Feed-In-Tariffs and Government Corruption: Another Look at the Diffusion of Renewable Energy Technologies

Genius Margarita.* Vergis V. Emmanouil†, June 2012

Abstract

This paper analyzes the factors that contribute to the diffusion process of solar photovoltaic and wind systems. The literature on diffusion theory argues that government‟s intervention could act as a lever for the successful deployment of RETs by eliminating technology lockout barriers. Therefore, this research‟s main objective is to shed light on how gradual deregulation of the electricity sector, policy variables, such as the level of Feed-in Tariffs (FITs) and other support policies, and variables related to the perceptions of a country‟s level of corruption affect the propagation of Renewable Energy technologies (RETs). A panel dataset for the EU-15 covering the period 1997-2009 is analyzed using instrumental variables due to the endogeneity of some policy variables. The results show that Feed-in Tariffs and the deregulation process were successful in boosting investments towards solar photovoltaic and wind systems. However the strength of the effect diminishes as the perceived levels of corruption are lower which means that more policy intensity is required when corruption perceptions improve.

Keywords: Diffusion, Panel Data IV, Corruption Perceptions, EU-15 JEL: Q42, O38, C26

*

Assistant Professor of economics at the University of Crete, Greece, Department of Economics, Gallos Campus, 74100 Rethymno, Tel.: +30 28310 77414, Fax: +30 28310 77406, E-mail: genius@econ.soc.uoc.gr

PhD student at the University of Crete, Greece, Department of Economics, Gallos Campus, 74100, Tel.:+30 2810 322733 Rethymno, E-mail: mvergis@econ.soc.uoc.gr (Presenter). 1

1. Introduction

The frail economic situation experienced at present by most European economies seems to have diminished the importance of global warming concerns in the political agenda. The spearhead of the struggle to preserve the environment, is the substitution of conventional energy production technologies with renewable energy technologies (RET). The levels of Green House Gas (GHG) emissions as well as the temperature of the planet were balanced for thousands of years until the beginning of the past century. Human activity together with burning of fossil fuels raised the concentration of GHG emissions by more than one third over the past several decades causing the global warming effect. The world energy demand increased by 39% from 1990 to 2008 (IEA/OECD)3, which in turn boosted the produced CO2 emissions. In response to this situation world leaders started taking combined action which led to the adoption of the Kyoto Protocol on 11 December 1997. At the time that the protocol was firstly adopted it only encouraged the signatory countries to reduce GHG emissions, however, on February 16 2005 the protocol became mandatory for the 37 industrialized countries while the European community ratified it. Thus, the parties that ratified the Kyoto Protocol were obliged to reduce GHG emissions by an average of five per cent below their 1990 levels over the five-year period 2008-2012. The EU took action to promote RET diffusion towards the fight against global warming even earlier than the adoption of the Kyoto protocol. Although Europe initiated a strategic approach for its energy future with the publication of the White Paper on November 26 1997, the member targets were not compulsory. Even when the EU set Directive 2001/77/EC on September 27 2001 in order to promote renewable energy use in electricity generation, the established ambitious national targets - 12% of the EU energy consumption ought to be derived from RES by the year 2010 - were left to the environmental consciousness of each member state. In the light of the unsatisfactory progress made by the European community, the EU decided over mandatory and comprehensive Energy European Policy on October 27 2005 and proceeded to establish Directive 2009/28/EC with binding targets henceforth. The share of energy produced from RES and the share of RES used in the transport sector should increase to 20% and 10% respectively by the year 2020. In the technology diffusion literature, environmental regulation constitutes the main driver for the adoption of technologies with low and zero carbon emissions (Gray and Shadbegian, 1998; Kerr and Newell, 2003). All of the aforementioned studies argue that in order for investors to allocate their funds to environmentally favourable technologies, environmental regulation must be both stringent and coherent. The lack of stringency, which

3

IEA Key energy statistics 2010

2

al. Fredriksson. government intervention must not be recurrent and must be adjusted according to the energy technology needs. Institutional factors and particularly government corruption can negatively affect investment activities (Everhart et. It could for instance lead to a reduction in red tape or bureaucracy and speed up certain procedures. 2009. 1995) and distort incentives provided by the policy maker (Fredriksson. 2011) by providing valuable incentives to new RET investments in the energy sector. the fact that the energy industry is highly capital intensive makes it more difficult for RET technologies to replace conventional ones in the immediate future in the absence of environmental regulation especially in the cases where projects involving other sources such as oil. However. Vollebergh and Dijkgraaf (2004).is partly due to the lack of conviction among policy makers on the ability of new technologies to supersede the old ones. The effectiveness of a policy scheme is depicted through the long term achieved targets and is obviously affected from the outcome on the diffusion process. Due to the urgency to confront the global warming problem. Mauro. the diffusion of RET should follow a dynamic process where policy schemes are designed so as to provide security for investors (Jaccobson and Bergek. confirmed that 3 . 2004). On the other hand. The energy producing technologies are far more cost competitive than RETs. 1995. List and Millimet. 2004). In addition to policy instruments applied by each country. On the other hand. al. policy instruments can also be directed to the encouragement of RET (Popp et. governments could just eradicate the use of technologies using conventional sources in the energy sector. Söderholm and Klaassen. 2003. From another standpoint. al. (1999) in the case of Asian countries since potential investors might find investments rather attractive in a corrupt environment. Vollebergh and Dijkgraaf. eventually constitutes an obstruction to the implementation of policy mechanisms (Nilsson et. government corruption. al 2004). if highly organized. other factors can contribute to the success or failure of the targets set by the EU members in Directive 2009/28/EC. gas and nuclear power have not yet paid back their investments. thus the provision of incentives inspiring stability and coherency is a prerequisite for RETs to diffuse (Ringel 2006. a policy maker that abruptly harms investment funds will hardly be trusted by investors in the future. Fredriksson. 2007). could have a positive effect on investment as shown by Campos et. Thus. The literature on diffusion theory suggests that both innovation and diffusion processes should be considered as endogenous in the sense that they simultaneously cause cost reductions to new technologies entering a market (Jaffe and Stavins. Angoluchi 2008). Beyond the discouragement of energy production from conventional sources. policy supports such as Feed-in Tariffs could also be endogenous since they are revised according to changes in investment costs as the need for policy supports is lower as investment costs fall. Moreover. However..

