Consumption drivers of renewable energy:
Evidence from the European Union
Master thesis
University of Groningen Faculty of Economics and BusinessZsombor Endrődi
Student number: S2576600 (Groningen)
Email: z.endrodi@student.rug.nl
Supervisor: Dr. Tarek Harchaoui
Dr. Ger Lanjouw
Co-Assessor: Prof. dr. Hans van Ees (University of Groningen)
Prof. dr. András Sugár (Corvinus University of Budapest)
Degrees: Master of Science, International Economics and Business (Groningen)
Table of Content
I. Introduction ... 5
II. Literature Review ... 7
II.1. Literature on drivers of renewable energy ... 7
II.2. Literature on general drivers of investment. ... 9
II.3 The aim of this paper ... 11
III. Data and Model ... 12
III.1. Hypotheses and the model ... 12
III.2. Variables and data sources. ... 14
IV. Econometric implementation and Empirical Results ... 17
IV.1.Estimation Technique and general statistics... 17
IV.2.Econometric Issues and their statistical results ... 19
IV.3 Results and interpretation of the regression ... 24
V. Conclusion ... 26
VI. Limitations ... 27
VII. Reference ... 29
List of Figures and Tables
List of Figures
Figure 1.Investment in Clean Energy vs. Conventional Capacity ... 33
Figure 2.Global Investment Decisions in New Renewables and Nuclear Power ... 33
Figure 3: Histogram for Normality of error distribution of the dependent variable ... 20
List of Tables
Table 1: Descriptive statistics of the variables ... 18Table 2: Correlation between the variables ... 19
Table 3: Modified Wald test for groupwise heteroskedasticity ... 20
Table 4: Wooldridge test for autocorrelation ... 21
Table 5: Breusch-Pagan LM test for independence ... 21
Table 6: Pesaran test for cross sectional independence ... 22
Table 7: Variance Inflator factor (VIF) ... 22
Table 8: Hausman test for panel data ... 23
Abstract
The importance of renewable energy has drawn great global attention recently, which is also indicated by the proliferated treaties and agreements in the 21st century. Several empirical papers use various methods and variables to determine the consumption and investment drivers of renewable energy. However papers fail to include several characteristics that are emphasized by the literature on general drivers of investment. The aim of this paper is to merge together the literature on drivers of renewable energy with the literature on general drivers of investment, and extend previous models, providing better insights of the phenomenon within the EU, thus
improve our understanding of the characteristics promoting renewable energy.
I. Introduction:
The increasing importance of the usage of renewable energy (e.g., wind, solar,
geothermal, hydro, biomass, wave and tidal) has become a major issue dating back to 1911, when in an article of “Scientific American” authors stated that the natural fuels are getting exhausted and urged the need to apply renewable, sustainable energy sources (at that time solar energy was emphasized as the main subject of renewable energy experiments). The issue reemerged with the oil crisis in the 1970s, which resulted in the first investments in electricity generating wind turbines. From that point the significance of renewable energy and investment grew gradually over the years.
This investment tendency in clean energy rapidly recovered from the financial crisis of 2008 and even though investment in renewables decreased in 2012-2013, according to IEA ,“policy uncertainties, economic challenges, incentive reductions and competition from other
energy sources clouded the investment outlook for some markets” (International Energy Agency
Medium Term Market report (2013)), total renewable power capacity continued to grow in 2012, up about 8.5% from 2011. Wind power accounted for about 39% of renewable power capacity in 2012, followed by hydropower and solar PV, each accounting for approximately 26%. These trends continued in 2013 as well (Bloomberg report (2014)).
According to Wüstenhagen and Menichetti (2011) the three most important reasons that draw attention to the phenomenon are 1) the recognition of the depletion of fossil (conventional) fuels, 2) the danger associated with the use of nuclear reactors (Fukushima, Chernobyl) and 3) the commitment of countries to reduce CO2 emission to promote a sustainable future.
Increasing number of studies address the issue of depletion of fossil fuels. For instance, Singh and Singh (2010) argue that fossil fuels could be depleted as soon as 2050. Other studies use exact formulas to predict the depletion of this type of energy source. For instance, in the paper of Shafiee et al. (2009) authors use one recently developed formula, taken from the Klass model, to predict for how long conventional sources can satisfy energy needs. The estimates of these authors for the depletion of oil, coal and gas are 35, 107 and 37 years, respectively1. This
1
means that 30 years from now coal reserves will be the only fossil based energy sources, which strongly emphasizes the significance of renewables.
Issues of nuclear energy have reemerged after the catastrophe of Fukushima in 2011. Opposition of this alternative energy source argues that the disadvantages of nuclear energy far exceed its perceived benefits. For instance, according to Greenpeace UK the problems are:
climate change, safety, energy security, terrorism, cost, reprocessing, nuclear proliferation and radioactive waste. Summing up briefly their remarks; the problems with nuclear energy are not
only the lack of safety due to possible catastrophes and terrorism, but also the cost of building a reactor and dealing with its byproduct created by the reactor during the electricity generating process, which make it an expensive and an inefficient energy alternative. Furthermore, as this alternative emits CO2 as well, it does not provide a real solution for the issues of climate change (e.g. global warming). Arising matters with this type of energy source also point towards the need for an increased role of renewable energy.
The third reason is the commitment of the countries to reduce CO2 emission. As
developed nations realized the problem, they had started to pay increasing attention to this issue and took real actions from the beginning of the 2000s. They have realized that for a sustainable future CO2 reduction is essential. One of the most publicly known agreements is the Kyoto Protocol. This treaty dates back to the planning phase, from the UN Conference on the
Environment and Development in Rio de Janeiro in 1992, to the actual implementation phase2,
coming into force at 2005. It was initially an agreement created by developed nations, later emerging nations joined as well. To reduce CO2 emission the key factor is to reduce the role of fossil based energy sources, which are the main cause of excessive greenhouse gas emission (CO2, N2O etc..). This also highlights the emerging need to change from conventional energy sources to renewable ones.
One additional cause, not mentioned by Wüstenhagen and Menichetti (2011), is the dependence on foreign energy sources. The power of the middle-eastern countries is basically determined by their oil reserves. In the case of the EU it is even more emphasized as significant amount of gas comes from Russia, which is sometimes used as a political weapon, especially
2
against Eastern European EU members. The need of stable and predictable energy supply also highlights the importance of renewable energy sources that promote the country’s self
-supporting and independence.
