The Relationship between Hydropower Energy Consumption and Economic Growth

Available online at www.sciencedirect.com

ScienceDirect Procedia Economics and Finance 38 (2016) 264 – 270

Istanbul Conference of Economics and Finance, ICEF 2015, 22-23 October 2015, Istanbul, Turkey

The Relationship Between Hydropower Energy Consumption and Economic Growth Melike Bildiricia1 a

Prof.Dr.Yıldız Technical University, Department of Economics, Istanbul, Turkey

Abstract This paper will analyse the relationship between economic growth and hydropower energy consumption. According to the results of the short run causality, there is evidence to support the growth hypothesis in OECD countries with high incomes. There is evidence to support the conservation hypothesis for Brazil, Finland, France, Mexico, the U.S. and Turkey. The unidirectional causality goes from economic growth to energy consumption and suggests that the policy of conserving hydropower energy consumption may be implemented with little or no adverse effects on economic growth in less energy-dependent economies. © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2015 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Committee of ICEF Peer-review underresponsibility responsibility of the Organizing of the Organizing Committee of ICEF 2015.2015. Peer-review under Keywords: Economic Growth, Hydropower Energy Consumption, ARDL

1. Introduction Pollution caused by non-renewable energy is an important issue because humans are facing the dual pressure of economic growth and environmental protection (See Zhang et.al: 2011). Pollution caused by non-renewable energy revealed the dual pressure of economic growth and environmental protection (Zhang et.al 2011). In the late 1980s, the analytical paradigm was altered with concerns about environmentally sustainable economic growth. Sustainable economic growth policies are based on the level, quality and management of renewable and non-renewable natural resources. The states of the environment depends on the level and growth of pollution or waste streams, the environment’s natural assimilation of pollution or through clean up expenditures. As the concept of sustainable development emerged with the rise of green movements, hydropower energy became popular. 1

Corresponding author. Tel.:+90-212-383-6818. E-mail address: [email protected]

2212-5671 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Organizing Committee of ICEF 2015. doi:10.1016/S2212-5671(16)30198-8

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270

As electricity demand especially grew, the number and size of non-renewable energy increased. Early hydropower plants were much more reliable and efficient than the non-renewable plants of the day (Baird:2006; Kumar et.all:2011). Unlike non-renewal energy sources, renewable hydropower energy can continuously produce energy (Margeta and Glasnovic. 2011). According to the report of world bank (WB:2009), hydropower has a powerful contribution to make to regional cooperation and development and to the allocation of increasingly scarce water resources and it is complex and brings a range of economic, social and environmental risks (Schumann et.all R.2010). This paper, will analyse the relationship between economic growth and hydropower energy consumption. The next section of the study will present studies about hydropower energy consumption and economic growth in the literature. Econometric theory and methodology is identified in the third section. The fourth section consists of the empirical results while the last section includes conclusions and policy implications. 2. The Literature of Energy Consumption and Economic Growth Kraft and Kraft (1978), Berndt (1978), Akarca and Long (1980), Yu and Hwang (1984), Yu and Choi (1985), and Erol and Yu (1987) were among the first researchers examined the relation between GDP and energy consumption. In pursuit of their work, many paper analysed the relationship between energy consumption and economic growth. As result of these papers, it was obtained the different result on direction of causality. Different results on direction of causality allowed for four hypotheses: 1) the “neutrality hypothesis”; 2) the “conservation hypothesis”; 3) the ”growth hypothesis” and 4) the “feedback hypothesis’’. The literature about the causal relationship between hydropower energy consumption and economic growth is very little when compared with the number of papers on other forms of renewable energy. Sadorsky (2009) examined the relationship between renewable energy consumption and real GDP for 18 emerging countries and determined that an increase in real GDP should have a positive impact on renewable energy consumption. Chien and Hu (2008) used the structural equation modeling approach to show the effects of renewable energy on the GDP of 116 countries. They determined that renewable energy has a positive and significant effect on capital formation and that there are relationships between renewable energy and GDP through the increase in capital formation. Chien and Hu (2007) examined the effects of renewable energy on the technical efficiency of 45 countries in 2001 and 2002. They found that the technical efficiency in OECD countries was higher than non-OECD countries, but non-OECD countries have a higher share of renewable energy. Solarin and Öztürk(2015) examined the relationship between hydroelectricity consumption and economic growth in seven Latin America countries including Argentina, Brazil, Chile, Colombia, Ecuador, Peru and Venezuela. Their results determined long run bidirectional causality between hydroelectricity consumption and economic growth in Argentina and Venezuela. There is the evidence for long run unidirectional causality from hydroelectricity consumption to economic growth in Brazil, Chile, Colombia, Ecuador and Peru. Bildirici (2015) investigated the relationship between CO2 environmental pollution, hydropower energy consumption, and economic growth in for Austria, Belgium, Denmark, Finland, France, Germany (1970–2011), Iceland, Italy, Ireland, Portugal, Spain, Sweden, Switzerland (1981–2011), Turkey, and the United Kingdom utilizing the ARDL method in the period from 1961 to 2011. According to her causality results, for Germany the conservation hypothesis determines the unidirectional causality running from GDP to hydropower energy consumption. For Austria, the growth hypothesis suggests unidirectional causality running from hydro energy consumption to GDP. For the United Kingdom, the neutrality hypothesis was determined. For other countries, a bidirectional causality hypothesis was determined.

