Beruflich Dokumente
Kultur Dokumente
ABSTRACT
KEYWORDS:
Economic growth, renewable energy, (in)efficiency, European Union, data envelopment
analysis, window analysis
INTRODUCTION
1
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
the share of renewable energy in total final energy consumption and by the consequent
decrease of greenhouse gas (GHG) emissions.
Renewable energy is collected from solar, wind, rain, biomass and waste,
geothermal, waves and tides. Unlike the conventional energy, it is safe, clean and its
sources (also called renewables) are naturally replenished on a human timescale.
Therefore, its share in total energy consumption has been growing fast worldwide in
recent years and is expected to occupy a leading position. Renewable energy provided
an estimated 19.2 % of global final energy consumption in 2014. The contribution of
modern renewable technologies increased to approximately 10.3 %, while the
traditional use of biomass accounted for about 8.9 %. An unprecedented growth in
renewable capacity and generation occurred in 2015 despite tumbling global prices for
all fossil fuels, ongoing fossil fuel subsidies and other challenges faced by renewables,
including the integration of rising shares of renewable generation, policy and political
instability, regulatory barriers and fiscal constraints (REN21, 2016).
The necessity to reconcile energy demand and implications for the environment
with sustainable economic growth led to the understanding of the importance of
renewables and, consequently, to comprehensive administrative efforts in terms of
EU’s renewable energy policy. These endeavours have made the EU one of the leading
participants in the world in sustainability and renewable energy implementation, and
resulted in various official publications of different EU institutions (EU commission,
EU parliament, Council of ministers, etc.), including white papers, directives,
strategies, communications, reports and resolutions. The use of renewable energy in
the EU is projected to increase substantially by 2020. Although this horizon is still a
long way off, according to the latest reports significant progress has been made since
2005 and the EU is on track to reach its 2020 renewable energy targets.
A comprehensive review of the policy framework for renewable energy in the
EU and an analysis of the progress made by the use of renewable energy are given by
Scarlat et al. (2015) and European Parliament (2017). Although of indisputable
importance, these numerous documents, and the resource-specific issues they
consider, are not the focus of this research. Hence only a short chronological overview
2
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
of the main acts, along with set climate and energy goals and targets, is provided
(Table 1).
3
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
2013 Delivering the internal Communication from the - completely overhauling the
electricity market and commission on the further subsidies that Member States are
making the most of public guidance on renewable allowed to offer the renewable
intervention energy support schemes and energy sector, preferring tendering,
on the use of cooperation feed-in premiums and quota
mechanisms to achieve obligations to commonly used feed-
renewable energy targets at a in tariffs
lower cost
2014 Guidelines on State aid Communication from the - establishing the support by a
for environmental commission on the further competitive bidding process for at
protection and energy- shaping of the new least 5% of the new renewable
related objectives for framework for renewable capacity (in a transitional phase
2014-2020 energy support schemes covering the years 2015 and 2016)
Conclusion
2016 Renewable energy A resolution adopted by the - presenting a more ambitious
progress report European Parliament climate and energy package by
2030 which increases the target for
renewable energy sources to at least
30% to be implemented by means
of individual national targets
Source: Compiled by the authors based on European Parliament (2017) and Scarlat et al. (2015).
This brief survey is not intended to completely cover this voluminous literature,
but to show researchers and practitioners the major EU policy initiatives concerning
renewable energy and its share in total energy consumption. It chronologically covers
all the period since 1997 when the European Commission set the basis for the EU policy
on renewable energy. At the same time, the scenarios contained in these documents
extend even to 2050.
The purpose of this study is to articulate the nexus between economic growth
and renewable energy consumption in the case of EU-28 for the period 1995-2012. As
a cross-country time-series nonparametric analytical tool, DEA is used assuming both
constant and variable returns to scale and input orientation.
The paper is organized as follows. A succinct overview of the literature on the
role of renewable energy in economic growth, using DEA approach, is provided in the
next section. Section 3 briefly describes the chosen socio-economic indicators along
with criteria for their selection, and the data collection, also presenting the DEA
approach setting up methodological framework used in this investigation. Section 4
shows the empirical results of the DEA-based efficiency analysis of EU-28 members.
The findings of the study and general remarks about their implications for
policymakers are summarized in the last section of this paper.
4
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
5
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
6
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
1Although Malmquist productivity index is defined based on the concept of distance function, it can
also be directly represented by DEA efficiency measures (Zhou et al., 2008).
7
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
2.1. Indicators
2 For simplicity, the names of the variables will not be used in their full forms which include the
aforementioned adjustments. For example, unemployment, total (% of total labour force) will be simply
referred to as unemployment.
8
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
for changes in prices and population size during the observed period, thus offering
more appropriate cross-country comparisons and simpler interpretations of the
results. These six adjusted variables will be merged into a single performance scale
which synthesizes separate indicators into one single statistic.
To build a model adequate for capturing the dynamic relationship between each
of the selected indicators and GDP as a standard measure of national economic
performance, with special emphasis on renewable energy-GDP nexus, only GDP is
considered an output while other five indicators are considered inputs.
3This division is based on the International Monetary Fund’s classification of countries into ‘emerging
market and developing’ economies, on one side, and ‘advanced’ economies, on the other. Since some of
the now advanced economies were considered developing until recently, they are classified into a
separate subgroup designated as ‘graduated developing’, regardless of the year in which they became
advanced. With the exception of Cyprus and Malta, these are all transition economies. The rest of the
advanced economies are labelled as ‘developed’.
9
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
10
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
selected inputs and outputs of all analysed entities into a linear program that
represents the DEA model and provides a single relative performance efficiency index.
Because of its unique robust characteristics, DEA has been broadly practiced in a
variety of both profit and non-profit fields, and on both organizational and country
level, over nearly four decades. The DEA concept is based on the seminal work of
Farrell (1957). Since it was initially developed in 1978 by Charnes et al., and up to the
year 2014, about 6,500 DEA papers of both theoretical and practical interest have been
published in ISI Web of Science database (Liu et al., 2016). Thus, this approach is
proven to be considered as an internationally and academically recognized tool for
decision support.
