UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effects of green energy policies on innovation Prescott, Ryan 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2009_fall_prescott_ryan.pdf [ 1.79MB ]
Metadata
JSON: 24-1.0068271.json
JSON-LD: 24-1.0068271-ld.json
RDF/XML (Pretty): 24-1.0068271-rdf.xml
RDF/JSON: 24-1.0068271-rdf.json
Turtle: 24-1.0068271-turtle.txt
N-Triples: 24-1.0068271-rdf-ntriples.txt
Original Record: 24-1.0068271-source.json
Full Text
24-1.0068271-fulltext.txt
Citation
24-1.0068271.ris

Full Text

The Effects of Green Energy Policies on Innovation  by  Ryan Prescott  BSc, University of Victoria, 2006  A thesis submitted in partial Fulfillment of the requirements for thó degree of  Master of Science  in  The Faculty of Graduate Studies (Agricultural Economics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2009  © Ryan Prescott, 2009  ABSTRACT  Over the last twenty years, climate change and energy stability have increasingly played a role in government policy. Innovation is a key ingredient to increased growth while achieving climatic goals. The policy paths taken by different countries have been extremely diverse. What effects do these policies have on innovation and what policies are best at encouraging the market? This thesis uses a theoretical model to examine the effect that different policies have on innovation and to empirically test a series of hypotheses obtained from the model. Both the theoretical and empirical results support the hypothesis of a “home bias”, whereby innovation in domestic markets is impacted more than innovation in world markets when a domestic policy is initiated. A second result is that mandatory renewable energy minimum levels for power companies have relatively strong impacts on installed wind capacity. In general, a country’s wind capacity is determined by both economic and non-economic considerations.  11  TABLE OF CONTENTS  Abstract  II  Table of Contents  iii  List of Tables  iv  List of Figures  v  1  Introduction Background 1.1 1.2 Problem Statement Research Objectives 1.3 Thesis Description 1.4 2 An Overview of Green Energy Markets 2.1 Technology Background Policy Background 2.2 2.3 Literature Review of R&D Policy Theoretical Model 3 Introduction 3.1 3.2 Stage Two Cournot Competition Stage One R&D Race 3.3 4 Empirical Empirical Model 4.1 Patent Data 4.2 4.3 Policy Variables Other Variables 4.4 Results 5 5.1 Patenting Regression One Market Size Regression Two 5.2 6 Brief Country Review 6.1 Germany France 6.2 Denmark 6.3 Conclusion 7 References —  —  Appendix  •  1 1 3 5 6 7 7 12 17 23 23 26 33 41 41 47 50 51 53 53 56 60 60 61 62 63 65 68  111  LIST OF TABLES Table 4.1: Table 5.1: Table 5.2: Table 5.3: Table 8.1: Table 8.2: Table 8.3:  IPC codes for Wind Results of Patent Regression Fixed Effects Market Size Results Random Effects Market Size Results Change in Research as Feed-in-Tariff Chges Data Descriptive Statistics  49 54 57 58 68 68 78  iv  LIST OF FIGURES Figure 4.1: Patents from 1990-2006 for Germany, Dnmark, Great Britain and Austria in Wind Energy 50 Figure 4.2: Electricity Output (GWh) for Germany, Denmark, Great Britain, Austria and OECD 52  V  1  INTRODUCTION 1.1 Background With ongoing worldwide concern about the effects of climate change and large  fluctuations in oil prices, governments are increasingly looking for policies and incentives to help curb carbon emissions and allow continued economic growth. Europe is leading the way with a mandated decrease in greenhouse gas levels by 20% compared to 1990 levels and a mandated 20% increase in renewable energy by 2020 (BBC News, 2009).  As developing countries such as China and Irdia continue to grow, world energy  demand is expected to increase substantially. Renewable energy will undoubtedly play a key role in meeting both the world’s climatic goals and allowing increased energy price stability. Energy price instability is due to increased worldwide demand, political instability in large oil producing countries, and decreases in easily accessible oil stocks. There has also been speculation that current oil stocks may have peaked already and that worldwide production may decline in the near future (Bentley, 2002). Currently most renewable energy sources (excluding large scale hydro) are in their infancy of development and are not economically feasible without government incentives. Therefore if governments want to meet their future growth and climate objectives without large subsidies, innovation will have to play a key role.  The effects of government policies on the growth of renewable energy sources can be seen in annual growth rates of renewable energy sources of countries such as Denmark, Spain and Germany (OECD/IEA, 2004). Denmark was one of the first to  1  implement a carbon tax in the early 80’s, while Germany and Spain introduced several policies including feed-in tariffs in the early 90’s. Knowing how different types of government policies affect incentives for innovation by different sources of renewable energy will better allow government officials to design policies which leverage their individual country’s competitive advantage at a comparatively low cost.  There are several fundamental economic reasons why policies should be introduced to encourage research and development, and innovation. Economists generally agree that research and development (R&D) should be encouraged due to the spillover effects associated with R&D and innovation. The spillovers are a result of the innovating firm not being able to capture all the potential benefits of the innovation because of imperfect property rights, which results in a social valuation greater than the private valuation. This can be the result of such things as the innovation not being patentable or the inability of the innovating firm to protect its innovation from copying. However, if through a patent or some other form of intellectual property rights an innovator is protected these spillovers would be eliminated. This would remove the need for a government policy intervention.  It is generally agreed that private industry on its own does not supply a sufficient amount of R&D because private valuation of a product or innovation is less than the social valuation of the product or innovation because innovators cannot capture all of the gains from the innovation due to imperfect property rights. When looking at the issue of green technology, the pollution externality also plays a significant role in social benefits calculations. From a social benefit perspective there is an added external benefit in  2  encouraging the growth and innovation in green technology in the form of less pollution. Therefore it is in a government’s interest to examine ways to encourage innovation and R&D in renewable energy to both address spillover effects and reduce the social cost of pollution.  A variety of policies have been used throughout the world to try to raise the level of R&D and growth in alternative energy to achieve a more socially optimal outcome. The most basic policy is the provision of an R&D subsidy to increase R&D to a more socially optimal level. Other policies, which are more directly related to the environmental externality, are designed to increase reliance on renewable energy or to decrease reliance on fossil fuel based energy. Permits and taxes have been used in the area of energy supply in much the same manner as any other form of environmental regulation. Price guarantees or price tariffs have been used to raise the price paid for renewable energy above the price paid for traditional energy, to allow for continued growth and innovation in renewable energy.  1.2  Problem Statement  The market structure for renewable (“green”) energy is unique in many ways. Green energy firms compete in a race with each other to develop new green technologies. The winner of the race then uses their newly developed technology to compete with traditional forms of energy in the energy markets, often relying on government policies to make them more cost competitive. Given the broad array of innovation and green energy policies which currently exist in global energy markets, it is imperative that governments  3  understand the comparative impacts of their policies on domestic versus foreign firms and understand which policies are the most effective in absolute terms. Governments may want to implement policies that encourage domestic innovation more than international innovation in order to grow their domestic firms and give them a comparative advantage in the future. The reduction of the pollution externality and the encouragement of domestic firms may in practice have conflicting results. Domestic firms may not be the most efficient at producing green energy resulting in less pollution abated than if an international firm was supplying the green energy.  Underinvestment in private R&D and innovation due to spillovers is widely acknowledged; however, for green energy, additional environmental benefits are associated with innovations which result in less polluting forms of energy production. These social benefits, which extend beyond private benefits, are the main reasons governments have been increasingly designing policies which encourage innovation in the area of green energy. With increased globalization, governments also need to look at whether it is domestic green energy firms or foreign green energy firms that are being encouraged by their regulations. This could have a direct impact on what type of policy should be implemented, depending on the goal of the regulation.  Due to the imperfect competition nature of green energy and the domestic and international components there are few questions which should be asked: How are the R&D incentives of domestic and international firms impacted differently by the introduction of a particular R&D policy? Which policies have been the most effective in stimulating the growth in green energy? Knowing the answers to these questions will  4  better enable governments to implement policies which encourage domestic innovation and growth in green energy.  1.3  Research Objectives  The aim of this thesis is to examine the comparative impact of different policies on the innovation incentives facing domestic and international firms. The specific research objectives are to:  • •  • •  review current literature on innovation and the environment develop a theoretical model of a R&D race between innovation firms, the winner of which goes on to compete in a non-competitive energy market use the theoretical model to generate a set of testable hypotheses use empirical analysis to test the various hypotheses Innovation in renewable energy is the key to continued growth and addressing  climatic objectives. Unfortunately, innovation is inherently difficult to measure. Most research on innovation resort to use patent counts as a measure of innovation activity. Patents allow innovators temporary monopoly power and thus there is a major incentive to patent any new innovation. For innovations to be patentable they must be novel, involve inventive steps and be capable of industrial application (Dernis & Guellec, 2001).  Patent counts from European countries will be used in the empirical analysis. Specifically, results from patent count empirical analysis will be used to assess the absolute effectiveness of policies which are designed to encourage growth of renewable  energy sources. Patent count data will also be analyzed to examine the comparative impact of an R&D policy on innovation activity in domestic and foreign countries. Later  5  in this thesis I conduct a case study on wind power, the most advanced of the developing renewable green energy sources.  1.4  Thesis Description  The thesis will be divided into seven different chapters. Following this introductory chapter, Chapter 2 provides an overview of policy initiatives and technology changes in the green energy markets, and also reviews the relevant literature. The development of the theoretical model is contained in Chapter 3. Chapter 4 presents the data and empirical model. The empirical results are presented in Chapter 5. A brief discussion of specific green energy policies countries is contained in Chapter 6. Chapter 7, the final Chapter, concludes the thesis.  6  2  AN OVERVIEW OF GREEN ENERGY MARKETS This overview chapter will be broken up into four sections. Section 2.1 will focus  on the current technology of the four most developed renewable energy sources: wind power, ocean power, geothermal power, and biomass. Section 2.2 will look at the various policies that have been used throughout the world to encourage growth and innovation in alternative energy sources. The final section, 2.3, will be a literature review relating to R&D policy and the economic rational proposed to introducing government policy to encourage growth and innovation in alternative energy.  2.1  Technology Background  Renewable energy sources can be grouped into three main categories based on  maturity levels: first generation energy sources, which emerged from the industrial evolution, second generation energy sources, which are now becoming commercially viable, and third generation sources, which are those sources still under development (International Energy Agency, 2006). Excluding hydropower, which is widely used throughout the world, the remaining renewable energy sources fall into the following categories: biomass and geothermal (first generation); wind, solar hot water, solar photovoltaics and advanced bioenergy (second generation); and third generation, which consists of ocean energy, advanced geothermal, biorefinery technologies and advanced biomass. These technologies, which are still under development, have the potential to be important sources of energy in the future.  7  From the beginning, and currently in many developing countries, combustion of wood or biowaste combustion was used for trivial tasks such as cooking food and heating living quarters. However advances in technology have produced an array of new options for producing bioenergy. Bioenergy includes biogas production, combustion, gasification  and proylsis.  Biogas: The major source of biogas is through anaerobic digestion which involves the breaking down of biodegradable material using microorganisms in the absence of oxygen to produce methane and carbon dioxide rich biogas. The methane is then used in the same maimer as natural gas. The most common source of fuel for anaerobic digestion is agricultural products and waste such as landfill waste or sewage. This technology is often used for the treatment of waste water with high organic content.  Biomass combustion: The combustion of biomass is used in several different ways to produce bioenergy. The combustion of biomass such as wood or wood pellets has been used for heating of residences and with automation and standardization has lead to increased efficiency. The combustion of biomass has also been used in combined heat and power (CHP) plants throughout Scandinavia for district heating. Biomass, co-fired with coal, has also been used in large scale combustion plants for the production of electricity.  