265

266

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270

3. Data and Econometric Methodology 3.1. Data The annual data used in this study cover the period from 1980 to 2011 for Brazil, Canada, Finland, France, Japan, Mexico, USA, UK, Turkey, countries with high growth in world, countries:Non-OECD with high growth in world and countries with high Income in OECD countries. Since data did not obtained for all of high growth countries, average value for countries with high growth in world, countries:Non-OECD with high growth in world and countries with high Income in OECD countries was used to analyses The variables in this study are hydro energy consumption (hec) and real GDP (py). Data are taken as hy=log(hec) and py=log(py). Hydro energy consumption is obtained from Word Bank and Real GDP measured in constant 2005 US dollars comes from the World Bank World Development Indicators (WDI 2012). 3.2. Methodology In the ARDL analysis, the variables of the model are allowed to possess mixed order of integration. The ARDL model for the standard log-linear functional specification of long-run relationship between variables with OLS estimation technique is presented as, 'y

m

n

D 0  ¦ E i 'yt i  ¦ Ii 'hect i  G1 yt 1  G 2 hect 1  H t i 1

i

(1)

0

where ' and H t are the first difference operator and the white noise term. The bounds testing procedure is based on the joint F-statistic or Wald statistic that tests the null hypothesis of no cointegration. The null hypothesis of no cointegration among the variables in Eq. (1) are H 0 : G1 G 2 0 against the alternative hypothesis

H1 : G1 z G 2 z 0 .One set of critical values assumes that all variables in the ARDL model are I(0), while the other is calculated on the assumption that the variables are I(1). In the second step, if cointegration is established, the conditional ARDL long-run model for hec can be estimated as: m

n

i=1

i=0

y=O0 +¦ D i yt-i +¦ -i hec +u t

(2)

In the third stage, the short-run dynamic parameters are obtained by estimating an error correction model associated with the long-run estimates: m

n

i=1

i=0

Δy=F 0 +¦ E i Δy t-i +¦ T i Δhec t-i +] ECM t 1 +e t

(3)

where residuals et is independently and normally distributed with zero mean and constant variance and ECMt-1 is the error correction term. ] is a parameter that indicates the speed of adjustment to the equilibrium level after a shock. 3.3. Granger Causality ARDL approach tests if the existence or absences of long-run relationship but it doesn’t determine the direction of causality. It was used the two-step procedure from the Engle and Granger (1987) model to examine the causal relationship between hydroenergy consumption and the real GDP (see Bildirici: 2015; Bildirici and Kayıkcı:2012).

267

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270

Vector Error Correction model that was used to analyze the short run relationships between the variables is constructed as follows, 'y 'hec

m

n

i 1

i 0

D 0  ¦ b1i 'yt i  ¦ b2i 'hect i  b3 ECM t 1  et p

q

i 1

i 0

(4)

D 0  ¦ d1i 'hect i  ¦ d 2i 'yt i  d3 ECM t 1  et

(5)

where residuals, et is independently and normally distributed with zero mean and constant variance and ECM t-1 is the error correction term resulting from the long-un equilibrium relationship and d’s are parameters to be estimated. Granger causality can be examined in two ways in the paper. First, short run or weak Granger causalities are tested by H 0 : b2i 0 in Equations (4) and (5). Second, long run Granger causalities are tested 0 and H 0 : d 2 i from the ECTs in those equations. Long-run causalities are tested by H 0 : b3 causalities are tested by H 0 : b2i

b3

0 and

H 0 : d 2i

d3

0 and H 0 : d3

0 . Strong

0

4. Econometric Result According to the F-statistics, we have enough evidence to reject the null hypothesis of no cointegration at a 1 percent significance level between hydropower energy consumption and economic growth for Brazil, Canada, Finland, France, Japan, Mexico, the U.S., UK, and Turkey, countries with high growth rates, non-OECD countries with high growth rates, and OECD countries with high income. Table 1: Bounds Testing for Cointegration Fy(y|hec) Brazil Canada Finland France Japan Mexico UK USA Turkey High Income ++ High Income: Non-OECD High Income: OECD