ii) control for the effect of policy schemes through different supporting schemes. Literature on the Diffusion of Environmental Technologies Diffusion of new technologies is regarded as the process by which old technologies are gradually substituted by new more cost-effective and efficient ones. and iii) investigate the interplay between the effectiveness of policy instruments as a mechanism to enhance investments on SP&W and the prevailing perceptions about government corruption. 2006). Our contribution to the literature of RET diffusion is that we i) empirically scrutinizing factors proposed from the literature of diffusion theory. The duration needed for one new technology to reach the critical mass and to be established in the market. Lastly. in the case of the electricity industry where conventional source technologies are generally used. For instance. the process of RETs diffusion has taken more than two decades and its progress has been slower than expected. the main determinants proposed in the literature related to technology diffusion theory of RE systems will be considered with special emphasis on policy schemes enhancing RE and corruption perceptions. Section 4 the main variables. The remainder of the paper is organized as follows: section 2 provides a background review concerning the role of policy schemes in the diffusion theory RETs and examines the existing empirical research dealing with the factors affecting the diffusion of RETs.coordination between lobby groups may result in policy schemes effectiveness. this research examines the way in which the gradual deregulation in the electricity market of the EU-15 countries influences the diffusion of SP&W technologies. In doing so. while section 3 provides some motivation for the use of corruption perceptions as a determinant. Angolucci. 2008. the present paper offers a thorough analysis of the determinants of SP&W proliferation based on the literature dealing with technology diffusion and empirically analyzes the importance of these determinants using panel data for the EU 15 during the period 1997-2009. Overall. Rational investors behavior is driven by 4 . The aim of the present paper is to scrutinize the determinants of RE generation proliferation and in particular Solar Photovoltaic and Wind generators (SP&W hereafter) in the EU-15. is an issue closely related to the technology characteristics and the market for which it is intended. Section 6 presents the main conclusions and limitations of our methodology and data. The reason behind this lies in the fact that investors act in a highly capital intensive market and have not yet been presented with the right incentives (Pizer and Popp. the econometric model and presents the basic results while section 5 offers a discussion of the results. It is therefore not clear at the outset whether corruption could speed up or slow down the diffusion of RETs. 2. Moreover.

where new technological advances result in cost reductions of new technologies (Söderholm and Klaassen. al. Jaffe and Stavins (1995) develop a theoretical model explaining the adoption of new technologies in the presence of different policy schemes and apply it to empirically analyze the effect of building codes and energy taxes on the adoption of thermal insulation by US households. This effect is confirmed by Popp et. Using patent data. The main disadvantage of imposing policy mechanisms to reduce the competiveness of existing more polluting technologies is the downsizing of the profitability of the industry which may have adverse effects and scare new investment funds (Xepapadeas and Zeeuw. investors do not have any incentives to allocate their funds towards RETs (Jaccobson and Bergek 2004). Examples of such environmental regulation are carbon tax emissions and technology standards among others. in the form of cumulative R&D expenses in the wind industry. For instance. Government intervention through policy mechanisms results in a reduction of RETs costs which is the determining factor that influences investors‟ decisions to adopt RETs (Söderholm and Klaassen. The literature on diffusion of environmental friendly technologies argues that there are three main government support schemes to enhance the competitiveness of new technologies. 2007). The authors empirically test their model using pooled time series and conclude that investments in R&D translated into technological cost reductions and thus assisted in the diffusion process of windmills in Europe. Söderholm and Klaassen (2007) develop a rational choice model in which profit maximizing firms choose to invest in windmills in Europe and learning by doing effects are introduced in their model. 1998. (2011). On the other hand. 2011. 2007. in a panel data analysis of 26 OECD countries. etc). Further analysis on the effects of technological change can be found in the review provided by Pizer and Popp (2008). 1999). however. The second type of schemes includes the regulation imposed to reduce GHG emissions produced from the existing technologies by simply raising their production cost (Jaffe and Stavins. the authors created knowledge stocks which found them to have small effect in RETs investments. 1999. Kerr and Newell. Kemp. 1995. al. 2003). thus making them less competitive. Popp et. they find no evidence that building codes had a significant effect on insulation adoption. 2008.investment returns‟ maximization and not by social welfare considerations such as reducing CO2 emissions. Gray and Shadbegian (1998) and Kerr and Newell (2003) find evidence that stringent environmental regulation could raise the 5 . without policy intervention. Xepapadeas and Zeeuw. 1998. Pizer and Popp. In this sense. Gray and Shadbegian. In the same direction are the results of Snyder et. al (2003) who find that regulatory factors had no effects on adoption of new technologies in their particular example of membrane-cell technology. The first is through innovation.

Finally. Ringel (2006) argues that policy schemes must be constructed in compliance with both economic effectiveness and ecological efficiency raising doubts on how FITs can be compatible with a common European electricity market. the author acknowledges that FITs constitute the proper mechanism for a country seeking to rapidly adopt RETs. al. However. Excluding Finland. Mulder (2008). In the same line.profitability of new less polluting technologies with respect to the existing ones and thus enhance their diffusion process. or lack of it. Nevertheless. whose predominant policy instrument is investment incentives. EU countries using predominantly Feed-in-Tariffs instead of TGCs showed an obtrusive appraisal of electricity produced from SP&W (Maza et. a predefined bidding process for fix amounts of power installations which offers winners propitious price or investment conditions. are usually applied by EU countries as complementary instruments. A number of policy schemes enhancing the diffusion of SP&W systems have been implemented in the European Union during the last two decades. Quotas/Tradable Green Certificates (TGC) upon which a parallel market of renewable energy certificates is established with producers benefiting from the sale of certificates and Investment incentives where a proportion of the overall investment cost of RES electricity production projects is financially supported. the rest of EU countries‟ policy mechanisms rely on either Feed-in Tariffs or TGCs. FITs and TGCs have generated a great deal of discussion among scholars concerning their effectiveness. Fiscal and tax reduction incentives along with tenders. 2010). The positive effect of FITs on the expansion of wind capacity in four European countries has also been uncovered by Söderholm and Klaassen (2007) who use a rational 6 . but they do not make it possible for policy makers to accurately forecast the amount of the energy produced from RES in a given period. EU countries have devised economic instruments aiming at either the prompt increase of electricity production from renewable energy sources (RES) or at the long term viability of RES investments. The main policy mechanisms implemented in EU countries are Feed-in Tariffs (FITs) where RES electricity producers are paid an additional premium on top of the electricity market price. Denmark and Spain. Wang (2006) argues that the TGCs supporting mechanism does not provide stimulation for high capital cost technologies such as SP&W systems. in enhancing RETs. studied policies implemented to promote wind energy in EU countries and argued that although relatively low Feed-in-Tariffs has been an effective policy scheme for Germany. Moreover. its effectiveness relied on the early and consistent way that were implemented. the success of the implementation of such policy instruments in the EU is not yet clear (OPTRES 2007). The main policy schemes. Falconett and Nagasaka (2010) argue that FITs is the best instrument to support less mature RETs while TGCs are more suitable for more mature technologies. governments can promote the diffusion of new technologies by providing incentives for their deployment.

