Master’s thesis Public Administration: Economics and Governance
The effect of market restructuring on consumer and industrial sector electricity prices: an EU panel data analysis between 1995 and 2013
Jasper Schoones Leiden University Student number: s1850563
Date: 09-06-2020 Supervisor: Dr. P.W. van Wijck
Wordcount: 21615
Abstract
Since the creation of the internal market, the European Commission has been committed to an internal energy market of which electricity is an important part. The restructuring process imposed by the EU with the adoption of three electricity market directives in 1996, 2003 and 2009, was intended to take companies out of the public sphere, introduce more competition and to create one internal electricity market (European Parliament and Council, 1996, 2003, & 2009).
In this study, the effect of market restructuring on both consumers and industrial electricity prices is investigated. Because the market restructuring process consists of many components that cannot be summarized and explained in one relationship, I used six components that together represent market restructuring by relying on earlier literature i.e. vertical unbundling, introduction of a wholesale market, openess of the market, degree of public ownership, degree of wholesale market concentration and level of electricity imports. These six individual components led to six separate hypotheses which were tested in a static and dynamic fixed effects regression model, by taking into account delayed policy outcomes and path dependency.
The analysis showed that of the six components only three have a highly significant effect on prices, namely market openess, public ownership and electricity imports. The other components have had no or a very weak effect within a single model. Furthermore, it has been demonstrated that the electricity market is not yet functioning optimally, in particular for consumers. This needs to be addressed by policy makers at EU and national level in the future in order to allow everyone to benefit from increased competition in the market. For future research it is recommended to use a concentration measure such as a Hirsch Herfindahl Index (HHI) instead of using market share because the latter variable was to rough and had to little observations. It is also recommended to carry out the research over a longer period of time so that the development of prices can be estimated better and more accurately.
Page 2 of 66
Table of contents
1. Introduction ... 4
2. Market characteristics of the electricity sector in European counties and the institutional change ... 7
2.1 Technical aspects of the EU electricity market ... 7
2..1.1 Electricity generators ... 7
2.1.2 Transmission and distribution networks ... 9
2.2 Institutional change ... 9
3. Literature review ... 11
3.1 The old status quo ... 11
3.2 The changing status quo – towards the introduction of market restructuring ... 12
4. Theoretical framework ... 14
4.1 Vertical unbundling... 14
4.2 Introduction of a wholesale market ... 16
4.3 Openess of the market ... 17
4.4 Level of public ownership ... 18
4.5 Degree of wholesale market concentration ... 19
4.6 Degree of electricity imports ... 20
4.7 Hypotheses ... 21
5. Research Design ... 22
5.1 Variables ... 22
5.1.1 Dependent variables ... 22
5.1.2. Explanatory variables... 23
5.1.3. Country specific effects ... 24
5.2 Estimation methodology ... 25
5.2.1 Cases ... 26
5.2.2. Time-period ... 27
6. Descriptive Statistics ... 28
6.1. Analysis of descriptive statistics... 28
7. Analysis ... 32
7.1 Consumer electricity prices ... 32
7.1.1 Vertical unbundling ... 34
7.1.2 Liberalized wholesale market ... 35
7.1.3. Market Openess ... 35
7.1.4. Public Ownership ... 36
Page 3 of 66
7.1.6. Electricity imports ... 37
7.2 Industrial electricity prices ... 37
7.2.1. Vertical unbundling ... 39
7.2.2. Liberalized wholesale market ... 39
7.2.3 Market Openess ... 40
7.2.4. Public Ownership ... 40
7.2.5 Wholesale Market Concentration ... 41
7.2.6. Electricity Imports... 41
7.3 Interpretation of the control variables ... 41
7.4 Synthesis ... 42
8. Conclusion ... 44
Bibliography ... 47
Appendices ... 56
APPENDIX 1 Basic Theoretical model for competitive wholesale and energy market (Joskow, 2005, pp. 38) ... 57
APPENDIX 2 Main measures of the three EU energy packages ... 58
APPENDIX 3 Theoretical benefits and costs of ownership unbundling based on work by Baarsma et al. (2007) (Pollit, 2008, p. 706) ... 60
APPENDIX 4: Graphical representation of the dependent and main explanatory variables over years (1995-2013) ... 62
Page 4 of 66
1. Introduction