All these reasons can be categorized under the term of sustainability. According to the
United States Environmental Protection Agency (EPA)“Sustainability creates and maintains the
conditions under which humans and nature can exist in productive harmony, that permit fulfilling the social, economic and other requirements of present and future generations”. As the role of corporates started to be realized, more and more actions were taken by responsible firms, which later became a new phenomenon called CSR (Corporate Social Responsibility).
Due to the fact that the importance of renewable energy has been realized investment also took place, increasing gradually year by year especially from the 2000s. To illustrate this, it is enough to take a look at Figure 1 and Figure 2. Figure 1 in the appendix shows us that investment in renewable energy started to increase from 2004 compared to conventional resources. Figure 2 shows how the role of nuclear energy, renewable energy’s main competitor, started to decrease in 2010 and reached its decade nadir in 20113.
According to the above mentioned information, the theme of my master thesis is the drivers of renewable energy. Within this framework my research question is; what country specific factors affect the use of renewable energy, and how?
II. Literature review:
II.1.: Literature on drivers of renewable energy
The literature on drivers of renewable energy has proliferated significantly in the last couple of decades. One way to categorize them is to differentiate between the countries’ level of development. My analysis focuses on the nations of the European Union, including several developed nations, but also some emerging ones, thus it is important to focus on the literature on developing countries as well. Energy consumption is quite stable in developed economies, while energy demand of emerging economies is growing fast, meaning that increasing the role of renewables is essential for these countries to reduce the rising CO2 emission (Sadorsky (2009)).
3
There are also several studies examining the drivers in developed countries. This is an important issue from various aspects. For example, countries that are member of the OECD were major emitters of greenhouse gases in the 1980s. According to Hamilton and Turton (2000) energy related gas emission in OECD countries grew due to three factors. The increasing GDP per capita, the rising number of the population in these countries and the increase in the amount
of primary energy needed to be delivered for final consumption4. The volume of emission is not
the only reason why investigating developed countries is important, the fact that most renewables are costly is a relevant factor as well. It follows that developed countries have more funds and are also more willing and able to invest money in this alternative. Therefore, drivers of investment are especially important in developed nations. The issue is addressed by authors such as Waldau (2011); Gan and Smith (2011).
Furthermore, novel studies show that investment in renewables has become important in countries that are considered rich in traditional energy sources. For example, the paper of Romano and Scandurra (2013) shows that investment in and electricity consumption from this alternative have risen greatly in OPEC member countries, which are known to be traditionally oil rich and are major exporters of natural energy sources (countries such as; Republic of Iran, Iraq, Kuwait, Saudi Arabia). It is striking as these nations are the main owners of at least 50 % of the world’s proven oil reserves (Sadorsky 2009)), thus they would not consider investing in
renewables unless it is absolutely necessary. This is in line with the findings on depletion of fossil fuels, which I had already mentioned at the introduction part. Nations realized that depletion of conventional sources is no longer a future threat, but an academically proven issue that should be tackled cautiously.
Literature can be scrutinized also by distinguishing between financial and non-financial drivers of investment. Regarding the financial drivers, studies tend to focus on a cost benefit analysis. As an example, in the paper of Böhme et al (2008) authors state that German policy makers ought to support investment in renewables. According to their cost benefit analysis it would provide 9.4 billion euro net benefit in the EU. From this amount 5 billion euro would be due to efficiency-increase, 1 billion euro due to the decreased amount of imports and 3.4 billion euro due to lower pollution. The paper also draws attention to the fact that studies frequently fail
4
to take into account the efficiency increase and lower pollution as benefits. It is realized by most studies that these perceived benefits (energy security, jobs) and social costs (transport and conversion losses) are hard to be measured.
Another cost benefit analysis by Krozer (2012) approaches from a policy perspective. He intends to show whether the use of renewable energy for electricity production in the EU is beneficial or not, throughout the cycle of low oil prices 1998-2002 (demand for renewables are low) and high oil prices 2003-2009 (demand for renewables are high). In this study Krozer highly supports investment in renewable energy and as a conclusion he argues that during the period of 1998-2008 the EU had 41 billion euro net benefit annually. Therefore, he encourages EU policies subsidizing plantation of renewable energy assets.
On the other hand, recent literature started to focus on non-financial drivers of renewables as well. As renewable energy is still a costly and relatively uncertain investment type these drivers are also highly relevant. Masini and Menichetti (2012) take into account factors like a priori believes on investment, institutional pressure, attitude towards radical technological innovations and knowledge of the operational context. Results show that several of these cognitive and behavioral elements have great influence on investment in and thus consumption from renewable sources.
II.2.: Literature on general drivers of investment.
Although literature on investment in renewable energy has improved significantly regarding the measurement of the phenomenon, there are several factors that are not taken into account in these studies. According to Bausmann and Price (2007) there are general drivers of investment that ought to be included in every study that analyzes this field. Authors highlight four main ingredients. These are 1) adjustment cost of capital, 2) uncertainty, 3) financial constraints and 4) elasticity of substitution (in our case substitution of renewable energy with other alternative energy source).
investment would predict, but it is recovering slowly. Therefore, the adjustment cost captures the long period when the capital is in disequilibrium (Bausmann and Price (2007)). Examples of adjustment costs are; disruption costs, installing new machines that usually affect production process or moving to other production locations.