1.8569 1.6029 1.9502 20.6718* 23.7512* 1.9371 2.5251 1.1142 .13309 11.8062* 10.8380* 6.8997*

Fhec(hec|y) 6.2448* 6.1349* 7.6400* 2.0013 1.3036 7.5321* 0.1145 9.8998* 7.3900* 1.7235 1.11425 2.1174

This simply means that the computed F-statistics for these models are above the upper bound critical value. Fy (y|hec) was determined for France, Japan, high income: non-OECD and high income: OECD Fhec(hec|y ) for Brazil, Canada, Finland, Mexico, UK, U.S. and Turkey. The results of the ARDL bounds test suggest the rejection of the null hypothesis of no long-run relationship at the 1 percent level of significance. 4.1. Results of Long-Run and Short Run Elasticities The ARDL co-integration analysis assumes the existence of a unique long-run relationship among variables.

268

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270 Table 2. Long and Short Run Coefficients for ARDL Long Run Coefficients hec Brazil Canada Finland France Japan

Turkey

High Income: OECD

dhec

-.85292 (-4.7628) -.87436 (-5.1348) -1.5541 (-5.3136)

dY -.10917 (4.0887) -.11289 (2.31961) -.72376 (3.86519

-.013822 (2.2023) -.015006 (2.41528) -2.2836 (3.4309) -1.1981 (15.4839) .55197 (2.1475)

USA

High Income: Non-OECD

y -.49381 (3.4143) -.38044 (8.1172) -1.2550 (12.9174)

-.22772 (1.98752) -.21309 (0.54973)

Mexico

High Income

Short Run Coefficients

-

-.68919 (1.8724) -.74160 (4.518) 1.3168 (1.827) -.25428 (4.3928) -.25082 (4.3172) -.56927 (-2.5039)

ECM -.22108 (2.8808) -.29673 (2.9025) -.57669 (4.1633) -.060699 (3.2916) -.081952 (2.6108) -.30180 (2.4041) -.61898 (4.9005) -.31015 (2.9235) -.092471 (-1.983) -.099318 (-1.9830) -.36631 (3.1044)

It is given t value in parenthesis

It is possible to forecast the long-run relationships and short-run dynamic effects by using ARDL approach. Table 2 shows the long-run elasticities for the ARDL model. In Table 2, when it is analysed the relation between hydropower energy consumption to income, it is seen that the income elasticity of hydropower energy demand has a negative sign for Brazil, Canada, Finland, Mexico, USA. For Brazil, Canada, Finland, Mexico, USA, the estimated income elasticities of hydropower energy demand has a negative sign. For Turkey, a positive income elasticity of hydropower energy demand is associated with normal good. The ECMs in Table 2 indicate that there is a mechanism to correct the disequilibrium between economic growth and hydropower energy consumption. The sign of the coefficient of the error correction term must be negative. The ECMs change between -.092471 and -.61898 to provide stability for the model. An interesting result for ECM was found as <1 % for high income countries and non-OECD high income countries. The error correction term was negatively and statistically significant, showing a low speed of adjustment of any disequilibrium toward a long-run equilibrium state. 4.2. The Results for Granger Causality Test Table 3 summarizes the causality relationship between hydropower energy consumption and economic growth. Table 3. Granger Causality Δy→Δ hec, Δ hec →Δy, Brazil Canada Finland

51.859 0.90876 34.7937 15.4397 18.598 2.4448

ECT→ Δy ECT→ Δ hec 16.0012 0.19896 11.7713 16.0245 21.510 16.7442

' y, ECT → ' hec ' hec , ECT→ ' y 41.0801 35.189 56.6688 55.6278 85.1171 94.987

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270

France Japan Mexico Turkey USA High Income in World High Income: Non-OECD High Income: OECD

92.8843 .45038 31.3367 173.3739 22.4077 1.2756 14.5076 1.0075 8.7671 2.1829 58.0 2.5076 1.9354 79.760 1.934 79.760

.45038 28.7280 13.543 34.9095 14.652 30.0789 18.972 17.023 21.7467 11.1167 56.9161 22.7322 29.1598 22.5291 139.4769 237.1658

269

48.9276 98.045 74.2482 93.1996 68.1345 67.9913 47.219 56.843 40.61464 26.8758 58.77254 38.0418 87.0554 110.875 91.1517 75.685