Already in the eighties, energy markets around the globe started to reform. This change started in Chile in 1987, followed by the United Kingdom in 1989 and has since spread to many other parts of the world (Sioshansi, 2006, Sioshansi &Pfaffenberger, 2006). Reasons for change are very diverse per country and region: political ideology, to improve government finances, and in many cases in the context of overall liberalization of the services sector (Hyland, 2016). Yet the most important rationale behind the energy sector-specific liberalization was the expectation that the introduction of competition resulted in lower energy prices for both companies and consumers creating additional room for economic growth. Also, more innovation in the sector would become stimulated and the sector would change to a more consumer-centred approach, in the end making the energy sector more efficient (Pepermans, 2018). In order to realize this, governments around the world found their salvation in liberalization, which sought to remove or reduce restrictions placed upon a particular sphere of economic activity, like the energy markets (Dunne, 2017). The most important step in energy market liberalisation program was the market restructuring of the electricity sector. The definition of market restructuring is the separation (or unbundling) of vertically-integrated monopolies, privatisation of certain segments of the electricity market and incentive-based regulation for those parts of the industry not generally amenable to competition (Hyland, 2016). These three elements of market restructuring are achieved by vertical disintegration of energy companies (splitting up of activities), diminishing the role of the state as owner in this sector and introducing elements that can reinforce competition such as a liberalised wholesale market.
Also within the European union motivation to reform the sector was mainly driven by factors such as escalating operation costs of power plants, construction costs that were far too high, and retail prices that started to raise for the first time in centuries. Policymakers sought to overcome these growing problems and achieve overall cost reductions in the sector that could be passed on to consumers (Joskow, 2008). The European Commission expresses this goal as follows:
“An integrated EU energy market is the most cost-effective way to ensure secure and affordable energy supplies to EU citizens. Through common energy market rules and cross-border infrastructure, energy can be produced in one EU country and delivered to consumers in another. This keeps prices in check by creating competition and allowing consumers to choose energy suppliers.” (European Commission, 2020)
It was decided it the nineties to proceed with the restructuring of the energy market on EU level. In addition to the actors described above, the expanding of the internal market was an important driver. As Jamasb and Pollit (2005) wrote in their paper electricity market reform in the European Union that this reform was one of the most significant and far-reaching cross-jurisdiction restructuring of the electricity supply industry to this day. In order to achieve this change, the EU introduced three single electricity market directives in 1996 (96/92/EC), 2003 (03/54/EC) and the last in 2009 (09/72/EC). These
Page 5 of 66 three directives moved the member states (and Norway as part of the internal energy market) to a Pan-European policy whereby certain criteria must be met (pollit, 2019). The fist package introduced the idea of competition, the second accelerated the move towards a market-based system (Talus, 2013) and the third attempted to tackle substandard performance from the previous two directives.
However, while lower costs were one of the main reasons to liberalize energy markets, we see in the literature that this common wisdom is heavily disputed. As already indicated, it was believed that vertical unbundling of the industry was necessary to introduce competition. This introduction would lead to more cost-efficient operating of the sector, therefore lowering prices for consumers and industrial buyers (Pollitt et al., 2007). It has also been argued that unbundling is associated with increased operational costs and loss of scope or coordinated economies, making the effect of unbundling ambiguous (Hyland, 2016). In addition, the introduction of privatization is also being questioned as to whether this will bring efficiency benefits and therefore lower costs. Fiorio and Florio (2013) note in their paper that while private ownership may decrease inefficiencies and costs, this will not result in lower prices in the absence of effective regulation due to the inelastic nature of electricity demand. Later Florio (2014) adds that regulatory oversight on these privatised markets is important to protect the consumer from companies that, due to the provision of an important need, have a powerful position. As the electricity sector becomes increasingly subjected to both privatisation and market forces, there is a clear need for regulation to ensure that the benefits from restructuring are passed on to consumers and to protect more vulnerable consumer groups (Florio, 2014; Hyland, 2016). This additional regulatory oversight can itself ensure that the costs of privatization and hence market restructuring will increase, showing that market restructuring can have opposing effects on prices. Therefore the aim of this research is to show what the net effects are of market restructuring on prices for both consumers and industrial buyers by posing the following research question:
What is the effect of market restructuring as initiated by the European single electricity market directives on net consumer and industrial sector electricity prices in the period
between 1995 and 2013?
