.org/research/WP/WP548.pdf.
  9. Battese, G. E., & Coelli, T. J. (1995). A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empirical Economics, 20(2), 325–332. doi: 10.1007/bf01205442.CrossRefGoogle Scholar
  10. Battese, G. E., Heshmati, A., & Hjalmarsson, L. (2000). Efficiency of labour use in the Swedish banking industry: A stochastic frontier approach. Empirical Economics, 25(4), 623–640. doi: 10.1007/s001810000037.CrossRefGoogle Scholar
  11. Berndt, E. R., & Wood, D. O. (1975). Technology, prices, and the derived demand for energy. The Review of Economics and Statistics, 57(3), 259–268.CrossRefGoogle Scholar
  12. Berndt, E. R., & Wood, D. O. (1979). Engineering and econometric interpretations of energy-capital complementarity. American Economic Review, 69(3), 342–354.Google Scholar
  13. Bhattacharyya, S. C., & Timilsina, G. R. (2009). Energy demand models for policy formulation : a comparative study of energy demand models (Policy Research Working Paper WPS4866). Retrieved from World Bank website: http://econ.worldbank.org/external/default/main?pagePK=64165259&theSitePK=469372&piPK=64165421&menuPK=64166093&entityID=000158349_20090317093816.
  14. Boyd, G. A., Hanson, D. A., & Sterner, T. (1988). Decomposition of changes in energy intensity—a comparison of the divisia index and other methods. Energy Economics, 10(4), 309–312. doi: 10.1016/0140-9883(88)90042-4.CrossRefGoogle Scholar
  15. Chichilnisky, G., & Heal, G. M. (1993). Energy-capital substitution: a general equilibrium analysis. In G. M. Heal (Ed.), Critical Writings in the Economics of Exhaustible Resources (pp. 339–390). London: Edward Elgar.Google Scholar
  16. Choi, K. H., Ang, B. W., & Ro, K. K. (1995). Decomposition of the energy-intensity index with application for the Korean manufacturing-industry. Energy, 20(9), 835–842. doi: 10.1016/0360-5442(95)00041-E.CrossRefGoogle Scholar
  17. Christopoulos, D. K. (2000). The demand for energy in Greek manufacturing. Energy Economics, 22(5), 569–586. doi: 10.1016/S0140-9883(99)00041-9.CrossRefGoogle Scholar
  18. Clifton, T. J. (1995). A dynamic analysis of interfuel substitution in US industrial energy demand. Journal of Business & Economic Statistics, 13(4), 459–465. doi: 10.2307/1392391.CrossRefGoogle Scholar
  19. Denny, M., May, J. D., & Pinto, C. (1978). The demand for energy in Canadian manufacturing: Prologue to an energy policy. The Canadian Journal of Economics, 11(2), 300. doi: 10.2307/134350.CrossRefGoogle Scholar
  20. Diewert, W. E. (1974). Functional forms for revenue and factor requirements functions. International Economic Review, 15(1), 119. doi: 10.2307/2526093.MathSciNetCrossRefzbMATHGoogle Scholar
  21. Dougherty, C. (2007). Introduction to econometrics. New York, USA: Oxford University Press.Google Scholar
  22. Field, B. C., & Grebenstein, C. (1980). Capital-energy substitution in US manufacturing. The Review of Economics and Statistics, 62(2), 207. doi: 10.2307/1924746.CrossRefGoogle Scholar
  23. Filippini, M., & Hunt, L. C. (2011). Energy demand and energy efficiency in the OECD countries: A stochastic demand frontier approach. The Energy Journal, 33(2), 59–80.Google Scholar
  24. Frondel, M., & Schmidt, C. M. (2002). The capital-energy controversy: An artifact of cost shares? Energy, 23(3), 53–79.Google Scholar
  25. Griffin, J. M., & Gregory, P. R. (1976). An intercountry translog model of energy substitution responses. The American Economic Review, 66(5), 845–857. doi: 10.2307/1827496.Google Scholar
  26. Harper, C., & Field, B. C. (1983). Energy substitution in US manufacturing: A regional approach. Southern Economic Journal, 50(2), 385. doi: 10.2307/1058213.CrossRefGoogle Scholar
  27. Hartman, R. S. (1979). Frontiers in energy demand modeling. Annual Review of Energy, 4(1), 433–466. doi: 10.1146/annurev.eg.04.110179.002245.CrossRefGoogle Scholar
  28. Hazilla, M., & Kopp, R. J. (1982). Substitution between energy and other factors of productions: US industrial experience 1958–1974. Final Report EPRI EA Research Project.Google Scholar
  29. Heshmati, A. (2003). Productivity growth, efficiency and outsourcing in manufacturing and service industries. Journal of Economic Surveys, 17(1), 79–112. doi: 10.1111/1467-6419.00189.MathSciNetCrossRefGoogle Scholar
  30. Hjalmarsson, L., Kumbhakar, S. C., & Heshmati, A. (1996). DEA, DFA and SFA: A comparison. Journal of Productivity Analysis, 7(2–3), 303–327. doi: 10.1007/Bf00157046.CrossRefGoogle Scholar
  31. Hu, J. L., & Wang, S. C. (2006). Total-factor energy efficiency of regions in China. Energy Policy, 34(17), 3206–3217. doi: 10.1016/j.enpol.2005.06.015.CrossRefGoogle Scholar
  32. Hudson, E. A., & Jorgenson, D. W. (1974). US energy policy and economic growth, 1975–2000. The Bell Journal of Economics and Management Science, 5(2), 461. doi:  10.2307/3003118.