Biomassfer,nentation: Fermentation involves the conversion of sugars from such sources as corn, sugarcane or sugar beet into ethanol and another byproduct such as methane using microorganisms. The fermentation of anything that contains sugar can  8  result in ethanol; however, sources like sugar cane, sugar beets, molasses and fruits are most efficient because of their ability to produce ethanol directly from the sugars. Sources such as grains or cellulose using wood waste need an intermediate step to convert the primary source to simple sugars.  Prylosis: Prylosis is process in which biomass is converted to liquid (bio-oil), gaseous and solid (charcoal) through the super heating of the biomass in the absence of oxygen. There however has been little market implementation of this technique at the present time.  Geothermal: Geothermal energy comes from the harnessing of super heated underground aquifers or reservoirs. These reservoirs are heated via an outward transfer of heat from the earth’s core. Rainwater is usually used as the medium to transfer heat to the surface. When the rock characteristics are appropriate (i.e., permeable) the water will pool forming underground reservoirs of high pressure and temperature, essentially  geothermal fields. These fields can be grouped into two types: either water-dominated or vapour-dominated. In the simplest case, steam from the geothermal well is passed through a steam turbine to produce electricity. If the well produces hot water instead of steam, another working fluid is used with a low-boiling point. Water is used instead to vaporize the working fluid which is then passed through a turbine, condensed in water  and reused.  Solar: Solar energy can be broken up into two streams, photovoltaics (PV) and solar thermal power plants. Photovoltaics involves the use of sunlight to split positive and  9  negative charge carriers which produces an electrical voltage. The DC voltage is then converted to AC and distributed to the grid. The use of PV has increased substantially with installed capacity increasing from 110 MW in 1992 to 1809 MW in 2003 (International Energy Agency, 2006). The second stream of solar power is the solar thermal stream which involves the concentration of the sun’s energy to produce either a super heated liquid or gas. Small scale solar thermal is used to heat water and heat an individual residence. On the large scale there have been several proposed designs for large scale solar thermal plants such as the parabolic trough and tower plants. However the main components are all the same. There is a concentrator to concentrate the solar radiation. That concentrated solar energy is then used to heat a medium like water, resulting in steam. That steam is then used to produce energy like any other thermal plant. A major advantage of the solar thermal is that in times of bad weather a fossil fuel burner can drive the water steam cycle guaranteeing continuous supply of energy.  Wind: Wind is by far one of the most established renewable energy sources. The design of wind turning a horizontal or vertical turbine has remained constant for decades; however, since the 1980’s major advances in aerodynamics, structural design and resource assessment have resulted in 5% annual increases in energy yield of turbines (Herbert, Iniyan, Sreevalan, & Rajapandian, 2007). This has resulted in wind turbines increasing in size from less than 100kW in 1980’s to in excess of 3.5 MW presently.  Ocean: Ocean energy can be broken down into three areas: tidal energy, wave energy, and ocean thermal energy. Tidal, wave and ocean geothermal energy are all still  10  in their infancy and in the research, design and development stage. Therefore there is no leader at the moment.  Many companies have come up with very different ideas to harness the power of the tides. The most basic is a turbine in the water, whereby the tide rotates the turbine producing power. Currently the main source of tidal power is the result of barrage tidal power, where a barrage or dam is placed across an estuary. The barrage has openings for tidal water to flow into on its way to the estuary; the water turns a turbine resulting in power.  Wave power also has several competing technologies with no clear winner depending on whether it is offshore, near shore or shoreline waves. Presently onshore devices include the oscillating water column which consists of a partially submerged structure where air is trapped above the water free surface. Waves force water in from below forcing the trapped air outwards spinning a turbine. Another onshore device uses a convergent channel which water rushes into it until it overflows into a reservoir. That reservoir is then used as a conventional low head dam. Offshore devices include the Palemis, a floating cylindrical snake like device. Its hydraulic joints extract energy with the movement of the waves.  Ocean thermal uses warm water from the surface to produce steam in a working medium with a low boiling point to produce steam and run a generator with cold water from the depths to condense the working medium. This is, however, extremely inefficient and there are presently only demonstrations plants producing electricity of this nature.  11  With so many competing technologies and with ocean technology still in its infancy, it is likely that the best technology has yet to be discovered.  2.2  Policy Background  Countries throughout the world have taken vastly different approaches to encouraging renewable energy growth and innovation. Europe has a long history of encouraging alternative energy through not only practical demonstrations, but also policy intervention. This started in the early 1970’s and before, coinciding with the fear over fossil fuel supplies (OECD/IEA, 2004). In the 1970’s the main policy interventions throughout Europe, Japan and United States involved the funding of research and development and implementation of demonstration plants.  The pace of policy interventions throughout Europe and the United States continued slowly throughout the late 70’s and early 80’s as prices of oil receded following the oil shocks of 1973-1974. This changed in the late 80’s at which time the European commission, which previously had minimal effect on establishing a place in European energy policy making, started to establish a position (McGowan, 2000).  Many different policies have been used over the last thirty years to encourage the growth and subsequently the innovation in renewable energy sources. From country to country, the type and target of policies vary substantially depending on their needs. For example, Denmark has targeted growth in wind energy with capital grants incentives for wind turbines implemented in 1979 followed by guaranteed prices in 1981 (OECD/IEA,  12  2004). Canada did little to encourage any form of growth until the early 1990’s most likely due in part to substantial cheap hydro electricity.  The theoretical and empirical literature on the effects on these policies can be taken from not only environmental theory, but also trade theory. The environmental literature looks at the effects of policies such as taxes and pollution permits on the pollution. Trade theory looks at the effect of protectionist policies to protect a domestic industry. This is similar to renewable energy policy when a government is trying to protect the renewable industry and increase its output while competing with the traditional energy source. The most recent literature is from the biofuels debate which uses both branches of literature. Some approaches used by countries to encourage green energy are producer grants, individual or company grants, obligations, carbon taxes, green certificates, targets and feed-in-tariffs.  Tax incentives such a carbon taxes, charge carbon polluters in an attempt to internalize the effect of the externality resulting in less pollution. The optimal level of pollution can be achieved if the marginal damage of the pollution is known. Baumal and Oates (1971) discuss the difficulties of implementing a tax system to properly capture the damage of the pollution externality. Firms decrease the level of pollution until the marginal cost of abatement of another unit is equal to the marginal benefit. This does not directly subsidize non polluting firms but rather increases the cost of polluting firms resulting in non polluting firms being more cost competitive. The carbon tax increases the price of electricity resulting in a user pay system whereby the more you consume the more you pay in taxes. Empirical literature on pollution taxation is widespread. For  13  example Dahi (1979) looked at the effects of a gasoline taxation and finding the elasticity of demand to be -0.442.  Producer grants are given to producers to produce renewable energy. These grants are subsidies to increase the amount of renewable energy used. The subsidy decreases the cost of producing renewable green energy resulting in an increased share of the market. It does not, however, give the traditional sources of energy incentive to abate emissions. The resulting market price of electricity decreases because of the subsidy. This kind of subsidy results in everyone paying an equal portion of the subsidy. There is also a subsequent decrease in price of electricity.  Obligations require that a certain amount of renewable energy must be used in a country’s energy supply. This is usually done through a bid system whereby firms bid to supply some much renewable energy at a given price per kilowatt hour. The lowest cost bids are then selected until the amount equal to the obligation is filled. Each bidder receives a price equal to their bid price of supplying energy. This implies that the subsidy to each generator may be different as the price paid to each generator should be equal to its marginal cost (Menanteau, Finon, & Lamy, 2003). The additional price is passed along to consumers in the form of higher electricity bills implying a user pay system of financing. Obligations are equivalent to a subsidy on renewable output and a tax on traditional energy output (Fischer & Newell, 2004).  Recent research in the area of obligations has focused on ethanol mandates. Joust and de Gorter (2009) look at the impact of mandates combined with other incentives such  14  as biofuel tax credits and find they may work the same as consumption subsidies with no  effect on ethanol consumption. Different models have been used to look at the effects of the bio-fuel mandate on reaching national pollution targets and the overall effect on the economy (Dixon & Osborne, 2007).  Another example of obligations is domestic content protection. In this case there is a requirement to use a certain amount of domestic content in the production of a good. If the domestic good and the international good are perfect substitutes similar market reactions that would occur as a result of the introduction of renewable obligations.  Krishna and Itoh (1988) develop a theoretical model to look at content protection in oligopolistic markets and find that domestic firms benefit from content protection if the  two inputs are perfect substitutes. Other literature looks at other market structures such as perfect competition and monopolies and the effects of goods which are compliments or not perfect substitutes. Empirical literature from Beghin and Lovell (1993) looked at the effect of domestic content protection on the Australian cigarette and tobacco industry, finding that increased domestic content protection led to increased use of domestic Australian tobacco.  Feed-in-tariffs or price guarantees, guarantee a price for renewable energy above the price of traditional energy sources. Electricity is generated from the renewable energy source until its marginal cost equals the price given by the feed-in-tariff (Menanteau Ct al., 2003). Since the feed-in-tariff is greater than renewable energy suppliers would receive under normal conditions there is more renewable output. The benefit to renewable firms varies with their cost of production. Firms with low cost of production  15  collect large economic rents. The increased price given to the renewable energy source is  funded through an increase in the price of electricity to consumers resulting in a user pay system.  Consumer/individual grants are grants to individuals or companies to install renewable energy sources in their homes or businesses. The result is an increase in personal sources of renewable energy which might result in a small increase in renewable energy.  Targets are normally unbinding renewable energy targets set by government and depend on other policies to reach to the targeted goal. If targets were binding they would work in the same manner as obligations.  Green certificates function by allotting a certain amount of renewable energy that must be produced by each producer. This producer can either produce that output themselves or purchase it from another producer or from a specialized renewable energy firm in the form of green certificates (Menanteau et al., 2003). The renewable energy firm receives both the price received for the electricity and the price of the green certificate. The increased price is passed along to consumers in the form of higher electricity prices.  When all necessary information is available, price approaches such as taxes and quantity approaches such as obligations or green certificates can have similar effects on reducing emissions and increasing renewable energy output. The key difference is that  16  that with a tax or feed-in-tariff the exact amount of pollution abated or new renewable energy supplied won’t be known until ex post, while in the case of obligations or green certificates it will be exactly known how much renewable energy will be added or pollution abated.  2.3  Literature Review of R&D Policy  The literature on innovation relates to two main areas of renewable energy. The broadest area involves public good aspects of research and development (R&D) and the spillovers resulting from R&D. Economists have long argued for R&D tax subsidies or tax credits in order to achieve a more socially optimal level of R&D. The second area is the literature on innovation and the environment, which falls under the broader umbrella of induced innovation. Policies are implemented leading to a change in relative prices. This change in relative prices leads to the environmentally friendly technology becoming comparatively cheaper, resulting in innovation.  Increased research and innovation in alternative energy sources will lead to increased growth in alternative energy sources resulting in decreased use of traditional polluting sources of energy. This increased growth may come as a result of the creative destruction of innovations, a notion pioneered by Schumpeter (1942). The idea is that new innovations come along and make old technology obsolete. In the case of renewable energy innovation this could be a new way harnessing wind energy or turbine that produces more energy. Either will result in increased growth in output of renewable energy.  