For instance. Market deregulation in the electricity sector can affect electricity generation. Tenders. Popp et. While the price received by the RETs generator plays an important role for profit maximization it is the case that the price of alternative electricity generation sources will also affect his choice. when these are used as determinants of SP&W can lead to erroneous conclusions if this is not taken properly into account. on the other hand. Quotas. In the same direction are the models developed by Kemp (1998). For instance. Söderholm and Klaassen (2007) include the price of coal in the profit function of their rational simultaneous choice model. From a rational investor point of view. the present paper tries to shed some light about the effects of policies targeting to SP&W sources. From an econometrics point of view the endogeneity of policies. namely Feed-in Tariffs.choice model of technological diffusion. Other researchers in the literature of RETs diffusion have also examined the effect of policy mechanisms. thus creating a more attractive environment for the option of wind generators. (2011) find that both FITs and other policies such as TGCs. Investment Subsidies. 2007. which for the countries examined was considered the main competitor of windmills. In the same direction the results of Johnstone et. Mazaa and Winden. Tax and fiscal incentives while at the same time it investigates whether policies are endogenous to the deployment of SP&W as suggested in the literature of technology diffusion and policy making (Söderholm and Klaassen. The effect of prices of conventional sources was also examined by Van Ruijven and Van Vuuren (2009) using the energy model TIMER where they argue that higher prices of traditional sources could favour the deployment of alternative energy sources including RES. Although they find that ceteris paribus FITs increase investments in capacity. Generation 7 . Jaffe and Stavins. al. Gray and Shadbegian (1998) and Kerr and Newell (2003). 2004). they show. From a general perspective. Apart from policy schemes which provide the incentives for new investments to be directed towards RETs. Following the contradictory results in the literature. 1995. have an insignificant effect on RE investments . al. The idea behind the inclusion of coal prices lies in the fact that increasing coal prices result in increasing costs of electricity production from coal-fired technologies. the motivation under which deregulation of electricity markets took place lies in the formation of cost efficient grids and in providing lower prices for the end user. market deregulation may act as a facilitator to attract investment interest to the electricity industry. investment incentives etc. (2010) show that the FITs effect is insignificant on the proliferation of RETs patents. supply transmission and distribution activities. the decision to invest in RETs will be solely based on profit maximization considerations. The process of deregulating the electricity sector started in Europe in 1990 and is still ongoing. that high Feed-in-Tariffs can discourage investments on innovation activities due to the fact that there is no need to further reduce the costs of a new technology.

Delmas and Sancho. Deregulation has been studied as a factor affecting renewable electricity investments and production in the US (Delmas et al. The aforementioned empirical studies confirm that the electricity sector deregulation has a positive impact on the proliferation of renewable base electricity production by enhancing investments on new low and zero carbon emission technologies. is examined in the present study. In particular it is argued that firms switch their investments decisions towards RETs when deregulation of the electricity market along with customers with developed environmental consciousness is present. several impacts of these actions remain unexplored. However. firms may be forced to downsize their fixed assets and invest in new ones on account of this strategic management differentiation. The Role of government Corruption Although the European Union has a strong commitment to reducing CO2 emissions through the deployment of RE sources as well as achieving price reductions for consumers. government corruption could cause deficiencies in the behaviour and strategies of investors in the electricity sector (i. supply activity consists of selling electricity to the end users of the grid and transmission and distribution involves transporting electricity to high and low voltage electricity networks respectively. given that environmental concerned electricity customers can decide on their provider. found that deregulation can stimulate existing firms environmental differentiation. Generally speaking. Jamasb (2006) argued that lack of competition in either part of the electricity sector can produce distortions on investments towards new capacity installations. 3. under market deregulation conditions firms tend to discriminate their position towards competition by investing on RETs. On the other hand. government corruption is defined as the abuse of authority by public servants in order to satisfy personal gains or special interest groups‟ agenda and constitutes one of the most severe hazards to economic growth (see Blackburn and Forgues-Puccio. Although there exists an abundant literature on the effects of 8 . 2001. In other words. literature on strategic management argues that firms tend to differentiate their strategic choice when the market where they act is under deregulation. in the sense that actions concerning corruption remain secret. Henisz and Zelner. (2007). because bureaucrats‟ and public servants‟ corruption cannot be documented. the impact of the ongoing electricity market deregulation on the proliferation of SP&W systems in the EU-15 members. 2007. 2011). It has not been until recently that the literature has examined corruption as a threat to different economic sectors. Delmas et. Carley 2009. al. In this direction.e. According to the author. Following the above discussion.includes the production of electricity. 2010). Jamasb 2006).

1995. al. On the other hand. 1999. 2009). Vollebergh and Dijkgraaf. Campos et. (2004). Yet.. Fredriksson. Mazaa and Winden. Investment activities can be negatively affected by the existence of corruption since it creates an environment where returns to investment are harder to predict (Everhart et al. 2004.. its relationship to the diffusion of RETs has not been scrutinized yet. Vollebergh and Dijkgraaf.. 2010). In addition. When corruption is predictable or is organized then it could have a positive effect on investment as shown by Campos et al. Mauro (1995) finds evidence that corruption lowers private investment for a sample of 67 countries that includes several of the countries considered in our study. some authors claim that corruption could be investment enhancing since redtape can be bypassed in heavily bureaucratic societies as explained by Mauro (1995). the level of corruption could negatively affect the effectiveness of policies that enhance electricity production from RES because potential investors might not find RES investments as attractive in a corrupt environment where policy makers are influenced or are thought to be bribed by lobbies supporting conventional sources. corrupted policy makers which act under greater lobbying groups coordination. Everhart et al. To this extent. On the other hand. Main Variables Dependent variable 9 . Evrensel. the political and policy design literature identifies that government policies result from the cooperation of different agents which commonly are bribed or influenced by the existence of lobbying parties in the market (Fredriksson. that finds environmental policies to be undermined when higher corruption is observed. List and Millimet 2003. In his seminal paper. constructed policies tend to be more stringent. Econometric Model and Results 4.1. may produce more effective policies or it could be the case and investors might be willing to undertake investments in a corrupted environment if the necessary policy incentives are offered. The effect of corruption on environmental regulation has been examined by Leitão (2010). 2009. (1999) in the case of Asian countries. Blackburn and Forgues-Puccio (2009) make the claim that countries with organized corruption networks are likely to display higher rates of growth than countries with disorganized corruption arrangements. In this direction. Main Variables. 4.corruption on investment and growth (Mauro. the authors study the effect of lobby group‟s coordination and found that when lobbies are organized under the identical objectives. 2008). provided a simple model of bribery and empirically scrutinized the effect that government corruption has on the stringency of energy policies implemented in 11 industry sectors. Fredriksson.