The second important element that is usually left out from the literature is the country level uncertainty affecting the promotion of renewable energy. Country level risk is measured by various global rating agencies e.g. Standards & Poors, Moody’s or Fitch. These institutions take into account various aspects to decide on the annual rating, including the level of democracy, attitude towards foreign investors etc… Additionally, the level of corruption recorded by for instance, the Transparency International is sometimes used by authors as a proxy to measure uncertainty in a specific country, as some empirical evidence support this, although the two things are not the same. After the financial crisis in 2008 the literature on uncertainty proliferated largely and is still one of the major concerns for authors nowadays (Marfatia 2014, Carrière-Swallow and Céspedes 2013, Calmés and Théoret 2014). One of the outstanding studies has been written by Bloom (2007) who connects uncertainty with the adjustment cost and examines the effects post crisis. He analyzes how output, productivity and employment dropped due to uncertainty after catastrophes such as the Cuban missile crisis (1962), assassination of JFK (1963), OPEC oil price shock (1973) and the 9/11 (2001) terrorist attack. He comes to the conclusion that uncertainty is the highest after these shocks, and that drops are caused by the firms’ temporary decline in investment, hiring and reallocation across units in these times. Another relevant study by Bond et al (2004) investigates the importance of cash flows for UK firms when deciding on investment. In this paper the authors use various measures for uncertainty such as “(1) the volatility in the firm’s stock returns; (2) disagreement among securities analysts,
forecasting the firm’s profits in recent past; and (3) the variance of the errors made by analysts while forecasting the firm’s profits” (Bond et al (2004)). He concludes that all his measures of
The third factor articulated by Bausmann and Price (2007) is the financial constraint. In this paper I presume that country level financial constraints are affecting investment decisions, thus investment in renewables as well. Several authors mention that the level of development of a country’s financial market affects economic growth (Gorodnichenkoy and Schnitzerz (2011)). Literature on this factor is quite thorough, though mainly focuses on micro level as well. For instance, in their paper Wang et al (2014) analyze the mechanisms through which finance affects investment and capital accumulation of companies. They come to the conclusion that larger firms tend to be affected by improved financial conditions, while smaller firms are rather influenced by the financial development.
The fourth factor is the elasticity of substitution. In our case the substitution of renewable energy with other energy alternatives (e.g. conventional sources, nuclear energy def.: how elastic are the prices of alternatives; what level of change in the prices would promote the use of
renewables, vice versa). The previous three factors determined the short run dynamics of investment; however, the elasticity of substitution is only meaningful to policymakers when it holds on the long run.
II.3: The aim of this paper
The aim of this paper is to merge together the existing literature on investment and consumption drivers of renewable energy with the literature on general drivers of investment emphasized by Bausmann and Price (2007). I would implement this by extending the models on drivers of renewable energy by factors from general drivers of investment. In my thesis I will follow the analysis of Marques and Fuinhas (2011), whom examined drivers of renewable energy in 24 European countries and Romano and Scandurra (2013), whom analyzed the investment drivers of the phenomenon in OPEC member countries. However, I will concentrate on members of the European Union. From this integration my analysis includes 18 member countries, where data was available (Austria, Belgium, Czech Republic, Denmark, Finland ,France ,Germany, Greece, Hungary, Ireland, Italy, Luxembourg , Netherlands, Poland, Portugal, Spain, Sweden ,United Kingdom,)
European Union it is useful to examine this group, in order to identify factors affecting
investment in and consumption from renewable sources in this mixed integration. Furthermore, the difference between my paper and the paper of the previously mentioned authors is that
Marques and Fuinhas (2011) included European countries that are not part of the European Union (e.g Iceland, Switzerland and Turkey), which thus does not provide information that could be used for policy analysis for the Union.
From a policy perspective European Union is the most interesting examination field as it is a leader in the fight against climate change. This is confirmed by the creation of the largest multinational experiment of emission trading systems, the European Union Emission Trading
Scheme (which is a carbon market) to fight global climate issues. Meanwhile the U.S.has
adopted the Renewable Portfolio Standard (RPS), which sets standards and requirements at state level. However, Elliott (2013) stated that despite the fact that all American presidents from the mid-1900s highlighted the importance of renewables, it only accounts for 12 % of total electricity generation in America, while in Europe Portugal has around 45% and this pattern is similar in several members of the EU (e.g. Sweden, Germany) (Elliott (2013)). This fact emphasizes the significance of the academic papers focusing on the EU compared to the U.S. based papers in the field of clean energy. Additionally the EU members not only commit themselves to the
promotion of renewable energy by signing the Maastricht treaty, but the community sets the Renewables Directive, which is a common policy for members to reach the goal that by 2020, 20% of electricity consumption in the EU will come from renewable sources.
III. Data and Model
III.1 Hypotheses and the model
Due to the fact that I use consumption to proxy investment, which will be explained throughout the paper, my hypotheses will be concerning the effects of variables on investment in renewable energy, although the main results will reflect the impact on country level consumption of electricity from renewable sources.
As mentioned previously, country level development matters, as some authors already highlighted. Sadorsky (2009) finds “that real income and carbon dioxide emissions are both
important drivers of renewable energy consumption” (Sadorsky (2009)). He concludes in his
energy, even though investment in renewables is more crucial for less developed countries, due to their high CO2 emission. Therefore, my first hypothesis is:
Hypothesis 1: The level of development, measured by GDP, increases investment in renewable
energy
Following Marques and Fuinhas (2011), the prices of other alternatives matter as well. Literature suggests that nations are willing to invest more in renewable energy if the imported energy is getting too expensive, for instance, due to a crisis (e.g. oil crisis in 1973). Therefore, my second hypothesis is:
Hypothesis 2: Investment in renewables is affected by the prices of alternatives, namely natural
gas, oil and coal.
As mentioned within the general drivers of investment, financial constraints enormously affect any kind of investment, including renewable energy (Bausmann and Price (2007)). As financial constraints are a bigger issue in countries where there is less financial market development my third hypothesis is:
Hypothesis 3: Countries with more developed financial market will invest more in renewable
energy sources.
According to Gorodnichenkoy and Schnitzerz (2011) and Wang et al (2014) uncertainty has a significant impact on investment on firm level. Therefore, I assume that the case is identical on a country level. Uncertainty is measured by the level of corruption (detailed explanation in section III.2). High level of corruption in countries also reduces the willingness to invest in general. Therefore, my fourth hypothesis is:
Hypothesis 4: Higher levels of uncertainty will have a significant impact on investment in
renewable energy sources.
Some studies also highlight national security concerns by addressing dependence on foreign natural resources. As emphasized by Huang et al (2007), who analyzed the case in the US, national dependence on foreign resources has risen from the mid-1990s. Strengthening national energy security is a major concern for all countries. Therefore, my fifth hypothesis is:
Hypothesis 5: More dependence on foreign resources positively affects investment in renewable
energy.
Hypothesis 6: Country level CO2 emission positively affects investment in renewable energy.