The decisive factor in the selection of this method over traditional benchmarking
techniques for the present study was its possibility of dynamic efficiency
measurement, with inputs and outputs expressed in different and often incongruous
measurement units, and without the necessity of a priori determination of variable
weights or the exact knowledge of the functional form linking inputs and outputs. As
a matter of fact, the weights are set by the model itself, in a manner that maximizes the
efficiency rating for each evaluated entity, thus circumventing the subjectivity
connected with the estimation of each weight’s significance. The production possibility
frontier is constructed by simply enveloping observed inputs from below and outputs
from above. The DMUs on this ‘best practice’ frontier are classified as efficient (i.e.
benchmarks), compared to the rest of the analysed DMUs, and are all assigned the best
efficiency measure of one (or 100%). The non-frontier DMUs turn out to be inefficient
and are scored somewhere between zero and one, depending on their distance from
the frontier. This inefficiency is a result of input surpluses (i.e. using excessive inputs
at a given output level) and/or output shortfalls (i.e. producing poor output at a given
input level), and can be removed by reaching a model-calculated efficient projection
point of operation on the efficient boundary. Since empirically generated, this
boundary appears as an objectively achievable goal for each inefficient DMU and, at
the same time, serves as the basis for identifying its inefficiency sources and their
amounts, improvement directions and reference DMUs to most directly compare with.
11
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
The basic DEA models are characterized above all by the assumption on returns
to scale (constant or variable) and by the model orientation (to input minimization or
to output maximization). The numerous advanced models, built upon the basic ones,
are distinguished by miscellaneous extensions and relaxations.
Nevertheless, the first issue the analyst must face after the selection of indicators
is to determine the type of returns to scale in order to estimate the production frontier.
In this sense, two basic models most commonly exploited in DEA applications are CCR
(Charnes et al., 1978) and BCC (Banker et al., 1984), named after their authors’ initials.
The first model, under constant returns to scale (CRS) assumption, measures
inefficiencies due to the size of operations and the input-output configuration,
resulting in an overall efficiency score. One of its various theoretical extensions
represents the second model that, under the assumption of variable returns to scale
(VRS), estimates pure technical efficiency which purely reflects managerial
underperformance. The CCR score takes no account of scale effect and is therefore
called the (global) technical efficiency (TE). The BCC model, on the other side, separates
the scale effect thus expressing the (local) pure technical efficiency (PTE). Particular
previous studies and preliminary surveys on the process to be investigated and its
properties could provide a first indication on the type of returns to scale and confront
the validity of its assumption. Despite this, the characteristics of the production
frontiers sometimes, as in our application, cannot be determined with certainty, which
complicates the selection of model type. Namely, in these cases, it may be risky to rely
on only one particular model. A potential solution is to try some alternative methods
before arriving at a definitive conclusion, such as using statistical regressions and DEA
to cross check each other. Another possible solution is to try models under each of the
two assumptions, compare the similarity between their results and, based on the
magnitude of their differences and utilizing expert knowledge of the problem, find the
most suitable type of assumption for the analysis. In the thus conducted initial phase
of our research, the results received from the said models appeared to have significant
discrepancies. They may be ascribed to the return effect with respect to the range of
activities, which makes the BCC model more appropriate for describing the
investigated process, and consequently is the reason why it receives somewhat greater
12
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
13
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
Definition 1 (Efficiency):
If an optimal solution + ∗ , λ∗ , - .∗ , - /∗ , obtained in this two-phase process satisfies
∗
= 1 and has no slack +- .∗ = 0, - /∗ = 0,, then the is called efficient, otherwise
it is inefficient.
The provided information can be used as a basis for goal setting for the DMU
being assessed as inefficient. A first step in determining its targets is to compare it with
the efficient DMUs that make up its reference set.
14
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
The constraint (4) distinguishes the BCC from the CCR model, causing notable
differences in their production frontiers. Consequently, in practice, the projection of
any inefficient DMU on the CRS frontier is farther than its projection on the VRS
frontier. CCR efficiency is therefore more difficult to accomplish, resulting in lower
scores. Moreover, it may happen that a DMU is BCC efficient, but has a low CCR score,
which means that it is locally, but not globally efficient due to its scale size. It is
therefore recommended to evaluate the scale efficiency (SE) of a DMU as the ratio of the
CCR and BCC efficiency scores.
Obviously, this number cannot exceed one and, for a CCR efficient DMU, it is
equal to one. This formula reveals an efficiency decomposition as
15
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
In that case data for several periods are incorporated into analysis for each entity,
which is regarded as a different entity in each of the observed periods. Efficient frontier
of the window analysis model is built analogously to the one in the basic model with
the difference in sample size, also depending on the window length which can range
from one year to all available years. If in our case, for example, we choose a window
length of five years, there would be 14 windows (1995-1999, 1996-2000, ..., 2008-2012)
provided by dropping one year and adding another each time. In this way, each of the
years from 1999 to 2008 is tested ten times. On the other hand, the initial and final years
are tested only once. This makes the overview of the obtained results over time
somewhat unbalanced, which could be circumvented by equal treatment of all years
from the observed period. It is possible only in the cases with minimum and maximum
window length. In the first case there are 18 one-year windows, while in the second
one there is one 18-year window. Since the countries are mutually compared only
within and not across windows, the first case implicates 18 sets of 28 entities, while the
second implicates one set of 504 (18 x 28) entities4. The opportunity to compare each
country’s performance in a given sub-period not only to other countries’
performances, but also to its own performance in any of the other sub-periods is of
utmost importance. Therefore, the purpose of this research will be fulfilled by
examining the second of the above cases.
4Among these entities, 270 (18 x 15) correspond to developed, 144 (18 x 8) to graduated developing and
90 (18 x 5) to developing countries.
16
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
Since these valuable facts are available with the basic DEA models, but not with the
window analysis, the new combined models were built to compute them while leaving
the relative efficiency scores unaltered (Rabar, 2013). For brevity, these two models are
hereafter referred to as CCR-I-W1 and BCC-I-W1.
To increase the efficiency of an inefficient entity, the input-oriented DEA models
first tend to reduce its inputs and, if that is insufficient, to enlarge its outputs. Since in
this study gross capital formation and renewable energy consumption are specified as
inputs but, in practice, preferred to have larger amounts, in the model calculation their
reciprocals are taken. This at the same time preserves the relationship between inputs
and outputs and enables the assessment of countries’ performance regarding their
capability to minimize unemployment, energy use and GHG emissions, and to
maximize gross capital formation, renewable energy consumption and GDP.
However, recalculation of the obtained results and an additional caution in their
interpretation are necessitated. The evaluation of the relative performance is computed
by the software package DEA-Solver-Pro 7.0F from the Saitech Company, and further
processed by the authors.