17  The idea of spillovers associated with R&D has been pointed out by many economists starting with Arrow (1962) and Nelson (1959). Nelson noted that private firms have little incentive to undertake basic research because some basic research is not patentable and because of the narrow technology base and therefore limited applications of technology. In many cases the significance of the basic research will not be realized for some time. With innovation being the result of R&D, policies aimed at increasing R&D will eventually increase innovation.  Theoretical research into R&D spillovers starts with Spencer (1984), who proposed a model with varying amounts of spillover, ranging from no spillovers in the case where patents are 100 percent effective to complete spillovers, where all firms share equally in the benefits flowing from the innovation. He finds that when R&D is costly and can lead to significant cost reduction, social welfare can be raised by providing firms with R&D subsidies. If firms know spillovers will occur there is less incentive for them to invest in R&D and as the number of firms increases the spillovers will increase, thereby decreasing an individual firm’s incentive to innovate. Spencer suggests that the most direct way to deal with low incentives to innovate is to subsidize research.  Grilches (1992) reviews empirical evidence showing that private rates of return on R&D have been significantly less than the social rate of return. The studies reviewed by Grilches use essentially two techniques: the simple measurement of the improvement of an innovation, such as the yield of farm product; or the regression based estimates of spillover benefits to outside sources from R&D expenditures. Much of the recent work on spillovers has used patent citations. Patent data facilitates an examination of cross  18  citations to look at how patents in one sector are cited by patents in another sector. These results have lead many countries to implement market based solutions such as subsidizing R&D or giving tax credits designed to increase R&D to more socially efficient levels.  A large body of empirical evidence documents the effects of R&D policy on R&D expenditures. Hall and Van Reenen (2000) review previous studies of this market based approach. They review 12 US studies and 10 non-US studies, published between 1983 and 1998, looking at the effects of different tax treatments on R&D spending. A variety of methods from survey to econometric methods were used to determine the price elasticity of R&D and a cost-benefit ratio. The econometric methods were usually based on one of two methods. The first method is the quasi experimental approach regressing R&D investment on explanatory variables and a policy dummy, allowing one to examine how R&D changes as policies change. The second is very similar to the first except instead of a policy dummy there exists a user cost of R&D. For survey results, the researchers ask the senior managers of leading firms how things have changed. The estimated elasticities range from insignificant or negative to close to 2, with the majority being less than 1.  It is widely agreed that induced innovation was first described by Hicks (1966):  “The real reason for the predominance more of labor saving inventions is surely that which was hinted at in our discussion of substitution. A change in the relative prices of the factors of production is itself a spur to invention, and to invention of a particular kind directed to economising the use of factor which has become relatively expensive.” -  19  The extension of induced innovation to environmental policy is only a recent revival of what Hicks first pointed out. If the environment is included as an input to producing a good and that input changes value, there is incentive to innovate. Environmental policy stimulates innovation by creating these changes in input prices and as a result several empirical papers look at the effects of environmental policy on innovation.  Induced innovation as a result of energy price changes has been looked at by several authors. Newell, Jaffe and Stavins (1999) look at the effect energy prices have on room air conditioners, central air conditioners and gas water heaters.  Popp (2002) uses patents from the US patent and trademark office from 1970 to 1994 to look at the effect of energy prices on energy efficiency innovations. He limits his research to American inventors. He regresses the log of the ratio of the number of patents in a particular field and year over the total number patents filed that year against energy prices, stock of knowledge, federal energy R&D and other explanatory variables. Instrumental variables are used for federal energy R&D because of its correlation to energy prices. A knowledge variable is created using a second regression looking at the number of citations a patent receives. The results show the expected positive effect that energy prices have on innovation.  Popp (2006) used patent data on patents filed by the US, Germany and Japan in their home countries to look at innovation spillovers. He looks at whether latecomers of 2 regulation adopt innovation of early regulators or continue to innovate on NO and SO  their own. Although most of the scrubbers used to reduce NO and SO 2 are produced  20  domestically for the three countries, the author goes a step further, looking at whether the innovations from the country that first regulate the emissions influences the innovations from the countries which regulate at a later date. This is done using citations counts, specifically by looking at whether innovations resulting from early regulation get more citations. He finds that foreign innovations are not being directly adopted and therefore domestic R&D will continue to play a substantial role in innovation.  The work most closely related to the model developed in this thesis is Johnston et. al (2008) who were among the first to look at policy changes over countries in a panel data setting. They looked at the effect of policy changes within most industrialized nations on innovation using country patent counts. Using a fixed effect binomial maximum likelihood estimate to regressing patents per country per time period against explanatory variables such as public R&D expenditure, electricity price, electricity consumption and policy dummies, they look at the effect policy changes have on the number of patents. Their results vary significantly from policy to policy and from renewable energy source to renewable energy, with obligations and tradeable permits having the greatest positive effect on patenting.  In a race situation like a patent race where multiple firms compete to discover the same or very similar ideas, there may actually be too much R&D and firms may duplicate R&D especially if there is a limited amount of spillover. Loury (1979) and Lee and Wilde (1980), for example, create a theoretical model that produces several interesting results, one of which says that in the short-run with a fixed market structure each firm may be researching more than the socially optimal due to duplication.  21  Dasgupta and Stiglitz (1981) developed a theoretical model that included another important factor, the size of the innovation. They use their model to examine oligopoly  and monopoly market structures and the frequency and size of market innovations. In the oligopoly case, they determine that there may be more innovation than is socially optimal. In the monopoly case there may be not enough innovation and the innovation that does occur is larger than that which is socially optimal.  Nordhaus (1972) looked at the optimal length of patents. He suggests that different patent lengths should be designed for different types of patents in order to maximize social welfare. He suggested that society may want to error on the side of a longer patent length rather than a patent length that is too short.  Along with subsidies, other approaches such as public R&D, prize competitions or public private partnerships have been examined for more socially optimal results. But determining how these policies address inefficiently low levels of research is complicated, with no clear cut best response.  22  3  THEORETICAL MODEL This theoretical chapter section will be divided into four sections. Section 3.1 will  describe the basic model including variables, structure and timing of the relevant events. Section 3.2 will focus on solving the second stage of the model, which involves a Coumot-Nash game between the renewable energy firm and the traditional energy firm. Section 3.3 will examine stage one of the game, where two renewable energy firms from different countries compete for a patent which allows them to be the sole provider of renewable energy in stage two. The final section 3.4 presents comparative statics results on how different government policies or changes in demand for energy will increase or decrease a firm’s propensity to patent.  3.1  Introduction  The main way to protect a new innovation is through the filing of a patent. A patent allows filers protection against other individuals or companies using their idea. For simplicity we will assume that patents give permanent and total protection to the filer of the patent. In reality this is typically not true as patents can quite often be innovated around.  Different technologies typically have different marginal costs (MC) of producing a MW/h of energy. Innovation lowers a firm’s MC and normally dominates the old technology. A pair of competing innovating firms are assumed to take part in a two stage game. In stage one, the two firms from two different countries (countries m and n) compete in an R&D race knowing the type of intervention that the governments of m and  23  n have chosen. In the final stage, the winner of the patent race, which has the lowest marginal cost amongst all renewable energy sources and therefore completely dominates the market for renewable energy, takes its new technology and competes in a Cournot competition with a traditional supplier of electricity (coal, natural gas, hydro).  The winner has two markets which it has an opportunity to service: a domestic market and an international market. International markets and domestic markets have different levels of accessibility. Accessibility is the ease at which a firm can enter a particular market. It is reasonable to assume that the cost of entering an international market exceeds the cost of entering a domestic market. For this analysis, accessibility is exogenously determined and is a result of government policies and regulations. The traditional (fossil fuel) supplier is expected to have a lower cost than the renewable energy supplier and therefore will supply most of the market, much like what is currently observed.  The renewable energy firm supplies q MW/h of electricity to the market in country i while the traditional energy firm supplies q MW/h of electricity to the market in country i, where i is either country m or n. Since a MW/h of electricity from a renewable source is equivalent to a MW/h of traditional source in the minds of consumers, the two goods are homogenous. We will assume that both traditional and renewable sources of energy have constant marginal cost curves cr and ct; however the cost of renewable energy will be higher in the international market because of the added accessibility cost 8.  24  The demand for MW/h of power is determined by the utility companies. For simplicity we will assume a linear demand curve, where P(Q) is the price per MW/h of electricity:  (3.1)  P’(Q)=a’—/3’Q  Q=q-i-q  i—m,n  Assume that a firm’s ability to service one market does not affect its ability to service another market. There are several policy interventions that government can use, one of which is guaranteed prices or feed-in-tariffs which usually involve the government guaranteeing a price above the current market value. Therefore renewable energy sources become more cost competitive.  To establish a relationship between the prices of renewable and traditional energy sources, assume that:  (3.2)  1.’ (Q) = sP’ (Q) If s>l then the government is using feed-in-tariffs or price guarantees and the price  received by the renewable energy source is greater than the price received by the traditional energy source. If s= 1 then there is no government intervention. Alternatively s  =  (1+ j) wherej is the percentage increase in price of renewable energy sources over  traditional energy sources.  25  3.2  Stage Two Cournot Competition  There are two subgames that must be solved in stage two, depending on whether it is the domestic finn or the international firm which innovates. If a firm from country m wins the patent race, it will compete in its domestic market m and an international market n. Whereas if a firm from country n wins it is a mirror case with its domestic market being n and international market being m. Therefore only one of the cases will be solved.  Assumption: Firm in countly m wins and competes with traditional energy source.  In this case the domestic firm competes in its domestic market and receives profits  and it also competes in country n’s market and receives a profit pr,”. Therefore the profit for the firm from country m should it innovate is:  (3.3)  2rm =  + 2Z  Within country m the firms supplying the traditional and renewable sources of energy solve the maximization problems:  (3.4a) max(q,q) (3.4b) max(q,q) where: (3.5a)  (3.5b)  ifl  =I ( m Q)q —cq = prn (Q)q cq —  sm(am =  (am  —  _fim(qrn rn  +q)) —cq (q + q” )‘‘ c7q —  26  Recall from equations (3.1) and (3.2) that P’(Q)  =  &  —  /3Q and P’(Q) = sF (Q). 1  The domestic market and international market can be solved separately due to their constant marginal cost structure. If the renewable energy firm had increasing marginal costs then the maximization of profits from the domestic market and international market would have to be solved jointly.  