The basic model on which I performed the test is: LRECONTi,t = +1*GDPi,t + 2*CO2i,t
*COALPi,t 4*GASPi,t 5*OILPi,t 6 * FINDEVi,t 7*CPIi,t 8*SUFFi,t +
i,t
WhereLRECONT is the dependent variable, is the intercept parameter, s are the coefficients of the different variables (drivers of renewable energy) and
is the error term. i and t measure different countries in different time periods respectively.III.2 Variables and data sources
Annual data was gathered from the year 1994 to 2011 for 18 members of the European Union. Excluding Latvia, Lithuania, Croatia, Romania, Bulgaria, Cyprus, Malta, Slovakia, Estonia and Slovenia, due to lack of data availability, and the candidates: Iceland, the former Yugoslav Republic of Macedonia, Montenegro, Serbia and Turkey, as in 2014, at the time of my thesis, these countries are not part of the European Union. Although the literature on investment drivers of renewable energy is useful, my analysis explains investment by consumption as a proxy, thus it is about the consumption drivers of renewable energy. For an appropriate data gathering I scrutinized all available databases, especially energy related statistics (e.g., International Energy Agency, OECD). For other country specific characteristics I used the database of the World Bank. Due to the lack of data availability, proxies were used in some other cases as well to measure the different kinds of variables, which are now explained.
Dependent variable:
For an accurate measurement of investment in renewables, my paper should have been based on reliable time series data on the amount of investment in renewable energy, for a
sufficiently broad time period (in my case 1994-2011). This could be done through examining the country level expenditure, both public and private, on energy related assets. Within this, the precise amount of expenditure on renewable energy assets on current prices (e.g. hydro plants, photovoltaic cells or wind turbines) is required.
information might be useful in countries where it is a significant source of clean energy (biofuels in Germany, photovoltaics in Spain, Portugal), however, it is not sufficient to serve as a
measurement of investment in renewable sources. One special publication that I have found, regarding the Netherlands, is the Green Growth in the Netherlands, published annually by Sjoerd Schenau. Similar publications for other EU countries would be helpful to acquire useful data.
Subsequently to gather information from sources specifically dealing with energy I have turned to the database of International Energy Agency, where they also do not provide
information on the amount of investment in renewable energy or related assets. Although they do have data on the R&D expenditure on renewable energy, the data was only available for the most developed countries of the EU (e.g. United Kingdom, Denmark, Sweden, France, Netherlands, Germany), moreover, the time series was only available for a limited time period and even within that period there had been huge gaps at some dates, where no information was available.
As a last major source I examined the Bloomberg energy reports, published annually. These provide information on the amount of investment in renewables, however, only on a global or regional level. It had data on the US and Europe, however, no further information was
available for the European Union, and most importantly no datasets on individual country level. The lack of data on the countries’ investment level in renewables is confirmed by research papers of several scientific authors of the field, whom used proxies to measure the drivers of renewable energy (both investment and consumption). For instance, in the paper of Romano and Scandurra (2013) authors measured investment in renewables by the ratio of renewable
generation per total net electricity generation. Additionally Marques and Fuinhas (2011) used the contribution of renewables to total energy supply to measure the consumption from renewable sources.
To sum up, at the time of my thesis there is no available data on country level investment in renewables, thus rationality dictates that in my analysis I will use a proxy that has been
developed before. Therefore, I will follow Marques and Fuinhas (2011) and their proxy; contribution of renewables to total energy supply. Although one might fairly argue that
reveals information on the consumption drivers of renewable energy. This variable is used sometimes by authors to proxy investment, as in the work of Romano and Scandurra (2013), however, it is not an academically accepted proxy. Therefore, my dependent variable is:
Investment in renewable energy (LRECONT): Investment in renewables is measured by
contribution of renewables to total energy supply, in percentages. Source: OECD Factbook 2013.
Independent variables:
In the case of the independent variables, most variables measure exactly the subject of my analysis, in other cases such as uncertainty and financial development I used proxies supported by previous literature.
Level of development (GDP): Gross Domestic Product measures the level of development of the
country in billion USD. Source: World Bank Development indicators.
CO2 emission of countries (CO2): This variable measures the CO2 emission of the country
from fuel combustion, by million tons annually. Source: OECD Factbook 2014, OECD Factbook
2005
The price of oil (OILP): This variable measures the price of oil in US dollars/barrel as an
alternative conventional energy resource. Source: Database of International Energy Agency, oil
import prices.
The price of gas (GASP): This variable measures the price of gas in US dollars/million BTU on
average German import price, as an alternative conventional energy resource. Source BP.
Statistical Review of World Energy.
The price of coal (COALP): This variable measures the price of coal in US dollars/ton on North
West Europe market price as an alternative conventional energy resource. Source BP. Statistical
Review of World Energy.
Country level uncertainty (CPI): This variable aims to measure country level uncertainty,
the high correlation between the two using several robustness checks. As a brief explanation they state that “higher uncertainty may shift the economic agents’ focus to the present, induce
‘impatience’ and strengthen the propensity to offer and demand bribes” (Goel and Ram (2012)).
To measure the level of corruption I am going to use the CPI index. The abbreviation stands for corruption perception index. This index is made by Transparency International annually. It measures the public sectors’ perceived corruption in 177 countries. The higher the ranking the less corruption is in the given country, it scales from 1-10. Source: Transparency International
Corruption Perception Index.
Financial development (FINDEV): This variable measures the financial development of the
country. It is based on the work of Lynch (1996), where the author mentions several ways to measure a country’s financial development. One way to express this is the domestic credit to private sector. Domestic credit to private sector refers to the financial resources provided to the private sector by monetary enterprises, through loans, purchases of non-equity securities, trade credits and other accounts receivable that establish a claim for repayment. The monetary
enterprises include monetary authorities, deposit banks; insurance banks, pension funds, leasing companies, companies dealing with foreign exchange etc. Source: World Bank Development
indicators.
Energy Dependency (SUFF): This variable measures total self-sufficiency in energy, which is
an indicator to measure dependence on foreign energy sources. It is measured by total self-production of energy divided by total primary energy supply. Total primary energy supply is made up of; production + imports - exports - international marine bunkers - international aviation bunkers ± stock changes. Source: International Energy Agency (IEA) World Energy Balances
(2013 edition).
IV: Econometric implementation and Empirical Results
IV.1: Estimation Technique and general statisticsTo deal with multiple countries and varying time dimensions I used the panel data
both on time dimension and country dimension. Countries with insufficient data were left out from the analysis, as mentioned at the previous section.
The basic form of panel data is:
Yiti + t +’x
it + it.. Where i =1, 2,., N for individuals (e.g., countries in our case); t = 1, 2, ., T for time; Y= the dependent variable (contribution ofrenewable energy to total primary energy supply); x = independent variables; = coefficient of
each independent variable; = the individual effect;
the time effect; and
= the error term.Within the panel framework there are three estimation techniques that can be used to estimate the results:
1) Pooled regression; method to use: Ordinary Least Squares (OLS): This method is the
basic estimation technique used at cross sectional or time series analysis.