The technical and pure technical efficiency results respectively generated by the
models CCR-I-W1 and BCC-I-W1, along with the calculated scale efficiency results, are
classified by subgroups of countries (developing, graduated developing and
developed). Due to space limitations, it would be far too extensive to analyse each
country in each year, so the focus will be placed on the average efficiency scores that
are provided in Table 3 and Figures 1 and 2 (for detailed results see the supplement).
17
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
Figure 2. Relative efficiency scores averaged by year: (a) developed countries; (b)
graduated developing countries; (c) developing countries; (d) EU-28
(a)
(b)
18
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
(c)
(d)
Among other things, both similarities and differences between the results of the
models are revealed. The differences between efficiency scores are considerable and,
in average terms, most prominent in the case of developing countries. Average
countries’ efficiencies are, without exception, below one, which is indicating some
degree of inefficiency for each country. This means that none of the 28 countries was
efficient throughout the entire period. Moreover, 13 of them were continuously pure
technically inefficient, and even 23 exhibited permanent technical and scale
inefficiency. We note as a curiosity that 2003 is the only year with no pure technically
efficient country. Only 13 entities were technically efficient, which is every fifth pure
technically efficient entity. Interestingly, all technically efficient entities are from the
subgroup of developed countries. Moreover, the developed countries are on average
the most efficient, followed by the graduated developing countries, regardless of
efficiency type. The differences between the subgroups of countries are however much
more prominent in technical and scale than in pure technical terms, both in each year
(Figure 3) and averaged over all years (Table 3). According to both models, on average
the least and most efficient years were 1996 and 2008 respectively. This may seem
19
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
surprising, given that 2008 was the first full year of the Great Recession. It should
therefore be underlined that five out of six indicators upon which these results are
based had better than average values in that year.
The standard deviations in both models are significantly lower when considering
a particular country in different years than when considering different countries in a
particular year. This testifies to a well-balanced countries’ performance over time and,
at the same time, indicates large differences across countries. These differences are
especially pronounced among graduated developing countries, which comes as no
surprise given their heterogeneity. Namely, only this subgroup includes both
transition and non-transition countries, also with considerable variation in the dates
of their inclusion to the list of advanced economies.
The efficiency scores of inefficient entities allow their immediate ranking, while
efficient entities cannot be ranked directly because of their maximum efficiency. Since
the reference set of an inefficient entity is composed of efficient ones, one among many
approaches proposed by researchers for ranking an efficient entity is to sum the
frequency of its appearance in each reference set. Consequently, the higher the
frequency, the more robust the entity is. Table 4 shows the results of the BCC-I-W1
model in these terms. Because of the large sample size, the findings are not presented
at the entity level, but at the country level. In this sense, it should be noted that a
country may occur as a reference, even in the same reference set, through its
performance not only in one year but in any year in which it was efficient. At the same
time, any of these efficient performances may become a reference not only to a number
of different inefficient countries, but to the same country (including itself) in different
years. Nevertheless, 15 out of 28 countries have been selected as referential for the
20
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
proposed model. The total number of their appearances in the reference sets (1,723) is
divided into four groups – the one named ‘Self’ for the cases in which a country’s
performance in a particular year is a reference to its own performance in another year,
and one for each subgroup of countries (excluding self-comparison). Latvia is the
country with the highest average pure efficiency score and, although efficient in only
eight years, sets an exemplar by serving as a reference for the greatest number of
inefficient entities (370) that mostly represent developed countries (45.68 %). On the
other side, the Netherlands is a reference in only nine cases, out of which in even eight
to itself. By contrast, Lithuania is the only country that is no reference to itself, but only
to other graduated developing countries.
The differences between target and actual values provide the most direct insight
into the extent of each indicator’s contribution to inefficiency and dictate performance
improvements of the countries. The improvements needed according to the model
BCC-I-W1 are averaged over time and across countries, and expressed as percentages
in Table 5 for each indicator (for detailed results see the supplement). Thus represented
inefficiencies can be decomposed into radial and non-radial components and removed
by carrying out the aforedescribed two-phase procedure. The overall pure efficiency
score, averaged over time and across inefficient countries, is 0.835 (Table 5) which
21
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
suggests that on average the countries have 16.5 % inefficiency5. To perform efficiently,
∗
the first phase enjoins the countries to eliminate radial inefficiency by (0.835), i.e.
through the maximum possible equiproportional reduction (-16.5 %) of all five inputs6,
while preserving the output quantities at their actual levels and the proportions in
which the inputs are utilized. All the percentage differences in the last column of Table
5 indicate that the inefficiency cannot be completely eliminated by solely removing the
radial inefficiency. Hence, after radial reduction of all five inputs, the countries should
eliminate their non-radial inefficiency, through further decrease of inputs in the
second phase, which alters their proportions. The different percentages of
improvements needed to remove inefficiency signify that for all inputs, except for
renewables, the radial inefficiency is much more pronounced than non-radial for the
majority of inefficient countries in the observed year. Conversely, GDP, as the single
output, causes solely non-radial inefficiency.
22
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
insufficiently practiced fact that the structure of the energy consumed and its
environmental consequences should indeed become a major issue for policy makers,
especially in developed nations. On the other side, GDP has the least impact on
efficiency. Consideration of GDP by subgroups of countries leads to quite the opposite
conclusion from the one concerning renewables, i.e. the more the country is developed,
the lower the required increase. Moreover, it causes no inefficiency at all in eleven
years for developed and in three years for graduated developing countries. More
specifically, the inefficient entities’ unemployment is 26.76 % above target, gross
capital formation is 21.23 % too low, renewable energy consumption is an astonishing
48.63 % below target, energy use is 19.41 % too high, GHG emissions are 23.04 %
higher, while GDP is 8.86 % lower than required, all in average terms. Viewing the
variables’ contribution to inefficiency in proportion, as given in Figure 4, renewables
have the largest (33 %) while GDP has the smallest (6 %) share. A stronger impact of
inputs rather than output on inefficiency is one of the predicted consequences of model
orientation selection.
23
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
scale inefficiency. To portray this problem, two contrary examples are briefly
introduced. On one side, pure technical scores are extremely higher than scale scores
in the case of Romania during the whole period under analysis. The difference is most
pronounced for the period 1997-2001, meaning that the major part of its overall
inefficiency can be attributed to scale inefficiency. The investigation also reveals
continuous increasing returns to scale, which means that Romania performs at sub-
optimal scale. To achieve the optimal scale, it obviously needs to increase the size of
operations. On the other side, the influence of scale inefficiency on overall inefficiency
is somewhat lower than the influence of pure technical inefficiency for Germany, also
throughout the whole analysed period. To remove the pure technical inefficiency
policy makers should embrace the strategies from benchmarking countries, of course
in the sense of here selected input and output variables. Nevertheless, regardless of
the type and extent of inefficiency, less than optimally demanded outcomes with
respect to the employed resources were produced. The efficiency assessment of
Germany may seem improper with regard to the overall socio-economic condition of
this country. It should therefore be noted that such result is to the greatest extent
related to renewable energy consumption, which also puts this issue in the context of
this research.
