Next Article in Journal
Agency-Linked Risk Management with Ownership and Board Sub-Committee Governance: Evidence from an OECD Economy
Next Article in Special Issue
The Nexus between Financial Development and Economic Growth: Panel Data Evidence from Developing Countries
Previous Article in Journal
Gender Differences in Intra-Household Financial Decision-Making: An Application of Coarsened Exact Matching
Previous Article in Special Issue
Revisited: Monopoly and Long-Run Capital Accumulation in Two-Sector Overlapping Generation Model
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Causality between Energy Consumption and Economic Growth in the Presence of Growth Volatility: Multi-Country Evidence

1
Bond Business School, Bond University, Gold Coast, QLD 4229, Australia
2
Economy, Planning and Development, Gold Coast, QLD 9726, Australia
*
Author to whom correspondence should be addressed.
J. Risk Financial Manag. 2021, 14(10), 471; https://doi.org/10.3390/jrfm14100471
Submission received: 17 September 2021 / Revised: 30 September 2021 / Accepted: 1 October 2021 / Published: 7 October 2021
(This article belongs to the Special Issue Financial Development and Economic Growth)

Abstract

:
Falling energy intensity (increasing efficiency) is believed to be a result of more efficient production methods that have evolved over time, indicating overall sustainability in the production process. The objective of this study is to investigate the diminishing trend of energy intensity and the related volatilities in growth of energy consumption and income growth through the energy–growth nexus. The country specific long-run and short-run causal relationships among real energy consumption per capita, real GDP per capita, and the volatilities of growth in income and the growth in energy consumption are established using the method proposed by Yamamoto–Kurozumi within a cointegration framework in 48 countries. The overall findings suggest that energy intensity is falling, in conjunction with the existing evidence on the energy–growth nexus in most of the countries studied; hence, implicitly this confirms sustainability. The results based on volatility analysis show a significant decrease in energy use in response to increasing income growth volatility. The negative effects of income growth volatility on energy consumption are usually countered through compensation measures, with subsidies provided to households and producers in order to smooth the energy consumption behaviours in those economies.

1. Introduction

The main purpose of this paper, by controlling for volatility in income growth and energy use, is to shed light on the global trend of falling energy intensity using the energy–growth nexus. Falling energy intensity (increasing efficiency) is believed to be a result of more efficient production methods that have evolved over time, indicating overall sustainability in the production process. Furthermore, energy intensity is also considered an expression of the price/cost of transferring energy into GDP (Menegaki and Tsani 2018). Hence, a reduced amount of energy usage with a greater GDP has been shown to be a great success. Many studies have examined all the possible interactions between energy and GDP growth for many countries in the form of country-specific time series analyses or in a panel data framework. They have observed four types of relationship—GDP growth causing energy consumption, energy consumption causing GDP growth, a feedback relationship, or no causality. However, the aspect of energy intensity has been explicitly ignored in the energy–growth literature. Furthermore, there is no explanation available on the negative causal effects of income on energy consumption, indirectly indicating that falling energy intensity may provide a pathway towards total sustainability. The falling energy intensity has been explained to some extent by Agovino et al. (2018) and Shahbaz et al. (2018) for European countries and for the top 10 energy consumers, respectively, through the negative causal relationship of GDP to primary energy consumption. Nepal et al. (2014) found that market liberalisation, the financial sector, and most infrastructure industries drove energy intensity to fall. In contrast, Salim et al. (2019) found that increasing population and non-renewable energy usage increased energy intensity, according to evidence from Asian countries. This study moves the discussion forward by investigating energy intensity through the energy–growth nexus, using a set containing many countries.
Figure 1 presents the falling trend in energy intensity across regions over more than one and a half decades. The overall energy intensity decreased by an average of 1.58% per annum from 1990 to 2016, with the highest decline of 2.75% in the BRICS countries and the lowest decline of 1.46% in the OECD countries. These diminishing trends in energy intensity show that energy consumption has declined relative to per-capita production levels—potentially, this may lower the adverse environmental impacts of energy use and production costs, hence achieving sustainability. According to the International Energy Agency (IEA), although the global energy consumption has declined by 12 percent, the energy efficiency has increased by 13%. Along similar lines, Rühl et al. (2012) have discovered that aspects including economic systems, resource endowments, and technology have supported the lowering of energy intensity through supplementing conversions and end-use efficiency. This can be inferred as an unconventional revolution in terms of humans’ environmental impacts.
Figure 2 shows the trends in energy intensity (panel a) and real GDP per capita (panel b) for the sample of 48 countries (covering both energy exporting and importing countries, as in Jalil (2014)) included in our empirical study. Historically, the energy intensity presents wider fluctuations—although it consistently fell during the last one and a half decades—matched with irregular swings in per-capita income. Therefore, volatility analysis becomes relevant in the energy–growth nexus, a fact which has been grossly ignored in the previous literature on this topic. The causal relationship between economic growth and income volatility has been reported by several conspicuous studies (Aizenman and Marion 1993; Bernanke 1983; Pindyck 1991; Hnatkovska and Loayza 2005; Black 1987). However, there is no study available on energy–income volatility, according to the best of the author’s knowledge.
Tiba and Omri (2017) provide a comprehensive literature survey on energy growth nexus covering a range of empirical studies which are conveniently classified into country-specific, multi-country, and panel data analysis. Most of the literature investigates direction of causality between energy consumption and GDP growth, which remains a matter of concern among policy makers. Moreover, the available literature lacks sufficient consensus, hence it invites all possible corners of criticism. The current study takes a different approach and makes energy intensity the foundation of analysis, which has been grossly ignored in the previous literature on this topic. Furthermore, this study incorporates the uncertainty (volatility) in income growth on the energy–growth nexus and offers a novel contribution to the literature. The current study contributes to the energy–growth debate in many ways. First, we describe the recent shifts in the conventional energy–growth nexus in terms of diminishing energy intensity. We estimate the coefficient of income on energy intensity within the energy–growth framework, which explains the variations in energy intensity. Therefore, we examine the following four hypotheses within the energy–growth nexus in conjunction with the notion of falling energy intensity:
Hypothesis 1 (H1)
. Growth hypothesis—a unidirectional causality from energy consumption to economic growth.
Hypothesis 2 (H2).
Conservation hypothesis—a unidirectional causality from economic growth to energy consumption.
Hypothesis 3 (H3).
Feedback hypothesis—the feedback causal relationship between energy consumption and income (often referred as feedback hypothesis in the literature). and
Hypothesis 4 (H4).
Neutrality hypothesis—no causality between income and energy consumption (known as neutrality hypothesis in the literature).
In H1, the primary energy consumption helps to maintain economic growth and the energy conservation policies may be jeopardised. Furthermore, the energy efficiency would be compromised along with environmental costs. The above might also be reflected through increasing trends of the energy intensity. However, with significant falling energy intensity, the energy conservation process does not necessarily halt growth. In other words, growth remains sustainable with maintaining the given levels of energy consumption. This also refers to increasing energy efficiency, which in turn contributes towards stronger growth.
In H2, energy consumption follows economic growth. This hypothesis allows us to compare the evidence in the light of currently falling energy intensity. H3 highlights a two-way causality between energy and growth. This helps to understand the multiplier effect of primary energy consumption in terms of increasing economic growth, has some obvious environmental implications—this is beyond the scope of this study. In H4, in terms of energy-intensity, energy efficiency will not be related to growth; therefore, energy intensity is not significantly falling.
Second, we analyse income growth and energy consumption growth volatilities using exponential generalised conditional heteroscedasticity (EGARCH) models for each country and incorporate them into the standard energy–growth nexus. The connections between the volatility of income growth and energy consumption will have useful policy implications. Hacker et al. (2014) argued that income volatility causes economic insecurity. Economic insecurity may have grave implications in regard to meeting intensity-based targets of greenhouse gas emissions and sustainable economic growth. Carmona et al. (2017) observed a pro-cyclical movement in energy consumption and economic growth cycles in the USA. Gately and Huntington (2002) explored the asymmetric effects of changes in income and price on energy demand. Liddle and Sadorsky (2020) analysed the effect of asymmetric changes in income and energy prices on energy demand for the panel of 91 OECD and non-OECD countries in a non-linear cointegration framework. Their study did not find any evidence of asymmetry due to income. Their model captures the asymmetry through increases or decreases in GDP rather than the income growth volatility. Liddle et al. (2020) analysed the time-varying income and price elasticities for energy demand using a dynamic model framework. Owyang et al. (2008) discovered in the USA that high energy consumption decreases macroeconomic volatility. Rashid and Kocaaslan (2013) empirically examined the connections between income and energy volatilities in the United Kingdom They found a significant relationship between energy consumption and GDP volatilities. Furthermore, they observed that volatility in energy consumption determines volatility in income—particularly, regimes with higher volatility present a stronger relationship between energy and income volatilities. Though Rashid and Kocaaslan explore the causal relationship between these volatilities, they failed to analyse the causal link extending from the volatilities of economic growth and energy consumption growth to the economic growth and energy consumption growth and vice versa. Therefore, we investigate whether unexpected variations (volatility) in energy consumption growth are related to economic growth and energy consumption growth and vice versa. We examine the following two hypotheses within the energy growth nexus:
Hypothesis 5 (H5).
There exists a positive feedback relationship between the income growth volatility and the energy consumption volatility.
Hypothesis 6 (H6).
Uncertainties in income growth (income growth volatility) will have adverse effect on energy consumption.
In summary, we establish the causal relationship between (i) energy consumption and income (transmission between the means); (ii) volatilities of economic growth and income growth (transmission between the variances); and (iii) energy consumption and income and the volatilities of energy consumption and economic growth (transmission between the means and the variances).
Third, we establish the long-run and short-run causality relationships among real energy consumption per capita, real GDP per capita, and the volatilities of growth in income and the growth in energy consumption through the Yamamoto and Kurozumi (2006) technique within a cointegration framework. Since the results based on earlier literature are sensitive to the effects of sampling frequency, the sign rule by Rajaguru and Abeysinghe (2008) is used to determine the non-spurious causal inferences. Fourth, we estimate the country-wise results for all 48 countries. This helps us to understand the correct relationship between energy and growth by maintaining the respective drivers of energy and environment-related effects for each country. The model can also be estimated using appropriate panel data methods. However, as we shall see later, the estimated elasticities are not invariant in relation to individual countries. The results could be misleading if we assume homogeneity across the countries or regions if the model is estimated in a panel framework. Furthermore, it has been reported in Menegaki and Tsani (2018) that country-specific estimates are mostly different from the results of pooled (panel) regression models, such as those discussed in the main study by Apergis and Payne (2012). They have noted that almost 73% of the individual studies on the clean energy–growth nexus is different from their principal study.
The novel contribution to the literature is the implementation of the volatility in the energy–growth nexus to explain the falling trend in the energy intensity. At the same time, the paper also contributes to the energy–growth literature by establishing the non-spurious long-run causal inferences using a sign rule, along with the Yamamoto and Kurozumi technique, which is invariant to sampling frequency.
The rest of the paper is organised as follows. The review of literature review is presented in Section 2. Data and methods are discussed in Section 3. The results and discussion are presented in Section 4. Finally, Section 5 presents concluding remarks, along with important policy inputs.

2. Literature Review

Tiba and Omri (2017) offered a systematic review of more than 250 studies on the energy–growth nexus, covering the period from 1978 to 2014. They included studies dealing with country-specific, multi-country, and panel data analysis. Many studies have used panel data analysis to analyse the causal inferences between income and energy consumption (for example, Apergis and Payne 2012; Belke et al. 2011; Eggoh et al. 2011; Yildirim and Aslan 2012; Hossein et al. 2012; Damette and Seghir 2013; Mohammadi and Parvaresh 2014; Śmiech and Papież 2014; Larissa et al. 2020; Batrancea et al. 2020; Batrancea 2021). Although a panel data study gains from a statistical power viewpoint, it fails to capture the country-specific characteristics in many instances (Menegaki and Tsani 2018). On the other hand, country-specific studies offer a range of conclusions, subject to the methods used and the availability of data. On occasion, we have found studies with different inferences for the same country. Such differences are largely attributed to methods of analysis and the availability of time-series data. Extensive research has shown that the various sampling frequencies due to temporal aggregation and systematic sampling may distort the causal inferences and exogeneity of results (Rajaguru et al. 2018). For example, in the case of China, some have observed evidence in favour of the causality from economic growth to energy consumption (also known as conservation hypothesis in the literature) (Chang 2010; Wang et al. 2011a), others have witnessed the causality from energy consumption to economic growth (commonly known as growth hypothesis in the literature) (Wang et al. 2011b; Zhang 2011; Zhixin and Xin 2011). Similarly, many studies have estimated contradictory results from United States (Kraft and Kraft 1978; Akarca and Long 1980; Yu and Hwang 1984; Abosedra and Baghestani 1989; Yu and Jin 1992; Stern 1993; Cheng 1995; Stern 2000; Soytas et al. 2007; Ewing et al. 2007; Payne 2009; Bowden and Payne 2009; Fallahi 2011), India (Cheng 1999; Paul and Bhattacharya 2004), Malaysia (Ang 2008; Tang 2009), New Zealand (Fatai et al. 2004; Bartleet and Gounder 2010) and Pakistan (Aqeel and Butt 2001; Jamil and Ahmad 2010; Shahbaz et al. 2012). To overcome the spurious results due to sampling frequencies, the sign rule by Rajaguru and Abeysinghe (2008) is used to determine the non-spurious causal inferences between the variables of interest.