(2003) test and the Fisher type tests using ADF and PP tests proposed by Maddala and Wu (1999) and Choi (2001).13 0. making SP&W more attractive for Europe‟s RE future.01 0. led us to apply a cross section panel data analysis from 1997 to 2009 to a set of the EU-15 countries.61 14.07 -0.75 0. (2002) test. The first 10 .03 Source: Eurostat Energy Statistics . Table 1: Electricity production in Gw/h . the Im et al.64 1. consumption -nrg_105a The dependent variable is set to be the amount of electricity produced from SP&W as a percentage of total electricity production.We decided to examine only electricity production from SP&W mainly due to their higher growth rate in the past decade (see Table 1). EPC and CORPI. wave and tidal and cogeneration because of their most recent appearance in electricity production and the lack of data before 2005.38 0.21 1990-1995 21.08 0.64 -0. in the Appendix indicates that a nonlinear increasing trend is present. Furthermore. Based on this lack of support for a unit root we decided to proceed and use LSNWTT without differencing but taking into account that it has a deterministic trend. namely the Levin et al.Percentage change Wind Solar Hydro period/year Systems Photovoltaics Electric 4. LCOALPPC. On the other hand the Breitung (2000) test fails to reject the null of a unit root. so the logarithm of this variable was formed (LSNWTT). OTHPOL. in contrast to the aforementioned technologies‟. transformation. The graphical representation of the dependent variable depicted in Figure A.16 2005-2009 0.63 1. while Table A1 in the Appendix shows their descriptive statistics.1.67 0. The EU policy schemes and measures concerning the stimulation of investments on Renewables date back to the 1990s. ELEC. Panel unit root tests with a linear trend and an intercept were applied to LSNWTT and the results are reported in Table A2 in the Appendix.94 0. However.85 8.supply.01 -0. LGASPPCR. Four out of five of the tests reject the null hypothesis of a unit root. We excluded Biomass. most of the EU members appear to have a clear-cut policy scheme supporting SP&W systems. LOILPPCR.13 Tide and wave -0.07 Geothermal 0. limitations on finding credible policy data along with the delayed adoption of SP&W from several European members. which are presented in Table 2 below.18 1995-2000 2000-2005 2. Explanatory variables The main explanatory variables used in the present research are FITS.

FITS. used Feed-in-Tarrifs and TGCs distinctly and dummy variables to depict the effect of other policies on RETs investments. as level-based disjointedly. In other words. (2011). This difficiency eventually prohibited us to treat the main policy instruments (OPTRES. applied for each source in each of the country examined. elaboration of our dataset showed that 12 of the EU-15 countries used Feed-in-tariffs as their main policy mechanism and complements their policy strategy using one of the rest instruments. LOILPPCR and LGASPPCR for Coal. However. FITs and TGCs. The authors used Feedin-tariffs and RE targets as level-based mechanisms and dummy variables to depict the effect of other policies on the enhancement of new patents. For instance. determined a 5% annual tariff reduction. which are proposed from the diffusion literature are policy schemes (FITS and OTHPOL). tenders and investment subsidies used in each country. 11 . As a result of not finding such data we proceeded in with calculating the tariffs of each source using the un-weighted average.two explanatory variables. consisted of information of just being implemented and not on their policy level.res-legal. However. Germany‟s Feed-in-Tariff mechanism which officially introduced in 1991 for wind and 1998 for solar. (2010). several European countries have designed complex tariff schemes oriented on the level of production from SP&W. 2007). Oil and Gas respectively) are introduced in order to illustrate the effect of how increased prices of the most competitive alternatives of SP&W systems influence the choice of investors. OTHPOL demonstrates the number of the implemented policy mechanisms but FITs. controlled to range from 0 to 1. al.de. Finally. for energy security reasons. tax and fiscal reductions. countries 4 Data on FITS and the OTHPOL were found through consulting the legislative frameworks of each country in the EU-15 countries along with references on the site http://www. we should bear in mind that the EU and the rest of the world gradually diminishes traditional energy sources such as oil and coal and goes over less polluting alternatives like gas. In addition. In that sense. Popp et. with higher production levels receive rather smaller tariffs. We should expect that higher prices would result in higher levels of the share of solar PV and wind systems confirming diffusion theory. In this direction was the methodology followed by Johnstone et. include the price producers of SP&W systems are compensated at the year of adoption without taking into account any possible digression imposed to producers fees any year later. In addition to policy proxies described earlier. An increase of the OTHPOL index signifies the adoption of one or more mechanism supporting SP&W. al. Furthermore. the natural logarithm the real prices of conventional sources (LCOALPPCR. on the other hand. It would be useful to mention that the data found 4 concerning the other mechanisms apart from FITS. Italy imposed a tariff reduction of 2% for solar photovoltaic systems since 2007. where FITS represent the source weighted average of Feedin-Tariffs offered in SP&W systems in each of the country of our dataset and OTHPOL is the source weighted average of the number of other policies namely TGCs.

transformation. 5 The Corruption Perceptions Index (CORPI) is an aggregate indicator that combines different sources of information about government corruption.de. ownership (private versus public) and the degree of vertical integration. It captures information about the administrative and political aspects of corruption. According to CORPI. taken from Transparency International (http://cpi.may support both RETs and less polluting technologies using gas. The regulation index is a composite index that takes into account several aspects of the electricity market such as entry regulation (possibility of third party access to the electricity grid). P.res-legal. Table 2: Variables Definition. kickbacks in public procurement. Furthermore. in order to examine the manner in which SP&W power generation is affected from the different regulation frameworks (ELEC) of the EU-15 countries we use an electricity regulation index formed from Conway. market demand condition is controlled using the energy consumption per capita (EPC) which is translated in KToe per capita. Given that policies have managed to provide a competitive advantage for SP&W investments. DG TREN COALPPCR GASPPCR Gas Prices per country in real terms in US dollars OILPPCR FITS OTHPOL CORPI Oil Prices per country in real terms (US dollars) Level of applied Feed-in-Tariffs weighted by the percentage of each source Number of policies supporting each source. in an attempt to control for the institutional factor of corruption we use the corruption perceptions index5 (CORPI). EC. Own research http://www. and questions that probe the strength and effectiveness of public sector anticorruption efforts (Transparency International).transparency. making it possible to compare countries. the EU-15 countries that rank the best are Finland. 12 . enhance the diffusion of SP&W.org) which ranks countries according to the perceptions of corruption in the public sector. Nicoletti (2006) for OECD countries.res-legal. weighted by the percentage of each (index 0-1) Country corruption perception index (index 0-10) larger: better performance 0 unregulated ELEC Electricity Regulation Index (index 0-6) EPC Energy consumption per Capita kg/capita Finally. Consequently. increase of EPC may direct new investments on RES thus. Own research Transparency International OECD Indicators of regulation in energy. embezzlement of public funds.de. Broadly speaking. the surveys and assessments used to compile the index include questions relating to bribery of public officials. and G. transport and communications (ETCR) Eurostat December 2010. consumption -nrg_105a BP Statistical Review of World Energy 2011 BP Statistical Review of World Energy 2010 BP Statistical Review of World Energy 2009 http://www. units and source Definition LSNWTT Logarithm of Spv and Wind electricity production to Total Coal Prices per country in real terms in US dollars Description Dependent variable real terms using CPI IMF real terms using CPI IMF real terms using CPI IMF €cents Source Eurostat Energy Statistics supply.