However, the following consistent positive trends are noted based on the data on
this indicator. The share of renewables on the EU level has increased from 10 % at the
beginning to 17.5 % at the end of the investigated period. Thanks to a greater relative
shift upward, from 2.2 % to 12.4 %, that Germany managed to make during the same
period, the unfavourable ratio of these shares on the EU and German level declined
from 4.7 to 1.4.
Relatively similar average pure technical efficiency scores and, at the same time,
mutually quite different average technical efficiency scores, on the subgroup level,
result in significant differences between the subgroups of countries regarding their
scale efficiencies. Thus the scale efficiency scores are 0.809, 0.417 and 0.264 for
developed, graduated developing and developing countries respectively. On average,
the highest scale efficiency scores in these three subgroups of countries are achieved
by Denmark, Cyprus and Croatia, respectively.
24
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
4. CONCLUDING REMARKS
The dynamic relative efficiency assessment of the European Union countries was
carried out through the mutual performance comparison of all 28 current member
states classified into three subgroups, covering the period 1995-2012. The analysis was
conducted using input-oriented DEA window analysis models under the assumptions
of both constant and variable returns to scale. The research explores countries’
efficiency and its dynamics over an 18-year period, relying on six energy and non-
energy indicators – five inputs and one output. Although the empirical findings could
25
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
have not been discussed in detail due to space limitations, some challenging
implications for policy and practice can be drawn. The efficiency scores reveal
significant disparities, both within and among the subgroups of countries, with
different intensities depending on the type of efficiency. Nevertheless, the results
suggest that all countries have certain potential to raise the scores by improving their
inputs. Thus the overall technical inefficiency, averaged over time and across
countries, could be reduced by 48 %. This percentage can be ascribed to pure technical
inefficiencies and to suboptimal scale operations, with the impact of the latter being
more pronounced, particularly in the case of less developed countries. Renewable
energy consumption is the most influential input variable that contributes significantly
to the inefficiency of each country. These facts pose a major threat to sustainability and
an imperative to policy makers to perceive the problems comprehensively and to tailor
their decisions on energy strategies to address contemporary challenges.
To gain a more comprehensive and, at the same time, more detailed insight into
the energy-growth nexus, the analysis relying on the DEA method should be
broadened by covering more countries from Europe and other parts of the world, by
extending the time frame and by including more indicators that would reflect some
additional energy and environmental burdens. In this way, the limitations of the
present study can serve as opportunities for future research.
REFERENCES
1. Apergis, N., Aye, G.C., Barros, C.P., Gupta, R. and Wanke, P. (2015) ‘Energy
efficiency of selected OECD countries: A slacks based model with undesirable
outputs’, Energy Economics, 51, pp.45-53.
2. Banker, R.D., Charnes, A. and Cooper, W.W. (1984) ‘Some models for estimating
technical and scale inefficiencies in data envelopment analysis’, Management
science, 30(9), pp.1078-1092.
3. Ceylan, D. and Gunay, E.N.O. (2010) ‘Energy efficiency trends and policies:
Cross-country comparison in Europe’, paper presented in International
26
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
4. Charnes, A., Cooper, W.W. and Rhodes, E. (1978) ‘Measuring the efficiency of
decision making units’, European journal of operational research, 2(6), pp.429-
444.
6. Chien, T. and Hu, J.L. (2007) ‘Renewable energy and macroeconomic efficiency
of OECD and non-OECD economies’, Energy Policy, 35(7), pp.3606-3615.
27
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
13. EU Commission. (2011) ‘Energy Roadmap 2050’, COM (2011) 885 final, Brussels,
15th December 2011. Available at: http://eur-lex.europa.eu/legal-content/EN/
TXT/PDF/?uri=CELEX:52011DC0885&from=EN (Accessed: 1 March 2017)
16. EU Commission. (2015) ‘Renewable energy progress report’, COM (2015) 293
final, Brussels, 15th June 2015. Available at: http://eur-lex.europa.eu/
resource.html?uri=cellar:4f8722ce-1347-11e5-8817-01aa75ed71a1.0001.02/
DOC_1&format=PDF (Accessed: 1 March 2017)
18. European Parliament. (2017) ‘Renewable Energy’, Fact Sheets on the European
Union – 2017, Brussels. Available at: http://www.europarl.europa.eu/ftu/pdf/
en/FTU_5.7.4.pdf (Accessed: 1 March 2017)
28
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric
Evidence for Developing, Graduated Developing and Developed European Union Countries.”
Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
19. Farrell, M.J. (1957) ‘The measurement of productive efficiency’, Journal of the
Royal Statistical Society. Series A (General), 120(3), pp.253-290.
20. Halkos, G.E. and Tzeremes, N.G. (2013) ‘Renewable energy consumption and
economic efficiency: Evidence from European countries’, Journal of Renewable
and Sustainable Energy, 5(4), 041803.
21. Liu, J.S., Lu, L.Y. and Lu, W.M. (2016) ‘Research fronts in data envelopment
analysis’, Omega, 58, pp.33-45.
22. Menegaki, A.N. (2013) ‘Growth and renewable energy in Europe: Benchmarking
with data envelopment analysis’, Renewable Energy, 60, pp.363-369.
24. Renewable Energy Policy Network for the 21st century – REN21 (2016)
Renewables 2016: Global Status Report. Paris: REN21 Secretariat. Available at:
http://www.ren21.net/wp-content/uploads/2016/10/REN21_GSR2016_
FullReport_en_11.pdf (Accessed: 1 March 2017)
25. Scarlat, N., Dallemand, J.F., Monforti-Ferrario, F., Banja, M. and Motola, V. (2015)
‘Renewable energy policy framework and bioenergy contribution in the
European Union–An overview from National Renewable Energy Action Plans
and Progress Reports’, Renewable and Sustainable Energy Reviews, 51, pp.969-
985.
26. Woo, C., Chung, Y., Chun, D., Seo, H. and Hong, S. (2015) ‘The static and dynamic
environmental efficiency of renewable energy: A Malmquist index analysis of
OECD countries’, Renewable and Sustainable Energy Reviews, 47, pp.367-376.