In the international market, it is assumed that the marginal cost for the firm from country m is equal to c 8  = Cr  +6, where ô is an international accessibility parameter. If  0 there is no additional cost of supplying the international market relative to the  domestic market. If 8> 0 there is an additional cost of doing business in the international market as a result of such things as increased transportation costs or non tariff barriers such as a complex regulatory environment. The renewable energy and traditional energy firms solve the maximization problems:  (3.6a) m(q,q) q (3.6b) max(q,q) 1 q where: (3.7a) r’ =F(Q)q —cq =s(a’ —fi”(q +q’))J +S)q = (3.7b) r’ P (Q)q cq,” = (a” p” (q + q cq” —  —  —  )“  —  Equilibrium quantities: Solving the first order conditions for the various profit maximization problems, we get the reaction functions for the renewable energy firm and the traditional firm in the domestic market of country m:  27  am -f m q 3m  (3.8a) q =  2f3”  =  prnn  (3.8b) q, m  a  -  C  In firm m’s international market the reaction functions are:  a _fiflq; (3.9a) q =  -i-S  2/J’  =  (3.9b) q;’  _Cr  a  6q,. —c’ 1  The reaction function allows one to determine the optimal production response of MW/h of electricity supplied by one firm in response to the other firm’s supply. In solving the reaction functions we are looking for the Cournot-Nash equilibrium quantity on MW/h produced by each firm. The Cournot Nash equilibrium production by the renewable energy and traditional firms in the domestic market is given by:  2c a m +c m ..z (3.1Oa)q*= tm a (3.lOb)q”*=  The corresponding quantities in the international market are:  28  a”  2(Cr +c’  +8)  —______  Sn  (3.lla) ql’* =  3,8” a”  +  (Cr  +8) S  (3.llb)q’*=  3,8”  Substituting equations (3.lOa)  and (3.lOb)  into  equation  (3.5a)  and equations  (3.1 la) and (3.1 ib) into equation (3.7a) we can determine the equilibrium level of profit from the renewable energy source in each market. In the domestic market,  lam *  (3.12)  =  lam  m +c  lam  5  I  s[  +Cr [[Smam  ++c  I  3fi  3  N1’am  +SmCfl  —  Cr  +m  I  I  3fi  J  m +C  m  am +C 1 m ——--fl  Is  3fi  3  ‘  j s”  firm m earns:  3 J  In the international market,  firm m  earns:  (3.13)  +  Cr + Sn  * = S  + ce”  1 a ”  I  +c  —  2(Cr + 8”)  (C + 2  Sn  (Cr  + 8)  JL  J  + +(c  +8)+s”C J_(Cr+8)  2(Cr  a”  +8  (  IsnIa+t  3flfl  ,8n  [[snan  2(Cr  +ô)  J  3  J  29  Therefore the combined equilibrium profit for the firm from country m is:  (3.14) r  +  =  Policy impacts: The issue of interest is the differential impacts of a domestic versus an  international feed-in tariff policy on the profits of firm m. The impact of the policy on profits is important because, as shown below, the sie of the profits earned in stage two determines a finn’s incentives to innovate in stage Ône.  The following two results can be established.  An increase in domestic feed-in-tariff or price guarantee increases profits n.  (3.15)  ô2tm*1  m a  2c tm s  2  2c  am+cm_! t  tm fi  m  ‘  tm fJ  3  I  3  2cr  1>o  An increase in international feed-in-tariff or price guarantee increases profits r.  (3.16)  m1  as”  m jj  2(c +  a +c  5fl)  2  a  2(cr  + —  + 8’s) “ ,n  —  /3”  3  As we can see from equations (3.1 Oa) and (3.11 a), an increase in feed-in tariff or price guarantee shifts the reaction curve of the renewable energy source outwards leading  30  to an increased market share for the renewable energy firm. This increased market share results in increased profit for the renewable firm.  Another issue is what happens to the price faced by the traditional energy provider and the renewable energy provider when a feed-in tariff or price guarantee is implemented. Using equation (3.1) the equilibrium price faced by traditional energy firms is:  pm*  _fim(q* qm*)  m =a  In the domestic market and  P, =a’  _flfl(q;* q*)  in the international market.  If there is an increase in the domestic feed-in tariff or a price guarantee there is a decrease in traditional energy prices  ôpm* = _m  m a  (oqm*  + oq*  tm ôs  Pm*.  c  =  )  rn  L  2 3pm(srn)  +  2Cr ) 2 313m(Sm)  <  o  An increase in the international feed-in tariff or price guarantee results in a decrease in traditional energy prices P ” in the international market. 1  31  =  (nfl  +  6?  1 ãs’  ,)  tm p  (L  Cr  +8  2 3/3m(sm)  + 2(Cr + 8)  ) o 2 313m(sm)  Therefore an increase in price guarantee or feed-in tariff decreases the price that traditional energy sources receive. This would likely be small because the share of the energy market that renewable energy sources hold is very small compared to traditional energy sources. The renewable energy firm has two forces affecting it, the decrease in the price of the traditional source which it is tied to and the increase in price as a result of the feed-in tariff or price guarantee. Using equation (3.2) the equilibrium price faced by the renewable energy firm is:  pm*(Q) = smPtm*(Q)  In the domestic market and  P*(Q) = s7iP*(Q)  in the international market.  If there in an increase in domestic feed-in tariff or price guarantee there is a an increase in the renewable energy price  =  ôpm*  m a  +s’ t P m ‘ tm as  tm a  1 ++C m’  Cr  3  32  An increase in international feed-in tariff or price guarantee results in an increase in the renewable energy price P, in the international market.  =  a  +  c+8  + Sn  =  ôs”  ‘  +c 1  + —  as”  3  3(s”))  Note that an increase in the feed-in tariff decreases the price received by the traditional energy source but this is counteracted by the feed-in tariff, resulting in an increase in the price renewable energy.  3.3  Stage One R&D Race  For stage one the two renewable firms from two separate countries (m and n) compete to try and discover the new technology. This technology takes the form of a drastic innovation that obsoletes other innovations and allows the innovator to be the sole supplier of renewable energy to the energy markets. Assume Firm i’s probability of discovering the new technology P(x) is determined only by x ,, which is the amount of 1 money spent on research. Specifically, assume that:  O<P(x<1 = 1/k> 0 P”(x) = 0  i=m,n  This assumption implies an increase in x increases the probability of a successful discovery at a constant rate 1/k. Finally if both firms discover the new technology then  33  the winner will be determined by a coin flip. Consequently, the probability of Firm i winning is: (3.17) Pr(flrm i is the winner) = Pr(flrm i discovers)Pr(flrm j doesn’t discover) + 0.5 Pr(flrm i discovers)Pr(flrm j discovers) In the coin flip situation there is a negative effect associated with both firms innovating at the same time resulting in one firm’s research investment being completely wasted. In reality their effort may not be completely wasted as they may have the opportunity to file a patent in a smaller less lucrative market which would allow them to capture some of their lost research investment.  1-f,.  Both firms have the same quadratic R&D cost function is a government subsidy of research. If  =  —-  (x.  2,  where 0  <  1  0 then there is no government research  subsidy. There may be a desire to subsidize research on the part of the government due to the traditional spillovers associated with imperfect property rights or in a desire to increase the speed in innovation.  Recall from equation (3.14) that r is the stage two profit that the winner will receive with the patent for supplying renewable energy to the domestic energy markets and the international market. Consequently, expected stage one profits for the pair of firms can be expressed as:  (3.18a) E(profit ) m  =  [P(x  (3.18b) E(profit )  =  [P(x )(1  m  )(1  —  —  P(x  n  )) + O.5P(x m )P(x n )],r m  P(Xm)) + O.5P(x )P(Xm )J  *  *  —  2  m (x )2  — —-  m  (x)  2  34  By assumption, the loser of the patent race will receive zero profits.  The two firms choose the amount of money to dedicate to R&D simultaneously. Therefore we can determine a Nash equilibrium level of expenditure on research. Solving the maximization profit we get the first order conditions, where ,r, is the exogenous profit earned in stage three for being the sole supplier of MW/h of renewable energy:  (3.19a) P’(Xm )(1 (3.19b) P’ (x )(1  —  —  O.5P(x ))r. —(1— m )Xm = 0 0 .5P(Xm ))r, —(1 =0 —  For simplicity the left hand side of the first order equations (3.1 9a) and (3.1 9b) will be described as a function of their variables. This allows for easier notation when taking the total derivatives.  (3.20a) P’(Xm)(1—0.5P(Xn))2r: (1(Om)Xm  (3.20b)P’(xn)(1—o.5P(xm))2r: —(l—q)x  fm@m,Xn,2,,tPm) = fn(xm,n,2mn*,c0n) =  0  0  Taking the total differential of equations (3.20a) and (3.20b) with respect to Xm,  X  and Sm and dividing by dSm we get:  (3.21a)  + km  -+ 4 (3.21b)dXm m  1 dSm dx, dx dSm  =0  +  d7Zm +  dç dSm  0  Solving the system of equations the resulting equations are:  35  (322a) ! 4 _ •  dltm dSm dx  dSm dXm dx,  (3 22b) •  dx diç dSm  dx  —  dx dXm  d,r dSm thm  thCm d2rm dSm  dSm  L  m dx  (‘Xn dXm  where:  (3.23) (3.24) (3.25)  (3.26)  = —O.5P’(Xm)P’(Xn)2rm =  <0  2 = = 0.5P’(Xm )P’(x )r,  —0.5- r, <0  dXm  (3.27)  = P’(Xm)(l  —  0.5P(x))  =  d2rm  (3.28)  ---  dr  = P’(x)(1  —  ---(1 k  0•5P(Xm)) = (1  k  —  —  )) 11 0.5P(x  >  0  0•5P(Xm)) > 0  Substituting equations (3.23) to (3.28) into (3.22a) and (3.22b) the result is:  (3.29a) 4 -= dSm  k  dSm  (‘Pm  —  1)(çz —1)— (—0.5  k  k  dSm  -r)(—0.5  36  (3.29b)-= dSm  —(l—0.5P(Xm)) ( 4 cOm k dSm  ((Om  —  k  k  1)(q —1) —(—0.5  dSm  )(—0.5  r)  Result 1: Ifboth countries are completely accessible (c môn=O) and have equal 5 research subsidies then an introduction ofafeed-in-tarffby one country will result in an equal increase in both firms’ equilibrium level ofresearch.  Proof Suppose a feed-in-tariff is introduced by country m.  From equations (3.12) and (3.13) it can be seen that if c5  and from (3.15) and (3.16) ifc5  =  = d,r  0 then d5m  subsidies are equal then q = X  =  Xm =X.  d5m  =  .  =  d5m  d5m  =  0 then r  r, =  >o• Ifbothresearch  Since equations (3.20a) and (3.20b) are symmetric  Substituting equations into (3.29a) and (3.29b) we get:  0 2 (ç—1)  (1—0.5P(x))—(—(ço—1)—0.5---r) d5m  2 (ç—1)  37  Therefore if  2 > (O5±)2 (q—1)  k  =  then dSm  0  Q.E.D  dSm  With two completely accessible economies a feed-in-tariff will result in both firms increasing their research equal amounts. Since both firms can capture all of the potential profits from both markets, if the potential profits increase in one market they should increase their research to attempt to capture the increased profits.  Taking the total differential of equations (3.20a) and (3.20b) with respect to x, x,  and corn and dividing by dom:  (3.30a)  +  dXm dcom  dx dm  (3.3Ob)4 -+ dx dtbm dXm dm  =0  +  dm 0  Solving the system of equations:  dfmdfn dm dXn -— 4 (331a) dmddJdf; dXmdXn  ‘nm dfndfm dXm dbrn (331b)--— d6mdfmdfn dfmdfn 1 dx dXm dXm dx,  38  (3.32) d(Om  Result 2: Ifone country ‘s government implements a research subsidy then the firm from that country will increase its equilibrium level ofmoney spent on research while the firmfrom the other country will decrease its equilibrium amount ofmoney spent on research.  Proof Suppose  q.’  decreases as a result of a research subsidy.  Substituting in equations (3.23) to (3.26) and (3.32) into (3.31a) and (3.31b).  (Xm)Q,Pnl)  dXm dbm  (corn  —  l)(ço —1)— (—0.5-r )(O•5_2Zm)  (Xm)(—O.5-2rn) døm —  (‘Pm —l)(ç,  —l>(0.5---2r) then 1 Thereforeif  0  ---<0  Q.E.D  An increase in research subsidy will increase the equilibrium level of money spent on research by the firm in the country which implemented the research subsidy and decrease the equilibrium level on money spent by the firm whose country didn’t  39  implement the research subsidy. The research subsidy lowers the cost of research to the domestic firm and therefore the domestic firm will do more research. The international firm’s best response to the domestic firm’s increase in research without any additional profits from stage two is to decrease their research. This is a unique result to this model which results in the domestic firm acting like a first mover and the international firm as a follower.  The following results were developed using artificial data to attempt to determine the effects on research in asymmetric situations. All solutions must satisf’ the expected profit first order conditions equations (3.1 9a) and (3.1 9b). Several different parameter values were used in order to try and determine sign and relative size of several derivatives in an asymmetric environment (See appendix).  Result 3: Ifa country is not perfectly accessible, afeed-in-tarffwill increase both  domestic and international research; however, the increase to domesticfirms will be greater.  Proof From table 1 in the appendix it is clear that in all simulated situations  dx dx dx dx —-,—-->O and —-——----> 0. dSm dSm  dSm  dSm  If a country is not completely accessible, policies which should affect all firms equally result in an unequal level of research.  40  4  EMPIRICAL This empirical Chapter is divided into four subsections. Section 4.1 describes the  basic empirical model. Sections 4.2-4.4 describe the data which will be used. Section 4.2 is a description of the patent data, followed by section 4.3 which describes the policy data and section 4.4 which provides a description of the other variables.  4.1  Empirical Model  This section looks at the effects that domestic output and world output have on innovation. To do this the effect that domestic output of wind energy (GWh) has on patents is compared to the world wind energy output (GWh). If there is a home bias domestic output (GWh) should have a greater effect on patenting than world output (GWh). Following from the theoretical section, one possible explanation for a home bias would be that policies implemented to encourage the increased output of wind energy encourage domestic firms to innovate more than international firms. However, the policy effects on innovation may not be the only explanation captured by this regression. Alternatively there may be a government bias in the procurement process towards domestic firms resulting in the growth of domestic firms. This would result in large wind producing countries having large domestic firms which would have incentive to invest in R&D and innovate. This bias and its effects on domestic growth of firms has been pointed many (Branco, 1994; Miyagiwa, 1991) and may still be prevalent throughout Europe (Nielsen & Hansen, 2001).  41  The literature on panel patent count data started from Hausman et a! (1984) which looked at the effects R&D expenditures had on patent counts of firms and was the first to develop random and fixed effects models for Poisson and negative binomial distributions. This was followed up by Cameron and Trivedi (1998) who pulled together all the applicable research on count data into one source. To test if there is a “home bias”, a simple model is constructed:  (4.1)  =  DS, 1 exp(/3  Where  +  WS 2 f3  + u.)  is the exponential mean function and is a function of the domestic size  of market (DS ,) in country i in time t, world size of market (WS) and a fixed effect or 1 random effect (uj). The domestic size and world size will be measured in tenns of thousands of gigawatt hours of electricity (GWh). For simplicity it will now be written in matrix form:  (4.2)  & =  =  )1 1 exp(u DS exp(fl =  + /2 WS)  =  1 exp(fi v DS, 1  ) 1 + /32 WS  , x. I 1 E[pat  In order to test for “home bias” both Poisson and negative binomial distributions will be tested due to the significant amount of zeros. If there is a “home bias” we would expect to see either the coefficient on domestic size being larger than that of world size if both are significant or the coefficient on domestic size being positive and the coefficient on world size being insignificant. The major difference between the Poisson and the negative binomial model is that the Poisson model forces the mean and variance to be 42  equal not allowing for overdispersion of the data. Overdispersion is often seen in cases where there are a lot of zero counts resulting in a variance larger than the mean. The negative binomial distribution allows for overdispersion. Fixed effects and random effects models will be tested for each distribution using the random and fixed effects developed by Cameron and Trivedi (1998).  