2) Fixed effects model; method to use: Weighted Least Squares (WLS): This type of method accounts for heterogeneity across individuals in a deterministic way, thus making it a more precise estimation technique than the pooled regression.
3) Random effects model; method to use: Generalized Least Squares (GLS): This method also accounts for heterogeneity across individuals, however, in a random way. Within random effects model it is assumed that unobserved variables are uncorrelated with observed ones and we treat the individual differences random, as they were gathered randomly.
Before turning to the possible econometric issues and later deciding on the estimation
technique that I am going to use, I will provide some general statistics and a correlation matrix to get better insights of the parameters. Table 1 shows us the description of the variables.
Table 1: Descriptive statistics of the variables
Variable Obs. Mean Stand.Dev. Min. Max.
From Table 1 we can see the minimum and maximum values of the variables and also that we have 324 observations for 18 members of the European Union, which seems to be enough for a precise estimation.
Table 2: Correlation between the variables
Table 2 provides information on the correlations. We can see that most variables are not correlated to each other; exceptions are the prices of the alternatives. GASP has a 0.71 correlation to OILP and the variable COALP has a 0.86 correlation to GASP. This might indicate that there is multicollinearity, which should be tackled in order to get unbiased results. Correlation matrix in itself, however, does not tell the whole story. It is not surprising that the prices of the
alternatives are in a way related as any changes in them have a huge impact on the other. However, in the next section, among other econometric issues, I will examine whether multicollinearity is an issue or not.
IV.2. Econometric issues and their statistical results
Before running any regressions and deciding on the technique, there are some econometric issues that should be addressed. The most well-known issues of panel data are the issue of normality of error distribution, heteroskedasticity, serial autocorrelation, cross sectional dependence and multicollinearity
Normality of error distribution
Testing the normality of the dependent variable’s error is essential to perform a reliable test statistics. First I am going to examine normality graphically, and then I am going to check the robustness of the visual examination by the Jarque-Bera test for normality.
Figure 3: Histogram for normality of error distribution of the dependent variable
Figure 3 strongly suggests that the errors of the dependent variable are not normally distributed, but the test is needed to approve this.
The Jarque Bera test statistics has a 2 distribution with 2 degrees of freedom, (one for skewness and one for kurtosis). The null hypothesis is: H0: The errors are normally distributed against the alternative H1: The errors are not normally distributed. The critical value of the 2 with two degrees of freedom and at 5 % level is 5.99. The result of the Jarque- Bera test is 27.216791 which means that JB> 2critical . According to the results I reject the H0 and accept the alternative H1, thus the errors of the dependent variable are not normally distributed. To tackle this I am going to use the natural logarithm of the dependent variable.
Heteroskedasticity
The problem of heteroskedasticity means that the variance of the errors differs across
observations(Hill et al (2012)). Although the estimates of the coefficients are not affected by it, it can bias the significance of the F test and t test. To identify whether my model suffers from heteroskedasticity I applied the Modified Wald test for groupwise heteroskedasticity, to test for the issue between the selected countries.
Table 3: Modified Wald test for groupwise heteroskedasticity
Modified Wald test for groupwise heteroskedasticity in fixed effects regression model
2 (18)
1606.96*** *Significant at 10 %, ** Significant at 5%, *** Significant at 1% level
This null hypothesis here is that H0: Homoskedaticity (constant variance) across countries and the alternative hypothesis is: H1: Heteroskedasticity is present across countries. Due to the result of Table 3, I reject H0 and accept the alternative H1 as 2 is significant on 1%, and conclude that the model suffers from heteroskedasticity. To tackle this issue I will use White’s Standard Errors at the final regression. It is essential to include robust standard errors in the regression because without this inclusion, incorrect standard errors tend to understate the
precision of the estimation,; by getting confidence intervals wider then they should be (Hill et al (2012))
Serial autocorrelation
As the same observations are being examined over time, serial correlation is a potential threat to my model. To identify whether the model suffers from serial auto correlation I will perform a Wooldridge test, where the null hypothesis is H0: No first order auto correlation against the alternative hypothesis H1: There is serial auto correlation.
Table 4: Wooldridge test for autocorrelation.
According to the result of the F statistics in Table 4 I reject the H0 and accept the alternative
H1, thus conclude that my model also suffers from first order autocorrelation as expected.
Cross sectional dependence
There are two methods to determine whether cross sectional dependence is apparent, meaning that the residuals across entities are correlated. It is important to address this issue as it can lead to serious biases in the test results. The first one is the Breusch Pagan LM test and the second one is the Pesaran Cross sectional test for independence. I am going to perform both tests to ensure that my results are robust. Both tests share the same null hypothesis H0: The residuals are not correlated against the alternative H1: The residual across entities are correlated
Table 5: Breusch-Pagan LM test for independence.5 Breusch-Pagan LM test for independence
2 (153)
524.711*** *Significant at 10 %, ** Significant at 5%, *** Significant at 1% level
5
For the results of the whole test see Appendix Table 5 Wooldridge test for autocorrelation in panel data
F(1,17) 50.855***
According to the probability value of Table 5 I reject the H0 and accept the alternative H1. Therefore, my model suffers from cross sectional dependence, though I will check it with the other test as well.
Table 6: Pesaran test for cross sectional independence
Pesaran test confirms the Breusch- Pagan test (significant on 5 %), meaning that there is solid evidence of cross sectional dependence in my model. Table 6 also shows that on average there is a 0.365 correlation between the residuals of the two stocks.
Multicollinearity
Multicollinearity exists if two or more variables are near perfect linear combinations of each other (Hill et al (2012)). The main problem with this is that as the level of multicollinearity increases, the estimates of the coefficients become unstable, thus standard errors for the coefficients can get wildly inflated. To determine whether multicollinearity is an issue in my model or not, I will use the variance inflator factor (VIF). As a rule of thumb, multicollinearity is present if the VIF number exceeds 10 or the tolerance (1/VIF) is lower than 0.1. From Table 7 we can see that none of the variables violate the rule of thumb, thus I conclude that multicollinearity is not present in my model.
