Factors influencing renewable energy generation in developed and developing countries

Размещено на http://www.allbest.ru/

ПЕРМСКИЙ ФИЛИАЛ ФЕДЕРАЛЬНОГО ГОСУДАРСТВЕННОГО АВТОНОМНОГО ОБРАЗОВАТЕЛЬНОГО УЧРЕЖДЕНИЯ ВЫСШЕГО ОБРАЗОВАНИЯ

«НАЦИОНАЛЬНЫЙ ИССЛЕДОВАТЕЛЬСКИЙ УНИВЕРСИТЕТ

«ВЫСШАЯ ШКОЛА ЭКОНОМИКИ»

Факультет экономики, менеджмента и бизнес-информатики

Выпускная квалификационная работа

Factors influencing renewable energy generation in developed and developing countries

Седова Лилия Вячеславовна

Пермь, 2020



Abstract

Building a carbon-neutral economy with the use of clean and renewable energy sources is extremely important due to the necessity of climate change consequences alleviation. This study explores factors affecting the transition to non-hydro renewable energy sources in high-, middle- and low-income economies. The growth of renewable energy considers in term of non-hydro renewable electricity generated and its share in total electricity supply. The results demonstrate that the leading factor of non-hydro renewable power generation is rising oil price, while a country's energy dependency indicates as a primary barrier towards green energy future. Economic development is revealed to have a feeble impact on renewable energy adoption. The effect of trade openness and the level of CO2 emissions on renewable power growth depends on the income group and chosen model specification. Finally, gas and coal prices are found to be a complementary energy sources in addition to non-hydro renewable electricity generation.

Построение углеродно-нейтральной экономики с использованием чистых и возобновляемых источников энергии чрезвычайно важно в связи с необходимостью смягчения последствий изменения климата. В данном исследовании рассматриваются факторы, влияющие на переход к возобновляемым источникам энергии в странах с высоким, средним и низким уровнем доходов. Рост сектора возобновляемой энергетики рассматривается с точки зрения производства возобновляемой электроэнергии, получаемой из неводных источников, и её доли в общем энергобалансе стран. Результаты исследования показывают, что драйвером распространения возобновляемых технологий является рост цен на нефть, в то время как высокая энергетическая зависимость стран препятствует их развитию. Выявлено, что уровень экономического развития оказывает слабое влияние на генерацию зеленой энергии, а влияние торговли и уровня выбросов CO2 меняется в зависимости от группы доходов стран и выбранной спецификации модели. Наконец, цены на газ и уголь являются комплементарными источниками энергии в дополнение к возобновляемым технологиям.



Table of contents

Introduction

1. Factors influencing renewable energy transition

1.1 Economic factors

1.2 Environmental factors

1.3 Other factors

2. Data

3. Methodology

3.1 Short-run dynamics

4. Pretesting

4.1 Long-run dynamics

5. Results

5.1 Panel unit root tests

5.2 Long-run dynamics

Conclusion

References

Appendix



Introduction

The rate of renewable energy dissemination has become rapid in recent years. This tendency is mainly affiliated to appreciation that renewable energy is one of the crucial means of decarbonizing the energy system and climate change effects assuaging. Climate scientists mostly agree that the use of fossil fuels in electricity production is the key reason of the rapid growth in greenhouse gas emissions and the leading cause of anthropogenic climate change. According to the Intergovernmental Panel on Climate Change (IPCC), the amount of carbon dioxide emissions released annually in the atmosphere increased by about 2,5 times during the period from 1970 to 2017 and composes 36.15 billion tones. Simultaneously, the global average temperature has increased by more than 1 o C since pre-industrial times (Met Office Hadley Centre, 2019).

This climate concern led to the adoption of the resolutions of the Kyoto Protocol, the Paris climate agreement, and the United Nations Sustainable Development Goals (SDGs), which urged to stem climate change. The Kyoto Protocol was adopted in 1997 and became the first global agreement aimed at ?ghting global warming and environmental degradation by setting targets striving to reduce greenhouse gas emissions. The Paris climate agreement (2015) resolves to limit an increase in global temperature within 2 oC above pre-industrial levels and achieve zero net carbon emissions in the second half of this century, mainly by reducing the share of fossil fuels in global energy use. Similarly, the SDGs (2015) recognize that clean and affordable energy is a vital element of sustainable development and growth of countries. Thus, one of the critical measures of global warming alleviation is increasing the share of renewable energy (or reducing the share of fossil fuel energy) in the world energy mix.

Some of the countries made considerable success towards decarbonization of energy system, while the level of renewable energy adoption in most developing economies remains relatively low. The development of renewable energy within industrialized and emerging countries varies considerably as well. Therefore, there is a need for analysis of factors influencing renewable energy growth. Nevertheless, the effect of the same factors may differ for affluent and low-income economies. Thus, it seems essential to investigate the drivers of renewable energy growth and the barriers to this transition in economies with different income level. Depending on the results, specific measures of energy policy should be improved or changed to make the transition to green energy faster and easier.

Thus, the research question of this study is which factors and to what extent influence the adoption of renewable energy sources, depending on the country's level of income?

The objectives of the ongoing research are:

To study foreign literature and to identify the main factors affecting renewable energy development;

To find available data on chosen determinants by country and analyze it;

To design an econometric model for renewable energy generation;

To calculate the effect of analyzed factors on renewable electricity generated and the share of renewable electricity in total energy supply;

To describe the results and make conclusions.

Unlike the previous papers, this study based on an extended sample of industrialized and emerging economies, not only on the most developed ones. Moreover, this research investigates a transition to renewable energy in terms of both renewable electricity generated and its share in the total energy mix.

The energy mix of a country is a combination of primary energy sources (fossil fuels, renewables, and nuclear) that makes up its total primary energy supply. It is essential to analyze renewable energy generated in the context of its percentage because countries should achieve a specific share of renewable energy until 2050 to prevent hazardous climate change consequences.