Fixed Effects Models: Both a negative binomial and Poisson fixed effects models will be used to look for “home bias”. The fixed effect model allows one to attempt to capture unobserved heterogeneity. In the fixed effects model u, is allowed to be correlated with x’s, and is assumed to be fixed. The fixed effects model results in different intercepts for each country. The coefficients are estimated using a conditional maximum likelihood estimates. For the Poisson distribution it is assumed pats, are independently identically distributed with Poisson distribution and mean conditional joint density for the  ,. 1 Prpat,  . .  ,  path.  I  the  observation is:  T (4.3)  H 2  (pat.)! , 1 pat  =  (  2.  pat  H pat::!  For the negative binomial model it is assumed that the assume  pat,,  is  independently identically distributed with a negative binomial distribution and mean and variance 2,, (v,  + v),  the conditional joint density for the ith country is:  43  (4.4)  ( Prpat 11  1 pat 1 + i) 1 2, )r( r’(  T ,. • .,  pat  I  11 pat  j  +  =  path)  + pat ,) 1  , ipat 1 r( 11 + i)  Random Effects Models: In the random effects model, v. is a country specific random variable. With the random effects model, u. must be independent of x’ . For the 1 Poisson model, v, is assumed to be distributed as a gamma random variable with parameters  (o, 8) normalized so that E[v]  =  1 and Var[v ] = 1 / 8. This leads to the joint 1  density for country i of:  (4.5) ,...,pat I x Pr(pat 1 ,...x,)= 1 1  T pat 11 +8)  ]S[ pat.t 61  :‘;  For the negative binomial model,  1 is assumed to distributed as beta (a,b) 1+ V 1  random variable. This leads to the joint density for country i of:  (4.6) P rat,i,.  . .,  1 I1 pat x , , .  . .  —  (  , + pat,, 1 r ) 1  ra + b)11a + ,1 1 F(2 )Fat + i),J r’(a)r’(b)r’(a + b + ,  +  ) 11 pat  +  ) 1 pat  Equations (4.2) to (4.4) are maximized to obtain the coefficient estimates.  Another interesting question that arises is what determines the domestic output (GWh) of a market? Some small countries have a disproportionately large domestic  44  output while some large countries have a disproportionately small domestic market. What will encourage the growth in renewable energy? There are several factors that could determine the output of a domestic market. One of the major factors that will determine the domestic supply of output will be the policies within that country. Ideally one would like to look at the effect that policies have on the supply of renewable wind energy. However, the supply of wind power (GWh) has a random component to it in that the supply of wind is determined in part by the weather. Therefore instead of using the supply of wind power the change in capacity over the year is used. Change in capacity is nonrandom and policies should have the same effect on the change in capacity as the change in supply.  The policy data available through OECD/IEA (OECD/IEA, 2004) poses a major problem in that it doesn’t provide a measure of the size of the policy. This could lead to misleading or incorrect results because it treats all policies as being of equal size. Another major issue is that in many cases multiple policies are implemented at the same time resulting in multicollinearity between policies. This will result in large standard errors between policies most likely resulting in many policies being insignificant in the results. The policies are broken up into seven subfields: tax incentives, feed-in-tariffs, green certificates, producer grants, consumer/individual grants, obligations and targets. From the theoretical literature all policies, except targets which tend be non-binding, should result in positive changes in supply of wind energy. This should result in their coefficients being positive implying that these policies increase the change in wind capacity. We would predict the expected price of oil to have an impact on the growth of renewable energy sources. As the expected price of oil rises, there should be more 45  incentive to invest in green energy as there is an expected relative drop in price. If this is true there should be a positive coefficient on the price of oil. For the expected price of oil, a simple change in oil price from the current year to the previous year is used. This is used as an approximation to what the expected change in oil prices will be in the next year.  Other factors that could affect the size of the domestic green energy market would be the size of the economy (GDP) and the size per capita (GDP per person). In general one would expect a large economy with a large demand for electricity would have an increased use of all forms of energy including green energy sources and traditional energy sources resulting in a positive coefficient on GDP. It has also been speculated amongst economists that as individuals’ wealth rises they tend to put more value on a clean environment. Since green energy isn’t polluting it would seem likely that as per capita income increases so too should the demand for renewable energy. This would be seen in a positive coefficient on GDP per capita. Due to multicolinearity with GDP an average of the past three years is used for GDP per capita. This eliminates the multicolinearity issue which arises from using GDP and GDP per capita from the same year and smoothes GDP per capita from business cycles. With this in mind a second model is created to look at the determination of domestic markets, Where Icapj,t is the change in capacity.  46  (4.7)  , 1 Acap  =  pol, 1 fi  +  111 fl in gdp 2  + /34 (In 1 ,,_ oil  —  gdpcap, +  in 2 ,_ + a, 1 oil )  +  5 c  +  +  113 gdpcap  +  14 gdpcap  11 e  The model includes either fixed or random country effects a 1 and time effects 6. If a. is a fixed effect, it is assumed that it is nonrandom and can be estimated using dummy variables. If a. is random it is assumed that it comes from a normal distribution which must be accounted for in the error covariance matrix. If a 1 is considered to be a random effect, two additional variables are included: a coastline dummy and an area variable. One might expect a country with a larger area to have a large change in wind capacity due to the potential of better locations for wind turbines. Additionally with coastline providing some of the best wind potential, countries with coastline may have larger changes in wind capacity. These are not included in the fixed effects model because they  are constant and therefore captured by the fixed effect. Time fixed effects are used to capture effects that are different across time periods but not countries.  4.2  Patent Data  Country count data on patents filed with the European Patent office seeking protection within the European Region will be used to measure the effects of policy changes on innovation. The European Patent Office maintains a searchable patent database (European Patent Office, 2008) of European, worldwide and US patents from 1978 to the present. It allows a search of the database using 10 fields such as keyword(s)  47  in title or abstract, application number, publication number, priority number, inventor, applicant, European classification (ECLA) and International Patent Classification (IPC).  There are several issues associated with using patent data as a way of measuring innovation. One is that all patents are equal regardless of how minor or major the innovation is. One way to alleviate this effect is to only look at patents filed to the European Patent Office rather than individual country patent offices. Since all patent applications require expenditures of time and money one could assume that only the truly worthwhile inventions would take the time and the money to file with the broader European Patent Office. Another major issue is determining the date of the invention, which will be crucial to determine the effect of policy changes. Although this information isn’t expressly given, the priority date does allow one to know the first date that the patent was filed with any patent office and thus is the closest date to the invention date. One would also assume that a patent would be filed soon after invention to allow for protection from competitors. This is more useful than the publication number or application number because many inventors will first file in their home office or the office of one of the larger European countries (Great Britain, Germany or France) before filing for a European patent.  Patents related to the use in producing wind energy can be determined using International Patent Classification (IPC) codes (World International Patent Organization, 2008). The hierarchal coding system allows the isolation of relevant classes or subclasses suitable for each form of alternative energy. For wind power the relevant classification are displayed in table 4.1.  48  Table 4.1: IPC codes for Wind Wind Wind motors with rotation axis substantially in wind direction Wind motors with rotation axis substantially at right angle to wind direction Controlling wind motors Adaptations of wind motors for special use; Combinations of wind motors with apparatus driven thereby Details, component parts, or accessories not provided for in, or of interest apart from, the other groups of this subclass ie Transmission of power, e.g. using hollow exhausting blades or Mounting structures  Class FO3D FO3D FO3D  Sub-Classes 1/00 3/00 7/00  FO3D  9/00  FO3D  11/00  Once the categories are determined, the database is searched using the IPC classification and the results searched to determine that the patent is relevant. A patent result is relevant if its intended use is in producing wind energy. The description section of each patent allows the innovator to discuss the intended use of their innovation. This minimizes the risk of counting patents which should not be included. The risk of not counting a patent is small because most patents are filed under numerous IPC codes.  Patent counts were determined for all major patenting countries along with patent priority dates and country of inventor. A panel data set is constructed for 1990 to 2006. From the twenty countries of the European Union, sixteen were found to have patented during that period of time. As one can see from figure 4.1 each country has significantly different rate of patenting and there has been a substantial increase in overall patenting in the last several years.  49  Figure 4.1: Patents from 1990-2006 for Germany, Denmark, Great Britain and Austria in Wind Energy 140  C  1.......  120  100  Germany Denmark Great Britain Austria AllEurope  80  ! I  a. 60  1990  1991  1992  1993  1994  1995  ii  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  Year  The drop off in patenting in the final years is a result of patent counts being based on priority date, the first date that a patent is filed at any patent office. As a result there will be many patents that have a priority date from 2005-2006 that have not yet been filed with the European patent office or have not yet made it through the European patent office system.  4.3  Policy Variables  Policy variables are determined using the International Energy Agency Global Renewable Energy Policies and Measures Database (OECD/IEA, 2008) and their  50  published policy reviews (OECD/IEA, 2004). The database allows the searching of policies based on type of policy, country and policy target. Policy target is the type of renewable energy that the policy is supposed to target since different policies have been used to target specific renewable energy sources. Policy type will allow the grouping of policies into the following categories: grants to producers, grants to companies/individuals to invest in wind energy, tradeable green certificates, tax incentives, feed-in-tariffs, obligations (producers are required to use dedicated amounts of certain renewables) and targets. For each policy type, dummy variables are created. The policy dummy is 1 if it is implemented in a given country and remains a 1 until it is phased out. The policy dummy will only be connected with the renewable energy source that they are supposed to be targeting, since usually policies are designed to target only certain renewable energy sources. The lEA policy database starts with policies implemented as early as 1974 and has been updated up to July 2008. It would be desirable to know the strength of these policies, such as the size of the feed-in-tariffbut this information is not readily available.  4.4  Other Variables  Wind capacity and supply is available from 1990-2006 through the lEA database (International Energy Agency, 2009). Like the trend in patenting there has been a steady increase in the use of wind energy. Some countries like Germany have seen a significant growth in output from wind energy over the 16 year period from 1990 to 2006 (see Figure 4.2).  51  Figure 4.2: Electricity Output (GWh) for Germany, Denmark, Great Britain, Austria and OECD 140000  120000  100000  C,  80000  Germany Denmark U Great Britain Austria 0ECD  0 60000  40000  20000  0  I,1990  i.ijjJJJJLti  1991  1992 1993 1994  1995  1996  1997  1998  1999 2000 2001  2002 2003 2004 2005 2006  Years  The lEA database also provides the cost of crude oil imports to Europe. GDP and GDP per capita are available from the International Monetary Fund (2009). Coast dummies and land area are available through the Central Intelligence Agency (2008). For complete descriptive statistics see tables 8.2 and 8.3.  52  5  RESULTS The results chapter is broken up into two sections. Section 5.1 examines the results  of the first regression to determine whether there is a “home bias” and section 5.2 examines the second regression, which looks at what determines the size of a market.  5.1  Patenting  —  Regression One  The results to both the Poisson and negative binomial regressions are supplied in table 5.1. In both cases domestic output and total output have a positive impact on the amount of patenting done by a given country. A dummy variable is included for 2006 to capture the decrease in patents due to the priority lag. In all but one case the coefficient on domestic output is greater than the coefficient on total OECD output.  