3. Data and Methods

As in Jalil (2014), we also use a sample of 48 countries. We use the World Development Indicators (2017) as the primary source of data information. According to the World Development Indicators, energy consumption refers to the use of primary energy before transformation to other end-use fuels, which is equal to indigenous production plus imports and stock changes, minus exports and fuels supplied to ships and aircraft engaged in international transport. Real GDP per capita (RGDPPC) in 2011 constant US dollars is used as a measure of income. All variables are converted into real constant prices per capita.
We use the exponential autoregressive conditional heteroskedasticity (EGARCH) model for income growth and energy consumption growth to generate the volatility measures. The EGARCH model ensures that the variances of income growth and energy consumption growth are non-negative.
The unit root properties of real GDP per capita, energy consumption per capita, income growth volatility, and energy consumption growth volatility are examined by applying the traditional unit root tests, such as the augmented Dickey–Fuller (ADF), Phillips–Perron (PP), and Kwiatkowski–Phillips–Schmidt–Shin (KPSS) unit root tests. However, a stationary variable can be misinterpreted as I(1) in the presence of structural breaks. For this purpose, we use the break-point unit root test developed by Carrion-i-Silvestre et al. (2009). The unit root test results are not reported in the Appendix A (Table A1, Table A2, Table A3 and Table A4). The results show that the real GDP per capita (RGDPPC) and energy consumptions per capita (ECPC) are I(1) at the five-percent level of significance. The results also indicate that the volatility measures of both income growth (INVOL) and energy consumption growth (EVOL) are stationary, I(0). In addition, these volatility measures will not exhibit a long-run relationship between them and others.
Since RGDPPC and ECPC are I(1), the existence of long-run relationships between RGDPPC, ECPC, INVOL, and EVOL are examined through λ t r a c e and λ max cointegration tests. When we include the two stationary variables (INVOL and EVOL) in the cointegration test, we would expect at least two cointegrating vectors. We would expect the number of cointegrating vectors to be exactly three to establish the long-run relationship between RGDPPC and ECPC in the presence of INVOL and EVOL.
The vector autoregression (VAR) model to conduct the cointegration test can be written as:
( Δ R G D P C t Δ E C P C t I N V O L t E V O L t ) = μ + Π z t 1 + i = 1 p 1 Γ i ( Δ R G D P C t i Δ E C P C t i I N V O L t i E V O L t i ) + ( ε 1 t ε 2 t ε 3 t ε 4 t )
where μ is a vector of constants and ε t has a covariance matrix Σ . The long-run 4 × 4 matrix is Π = α β and it determines how many independent linear combinations of the elements of z t = ( R G D P P C t , E C P C t , I N V O L t , E V O L t ) are stationary. Here, α and β are n × r matrices of rank r. To establish the long-run relationship in the presence of two stationary variables, we expect that r = 3. In such cases, it has the following representation with the error correction (EC) terms:
e 1 t = β z t = ln ( E C P P C t ) β ln ( R G D P P C t )
  e 2 t = I N V O L t   and   e 3 t = E V O L t
Δ ln ( R G D P P C t ) = μ 1 + γ 11 e 1 t 1 + γ 12 e 2 t 1 + γ 13 e 3 t 1 + i = 1 p 1 δ 11 , i Δ ln ( R G D P P C t i ) + i = 1 p 1 δ 12 , i Δ ln ( E C P C t i ) + i = 2 p 1 δ 13 , i I N V O L t i + i = 2 p 1 δ 14 , i E V O L t i + ε 1 t
Δ ln ( E C P C t ) = μ 2 + γ 21 e 1 t 1 + γ 22 e 2 t 1 + γ 23 e 3 t 1 + i = 1 p 1 δ 21 , i Δ ln ( R G D P P C t i ) + i = 1 p 1 δ 22 , i Δ ln ( E C P C t i ) + i = 2 p 1 δ 23 , i I N V O L t i + i = 2 p 1 δ 24 , i E V O L t i + ε 2 t
I N V O L t = μ 3 + γ 32 e 2 t 1 + γ 33 e 3 t 1 + i = 1 p 1 δ 31 , i Δ ln ( R G D P P C t i ) + i = 1 p 1 δ 32 , i Δ ln ( E C P C t i ) + i = 2 p 1 δ 33 , i I N V O L t i + i = 2 p 1 δ 34 , i E V O L t i + ε 3 t
E V O L t = μ 4 + γ 42 e 2 t 1 + γ 33 e 3 t 1 + i = 1 p 1 δ 41 , i Δ ln ( R G D P P C t i ) + i = 1 p 1 δ 42 , i Δ ln ( E C P C t i ) + i = 2 p 1 δ 43 , i I N V O L t i + i = 2 p 1 δ 44 , i E V O L t i + ε 4 t
The parameter γi1 denotes the speed of the adjustment parameter for the i-th equation. Each equation in the system described above can also be viewed as an autoregressive distributed lag (ARDL) model.
The long-run and short-run Granger causality between the variables of interest are examined through the Yamamoto and Kurozumi tests. (see Appendix B for the methodological note). The technical econometrics components are available in the Supplementary Materials. Furthermore, we also incorporate the sign-rule to non-spurious long-run causal relationships which are invariant to sampling frequencies. The sign rule demonstrates that the established long-run Granger causality is non-spurious if the sign of the error correction coefficient, γi, is opposite to that of the sign of βi in the cointegrating vector. On the other hand, the short-run Granger causality will be based on testing the restrictions on δij in the short-run Equation (3a–d). For example, the short-run Granger causality from growth in income to growth in energy consumption is examined by testing the null hypothesis that δ 21 , i = 0 , for all i. Furthermore, the sign of the short-run Granger causality from income growth to energy consumption growth is determined by i = 1 p 1 δ 21 , i . Similarly, the short-run Granger causality from income growth volatility to energy consumption growth is determined by testing the null hypothesis that γ 22 = 0   and   δ 23 , i = 0 , for all i. Correspondingly, the sign is determined by γ 22 + i = 2 p 1 δ 23 , i . All estimation and testing except the long-run Granger causality testing using Yamamoto and Kurozumi testing are carried out in Eviews 11. The Yamamoto and Kurozumi testing long-run Granger causality testing is implemented in Gauss 18.

4. Results and Discussion

4.1. Descriptive Evidence

The energy–growth nexus is complex in many dimensions of the relationship. Although we observed that the income per capita has increased for all countries, Figure 3a shows that the per-capita energy consumption has increased in some countries and decreased in other countries. In particular, we observed a positive relationship between the real GDP per capita and per-capita energy consumption in countries with a relatively lower energy consumption level. On the other hand, countries with a higher energy consumption level tend to display a negative relationship between these two variables. More interestingly, we observed that the energy-intensity is falling in both groups of countries, motivating us to estimate the correct magnitude of the relationship between per-capita energy use and economic growth. Furthermore, Figure 3b depicts the contemporaneous negative relationship between the energy intensity and income per capita across all countries. However, the direction of the contemporaneous (long-run) relationship is not apparent from these descriptive measures. This postulates the basis of our further analysis on the energy-growth nexus through the lens of long-run analysis and the variety of the short-run analyses.
The uncertainty (volatility) in income growth was computed through the EGARCH model for all countries. Figure 4 depicts the relationship between income growth volatility and the growth in energy consumption. We observed that the uncertainty in income growth (higher volatility) reduces the energy consumption. All of the above provide important insights into the energy-growth nexus, which is further explored in the following sections.

4.2. Unit Roots and Cointegration

The general findings from the unit root test (Appendix A Table A1, Table A2, Table A3 and Table A4) demonstrate that the real GDP per capita (RGDPPC) and energy consumption per capita (ECPC) are I(1) and the volatility measures of both income growth (INVOL) and energy consumption growth (EVOL) are stationary, I(0). The summary of the cointegration test results is presented in Table 1 (see Appendix A Table A5 for the detailed cointegration test results). The results suggest that there was no long-run equilibrium relationship between the per-capita income, the per-capita energy consumption, the volatility of income growth, and the volatility of energy consumption growth in 17 countries. On the other hand, a long-run equilibrium relationship between per-capita income and per-capita energy consumption was established in 31 countries.

4.3. Long-Run Causal Inferences

Equation (3a,b) can be used to gauge the long-run causal relationships between per-capita real GDP (income) and per-capita energy consumption. Columns 2 and 3 of Table 2 present the coefficient β from the long-run Equation (2) and the corresponding standard errors, respectively. The magnitude of β helps to determine if the energy intensity is falling with respect to income. The speed of adjustment coefficients for the per-capita energy consumption equation (γ21) and per-capita real GDP equation 11) are reported in columns 4 and 6 of Table 2, respectively. The Yamamoto–Kurozumi test statistic, based on the Chi-squared distribution, used to examine the long-run causal relationship between income and energy consumption is reported in column 8 and its reverse causality is reported in column 9 of Table 2. Non-spurious causal inferences require the speed of adjustment for the energy equation (γi1) to be negative and the speed of adjustment for the income equation (γi2) to be the same sign as β (or the opposite sign to −β in the cointegrating vector) (Rajaguru and Abeysinghe 2008). The signs of γ11 and γ21 were found (as was to be expected) for all countries except for the case of Egypt (the sign of γ21 was positive (0.11) where it should have been negative). However, the test statistic for the long-run causality from income to energy consumption is insignificant with the Yamamoto–Kurozumi test statistic of 0.03. The wrong sign will not affect the general conclusion of Granger non-causality from income to energy consumption in Egypt. These findings based on the Yamamoto–Kurozumi test and the sign rule suggest that all the long-run causal inferences were non-spurious. Finally, the sign of β and the corresponding speed of adjustment coefficients, along with the statistical significance as assessed by the Yamamoto–Kurozumi test, determines the nature of the causal relationship (positive or negative or none). These are summarised in last two columns of Table 2. For the ease of interpretation, the results of the Table 2 are summarised in Table 3. The summary results demonstrate the existence of long-run causality between real GDP per capita and energy consumption per capita in 31 countries.
Overall, the results validate different causal relationships for different countries. The results show negative unidirectional long-run causality from income to energy consumption in Belgium, Chile, Denmark, Finland, Germany, Italy, Japan, the Netherlands, Sweden, Thailand, and the USA. This implies that increasing income decreases energy consumption and hence energy intensity has been significantly decreasing. In contrast with the above results, Cheng observed evidence for the causality from energy consumption to economic growth for Japan and many have observed mixed evidence in the case of the United States (see Kraft and Kraft 1978; Akarca and Long 1980; Yu and Hwang 1984; Abosedra and Baghestani 1989; Yu and Jin 1992; Stern 1993; Cheng 1995; Stern 2000; Soytas et al. 2007; Ewing et al. 2007; Payne 2009; Bowden and Payne 2009; Fallahi 2011). These results are according to the expectations that the magnitude of the energy–growth nexus has changed in favour of falling energy intensity in the past one and a half decades. On the other hand, a negative causality from energy consumption to growth also confirms the implied explanation of falling energy intensity in Egypt and Vietnam. Furthermore, a range of countries, including Australia, Canada, Iran, New Zealand, the UAE, and the UK, exhibit negative bi-directional causality, which generates the multiplier effect on energy consumption in the realm of falling energy intensity.
As confirmed through the Yamamoto and Kurozumi (2006) test and the sign rule developed by Rajaguru and Abeysinghe (2008), the overall evidence confirms the long-run energy–growth nexus. It indicates falling energy intensity in most sample countries. The long-run Equation (2) can be written as ln ( E C P P C t ) = β ln ( R G D P P C t ) + e 1 t . It can be re-rearranged to represent its results in energy intensity form as a function of real GDP per capita: ln ( E C P P C t / R G D P P C t ) = ( β 1 ) ln ( R G D P P C t ) + e 1 t . The energy intensity E C P P C t / R G D P P C t falls with respect to income when β < 1. Likewise, causality from consumption to income requires β > 1. Correspondingly, concerning unidirectional causality from income to energy consumption, evidence shows that β < 1, with significantly falling energy intensity in 15 countries. Furthermore, we can decompose the results into two broad categories 0 < β < 1 and β < 0. When the coefficient of income is between 0 and 1, regarding the conservation hypothesis, increasing income leads to a smaller increase in energy consumption. The above has been witnessed in Albania, Bangladesh, Brazil, and Columbia. Furthermore, coefficients of income with β < 0 have been observed in Belgium, Denmark, Chile, Finland, Italy, Germany, Japan, the Netherlands, Thailand, Sweden, and the USA. Therefore, the estimated coefficient of β ranging from 0 < β < 1 and β < 0 confirm the conservation with falling energy intensity in all corresponding countries. The above findings can be broadly reconciled with earlier studies reporting that income coefficients of energy consumption significantly decrease with the increasing development of corresponding economies (Joyeux and Ripple 2011; Medlock and Soligo 2001; Van Benthem and Romani 2009; Judson and Orphanides 1999).
The results confirm the causality from energy consumption to income in Austria, Egypt, India, South Korea, Sri Lanka, and Vietnam, where energy use leads per capita income. Egypt and Vietnam demonstrate that increasing energy consumption significantly decreases income—supporting conservation policies. Further results from Austria and South Korea show that a small increase in energy consumption increases income by many folds, strongly supporting the conjecture of falling energy intensity. However, results from countries such as India and Sri Lanka indicate that increasing energy consumption increases income less proportionately—this is mainly because of energy-intensive methods of production with high energy leakages. Furthermore, we found evidence supporting the feedback causal relationship between income and energy consumption in Australia, Algeria, China, Canada, France, New Zealand, Iran, Turkey, the UK, and the UAE. All the above countries have an estimated β of income to energy consumption that is less than one, implying a significant decrease in energy intensity in those countries. This is consistent with the results described in Wang et al. (2011a) and Zhixin and Xin (2011) for China and Erdal et al. (2008) and Acaravci (2010) for Turkey. The remaining 17 countries show no evidence of the relationship between energy consumption and economic growth.