Literature review

Renewable energy has become widespread modern technology in recent years. It is a more affordable and cleaner source for electricity generation than traditional energy derived from fossil fuels. Additionally, it is renewable and does not produce waste, which is essential within the context of a future transition to a cyclical economy. Therefore, renewable energy has become the primary tool for reducing carbon dioxide emissions.

Renewable energy is electricity and heat derived from natural processes and replenished faster than is consumed. Renewables include bioenergy, geothermal energy, hydropower, solar energy, wind energy and energy of the ocean. The term renewable energy for brevity is often replaced by the abbreviation RE or synonym «green energy» (REN21, 2018).

Modern renewable energy is used in power generation, heating and cooling, and transport sector. Power generation sector employs renewable energy sources to produce clean electricity, which then unitizes in both the heating and transport sectors. The heating and cooling sector could apply renewable energy in three ways. Primarily, through the direct combustion of biomass (both modern and traditional). On the other hand, through the direct use of geothermal and solar thermal energy, and, lastly, by using renewable electricity to heat or cool buildings. Finally, the transport sector utilizes renewable energy in the form of liquid biofuels and biomethane and as renewable electricity directly for charging electric vehicles (REN21, 2018).

According to the Sustainable Development Goals, the Paris Agreement, and the European Union climate strategy, all three above-mentioned energy sectors needed to be transformed into carbon-neutral ones by the year 2050 (IRENA, 2019). This research work is focused on the renewable energy power sector as it is the most developed field today.

In general, under the term «renewable energy» is understood both hydropower and non-hydro renewable energy. Majority of previous studies explore renewable energy, combining hydro and non-hydro renewable energy (Aguirre, Ibikunle, 2014; Cadoret, Padovano, 2016; Carfora et al., 2018; Chen, 2018; Lin et al., 2016; Marques et al., 2010; Omri, Nguyen, 2014; Rafiq et al., 2014; Sadorsky, 2009; Salim et al., 2014). However, some studies focus exclusively on non-hydro renewables (Lin, Omoju, 2017; Pfeiffer, Mulder, 2013). Lin and Omoju (2017) point out three main reasons for this option.

Firstly, non-hydro renewable sources of energy, especially solar and wind, are more affordable than hydro energy as it does not need the requirement of river flows accessibility. Secondly, hydro energy has already been extensively exploited in countries with rich water sources, while other renewables began to proliferate only at the end of the 20th century. For example, the first hydropower plants in Russia were built in the period 1970-1975 to supply energy-intensive industries. Thus, it is economically feasible for hydro energy-oriented countries to apply the power of water because it is cheaper for electricity production than a fossil-based generation. While the usage of non-hydro renewable energy is primarily driven by the idea to slow down the impact of climate change. Thirdly, there is an issue about the environmental sustainability of hydro energy as it harms the local environment causing disruptions in rivers ecosystems and habitat.

Furthermore, hydropower is highly developed all over the world. That is why the great part of new hydropower projects have comparatively low installed capacity (IHA, 2019). Besides, the potential of hydropower is limited due to the presence of natural boundaries. At the same time, the geographical potential of wind and solar energy is immensely higher (REN21, 2018). For this reason, the level of hydroelectricity production practically remains stable, and future growth is eliminated with geographical conditions. Therefore, in the current study, the hydroelectric power generation will not be considered.

Most of the literature focuses on the level of renewable energy produced or consumed as it allows to analyze renewable energy development from the side of supply or demand (Carfora et al., 2018; Chen, 2017; Omri, Nguyen, 2014; Pfeiffer, Mulder, 2013; Rafiq et al., 2014; Sadorsky, 2009; Salim et al., 2014). However, fewer studies are investigating renewable energy transition from the perspective of renewable energy share in the total energy mix (Aguirre, Ibikunle, 2014; Cadoret, Padovano, 2016; Lin et al., 2016; Lin, Omoju, 2017; Marques et al., 2010). The energy mix of a country is a combination of different primary energy sources that forms its total primary energy supply. In other words, it is a country's total electricity generation, which comprises power produced from fossil fuels, nuclear and renewables.

Aguirrea and Ibikunle (2014) maintain two fundamental reasons for choosing a percentage of renewable energy instead of renewable energy generation in their study. First, energy policy targets related to the renewable energy transition usually focus on achieving a certain share of renewable energy in the total energy mix. For instance, the European Union aims to produce 32% of energy from renewables by 2030 and achieve net-zero emissions by 2050. Secondly, climate policy aims to displace fossil fuel electricity generation by renewable non-polluting technologies or to decarbonize the energy system. The achievement of these goals will reflect as an increase in renewable energy share in the total energy mix.

Lin and Omoju (2017) hold the same opinion regarding the choice of RE share as a dependent variable. Authors prove empirically that determining factors have a diverse effect on non-hydro renewable energy generated and its share in total energy supply. Economic development has a positive effect on renewable energy power but affects its share. The authors consider that economic development increases not only the renewable energy generation but to a greater extent, it strengthens the generation of traditional energy sources. As a result, renewable energy generation rises but its share in energy supply declines.

Thus, the share of renewable energy should be included in the study as a dependent variable because the impact of chosen factors may vary, subject to the measuring unit of RE generation. In this research work, like in the study by Lin and Omoju (2017), the level of renewable electricity generated and the share of renewable electricity in total energy supply will be applied as explained variables.