Net consumer and industrial sector electricity prices is defined by the price without taxes and subsidies so that the effects of tax increases or the use of subsidies are avoided. As already noted, the European restructuring project was the biggest of its sort around the globe to this day, making the EU and interesting region to study. The core focus of this research is to estimate the relevant effects for 21 EU member states and Norway as part of the internal electricity market. The period between 1995 and 2013 is chosen because during this time interval the European energy market restructuring project took place and research data is available.
The focus in this paper relies heavily on the concept of market restructuring, this needs additional explanation. In many sectors where liberalization has taken place or the state has withdrawn, this was initiated by deregulation. Deregulation is defined as: the process of removing government controls or rules from a business or other activity
Page 6 of 66 (Cambridge, 2020). Examples of deregulated markets include financial markets, air industry or railways. The emphasis here is on removing barriers to open up business to all interested actors. This is different in the electricity sector. Here one can speak less of deregulation. After all, the retreating state in the electricity markets has in many cases ensured that more regulation has been added. Therefore the term market restructuring is better suited than deregulation (Hogan, 2002).
This study contributes to the current literature in three ways. In the first place, research so far never investigated the effect of market restructuring on the consumer as well as industrial electricity prices. Second, this study, unlike other studies in this area, takes delayed policy outcomes as a real form of danger. Thirdly, not many studies have investigated the effect of market restructuring more widely than the EU-15 countries. Taking together, this study enhances the understanding of the behaviour of basic economic theory in the energy sector creating scientific relevance which can potentially give policymakers new insights into the working of the energy markets and therefore
create better policies. `
The importance of this research for society is to create awareness about the effects of the initiated market restructuring and opening of the electricity market on the prices that consumers and producers pay for their electricity. Society must realise that this shift and changes in the market will lead to changes in prosperity within a given country and therefore also have a direct effect on welfare of the individual. Something that has received little to no attention at the moment in the current literature.
The structure of this thesis will be as follows. First, there will be a brief introduction to the European electricity sector. This will provide information on the specific market structure of the electricity sector and the adjustments proposed in the three EU energy market directives. Second, a literature review will describe the old status quo and why the introduction of market restructuring has taken place. Thirdly, the theoretical framework will be described. Here, different theories will be presented that interfere with the scientific debate of the effects of market restructuring on electricity prices for both consumers and the industrial sector. The theoretical part will be followed by a description of the methodology in the methodological framework. Here the emphasis will be on conceptualization and operationalization of the variables and a presentation of the methods used and the cases selected to analyse the regressions. Then the analysis will be performed, whereby the focus is on the presentation and interpretation of results based on the hypotheses. I will in the end conclude by answering the research question and discuss the policy implications.
Page 7 of 66
2. Market characteristics of the electricity sector in European counties and the institutional change
To understand further elaboration of this research it is necessary to give some introduction to the European electricity market and the institutional change induced by the EU electricity market directives. The European market is characterised by many important players that need to be taken into account. Moreover, as a result of the various EU internal market directives, this energy landscape has become considerably more complex with the addition of even more actors in this playing field. The focus will be on the technical description in the first part, followed by a description of the changed institutional processes.