4.4. Short-Run Causal Inferences

We analysed the short-run dynamic relationship among variables with a particular focus on energy consumption and income volatilities. The aspect of energy consumption and income volatility has been grossly ignored in the energy–growth literature. A summary of the short-run causality test results is presented in Table 4. The relevant test statistics and the corresponding levels of significance are presented in Appendix A Table A6.
The results show 12 countries with short-run unidirectional causality from real GDP per-capita growth to energy consumption growth, nine countries with reverse causality, and another nine countries with bi-directional causality. The results show that increasing energy consumption increases income in China, France, Italy, the Philippines, South Africa, South Korea, the UK, the USA, and Vietnam. The remaining 18 countries show no short-run relationship. In summary, we found 44 countries with either long-run (short-run) causality in at least one direction. However, the remaining four countries (the Czech Republic, Nigeria, Norway, and Trinidad and Tobago) showed no evidence causal relationship between income and energy consumption in both the short-run and long-run analysis.
The uncertainty (volatility) in income growth is expected to reduce energy consumption. The results show a significant decrease in the energy use in response to increasing the income growth volatilities in 28 countries, including Albania, Algeria, Belgium, Bolivia, Brazil, Canada, Chile, Colombia, Denmark, Ecuador, France, Gabon, Germany, Hungary, Indonesia, Iran, Japan, Korea, Netherlands, Nigeria, the Philippines, Portugal, South Africa, Spain, Syria, the UAE, the UK, and the USA. These results indicate that uncertainty (volatility) in income growth discourages expenditure by lowering energy consumption. The results for Canada, Chile, the Czech Republic, Indonesia, Iran, New Zealand, Portugal, Spain, the UK, and Vietnam show evidence in support of Bernanke (1983), Pindyck (1991) and Hnatkovska and Loayza (2005). They indicate that growth is negatively influenced by income volatility. We also found evidence supporting the claims of Black (1987), indicating that growth is positively related to income volatility. In general, increases in income tend to lower the income volatilities in many countries, rather than increasing income volatility.
On the other hand, the results also indicate that the volatilities in energy use growth are significantly related to income changes in China, Italy, Thailand, and Turkey. Furthermore, the results suggest that income growth volatilities increase volatilities in energy consumption growth in countries such as Bolivia, the Czech Republic, Denmark, Ecuador, Germany, India, New Zealand, Nigeria, Norway, Portugal, Thailand, and the USA. On the other hand, we found that the volatilities of energy consumption growth increased volatilities in income in Albania, Austria, China, Hungary, Indonesia, Japan, and South Africa. Interestingly, feedback effects of income and energy consumption growth volatilities were observed in Pakistan, Sudan, Syria, and the United Kingdom. The short-run negative effects of income volatility are usually countered by compensating people with subsidies to households and producers from many developing countries in order to smooth energy consumption behaviour.
We find the presence of long-run and/or short-run causality between growth and energy use in terms of energy intensity in a majority of the 48 countries included in the sample. There are only five countries exhibiting no causality between income and energy use. We observe 13 countries showing a valid causality running from energy consumption to economic growth and this has been confirmed through falling energy intensity. The above findings, in general, align with the sustainability argument that income growth does not require a proportional increase in energy consumption. Moreover, six countries observe causality from economic growth to energy use, concurrently with falling energy intensity. However, the falling energy intensity together with causality from economic growth to energy consumption indicate energy efficient methods of production, perhaps on the back of successful energy conservation policies. In the same vein, our results lend support lend support to the feedback relationships, in terms of falling energy intensity in 10 countries.

5. Concluding Remarks and Policy Implications

This study investigates energy-intensity through the energy–growth framework. We found that the historical evolution of energy intensity provides a new explanation for the energy–growth nexus, unlike the previous literature that merely relies on causality between the two. Furthermore, this study offers the unique feature of expressing the energy–growth nexus in terms of volatility analysis, which adds great value to the existing literature on this topic.
As confirmed through the Yamamoto–Kurozumi test, the evidence suggests the presence of long-run and/or short-run causality between growth and energy use in terms of energy intensity in 43 out of all the 48 countries included in the sample. The remaining five countries were confirmed to exhibit no causality between income and energy use. In particular, 13 countries appear to indicate a valid causal relationship from energy use to economic growth, with falling energy intensity captured through β < 1. The above results indirectly confirm the sustainability argument that income growth does not require a proportional increase in energy consumption in those countries. In addition, six countries showed the presence of causality from economic growth to energy use, with falling energy intensity. The above findings indicate improved methods of production, successful energy conservation policies, or switching to alternative energy sources. In Austria and South Korea the intensity coefficient appeared to be β > 1. The results also lend support to the feedback causal relationships, in terms of falling energy intensity in 10 countries. Policymakers and applied researchers may benefit from the current findings in relation to energy conservation efforts to overcome the serious environmental issues that the world is passing through. There are only five countries exhibiting no causality between income and energy use. It is possible that these countries may switch between the positive causal relationship and negative causal relationship in the sub-sample periods. These effects are likely to be nullified (i.e., no causality between the energy consumption and income) when the model is estimated for the full sample. Future research could extend this framework to the regime-switching approach to analyse the time-varying nature of the causal inferences. It will help policy makers to align themselves and so maintain the optimal energy mix to achieve the sustainable economic growth. Countries with a unidirectional causality from income to energy consumption may contribute to the fight against global warming directly implementing energy conservation measures. On the other hand, countries with a unidirectional causality from energy consumption to income may focus on technological developments and mitigation policies. However, for the countries with the feedback causal relationships, a balanced combination of alternative policies seems to be appropriate (see Soytas and Sari 2006).
Overall, energy efficiency policies are becoming progressively more important all over the world. Our results reveal that (for the majority of countries) energy consumption decreases or moderately increases along with increasing income; hence, energy efficiency has increased over the period of study. Thus, consuming less energy reduces the demand for available energy resources, such as fossil fuels. Energy efficiency programmes, however, ought to be created based on the promotion of suitable schemes. For instance, policymakers should promote incentive pricing for a successful energy efficiency plan. In the same vein, packages of measures including financial incentives and regulations should be implemented simultaneously, rather than one after the other. Moreover, public–private partnerships should be promoted in order to achieve set targets of energy efficiency. A favourable and stable institutional framework is also needed in order to ensure policy continuity with quantitative energy efficiency targets. Furthermore, in order to be efficient, energy efficiency strategies should be supervised regularly and adequately enforced. Certification and testing facilities should be available to promote innovations in energy-related tools and appliances. Innovative measures should be promoted in disadvantaged countries based on the experiences of developed countries. Finally, the successful implementation of energy policies reduces energy volatilities, which are strongly linked with income volatilities and vice versa.
Regarding short-run causality, the results also confirm falling energy intensity, along with the causal relationship between economic growth and energy use. The results indicate a significant decrease in energy use in response to increasing income growth volatility. Furthermore, the results suggest that income volatilities increase volatilities in energy consumption in a few countries. This occurs because the short-run negative effects of income volatility are usually countered by the provision of compensation with subsidies to households and producers in many developing countries to smooth energy consumption behaviour.
The country-specific long-run and short-run causal inferences are found to be not specific to any cluster of countries. In other words, the causal inferences are not robust to any regional countries. The results could be misleading if we assume homogeneity across the countries or regions if the model is estimated in a panel framework. Furthermore, it reconfirms the argument of Menegaki and Tsani (2018) that country-specific estimates are mostly different from the results of pooled (panel) regression models, such as those discussed in the main study by Apergis and Payne (2012). The policy recommendations should be made at the country level rather than the regional level.
In this study, we assumed that the volatilities of economic growth and energy consumption growth had symmetrical effects on income and energy consumption. The causal effects from volatility to energy consumption could be asymmetrical for both positive and negative deviations. Future studies could incorporate such time-varying asymmetrical non-linear causal relationships into their analysis. The study could also be extended to analyse the role of renewable energy and related policies on sustainable economic growth.

Author Contributions

Conceptualisation, S.U.K. and G.R.; methodology, G.R.; software, G.R.; validation, S.U.K. and G.R.; formal analysis, S.U.K. and G.R.; investigation, S.U.K. and G.R.; resources, S.U.K. and G.R.; data curation, S.U.K. and G.R.; writing—original draft preparation, S.U.K. and G.R.; writing—review and editing, S.U.K. and G.R.; visualisation, G.R.; supervision, G.R.; project administration, G.R. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

We used the World Development Indicators (2017) as the primary source of data.