1. Factors influencing renewable energy transition

A great number of renewable energy studies was written in the last few years. Among them stands out a range of research works, which focus on factors driving renewable energy transition (Aguirre, Ibikunle, 2014; Cadoret, Padovano, 2016; Chen, 2018; Lin et al., 2016; Lin, Omoju, 2017; Marques, 2010; Omri, Nguyen, 2014; Pfeiffer, Mulder, 2013; Rafiq et al., 2014; Sadorsky, 2009). The authors usually maintain several groups of factors in analyzing renewable energy transition: economic, environmental, political, social, and others. In this paper, variables are divided into three groups: economic factors, environmental and variables reflecting a country's energy dependency. Table 1 illustrates the division of factors analyzed in the previous papers and current study.

Table 1 The division of determinants used in the analyzed literature

Previous studies	Variables
	Economic	Environmental	Fossil fuel dependency
	GDP per capita	Trade openness	CO2 emissions	Oil price	Gas price	Coal price	Natural resources dependence
Aguirrea, Ibikunle (2014)	+	?	+	+	+	+	?
Alam, Murad (2019)	+	+	?	?	?	?	?
Cadoret, Padovano (2016)	+	?	+	?	?	?	?
Omri, Nguyen (2014)	+	+	+	+	?	?	?
Lin et al. (2016)	+	+	?	?	?	?	+
Lin, Omoju (2017)	+	+	?	+	?	?	+
Marques et al. (2010)	+	?	+	+	+	+	+
Pfeiffer, Mulder (2013)	+	+	?	?	?	?	?
This study	+	+	+	+	+	+	+

Economic factors in the previous research works often include GDP per capita, financial development, foreign direct investments, trade openness, and oil price. To reveal the aggregate effect of fossil fuels, some authors comprise gas and coal prices in the analysis (Aguirre, Ibikunle, 2014; Marques et al., 2010). The level of carbon dioxide emissions usually considers as an indicator describing the environmental situation of the country. It is generally used in per capita terms as the measure of environmental concern among the population. Some research works mainly focus on political factors such as the level of corruption, government ideology, the type of government system, certain measures of energy policy and institutional support like the ratification of the Kyoto protocol or commitment to green energy policy or target (Aguirre, Ibikunle, 2014; Cadoret, Padovano, 2016). While other renewable energy studies try to capture the influence of lobby effect on the renewable energy transition, or in other words, the impact of country's dependence on its natural resources (Lin et al., 2016; Lin, Omoju, 2017; Marques et al., 2010).

1.1 Economic factors

All factors listed in the previous paragraph are accurately described underneath. GDP per capita often use as an indicator of a country's economic development. In the greater part of previous works, authors empirically confirm the assumption that GDP per capita affects positively on renewable energy (Sadorsky, 2009; Pfeiffer, Mulder, 2013; Omri, Nguyen, 2014; Lin et al., 2016; Lin, Omoju, 2017). Lin and Omoju (2017) reveal a diverse impact of economic development on the level of renewable power generated and its share in total energy generation. Higher economic development leads to a higher level of renewable electricity produced, but at the same time, the share of renewable power in the total energy mix will diminish as energy generation from fossil fuels would increase.

Trade openness often represents as an economic factor, reflecting the rate country's trade intensity with other economies. In theory, this factor should facilitate the movement of goods and services between countries and, particularly, the exchange of renewable technologies. Thus, the more opened the country is in the international market, the higher level of energy supply it will derive from renewable energy sources. However, the results from the previous research works show that the impact of trade openness on renewable energy transition is contradictory. The outcomes of Omri and Nguyen study (2014) show that trade openness has a positive effect on renewable energy consumption, except for high-income sample of countries. While other authors find that with strengthening trade openness, the probability of renewable energy adoption lessens (Pfeiffer, Mulder, 2013; Lin et al., 2016).

Foreign direct investments (FDI) have a similar effect on the development of renewable energy. This indicator is measured as the ratio of foreign investment to GDP. Theoretically, FDI should enhance technology and knowledge transfer, which would encourage renewable electricity generation. Nevertheless, the great part of authors reveals an undermining or insignificant effect of this factor on renewable energy growth (Lin et al. 2016; Pfeiffer, Mulder, 2013; Lin, Omoju, 2017). This research excludes FDI from the analysis as it partially accounts through trade openness variable.

Financial development specifies the country's finance market. It is supposed that a higher level of a country's financial development facilitates the development of renewable energy by allocating financial resources to renewable energy projects. Several studies find that financial development has a positive impact on both renewable energy share and renewable energy generation (Lin et al., 2016; Lin, Omoju, 2017).

1.2 Environmental factors

Environmental factor, which widely uses in modern literature, is carbon dioxide (CO2) emissions. The significant positive impact of CO2 emissions on renewable energy is empirically proven in a range of studies (Aguirre and Ibikunle, 2014; Chen, 2018; Omri and Nguyen, 2014; Rafiq et al., 2014; Cadoret and Padovano, 2016). All these papers reveal that level of CO2 emissions is one of the major drivers of renewable energy adoption. Besides, Marques (2010) find the negative influence of carbon dioxide on renewable energy share on the sample of 24 European countries. Most of the papers note that developing economies, in general, emit more CO2 emissions than developed countries. Hence, this factor may have a significant impact on renewable energy transition, especially in emerging economies within the current study.

1.3 Other factors

Some studies include fossil fuels prices to measure the effect of fossil fuel dependency impeding renewable energy transition (Aguirre, Ibikunle, 2014; Lin, Omoju, 2017; Marques et al., 2010; Omri, Nguyen, 2014). It is considered that renewable energy is a substitute for crude oil and oil refining products, coal, and gas. In theory, the rise in fossil fuel prices will lead to an increase in energy generation from renewable energy sources. Most authors prove the assumption that the rise in crude oil prices facilitates renewable energy growth (Aguirre, Ibikunle, 2014; Lin, Omoju, 2017). However, Marques (2010) reveal a negative impact of oil price on RE on the sample of 24 European countries. Nevertheless, the same study affirms the existence of a positive and statistically significant relationship between gas and coal prices and renewable energy growth.