27. Zhou, P., Ang, B.W. and Poh, K.L. (2007) ‘A mathematical programming
approach to constructing composite indicators’, Ecological economics, 62(2),
pp.291-297.
28. Zhou, P., Ang, B.W. and Poh, K.L. (2008) ‘A survey of data envelopment analysis
in energy and environmental studies’, European Journal of Operational
Research, 189(1), pp.1-18.
29
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric Evidence for Developing, Graduated Developing and
Developed European Union Countries.” Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
APPENDIX
CCR efficiency scores
Year St.
EU-Country Mean Min. Max.
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 dev.
Austria 0.725 0.701 0.723 0.740 0.808 0.851 0.832 0.794 0.747 0.760 0.774 0.834 0.913 0.991 0.904 0.922 0.982 1 0.833 0.701 1 0.098
Belgium 0.483 0.475 0.504 0.509 0.537 0.565 0.563 0.567 0.559 0.601 0.633 0.656 0.691 0.691 0.646 0.652 0.681 0.686 0.594 0.475 0.691 0.075
Denmark 0.753 0.699 0.781 0.813 0.849 0.913 0.899 0.906 0.861 0.915 0.961 0.967 1 1 0.948 0.894 0.972 1 0.896 0.699 1 0.089
Finland 0.490 0.494 0.559 0.622 0.632 0.708 0.699 0.681 0.693 0.746 0.815 0.826 0.913 0.914 0.716 0.749 0.827 0.793 0.715 0.490 0.914 0.124
France 0.589 0.568 0.585 0.611 0.645 0.682 0.683 0.682 0.678 0.704 0.722 0.755 0.788 0.785 0.747 0.756 0.813 0.812 0.700 0.568 0.813 0.078
Germany 0.518 0.506 0.523 0.554 0.579 0.601 0.591 0.575 0.566 0.566 0.577 0.603 0.646 0.648 0.598 0.639 0.691 0.685 0.593 0.506 0.691 0.053
Greece 0.497 0.495 0.498 0.491 0.504 0.506 0.513 0.523 0.555 0.563 0.539 0.595 0.619 0.599 0.577 0.581 0.545 0.509 0.539 0.491 0.619 0.041
Ireland 0.521 0.541 0.586 0.617 0.661 0.703 0.723 0.753 0.805 0.857 0.938 1 0.973 0.884 0.824 0.823 0.900 0.890 0.778 0.521 1 0.148
Developed
Italy 0.628 0.637 0.640 0.632 0.633 0.648 0.658 0.665 0.645 0.655 0.657 0.681 0.704 0.702 0.714 0.711 0.726 0.710 0.669 0.628 0.726 0.033
Luxembourg 0.733 0.718 0.810 0.896 0.961 1 1 0.954 0.932 0.927 0.954 0.898 1 1 0.919 0.953 0.989 0.978 0.923 0.718 1 0.087
Netherlands 0.523 0.536 0.574 0.610 0.668 0.734 0.794 0.760 0.687 0.666 0.687 0.747 0.828 0.886 0.811 0.741 0.792 0.765 0.712 0.523 0.886 0.102
Portugal 0.481 0.503 0.499 0.489 0.484 0.503 0.508 0.482 0.481 0.476 0.468 0.500 0.509 0.535 0.515 0.551 0.549 0.556 0.505 0.468 0.556 0.027
Spain 0.500 0.521 0.509 0.523 0.542 0.561 0.581 0.579 0.590 0.598 0.613 0.650 0.661 0.682 0.644 0.656 0.653 0.644 0.595 0.500 0.682 0.059
Sweden 0.608 0.599 0.635 0.675 0.712 0.840 0.908 0.889 0.842 0.838 0.860 0.931 1 1 0.961 0.957 1 1 0.848 0.599 1 0.142
United Kingdom 0.456 0.454 0.474 0.497 0.529 0.558 0.580 0.602 0.620 0.645 0.664 0.668 0.698 0.703 0.672 0.658 0.718 0.706 0.606 0.454 0.718 0.091
Mean 0.567 0.563 0.593 0.619 0.650 0.692 0.702 0.694 0.684 0.701 0.724 0.754 0.796 0.801 0.746 0.749 0.789 0.782 0.700
Minimum 0.456 0.454 0.474 0.489 0.484 0.503 0.508 0.482 0.481 0.476 0.468 0.500 0.509 0.535 0.515 0.551 0.545 0.509 0.454
Maximum 0.753 0.718 0.810 0.896 0.961 1 1 0.954 0.932 0.927 0.961 1 1 1 0.961 0.957 1 1 1
Stand. deviation 0.101 0.088 0.106 0.121 0.136 0.152 0.154 0.145 0.130 0.136 0.155 0.150 0.162 0.161 0.142 0.133 0.155 0.162 0.155
Cyprus 0.666 0.609 0.655 0.626 0.652 0.654 0.705 0.737 0.689 0.762 0.762 0.780 0.783 0.767 0.736 0.747 0.762 0.781 0.715 0.609 0.783 0.058
Czech Republic 0.237 0.248 0.236 0.233 0.242 0.253 0.258 0.256 0.255 0.265 0.283 0.302 0.327 0.335 0.309 0.308 0.319 0.316 0.277 0.233 0.335 0.035
Estonia 0.127 0.145 0.187 0.188 0.166 0.187 0.205 0.236 0.268 0.283 0.298 0.361 0.401 0.326 0.237 0.241 0.281 0.325 0.248 0.127 0.401 0.076
Graduated developing
Latvia 0.149 0.155 0.172 0.194 0.213 0.235 0.237 0.246 0.261 0.279 0.314 0.356 0.394 0.364 0.302 0.283 0.318 0.324 0.266 0.149 0.394 0.072
Lithuania 0.119 0.116 0.133 0.151 0.168 0.188 0.196 0.199 0.219 0.227 0.239 0.260 0.297 0.286 0.243 0.282 0.289 0.297 0.217 0.116 0.297 0.062
Malta 0.420 0.443 0.415 0.495 0.476 0.566 0.489 0.541 0.492 0.492 0.481 0.518 0.512 0.649 0.680 0.676 0.696 0.734 0.543 0.415 0.734 0.100