(2) below. Note that as discussed previously. Likewise.2. 4. Italy and Portugal because their average index for the same period is worse than 6. Vollebergh and Dijkgraaf . These type of indices have been empirically used by scholars on political theory. In order to check for possible autocorrelation of the error term in (1) we follow Wooldridge (2002). we introduce in our specification. We expect that low corruption perceptions (high values of the CORPI index) will most likely be linked to higher investment funds allocated to RE and especially to wind and PV systems.Denmark and Sweden because their average index from 1997 to 2009 exceeds 9 and the worst are Greece. Econometric Model and Results In order to uncover which of the above discussed factors affect the diffusion of SP&W technologies we use the panel data specification given by eq(1) below. 13 .073 and therefore the null hypothesis that the corresponding population parameter is equal to -0. Note that following our previous discussion a deterministic time trend has been introduced as well. 2004. 2006. we have introduced an interaction term between the corruption index and the Feed-in-Tariff value in order to test whether government corruption perceptions affect the effectiveness of Feed-in Tariffs as a policy instrument. but never before in the literature of RET diffusion. and estimate the first difference (FD) transformation given by eq. Ederveen et al. Moreover since it is possible that SP&W electricity generation is not only responsive to current levels of FITs but also to its past levels. investments and growth (see Fredriksson. where LSNWTTit is the natural logarithm of the percentage of electricity produced from SP&W sources in country i at time t. Evrensel 2010). αi represents unobserved individual (country) heterogeneity and εit is the usual idiosyncratic error. the interaction term between corruption and the level of regulation in the electricity market tries to capture any effects of corruption on the effectiveness of deregulation towards the diffusion of SP&W technologies.5.5 is rejected. The estimated first order correlation coefficient of the FD residuals is with a standard error of 0.

0557* 0.0415 5. According to the diffusion literature discussed in section 2.0016* 0.0016* 0.0928 0.0342 165 1.0605 5.0676 0. Jaffe and Stavins.0251 0.0168 165 0. In order to confront this endogeneity problem we estimated 14 .Another issue that is of great importance for the validity of the estimation results is whether there are some issues of endogeneity in our model.0672 0.1038 -0.0031 -0.3644 0.0029 -0.0113 0.0238*** 0.55E-05** 0. Table 3.0882* 0. Error Estimate 0.0593 0.0110 0.53E-05 -0.0142 0.0346 -0.0651* 0.0310 -0.1042*** 0. 1995). it could be the case that higher levels of deregulation could be set so as to achieve a higher deployment of SP&W.0415 0.0181 -9.0925 0.0183 0.13E-05* 0.0217 0.0069 0.0323 0.5691 (0. Error Estimate Std. Estimation results for FD / Model I Independent Variables Constant Term D(LOILPPCR) D(LCOALPPCR) D(LGASPPCR) D(FITS) D(FITSt-1) D(OTHPOL) D(CORPI) D(EPC) D(FITS*CORPI) D(ELEC) D(ELEC*CORPI) N*T OIR (p-value OIR) */**/*** Significant at the 10%/5%/1% levels FD 2SLS / Model II Std.0001 -0.0462 0. For instance support mechanisms such as FITS and OTHPOL could be adjusted downward as the share of SP&W increases because investment costs are lower.66) 0.0771 0.0105** 0.0076 -0.0099 -0. We test for strict exogeneity of FITS.0290 0.0111*** 0.0063* 0. government intervention through policy mechanisms should be treated as (or at least tested for) endogenous (Söderholm and Klaassen. OTHPOL and ELEC and the interaction terms involving them.0129*** 0. 2007.0104 -0. In the case of ELEC.1464 0.0177* The results of the testing procedure show that only ELEC and ELEC*CORPI violate the strict exogeneity assumption.

Note that the standard errors in Table 3 are the White period robust standard errors which are robust to serial correlation.5691 which is distributed as a chi-square with 3 degrees of freedom with a p-value of 0.66. towards the mitigation of GHG emissions. the differences between the robust standard errors of the two estimated models are reasonable although as expected they are bigger in the case of IV estimation. We use two dummy variables concerning the two most significant dates of European legislation. 15 . thus necessitated the rapid deregulation of the EU members electricity markets. as instruments for the regulation variable. Furthermore. (2) and on the two instruments we use shows that both of them are significant. we fail to reject the null hypothesis that the over-identifying restrictions are valid since the test statistic is OIR=1. which has been treated as endogenous. For the interaction term CORPIit interacted with each one of the two above dummies and also with LOILPPCRit. As depicted on Table 3. Finally. Both of the dates were selected in the sense that both events enhanced the adoption of RETs through the imposition of targets.eq. has a significant positive impact on the share of SP&W but this effect gets smaller when CORPI increases (lower perceived levels of corruption). above. energy consumption per capita has a significant (at 10% level for Model II) negative and a rather small effect to LSNWTT indicating that increasing energy needs are covered from the use of other conventional or renewable sources and not from SP&W. Continuing with the results for Model II. The regression of the dependent variable on all strictly exogenous variables in eq. we find that an increase of conventional sources prices does not have any impact on accelerating the diffusion process of the technologies under study. and worldwide action. while FITSt1. Turning now to the parameter estimates. If we focus on the FD 2SLS results we find that none of the prices of traditional energy sources has a significant effect on the share of SP&W. OTHPOL has a significant effect at the 10% level and negatively affects the deployment of SP&W systems. In other words. we find FITS to have a positive and significant effect on the proportion of SP&W to total electricity production at the 1% significance but this effect is smaller for higher levels of corruption perceptions (lower perceived levels of corruption). (2) by FD 2SLS methods using strictly exogenous instruments for both endogenous parameters. Additionally. turns out to be significant (at the level of 10%) and positive in the FD 2SLS model. Directive 2001/77/EC (EU01). the process of gradual deregulation in the electricity market (lower index of ELEC). In addition. we note that correcting for endogeneity has mostly affected the parameter estimates for the regulation variable and its interaction effect which more than double in magnitude. Kyoto Protocol (KYO05) ratification on 16 February 2005.