Most investigated renewable energy studies focused mainly on the economic factors of RE growth. However, some papers explore the effect of political factors starting with the form of government organization and the ideology of parties in the parliament and ending with the effect of certain energy policy on RE development. Nevertheless, there is a barrier of including political data on the research. Mostly because energy policy statistics contain data limited in a period and countries included, that makes it challenging to collect panel data. Thereby, this study does not analyze political variables.

Several works (Lin et al., 2016; Lin, Omoju, 2017; Marques et al., 2010) investigate the country's dependency on its' natural resources as a barrier on the way to support renewable energy development. This phenomenon is also known as the lobby effect. To capture this outcome, Marques (2010) and Lin (2016) use the share of fossil fuel in total energy consumption. They reveal an adverse impact of this factor on renewable energy adoption. Lin and Omoju (2017) use the share of natural resources rent in GDP as an indicator of countries' dependence from fossil fuels. However, in contrast to previous studies, the authors investigate no significant negative impact on renewable energy generation.

Majority of studies on renewable energy analyze renewable energy on the sample of Organization of Economic Cooperation and Development (OECD) countries (Aguirre, Ibikunle, 2014; Cadoret, Padovano, 2016; Marques et al., 2010) and some emerging economies like India and China (Chen, 2018; Lin et al., 2016; Rafiq et al., 2014). Nevertheless, the determinants of the renewable energy transition in industrialized and emerging countries could differ. In this instance, analyzing the chosen factors on the global sample could lead to inaccurate results and distort the real effects. That is why the division of a sample into income groups is coarse as countries are more heterogeneous in the level of prosperity.

Only a few papers compare the factors' impact on different groups of countries (Lin, Omoju, 2017; Omri, Nguyen, 2014). Omri and Nguyen (2014) divide the global sample of economies into three groups by income: low-, middle- and high-income groups. The results of the study show that changes in economic factors affect differently on renewable energy consumption that depends on the countries' wellbeing. Hence, dividing the sample into several subsamples by income level, and separately analyzing them may prevent incorrect results.

The comparative results of the literature review are reflected in Table 2. Besides, several prerequisites have been determined. Firstly, like in the research of Lin and Omoju (2017), this study is focusing on non-hydro renewable energy generation and the share of non-hydro renewable energy in total generation. Thus, the first hypothesis sounds as follows: the coefficients for non-hydro renewable energy generated and non-hydro RE share differ for the same factors. Secondly, as the effects of driving factors can vary across countries with unequal incomes, the sample of countries is divided into three groups by income. The second hypothesis states that the effects for various income groups will differ. Thirdly, it is supposed that gas prices will influence renewables transition more intensively in developed economies like European countries, whereas an impact of coal prices will be more substantial for economies at the stage of transition. Thus, the last hypothesis asserts that the influence of energy prices on renewable energy varies depending on the country's income group.

Table 2 The comparative results of literature review on RE growth

Previous studies	Dependent variable	Time period	Division on subsamples	Sample
Aguirrea, Ibikunle (2014)	Contribution of renewable energy to energy supply (%)	1990-2010	--	38 high-income
Alam, Murad (2019)	Renewable energy use (TJ)	1970-2012	--	25 OECD countries
Cadoret, Padovano (2016)	Share of RE in gross final energy consumption (%)	2004-2011	--	26 European
Chen (2018)	Renewable energy consumption growth (%)	1996-2013	Eastern, central and western regions	30 provinces in China
Omri, Nguyen (2014)	Renewable energy consumption (KWh)	1990-2011	High-income, middle-income, low-income, global	64 countries
Lin et al. (2016)	Share of RE in total final energy consumption (%)	1980-2011	--	China
Lin, Omoju (2017)	Non-hydro RE (Kwh) Share of non-hydro RE (%)	1980-2011	--	46 countries
Marques et al. (2010)	Logged share of contribution of renewables to energy supply (%)	1990-2006	EU-members, non-EU-members	24 European
Pfeiffer, Mulder (2013)	Renewable electricity generation (kWh per capita)	1980-2010	--	108 developing
Ra?q et al. (2014)	Renewable electricity generation (kWh)	1972-2011	--	China and India
Sadorsky (2009)	Renewable energy consumption (kWh)	1994-2003	--	18 emerging
Sikder et al. (2019)	Renewable energy consumption (KWh)	1991-2013	--	G20 countries
This study	Non-hydro RE (Kwh) Share of non-hydro RE (%)	1988-2014	High-income, middle-income, low-income	51 countries



2. Data

This study utilizes annual data on non-hydro renewable electricity generated and renewable energy share in energy mix on a sample of 51 industrialized and emerging economies over the period 1988-2014. The time series of the panel began in 1988, the starting year of recording data for coal prices, and ends in 2014 as it is the last year when World Bank contains data for the bulk of countries. Countries with missing data and zero levels of renewable energy generation for more than 40% of the considered period both are excluded from the sample.

Selected countries make up groups by income, like in the study Omri and Nguyen (2014). They are the high-income group, middle-income group, and low-income group. The countries' division into subsamples is carried out following the World Bank list of economies for 2014. Unlike the previous study, the middle-income group in this report contains upper-middle-income economies of the World Bank classification. Similarly, the low-income group includes lower-middle-income countries with the addition of Kenya. This change in classification is done because, in the period from 2011 to 2014, some countries received the status of lower-middle-income economies, and the sample of low-income countries in 2014 contained only Kenya. According to new classification, the high-income group (1) consists of 26 countries, middle-income (2) contains 16 countries, and low-income (3) - 9 countries. Appendix 1 illustrates the membership of each country to income groups.