Slovak Republic 0.163 0.193 0.204 0.215 0.201 0.203 0.220 0.228 0.227 0.247 0.268 0.290 0.323 0.337 0.297 0.313 0.332 0.328 0.255 0.163 0.337 0.056
Slovenia 0.289 0.287 0.293 0.310 0.336 0.351 0.343 0.345 0.362 0.387 0.397 0.424 0.469 0.479 0.406 0.397 0.399 0.372 0.369 0.287 0.479 0.057
Mean 0.271 0.274 0.287 0.301 0.307 0.330 0.332 0.348 0.347 0.368 0.380 0.412 0.438 0.443 0.401 0.406 0.424 0.435 0.361
Minimum 0.119 0.116 0.133 0.151 0.166 0.187 0.196 0.199 0.219 0.227 0.239 0.260 0.297 0.286 0.237 0.241 0.281 0.297 0.116
Maximum 0.666 0.609 0.655 0.626 0.652 0.654 0.705 0.737 0.689 0.762 0.762 0.780 0.783 0.767 0.736 0.747 0.762 0.781 0.783
Stand. deviation 0.189 0.171 0.172 0.170 0.174 0.182 0.180 0.191 0.165 0.182 0.173 0.170 0.158 0.176 0.197 0.195 0.192 0.201 0.179
Bulgaria 0.074 0.075 0.082 0.088 0.091 0.094 0.094 0.103 0.105 0.115 0.122 0.132 0.146 0.162 0.159 0.152 0.144 0.151 0.116 0.074 0.162 0.030
Croatia 0.301 0.310 0.307 0.303 0.303 0.322 0.326 0.331 0.327 0.340 0.351 0.366 0.372 0.393 0.373 0.373 0.378 0.394 0.343 0.301 0.394 0.033
Hungary 0.196 0.191 0.205 0.224 0.227 0.244 0.245 0.256 0.258 0.282 0.285 0.299 0.307 0.313 0.316 0.314 0.329 0.334 0.268 0.191 0.334 0.047
Developing
Poland 0.136 0.139 0.151 0.169 0.181 0.197 0.198 0.204 0.206 0.216 0.221 0.223 0.243 0.249 0.263 0.256 0.267 0.281 0.211 0.136 0.281 0.044
Romania 0.128 0.130 0.132 0.141 0.159 0.163 0.169 0.172 0.175 0.193 0.202 0.211 0.226 0.250 0.260 0.257 0.254 0.262 0.194 0.128 0.262 0.048
Mean 0.167 0.169 0.175 0.185 0.192 0.204 0.206 0.213 0.214 0.229 0.236 0.246 0.259 0.273 0.274 0.270 0.274 0.284 0.226
Minimum 0.074 0.075 0.082 0.088 0.091 0.094 0.094 0.103 0.105 0.115 0.122 0.132 0.146 0.162 0.159 0.152 0.144 0.151 0.074
Maximum 0.301 0.310 0.307 0.303 0.303 0.322 0.326 0.331 0.327 0.340 0.351 0.366 0.372 0.393 0.373 0.373 0.378 0.394 0.394
Stand. deviation 0.086 0.089 0.086 0.082 0.079 0.086 0.086 0.086 0.084 0.086 0.087 0.089 0.085 0.086 0.079 0.082 0.088 0.091 0.086
Mean 0.411 0.410 0.431 0.451 0.470 0.501 0.508 0.509 0.504 0.522 0.539 0.565 0.598 0.605 0.563 0.566 0.593 0.594 0.519
Minimum 0.074 0.075 0.082 0.088 0.091 0.094 0.094 0.103 0.105 0.115 0.122 0.132 0.146 0.162 0.159 0.152 0.144 0.151 0.074
EU
Maximum 0.753 0.718 0.810 0.896 0.961 1 1 0.954 0.932 0.927 0.961 1 1 1 0.961 0.957 1 1 1
Stand. deviation 0.214 0.205 0.218 0.227 0.242 0.259 0.262 0.254 0.241 0.245 0.255 0.257 0.268 0.269 0.252 0.250 0.268 0.265 0.252
Source: Authors’ calculations using DEA-Solver-Pro
30
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric Evidence for Developing, Graduated Developing and
Developed European Union Countries.” Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
Italy 0.819 0.818 0.818 0.814 0.824 0.829 0.836 0.844 0.824 0.830 0.831 0.854 0.878 0.884 0.902 0.898 0.915 0.899 0.851 0.814 0.915 0.035
Luxembourg 0.844 0.781 0.916 0.957 1 1 1 0.962 0.955 0.960 0.973 0.905 1 1 0.959 0.973 1 1 0.955 0.781 1 0.060
Netherlands 0.683 0.717 0.760 0.808 0.874 0.941 1 0.942 0.835 0.775 0.786 0.849 0.935 1 0.923 0.814 0.860 0.819 0.851 0.683 1 0.092
Portugal 0.984 1 1 1 0.981 1 1 0.934 0.917 0.897 0.864 0.901 0.916 0.964 0.927 0.956 0.942 0.970 0.953 0.864 1 0.043
Spain 0.840 0.859 0.833 0.825 0.830 0.837 0.850 0.844 0.853 0.859 0.887 0.930 0.937 0.941 0.912 0.932 0.919 0.904 0.877 0.825 0.941 0.042
Sweden 0.787 0.748 0.797 0.823 0.838 0.970 1 0.975 0.930 0.901 0.892 0.942 1 1 1 0.972 1 1 0.921 0.748 1 0.086
United Kingdom 0.643 0.640 0.653 0.675 0.697 0.724 0.758 0.763 0.777 0.801 0.810 0.793 0.819 0.827 0.799 0.784 0.844 0.828 0.758 0.640 0.844 0.068
Mean 0.793 0.779 0.805 0.823 0.842 0.877 0.880 0.858 0.843 0.849 0.864 0.883 0.922 0.929 0.882 0.876 0.913 0.909 0.863
Minimum 0.643 0.640 0.653 0.675 0.697 0.724 0.737 0.694 0.693 0.701 0.709 0.727 0.767 0.765 0.737 0.759 0.814 0.804 0.640
Maximum 0.999 1 1 1 1 1 1 0.975 0.955 0.960 0.978 1 1 1 1 0.973 1 1 1
Stand. deviation 0.110 0.107 0.099 0.094 0.095 0.093 0.098 0.090 0.077 0.070 0.076 0.070 0.073 0.074 0.082 0.073 0.067 0.078 0.093
Cyprus 1 0.946 1 0.926 0.939 0.918 0.975 1 0.928 1 1 1 1 1 1 1 1 1 0.980 0.918 1 0.032