  33. Hunt, L. C. (1984). Energy and capital—substitutes or complements—some results for the UK industrial sector. Applied Economics, 16(5), 783–789. doi: 10.1080/00036848400000027.CrossRefGoogle Scholar
  34. Iqbal, M. (1986). Substitution of labour, capital and energy in the manufacturing sector of Pakistan. Empirical Economics, 11(2), 81–95. doi: 10.1007/bf01987506.CrossRefGoogle Scholar
  35. Jondrow, J., Lovell, C. A. K., Materov, I. S., & Schmidt, P. (1982). On the estimation of technical inefficiency in the stochastic frontier production function model. Journal of Econometrics, 19(2–3), 233–238. doi: 10.1016/0304-4076(82)90004-5.MathSciNetCrossRefGoogle Scholar
  36. Jung, T. Y., & Park, T. S. (2000). Structural change of the manufacturing sector in Korea: Measurement of real energy intensity and CO2 emissions. Mitigation and Adaptation Strategies for Global Change, 5(3), 221–238. doi: 10.1023/a:1009661000859.CrossRefGoogle Scholar
  37. Kander, A., & Schön, L. (2007). The energy-capital relation—Sweden 1870–2000. Structural Change and Economic Dynamics, 18(3), 291–305. doi: 10.1016/j.strueco.2007.02.002.CrossRefGoogle Scholar
  38. Kemfert, C., & Welsch, H. (2000). Energy-capital-labor substitution and the economic effects of CO2 abatement: Evidence for Germany. Journal of Policy Modeling, 22(6), 641–660. doi: 10.1016/S0161-8938(98)00036-2.CrossRefGoogle Scholar
  39. Khayyat, N. T., Lee, J., & Heshmati, A. (2014). How ICT investment and energy use influence the productivity of Korean industries. IZA Discussion Paper No. 8080. Retrieved Feb 17, 2015, from SSRN: http://ssrn.com/abstract=2424179.
  40. Kim, B. C., & Labys, W. C. (1988). Application of the translog model of energy substitution to developing-countries—the case of Korea. Energy Economics, 10(4), 313–323. doi: 10.1016/0140-9883(88)90043-6.CrossRefGoogle Scholar
  41. Koetse, M. J., de Groot, H. L. F., & Florax, R. J. G. M. (2008). Capital-energy substitution and shifts in factor demand: A meta-analysis. Energy Economics, 30(5), 2236–2251. doi: 10.1016/j.eneco.2007.06.006.CrossRefGoogle Scholar
  42. Koschel, H. (2000). Substitution elasticities between capital, labour, material, electricity and fossil fuels in German producing and service sectors (Discussion Papers No. 00–31). Mannheim. Retrieved from ZEW the Centre for European Economic Research website: http://www.zew.de/en/publikationen/publikation.php3?action=detail&nr=435.