The dependent variables are non-hydroelectricity generated, measured in gigawatt-hours (GWh), and the share of non-hydroelectricity in energy supply (in %). Figure 1 illustrates non-hydro renewable power generation in four samples of countries. According to the graph, non-hydro RE generation increases over time in all income groups. Three graphs represent non-hydro renewable energy generated in each income group separately (Figures 2-4). There is an upward trend in green energy production almost in all economies. Simultaneously, the level of electricity production varies significantly within income groups. In some nations, non-hydro renewable power generation is minimal, while in others, it is tremendous. Appendix 2 illustrates non-hydro renewable electricity generated by the country. Similarly, Appendix 3 displays the share of non-hydro renewable electricity in countries' energy supply.

Fig. 1. Cumulative non-hydro RE generation by income group (GWh)

Fig. 2. Non-hydro RE generation in high-income countries (GWh)



Fig. 3. Non-hydro RE generation in middle-income countries (GWh)

Fig. 4. Non-hydro RE generation in low-income countries

There are three groups of independent variables: economic factors, environmental factors, and factors reflecting energy dependency. Economic factors in this study contain GDP per capita and trade openness. The study does not include FDI and financial development because, for a significant part of countries, data on these factors is available since 1990 and later. Otherwise, that would have significantly reduced the sample. Among non-economic factors, there are the amount of CO2 emissions per capita, energy prices including oil, gas and coal prices, and the factor reflecting the country's natural resources dependence. Table 3 contains the definition, designation, and units of measurement of all considered variables. Whereas the descriptive statistics for each variable and income group are reported in Table 4.

GDP per capita is an economic variable, measured in US dollars. It represents the level of economic development and people's capacity to invest in renewable energy technologies. It is expected that higher per capita income stimulates people to invest in environmental protection. As a result, non-hydro renewable electricity generation is supposed to rise. Data on GDP per capita is obtained from the World's Bank database of World Development Indicators.

Trade openness in this study is the average share of export and import trade in GDP, measured in percent. Trade openness characterizes the intensity of trade in a specific country. An increase of trade should lead to the rise in technological and knowledge transfer, which will cause the growth in non-hydro renewable power generation. Data on trade openness is calculated by taking an average of the sum of the exports and imports shares, collected from the World Bank database.

The rate of carbon dioxide emissions is an indicator of environmental conditions. Higher levels of carbon dioxide emissions are usually a consequence of worse ambient conditions due to the dirty and technologically obsolete industry and transport. In this research, the level of CO2 emissions measured in metric tons per capita. In theory, a higher level of carbon dioxide in the atmosphere may lead to a rise in public environmental concern, which will increase the investments in the green energy sector. Eventually, the generation of non-hydro renewable energy should rise. Data on CO2 emissions is obtained from the World's Bank database of World Development Indicators.

Variable RENT implies a country's dependence on natural resources, which is also known as lobby effect. Like in the study of Lin and Omoju (2017), this variable is measured as the share of natural resources rent in GDP. Rent in this study is defined as the revenue of mining companies from the realization of natural resources, net of production costs. It is supposed that countries, which primarily depend on fossil fuels, are less inclined to support the development of the green energy sector. Thus, a higher share of natural resources rents in GDP should have a negative effect on non-hydro renewable energy growth. Data on the share of natural resources rent in GDP is sourced from the World's Bank database of World Development Indicators.

Oil, gas, and coal prices imply the price of substitutes for electricity generation. In theory, an increase in fossil fuel prices should reduce the demand for traditional electricity generation and possibly could stimulate the generation of non-hydro renewable power. Fossil fuel prices are the same for all countries but change over time, so they represent as time-specific fixed effects. Hence, a rise in fossil fuel prices will lead to the growth of non-hydro renewable energy generation. Oil price is proxied by West Texas Intermediate (WTI) spot prices, gas price by Average German Import gas price, and coal price by Northwest Europe marker coal price, measured in US dollars. All data on fossil fuel prices are sourced from the British Statistical Petroleum Review of World Energy.

Table 3 Definition of the variables used in the study

Variable	Definition	Units of measurement
snhe	Share of non-hydro renewable power in total electricity generation	Percent
nhe	Total non-hydro renewable power generated	GWh
GDP	GDP per capita	US$
TO	Trade openness (average of export and import % in GDP)	Percent
CO2	Carbon dioxide emissions	Metric tons per capita
OIL	WTI spot oil price	US$
GAS	Average German Import price	US$
COAL	Northwest Europe marker price	US$
RENT	Share of natural resources rents in GDP	Percent

The descriptive statistics for the variables in Table 4 represent a variation existing between income groups. Generally, the outcomes received are expected, but few of them require more detailed comments.

As expected, the mean for variable nhe is the largest for the high-income group and the lowest for the low-income sample. However, the high-income group characterizes by a higher standard deviation. That indicates about the gap in renewable energy generation capacities among countries of this group.

Surprisingly, the highest mean of snhe variable relates to the low-income group, not the high-income one, as was expected before. Probably, such a significant average rate of renewable energy production in total electricity supply occurs due to the wide usage of traditional bioenergy for meeting energy needs in these countries. Either it may be due to the entirely low level of electricity access among the population, which accomplishes mostly with new solar energy grids.

The mean of GDP variable is approximately 23 times higher for the group of high-income economies than the mean for the low-income group of nations. These numbers reflect how countries of considered samples differ dramatically regarding per capita income. Analogically with nhe, the standard deviation increases with raising the income of the group.

In the opposite, the variance between the means of TO variable is not too large comparing among income-groups. Trade earnings form roughly 40% of GDP on average for all countries. However, richer economies generally characterized by a higher rate of trade intensity than less developed ones.