5. which represents the level of €cent per Kilowatt that producers receive. Our results indicate that implementing an additional policy mechanism will induce a reduction in the proportion of electricity generated from SP&W systems thus it will act as constraint in the allocation of funds. potential investors hold beliefs that corruption will also “leak” to the SP&W market and that they can secure certain returns through parallel channels to Feed-in Tariffs. who gives the example of Spanish solar panel operators being paid “for producing solar energy during the night”. Maoro 1995). Fredriksson. when entrepreneurs believe their country is highly corrupted and that they can easily conduct fraudulent activities then any small increase in the levels of Feed-in-Tariffs will be more effective than in less corrupted environments. we noted that the EU-15 members use either Feed-in-Tariifs or TGCs as their main policy instrument and complements their scheme with other secondary mechanisms. we confirm the findings of Söderholm and Klaassen. except Feed-in-Tariffs. In the present study we capture both of these effects. Vollebergh and Dijkgraaf. that Feed-in-Tariffs do have a significant positive effect on renewable energy technologies diffusion. Discussion The literature dealing with the diffusion of environmental friendly technologies argues that government intervention through policy schemes can create both incentives and constraints to the process of technological substitution. Introducing the variable FITS. we find that an increase in the level of FITS indeed spur investments towards SP&W systems in the EU-15 countries. An exception to this rule is Finland that uses investment subsidies as its main policy mechanism in the time span 1997 to 2009. and in addition negatively influences the efficiency of policies (Fredriksson. 2004). In order to further support this view we refer to White (2010). is less strong when corruption is perceived as being low. In any case. A parallelism exists here with the topic of agricultural subsidies whereas increasing subsidies could lead to an increase of farming activity which is especially strong in countries where producers misreport and inflate the level of their production and authorities fail to combat fraud. in order to attract investment funds towards SP&W. Thus. since corruption acts as a deterrent in allocating investment funds (i. To begin with. This may look as a paradox but a closer look at the policy data could provide us with an explanation. However the magnitude of this effect decreases with the level of the corruption index or in other words. This result comes in contradiction with the literature studying government corruption. (2007).e. List and Millimet 2003. In order to capture the effect of the other policy mechanisms implemented in the EU15 we introduced the index OTHPOL which represents the number of policies applied to support SP&W systems. It is also clear that the 16 . in a previous discussion on policy mechanisms. At a first sight this result might seem counterintuitive but it could just portray a situation where in countries which are prone to corruption.

suggests a positive relation between competition enhancement and investment decision differentiation towards SP&W. it could be the case that a country which considers its policy scheme inefficient and implements another mechanism instead of improving the efficiency of the existing ones. As diffusion theory suggests. al. would distort investment incentives towards the new technologies. the EU countries with high levels of energy consumption could be facing either high costs or production limitations in producing electricity from SP&W and as a result turn to other renewable or conventional sources. Thus. The negative sign though. altering either their main and secondary supporting instruments. Angolucci. This effect is described in the diffusion literature as lack of policy coherency resulting in raising the investment decision risk (Jaccobson and Bergek. found that electricity demand growth has a negative but insignificant effect on investments made in RETS. In our case. Popp et. oil and gas appear to have no impact on the investors‟ choice towards SP&W systems. 17 . in other words deregulation‟s positive effect is smaller in less corrupted environments. apart from the incentives and constraints government intervention produces. 2007). it could be the case that several countries analyzed in the present research reached their present policy strategy through experimenting. which comes in accordance with the literature of firms strategic choice (Delmas et. Finally. according to our findings the prices of alternative sources of electricity production such as coal. This again could be explained by the fact that deregulation in more corrupted environments might attract more “investors” trying to benefit from less stringent conditions against fraud. al. In this direction are the findings of Söderholm and Klaassen. On the other hand. (2007). One policy implication of this result is that limiting energy consumption per capita might be crucial when devising energy policies. 2011. Yet again.. Nevertheless. the magnitude of this effect decreases as perceptions on government corruption are reduced. that discuss about the impact of liberalization in the electricity sector and agreed that its impact on investments cannot yet be fully resolved. controlling for the energy market demand conditions we find a small negative effect on the proportion of electricity produced from SP&W systems. the estimate we find for the electricity sector regulation reveals a clear-cut result for the impact it has on the deployment of SP&W.EU-15 countries each time they form their policy strategy include both main and secondary mechanisms. 2004. the decision of investors to install SP&W technologies is subject to the profitability of other alternatives. that found coal price to be insignificant to the diffusion process of windmills in the EU. 2008). Continuing with the factors our research controls for. This comes in contrast to Sunaidy and Green (2006).

forging changes on the efficiency of the implemented mechanisms and not on the concreteness of the whole scheme.6. Furthermore. The compulsory time period where the EU-15 members experimented to find the best policy scheme appears to have acted as a drawback in the SP&W diffusion process. we find that when corruption levels are perceived as being low then markets should be more deregulated in order to promote investments than when corruption levels are high. by raising investment risks. Conclusions and Limitations Diffusion theory. One of the leading characteristics that have promoted the enhancement of new investments in the electricity industry for the EU-15 has been market deregulation. the present research results provides some useful guidelines to policy makers. it is shown that the deployment of SP&W cannot catch up with a growing energy consumption and energy policies should also focus on the consumption side in addition to the production side. we show that the EU-15 countries should apply a coherent policy scheme strategy. Improving market competition can be considered as a prerequisite for new investments to take place and consequently to gain from the incentives effective policy schemes provide. we have only examined the efficiency of 18 . the results of the present research indicate that the diffusion process of SP&W systems is greatly determined by the intensity and the coherency of the applied policy schemes and the market characteristics to which they are targeted. In addition. Increasing competition in the electricity market results in new entrants having the opportunity to invest on RETs and in firms. which acted in the market before deregulation. A limitation of the present research is stemming from not including technological change and environmental regulation as mechanisms to proliferate RETs. This contradictory result could be partly stemm from to the fact that we included extreme cases of corruption in our analysis and its further analysis is left for a future study. differentiating their strategic choice towards RETs. it appears that policy mechanisms must be devised or adjusted according to the market characteristics which they a target. argues that government intervention could act as a lever to the successful deployment of RETs by eliminating technology lock-out barriers. Our results show that for the EU-15 increases in the Feed-in-tariff levels have a positive effect on the share of SP&W which spreads more than one period. Further. could provide a better understanding to policy makers and aid them in devising successful policy strategies. However the inclusion is delayed for future work. In addition. To this extend. On the other hand. Nevertheless. FITs seem to be more effective in economies in which investors believe corruption is high and we have linked this effect to the higher opportunities for fraud existing in corrupted environments. Overall. Therefore. the study of the factors that affect the diffusion of new environmental friendly technologies.