53  Table 5.1: Results of Patent Regression Fixed Effects Poisson Negative Binomial Output in thousands of GWh Domestic Output World Output  0.025 (3.87)*** 0.01 (7.63)***  Dummy2006 Constant Observations Number of Countries  Output in thousands of GWh Domestic Output World Output Dummy2006 Constant  0.015 (2.28)** 0.019 (11.62)*** -1.366 (9.74)***  0.023 (1.69)* 0.012 (5.90)***  0.025 (1.96)* 0.019 (8.18)*** -1.372 (5.00)*** -0.185 (0.89) 272 16  -0.265 (1.41) 272 272 272 16 16 16 Random Effects Poisson Negative Binomial 0.026 (4.18)*** 0.01 (7.64)***  0.018 (2.66)*** 0.018 (11 .64)*** -1.368 (9.78)***  0.027 (2.14)** 0.011 (5.69)***  -0.25 5 (1.37) 289 16  0.030 (2.41)** 0.018 (8.02)*** -1.353 (5.03)*** -0.181 (0.88) 289 16  Observations 289 289 Number of Countries 16 16 Absolute value of z-statistics in parentheses * significant at 10% level;** significant at 5% level; *** significant at 1% level  When looking at the two results one must also remember the major difference in the Poisson and negative binomial models. The Poisson model assumes that the mean is equal to the variance. This may not be true when it comes to patents which tend to display overdispersion. Overdispersion is caused by a large number of zero entries resulting in the variance being larger than the mean. In all regressions domestic size of market and world size of market are significant at 1%, 5% or 10% significance level and jointly significant. An F-test was run to check whether the coefficients for domestic size and world size were significantly different. In all the Poisson models the null hypothesis  54  that the coefficients on domestic size and world size were equal could not be rejected at 5%. Comparatively in all the negative binomial models the null hypothesis that the coefficients on domestic size and world size were equal could always be rejected in favor of the alternative that they are statistically different.  The Haussman test is run with the negative binomial fixed and random effect models with 2006 dummies to attempt to determine which model specification is most appropriate. The null hypothesis that the coefficients estimated by the efficient random effects estimator are the same as the ones estimated by the consistent fixed effects estimator are rejected (p-value 0.0006); therefore the fixed effects model should be used.  In order to interpret the coefficients the conditional mean is differentiated with respect to domestic market size (DS,) and world market size (WS) to look at the effects of marginal changes on the conditional mean.  , Ix 1 ÔE[pat  , 1 ows 1 Ix ãE[pat 11 I 11 8DS  2  —  =  1  2i,t  In all but one case a change in the size of the domestic market has a larger effect on the expected patent count than a change in the size of the international market. In some cases it has almost double the effect; however in only the negative binomial specifications are the coefficients significantly different. Another significant issue is that  55  the maximum likelihood negative binomial estimators are inconsistent under misspecification.  5.2  Market Size  —  Regression Two  The results of the market size regression can be seen in tables 5.2 and 5.3. The results show inconsistencies. GDP and GDP per capita which one would assume to be both be positive and significant in fact change signs and in some cases are significant and in other cases are not significant. The results as presented in tables 5.2 and 5.3 are also highly influenced by outliers. If Spain and Germany are excluded from the fixed effects regressions, GDP and GDP per capita become insignificant in regression 1 and in regression 2 GDP per capita remains significant.  56  Table 5.2: Fixed Effects Market Size Results Oil ((in oil,,_ 1  —  in oil )) 2 _ , 1  GDP per capita 1 gdpcap  +  113 gdpcap  (1)  (2)  Dropped  117.69 (1.46)  477.10 (2.14)**  375.92 (2.45)**  -379.40 (1.88)* -277.51 (3.62)***  33.86 (0.27) -188.79 (2.58)*** 24.40 (0.33) -81.18 (0.94) 578.90 (9.20)*** -45.96 (0.71) 16.46 (0.21) 51.62 (0.83) No -3825.24 (3.29)*** 272 17 0.34  + gdpcap 114  GDP (ln gdp, ) 11 Green Certificates Tax Incentives  -26.29 (0.34) -89.76 (1.04) 451.27  Targets Obligations Feed-in-tariffs Grants to Companies or Individuals Grants to Producers of Eiectricity Time Dummies Constant  -104.67 (1.58) -43.71 (0.55) 18.20 (0.29) Yes -2464.51 (1.48) 272 17 0.4  Observations Number of Countries R-squared Absolute value oft-statistics in parentheses * siificant at 10% ievei;** significant at 5% level;  ***  significant at 1% ievel  57  Table 5.3: Random Effects Market Size Results 1 Oil ((in oil_  —  GDP per capita 1 gdpcap  in oil_ )) 2  +  13 gdpcap  (1)  (2)  Dropped  115.75 (1.50)  106.58 (0.86)  216.61 (1.98)**  20.58 (0.27) -254.90 (3.58)*** -65.94 (0.96) -130.42 (1.66) 468.12 (7.12)*** -69.56 (1.14) 14.98 (0.21) 35.60 (0.62) Yes  52.68 (0.75) -155.19 (2.23)** -9.23 (0.13) -73.40 (0.92) 583.29 (9.66)*** -12.78 (0.21) 73.40 (1.00) 68.98 (1.16) No  46.59 (0.55) 96.65 (0.51) -1754.483 (1.08) 272 17  45.11 (0.55) -22.86 (0.12) -2896.45 (2.47)** 272 17  + gdpcap 114  L L GDP (lngdp ) _ , 1 Green Certificates Tax Incentives Targets Obligations Feed-in-Tariffs Grants to Companies and Individuals Grant to Producers Time Dummies ) 1 Area ln(area Coastline Dummy Constant  Observations Number of Countries Absolute value of z-statistics in parentheses * significant at 10% level; ** significant at 5% level;  ***  significant at 1% level  The only significant results that remain constant are that obligations result in an increase in the change in wind capacity and green certificates result in a decrease in the  58  change in wind capacity. It is not surprising that countries that set obligations to the amount of wind energy or renewable energy that must be used show increases in the change in wind capacity. This result holds even as different countries are removed from the regression to check if an individual country is driving the result. The result that green certificates are negative throughout all the regressions may be of some surprise; however if Italy is removed from the regression, green certificates are insignificant. Italy uses green certificates but has been very unsuccessful at encouraging growth in renewable energy sources. The lack of significance amongst other policies could be a reflection of how policies are introduced, with several policy types being introduced at the same time. The result is high correlation amongst policies resulting in large standard errors, which may cause none of the other policies to be significant.  The lack of significance of many of the economic variables that one may expect to be a factor in the increase in wind capacity, leads otie to look for other social, political or infrastructure constraints that may have an impact on the change in demand for wind energy. In the following section, several countries will be examined to determine the other factors that may have lead to increases in demand of wind energy.  59  6  BRIEF COUNTRY REVIEW If changes in wind power capacity can’t be explained by any of the typical  economic variables, what does determine changes in country’s wind power capacity? In the next sections we will look at a few countries which have been both leaders in the development of wind capacity and followers. In Europe the environment has become an issue when it comes to energy policy; however a country’s response varies significantly. In many cases a country’s response to the use of nuclear power plays a significant role in a country’s growth in renewables. Nuclear power is a cheap and reliable way of producing electricity without the emissions of greenhouse gases. Section 6.1 will look at Germany a world leader in wind capacity. Section 6.2 will look at France a country in many ways similar to Germany, but with little development of wind potential and section 6.3 will look at Denmark a relatively small country with large wind development.  6.1  Germany  Germany is the largest economy in Europe, the fifth largest consumer of the electricity in the world and second leading producer of wind energy in 2008 behind only the United States (NBC News, 2008). In 2006 5% of its total generation was from wind resulting in 30710 GWh of electricity (Energy Information Administration, 2009). Wind output has grown substantially from 71 GWh in 1990 to 30710 GWh in 2006. So what has prompted this rapid growth in wind energy? Most European countries started to take an increasingly serious look at their energy makeup following the oil crisis of the 80’s. This was followed in the early 90’s by a push for stronger environmental regulation. In the case of Germany there seems to be a political part to the story not captured in the  60  economic data. In 1998 the Social Democrats-Green coalition came into power with one of their prime objectives to phase out the use of nuclear power in Germany. Since 1997 the amount of nuclear power generated has, as a percentage of total generation, decreased from approximately 31% to approximately 27% in 2006 resulting in a 30000 GWh decrease in the use of nuclear. Germany has taken a very aggressive role in implementing and researching green energy sources which can be seen in the increase in share of all renewables in the past 15 years. The strength which the renewables are encouraged may not accurately be represented in the binary nature of the policy variables.  6.2  France  France is in many cases very similar to Germany. France and Germany are of similar size with France being slightly smaller in terms of GDP, but having a slightly larger per capita GDP. Both seem very similar politically with left leaning and right leaning political parties being in power over similar amounts of time; however the results have been very different when it comes to wind power. The French by comparison have been the largest European net exporter of electricity for the last 15 years (Energy Information Administration, 2009), the result of an overcapacity. Starting in the 1980’s the French diversified their energy portfolio by investing in nuclear power. By the 90’s 90% of their domestic electricity needs and 75% of there total generation were met by nuclear power resulting in substantial exporting of electricity. The reliance on nuclear has also resulted in a power grid designed for large consistent centralized power plants (Reiche & Bechberger, 2004), which may be poorly designed to handle the strains of wind energy. This combined with what seems to be a general acceptance of nuclear  61  power by the French people has resulted in the slow growth of wind power and has probably resulted in extremely modest targets for renewable energy sources such as wind power.  63  Denmark  Denmark is in a position quite different from France and Germany in that it is relatively small. Denmark consumed around 1% of Europe’s electricity in 2006 of which 18% of its electricity was produced from wind. In 1985 Denmark abandoned any idea of the use of nuclear power as part of their power portfolio (Climate Change and Power Economic Instrumentsfor the European Electricity, 2002). Denmark has taken an even stronger taxation approach than many European countries. As imported fuel prices decreased in the late 80’s taxation increased, resulting in a neutralization in the drop of energy prices. This coupled with the moratorium on nuclear power resulted in the perfect environment for increased use in wind power. This can be seen not only in the use of wind power and other renewables but also the total energy intensity (consumption per GDP) which has not increased since 1972 (Energy Information Administration, 2009). Denmark also has a distinct advantage compared to other countries in that it is connected to two large grids, the Nordic grid and the German grid and it has relatively small output. The structure of the Danish wind market is vastly different than most as well. Whereas as many countries are moving towards privatization there are over 3000 co-op wind farms in Denmark leading to 150,000 owners (Reiche & Bechberger, 2004). This has the effect of leading to increased acceptance amongst the Danish people.  62  7  CONCLUSION Based on the theoretical model developed, government regulation or other factors  which increase the cost of doing business in an international market may lead to domestic firms innovating more than international firms. Empirical evidence shows that domestic wind energy output might increase the expected number of patents in a country more than international wind energy output. The results of the regressions are inconclusive.  The answer to the second question, which policies have the most impact on the domestic change in wind capacity, is unclear. The only variable that is robust to the inclusion and exclusion of different countries is obligations which remains significant and positive throughout. This implies the implementation of a policy involving an obligation to use a certain amount of renewable energy results in a positive increase in the change in capacity of wind energy. This is not surprising due to the strength and ease of implementation of this policy. Other policies were generally insignificant for most model specifications, which could be due to the multicolinearity. When looking at several different countries one can see that the choice of whether to implement policies to encourage wind power and the strength of those policies is very individualistic. Exactly what contributes to the size of the market is difficult to determine because factors other than economic factors, such as social attitudes towards wind or nuclear power and constraints of the grid and domestic potential of other sources of reliable energy, all play a part.  63  A critical missing element to this research is the strength of the policy implemented; however, this is extremely hard to detemiine because within a country there are numerous regional governments implementing many different policies. As the data set gets longer and more green energy sources are widely used throughout the world, the factors leading to increases in capacity may become clearer. Future work should include some level of the stringency of the policies; however, at this time there is not enough information to produce any strong results on different policy effects on the change in wind capacity.  