The mean of CO2 variable is the highest for high-income economies, and it equals almost zero for the low-income group. Thus, a wealthier group of economies emits the largest amount of carbon dioxide than emerging economies in the determined period. This finding indicates that countries with a higher rate of economic development distinguished by a more extensive amount of carbon dioxide emitted in the atmosphere. carbon energy income trade

Variable RENT has the highest mean for low-income countries and the lowest for high-income. According to the numbers, richer economies are less energy dependent than poorer ones. Probably, the transition to green energy technologies will be more complex and long-lasting in the low-income group of economies.



Table 4 Descriptive statistics for different groups of countries

Variant
Global sample	High-income sample
Variables	Mean	Sd	Max	Min	Mean	Sd	Max	Min
nhe	6.93	21.6	298.02	0.00	9.88	26.12	298.02	0.00
snhe	5.92	7.8	55.85	0.00	5.29	7.09	55.85	0.00
GDP	22 377	22 366	111 968	542	39 600	19 162	111 968	5 397
TO	38.67	29.46	220.8	6.82	43.9	37.17	220.8	8.0
CO2	5.81	5.08	27.43	0.20	9.42	4.72	27.43	1.27
RENT	3.05	6.05	49.45	0.00	1.35	2.97	21.42	0.00
Middle-income sample	Low-income sample
Variables	Mean	Sd	Max	Min	Mean	Sd	Max	Min
nhe	4.03	17.68	229.84	0.00	3.54	8.2	66.93	0.00
snhe	3.61	5.56	25.01	0.00	11.86	9.96	45.73	0.002
GDP	6 014	2 698	13 278	694	1 712	728	3 693	542
TO	35.06	19.37	83.35	6.88	30.02	9.33	57.59	6.82
CO2	2.73	1.17	7.56	0.82	0.86	0.42	2.56	0.2
RENT	4.31	8.0	49.45	0.00	5.72	7.08	36.5	0.31
Invariant
Variables	Mean	Std. Dev.	Max	Min
OIL	44.61	30.42	100.06	14.39
GAS	5.14	3.26	11.6	1.86
COAL	58.55	29.32	147.67	28.79



3. Methodology

3.1 Short-run dynamics

The fixed-effects model is applied to estimate the coefficients of the analyzed variables in the short run. The choice of method is based on the decision to consider country-specific characteristics such as country area or geographical position. The fixed effects method allows to consider specific features of a country, making the results of the study more precise. Nevertheless, some of the considered variables could have a long-run equilibrium (Lin, Omoju, 2017). Without considering this feature, inconsistent and biased results could be produced. A significant part of the recent studies implements panel cointegration models, which consider long-run equilibrium. Among them are fully modified OLS, dynamic OLS, fixed effect vector decomposition, and others. Table 5 reflects more information about the techniques used in previous renewable energy studies.

Table 5 Comparison of methods used in the previous studies

Authors	Dependent variable	Model specification
Aguirrea, Ibikunle (2014)	Contribution of renewable energy to energy supply (%)	Fixed effects vector decomposition, PSCE, Generalized least squares
Alam, Murad (2019)	Renewable energy use (TJ)	ARDL, Pooled Mean Group estimation, Mean Group estimation, Dynamic Fixed effects, FMOLS, DOLS
Cadoret, Padovano (2016)	Share of RE in gross final energy consumption (%)	Least-squares dummy variables (LSDV), fixed effects, GLS-IV
Omri, Nguyen (2014)	Renewable energy consumption (KWh)	Generalized method of moments (GMM)
Lin, Omoju (2017)	Non-hydro RE (Kwh) Share of non-hydro RE (%)	Dynamic fixed effects, fully modified OLS, dynamic OLS
Marques et al. (2010)	Logged share of contribution of renewables to energy supply (%)	Fixed effects vector decomposition, pooled OLS, fixed effects, RE
Pfeiffer, Mulder (2013)	Electricity generation measured (kWh per capita)	Probit, OLS
Sadorsky (2009)	Renewable energy consumption (kWh)	Fully modified OLS, dynamic OLS, OLS
Sikder et al. (2019)	Renewable energy consumption (KWh)	Fully modified OLS, dynamic OLS

In general, the fixed effects model has the following specification:

,

where - dependent variable, I = [1, …, I] - country-specific index, I - number of countries in a sample, t = [1, …, T] - time-specific index (year), T - analyzed time period, - vector of independent variables, - individual-specific fixed effects, is the error term.

In the framework of this study, two econometric models are created for each group of countries, divided by income. In the first model, total non-hydro renewable electricity generated is a dependent variable. While a response variable for the second model is the share of non-hydro renewable electricity in the total energy mix. At first, these two models were created on the global sample of countries, containing 51 industrialized and emerging economies. Then, models with similar specifications are designed for each subsample of countries, divided by the level of income.

The model, which dependent variable is non-hydro RE generated, has the following specification:

,

where nhe - non-hydro renewable electricity generation in GWh; GPD - GDP per capita in US dollars; TO - trade openness; CO2 - carbon dioxide emissions in metric tons per capita; RENT - share of total natural resources rent in GDP; OIL - WTI spot oil price; GAS - Average German Import gas price; COAL - Northwest Europe marker coal price; trend - variable indicating model's time trend.

The second econometric model for the share of non-hydro RE in total energy mix differs from the previous one only by the responding variable:

,

where snhe - the share of non-hydro renewable electricity in total electricity generation.

3.2 Specification for income groups

Then, models with the similar specifications have been designed for each subsample of countries. The specification for non-hydro RE generation in income group j has the following view:

,

where I = {1, … , nj} - number of countries in income group j; j = {1, … , 4} - number of income groups; D - country-specific dummies; nj = {51, 26, 16, 9} - number of country-specific units in each income group.

While the structure of the model for the share of RE generation in income group j looks as follows:

,

where snhe - the share of non-hydro renewable electricity in total electricity generation.