19 .the Feed-in-Tariff mechanism. the rather small panel size with N=15 and T=12 precludes us from using dynamic effects or lagged dependent variables which could influence the results. Finally. excluding other important supporting mechanisms such as TGCs and Investment Subsidies because data was not available to us.

1. LSNWTT Aus -2 -4 -4 -5 -6 -6 -7 -8 -8 -10 -9 -10 90 92 94 96 98 00 02 04 06 08 90 92 94 96 98 00 02 04 06 08 -3.0 -2.5 -4.0 90 92 94 96 98 00 02 04 06 08 -12 90 92 94 96 98 00 02 04 06 08 Fr -4 -6 -2 Ger -2 Gr -2 Ire -4 -4 -4 -8 -6 -10 -8 -8 -8 -6 -6 -12 -14 90 92 94 96 98 00 02 04 06 08 -10 90 92 94 96 98 00 02 04 06 08 -10 90 92 94 96 98 00 02 04 06 08 -10 90 92 94 96 98 00 02 04 06 08 It -2 -4 -3.5 -10 -6.5 -2.5 -6 -5.0 -4.Appendix Figure A.5 -8 Bel -1.5 -10 -3.0 Den -4 Fin -6 -12 -4.0 -6.0 -8 -5.5 90 92 94 96 98 00 02 04 06 08 90 92 94 96 98 Lux -3 -4 Neth 0 -2 Por -5 -4 -6 -7 -6 -8 -12 -8 00 02 04 06 08 90 92 94 96 98 00 02 04 06 08 -10 90 92 94 96 98 00 02 04 06 08 Sp 0 -2 -4 -2 -4 -6 Swe -2 -4 -6 UK -6 -8 -8 -8 -10 -10 -10 90 92 94 96 98 00 02 04 06 08 -12 90 92 94 96 98 00 02 04 06 08 -12 90 92 94 96 98 00 02 04 06 08 20 .

97539 18. 2009.593908 4450. Lin & Chu t* Breitung t-stat Im.997 N Desc 195 195 Real prices 195 Real prices 195 Real prices 195 €cents 195 Index 0-1 195 Index 6-0 195 Index 0-10 195 Ktoe pc COALPPCR 58.88292 27. For the other three tests.64304 0. All other tests assume asymptotic normality.64568 46. Pages 1321-1339.780 Std. “The Local Power of Some Unit Root Tests for Panel Data”.036115 0. 161–178. The null hypothesis for the first two tests is a unit root (assumes common unit root process.18072 2.100 Method Table A2: Panel Unit Root test on Lsnwtt Statistic Prob.Fisher Chi-square PP . Issue 7.002.000 0..** N -7.04223 Median 0. Breitung.000000 0.800000 2137. Why is corruption less harmful in some countries than in others? Journal of Economic Behavior & Organization 72 797–810.. Amsterdam: JAI Press.232898 2.243 65.986 67.599163 0.967081 0. bureaucratic corruption and economic development. Paolo.026032 42. 2008. Nonstationary Panels. Volume 29.009570 31. 10.000000 10.980 Maximum Minimum 0. Journal of International Money and Finance.484639 4.. Automatic lag length selection (Schwarz Information Criteria) used. Financial liberalization. Blackburn K.34135 140..200000 2. Donald Lien and Sanjay Pradhan.jimonfin. 2010. pp.Gonzalo F.159962 2. 27.43817 4.000 193 178 193 193 194 Levin.700000 3988. the null hypothesis is a unit root (assumes individual unit root process. ISSN 0261-5606.262 0.15 2. Campos J.343333 7. Pesaran and Shin W-stat ADF .1016/j. 1999. 6. The Impact of Corruption on Investment: Predictability Matters.41758 2.00000 10310. Dev.151360 1. Renewable and Sustainable Energy Reviews 12 pp 141–161.400000 6. BP.040112 21. J. 2010. 15.001 0. p. No.Table A1: Descriptive statistics Mean SNWTT OILPPCR 0. November 2010.05.000000 7.189180 92.59657 GASPPCR FITS OTHPOL ELEC CORPI EPC 5. 2000.000 0.647 -0.58E-05 13.88042 31.2020 10. Factors influencing the likelihood of regulatory changes in renewable electricity policies. and Dynamic Panels. References Agnolucci. World Development Vol.2010. Advances in Econometrics. 1999.000000 0. 0.158988 4. Panel Cointegration. Forgues-Puccio.580723 1. Gonzalo F. Edgardo.560528 1694. 1059 1067.000000 3.639 -3. Statistical review of world energy. 21 .726 0. Forgues-Puccio. Individual intercept and time trend included in test regressions.Fisher Chi-square ** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution.191628 0. Blackburn K.