Acknowledgments

We acknowledge feedback and comments offered by the audience of a seminar at the Bond Business School, Econometrics Society Australasian Meeting, and the Australian Conference of Economists 2019. We have also benefited from the comments and suggestions received from Robert Wrathall, Bond Business School, Australia. The views expressed here are those of authors and not their affiliated institutions.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table A1. Unit root test—real GDP per capita (RGDPPC).
Table A1. Unit root test—real GDP per capita (RGDPPC).
CountriesLevelFirst Difference
ADFPPKPSSBreakpointBreakADFPPKPSSBreakpointBreak
Albania−2.36−1.420.16 **−3.792003−3.79 ***−3.03 ***0.32−5.29 **1990
Algeria−2.63−1.660.15 **−3.021990−2.98 **−4.52 ***0.15−4.64 **1979
Australia−1.74−1.990.19 **−3.552000−5.35 ***−5.35 ***0.11−6.33 ***2001
Austria−2.11−1.440.18 **−2.371981, 1986−3.02 **−3.17 **0.19−8.97 ***1985
Bangladesh−2.47−2.040.14 *−3.722003−4.59 ***−4.59 ***0.08−4.81 **2002
Belgium−1.01−0.890.15 **−3.631980−3.05 **−3.08 **0.26−6.94 ***2000
Bolivia−2.16−2.270.17 **−2.021999−4.29 ***−4.35 ***0.19−4.89 **2003
Brazil−2.53−2.130.14 *−3.881986−4.25 ***−4.24 ***0.09−4.77 **1985
Canada−2.35−2.270.17 **−3.351992−4.65 ***−4.66 ***0.08−5.69 ***1975
Chile−2.32−2.570.18 **−3.321997−6.95 ***−6.95 ***0.07−16.87 ***1998
China−2.34−2.250.18 **−3.011986−6.33 ***−6.34 ***0.19−9.92 ***1992
Colombia−1.68−1.650.17 **−3.111992−4.54 ***−4.56 ***0.18−5.89 ***1993
The Czech Republic−2.08−1.360.15 **−2.711985−4.53 ***−4.53 ***0.23−5.06 ***2012
Denmark−2.52−2.340.14 *−2.681980−4.79 ***−4.69 ***0.13−5.37 ***1985
Ecuador−2.39−1.740.12 *−3.261987−2.91 *−3.13 **0.16−4.31 *1979
Egypt−1.71−1.820.16 **−3.731985−4.25 ***−4.36 ***0.16−4.94 **1994
Finland−1.44−1.480.14 *−2.931982−4.31 ***−4.44 ***0.19−6.34 ***1986
France−2.66−2.270.17 **−2.222004−4.66 ***−4.66 ***0.06−5.25 ***1986
Gabon−1.8−1.80.22 ***−2.921986−5.95 ***−6.11 ***0.29−7.11 ***1983
Germany−1.93−1.210.21 ***−3.051985−4.92 ***−4.84 ***0.31−5.55 ***1981
Hungary−2.99−2.060.12 *−3.431985−4.47 ***−4.37 ***0.14−4.92 **1992
India−0.58−0.260.22 **−3.941974−4.91 ***−4.61 ***0.13−7.43 ***1975
Indonesia−1.93−1.370.12 *−4.111987−2.92 *−2.97 **0.21−4.29 *1994
Iran−1.88−0.820.23 ***−2.881978−3.26 **−2.64 *0.31−5.23 ***1978
Italy−0.89−1.050.16 **−1.042001−2.86 *−5.25 ***0.3−5.89 ***1996
Japan−1.75−3.33 *0.15 **−2.922004−6.64 ***−6.37 ***0.25−7.55 ***2008
Korea South−3.12−2.170.14 *−4.121985−4.81 ***−4.71 ***0.31−5.42 ***1981
The Netherlands−3.38 *−2.580.15 **−3.711985−5.00 ***−4.86 ***0.08−5.53 ***1981
New Zealand−3.11−2.580.12 *−3.582000−4.22 ***−4.13 ***0.06−4.62 ***1999
Nigeria−3.11−2.320.16 **−3.891984−4.81 ***−4.54 ***0.08−5.88 ***1993
Norway−3.11−2.510.15 **−3.571985−4.97 ***−4.92 ***0.09−5.34 ****1981
Pakistan−2.26−2.070.13 *−2.92 *−2.84 *0.28
Philippines−0.78−0.990.16 **−2.771988−5.89 ***−6.03 ***0.24−7.46 ***1991
Portugal−3.03−2.50.18 **−3.721980−4.95 ***−4.73 ***0.06−6.29 ***1986
South Africa−1.38−0.470.24 ***−3.421985−5.03 ***−4.83 ***0.33−5.77 ***1995
Spain−5.61 ***−5.52 ***0.05−4.69 **1986
Sri Lanka−0.64−0.710.19 **−2.561978−6.81 ***−6.89 ***0.32−11.02 ***1978
Sudan−2.51−1.810.15 **−3.711985−4.94 ***−4.87 ***0.18−5.54 ***1981
Sweden−2.98−2.140.19 **−3.922000−4.28 ***−4.32 ***0.24−4.84 **1984
Syrian−3.06−3.050.13 **−1.711982−5.27 ***−5.11 ***0.07−8.53 ***1973
Thailand−1.39−1.090.17 **−2.791981−4.49 ***−4.52 ***0.32−5.94 ***1984
Trinidad and Tobago−2.29−1.950.12 *−3.341979−3.74 ***−3.74 ***0.12−4.74 **1980
Turkey−3.05−2.580.13 *−3.382001−5.26 ***−5.08 ***0.05−5.69 ***1992
UAE−2.79−2.380.12 *−2.451986−4.91 ***−4.95 ***0.067.46 ***1998
UK−2.34−1.660.13 *−3.171979−2.82 *−3.79 ***0.195.23 ***2001
USA−3.01−2.620.15 **−3.561992−4.29 ***−3.74 ***0.07−6.29 ***2008
Venezuela−2.19−1.240.21 **−3.811982−5.03 ***−4.80 ***0.27−5.73 ***2009
Vietnam−2.09−2.30.15 **−2.66 *−2.72 *0.31
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively. ADF—Augmented Dickey–Fuller Test, PP—Phillips–Perron Test, and KPSS—Kwiatkowski–Phillips–Schmidt–Shin Test.
Table A2. Unit root rest—real energy consumption per capita.
Table A2. Unit root rest—real energy consumption per capita.
CountriesLevelFirst Difference
ADFPPKPSSBreakpointBreakADFPPKPSSBreakpointBreak
Albania−1.27−1.500.15 **−2.291989−6.18 ***−6.20 ***0.14−7.65 ***1992
Algeria−2.94−2.920.16 **−4.082002−4.29 ***−5.07 ***0.25−7.55 ***1982
Australia−1.43−1.290.21 **−4.121993−7.65 ***−7.61 ***0.31−8.67 ***2007
Austria−1.83−1.830.21 **−3.061,995−6.72 ***−6.74 ***0.31−8.11 ***1972
Bangladesh−0.93−0.620.20 **−0.692000−8.18 ***−8.18 ***0.28−9.16 ***2001
Belgium−1.61−1.620.19 **−3.532012−6.61 ***−6.61 ***0.32−7.29 ***1972
Bolivia−2.74−2.820.16 **−2.611993−7.74 ***−7.63 ***0.11−8.36 ***2001
Brazil−1.61−2.060.15 **−2.542003−5.59 ***−5.59 ***0.13−6.51 ***1981
Canada−2.13−2.110.22 ***−4.77 ***−4.57 ***0.35 *
Chile−3.02−2.660.15 **−3.341987−4.52 ***−4.48 ***0.27−5.58 ***1975
China−1.35−0.880.19 **−3.282002−3.57 **−3.57 **0.31−4.49 **2001
Colombia−1.76−1.860.17 **−1.971984−7.17 ***−7.14 ***0.09−8.18 ***1999
The Czech Republic−2.17−2.270.18 **−3.451990−7.12 ***−7.12 ***0.08−7.45 ***1999
Denmark−2.96−2.970.20 **−3.572009−7.18 ***−7.14 ***0.09−8.18 ***1999
Ecuador−2.78−2.720.15 **−3.071995−7.32 ***−7.45 ***0.15−7.81 ***1985
Egypt−1.09−1.180.15 **−2.732001−5.58 ***−5.17 ***0.37 *−6.42 ***1985
Finland−1.35−1.070.23 ***−7.27 ***−7.33 ***0.28
France−1.46−1.450.23 ***−6.15 ***−6.14 ***0.29
Gabon−1.07−1.080.19 **−2.232001−5.99 ***−6.00 ***0.21−6.61 ***1976
Germany−1.96−1.960.23 ***−3.282008−5.86 ***−5.82 ***0.27−11.34 ***1972
Hungary−1.97−1.970.21 **−3.562008−4.68 ***−4.66 ***0.18−6.83 ***1973
India−0.21−0.380.18 **−0.492005−4.81 ***−5.04 ***0.17−7.19 ***2003
Indonesia−1.24−1.240.15 **−3.161999−6.58 ***−6.58 ***0.19−8.50 ***1990
Iran−3.33−3.250.16 **−3.591988−8.34 ***−8.16 ***0.15−10.28 ***1977
Italy−3.11−3.120.20 **−2.261995−6.29 ***−6.35 ***0.27−7.99 ***2007
Japan−2.78−2.520.21 **−2.612008−5.85 ***−5.85 ***0.32−7.36 ***1972
Korea South−0.19−0.170.21 **−2.651985−5.34 ***−5.42 ***0.287.64 ***1998
The Netherlands−2.29−2.270.19 **−5.83 ***−5.84 ***0.32
New Zealand−2.06−1.980.24 ***−3.591983−7.37 ***−7.37 ***0.33−8.368 ***1979
Nigeria−2.63−2.460.18 **−3.041998−5.51 ***−5.44 ***0.24−6.42 ***1994
Norway−1.97−1.730.25 ***−9.44 ***−9.91 ***0.36 *
Pakistan−1.87−1.880.18 **−3.211986−5.16 ***−5.18 ***0.33−7.33 ***2007
Philippines−2.49−2.530.15 **−3.111985−8.63 ***−8.33 ***0.09−8.97 ***2009
Portugal−0.16−0.160.22 ***−5.41 ***−5.44 ***0.26
South Africa−1.98−1.950.16 **−2.612002−6.22 ***−6.23 ***0.16−6.87 ***2007
Spain−0.75−0.910.22 ***−4.31 ***−4.39 ***0.26
Sri Lanka−2.28−2.080.18 **−2.991994−7.32 ***−7.45 ***0.21−8.18 ***1996
Sudan−2.97−2.850.17 **−3.241985−7.01 ***−11.08 ***0.29−10.18 ***2002
Sweden−2.13−2.130.25 ***−3.452008−8.37 ***−8.45 ***0.31−9.15 ***1985
Syrian−0.26−0.510.21 **−1.982004−5.49 ***−5.55 ***0.35 *−6.75 ***2005
Thailand−1.91−2.010.19 **−3.241986−4.88 ***4.98 ***0.09−6.21 ***1983
Trinidad and Tobago−2.19−2.210.19 **−2.351995−3.47 **−6.33 ***0.15−7.38 ***1978
Turkey−2.52−2.570.17 **−1.792002−6.42 ***−6.42 ***0.09−7.35 ***1999
UAE−0.89−0.830.21 **−2.632003−6.39 ***−6.42 ***0.19−7.99 ***1987
UK−0.54−0.350.19 **−2.922008−7.27 ***−7.24 ***0.13−8.57 ***2005
USA−2.62−1.920.19 **−4.072008−5.05 ***−4.94 ***0.29−5.56 ***1978
Venezuela−2.92−2.970.15 **−3.572004−10.98 ***−10.46 ***0.31−13.16 ***1991
Vietnam−1.74−1.750.22 ***−1.961996−5.32 ***−5.51 ***0.327.43 ***1992
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively. ADF—Augmented Dickey–Fuller Test, PP—Phillips–Perron Test, and KPSS—Kwiatkowski–Phillips–Schmidt–Shin Test.
Table A3. Unit root test—income volatility.
Table A3. Unit root test—income volatility.
CountriesModelADFPPKPSSBreakpointBreak
AlbaniaEGARCH(1,1)−3.77 **−3.57 **0.07−7.64 ***2002
AlgeriaEGARCH(1,1)−4.51 ***−5.80 ***0.07−5.20 ***1992
AustraliaEGARCH(3,0)−6.74 ***−8.08 ***0.19−12.26 ***2010
AustriaEGARCH(1,0)−5.14 ***−5.15 ***0.1−23.59 ***1985, 1990
BangladeshEGARCH(1,2)−7.47 ***−5.56 ***0.08−9.71 ***2001
BelgiumEGARCH(1,0)−4.59 ***−4.65 ***0.06−6.01 ***2004
BoliviaEGARCH(1,0)−4.81 ***−4.48 ***0.1−5.09 ***1976
BrazilEGARCH(2,1)−3.11 **−4.04 ***0.1−4.45 **1994
CanadaEGARCH(1.0)−5.37 ***−5.22 ***0.15−10.46 ***1995
ChileEGARCH(2,2)−2.97 **−2.92 **0.11−5.06 ***2001
ChinaEGARCH(2,2)−5.71 ***−5.53 ***0.14−13.30 ***1992, 1999
ColombiaEGARCH(1,0)−5.76 ***−5.77 ***0.17−14.09 ***1994, 1997
The Czech RepublicEGARCH(1,2)−3.72 ***−3.81 ***0.22
DenmarkEGARCH(2,2)−6.74 ***−8.11 ***0.12−8.44 ***2004
EcuadorEGARCH(2,2)−4.92 ***−5.42 ***0.21−4.64 **1982, 1990
EgyptEGARCH(3,0)−5.23 ***−5.13 ***0.09−7.07 ***1996
FinlandEGARCH(1,1)−8.84 ***−8.77 ***0.17−10.96 ***1985, 1991
FranceEGARCH(1,0)−6.67 ***−6.66 ***0.31−8.92 ***1991, 2010
GabonEGARCH(1,1)−12.53 ***−11.76 ***0.08−14.89 ***1987, 1997, 2003
GermanyEGARCH(2,2)−4.09 ***−3.17 **0.21−4.49 **1994
HungaryEGARCH(2,2)10.31 ***−9.74 ***0.35 *−14.61 ***1978
IndiaEGARCH(1,2)−6.56 ***−7.92 ***0.28−6.95 ***1975, 1981
IndonesiaEGARCH(1,1)−2.93 **3.87 ***0.14−5.45 ***1983, 1992
IranEGARCH(1,1)−3.04 **−3.19 **0.23−11.73 ***1975, 1978
ItalyEGARCH(1,0)−5.83 ***−5.72 ***0.28−6.29 ***1988
JapanEGARCH(1,1)−6.73 ***−6.38 ***0.17.26 ***2008
Korea SouthEGARCH(1,0)−6.62 ***−6.62 ***0.13−6.91 ***1987, 2010
The NetherlandsEGARCH(2,0)−7.84 ***11.11 ***0.14−11.09 ***1989, 1992
New ZealandEGARCH(2,2)3.02 **−3.07 **0.14−5.67 ***1981
NigeriaEGARCH(2,2)−3.99 ***−10.23 ***0.11−11.53 ***1980
NorwayEGARCH(1,3)−7.79 ***−8.88 ***0.22
PakistanEGARCH(1,0)−4.39 ***−4.38 ***0.11−4.78 ***2004
PhilippinesEGARCH(1,0)−6.22 ***−6.36 ***0.16−7.01 ***1986, 2004
PortugalEGARCH(2,2)−11.18 ***−10.89 ***0.08−13.18 ***1979
South AfricaEGARCH(1,1)−3.09 **−3.14 **0.34−4.51 **2002
SpainEGARCH(2,2)−8.99 ***−5.92 ***0.28−10.33 ***1999
Sri LankaEGARCH(2,2)−7.81 ***−7.19 ***0.18−14.58 ***1978
SudanEGARCH(3,0)−7.92 ***−10.49 ***0.24−9.47 ***2000
SwedenEGARCH(1,0)−6.07 ***−6.08 ***0.19−6.91 ***2004
SyrianEGARCH(1,0)−6.55 ***−6.54 ***0.14−33.61 ***1975
ThailandEGARCH(1,0)−7.55 ***−7.61 ***0.19−8.28 ***1985
Trinidad and TobagoEGARCH(1,1)−2.65 *−2.92 ***0.18−4.96 ***1982
TurkeyEGARCH(2,0)−3.75 ***−3.75 ***0.26−4.73 ***1978
UAEEGARCH(1,0)−4.84 ***−4.80 ***0.21−12.85 ***1999
UKEGARCH(1,0)−5.94 ***−5.95 ***0.12−6.35 ***1995
USAEGARCH(2,1)−3.11 **−3.22 **0.41 *−5.27 ***1975
VenezuelaEGARCH(2,0)−4.73 ***−3.24 **0.09−5.34 ***1983