Table 7: Variance Inflator factor (VIF)
Pesaran’s test for cross sectional independence
Test statistic 2.076**
Average absolute value of the off-diagonal elements 0.365 *Significant at 10 %, ** Significant at 5%, *** Significant at 1% level
Variable VIF 1/VIF
Hausman test
After addressing the potential econometric issues I have to decide on the type of model to be used. As heterogeneity across individuals is a major concern in my analysis the only possible model that I can use within the panel data is either the fixed-effects or the random-effects model. Fortunately, there is the Hausman test that is able to tell which one to choose.
Within this test the zero hypothesis is: H0: The preferred model is random-effects, against the alternative hypothesis: H1: The preferred model is the fixed-effects model
Table 8: Hausman test for panel data
Hausman test is basically a regression of the saved parameters of fixed- effects regression on random-effects regression, from which we get the differences and standard errors to help us
decide on the appropriate method to use. Table 8 shows that the 2
became significant on 5% level, which suggests rejecting the H0 and accepting H1. Thus, Hausman test suggests the use of fixed-effects model.
Variables
Coefficients
Fixed Random Difference
IV.3: Results and interpretation of the regression
To account for the econometric issues determined in the previous section I will use the feasible generalized least squares method, which is an applicable method within fixed effects model and which provides an appropriate estimation even when these econometric issues are apparent in the model.
Table 9: Results of FGLS regression
Independent variables Dependent variable: contribution of renewables to total energy supply (LRECONT)
CPI -0.0023359*** (19.34) SUFF 0.0001021 (1.09) GDP 0.0004711*** (12.51) FINDEV 0.0004711*** (39.99) COALP -0.0074847*** (7.88) GASP 0.0139672*** (14.27) OILP 0.0000879** (2.10) CO2 -0.0004132*** (5.44) _cons 3.900696*** (98.67) Wald2 2469.55*** Number of observations 324
Panels Cross sectional correlation
Correlation Common (AR1) correlations for all panels (0.7844) Numbers in parentheses are absolute values of z-statistics
*Significant at 10 %, ** Significant at 5%, *** Significant at 1% level, Using White’s Standard Errors
GDP immediately turned from negative to positive. Next I will interpret the results and decide on the validity of my hypotheses.
Interpretation of the results
First of all it is important to state that some of my results confirmed the results of previous authors, and some of the results were surprising, considering my previous expectations. The level of development measured by GDP is positive and significant as expected, which confirms the result of Marques and Fuinhas (2011). Therefore, I accept the first hypothesis that the level of a country’s development positively affects investment in renewable energy. This serves as a proof that in general wealthier nations are more able and willing to focus on and invest in this
alternative resource, even if it is an expensive solution. The second hypothesis tested brought some surprising results. It is interesting to see that the price of natural gas, the price of oil and the price of coal affect electricity consumption from and investment in renewables in different ways. While the price of oil and the price of gas have a positive effect on the dependent variable, the price of coal has a negative effect on it. Although as all price levels of alternative resources have significant effect on the dependent variable at least on 5%, I accept the second hypothesis.
The level of financial development has a positive and significant effect on renewable consumption, which highly emphasizes the importance of the inclusion of variables from the literature on general drivers of investment. Therefore, I accept my third hypothesis that financial development of the given country does have a positive effect on electricity consumption from renewable energy sources. The fourth variable shows different results than expected. Results show that the variable of CPI, measuring uncertainty, is negative and significant. Since, it is significant I accept the fourth hypothesis. One possible explanation for this result is that more uncertainty means more yield in some cases, which makes investment and from that consumption desirable in specific sectors. The regression also revealed that dependence on foreign energy sources measured by SUFF is positive, however, it is not significant, which seems surprising. Thus, I reject the fifth hypothesis. One possible explanation for this is that due to the
commitments of EU members to reach a certain level of renewable capacity by 2020, energy dependence is no longer a major driver of renewable energy.
explanation is that some of the biggest CO2 emitters are reluctant to tackle this issue. There are some examples for this not only within but outside the European Union as well. For instance, this is the case with Canada, which despite of being a major greenhouse gas emitter, refused to join the Kyoto Protocol. This result is in contradiction with the findings of Marques and Fuinhas (2011), as these authors came to the conclusion that CO2 emission does not affect investment in renewables. Therefore, I reject the sixth hypothesis that more CO2 emission positively affects consumption from renewable energy sources.
V. Conclusion
As global climate change, depletion of fossil fuels and catastrophes with nuclear power plants gave an increasing challenge to countries and economies in the last decades, more attention had been drawn to the role of renewable energy. This type of alternative energy source does not only address these concerns, but also secures a country’s energy source, thus reducing the dependence and vulnerability from traditionally resource abundant nations.
In my analysis I was examining the drivers of renewable energy in 18 members of the European Union. I included those countries where data availability was sufficient to provide an appropriate estimation of the phenomenon. However, as in my analysis I had to proxy investment by consumption, my results are only valid for electricity consumption from renewable sources and it is not necessarily true for investment, despite consumption being the most reliable proxy at this time.
The main novelty of this paper is that it does not solely take into account factors that are usually considered drivers of renewables, but it includes factors that generally affect investment of any kind. For this reason, I included variables measuring the given country’s financial
development and country level uncertainty. Moreover, my thesis concentrates on the members of the European Union not on Europe as a whole, as most literature concentrating on this region does. My study also explains the rationale behind using the mentioned proxy to measure investment in renewable energy.
Marques and Fuinhas (2011) applied a different method. They used dynamic panel
the importance of CO2 emission. These authors emphasize that CO2 emission is not a relevant driver of renewable consumption; however, my results rather confirm previous literature where other authors argue the opposite.
The significance and impact of the variables taken from the literature on general drivers of investment highlighted the importance of these factors that are needed to be taken into account to measure investment in renewables more accurately by approaching from a different
perspective.
VI. Limitations
The most important limitation that should be mentioned at the first place is the lack of data on investment. As the quality and quantity of data improves over time, future analysts will have the opportunity to measure investment as it is, not by proxies that I had to turn to, due to lack of data availability. From there it will be a breakthrough to see whether new results confirm the results of previous literature that had to use proxies for the measurement of the phenomenon. Furthermore, novel studies with better data will also reveal whether these proxies were
sufficiently explaining the phenomenon or not.
This paper introduced several explanatory variables to increase the precision of
estimations of already existing models, although there are still gaps that can be filled by future analysts of the field.