4. Pretesting

Before performing the long run estimation, the pretesting procedure is needed. First, analyzed data should test for stationarity. This test is necessary due to preventing inaccurate and biased results. According to the previous studies, most series are non-stationary in levels, but after taking the first difference data becomes stationary (Alam, Murad, 2019; Lin, Omoju, 2017). Then, cointegration test should prove the presence of a long-run relationship between the responding variables and independent ones. If this relationship exists, further estimation will be conducted.

This study employs several tests for panel data stationarity testing. They are Im-Pesaran-Shin (IPS) test, Levin-Lin-Chu (LLC) test and Hadri LM test. Additionally, Pesaran (PES) test is applied to reveal the existence of cross-sectional dependence among panels. This test complements the results of the previous three ones. If the great part of tests shows the existence of stationarity on first differenced data, then panels potentially could be cointegrated. For cointegration testing, this study implements a bivariate Pedroni test.

There is a slight difference between IPS, LLC and Hardi LM tests. First two tests have a null hypothesis of no stationarity or a presence of a unit root, while the last one has as the null hypothesis that all the panels are stationary.

For cointegration testing of designed panel data models the Pedroni test is applied as it is one of the newest cointegration tests applied in the modern literature (Alam, Murad, 2019; Lin, Omoju, 2017; Sikder, 2019). This test checks for the H0 hypothesis of no cointegration. It was created for panel data cointegration testing by Pedroni in 1999. Since, Pedroni test was highly scored by a scientific community in its implementation accuracy for panel data.

4.1 Long-run dynamics

According to Table 3, there is a vast majority of different techniques to panel data estimation. Recently, the most popular ones are fully modified ordinary least squares (FMOLS) and dynamic ordinary least squares (DOLS) methods. Both methods enable estimating non-stationary panel data with an extended time series period. The main advantage of these methods is an avoidance of endogeneity, serial correlation, and cross-sectional heterogeneity of regressors, making the results more accurate in comparison with standard estimating techniques as pooled OLS or fixed effects. This study uses the dynamic ordinary least squares (DOLS) method for the estimation of coefficients in the long run.

The DOLS is a parametric method, which has been extended to panel analysis by Kao and Chiang (2000). This method corrects the errors by including the past and the future values of the differenced regressors. In general, the DOLS method has the following specification:

,

where q -- the number of leads/lags and - the coefficient of a lead/lag of first differenced explanatory variables.

DOLS estimator is defined as:

,

where = [] is 2(q+1)Ч1 vector of regressors.



5. Results

5.1 Panel unit root tests

Three tests for stationarity have been made on data in levels and the first differenced values. Table 6 shows the result of three tests on stationarity with the addition of the Pesaran test on cross-sectional dependence. In most cases, IPS and LLC test show the stationarity when taking the first difference, while the raw data is non-stationary. The only exception is variable RENT, which shows stationarity without taking the first difference. The Hardi LM test shows that all variables are non-stationary in levels. While the output of first differenced testing reports that nhe, snhe and GDP variables are non-stationary. Finally, the Pesaran test indicates an existence of cross-sectional dependence on both in levels and first differenced data.

Table 6 Panel unit root tests results

Order of integration	Variables	IPS	LLC	Hardi LM	Pesaran
In levels	nhe	-1.963	13.17	65.929***	141.37***
	snhe	-1.855	12.892	62.498***	84.417***
	GDP	-1.851	-0.958	59.863***	148.89***
	TO	-1.96	-6.374***	36.132***	75.773***
	CO2	-2.206	-1.256	52.045***	20.655***
	RENT	-2.693**	-8.136***	41.801***	51.913***
First difference	nhe	-3.003***	-9.62***	16.742***	41.981***
	snhe	-2.273	-15.37***	12.691***	31.659***
	GDP	-2.644**	-15.248***	10.883***	42.945***
	TO	-2.592*	-24.632***	-1.234	59.894***
	CO2	-2.783***	-27.269***	56.592	8.149***
	RENT	-3.124***	-18.316***	-1.995	47.839***

Note: *, **, *** denotes 10%, 5% and 1% signi?cance level, respectively.

Cointegration test

For cointegration, the bivariate Pedroni test is conducted. It shows the existence of the long-run relationship between GDP, TO, CO2 and two responding variables. Only three independent variables are taken for cointegration testing as there is clear evidence of an upward trend among their values, basing on the results of stationarity tests. The outcomes for the Pedroni test are illustrated in Table 7.

Table 7 Bivariate Pedroni cointegration test

Testing for non-hydro RE generation
H0: No cointegration	GDP	TO	CO2
Panel х-Statistic	-6.560534*	-6.60364*	-6.60364*
Panel с-Statistic	-22.352835*	-21.77878*	-21.77878*
Panel t-Statistic (non-parametric)	-27.986406*	-27.33530*	-27.33530*
Panel t-Statistic (parmetric)	-48.470754*	-47.96980*	-47.96980*
Group с-Statistic:	-21.645180*	-21.13864*	-21.13864*
Group t-Statistic (non-parametric)	-31.346662*	-30.47136*	-30.47136*
Group t-Statistic (parametric)	-31.231847*	-30.39713*	-30.39713*
Testing for the share of non-hydro RE generation
H0: No cointegration	GDP	TO	CO2
Panel х-Statistic	-2.619127	-3.170846*	-3.12994*
Panel с-Statistic	-24.825158*	-25.855853*	-24.96727*
Panel t-Statistic (non-parametric)	-28.2961...

Подобные документы

• Peculiarities of economic growth in developed countries at the end of XX-beginning of XXI century

  Defining the role of developed countries in the world economy and their impact in the political, economic, technical, scientific and cultural spheres.The level and quality of life. Industrialised countries: the distinctive features and way of development.
  
  курсовая работа [455,2 K], добавлен 27.05.2015
  

• Differentiation of family incomes

  Priority for the importance of Economy of Ukraine. Sources, functions, structure of income Household as a politico-economic category. Family income - the economic basis of reproduction. Levels of income of the population. The structure of family income.
  