VietnamEGARCH(2,2)−14.41 ***−13.36 ***0.34−54.22 ***1997
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively. ADF—Augmented Dickey–Fuller Test, PP—Phillips–Perron Test, and KPSS—Kwiatkowski–Phillips–Schmidt–Shin Test.
Table A4. Unit root test—energy consumption volatility.
Table A4. Unit root test—energy consumption volatility.
CountriesModelADFPPKPSSBreakpointBreak
AlbaniaEGARCH(1,1)−8.26 ***−8.43 ***0.21−14.68 ***1992
AlgeriaEGARCH(1,1)−4.08 ***−2.53 **0.317.56 ***1995
AustraliaEGARCH(1,0)−6.99 ***−6.98 ***0.08−7.57 ***2011
AustriaEGARCH(1,1)−3.60 **−3.55 **0.23−5.07 ***2004
BangladeshEGARCH(1,1)−3.98 ***−3.98 ***0.09−5.02 ***2011
BelgiumEGARCH(1,1)−6.09 ***−6.29 ***0.24−6.99 ***1982
BoliviaEGARCH(1,0)−6.69 ***−8.78 ***0.277.07 ***1999
BrazilEGARCH(3,0)−3.91 ***−3.82 ***0.15−5.83 ***1982
CanadaEGARCH(1,1)−3.65 ***−3.69 ***0.21−5.13 ***1981
ChileEGARCH(1,1)−9.72 ***−9.75 ***0.13−11.29 ***2012
ChinaEGARCH(1,0)−5.99 ***−5.85 ***0.18−7.21 ***2004
ColombiaEGARCH(4,0)−6.82 ***−6.97 ***0.11−15.35 ***1980
The Czech RepublicEGARCH(2,2)−9.99 ***−12.01 ***0.11−11.97 ***1993
DenmarkEGARCH(1,1)−4.76 ***−4.79 ***0.19−5.74 ***1996
EcuadorEGARCH(2,2)−7.47 ***−7.46 ***0.17−10.01 ***2000
EgyptEGARCH(2,2)−10.86 ***10.24 ***0.25−11.91 ***1985
FinlandEGARCH(2,1)−10.15 ***−10.16 ***0.09−10.21 ***1987
FranceEGARCH(1,1)−4.28 ***−4.31 ***0.16−5.68 ***1978
GabonEGARCH(2,1)−7.88 ***−8.08 ***0.19−14.26 ***1975
GermanyEGARCH(3,0)−6.64 ***−6.64 ***0.29−46.67 ***1974
HungaryEGARCH(2,0)−4.46 ***−4.47 ***0.12−5.16 ***1993
IndiaEGARCH(2,2)−6.62 ***−8.05 ***0.15−8.11 ***2010
IndonesiaEGARCH(1,1)−3.22 **−3.15 **0.28−5.63 ***1993
IranEGARCH(1,0)−6.61 ***−6.61 ***0.26−11.24 ***1984
ItalyEGARCH(1,1)−2.68 *−2.69 *0.21−4.92 ***2009
JapanEGARCH(1,1)−7.25 ***−7.34 ***0.327.61 ***2008
Korea SouthEGARCH(1,1)−3.59 ***−3.56 **0.114.96 **1998
The NetherlandsEGARCH(2,2)−6.09 ***−6.09 ***0.18−6.80 ***2011
New ZealandEGARCH(2,2)−7.18 ***−7.64 ***0.32−7.61 ***1981
NigeriaEGARCH(1,2)−7.21 ***−7.18 ***0.28−8.29 ***2010
NorwayEGARCH(2,2)−6.20 ***−9.74 ***0.15−7.21 ***1999
PakistanEGARCH(1,0)−6.22 ***6.23 ***0.29−7.45 ***2009
PhilippinesEGARCH(1,0)−11.42 ***−11.21 ***0.3−13.10 ***1984
PortugalEGARCH(1,1)−3.84 ***−3.84 ***0.11−4.57 **1989
South AfricaEGARCH(1,1)−3.31 **−3.23 **0.14
SpainEGARCH(1,2)−7.46 ***−7.42 ***0.24−12.17 ***1974
Sri LankaEGARCH(1,2)−7.49 ***−7.43 ***0.18−8.87 ***1997
SudanEGARCH(2,1)−7.63 ***−7.78 ***0.18−9.44 ***2000
SwedenEGARCH(1,1)−9.61 ***−10.64 ***0.25−10.21 ***2010
SyrianEGARCH(1,0)−5.16 ***−5.18 ***0.23−6.21 ***2011
ThailandEGARCH(2,2)−6.54 ***−6.54 ***0.09−8.14 ***1985
Trinidad and TobagoEGARCH(1,1)−15.05 ***−12.96 ***0.28−17.45 ***1999
TurkeyEGARCH(1,1)−7.33 ***−14.77 ***0.21−7.92 ***2006
UAEEGARCH(3,0)−6.43 ***−6.56 ***0.29−9.22 ***1999
UKEGARCH(1,2)−7.13 ***−7.12 ***0.12−7.65 ***2005
USAEGARCH(2,0)−7.36 ***−6.45 ***0.31−7.96 ***1994
VenezuelaEGARCH(2,0)−14.15 ***−14.52 ***0.18−14.94 ***2002
VietnamEGARCH(2,2)−8.37 ***−8.37 ***0.12−14.24 ***1974
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively. ADF—Augmented Dickey–Fuller Test, PP—Phillips–Perron Test, and KPSS—Kwiatkowski–Phillips–Schmidt–Shin Test.
Table A5. Cointegration test.
Table A5. Cointegration test.
CountriesTrace TestMax Eigen Value
r = 0r = 1r = 2r = 3r = 0r = 1r = 2r = 3
Albania103.58 ***41.51 ***12.53 **1.4662.08 ***28.97 ***11.22 **1.46
Algeria103.48 ***61.60 ***26.52 **0.1941.88 ***25.82 ***12.940.19
Australia67.73 ***37.05 **14.35 *1.4113.68 **22.69 **6.231.41
Austria182.32 ***27.38 **11.97 *0.16154.95 ***15.41 *11.80 **0.164
Bangladesh87.21 ***40.31 ***11.71 *0.9146.91 ***17.79 **11.22 *0.91
Belgium75.43 ***43.14 **22.914.1332.29 ***20.24 **18.77 ***4.13
Bolivia64.39 ***27.89 *6.710.6136.49 ***21.19 **6.10.61
Brazil52.38 ***28.89 **12.31 *2.4923.48 *16.58 *9.82 *2.49
Canada89.61 ***48.51 **24.96 *4.7141.09 ***23.55 **20.25 **4.71
Chile66.41 ***35.06 **14.40 *0.3931.35 **20.67 *14.01 *0.39
China117.78 ***34.91 **13.36 **0.3982.89 ***21.54 **12.97 **0.39
Colombia71.78 ***38.61 ***17.46 **2.6933.17 ***21.15 ***14.77 **2.69
The Czech Republic73.89 ***33.44 ***9.492.2440.46 ***23.95 ***7.252.24
Denmark59.77 ***23.51 *12.79 **1.0636.26 ***10.7111.73 **1.06
Ecuador 69.52 ***27.47 *6.520.9442.05 ***20.95 *5.580.94
Egypt74.63 ***41.39 ***19.97 ***1.0433.23 ***21.42 **18.93 ***1.04
Finland82.31 ***40.09 ***13.85 *1.9142.22 ***26.24 ***12.561.29
France58.54 ***33.34 ***11.41 *0.4325.21 **21.94 **11.22 *0.43
Gabon85.55 ***34.71 **8.412.1550.84 ***26.29 ***6.262.15
Germany89.73 ***43.08 ***18.26 **1.4246.65 ***24.81 **16.83 **1.42
Hungary60.64 ***29.63 *10.970.6631.01 **18.66 *10.310.66
India161.61 ***66.76 ***27.42 ***1.3794.84 ***39.34 ***26.06 ***1.37
Indonesia102.22 ***36.26 ***5.791.4865.86 ***30.57 ***4.311.48
Iran165.52 ***71.52 ***30.64 ***0.0393.99 ***40.87 ***30.62 ***0.03
Italy77.78 ***37.66 ***13.97 *1.740.11 ***23.69 **12.26 *1.7
Japan84.13 ***46.64 ***18.18 **2.0337.49 ***28.46 ***16.15 **2.03
Korea South115.84 ***69.29 ***26.67 ***6.5146.54 ***42.63 ***20.15 ***6.51
The Netherlands78.76 ***35.74 ***15.21 *2.4143.01 ***20.54 *12.79 *2.41
New Zealand86.91 ***49.65 ***20.38 **5.5337.26 ***29.27 ***14.82 *5.53
Nigeria75.61 ***37.85 ***9.020.0137.75 ***28.83 ***9.010.01
Norway78.37 ***40.93 **14.626.1237.43 ***26.31 **8.516.12
Pakistan54.76 ***24.72 **8.490.4930.04 ***16.23 *7.990.49
Philippines48.79 ***25.03 **4.450.0523.76 *20.58 **4.410.05
Portugal91.03 ***48.25 ***18.434.8942.77 ***29.83 ***13.524.89
South Africa103.99 ***44.73 **21.327.2659.27 ***23.41 *14.067.26
Spain--------
Sri Lanka77.29 ***40.47 ***17.23 **0.1436.81 ***23.24 **17.09 **0.14
Sudan80.24 ***43.67 ***16.785.8336.58 ***26.89 **11.055.73
Sweden74.98 ***36.27 ***16.18 **2.3138.10 ***20.09 *13.87 *2.31
Syrian142.56 ***32.12 **10.361.94110.45 ***21.76 **8.421.93
Thailand80.44 ***41.07 ***19.91 *3.1739.37 ***21.17 *16.74 **3.17
Trinidad and Tobago75.74 ***44.73 **15.414.2231.01 *29.32 **11.184.22
Turkey94.45 ***40.36 **18.91 *3.0254.09 ***21.54 *15.88 *3.02
UAE94.05 ***38.79 ***16.36 **2.4555.25 **22.44 **13.91 *2.45
UK89.98 ***36.93 ***14.06 *0.7853.05 ***22.88 **13.28 *0.78
USA79.53 ***34.72 **13.73 *0.0144.81 ***20.98 *13.72 *0.01
Venezuela75.08 ***43.58 ***16.874.1631.51 **26.71 **12.714.16
Vietnam180.79 ***71.44 ***35.20 ***0.49109.35 ***36.24 ***34.71 ***0.49
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively.
Table A6. Short-run Granger causality results (F Statistic).
Table A6. Short-run Granger causality results (F Statistic).
CountriesRGDPPC → ECPCINVOL
→ ECPC
EVOL
→ ECPC
ECPC →
RGDPPC
INVOL→ RGDPPCEVOL→ RGDPPCECPC → INVOLRGDPPC →INVOLEVOL→ INVOLECPC→ ECVOLRGDPPC → EVOLINVOL → EVOL
Albania18.32 ***19.15 ***3.025.23 *7.48 **7.78 **0.1928.18 ***5.95 **7.02 **13.77 ***2.79
Algeria1.995.71 **12.97 ***1.6418.42 ***4.74 *1.38116.27 ***0.5911.76 ***0.080.93
Australia0.470.640.090.470.590.780.0024.29 **0.00070.281.840.09
Austria0.10.020.781.212.21 ***6.11 **1.15382.67 ***8.16 **0.110.220.16
Bangladesh0.050.964.350.170.420.021.681.710.521.975.32 *1.78
Belgium6.03 **5.86 **0.150.981.721.060.395.41 *0.863.430.750.89
Bolivia10.85 ***5.81 *1.151.752.121.751.655.73 *1.861.170.249.51 ***
Brazil0.013.25 *9.28 ***0.320.140.750.1111.75 ***0.761.711.110.32
Canada9.29 **7.89 **3.2716.87 ***13.77 ***3.577.25 **20.27 ***3.944.226.07 *3.72
Chile0.373.87 *1.630.8516.75 ***12.17 ***0.8629.36 ***0.610.062.092.43
China0.670.770.260.1913.85 ***3.776.58 **3.78 *5.05 **36.06 ***0.060.45
Colombia0.364.48 *1.681.050.272.751.6253.96 ***1.533.320.591.07
The Czech Republic1.860.770.530.937.68 ***7.68 ***1.856.79 *2.7613.54 ***4.6511.72 ***
Denmark2.2314.62 ***4.894.223.217.133.047.65 *3.092.031.199.84 **
Ecuador6.95 **4.23 *6.91 **2.193.270.282.0747.46 ***2.666.68 **8.14 **9.19 **
Egypt2.720.744.824.433.344.420.113.190.433.242.013.16
Finland1.980.267.73 **0.594.371.351.0615.59 ***0.320.8212.78 ***2.99
France0.873.91 **0.054.73 **0.176.79 ***1.413.57 *0.060.330.0011.26
Gabon8.12 ***10.88 ***0.245.04 *0.292.010.4333.18 ***0.013.511.971.79
Germany11.11 **7.72 *29.11 ***4.095.034.094.964.134.257.121.058.86 *
Hungary7.85 *10.33 **5.2217.04 ***4.5913.46 **10.26 *5.5212.05 **3.952.544.36
India5.31 *2.623.31.4112.53 ***3.33.314.55 *2.412.089.09 ***6.69 **
Indonesia7.73 ***7.61 ***0.281.653.38 *0.993.55 **164.05 ***7.32 ***3.030.470.37
Iran2.97 *3.11 *0.0222.06 ***22.71 ***5.13 **0.5155.22 ***0.227.69 ***2.451.26
Italy0.0020.031.653.03 *0.013.74 *3.01 *24.15 ***0.511.682.540.01
Japan33.66 ***14.81 ***1.462.440.240.155.27 *2.268.27 **2.930.010.31
Korea, South3.188.15 *3.2919.87 ***13.34 **14.26 **3.023.93.333.970.923.15
The Netherlands6.59 **4.97 *5.99 **3.913.020.770.275.84 *1.038.38 **1.630.32
New Zealand0.240.010.010.0110.36 ***0.780.40.350.650.082.145.54 **
Nigeria0.417.26 **8.69 ***2.960.510.413.613.633.056.77 *9.16 ***7.01 **
Norway0.121.481.562.290.055.34 *0.810.393.685.28 *1.4917.56 ***
Pakistan4.60 *0.590.164.54 *0.010.010.080.093.08 *0.213.67 *6.24 **
Philippines0.054.21 **0.666.10 **2.72 *0.112.221.150.130.760.541.69
Portugal11.52 ***6.99 *9.59 **9.65 **10.63 **12.15 ***1.339.72 **3.6611.98 ***13.34 ***8.99 **
South Africa2.6410.14 **6.918.42 *1.872.272.050.388.37 *0.793.875.32
Spain11.88 **13.23 **2.478.61 *20.41 ***3.088.69 *19.87 ***4.036.273.553.17
Sri Lanka0.780.733.412.146.31 **2.290.015.41 *0.334.84 *0.340.63
Sudan3.48 *0.260.150.360.211.822.430.553.42 *3.78 *0.577.72 ***
Sweden3.71 *0.372.120.950.190.020.020.220.950.082.410.39
Syrian4.89 *4.89 *4.93 *2.710.165.43 *0.7270.84 ***4.92 *4.58 *4.91 *6.26 **
Thailand5.44 *1.182.821.411.523.955.44 *16.75 ***0.1910.04 ***3.695.21 *
Trinidad and Tobago1.660.213.85 **0.021.190.170.3662.80 ***0.380.640.650.27
Turkey1.490.912.91 *0.633.42 *0.229.21 ***13.52 ***0.4935.43 ***0.990.46
UAE0.726.48 *10.41 **3.911.763.496.51 *95.00 ***1.512.913.725.63
UK0.713.79 *0.022.99 *3.01 *1.151.721.7212.31 ***22.28 ***1.958.98 ***
USA2.736.02 *6.42 *8.15 **3.621.312.464.310.291.4910.45 **7.59 *
Venezuela11.43 ***3.732.4619.97 ***3.7521.35 ***1.8313.11 ***4.682.323.794.79
Vietnam0.870.090.387.52 **31.26 ***2.480.18229.89 ***0.4913.67 ***7.66 **2.46
Notes: *, ** and *** denotes significant at 10%, 5% and 1% respectively. EC—error correction term, ECPC—energy consumption per capita, RGDPPC—real gross domestic product per capita, INVOL—income growth volatility, and EVOL—energy consumption growth volatility.