pp 815-549. C. Fredriksson. Energy Policy 37:3071–3081. Journal of Industrial Economics 51 (3). Fredriksson. Delmas.. Pages 1135-1144. 1998. Issues 7–8. Volume 35. and technology choice. Ragwitz. W. Newell. 207-231. Rathmann... Volume 87. "Product market regulation in non-manufacturing sectors in OECD countries: measurement and highlights". & Robert M.eurostat. N. 6163–6187. 28: 189–209.. Bergek. Pesaran and Y. “Unit Root Tests for Panel Data”.. June. Johnstone N. environmental policy and inbound US FDI: theory and evidence. Journal of Environmental Economics and Management 29. 22 . Renewable energy policies and technological innovation: evidence based on patent counts. 249–272. Deregulation and environmental differentiation in the electric utility industry." Energy Policy. Russo. René. S. Falconett Irina. Klessmann. transformation. Shin. Mgmt. Pages 1407-1430. lead phase down. R. Testing for unit roots in heterogeneous panels.Choi. State renewable energy electricity policies: an empirical evaluation of effectiveness. P.. Everhart. Jacobsson. and A. 235–256. 39: 7637-57..H. supply. Hašcic I. 2011. S43–S63. 2007.europa. Delmas. Magali A. Environmental and Resource Economics 12: 113–136. Applied Economics. A. Environ Resource Econ 45:133–155. International Review of Economics and Finance 19 501–514. Number 5.. http://epp. The Political Economy of Private Electricity Provision in Southeast Asia. Jaffe. and R. Millimet. Zelner. I. 2007. 2004. August. Henisz.A.J. B. M. ISSN 0960-1481. Code: nrg_105a.J. Environmental regulation. Nicoletti (2006). Jamasb.S. Ken Nagasaka. pp. J. J. Journal of International Money and Finance. 2009. Renewable Energy. US State Policies for Renewable Energy: Context and Effectiveness. M. 2004. 10. John A. Eurostat: Energy Statistics.Vazquez J. EU (2009) Directive 2009/28/EC on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/30/EC.G. governance.. 2011. Policy-induced technology adoption: evidence from the U. S.. 2010. Maria J.. investment and growth in emerging markets. Bureaucratic corruption. Corruption and Energy Efficiency in OECD Countries: Theory and Evidence. 2010. and Montes-Sancho. Journal of Environmental Economics and Management. Per G. 1579-1594 Evrensel..B. Kerr. Strat.J. 41:13. M. „Dynamic Incentives of Environmental Regulations: The Effects of Alternative Policy Instruments on Technology Diffusion‟. K. 47.. Jones Center Working Paper WP 2001-02. 317–343. 115: 53-74. B. 35. Journal of Industrial Economics 46 (2). Reginald H. Econometrics. and Montes-Sancho. Tooraj & Pollitt. M. Carley S. January 12. 1998. Conway. Popp. Vol.Stavins.530. J. p. 20. and growth volatility. 2001. Michael. Im. W. consumption. 2003. Corruption. Kemp. The Wharton School. Per G. Held.. Shadbegian. 'Transforming the energy sector: the evolution of technological systems in renewable energy technology'. growth. "Incentive regulation of electricity distribution networks: Lessons of experience from Britain. Daniel L. Energy Policy. 2001... Journal of Public Economics. and Dijkgraaf. Issue 6. Gray. McNab (2009): Corruption. M. and G. List. D. 2003. Ayse Y. 2003. 2010 Comparative analysis of support mechanisms for renewable energy technologies using probability distributions.S.. Elbert. S. A. M. R. Industrial and Coorporate change. V. 1998. OECD Economics Department Working Paper No. The diffusion of biological waste-water treatment plants in the Dutch food and beverage industry. Vollebergh. Martinez.ec. 1995. investment timing. Herman R. University of Pennsylvania.. Volume 13. Status and perspectives of renewable energy policy and deployment in the European Union – What is needed to reach the 2020 targets? Energy Policy.eu/portal/page/portal/energy/data/database....

10. 2006. S. 1997 – 2009. Wind power in Europe: a simultaneous innovation– diffusionmodel. Volume 69. Oil and natural gas prices and greenhouse gas emission mitigation.” Journal of Econometrics. 431-435. Pizer A. J. 61.” Mimeo.. 1–17. and R. April 12. Stavins." Energy Economics. William & Popp David. 15 September. 2010. Maddala. Fostering the use of renewable energies in the European Union: the race between feed-in tariffs and green certificates. Issue 11. ISSN 0921-8009. Pages 648-662.. Klaassen. Snyder. Energy Economics. 1995. María Hierro. 1993. Nilsson L. Johansson B. 2006. OPTRES 2007. Volume 33. A. 2007. Pages 2980-2991. "The Effects of Environmental Regulation on Technology Diffusion: The Case of Chlorine Manufacturing. Tilburg University. and C. Seeing the wood for the trees: 25 years of renewable energy policy in Sweden. Econometric Analysis of Cross Section and. G. University of California. 2010. Assessment and optimisation of renewable energy support schemes in the European electricity market. A Comparative Study of Unit Root Tests with Panel Data and New Simple Test". Renewable electricity consumption in the EU-27: Are cross-country differences diminishing?. 10. 631-652. ISSN 0140-9883. and Neij L. Börjesson P. Panel Data. Söderholm. 2006. Elsevier.. Environmental and Resource Economics 36. José Villaverde. pp 67-81. S. 2011. MIT Press Wang Y. Issue 4. 2. Pages 2094-2101. Ringel Marc. Technology and the diffusion of renewable energy.. Wooldridge.transparency. Ecological Economics. vol. November.J. Energy Policy 37:4797–4808. No.org/cpi. Recommendations for implementing effective & efficient renewable electricity policies. 93. July. Paolo. Mauro. Van Vuuren D. Mulder Arjen. 2002. (2002). Patrik.M. http:// www.Levin. 2008. ISSN 0960-1481. Green. Ger. September. 108. Oxford Bulletin of Economics and Statistics. Anastasios and De Zeeuw. Maza Adolfo." The American Economic Review. Levin.. Corruption and the environmental Kuznets Curve: Empirical evidence for sulfur. Neelakshi Medhi.. Renewable Energy 31 p. Volume 30. 23 . and Robert N. 8(1). 769–787. 2006. Van Ruijven B. Report of the IEE project OPTRES. San Diego. Ivan Hascic. A. 2008. Lin. 2004. White T. 1999. Energy 31. 163–190. pp. Energy for Sustainable Development. Corruption and growth. Åstrand K. Vol. Alexandra Leitão. Vol. Do economic instruments matter? Wind turbine investments in the EU(15).” Bloomberg Businessweek. 2010. Nolan Miller. David Popp.. Issue 9.. "Endogenizing technological change: Matching empirical evidence to modeling needs. 30(6). 10. Electricity deregulation in OECD countries. Chu. “Unit Root Tests in Panel Data: Asymptotic and FiniteSample Properties. Pages 2191-2201. “Spanish Solar-Panel Trade Group Calls for Fraud Investigation. 1998-38. Issue 6. C. Volume 35. Lin. Energy Policy 34:1209–1220. 10. Sunaidy. 1–24.. Svennsingsson P. Energy Economics. Wang Y. Lori. 2009. Ericsson K. Center for Economic Research Discussion Series No. and Wu. and J. 2003. Environmental Policy and Competitiveness: The Porter Hypothesis and the Composition of Capital. ISSN 01409883. “Unit Root Tests in Panel Data: Asymptotic and Finite Sample Properties. Aart J.. Quarterly Journal of Economics 109. Xepapadeas. Energy Policy 34:1209–1220. Transparency International. Renewable electricity in Sweden: an analysis of policy and regulations. 1998. Renewable Energy.. 681-712. A.. Renewable electricity in Sweden: an analysis of policy and regulatiosns. pages 2754-2770.

Sign up to vote on this title
UsefulNot useful