Czech Republic 0.973 1 0.894 0.780 0.707 0.765 0.771 0.746 0.720 0.725 0.727 0.763 0.880 0.933 0.733 0.717 0.741 0.720 0.794 0.707 1 0.096
Estonia 0.690 0.701 0.831 0.815 0.661 0.694 0.725 0.797 0.849 0.831 0.805 0.970 1 0.869 0.645 0.661 0.701 0.738 0.777 0.645 1 0.104
Graduated developing
Latvia 0.958 0.946 0.918 0.977 1 1 1 1 0.997 0.999 1 1 1 0.998 0.987 0.921 0.956 1 0.981 0.918 1 0.028
Lithuania 0.761 0.702 0.764 0.781 0.821 0.815 0.805 0.829 0.838 0.836 0.842 0.948 1 0.873 0.719 0.769 0.790 0.770 0.815 0.702 1 0.073
Malta 0.999 0.971 0.872 0.964 0.891 1 0.886 0.970 0.871 0.873 0.847 0.893 0.864 1 1 1 1 1 0.939 0.847 1 0.061
Slovak Republic 0.650 0.865 0.867 0.845 0.705 0.687 0.755 0.740 0.678 0.702 0.718 0.721 0.729 0.728 0.693 0.688 0.714 0.709 0.733 0.650 0.867 0.063
Slovenia 0.739 0.725 0.743 0.720 0.750 0.771 0.803 0.756 0.752 0.805 0.792 0.846 0.967 0.997 0.767 0.718 0.703 0.702 0.781 0.702 0.997 0.083
Mean 0.846 0.857 0.861 0.851 0.809 0.831 0.840 0.855 0.829 0.847 0.841 0.893 0.930 0.925 0.818 0.809 0.826 0.830 0.850
Minimum 0.650 0.701 0.743 0.720 0.661 0.687 0.725 0.740 0.678 0.702 0.718 0.721 0.729 0.728 0.645 0.661 0.701 0.702 0.645
Maximum 1 1 1 0.977 1 1 1 1 0.997 1 1 1 1 1 1 1 1 1 1
Stand. deviation 0.150 0.128 0.083 0.095 0.124 0.127 0.103 0.116 0.107 0.110 0.109 0.107 0.099 0.097 0.151 0.142 0.136 0.142 0.117
Bulgaria 0.609 0.573 0.620 0.697 0.747 0.735 0.739 0.731 0.737 0.763 0.780 0.804 0.835 0.956 0.840 0.754 0.703 0.727 0.742 0.573 0.956 0.089
Croatia 1 1 1 0.937 0.945 0.940 0.944 0.961 0.947 0.950 0.954 0.973 0.946 0.985 0.945 0.920 0.902 0.938 0.955 0.902 1 0.027
Hungary 0.773 0.764 0.793 0.822 0.838 0.879 0.902 0.902 0.877 0.881 0.824 0.815 0.828 0.830 0.838 0.824 0.844 0.854 0.838 0.764 0.902 0.040
Developing
Poland 0.688 0.695 0.723 0.774 0.790 0.798 0.744 0.724 0.717 0.735 0.726 0.728 0.771 0.764 0.744 0.717 0.731 0.737 0.739 0.688 0.798 0.030
Romania 0.848 0.874 0.971 0.997 1 1 1 0.948 0.936 0.940 0.941 0.945 0.987 1 1 1 0.988 0.988 0.965 0.848 1 0.045
Mean 0.784 0.781 0.822 0.846 0.864 0.870 0.866 0.853 0.843 0.854 0.845 0.853 0.873 0.907 0.873 0.843 0.834 0.849 0.848
Minimum 0.609 0.573 0.620 0.697 0.747 0.735 0.739 0.724 0.717 0.735 0.726 0.728 0.771 0.764 0.744 0.717 0.703 0.727 0.573
Maximum 1 1 1 0.997 1 1 1 0.961 0.947 0.950 0.954 0.973 0.987 1 1 1 0.988 0.988 1
Stand. deviation 0.151 0.164 0.162 0.121 0.106 0.106 0.119 0.117 0.109 0.100 0.100 0.103 0.090 0.104 0.100 0.117 0.119 0.117 0.111
Mean 0.807 0.802 0.824 0.835 0.836 0.863 0.866 0.856 0.839 0.849 0.854 0.881 0.916 0.924 0.862 0.851 0.874 0.876 0.856
Minimum 0.609 0.573 0.620 0.675 0.661 0.687 0.725 0.694 0.678 0.701 0.709 0.721 0.729 0.728 0.645 0.661 0.701 0.702 0.573
EU
Maximum 1 1 1 1 1 1 1 1 0.997 1 1 1 1 1 1 1 1 1 1
Stand. deviation 0.127 0.124 0.106 0.096 0.103 0.104 0.100 0.098 0.089 0.085 0.087 0.085 0.083 0.083 0.108 0.104 0.105 0.109 0.104
Source: Authors’ calculations using DEA-Solver-Pro
31
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric Evidence for Developing, Graduated Developing and
Developed European Union Countries.” Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
Italy 0.768 0.779 0.782 0.777 0.768 0.782 0.787 0.788 0.783 0.789 0.791 0.798 0.802 0.794 0.791 0.792 0.793 0.790 0.786 0.768 0.802 0.009
Luxembourg 0.869 0.920 0.884 0.936 0.961 1 1 0.991 0.976 0.965 0.980 0.992 1 1 0.959 0.979 0.989 0.978 0.966 0.869 1 0.039
Netherlands 0.765 0.748 0.755 0.756 0.764 0.780 0.794 0.807 0.822 0.860 0.875 0.880 0.886 0.886 0.878 0.910 0.921 0.935 0.834 0.748 0.935 0.064
Portugal 0.489 0.503 0.499 0.489 0.494 0.503 0.508 0.517 0.525 0.531 0.542 0.555 0.555 0.555 0.555 0.576 0.582 0.573 0.531 0.489 0.582 0.032
Spain 0.596 0.607 0.611 0.634 0.653 0.670 0.683 0.687 0.692 0.696 0.691 0.698 0.706 0.725 0.706 0.704 0.710 0.712 0.677 0.596 0.725 0.040
Sweden 0.773 0.800 0.796 0.821 0.850 0.866 0.908 0.912 0.905 0.930 0.964 0.989 1 1 0.961 0.985 1 1 0.915 0.773 1 0.079
United Kingdom 0.710 0.710 0.725 0.736 0.758 0.770 0.765 0.789 0.798 0.805 0.820 0.842 0.853 0.850 0.841 0.840 0.851 0.853 0.795 0.710 0.853 0.052