One factor that was mentioned at the literature review section of the paper is the adjustment cost of capital. There have been some studies trying to measure cost of capital, but managed to do only at firm level (Bayer 2006; Gottai and Mastrolia 2014). There is a huge need for literature that accounts for adjustment cost of capital on a country level basis, thus estimations that take into account variation in time could provide more precise results. A formula that takes into account the country specific factors that enhance or impede the cost of capital to return to its equilibrium level after certain shocks would be a valuable tool that would improve existing estimations.
energy alternatives), which as mentioned before, was also not gained a lot of attention at macro level by experts on this field. Elasticity of substitution is especially important as it focuses on the long run dynamics of investment, which makes it also a valuable asset to policy makers, whom according to that can make real strategies to promote investment in renewable energy sources. To sum up, there are plenty of studies focusing on drivers of renewable energy and thus understanding of the phenomenon improved a lot especially after the millennium. However, there is still much work to do, studies should focus more on different aspects of investment and
consumption, applying useful measures and tools from other literature as well, as it is
VII. References
Aguilar.X.F., Zhen Cai.Z., 2010: Exploratory analysis of prospects for renewable energy
private investment in the U.S. Energy Economics, Vol. 32, pp. 1245–1252.
Bausmann, U., Price. S., 2007: Understanding investment better. Insights from recent research. Quarterly Bulletin.
Bayer, C., 2006: Investment dynamics with fixed capital adjustment cost and capital market
imperfections. Journal of Monetary Economics, Vol. 53, pp. 1909–1947.
Blok, K., 2006: Renewable energy policies in the European Union. Energy Policy, Vol. 34, pp.
251–255.
Bloom, N., Bond, S and Van Reenen, J., 2007:‘Uncertainty and investment dynamics’, Review
of Economic Studies, Vol. 74, pp. 391–415.
Bloomberg 2011: New Energy Finance. Bloomberg report 2011. Bloomberg 2014: New Energy Finance Bloomberg report 2014. BP. Statistical Review of World Energy 2014
Brecha.R.J. Mitchell.A. Hallinanc.K. Kissock.K., 2011: Prioritizing investment in residential
energy efficiency and renewable energy; a case study for the U.S. Midwest. Energy Policy, Vol.39, pp. 2982–2992.
Bond, S., Cummins, J., 2004: ‘Uncertainty and investment: an empirical investigation using data
on analysts’ profits forecasts’: Board of Governors of the Federal Reserve System, Finance and
Economics Discussion, Paper No. 20.
Boomsma, T.K., Meade, N., Fleten, S.E., 2012: Renewable energy investments under different
support schemes: A real options approach. European Journal of Operational Research, Vol. 220, pp. 225-237.
Böhme, D., Dürsschmidt, W., 2008:.Renewable energy resources in figures; National and
International Development. Nature Conservation and Nuclear Safety; pp. 29.
Calmès,C., Théoret, R., 2014: Bank systemic risk and macroeconomic shocks: Canadian and
U.S. evidence. Journal of Banking and Finance, Vol. 40, pp. 388–402.
Carrière-Swallow, Y., Céspedes, L.F., 2013: The impact of uncertainty shocks in emerging
economies. Journal of International Economics, Vol.90, pp. 316–325.
Chirinko, R. S., 1993: ‘Business fixed investment spending: modeling strategies, empirical
results, and policy implications’. Journal of Economic Literature, Vol. 31, pp. 875–911.
Detert, N., Kotani, K., 2013: Real options approach to renewable energy investments in
Mongolia. Energy Policy, Vol.56, pp. 136–150.
Donovan, C., Nunez, L., 2010: Figuring what’s fair: The cost of equity capital for renewable
Elliott. D.E., 2013: Why the United States Does Not Have a Renewable Energy Policy. Yale
Law and Economics Research Paper, No. 433
Ellis, C., Groth, C., 2003: ‘Long-run equilibrium ratios of business investment to output in the
United Kingdom’, Bank of England Quarterly Bulletin, pp. 177–187.
Ellis, C., Price, S., 2004:‘UK business investment and the user cost of capital’, Manchester
School, Vol. 72, pp. 72–93.
Gan, J., Smith, C.T., 2011: Drivers for renewable energy: A comparison among OECD
countries. Biomass Bioenergy, Vol.35, pp. 4497-4503.
Gottai, G., Mastrolia, S., 2014: Cost of Capital for Exempt Foreign Private Issuers: Information
Risk Effect or Earnings Quality Effect? It Depends. The International Journal of Accounting, Vol. 49, pp. 190–220.
Greenpeace, UK 2014: Nuclear Power: The problems
<http://www.greenpeace.org.uk/nuclear/problems> Last accessed 2014 Aug 13.
Groth, C., 2005: Calibrating capital adjustment costs in the New-Keynesian model. Economics
Letters, Vol. 93, pp. 26-30.
Groth, C and Khan, H., 2007:‘Investment adjustment costs: an empirical assessment’. Journal
of Money, Credit and Banking, Vol.42, No.8.
Gochoco-Bautista, M.S., Sotocinal, N.R.,Wang.J., 2014: Corporate Investments in Asian
Markets: Financial Conditions, Financial Development, and Financial Constraints. World
Development, Vol. 57, pp. 63–78.
Goel, R.K., Ram. R., 2012: Economic uncertainty and corruption: evidence from a large
cross-country data set. Applied Economics, Vol. 45, pp. 3462–3468
Gorodnichenkoy, Y., Schnitzerz, M., 2011: Financial constraints and innovation: Why poor
countries don't catch up? Journal of European Economic Association, Vol. 11, pp. 1115–1152.
Hamilton, C., Turton, H., 2000: Determinants of emissions growth in OECD countries. Energy
Policy, Vol. 30 (1), pp. 63–71.
Heal, G., 2009: The Economics of Renewable Energy. NBER Working Paper No. 15081. Hill, R.C., Griffiths, W.E., Lim, G.C., 2012: Principles of Econometrics, Asia: John Wiley and
Sons Pte Ltd.
Huang, M.Y., Alavalapati, J.R.R., Carter, D.R., Langholtz, M.H., 2007: Is the choice of
renewable portfolio standards random? Energy Policy, Vol. 35, pp. 5571–5575.
International Energy Agency 2013: Market Trends and Projections to 2018. Medium-Term
Market Report 2013
International Energy Agency 2013 World Energy Balances (2013 edition).
Lynch, D., 1996: Measuring Financial sector development: A study of selected Asia Pacific
countries. The Developing Economies, Vol.34 (1), pp. 1-33.