2.1 Technical aspects of the EU electricity market
As in every other developed country in the world, the EU electricity market consists of a physical infrastructure to make electricity generation and transportation possible together with an well-functioning economic market. The physical infrastructure for transport of electricity can further be elaborated by a transmission network which necessary for electricity transport over long distances, and distribution, which connects end-users like residential and industrial consumers to the transmission network. The well-functioning economic market is represented by the flow of money between these actors to make the electricity system possible. Figure 1 gives a schematic overview of this electricity system.
Figure 1: Model to organize the electricity sector (Erbach, 2016)
2..1.1 Electricity generators
The generators are taking the first step in order to make electricity delivery to the end-user possible. Whereas many people immediately think of the large electricity factories where coal, gas or hydropower is converted into electricity, we also often talk about consumers who have solar panels on their roofs as a generator. As a result, the range of electricity generators is often much larger than initially thought.
Page 8 of 66 There are two fundamental technological properties that affect electricity markets. First, Supply and demand must always be balanced, otherwise, blackouts will occur. Secondly, the flow of electricity on the transmission and distribution net can’t be controlled. It follows the path of least resistance and therefore consumers always receive electricity from mixed sources (Erbach, 2016).
Generators are divided into groups with the same generation capacity, which is the maximum power they can deliver to the infrastructure. Here we have the variable capacity, that is, capacity which is dependent on external (metrological) factors like sun or wind and therefore can only generate electricity when these external factors are present. Firm capacity refers to generators that can be switched on or off when demand asks to do so (Erbach, 2016). Examples of these firms can be found in table 1.
Another important distinction that needs to be made is the degree of flexibility of a generator with which they can operate. Demand for electricity is not always the same. For example, in night's demand nearly fades away completely but in the evening demand peaks. Also, on hot summers day electricity consumption is much higher due to the use of air-conditioning when compared to a regular day. In line with the first fundamental technological property that demand and supply always need to be balanced this degree of flexibility needs to be built in the electricity grid. Therefore the degree of flexibility of a generator determines whether it is a baseload, shoulder or peak generator.
Baseload generators are those with the lowest amount of flexibility. They deliver a stable amount of electricity to the grid 24/7 and therefore have low cost in the production. Shoulder generators are those with medium flexibility and are activated during day-time when demand slowly increases. Due to the start-up costs involved, the production costs for this type of generator are usually higher than those of a baseload generator. The last type, peak generators deliver electricity during peak hours only. They have a high degree of flexibility and can be started up in a very short period of time to meet the increased demand. Because of this high degree of flexibility, production costs for this type of generator are the highest (Lave, Apt, & Blumsack, 2004).
As shown in table 1, hydroelectric power is one of the most flexible sources and can be switched on or off within seconds. On the other hand, it is not possible for a nuclear reactor to shut down completely, which makes the flexibility very low. As a result, electricity from this type of generator must always be supplied to the grid.
Table 1: characteristics of different types of generators
Type Firm/Variable Flexibility
Coal Firm Medium
Natural Gas Firm High
Biomass Firm Medium
Nuclear Firm Low
Hydro (with a dam) Firm Very high
Solar Variable Very low
Wind Variable Very low
Page 9 of 66 2.1.2 Transmission and distribution networks
Transmission networks are used to transport electricity over long distances. This is done by using high voltages (between 220 kV and 1000 kV) which result in lower transmission losses. The EU transmission grid contains more than 300 000 km of powerlines. The distribution networks provide the last step in the delivering of electricity from the transmission network to the customer. Electricity from small generators, like consumers solar panels, is directly fed into the distribution network to deliver electricity directly to nearby customers (Erbach, 2016).
2.2 Institutional change
Before market restructuring took place in the EU a lot of vertically integrated companies dominated the electricity market in the member states. Vertical integration refers to the integration of the generator, transmission and distribution operator in one company. As will be pointed out in the literature review, this was initially an attractive way of market organisation, however since the 1980s this model no longer proved to be the best fit under current circumstances.