Appendix B

In the following, we present Yamamoto and Kurozumi procedures in line with our proposed VECM. To determine the long-run Granger non-causality from the ith component of z t to the jth component of z t , we define two 1 × 4 matrices, R L = [ r 1 r 2 r 3 r 4 ] and R R * = [ r 1 * r 2 * r 3 * r 4 * ] , such that r k = { 1   if   k = j 0   otherwise and r k * = { 1   if   k = i 0   otherwise . For example, to test long-run Granger non-causality from ECPC to REGDPPC, corresponding restrictions may take the following form R L = [ 1 0 0 0 ] and R R * = [ 0 1 0 0 ] . Furthermore, long-run Granger non-causality from the ith component of z t to the jth component of z t is established by testing the null H 0 : R L B ¯ R R = 0 . We construct the Wald-type statistic using the generalised inverse W = T v e c ( R L B ¯ ^ R R ) ( R L C ^ Σ ^ C ^ R L R R P ^ Σ ^ P ^ R R ) g v e c ( R L B ¯ ^ R R ) d χ s 2 . where T is the sample size, vec denotes the vectorisation of a matrix by constructing a column vector by appending each column of a matrix, and Σ ^ is a consistent estimator of Σ , given by Σ ^ = T 1 i = 1 T ε ^ t ( β ^ ) ε ^ t ( β ^ ) , where ε ^ t ( β ^ ) = [ ( β ^ z t 1 ) , Δ z t 1 ] .
Also, B ¯ = β β M + β E 12 ( I E 22 ) 1 L G K 1 , where β is a 4 × 3 matrix such that β β = 0 , M = [ I 4 0 ] ,   E = G D G = [ I 3 β α β Γ 1 H 0 1 + β α β Γ 1 H ] = [ I 3 E 12 0 E 22 ] , G = I 2 H , D = [ I 4 + β α Γ 1 α β Γ 1 ] , H = [ β , β ] , L = [ 0 I 5 ] and K = [ I 4 0 I 4 I 4 ] .
Further, we define the long-run impact matrix C = β ( α Γ β ) 1 α , where α is a 4 × 3 matrix such that α α = 0 and Γ = ( I + Π 2 ) , P = K 1 G L ( I 5 E 22 ) 1 [ I 0 0 I H ] . Note that Q g denotes the generalised inverse of matrix Q and s = r a n k ( R L β ) × { r a n k ( R R * β ) + 1 } .