  реферат [22,5 K], добавлен 28.10.2011
  

• Socio-economic development of Ukraine

  Prospects for reformation of economic and legal mechanisms of subsoil use in Ukraine. Application of cyclically oriented forecasting: modern approaches to business management. Preconditions and perspectives of Ukrainian energy market development.
  
  статья [770,0 K], добавлен 26.05.2015
  

• Costs and income of enterprises

  Resources of income for enterprises. Main ways of decreasing the costs Main ways of increasing the income. Any enterprise’s target is to make profit. In order to make it a company should understand where comes from the income and where goes out costs.
  
  курсовая работа [59,9 K], добавлен 09.11.2010
  

• The main principles of safety market economy

  Assessment of the rate of unemployment in capitalist (the USA, Germany, England, France, Japan) and backward countries (Russia, Turkey, Pakistan, Afghanistan). Influence of corruption, merges of business and bureaucracy on progress of market economy.
  
  реферат [15,5 K], добавлен 12.04.2012
  

• The market analysis of natural gas in the European Union

  Natural gas market overview: volume, value, segmentation. Supply and demand Factors of natural gas. Internal rivalry & competitors' overview. Outlook of the EU's energy demand from 2007 to 2030. Drivers of supplier power in the EU natural gas market.
  
  курсовая работа [2,0 M], добавлен 10.11.2013
  

• The basic principles of regulation by safety of economy

  The major structural elements of economic safety of a national economy branches. The structural analysis of economic activity. Share of wages in ВВП, of productivity of Russia and western countries. The essence of the economic taxes and their purpose.
  
  статья [166,3 K], добавлен 12.04.2012
  

• Refugees as a factor of economic development in host countries of migration

  The influence of the movement of refugees to the economic development of host countries. A description of the differences between forced and voluntary migration from the point of view of economic, political consequences. Supply in the labor markets.
  
  статья [26,6 K], добавлен 19.09.2017
  

• Shopping

  Identifing demographic characteristics of consumers shopping in supermarkets. Determine the factors influencing consumer’s way of shopping and the level of their satisfaction (prices, quality, services offered, etc in supermarkets and bazaars).
  
  доклад [54,4 K], добавлен 05.05.2009
  

• The costs of production. Income from factors of production

  Calculation of accounting and economic profits. The law of diminishing returns. Short-Run production relationships and production costs, it's graphic representation. The long-run cost curve. Average fixed, variable, total costs and marginal costs.
  
  презентация [66,7 K], добавлен 19.10.2016
  

• Stereotypes influences on economic relations between the European Union countries and Russian Federation

  Stereotypes that influence on economic relations between the European Union countries and Russia. Consequences of influence of stereotypes on economic relations between EU and Russia. Results of first attempts solving problem. General conclusion.
  
  реферат [19,0 K], добавлен 19.11.2007
  

• The Russian transition to a market system

  Concept and program of transitive economy, foreign experience of transition. Strategic reference points of long-term economic development. Direction of the transition to an innovative community-oriented type of development. Features of transitive economy.
  
  курсовая работа [29,4 K], добавлен 09.06.2012
  

• Factors of production

  Thematic review of the characteristics of each factor of production. The theories of main economists. The possible variants of new factors of production. Labor resources. "Elementary factors of the labour-process" or "productive forces" of Marx.
  
  реферат [437,4 K], добавлен 18.10.2014
  

• Globalization

  The use of computers in education. Improvements in health, education and trade in poor countries. Financial education as a mandatory component of the curriculum. Negative aspects of globalization. The role of globalization in the economic development.
  
  контрольная работа [57,9 K], добавлен 13.05.2014
  

• Predicting completion of cash acquisitions using option implied risk-neutral probabilities

  Estimate risk-neutral probabilities and the rational for its application. Empirical results of predictive power assessment for risk-neutral probabilities as well as their comparisons with stock-implied probabilities defined as in Samuelson and Rosenthal.
  
  дипломная работа [549,4 K], добавлен 02.11.2015
  

• The Development of foreign trade of Ukraine

  Principles of foreign economic activity. Concepts and theories of international trade. Regulation of foreign trade. Evaluation of export potential. Export, import flows of commodities, of services. Main problems and strategy of foreign trade of Ukraine.
  
  курсовая работа [603,8 K], добавлен 07.04.2011
  

• Comprehensive analysis of the economic activity of the company "Turboatom"

  Directions of activity of enterprise. The organizational structure of the management. Valuation of fixed and current assets. Analysis of the structure of costs and business income. Proposals to improve the financial and economic situation of the company.
  
  курсовая работа [1,3 M], добавлен 29.10.2014
  

• Economic Analysis of Italy and United States

  Short and long run macroeconomic model. Saving and Investment in Italy, small open economy. Government expenditure and saving scatterplot. Loanable market equilibrium in closed economy in the USA. Okun’s Law in the USA and Italy, keynesian cross.
  
  курсовая работа [1,6 M], добавлен 20.11.2013
  

• Economy of Kazakhstan

  A variety of economy of Kazakhstan, introduction of the international technical, financial, business standards, the introduction to the WTO. The measures planned in the new Tax code. Corporation surtax. Surtax reform. Economic growth and development.
  
  реферат [27,2 K], добавлен 26.02.2012
  

• Economy of the United States of America

  Socio-economic and geographical description of the United states of America. Analysis of volumes of export and import of the USA. Development and state of agroindustrial complex, industry and sphere of services as basic sectors of economy of the USA.
  
  курсовая работа [264,5 K], добавлен 06.06.2014
  

• главная
• рубрики
• по алфавиту
• вернуться в начало страницы
• вернуться к началу текста
• вернуться к подобным работам