Mahesha, A., Shoba Jasmin, K.S., 2013: Role of renewable energy investment in India: An
alternative to CO2 mitigation. Renewable and Sustainable Energy Reviews, Vol.26, pp. 414– 424.
Marfatia, A. H., 2014: Impact of uncertainty on high frequency response of the U.S. stock
markets to the Fed’s policy surprises. The Quarterly Review of Economics and Finance, Vol.54, pp.382-392
Marques, A.C., Fuinhas, J.A., 2011: Drivers promoting renewable energy: A dynamic panel
approach. Renewable and Sustainable Energy Review, Vol.15, pp.1601-1608.
Marques, A.C., Fuinhas, J.A., Pires Manso, J.R., 2010: Motivations driving renewable energy
in European countries: A panel data approach. Energy Policy, Vol.38, pp.6877–6885.
Masini, A., Menichetti, E., 2012: Investment decisions in the renewable energy sector: An
analysis of non-financial drivers. Technological Forecasting and Social Change, Vol.80, pp.510-524.
Ming, Z., Ximei, L., Yulong, L., Lilin, P., 2014: Review of renewable energy investment and
financing in China: Status, mode, issues and countermeasures. Renewable and Sustainable
Energy Reviews, Vol.31, pp. 23–37.
OECD Library database: OECD Factbook; 2005, 2013, 2014,
Rafiq, S., Alam, K., 2010: Identifying the Determinants of Renewable Energy Consumption in
Leading Renewable Energy Investor Emerging Countries. In: 39th Australian Conference of
Economists (ACE 2010).
REN21 2013: Renewables 2013 Global Status Report. Renewable energy policy network for the
21st century.
Romano, A.A., Scandurra, G. 2013: Investments in renewable energy sources in OPEC
members: a dynamic panel approach. Munich Personal RePEc Archive, Paper No. 50870.
Sadorsky, P., 2009: Renewable energy consumption and income in emerging economies.
Energy Policy, Vol.37, pp. 4021–4028.
Schenau,S., 2011: Green growth in the Netherlands. Statistics Netherlands. ISBN:
978-90-357-2030-5.www.cbs.nl. Last accessed 2014. Aug. 13.
Shafiee, S., Topal, E., 2009: When will fossil fuel reserves be diminished? Energy Policy, Vol.
37(1), pp. 181-189.
Shapiro, M., 1986 :‘The dynamic demand for capital and labor’, Quarterly Journal of
Economics, Vol.101, pp. 513–42.
Singh, B.R., Singh,O., 2012: Global Trends of Fossil Fuel Reserves and Climate Change in the
21st Century, Fossil Fuel and the Environment, Dr. Shahriar Khan (Ed.), ISBN:
978-953-51-0277-9, InTech, DOI: 10.5772/38655. Available from:
Summers, L., 1981: ‘Taxation and corporate investment: a q-theory approach’, Brookings
Papers on Economic Activity, Vol.1, pp. 67–140.
Transparency International. Corruption Perception Index
http://www.transparency.org/research/cpi/overview , Last accessed 2014. Aug. 13
The World Bank database: World development Indicators. http://data.worldbank.org/data-catalog/world-development-indicators , Last accessed 2014. Aug. 13.
United States Environmental Protection Agency: http://www.epa.gov/cleanenergy/ Last
accessed 2014. Aug. 13
Waldau, A.J., 2005: Photovoltaics and renewable energies in Europe. Renewable and
Sustainable Energy Reviews, Vol.11, pp. 1414–1437.
Wang, J., Gochoco- Bautista, M.S., Sotocinal, N.R., 2014: Corporate Investments in Asian
Emerging Markets: Financial Conditions, Financial Development, and Financial Constraints
World Development Vol. 57, pp. 63–78.
Wüstenhagen, R.,.Menichetti, E., 2011: Strategic choices for renewable energy investment:
VIII. Appendix
Figure 1.Investment in Clean Energy vs. Conventional Capacity
Figure 2.Global Investment Decisions in New Renewables and Nuclear Power
Table 5: Breusch Pagan test for cross sectional dependence
*Significant at 10 %, ** Significant at 5%, *** Significant at 1% level Correlation matrix for residuals:
e1 e2 e3 e4 e5 e6 e7 e8 e9 e10 e11 e12 e13 e14 e15 e16 e17 e18
e1 1.00 e2 0.58 1.00 e3 -0.26 -0.36 1.00 e4 -0.25 0.08 -0.56 1.00 e5 0.63 0.16 0.23 -0.46 1.00 e6 -0.48 -0.60 0.61 -0.50 -0.09 1.00 e7 -0.70 -0.34 0.15 0.43 -0.45 0.17 1.00 e8 -0.11 -0.47 0.61 -0.77 0.34 0.81 -0.13 1.00 e9 -0.72 -0.56 0.24 0.27 -0.32 0.30 0.94 0.08 1.00 e10 0.64 0.19 -0.29 -0.14 0.41 -0.36 -0.51 0.03 -0.47 1.00 e11 0.32 0.56 0.06 -0.06 0.28 -0.16 -0.24 -0.08 -0.40 -0.01 1.00 e12 0.72 0.47 -0.26 -0.21 0.32 -0.50 -0.58 -0.22 -0.57 0.55 -0.00 1.00 e13 0.22 -0.38 0.01 -0.49 0.42 0.12 -0.30 0.41 -0.12 0.48 -0.24 0.28 1.00 e14 -0.79 -0.66 0.55 -0.16 -0.36 0.86 0.60 0.53 0.66 -0.66 -0.27 -0.73 -0.07 1.00 e15 0.44 0.24 -0.01 -0.55 0.13 0.32 -0.74 0.41 -0.74 0.29 0.13 0.34 0.18 -0.13 1.00 e16 0.23 0.39 -0.46 0.03 -0-31 -0.11 -0.17 -0.25 -0.28 0.24 -0.21 0.50 0.01 -0.24 0.46 1.00 e17 -0.30 -0.28 0.63 -0.48 0.13 0.67 0.15 0.53 0.26 -0.59 0.04 -0.31 -0.06 0.64 0.08 -0.24 1.00 e18 0.87 0.67 -0.29 -0.20 0.50 -0.50 -0.70 -0.16 -0.73 0.69 0.25 0.84 0.14 -0.83 0.45 0.38 -0.28 1.00