References

  1. Abosedra, Salah, and Hamid Baghestani. 1989. New evidence on the causal relationship between United States energy consumption and gross national product. The Journal of Energy and Development 14: 285–92. [Google Scholar]
  2. Acaravci, Ali. 2010. Structural breaks, electricity consumption and economic growth: Evidence from Turkey. Journal for Economic Forecasting 2: 140–54. [Google Scholar]
  3. Agovino, Massimiliano, Silvana Bartoletto, and Antonio Garofalo. 2018. Modelling the relationship between energy-intensity and GDP for European countries: An historical perspective (1800–2000). Energy Economics 82: 114–34. [Google Scholar] [CrossRef]
  4. Aizenman, Joshua, and Nancy P. Marion. 1993. Policy Uncertainty, Persistence and Growth. Review of International Economics 1: 145–63. [Google Scholar] [CrossRef] [Green Version]
  5. Akarca, Ali T., and Thomas Veach Long. 1980. On the relationship between energy and GNP: A re-examination. Journal of Energy Finance & Development 5: 326–31. [Google Scholar]
  6. Ang, James B. 2008. Economic development, pollutant emissions and energy consumption in Malaysia. Journal of Policy Modeling 30: 271–8. [Google Scholar] [CrossRef]
  7. Apergis, Nicholas, and James E. Payne. 2012. Renewable and non-renewable energy consumption-growth nexus: Evidence from a panel error correction model. Energy Economics 34: 733–38. [Google Scholar] [CrossRef]
  8. Aqeel, Anjum, and Mohammad Sabihuddin Butt. 2001. The relationship between energy consumption and economic growth in Pakistan. Asia-Pacific Development Journal 8: 101–10. [Google Scholar]
  9. Bartleet, Matthew, and Rukmani Gounder. 2010. Energy consumption and economic growth in New Zealand: Results of trivariate and multivariate models. Energy Policy 38: 3508–17. [Google Scholar] [CrossRef]
  10. Batrancea, Larissa. 2021. An Econometric Approach Regarding the Impact of Fiscal Pressure on Equilibrium: Evidence from Electricity, Gas and Oil Companies Listed on the New York Stock Exchange. Mathematics 9: 630. [Google Scholar] [CrossRef]
  11. Batrancea, Ioan, Larissa Batrancea, Malar Maran Rathnaswamy, Horia Tulai, Gheorghe Fatacean, and Mircea-Iosif Rus. 2020. Greening the Financial System in USA, Canada and Brazil: A Panel Data Analysis. Mathematics 8: 2217. [Google Scholar] [CrossRef]
  12. Belke, Ansgar, Frauke Dobnik, and Christian Dreger. 2011. Energy Consumption and Economic Growth: New Insights into the Co-integration Relationship. Energy Economics 33: 782–89. [Google Scholar] [CrossRef] [Green Version]
  13. Bernanke, Ben S. 1983. Irreversibility, Uncertainty, and Cyclical Investment. Quarterly Journal of Economics 98: 85–106. [Google Scholar] [CrossRef]
  14. Black, Fischer. 1987. Business Cycles and Equilibrium. Cambridge: Blackwell. [Google Scholar]
  15. Bowden, Nicholas, and James E. Payne. 2009. The causal relationship between U.S. energy consumption and real output: A disaggregated analysis. Journal of Policy Modeling 31: 180–88. [Google Scholar] [CrossRef]
  16. Carmona, Mónica, Emilio Congregado, Julia Feria, and Jesús Iglesias. 2017. The energy-growth nexus reconsidered: Persistence and causality. Renewable and Sustainable Energy Reviews 71: 342–47. [Google Scholar] [CrossRef]
  17. Carrion-i-Silvestre, Josep Lluís, Dukpa Kim, and Pierre Perron. 2009. GLS-based unit root tests with multiple structural breaks under both the null and the alternative hypotheses. Econometric Theory 25: 1754–92. [Google Scholar] [CrossRef]
  18. Chang, Ching-Chih. 2010. A multivariate causality test of carbon dioxide emissions, energy consumption and economic growth in China. Applied Energy 87: 3533–37. [Google Scholar] [CrossRef]
  19. Cheng, Benjamin S. 1995. An investigation of cointegration and causality between energy consumption and economic growth. The Journal of Energy and Development 21: 73–84. [Google Scholar]
  20. Cheng, Benjamin S. 1999. Causality between energy consumption and economic growth in India: An application of cointegration and error-correction modeling. Indian Economic Review 34: 39–49. [Google Scholar]
  21. Damette, Olivier, and Majda Seghir. 2013. Energy as a driver of growth in oil exporting countries? Energy Economics 37: 193–99. [Google Scholar] [CrossRef] [Green Version]
  22. Eggoh, Jude C., Chrysost Bangaké, and Christophe Rault. 2011. Energy consumption and economic growth revisited in African countries. Energy Policy, 7408–21. [Google Scholar] [CrossRef] [Green Version]
  23. Erdal, Gülistan, Hilmi Erdal, and Kemal Esengün. 2008. The causality between energy consumption and economic growth in Turkey. Energy Policy 36: 3838–42. [Google Scholar] [CrossRef]
  24. Ewing, Bradley T., Ramazan Sari, and Ugur Soytas. 2007. Disaggregate energy consumption and Industrial output in the United States. Energy Policy 35: 1274–81. [Google Scholar] [CrossRef]
  25. Fallahi, Firouz. 2011. Causal relationship between energy consumption (EC) and GDP: A Markov-switching (MS) causality. Energy 36: 4165–70. [Google Scholar] [CrossRef]
  26. Fatai, Koli, Les Oxley, and Francis Gordon Scrimgeour. 2004. Modelling the causal relationship between energy consumption and GDP in New Zealand, Australia, India, Indonesia, the Philippines and Thailand. Mathematics and Computers in Simulation 64: 431–45. [Google Scholar] [CrossRef]
  27. Gately, Dermot, and Hiliard G. Huntington. 2002. The asymmetric effects of changes in price and income on energy and oil demand. The Energy Journal 23: 19–55. [Google Scholar] [CrossRef] [Green Version]
  28. Hacker, Jacob S., Gregory A. Huber, Austin Nichols, Philipp Rehm, Mark Schlesinger, Rob Valletta, and Stuart Craig. 2014. The Economic Security Index: A New Measure for Research and Policy Analysis. The Review of Income and Wealth 60: S5–S32. [Google Scholar] [CrossRef] [Green Version]
  29. Hnatkovska, Viktoria, and Norman Loayza. 2005. Volatility and Growth. In Managing Economic Volatility and Crises: A Practitioner’s Guide. Edited by Aizenman Joshua and Brian Pinto. Cambridge: Cambridge University Press. [Google Scholar]
  30. Hossein, Abbasinejad, Gudarzi Farahani Yazdan, and Asghari Ghara Ehsan. 2012. The relationship between energy consumption, energy prices and economic growth: Case study (OPEC countries). OPEC Energy Review 36: 272–86. [Google Scholar] [CrossRef]
  31. Jalil, Abdul. 2014. Energy–Growth conundrum in energy exporting and importing countries: Evidence from heterogeneous panel methods robust to cross-sectional dependence. Energy Economics 44: 314–24. [Google Scholar] [CrossRef]
  32. Jamil, Faisal, and Eatzaz Ahmad. 2010. The relationship between electricity consumption, electricity prices and GDP in Pakistan. Energy Policy 38: 6010–25. [Google Scholar] [CrossRef]
  33. Joyeux, Roselyne, and Ronald D. Ripple. 2011. Energy consumption and real income: A panel cointegration multi-country study. The Energy Journal 32: 107–41. [Google Scholar] [CrossRef]
  34. Judson, Ruth, and Athanasios Orphanides. 1999. Inflation, volatility and growth. International Finance 1: 117–38. [Google Scholar] [CrossRef] [Green Version]
  35. Kraft, John, and Arthur Kraft. 1978. On the relationship between energy and GNP. Journal of Energy Development 3: 401–3. [Google Scholar]
  36. Larissa, Batrancea, Rathnaswamy Malar Maran, Batrancea Ioan, Nichita Anca, Rus Mircea-Iosif, Tulai Horia, Fatacean Gheorghe, Masca Ema Speranta, and Morar Ioan Dan. 2020. Adjusted Net Savings of CEE and Baltic Nations in the Context os Sustainable Economic Growth: A Panel Data Analysis. Journal of Risk and Financial Management 13: 234. [Google Scholar] [CrossRef]
  37. Liddle, Brantley, and Perry Sadorsky. 2020. How much do asymmetric changes in income and energy prices affect energy demand? The Journal of Economic Asymmetries 21: e00141. [Google Scholar] [CrossRef]
  38. Liddle, Brantley, Russell Smyth, and Xibin Zhang. 2020. Time-varying income and price elasticities for energy demand: Evidence from a middle-income panel. Energy Economics 86: 104681. [Google Scholar] [CrossRef]
  39. Medlock, Kenneth, and Ronald Soligo. 2021. Economic Development and End-Use Energy Demand. The Energy Journal 22: 77–105. [Google Scholar] [CrossRef]
  40. Menegaki, Angeliki N., and Stella Tsani. 2018. Critical Issues to Be Answered in the Energy-Growth Nexus (EGN) Research Field. In The Economics and Econometrics of the Energy-Growth Nexus. Cambridge: Academic Press, pp. 141–84. [Google Scholar]
  41. Mohammadi, Hassan, and Shahrokh Parvaresh. 2014. Energy consumption and output: Evidence from a panel of 14 oil exporting countries. Energy Economics 41: 41–46. [Google Scholar] [CrossRef]
  42. Nepal, Rabindra, Tooraj Jamasb, and Clement Allan Tisdell. 2014. Market related reforms and increased energy efficiency in transition countries: Empirical evidence. Applied Economics 46: 4125–36. [Google Scholar] [CrossRef] [Green Version]
  43. Owyang, Michael T., Jeremy Piger, and Howard J. Wall. 2008. A State-Level Analysis of the Great Moderation. Regional Science and Urban Economics 38: 578–89. [Google Scholar] [CrossRef] [Green Version]
  44. Paul, Shyamal, and Rabindra N. Bhattacharya. 2004. Causality between energy consumption and economic growth in India: A note on conflicting results. Energy Economics 26: 977–83. [Google Scholar] [CrossRef]
  45. Payne, James E. 2009. On the dynamics of energy consumption and output in the US. Applied Energy 86: 575–77. [Google Scholar] [CrossRef]
  46. Pindyck, Robert S. 1991. Irreversibility, Uncertainty, and Investment. Journal of Economic Literature 29: 1110–48. [Google Scholar]
  47. Rajaguru, Gulasekaran, and Tilak Abeysinghe. 2008. Temporal aggregation, cointegration and causality inference. Economics Letters 101: 223–26. [Google Scholar] [CrossRef]
  48. Rajaguru, Gulasekaran, Michael O’Neill, and Tilak Abeysinghe. 2018. Does Systematic Sampling Preserve Granger Causality with an Application to High Frequency Financial Data? Econometrics 6: 31. [Google Scholar] [CrossRef] [Green Version]
  49. Rashid, Abdul, and Ozge Kandemir Kocaaslan. 2013. Does Energy Consumption Volatility Affect Real GDP Volatility? An Empirical Analysis for the UK. International Journal of Energy Economics and Policy 3: 384–94. [Google Scholar]
  50. Rühl, Christof, Paul Appleby, Julian Fennema, Alexander Naumov, and Mark Schaffer. 2012. Economic development and the demand for energy: A historical perspective on the next 20 years. Energy Policy 50: 109–16. [Google Scholar] [CrossRef]
  51. Salim, Ruhul, Shuddhasattwa Rafiq, Sahar Shafiei, and Yao Yao. 2019. Does urbanization increase pollutant emission and energy-intensity? Evidence from some Asian developing economies. Applied Economics 51: 4008–24. [Google Scholar] [CrossRef]
  52. Shahbaz, Muhammad, Muhammad Zeshan, and Talat Afza. 2012. Is energy consumption effective to spur economic growth in Pakistan? New evidence from bounds test to level relationships and Granger causality tests. Economic Modeling 29: 2310–319. [Google Scholar] [CrossRef] [Green Version]
  53. Shahbaz, Muhammad, Muhammad Zakaria, Syed Jawad Hussain Shahzad, and Mantu Kumar Mahalik. 2018. The energy consumption and economic growth nexus in top ten energy-consuming countries: Fresh evidence from using the quantile-on-quantile approach. Energy Economics 71: 282–301. [Google Scholar] [CrossRef] [Green Version]
  54. Śmiech, Sławomir, and Monika Papież. 2014. Energy consumption and economic growth in the light of meeting the targets of energy policy in the EU: The bootstrap panel Granger causality approach. Energy Policy 71: 118–29. [Google Scholar] [CrossRef]
  55. Soytas, Ugur, and Ramazan Sari. 2006. Energy consumption and income in G-7 countries. Journal of Policy Modeling 28: 739–50. [Google Scholar] [CrossRef]
  56. Soytas, Ugur, Ramazan Sari, and Bradley T. Ewing. 2007. Energy consumption, income, and carbon emissions in the United States. Ecological Economics 62: 482–89. [Google Scholar] [CrossRef]
  57. Stern, David I. 1993. Energy and economic growth in the USA. A multivariate approach. Energy Economics 15: 137–50. [Google Scholar] [CrossRef]
  58. Stern, David I. 2000. A multivariate cointegration analysis of the role of energy in the US macroeconomy. Energy Economics 22: 267–83. [Google Scholar] [CrossRef] [Green Version]
  59. Tang, Chor Foon. 2009. Electricity consumption, income, foreign direct investment, and population in Malaysia: New evidence from multivariate framework analysis. Journal of Economic Studies 4: 371–82. [Google Scholar] [CrossRef]