64  REFERENCES Arrow, K. 3. (1962). Economic Welfare and the Allocation of Resources for Invention. The RAND Corporation. Baumol, W. 3., & Oates, W. E. 0. (1971). The Use of Standards and Prices for the Protection of the Enivonment. The Swedish Journal ofEconomics, 73(1), 42-54. BBC News. (2009). EU climate package explained Retrieved March 16, 2009 Beghin, 3. C., & Lovell, C. A. K. (1993). Trade and Efficiency Effects of Domestic Content Protection: The Australian Tobacco and Cigarette Industries. The Review ofEconomics and Statistics, 75(4), 623-631. Bentley, R. W. (2002). Global oil & gas depletion:an overview. Energy Policy, 30, 189205. Branco, F. (1994). Favoring domestic firms in procurement contracts. Journal of International Economics, 37, 65-80. Cameron, A. C., & Trivedi, P. K. (1998). Regression Analysis of Count Data. New York: Cambridge University Press. Central Intelligence Agency. (2008). CIA world fact book. Climate Change and Power Economic Instrumentsfor the European Electricity. (2002). London: Earthscan. Dahi, C. A. (1979). Consumer Adjustment to a Gasoline Tax. The Review ofEconomics and Statistics, 61(3), 427-432. Dasgupta, P., & Stiglitz, J. (1981). Entery, Innovation, Exit Towards a Dynamic Theory of Oligopolistic Industrial Structure. European Economic Review, 15, 137-158. de Gorter, H., & Just, D. R. (2009). The Economics of a Blend Mandate for Biofuels. American Journal ofAgricultural Economics. Dernis, H., & Guellec, D. (2001). Using Patent Counts for Cross-Countiy Comparisons of Technology Ouçput: OECD. Dixon, P. B., & Osborne, M. T. R. (2007). The Economy-Wide Effects in the United States ofReplacing Crude Petroleum with Biomass. Paper presented at the GTAP Conference. Energy Information Administration. (2009). Official Statistics from the US Government. European Patent Office. (2008). esp@cenet. Retrieved September 19, 2008, from http://ep.espacenet.com/ Fischer, C., & Newell, R. G. (2004). Environmental and Technology Policies for Climate Change and Renewable Energy. Griliches, Z. (1992). The Search for R&D Spillovers. Scandinavian Journal of Economics, 94, 29-47. Hall, B. H., & Van Reenen, J. (2000). How effective are fisical incentives for R&D? a review of evidence. Research Policy, 29, 449-469. Hausman, J., Hall, B. H., & Griliches, Z. (1984). Econometric Models for Count Data with An Application to the Patents-R&D Relationship. Econometrica, 52(4), 909937. Herbert, G. M. 3., Iniyan, S., Sreevalan, E., & Rajapandian, 5. (2007). A review of wind energy technologies. Renewable and Sustainable Energy Reviews, 11(6), 11171145.  65  Hicks, J. R. (1966). The Theoiy of Wages. New York: St Martin’s Press. International Energy Agency. (2006). Renewable Energy: RD&D Priorities Insightfrom Technology Programmes. Paris: International Energy Agency. International Energy Agency. (2009). International Energy Agency Data Services. Retrieved January 5, 2009, from http://wds.iea.org/wds/default.aspx International Monetary Fund. (2009). World Economic Outlook Database. Retrieved January 5, 2009, from http://www.imf.orglexternallpubs/ft/weo/2008/02/weodatalindex.aspx Johnstone, N., Hascic, I., & Popp, D. (2008). Renewable Energy Policies and Technological Innovation: Evidence Based on Patent Counts. Krishna, K., & Itoh, M. (1988). Content Protection and Oligopolistic Interactions. The Review ofEconomic Studies, 55(1), 107-125. Lee, T., & Wilde, L. L. (1980). Market Structure and Innovation: A Reformulation. The Quarterly Journal ofEconomics, 94, 429-43 6. Loury, G. C. (1979). Market Structure and Innovation. The Quarterly Journal of Economics, 93, 395-410. McGowan, F. (2000). Reconciling EU Energy and Environmental Policy. In P. Jasinski & W. Pfaffeberger (Eds.), Energy and Environment: Multiregulation in Europe (pp. 1-2 1). Burlington: Ashgate Publishing Company. Menanteau, P., Finon, D., & Lamy, M.-L. (2003). Prices versus quantities: choosing policies for promoting the development of renewable energy. Energy Policy, 31, 799-812. Miyagiwa, K. (1991). Oligopoly and Discriminatory Government Procurement Policy. American Economic Review, 81(5), 1320-1328. NBC News. (2008). U.S. Edges Out Germany as World Wind Power Leader [Electronic Version]. Retrieved Febuary 18, 2009 from http://www.nbcnewyork.comlnews/green/US Edges Out_Germany_as_World _Wind_Power_Leader.html. Nelson, R. (1959). The Simple Economics of Basic Research. Journal ofPolitical Economy(67), 297-306. Newell, R. G., Jaffe, A. B., & Stavins, R. N. (1999). The Induced Innovation Hypothesis and Energy-Saving Technological Change. The Quarterly Journal ofEconomics, 114(3), 941-975. Nielsen, J. U.-M., & Hansen, L. G. (2001). The EU public procurement regime— Does it work? Intereconomics, 36(5), 255-263. Nordaus, W. (1972). The Optimum Life of a Patent: Reply. American Economic Review, 62(3), 428-431. OECD/IEA. (2004). Renewable Energy Market & Policy Trends in TEA Countries [Electronic Version]. OECDITEA. (2008). Global Renewable Energy Policies and Measures. Retrieved October 10, 2008, from http://www.iea.orgJtextbase/pm/?mode=re Popp, D. (2002). Induced Innovation and Energy Prices. The American Economic Review, 92(1), 160-180. Popp, D. (2006). Internatinal innovation and diffusion or air pollution control technologies: the effects of NOx and SO2 regulation in the US, Japan and Germany. Journal ofEnvironmental Economics and Management, 51,46-71. .  66  Reiche, D., & Bechberger, M. (2004). Policy differences in the promotion of renewable energies in the EU member states. Energy Policy(32), 843—849. Schumpeter, J. A. (1942). Capitalism, Socialism and Democracy. New York: Harper and Brothers. Spencer, M. (1984). Cost Reduction, Competition, and Industry Performance. Econometrica, 52(1), 101-121. World International Patent Organization. (2008). International Patent Classification. Retrieved October 18, 2008, 2008, from http://www.wipo.int/classifications/ipc/en!  67  APPENDIX  Table 8.1: Change in Research as Feed-in-Tariff Changes deltan 0.00 0.00 0.00  deltam 5.00 5.00 5.00  Cr 20.00 20.00 20.00  Ct 10.00 10.00 10.00  aipham 200.00 200.00 200.00  betam 0.50 0.50 0.50  aiphan 300.00 300.00 300.00  betan 2.00 2.00 2.00  sm 1.00 1.20 1.50  Phim 0.00 0.00 0.00  Phin 0.00 0.00 0.00  1/k 300 300 300  profit m 10472 12373 15254  profit n 9739 11606 14454  xm 33.12 38.76 47.08  xn 30.67 36.19 44.40  0.00 0.00 0.00 0.00  10.00 10.00 10.00 10.00  20.00 20.00 20.00 20.00  10.00 10.00 10.00 10.00  200.00 200.00 200.00 200.00  0.50 0.50 0.50 0.50  300.00 300.00 300.00 300.00  2.00 2.00 2.00 2,00  1.00 1.10 1.70 1,90  0.00 0.00 0.00 0.00  0.00 0.00 0.00 0.00  300 300 300 300  10472 11420 17186 19124  9050 9957 15581 17491  33.25 36.09 52.76 58.15  28.49 31.19 47.37 52.65  0.00 0.00 0.00 0.00  10.00 10.00 10.00 10.00  30.00 30.00 30.00 30.00  10.00 10.00 10.00 10.00  300.00 300.00 300.00 300.00  0.50 0.50 0.50 0.50  200.00 200.00 200.00 200.00  2.00 2.00 2.00 2.00  1.00 1.30 1.40 1.70  0.00 0.00 0.00 0.00  0.00 0.00 0.00 0.00  200 200 200 200  15139 21361 23453 29758  13006 19084 21141 27369  65.40 86.86 93.52 112.16  54.40 74.70  Q.99  change xm  change xn  5.63 8.33  5.52 821  2.84 16.68 5.39  2.70 16.17 5.28  21.46 6.66 18,64  20,30 6.29 17.48  Table 8.2: Data Countiy  Austria Austria Austria Austria Austria Austria Austria Austria  Land dummy Feed-in- Obligation&’ Targets Tax Output OECD Patents Change GDP US GDP US per Crude Oil Grant to Grants to b Area coastline Prices Producers companies taiiffs incentiv&’ Green Gwh total Wind Billiont Capitac d,e Capacitya certifacates output Gwh 0 21541.98 22,66 0 0 0 0 0 0 1990 0 3845 0 165.396 0 82444 82444 22357.74 0 0 0 0 4148 19.53 0 0 0 1991 0 0 0 173.382 0 0 82444 0 24883.82 18.54 0 0 0 0 0 1992 0 4537 0 0 195.107 0 82444 23996.95 16.14 0 0 0 0 0 0 1993 5473 0 0 189.709 0 0 0 0 82444 0 25701.69 15.63 0 0 0 0 0 1994 0 7056 2 0 203.972 0 0 82444 0 7349 1 1 238.55 30012.69 17.06 0 0 0 0 0 1995 1 0 82444 29429.96 20.67 0 0 0 0 0 1996 5 8380 0 9 234234 0 0 0 0 0 0 82444 0 20 10721 1 9 207.126 25994.61 18.88 0 0 0 1997 0  Time  68  Austria  1998 1999  45  14425  0  8  212.439  26632.13 26426.19  12.3 17.32  0  0  0  0  0  51  2000  67  19343 28551  0 2  8 19  211.206 191.761  0  0  0  0  0  0  23935.5  27.89 23.92  0  0  0  0  1  0  Austria Austria Austria  2001 2002  172 203  34632 48335  4 0  190.319 206.684  23662.5 25567.66  2003  366  58438  2  15 64 210  252.516  31106.68  24.3 28.39  0 0  1 1  1 1  0  1 1  0 0  0  1  1  0 0  1  0  Austria  2004  924  76864  1  217  289.419  35403.98  Austria Austria  2005 2006  1328 1722  93657 116182  2  267 142  304.529 323.071  36987.45 39098.61  36.53  0  1  1  0  1  0  51.74 62.77  1 1  1 1  1 1  0 0  1 1  0  Belgium  1990 1991  7 8  197.859  Belgium  4148  Belgium Belgium  1992  9  0  202.936 225.577  1993  8  4537 5473  0 0  0  19831.6  22.66  0  19.53 18.54  0  0 0  0 0  0  20263.23 22434.17  0 0  0  0  0 30278 0 30278 0 30278  21428.12  16.14  0  0 1  0 0  0 0  0  0  216.297  0  0  Belgium  1994  9  7056  0  0  0  1  23317.53  15.63  0  1  0  0  0  1  0 30278 0 30278  1  236.116  Belgium  1995  9  7349  0  0  277.151  1  1  0  0  0 0  275.708 249.639  17.06 20.67  0  0 0  27324.34 27109.89 24493.6  18.88  0 0  1 1  1 1  0 0  0 0  1 1 1  0 30278 0 30278 0 30278  Belgium Belgium  1996 1997  8 8  8380 10721  1 1 1  Belgium  1998  Belgium Belgium  1999 2000  11 13  14425 19343  0 2  1  25020.6 24792.54  12.3 17.32  0  1 1  1 1  0 0  0 0  1 1  16  28551  27.89 23,92  1 1  0 0  252.714  24405.05  24.3  1  1  1  Belgium Belgium  2003 2004  88 142  58438  3  28.39  0  1  1  1  0  1  0 1  3 10.526 359,3  29869.79  76864  36 29  0 30278 0 30278 0 30278  1  5  0 0  1 1  3  0 0  1 1  0  34632 48335  22669.1 22512.64  1 1 1  37  232.653 232.105  0 30278 0 30278 0 30278  2001 2002  6 3  4 4 12  255.56 253.851  Belgium Belgium  34395.94  36.53  0  1  0  1  0  1  71  376.399 398.192  35809.98  51.74  0  1  62.77  0  1  0 0  1  37618.56  1  0 0  1 1  52.128  5058.64  22.66  0  0  0  0  0  0  Austria Austria  57  3845  Belgium  2005  227  Belgium  2006  366  93657 116182  Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic  1990  0  1991  0 1  0 0  0  0  0  1  45  3845  0 0  0  4148  0  0  27.18  2636.3  19.53  0  0  0  0  0  0  1992  0  4537  0  0  31.898  3090,94  18.54  0  0  0  0  0  0  1993  0  5473  0  0  36.651  3551.5  16.14  0  0  0  0  0  0  1994  0  7056  0  0  42,538  4117.9  15.63  0  0  0  0  0  0  1995  0  7349  0  0  55.256  5348,65  17.06  0  0  0  0  0  0  0 82444 0 82444 0 82444  0 0 0 0  0 82444 0 82444 0 82444  0 0  0 82444 0 82444 0 82444  0 0 0 1 1 1 1  1 1  0 30278 0 30278  1 1  0 30278 77276 0 77276 0 77276 0 77276 0 77276 0 77276 0  0 0 0 0 0 0  69  Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Czech Republic Denmark  1996  0  8380  0  0  62.011  6011.54  20,67  0  0  0  0  0  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276  0  77276 0 0 42394 0 42394  0  0 1997  0  10721  0  0  57.135  5545.14  18.88  0  0  0  0  0  0 0  1998  0  14425  0  1  61.847  6033.68  12.3  0  0  0  0  0  0 0  1999  0  19343  0  0  60.192  5880.87  17.32  0  0  0  0  0  0 0  2000  0  28551  0  0  56.717  5548.48  27.89  0  0  0  0  0  0 0  2001  0  34632  0  0  61.843  6077.48  23.92  0  0  0  0  0  0 0  2002  2  48335  0  5  75.276  7401.06  24.3  0  0  0  1  0  0 0  2003  4  58438  0  5  91.358  8974.9  28.39  0  0  1  1  0  0 0  2004  10  76864  0  5  109.525  10742.47  36.53  0  0  1  0 0  2005  21  13  93657  2006  49  116182  0  1990  610  3845  Denmark  1991  740  4148  0 1  Denmark  1992  Denmark  1993  915 1034  4537 5473  0 1  Denmark  1994  1137  7056  Denmark  1995  1177  7349  Denmark  1996  1227  8380  Denmark Denmark Denmark  1997 1998  1934 2820  10721 14425  Denmark  1999 2000  3029 4241  19343 28551  10  Denmark Denmark  2001 2002  4306 4877  34632 48335  8 17  Denmark  5561  Denmark  2003 2004  6583  58438 76864  23 23  Denmark  2005  6614  Denmark Finland  2006 1990  6108 0  93657 116182 3845  0  15  124.549  12175.43  51.74  142.313  13863.89  62.77  136.174  26490.38  22.66  70 45  137.189 150.54  33  140.835  5  41  2  84  3 2 10 15  0  0  1  1  1  0 0  0  0  1  o  i  0  1  0 0  0  1  1  0 0  0 0  1  0 0  0  1 1  26618  19.53  29109.62 27141.09  18.54 16.14  0 0  1 1  153.901  29559.3  15.63  0  1  0  0  1  182.179  34810.19  17.06  0  1  0  0  1  226 288  184.481 170.642  35052.38 32287.96  20.67 18.88  0  0  0  1  0  1 1  1 1  313  173.902 174.172  32783,81 32726.84  12.3 17.32  1 I  160.533 160.583  30065.24 29967.94  27.89 23.92  0 0 0  394 225  174.42 212.981  32444.25  244.983  23  8 4  39506.68 45329.46  24.3 28.39  258.58  47717.25  11  6  276.283 139.83  50815.38 28042,04  316 633 106  36.53 5 1,74 62.77 22.66  1  1 1  0 1 1  0  1 1  1 1  1 1  1  0 0  0  0  1 1  0 0 0  1 1 1 1  0  0 0  0 0  0 0  1  0  0  1  0  0  1  0  0  1  0 0  0 0  1  0 0  0 0  0  0  0 42394 0 42394 0 42394 0 42394 0 42394 42394 42394 0 42394 1 42394 1 42394 1 42394 1 42394 1 42394 1 42394 1 42394 0 304473  1 1  70  Finland  1991  0  4148  0  1  126.422  25215.14  19.53  0  0  0  0  1992  2  4537  1  0  110.81  21977.51  18.54  0  0 0  0  Finland  0  0  0  Finland Finland  1993 1994  4  5473  2  4  87.421  17254.83  16.14  0 0  0 1  100.714 130,75  19793.1 25598,21  15.63 17.06  0 0  0 0  1995  7056 7349  0 0  Finland  7 11  0 0  0 1  0  0  0  0  1  0 0  Finland Finland  1996  11  8380  0  1  128.525  25080.17  20.67  0  0  17  10721  1  5  123.428  24014.02  18.88  0 0  0  1997  0  0  0  1 1  0 1  Finland  23 49  14425 19343  0 1  5 21  130.466  25315.98 25350.7  12.3  1  Finland  1998 1999  17.32  1  0 0  0 0  0 0  1 1  1 1  Finland  2000  78  28551  0  23612,3  27.89  1  0  0  1  2001  34632  0  0  0  1  1  2002  48335  4  23.92 24.3  1  Finland  70 64  0 0  1  Finland  0 1  1  0  0  0  1  1  Finland  2003  28.39  0  0  0  1  1  Finland  30 0  0 0  0  0  0 0 0  0 0 0  1 1  1 1  1 0  1 1  0  0  0  1  0 0  0 0  0 640053 o 640053  1  0 640053 0 640053  1 1  58438  2004  93 120  0 1  76864  2  Finland Finland  2005 2006  170 156  93657 116182  France France  1990 1991  0  3845 4148  France  1992  0 0  France  1993  France  1994  France  3  9  130.948 122,222 125.269  24145.89  135.972 165.031  26145.42 31657.45  189.411  36228,95  36.53  1 1  1  0  0 304473 0 304473 0 304473 0 304473  1 1 1 1 1  0 304473 0 304473  1  0 303 0 304473 0 304473  1 1 1  0 304473 0 304473 0 304473 0 304473  1 1 1  0 304473 0 304473 0 304473 0 640053  4 1  4  196.001 209.745  37361.28 39827.97  51,74 62.77  1 1 0  1  22015.74 21924.26  22.66  4537  0 0  1248.49 1249.15  0 0  1373.98  24003.98  0 0  0 0  2  5473  0  0 2  19.53 18.54  1292.04  22483.14  16.14  0  0  0 0  5  7056  0  0  1366.08  23692.46  15.63  0  0  0  0  0  1  1995  5  7349  1  0  1572.3  27181.59  17.06  1  0  0  0  0  1  France  1996  8380  1574.52  27134.71  20.67  1  1  1425.86 1474.31  24496.07 25245.93  18.88 12,3  0  0 0  0 0  1  10721 14425  1 1  0  1997 1998  3 3  3  France France  7 11 20  0 640053 0 640053  France  1999  37  19343  France  2000  France France  2001 2002  77 131 269  France France  2003 2004  391 596  France France  2005 2006  Gennany Gennany  1990 1991  1  1 1  0  8  1  0  1 1  0  0  1  640053 640053  3 39  1458.32  24854.93  17.32  1  0  1  0  0  1  640053  28551  3 2  1333.2  22577.69  27.89  1  22556.05 24450.75 29954.43  23.92 24.3 28.39  1 1 1  76864  1  89 141  1341.06 1463.7 1805,18  0 0 0  2061.09  33987.16  36.53  1  0 640053 0 640053 0 640053 0 640053 0 640053  1  26 50  0 0  1  5 4 4  0 0 0  1  34632 48335 58438  963 2150  93657 116182  3 1  360 665  3845 4148  1 4  62  35206.99 37019.84 19592.74 22692.8  51.74 62.77 22.66 19.53  1 1 1 1  640053 0 640053 0 349223 0 349223  1  71 215  2147.48 2271.26 1547.03 1815.06  0 0 0 0 1 1  1 1 1 1 1 1 1 1  0 0  0  0 0  0 0  0 0 0  0 0 0  1 1 1 1 1 1 0 0  1  1 1  o  71  291  73  2066.73  25523.16  18.54  1  1  1  0  0  0  4  151  2005.56  24657,05  16.14  1  1  1  0  0  0  6  309 494  2151.03  26380.45  15.63  1  1  1  0  0  0  2524.95 2439.35 2163.23  17.06 20.67  1 1  1 1  1 1  0  427 402  30860.75 29743.72  0 0  26362,45  18.88  I  1  1  0 0  0 0 0  0  706 1466 1957  2187.48  26664.6  12.3  1  0  0  0  26123.92 23168.07  17.32 27.89  1 1  1 1  1  2146.43 1905.8  1  1 1  0 0  0 0  76 45  2659  1892.6  22957.16  23.92  1  1  1  0 1 1  0  0  3247  2024.06  24523.16  24.3  1  1  1  1  0  0  69  2608  2446,89  29647.83  