The short-run causality results determined that there is unidirectional causality from y to hec for Brazil, France, Mexico and Turkey, countries with high growth and bidirectional causality for Canada, Finland, Japan and US, nonOECD countries with high growth. The short-run causality result found support for the growth hypothesis for OECD countries with high income. The long-run causality determined the conversation hypothesis for France, growth hypothesis for Brazil and feedback hypothesis for other countries. According to the strong Granger causality result, there is bidirectional causality for all countries. 5. Conclusion According to the results of the short run causality, there is evidence to support the growth hypothesis in OECD countries with high income. There is evidence to support the conservation hypothesis for Brazil, France, Mexico Turkey and countries with high growth. The conservation hypothesis suggests that the policy of conserving hydropower energy may be implemented with little or no adverse effects on economic growth, such as in a less energy-dependent economy. Therefore this is not the theoretically expected outcome for developing countries. According to the long-run causality result, there is evidence to support bidirectional causality for all countries except Brazil and France. Hydropower energy consumption and economic growth are complementary and an increase in energy consumption stimulates economic growth, and vice-versa. References Akarca, A.T., Long, T.V. (1980) On the relationship between energy and GNP: a re-examination, J Energy and Develop. 5;326-331. Baird, S. (2006). The future of electricity. Electri. Today, 18(5), pp. 9-11 Berndt, E. (1978) The Demand for Electricity:Comment and Further Results. MIT Energy. Laboratory 1978: WP. No. MIT-EL78-021WP. Bildirici, M., Kayıkçı, F., (2012) Energy Consumption and Growth in Eastern Europa: ARDL Approach, Economic Research, 25(3); 538-559. Bildirici M, (2015), Hydropower Energy Consumption, Environmental Pollution and Economic Growth, The Journal of Energy and Development, 40(1- 2) Chien, T., Hu J.L. (2008) Renewable energy: An efficient mechanism to improve GDP, Ener. Pol. 36: 3045-3052. Chien, T., and J. L, Hu, (2007) Renewable energy and macroeconomic efficiency of OECD and non OECD economies, Ener. Pol. 35; 3606-3615. Engle R.E. and Granger C. W. J., (1987), Co-Integration and Error Correction: Representation, Estimation, and Testing,, Econometrica, 55( 2): 251-276 Erol,U., Yu, E.S.H. On the relationship between electricity and income for industrialized countries. Journal of Electricity and Employment. (1987);13; 113-122. Kraft, J., Kraft A. (1978) On the Relationship Between Energy and GNP, Journal of Energy and. Development,spring:1978; 401-403 Kumar, A., T. Schei, A. Ahenkorah, R. Caceres Rodriguez, J.-M. Devernay, M. Freitas, D. Hall, Å. Killingtveit, Z. Liu, (2011) Hydropower. In IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation (eds) O. Edenhofer, R. Pichs-Madruga, Y. Sokona, K. Seyboth, P. Matschoss, S. Kadner, T. Zwickel, P. Eickemeier, G. Hansen, S. Schlömer, C. von Stechow, Cambridge University Press, Cambridge.

270

Melike Bildirici / Procedia Economics and Finance 38 (2016) 264 – 270

Margeta, J. and Glasnovic, Z. (2011). Role of Water-Energy Storage in PV-PSH Power Plant Development" J. Energy Eng., 10: 187-197. Menegaki, A. N., (2011) Growth and renewable energy in Europe: A random effect model with evidence for neutrality hypothesis, Ener. Econ. 33; 257-263. Pesaran MH, Pesaran B.(1987) Working with Microfit 4.0: interactive econometric analysis. Oxford University Press, Oxford. Pesaran, M. H., Y. Shin, and R.J., Smith, (2001) Bounds testing approaches to the analysis of level relationships, J App. Economet. 16; 289 – 326. Sadorsky, P., (2009) Renewable energy consumption and income in emerging economies, Ener. Pol. 37; 4021-4028. Solarin A.A and Ozturk İ., (2015), On the causal dynamics between hydroelectricity consumption and economic growth in Latin America countries, Renewable and Sustainable Energy Reviews, 2015; 52:1857-1868. The World Bank, 2009. Directions in Hydropower. Washington. Available at: http://siteresources.worldbank.org/INTWAT/Resources/Directions_in_Hydropower_FINAL.pdf. Schumann K., Saili L., Taylor R., AbdelǦ Malek R., Hydropower and Sustainable Development: A Journey Article submitted to the World Energy Congress, 15 February 2010, International Hydropower Association United States Bureau of Reclamation: www.usbr.gov/power/edu/history.html. Yu, E.S.H., Hwang, B. (1984) The relationship between energy and GNP: further results. Energy Economics. 6:186-190. Yu, E. S. H., Choi, J. Y. (1985) The Causal Relationship between Energy and GNP : An International Comparison. J. Ener. Develop. 10(2): 249– 272. Zhang Jun, Teng Wenrui, Wei Xinli (2011). Energy optimization of biomass pyrolysis and liquefaction system in CFB 2011 International Conference on Computer Distributed Control and Intelligent Environmental Monitoring)