  60. Tiba, Sofien, and Anis Omri. 2017. Literature survey on the relationships between energy, environment and economic growth. Renewable and Sustainable Energy Reviews 69: 1129–46. [Google Scholar] [CrossRef]
  61. Van Benthem, Arthur, and Mattia Romani. 2009. Fuelling growth: What drives energy demand in developing countries? The Energy Journal 30: 91–114. [Google Scholar] [CrossRef]
  62. Wang, Sisi. S., Dong Q. Zhou, Peng Zhou, and Qunwei Wang. 2011a. CO2 emissions, energy consumption and economic growth in China: A panel data analysis. Energy Policy 39: 4870–75. [Google Scholar] [CrossRef]
  63. Wang, Yuan, Yichen Wang, Jing Zhou, Xiaodong Zhu, and Genfa Lu. 2011b. Energy consumption and economic growth in China: A multivariate causality test. Energy Policy 39: 4399–406. [Google Scholar] [CrossRef]
  64. World Development Indicators. 2017. The World Bank Group. Washington: World Bank. [Google Scholar]
  65. Yamamoto, Taku, and Eiji Kurozumi. 2006. Tests for long-run Granger non-causality in cointegrated systems. Journal of Time Series Analysis 27: 703–23. [Google Scholar] [CrossRef] [Green Version]
  66. Yildirim, Ertugrul, and Alper Aslan. 2012. Energy consumption and economic growth nexus for 17 highly developed OECD countries: Further evidence based on bootstrap-corrected causality tests. Energy Policy 51: 985–93. [Google Scholar] [CrossRef]
  67. Yu, Eden S. H., and Been-Kwei Hwang. 1984. The relationship between energy and GNP: Further results. Energy Economics 6: 186–90. [Google Scholar] [CrossRef]
  68. Yu, Eden S. H., and Jang C. Jin. 1992. Cointegration tests of energy consumption, income, and employment. Resources and Energy 14: 259–66. [Google Scholar] [CrossRef]
  69. Zhang, Yue-Jun. 2011. Interpreting the dynamic nexus between energy consumption and economic growth: Empirical evidence from Russia. Energy Policy 39: 2265–72. [Google Scholar] [CrossRef]
  70. Zhixin, Zhang, and Ren Xin. 2011. Causal relationships between energy consumption and economic growth. Energy Procedia 5: 2065–71. [Google Scholar] [CrossRef] [Green Version]
Figure 1. Trends in energy intensity (1990–2016). Notes: Energy intensity at constant purchasing power parities (koe/$2005). This figure includes trends from all regions, including Africa, America, Asia, BRICS and CIS countries, Europe, the European Union, the G7, Latin America, the Middle East, North America, OECD countries, and the Pacific. All regions, except CIS and BRICS countries, are plotted against the primary axis. CIS and BRICS countries are plotted against the secondary axis. These two regions are plotted in yellow.
Figure 1. Trends in energy intensity (1990–2016). Notes: Energy intensity at constant purchasing power parities (koe/$2005). This figure includes trends from all regions, including Africa, America, Asia, BRICS and CIS countries, Europe, the European Union, the G7, Latin America, the Middle East, North America, OECD countries, and the Pacific. All regions, except CIS and BRICS countries, are plotted against the primary axis. CIS and BRICS countries are plotted against the secondary axis. These two regions are plotted in yellow.
Jrfm 14 00471 g001
Figure 2. (a) Log(energy intensity); (b) log(per-capita GDP). Note: The trends in energy intensity and real GDP per capita for the sample of 48 countries used in this study.
Figure 2. (a) Log(energy intensity); (b) log(per-capita GDP). Note: The trends in energy intensity and real GDP per capita for the sample of 48 countries used in this study.
Jrfm 14 00471 g002
Figure 3. (a) Real GDP per capita and energy consumption per capita; (b) income and energy intensity (period averages for 1970–2016).
Figure 3. (a) Real GDP per capita and energy consumption per capita; (b) income and energy intensity (period averages for 1970–2016).
Jrfm 14 00471 g003
Figure 4. Per-capita income growth volatility and energy consumption growth (averages).
Figure 4. Per-capita income growth volatility and energy consumption growth (averages).
Jrfm 14 00471 g004
Table 1. Summary of cointegration test results.
Table 1. Summary of cointegration test results.
# of Cointegrating VectorsCountries
r = 2Bolivia, the Czech Republic, Ecuador, Gabon, Hungary, Indonesia, Nigeria, Norway, Pakistan, the Philippines, Portugal, Spain, South Africa, Sudan, Syria, Trinidad and Tobago, and Venezuela
r = 3Algeria, Albania, Austria, Australia, Bangladesh, Brazil, Belgium, Canada, China, Chile, Colombia, Denmark, Egypt, France, Finland, Germany, India, Italy, Iran, Japan, New Zealand, the Netherlands, Sri Lanka, South Korea, Sweden, Turkey, Thailand, the UAE, the USA, the UK, and Vietnam
Note: Since the volatility measures are I(0), we expect at least r = 2. The number of cointegrating vectors r = 2 indicates no long-run relationships between per-capita income and the per-capita energy use and r = 3 indicates the presence of long-run relationships between real GDP per capita and energy consumption per capita.
Table 2. Long-run causal inferences.
Table 2. Long-run causal inferences.
Countries (1)Coefficient (β)
(2)
(se) (3)Per-Capita Energy Consumption Equation (3b)Per-Capita Real GDP Equation (3a)Yamamoto–Kurozumi TestLong-Run Causality
EC(-1)
21)
(4)
(se)
(5)
EC(-1)
11)
(6)
(se)
(7)
RGDPPC → ECPC
(8)
ECPC → RGDPPC
(9)
RGDPPC → ECPC
(10)
ECPC → RGDPPC
(11)
Albania0.13 ***(0.05)−0.26 ***(0.11)0.18(0.33)7.16 ***1.18NegativeNone
Algeria0.98 ***(0.08)−0.0008(0.01)0.05 **(0.02)1.645.19 **PositivePositive
Australia−0.58 **(0.28)−0.05 ***(0.01)−0.06(0.05)4.08 **0.67NegativeNegative
Austria7.66 ***(2.27)−0.00005(0.0002)0.01 ***(0.003)1.3416.02 ***NonePositive
Bangladesh0.43 ***(0.02)−0.05 ***(0.02)0.03(0.04)5.21 **0.08PositiveNone
Belgium−0.13 ***(0.04)−0.55 ***(0.14)0.07(0.43)9.74 ***0.54NegativeNone
Bolivia--------NoneNone
Brazil0.37 ***(0.06)−0.002 ***(0.0004)0.0008(0.001)5.96 ***0.18PositiveNone
Canada−0.25 ***(0.05)−0.20 **(0.09)−2.22 ***(0.53)5.29 **9.95 ***NegativeNegative
Chile−0.47 ***(0.04)−0.94 ***(0.33)−0.95(1.30)4.45 **1.09NegativeNone
China0.37 ***(0.11)−0.005 *(0.002)0.05 **(0.03)3.48 *6.47 **PositivePositive
Colombia0.06 ***(0.03)0.25 ***(0.07)0.13(0.51)6.35 **1.53PositiveNone
The Czech Republic--------NoneNone
Denmark−0.12 **(0.05)−0.06 **(0.18)−0.65(0.50)3.36 *1.41NegativeNone
Ecuador--------NoneNone
Egypt−0.38 ***(0.04)0.11(0.07)−0.51 ***(0.18)0.0350.09 ***NoneNegative
Finland−0.23 **(0.09)−0.15 **(0.06)−0.01(0.13)26.10 ***0.86NegativeNone
France0.81 ***(0.02)−0.03 **(0.01)0.09 **(0.03)3.45 *4.82 **PositivePositive
Gabon--------NoneNone
Germany−0.15 ***(0.03)−0.31 ***(0.10)−0.33(0.32)14.15 **0.26NegativeNone
Hungary--------NoneNone
India0.97 ***(0.03)−0.004(0.003)0.09 ***(0.02)1.456.38 **NonePositive
Indonesia--------NoneNone
Iran−0.17 ***(0.02)−0.14 ***(0.04)0.69 ***(0.017)14.39 ***10.56 ***NegativeNegative
Italy−0.32 ***(0.11)−0.06 ***(0.02)0.14(0.14)4.05 **1.73NegativeNone
Japan−0.17 ***(0.03)−0.49 ***(0.11)−0.07(0.82)3.94 **0.68NegativeNone
Korea1.82 ***(0.07)−0.11(0.12)0.65 ***(0.16)1.5810.86 ***NonePositive
The Netherlands−0.10 ***(0.05)−0.39 ***(0.14)0.004(0.34)4.51 **1.66NegativeNone
New Zealand−0.71 ***(0.14)−0.06 *(0.03)−0.34 ***(0.09)2.81 *8.27 ***NegativeNegative
Nigeria--------NoneNone
Norway--------NoneNone
Pakistan--------NoneNone
Philippines--------NoneNone
Portugal--------NoneNone
South Africa--------NoneNone
Spain--------NoneNone
Sri Lanka0.74 ***(0.03)−0.02(0.02)0.13 ***(0.04)2.5511.62 ***NonePositive
Sudan--------NoneNone
Sweden−0.20 ***(0.05)−0.51 ***(0.11)−0.26(0.38)8.06 ***2.54NegativeNone
Syrian--------NoneNone
Thailand−0.72 ***(0.10)0.10 **(0.05)-0.09(0.15)7.11 **0.29NegativeNone
Trinidad and Tobago--------NoneNone
Turkey0.79 ***(0.07)−0.02 **(0.01)0.06 *(0.04)12.62 ***5.73 *PositivePositive
UAE−0.61 ***(0.13)−0.14 ***(0.04)−0.08 *(0.04)16.72 ***4.72 **NegativeNegative
UK−0.07 ***(0.01)−0.26 **(0.11)−1.99 ***(0.71)9.53 **60.31 ***NegativeNegative
USA−0.32 ***(0.12)−0.21 **(0.08)−0.05(0.33)6.67 **1.63NegativeNone
Venezuela--------NoneNone
Vietnam−0.49 ***(0.03)−0.02(0.02)0.79 ***(0.06)1.3717.74 ***NoneNegative
Notes: *, **, and *** denote significance at 10%, 5%, and 1% respectively. ln ( E C P P C t ) = β ln ( R G D P P C t ) + e 1 t , where γ11 (column 6) and γ 21 (column 4) are the estimated speed of adjustment coefficients for the per-capita real GDP equation and the per-capita energy consumption equations, respectively. EC–error correction term, ECPC–energy consumption per capita, and RGDPPC – real gross domestic product per capita. RGDPPC  ECPC (in column 8) denotes long-run causality from RGDPPC to ECPC and ECPC RGDPPC (in column 9) denotes long-run causality from ECPC to RGDPPC. The sign of β determines the nature of the long-run causal relationships (positive, negative, or none).
Table 3. Summary of long-run Granger causality results.
Table 3. Summary of long-run Granger causality results.
CausalityPositiveNegative
RGDPPC → ECPCAlbania, Bangladesh, Brazil, ColumbiaBelgium, Chile, Denmark, Finland, Germany, Italy, Japan, the Netherlands, Sweden, Thailand, the USA
ECPC → RGDPPCAustria, India, South Korea, Sri LankaEgypt, Vietnam
RGPPC ↔ ECPCAlgeria, China, France, TurkeyAustralia, Iran, Canada, New Zealand, the UK, and the UAE
No causality between RGDPPC and ECPCBolivia, Ecuador, the Czech Republic, Gabon, Indonesia, Hungary, Norway, Nigeria, Pakistan, Portugal, the Philippines, Spain, South Africa, Sudan, Syria, Venezuela, Trinidad and Tobago
Notes: This is based on the results from Table 1.
Table 4. Short-run Granger causality results.
Table 4. Short-run Granger causality results.
CausalityPositiveNegative
RGDPPC → ECPCBolivia, Canada, Ecuador, Gabon, Germany, Hungary, India, Indonesia, Iran, Japan, the Netherlands, Pakistan, Portugal, Spain, Sweden, Syria, Thailand, VenezuelaAlbania, Belgium, Sudan
INVOL → ECPC Albania, Algeria, Belgium, Bolivia, Brazil, Canada, Chile, Colombia, Denmark, Ecuador, France, Gabon, Germany, Hungary, Indonesia, Iran, Japan, Korea, the Netherlands, Nigeria, the Philippines, Portugal, South Africa, Spain, Syria, the UAE, the UK, the USA
EVOL→ ECPCAlgeria, Brazil, Ecuador, Finland, Germany, the Netherlands, Nigeria, Portugal, Syria, Trinidad and Tobago, Turkey, the UAE, the USA
ECPC → RGDPPCCanada, China, France, Gabon, Hungary, Iran, Italy, South Korea, Pakistan, the Philippines, Portugal, South Africa, Spain, the UK, the USA, Venezuela, Vietnam
INVOL → RGDPPCAustria, India, South Korea, Philippines, Sri Lanka, TurkeyAlbania, Algeria, Canada, Chile, the Czech Republic, Indonesia, Iran, New Zealand, Portugal, South Africa, Spain, the UK, Vietnam
EVOL → RGDPPCAlbania, Algeria, Austria, France, Portugal, Syria Chile, China, Hungary, Iran, Italy, South Korea, Norway, Venezuela
ECPC → INVOLCanada, Indonesia, Italy, Spain, Turkey, the UKChina, Hungary, Japan, Thailand, the UAE
RGDPPC → INVOLBelgium, Bolivia, Chile, China, the Czech Republic, the UKAlbania, Algeria, Australia, Austria, Brazil, Canada, Columbia, Denmark, Ecuador, Finland, France, Gabon, India, Indonesia, Iran, Italy, the Netherlands, Portugal, South Africa, Spain, Syria, Thailand, Trinidad and Tobago, Turkey, the UAE, Venezuela, Vietnam
EVOL → INVOLAlbania, Austria, China, Germany, Indonesia, Japan, Pakistan, Portugal, South Africa, Sudan, Syria, the UK
ECPC → ECVOLSpain, Sri Lanka, Turkey, Albania, Algeria, China, the Czech Republic, Ecuador, Iran, the Netherlands, Nigeria, Norway, Portugal, Syria, Thailand, the UK
RGDPPC → EVOLAlbania, Canada, Ecuador, India, Portugal, Syria, the USA, VietnamBangladesh, Finland, Nigeria, Pakistan
INVOL → EVOLBolivia, Colombia, the Czech Republic, Denmark, Germany, India, New Zealand, Nigeria, Norway, Pakistan, Portugal, Sudan, Syria, Thailand, the UK, and the USA
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Rajaguru, G.; Khan, S.U. Causality between Energy Consumption and Economic Growth in the Presence of Growth Volatility: Multi-Country Evidence. J. Risk Financial Manag. 2021, 14, 471. https://doi.org/10.3390/jrfm14100471

AMA Style

Rajaguru G, Khan SU. Causality between Energy Consumption and Economic Growth in the Presence of Growth Volatility: Multi-Country Evidence. Journal of Risk and Financial Management. 2021; 14(10):471. https://doi.org/10.3390/jrfm14100471

Chicago/Turabian Style

Rajaguru, Gulasekaran, and Safdar Ullah Khan. 2021. "Causality between Energy Consumption and Economic Growth in the Presence of Growth Volatility: Multi-Country Evidence" Journal of Risk and Financial Management 14, no. 10: 471. https://doi.org/10.3390/jrfm14100471

Article Metrics

Back to TopTop