28.39  1  0  2748,82  33318.7  36.53  1  1 1  0  2020  1 1  1  62 50  0  0  1799  2794.48  51.74  1  2194  2914.99 92.195 99.422  1 1  1  40  33897.93 35422.06  1 1  1  1  0  1 0  0 0  0 0  0 0  0 0  18.54  0 0  0 0  0 0 0  0  0  0  0  9816,3  16.14  0  0  0  0  0  0  15.63  1  0  1  0  0  0  Gennany  1992  4537  1  Gemiany  1993  Germany  1994  674  5473  1428 1712  7056 7349  Germany Gennany  1995 1996  2078  8380  3 8  Germany  1997  3034  Germany Gennany  1998 1999  13 23  Germany Germany  4593 5528 9352  10721 14425  2000 2001 10456  19343 28551  34 46  34632  Germany  2002 15856  48335  Germany  2003 18859  Germany  2004 25509 2005 27229  58438 76864  Germany Germany Greece Greece Greece  2006 30710 1990 2 1991 2  93657 116182 3845 4148  1 0  9073.82 9705.08 10591.03  62.77 22.66 19.53  0  1992  8  4537  0 0  Greece  1993  47  5473  1  11  109.556 102.608  Greece Greece  1994  37  7056  0  0  109.824  10400.74  1995  34  7349  0  0  128.895  12096,48  17.06  1  1  1  0  0  0  Greece  1996  38  0  0  20.67  1  1  0  36  0  12311.98  1998 1999  14425  0 0  133.869 137,829  12308.92 12610.95  18.88 12.3  1 1  1 1  1  0 0  0 0  0 0  Greece  0 11 71  1 1  0  1997  136273 133.128  12688,9  Greece  8380 10721  17.32 27.89  1 1  1 1  1 1  0 0  0 0  15  0 0  Greece Greece  2000  70 162 451  127.604  11627.2  2001 2002 2003  756 651 1021  34632  0 1  117  Greece Greece Greece  44  1  1  1  0  0  0 2  17 84  11896.61 13482.68 17510.59  23.92  48335 58438  130.994 148.827 193.663  24.3 28.39  1 1  1 1  0 0  0 0  Greece  2004  1121  76864  0  230.291  20785.83  36.53  1  1  0  0  C)  Greece  2005  1266  93657  0  99 21  1 1 1  0 0 0  247.418  22290.45  51.74  1  1  1  0  0  0  Greece Ireland  2006  1699  116182  1  258  24157,31  62.77  1  1  1  0  0  1990 1991 1992  0  3845 4148 4537  0 0 0  268.69 47.753  13621.19 13727.88 15308.83  22.66 19.53 18.54  0 0 0  0 0 0  0 0 0  0 0 0  0 0 0  0 1 1 1  Ireland Ireland  0 5  19343 28551  0 6  48.4 54.415  0  0 349223 349223 0 349223  1 1 1 1  0 349223 0 349223 0 349223  1  0 3’193 0 349223 0 349223  1 1 1  0 349223 0 349223  1  0 349223 0 349223 0 349223  1  0 34fl23 0 130800 0 130800 0 130800  .1 1 1  0 130800 0 130800  1  0 130800 o 130800  1  1  1 1 1  1 1 1 1  0 130800 130800  1 1  0 130800 0 130800 0 130800  1 1  0 130800 0 130800 130800  1 1  0 130800 0 130800  1  68890 68890 0 68890  1 1 1  1 1  72  15  5473  0  Ireland Ireland  1993 1994  19  7056  Ireland  1995  16  7349  0 1  Ireland  1996 1997  14  8380 10721  0 0  0  50.421  14107.37  16.14  0  0  0  0  0 0  0  55.327  15429.05  15.63  0  0  0  0  0 0  18620.83  17.06  0  0  0  20411.26 22176.01  20.67 18.88  0 0  0 0  0 0  0 0  0 0  51  67.059 74.013 8126  0  0  1 1 1 1 1  0 68890 0 68890 0 68890 0 68890 0 68890 0 68890  1998  50 169  14425  0  5  88.227  23825.09  12.3  0  0  0  0  1999  187  19343  0  8  96.548  25803.86  17.32  0  0  0  0  0 1  1  Ireland Ireland  2000  244  1  45  27.89  0  0  0  0  1  1  1 68890 1 68890  2001  334  0  96.879 104.778  25565.25  Ireland  28551 34632  27234.97  0  0  1  1  0  0  0  0 0  1  31388.22  23.92 24.3  0  122.954  1  1  157.685  39620.27  28.39  0  0  0  0  1  1 68890 68890  36.53 51.74  0  0 0  0 1  0 0  1  1 1  0  1 0  0 0  1  1 68890 1 68890 1 68890  0  0  o294020  0 1  0 0  0 0  0 0 0  o294020 0294020 294020  Ireland Ireland  Ireland  2002  388  48335  0  8 13  Ireland  2003  454  0  74  Ireland  2004  0  131  185.221  45787.88  Ireland  2005 2006  655 1112  58438 76864 93657 116182  0 0  153 252  202.022 222.609  48870.67 52504.7  3845 4148  1  1135.54  20029.19  1198.99 1271.91  21129.68 22403.3 17997.61  Ireland Italy Italy Italy Italy Italy  1622  1990  2  1991 1992  3 2 4  1993 1994  7  4537 5473 7056  Italy  1995  9  7349  Italy  1996  33  8380  Italy  1997  Italy  1998 1999  118 231  10721 14425  403  19343  0 0 0 1  3 11 3  1022.66 1054.9  22.66 19.53  0 1 1 1  15.63  1  0  1 1  0  18557.95  0  0 0  0  o294020  17.06  1  0  1  0  0  0  o294020  0 0  0  0  0  0  o294020  0  0  0  o294020  0 1  0 0  0 0  1  0  1 1 1  0 0  1  1126.63  12  1259.95  22164.1  20.67  0  1  85  1193.62  2  45  18.88 12.3  0 0  0  0  0  68 131  1218.67 1202.4  20985.09 21433.7 21129,55  17.32  0  0  0  27.89 23.92  0  0 0  0 0  0 0  0 0 0  2000 2001  563 1179  28551 34632  1 2  Italy Italy  2002 2003 2004  1404 1458 1847  48335 58438 76864  8 3 3  1100.56  19293.4  301  1118.32  19541.11  116 94  1223.24 1510.06 1730.1  21317.51 26308.26 30119  253  24.3 28.39 36.53  0 0 0 0  Italy  2005  2344  93657  3  508  1779.41  30638.87  51.74  0  Italy Netherlands  2006 1990 1991  2971  116182  1858,34  31801.63  3845 4148 4537 5473  2 1  267  56  0 0  33 18 30  19760.56 20136.83 22119.79 21286.59  62.77 22.66  0 1  295.46 303.462 334.654  19.53 18.54 16.14  0 0 0  1992 1993  88 147 174  324.39  1 1  0 0 0  0 0 0  0 0 0  0 0  0  0  1 1 1  0  i294020  0  i294020 i294020 i294020  0 0  0  0 0  0  0 0  0 0 0  0 0 0  0 0 0 0  1  1  o294020 i294020  0 0  0 0  1  68890  0 0  18.54 16.14  0 1  Italy  Netherlands Netherlands Netherlands  0 0  19819.03  Italy Italy  Italy  1  62.77  0 0  1  i294020  1294020 0 33883 33883 0 33883 0 33883  73  Netherlands  1994  Netherlands Netherlands Netherlands  21  348.911  22742.86  15.63  0  0  0  7349  o o  98 46 28  27187.8 26985.21  17.06 20.67  0  2 2  419.348 418.106  0  8380 10721  24860.92  18.88  0 1  14425 19343  1 4  39 47  387.013 403.202  0 0  12.3 17,32  0  6  37  24250.65  27.89  0 0  1 1  28551  411.997 386.204  25756.79 26141.54  238  7056  1995  317  1996 1997 1998  437 475  0  0  0  0  0  0  1  0  0  0  1  0 0  0 0  0 0  1  0 0  0 0  0 0  1 1  1  1 1 1  33883 33883  1  1  1  33883  1  1  33883  1  1  33883  1  1  33883  1999  640 645  2000  829  Netherlands Netherlands  2001 2002  825  34632  400.998  0  1  0  185  439.357  24990.27 27206.57  23.92  48335  3 4  38  946  24.3  0  1  0  0 0  0 0  Netherlands  2003  1318  58438  6  236  539.343  33240.83  28.39  0  0  1  0  0  1  Netherlands  2004  1867  76864  2  167  610.691  37520.24  36.53  0  0  1  0  0  Netherlands  2005  2067  1  51.74 62.77  3845  0  0 0  0 0  0 0  0  Norway  1991  0 0  0 0  1 1  0  8  39210.61 41476.76  0  2733  639.579 677.961  0  2006 1990  151 334  0  Netherlands Norway  93657 116182 4148  0  0  0 0  0 0  1994  7056  0  0 0  0 0  Norway  3 9  4537 5473  0 0 0  0  1992 1993  0 0  0  Norway Norway  0 0 1  0  0  0  Norway  1995  10  7349  0  Norway  1996  8380  3  Norway  1997  9 10  Norway  1998 1999  11 25  Norway Norway Norway Norway Norway  117,865  27763,99  22.66  120.068  0  128.591 118.284  28143.58 29965.6 27404.05  1953 18.54 16.14  124.737  28732.29  15.63  0 0  0 0  3  149.007  34149.84  17.06  0  0  0  0  0  0  1  160.173  36521.51  0  0  0  0  0  10721 14425  0 1  0  35955.94 34075.61  0  0  0  0  12.3  159.093  35619.4  0 0  0 1  0 0  0 0  1  0 0  0 0  19343  0 10  158.55 151,156  20.67 18.88  0  37520.11 37840.31  0  1 1  0  0  23,92  0 1  1 1  1  0 0  0 0 0 1  17.32 27.89  2000  31  28551  2  -1  168.671  2001 2002  27  34632  o  o  75 218  48335 58438  2  84  170.982 193.175  42525.73  24.3  1  3  0  76864 93657  1  55 118  225.307 258.986  49316.72 56344.18  28.39  252  1 1  302.175  65604.6 72768.13  2003  Norway Norway  2004 2005  Norway  2006  506 673  Portugal Portugal Portugal Portugal Portugal  1990 1991 1992 1993 1994  1 1  0  116182  5 0  1  3845  0  1 4 11 17  4148 4537 5473  0  o  0 2 5  7056  0  0  0  14  337.426 75.967 85.975 103,394 90.982 95.335  7599.78 8623.01 10375.38 9121.33 9542.48  36.53 51.74 62.77 22.66 19,53 18.54 16.14 15.63  0  0 0  1  1 33883 1 33883 0 212460 o212460 212460 212460 0 212460 0 212460  1  0 212460 0 212460 0 212460 0 212460 0 212460  1  0 212460 0 212460 1 212460  0 0  1 212460 1 212460  0  1 212460 0 91951  0  1  0  0 0  0 0  0 0  0  0  0  1 1  0  0  0  0  0  0  0 0 0 0  0 0 0 0  0 0 0 0  0  0  0 0 0  0 0 0  0 0  1 1  1  33883 33883  1  Netherlands Netherlands Netherlands  0  0 33883 1 33883 1 33883  o  0 91951 0 91951 91951 0 91951  74  Portugal  1995  16  1996  21  7349 8380  0  Portugal Portugal  1997  38  10721  Portugal  1998 1999  89 123  14425 19343  0 0  168  28551  Portugal  2000 2001  256  34632  Portugal  2002  362  Portugal  2003  496  Portugal  2004  Portugal  2005 2006  Portugal Portugal  Portugal Spain Spain Spain  1990 1991 1992  0 10  113.017 117.658  11 19  0  9  0 0  26 42  48335  0  65  127.906  12382.79  24.3  0  58438  0  78  156.712  15057.65  28.39  0  816  76864  0  285  179.195  17107.4  36.53  0  1773  0  511 617  185.771 195.186  17643.35 18466.74  51.74  2925 14  93657 116182 3845  520.709  13407.9  15  4148  560.796 613.016  14401.9 15691.09  514.949 516.718  0  0 0 0 0  1  11281.89  17.06  0  0  0  0  0 0  0  20.67  0 0  0 0  0  112.134  11715.22 11132.65  0  0  0  0  0  11742.31  18.88 12.3  0  118.711  0  0  0  0  0  121.823 112.98  12003.54  17.32  0  0 0  0  0  27.89  0  0 0  0  11081.91  0  0  0  115.812  11291.37  23.92  0  1  0  0  0  1  1  0  0  1  1  0  0  0 0  1  1  0  0  0  0 0  1 1  1 1  0 0  0 0  0 0  0  19.53  1 1  0 0  0 0  18.54 16.14 15.63  1  0  13140 13149.72  0 0  1 1  0 0  1  62.77 22.66  1 1  0 0  0 0 0  0 0  o499542 0 499542  0  0  0  0 499542 0 499542  1994  175  7056  Spain  270  7349  15164.36  17.06  0  364  8380  0 1  597.278  Spain  1995 1996  0 0  129  622.65  15772.03  20.67  0  0  1  0  0  0  Spain  1997  742  10721  0  193  573.376  14485.63  18.88  0  0  1  1  0  0  Spain  1998  1352  14425  0  428  0  0  19343 28551  7 1  765 593  0  0  14464.24  0 0  0  582.377  17.32 27.89  1 1  0  2744 4727  1 1  0  1999 2000  15146.23 15495.85  12.3  Spain Spain  601.625 618.691  1  Spain  2001  6759  34632  609.631  14971.13  23.92  48335  1054  24.3 28.39  0  24493.52  36.53  2005 21176 2006 23040  93657 116182  19  1045.67 1132.13  0 0  0  2004 15700  4 16  16669.31 21077.65  0  58438 76864  688.676 885.358  0 0  0 1  2002 9342 2003 12075  1191 1494  0 0  Spain Spain  5 4  1 1  1233.43 242.848  26087.08 27989.03 28268.9  51.74 62.77  4  256.344  29655.26  22.66 19.53  8  266.224 202.589  30628.56  18.54  23166.01 24709.03 28753.02  16.14 15.63 17.06  Spain Spain Spain Sweden Sweden  1990  6  3845  3 1  1991  13  4148  0  Sweden  1992 1993 1994  31 48  4537  0 1  Sweden Sweden Sweden  1995  72 99  5473 7056 7349  0 1  7 57  2372 1601 1818  9 11 27  217.844 254.105  0  0 0 0 0 0 0 0 0 0 0  0  0  1  0 0  1 1  1 1  0  1 1  1 1  1  1  0  0  0 0  0  0  0  0  0 0 0 0 0  0 0 0 0  0 0 0 0  0  0 0 0  0 0 0 0  0 91951 0 91951  0 0  4537 5473  0  0 0 0 0  0 1 1  1  0 91951 0 91951 0 91951  0  103 116  0  0 91951 0 91951  499542 499542 0 499542  1993  30 1  0 91951 0 91951 0 91951  0 0  Spain Spain  0 1  0 91951 0 91951  0 499542 0 499542 o 499542 499542  1  1  p499542 0 499542 0 499542 0 499542  I 1  o499542  1  o499542 0 410934  1  0 410934 0 410934 0 410934 41O934 410934  1  75  Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland United Kingdom United Kingdom United Kingdom United Kingdom United  1996  144  8380  0  38  276,23  31231.81  20.67  0  0  0  0  0  1  410934  1  1997 1998  203  2  1  0  0  358 457  19343 28551  6 0  22 13  29027,48 27735.91  17.32 27.89  0 0 0  0 0 0  1 1  2001  482  34632  86  225.546  25316.3  23.92  0  0  0  0  1  2002  608  48335  7 2  1 1  0 0  0 0 0  0 410934 0 410934  1999 2000  1 1  1 1  1  12.3  0 0  0  253.154 257.225 246.372  28551.6 28590.99  18.88  0  18 51  252.614  317  10721 14425  62  249.378  27892.13  24.3  1  0  0  0  0  1  0 410934 0 410934 0 410934 0 410934  2003  679  58438  42  311.763  3473421  28.39  1  0  0  0  0  1  2004  850  76864  53 41  357.721  36.53 51.74  1  0  0  0  0  1  367.162  39696.58 40580.52  0  1  43190.45  62.77  0  0  1  239.332  35452.95 35361.64  22.66  0  0  0  0  19.53  0  0 0  0 1  0 0  0  393.606  1 1  0  23  0  0  1  0 0  0 0  0  36449.46  0 0  0  1 1  0 0 0  .0 0 0  2005  0 1 2  936 987  93657 116182  1990 1991  0  3845  0 0  0  4148  0  0  1992  0  0  1993  0 0  4537 5473  0 0  2006  1994 1995 1996  0 1  1997  2  1998  3 3  1999 2000 2001 2002  3 4 5 5  7056 7349  0 0 0  8380 10721  0  14425  0 1  19343 28551 34632  0  0  0 0 2  241.971 251.791 244198  35042.82  18.54 16,14  270.913 316.418  38597.06 44803.52  15.63 17.06  0  0 0  0  0  305.054  43078.48  20.67  0 1  264.703  37300.71  18.88  273.103  38338.08  12.3  0  268.572 250.195  37486.74 34802  255217 279.515  3  372.565 388.679  51218.67  0 2  48335  0 2 2 4  0 0 4 0  1  1 1  0 0 0 0  1 1 1 1 1  1 410934 410934  1  1 410934 1 410934  1  0 39770 0 39770  0  0 39770 .0 39770 0 39770 0 39770  0 0  0  1 1 0  0 0  0  0  0  1  0  0  0  0 39770 0 39770  1  0 0  0 0  0 0  0 39770 0 39770  0 0  0 0  0 0  17.32 27.89  0  0  1  0  0  0  0  0  23.92  0  0  0  0  0  0 39770 39770  0  35392.68  1 1  38659 44886.22 50051.87  24.3  0 0  1 1  0  0  0  39770  0  28.39  0 0  36.53  0  0  1  0 0  0 0  0 0  0 39770 39770  0 0  51.74 62.77  0  0  0  0 0  17782.06  22.66  0  0 1  0  1017.79  0 0  0  0  0 39770 0 39770 241590 0 241590 0 241590 0 241590 0 0 241590  0  0 0  1 1  0  53340.47  2003 2004  6  58438 76864  32527 363.428  2005 2006  8 15  93657 116182  1 1  1990  9  3845  2  1991  11  4148  0  4  1059.26  18441.42  19.53  0  0  0  1  0  0  1992  40  4537  0  36  1098.3  19072,63  18.54  0  0  0  1  0  0  1993  218  5473  4  81  982.615  17025.6  16.14  0  0  0  1  0  0  1994  342  7056  0  22  1061.38  18343.34  15.63  0  0  0  1  0  0  0  0  0 1 1  1  76  Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom  1995  391  7349  2  47  1157.44  19947.19  17.06  0  0  0  1  0  0  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  241590  1  0 1996  488  8380  1  38  1220.85  20989.85  20.67  0  0  0  1  0  0 0  1997  667  10721  1  84  1359.44  23312.43  18.88  0  0  0  1  0  0 0  1998  877  14425  0  9  1456.16  24902.18  12.3  0  0  0  1  0  0 0  1999  850  19343  0  26  1502.89  25609.93  17.32  0  0  0  0  0  0 0  2000  947  28551  2  55  1480.53  25142.25  27.89  0  0  0  1  0  0 1  2001  965  34632  2  15  1471.4  24891.24  23.92  0  0  0  1  0  1 1  2002  1256  48335  5  107  1614.7  27219.23  24.3  1  0  0  1  0  1 1  2003  1285  58438  10  208  1862.77  31278.68  28.39  1  0  0  1  0  1  2004  1935  76864  5  191  2199.25  36755.86  36.53  1  0  0  1  0  1 1  2005  2904  93657  7  632  2280.06  37863.45  51.74  1  0  0  1  0  1 1  2006  4225  116182  3  390  2435.7  40237.54  62.77  1  1  0  1  0  1 1  Sources: International Energy Agency, 2009 b European Patent Office, 2008 International Monetary Fund, 2009 d OECD/IEA, 2004 OECD/IEA, 2008 ‘ Central Intelligence Agency, 2008 a  77  Table 8.3: Descriptive Statistics Variable Output (GWh) World Output (GWh) GDP per capita GDP Crude Oil Prices Producer Grants Individual/Company Grants Feed in Tariff Obligation Targets Tax Incentives Green Certificates  Obs 289 289 289 289 289 289 289 289 289 289 289 289  mean 1358.384 31878.59 569.8027 25374.39 25.54529 0.314879 0.2 17993 0.456747 0.1903 11 0.121 107 0.33218 0.17301  sd 3988.772 34032.4 660.6475 10993.25 13.02466 0.465273 0.413599 0.49899 0.393228 0.326818 0.471812 0.378912  78  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0068271/manifest

Comment

Related Items