  43. Kumbhakar, S. C., Hjalmarsson, L., & Heshmati, A. (2002). How fast do banks adjust? A dynamic model of labour-use with an application to Swedish banks. Journal of Productivity Analysis, 18(1), 79–102.CrossRefGoogle Scholar
  44. Kumbhakar, S. C., & Lovell, C. A. K. (2000). Stochastic frontier analysis. Cambridge: UK.CrossRefzbMATHGoogle Scholar
  45. Ma, H., Oxley, L., Gibson, J., & Kim, B. (2008). China’s energy economy: Technical change, factor demand and interfactor/interfuel substitution. Energy economics, 30(5), 2167–2183.CrossRefGoogle Scholar
  46. Magnus, J. R. (1979). Substitution between energy and nonenergy inputs in the Netherlands. International Economic Review 465483, 20(2 SRC—GoogleScholar), 1950–1976.Google Scholar
  47. Mahmud, S. F. (2000). The energy demand in the manufacturing sector of Pakistan: Some further results. Energy Economics, 22(6), 641–648. doi: 10.1016/S0140-9883(99)00031-6.CrossRefGoogle Scholar
  48. Masso, J., & Heshmati, A. (2004). The optimality and overuse of labour in Estonian manufacturing enterprises. Economics of Transition, 12(4), 683–720. doi: 10.1111/j.0967-0750.2004.00199.x.CrossRefGoogle Scholar
  49. McNown, R. F., Pourgerami, A., & Hirschhausen, C. R. (1991). Input substitution in manufacturing for three LDCs: Translog estimates and policy implications. Applied Economics, 23(1 SRC—GoogleScholar), 209–218.Google Scholar
  50. Morrison, C. J. (1988). Quasi-fixed inputs in US and Japanese manufacturing A generalized Leontief restricted cost function approach. The Review of Economics and Statistics, 70(2 SRC—GoogleScholar), 275–287.Google Scholar
  51. Ozatalay, S., Grubaugh, S., & Long, T. V. (1979). Energy substitution and national energy-policy. American Economic Review, 69(2), 369–371.Google Scholar
  52. Pindyck, R. S. (1979). Interfuel substitution and the industrial demand for energy: An international comparison. The Review of Economics and Statistics, 61(2 SRC—GoogleScholar), 169–179.Google Scholar
  53. Pindyck, R. S., & Rotemberg, J. J. (1983). Dynamic factor demands and the effects of energy price shocks. American Economic Review, 73(5), 1066–1079.Google Scholar
  54. Reinhard, S., Lovell, C. A. K., & Thijssen, G. J. (2000). Environmental efficiency with multiple environmentally detrimental variables; estimated with SFA and DEA. European Journal of Operational Research, 121(2), 287–303. doi: 10.1016/S0377-2217(99)00218-0.CrossRefzbMATHGoogle Scholar
  55. Sahu, S. K., & Narayanan, K. (2011). Total factor productivity and energy intensity in Indian manufacturing : A cross-sectional study. International Journal of Energy Economics and Policy, 1(2 SRC—GoogleScholar), 47–58.Google Scholar
  56. Saicheua, S. (1987). Input substitution in Thailand’s manufacturing sector: Implications for energy policy. Energy Economics, 9(1 SRC—GoogleScholar), 55–63.Google Scholar
  57. Schön, L. (2000). Electricity, technological change and productivity in Sweden industry. European Review of Economic History, 4(2 SRC—GoogleScholar), 175–194 (1890–1990).Google Scholar
  58. Shephard, R. W. (1953). Cost and production functions. Princeton, USA: Princeton University Press.Google Scholar
  59. Siddayao, C. M., Khaled, M., Ranada, J. G., & Saicheua, S. (1987). Estimates of energy and non-energy elasticities in selected Asian manufacturing sectors: Policy implications. Energy Economics, 9(2 SRC—GoogleScholar), 115–128.Google Scholar
  60. Suzuki, K., & Takenaka, H. (1981). The role of investment for energy conservation: Future Japanese economic growth. Energy Economics, 3(4), 233–243. doi: 10.1016/0140-9883(81)90024-4.CrossRefGoogle Scholar
  61. Thompson, H. (2006). The applied theory of energy substitution in production. Energy Economics, 28(4), 410–425. doi: 10.1016/j.eneco.2005.01.005.CrossRefGoogle Scholar
  62. Turnovsky, M. H. L., Folie, M., & Ulph, A. (1982). Factor substitutability in Australian manufacturing with emphasis on energy inputs. Economic Record, 58(160), 61–72. doi: 10.1111/j.1475-4932.1982.tb00349.x.CrossRefGoogle Scholar
  63. Urga, G., & Walters, C. (2003). Dynamic translog and linear logit models: A factor demand analysis of interfuel substitution in US industrial energy demand. Energy Economics, 25(1), 1–21. doi: 10.1016/S0140-9883(02)00022-1.CrossRefGoogle Scholar
  64. Uri, N. D. (1982). The industrial demand for energy. Resources and Energy, 4(1), 27–57. doi: 10.1016/0165-0572(82)90019-6.CrossRefGoogle Scholar
  65. Watanabe, C. (1992). Trends in the substitution of production factors to technology—empirical-analysis of the inducing impact of the energy-crisis on Japanese industrial-technology. Research Policy, 21(6), 481–505. doi: 10.1016/0048-7333(92)90006-P.CrossRefGoogle Scholar
  66. Yi, F. (2000). Dynamic energy-demand models: A comparison. Energy Economics, 22(2), 285–297. doi: 10.1016/S0140-9883(99)00042-0.CrossRefGoogle Scholar
  67. Yuhn, K. H. (1991). Economic growth, technical change biases, and the elasticity of substitution: A test of the De La Grandville hypothesis. The Review of Economics and Statistics, 73(2), 340–346.CrossRefGoogle Scholar
  68. Zhou, P., Ang, B. W., & Zhou, D. Q. (2012). Measuring economy-wide energy efficiency performance: A parametric frontier approach. Applied Energy, 90(1), 196–200. doi: 10.1016/j.apenergy.2011.02.025.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2015

Authors and Affiliations

  1. 1.Technology Management, Economics, and Policy Program, College of EngineeringSeoul National UniversitySeoulSouth Korea

Personalised recommendations