Mean 0.720 0.730 0.740 0.753 0.772 0.787 0.795 0.807 0.810 0.824 0.834 0.850 0.859 0.859 0.843 0.855 0.861 0.857 0.809
Minimum 0.489 0.503 0.499 0.489 0.494 0.503 0.508 0.517 0.525 0.531 0.542 0.555 0.555 0.555 0.555 0.576 0.582 0.573 0.489
Maximum 0.922 0.925 0.936 0.944 0.962 1 1 0.991 0.976 0.985 0.995 1 1 1 0.986 0.995 1 1 1
Stand. deviation 0.110 0.114 0.111 0.121 0.125 0.130 0.129 0.126 0.121 0.123 0.128 0.130 0.130 0.131 0.118 0.123 0.129 0.133 0.129
Cyprus 0.666 0.644 0.655 0.676 0.694 0.712 0.722 0.737 0.742 0.762 0.762 0.780 0.783 0.767 0.736 0.747 0.762 0.781 0.729 0.644 0.783 0.045
Czech Republic 0.243 0.248 0.264 0.299 0.342 0.331 0.335 0.343 0.354 0.366 0.390 0.397 0.371 0.359 0.422 0.430 0.431 0.439 0.353 0.243 0.439 0.061
Estonia 0.184 0.207 0.225 0.231 0.251 0.270 0.282 0.296 0.316 0.341 0.370 0.372 0.401 0.375 0.368 0.365 0.400 0.441 0.316 0.184 0.441 0.076
Graduated developing
Latvia 0.155 0.164 0.187 0.198 0.213 0.235 0.237 0.246 0.262 0.279 0.314 0.356 0.394 0.365 0.306 0.307 0.332 0.324 0.271 0.155 0.394 0.071
Lithuania 0.156 0.165 0.174 0.193 0.205 0.231 0.243 0.240 0.262 0.272 0.284 0.275 0.297 0.328 0.338 0.367 0.366 0.386 0.266 0.156 0.386 0.072
Malta 0.420 0.456 0.476 0.514 0.534 0.566 0.553 0.558 0.565 0.564 0.568 0.580 0.592 0.649 0.680 0.676 0.696 0.734 0.577 0.420 0.734 0.085
Slovak Republic 0.251 0.223 0.236 0.254 0.286 0.296 0.291 0.309 0.334 0.351 0.374 0.402 0.443 0.463 0.428 0.455 0.465 0.463 0.351 0.223 0.465 0.087
Slovenia 0.392 0.396 0.394 0.430 0.449 0.455 0.427 0.456 0.482 0.480 0.501 0.502 0.485 0.481 0.529 0.553 0.567 0.529 0.473 0.392 0.567 0.053
Mean 0.308 0.313 0.326 0.349 0.372 0.387 0.386 0.398 0.415 0.427 0.445 0.458 0.471 0.473 0.476 0.487 0.502 0.512 0.417
Minimum 0.155 0.164 0.174 0.193 0.205 0.231 0.237 0.240 0.262 0.272 0.284 0.275 0.297 0.328 0.306 0.307 0.332 0.324 0.155
Maximum 0.666 0.644 0.655 0.676 0.694 0.712 0.722 0.737 0.742 0.762 0.762 0.780 0.783 0.767 0.736 0.747 0.762 0.781 0.783
Stand. deviation 0.176 0.171 0.169 0.175 0.174 0.175 0.172 0.175 0.169 0.168 0.158 0.160 0.153 0.157 0.159 0.157 0.158 0.163 0.169
Bulgaria 0.121 0.132 0.133 0.127 0.122 0.128 0.128 0.141 0.143 0.151 0.156 0.164 0.175 0.170 0.189 0.201 0.204 0.207 0.155 0.121 0.207 0.030
Croatia 0.301 0.310 0.307 0.323 0.321 0.342 0.346 0.345 0.345 0.358 0.368 0.376 0.393 0.399 0.395 0.406 0.419 0.420 0.360 0.301 0.420 0.039
Hungary 0.254 0.251 0.259 0.272 0.270 0.277 0.271 0.283 0.294 0.320 0.346 0.367 0.370 0.376 0.377 0.382 0.390 0.391 0.319 0.251 0.391 0.054
Developing
Poland 0.198 0.199 0.208 0.218 0.228 0.247 0.266 0.282 0.288 0.294 0.304 0.307 0.315 0.326 0.354 0.357 0.365 0.380 0.285 0.198 0.380 0.059
Romania 0.151 0.149 0.135 0.141 0.159 0.163 0.169 0.181 0.187 0.206 0.215 0.223 0.230 0.250 0.260 0.257 0.257 0.265 0.200 0.135 0.265 0.046
Mean 0.205 0.208 0.209 0.216 0.220 0.232 0.236 0.246 0.251 0.266 0.278 0.287 0.297 0.304 0.315 0.321 0.327 0.333 0.264
Minimum 0.121 0.132 0.133 0.127 0.122 0.128 0.128 0.141 0.143 0.151 0.156 0.164 0.175 0.170 0.189 0.201 0.204 0.207 0.121
Maximum 0.301 0.310 0.307 0.323 0.321 0.342 0.346 0.345 0.345 0.358 0.368 0.376 0.393 0.399 0.395 0.406 0.419 0.420 0.420
Stand. deviation 0.074 0.073 0.076 0.084 0.081 0.086 0.087 0.083 0.083 0.085 0.090 0.092 0.093 0.094 0.088 0.088 0.092 0.092 0.089
Mean 0.511 0.518 0.527 0.542 0.559 0.573 0.579 0.590 0.598 0.611 0.624 0.638 0.648 0.650 0.644 0.654 0.663 0.665 0.600
Minimum 0.121 0.132 0.133 0.127 0.122 0.128 0.128 0.141 0.143 0.151 0.156 0.164 0.175 0.170 0.189 0.201 0.204 0.207 0.121
EU
Maximum 0.922 0.925 0.936 0.944 0.962 1 1 0.991 0.976 0.985 0.995 1 1 1 0.986 0.995 1 1 1
Stand. deviation 0.263 0.265 0.266 0.268 0.272 0.274 0.276 0.276 0.271 0.271 0.269 0.272 0.270 0.269 0.256 0.258 0.258 0.255 0.267
Source: Authors’ calculations using DEA-Solver-Pro
32
Rabar, Danijela. “Renewable Energy Consumption-Economic Growth Nexus: A Nonparametric Evidence for Developing, Graduated Developing and
Developed European Union Countries.” Modelling Economic Growth: Issues and New Insights. Juraj Dobrila University of Pula, 2018.
33