In order to operate the market more efficiently by introducing competition and implement the criteria from Single European Act (1987) concerning the creation of the internal market, the EU used its legislative means to change national electricity markets. It introduced three single electricity market directives in 1996, 2003 and 2009. The basic architecture for restructuring and the development of competitive markets for power involves several key components which can be found in appendix 1 (Joskow, 2005). Based on this architecture, the three electricity directives took measures to create an internal electricity market with competitive elements so that it was in the benefit of the consumer to introduce.
The first directive drafted and signed 1996 had to be implemented before the end of 1999. It arranged that generation would be subjected to free entry by third parties on the wholesale market or had to be competitively tendered and licensed to a single buyer. Access to the transmission and distribution networks (that were still vertically integrated in a lot of countries) should be subjected to regulated or negotiated acces by a third party in the absence of a single buyer. Large (industrial) consumers could freely choose their preferential electricity supplier and those supplier were, therefore, allowed to compete for customers. Also, a first step was taken to unbundle the market by separate the accounts of transmission and distribution activities (also called accounting unbundling). Member states where encouraged to take more competitive enhancing measures, but at the time there was no need to do so.
The second directive in 2003 with implementation date before the end of 2007 forced slow-moving member states to change their market structures and it extended the scope by establishing a wholesale and retail market (Pollit, 2019). In the wholesale market, generators, electricity suppliers and large industrial consumers became participants.In the retail market, the electricity purchased by the electricity suppliers on the wholesale market is sold to the consumer. Because electricity cannot be stored easily, it must be produced at the moment when demand is there. Therefore, electricity
Page 10 of 66 transaction on the wholesale market is focused on the delivery of electricity in the future. This process is guided by different types of contracts ranging from an intra-day market (delivery within a specific moment of a day), a day-ahead market, to a multi-year contract. Most of the electricity is traded through the energy exchange which brings together buyers and sellers in a transparent way (Erbavh, 2016). To control the functioning, stability and security of this market a Transmission Systems Operator (TSO) is initiated. In the 2003 directive, it was decided that the retail market would be opened up to all customers. And that a national regulatory office that was independent from the government was responsible for regulating entree to the network and the charges for transmission and distribution activities. Also, the establishment of more boundary-crossing electricity connections was encouraged (European Parliament and Council, 2003).
One consequence of the first two directives was a massive wave of mergers throughout the EU. Companies such as Vattenfall, RWE and EON expanded considerably beyond their own national borders and became pan-European companies (Leprich, 2005).
What was observed was that electricity companies still had a lot of power in the market. For fear of losing the security of supply, national member states had not taken steps to weaken companies so that competition finally could take place. Therefore the European Commission started a major electricity sector competition inquiry in 2005 which spurred competition enforcement by member states and also the establishment of the third electricity directive in 2009 (European Parliament and Council, 2009; Sadowska, 2011; Pollit, 2019).
This directive enforced unbundling requirements for transmissions activities even further and created a preferential model of ownership unbundling (complete separation of the transmission network from generation activities). Also, the Agency for the Cooperation of Energy Regulators (ACER), a pan-European regulator, was established. It is founded to settled disputes between national regulators and to supervise boundary-crossing competition. As a result of the measures taken by the EU, the electricity landscape has been systematically and permanently altered throughout the European Union.
*A further elaboration of measures taken by the three electricity EU market directives can be found in appendix 2.
Page 11 of 66
3. Literature review
The motivations to change the electricity market structure vary from one country to another, as do the expected outcomes of reforms. One overall goal is identified: to increase the efficiency of the sector by restructure the markets in segments where it is possible to introduce so (Clifton, Comín & Díaz, 2006; Sioshansi, 2006). But before we delve into the market restructuring we will take a step back. What was the status quo before 1980 and under which developments change was induced?
3.1 The old status quo
As Tenenbaum, Lock, and Barker (1992) have described in their paper on electricity privatization and liberalization, there are two different models that were commonplace as a way to organize the electricity market before the liberalization of the market took place. The first is organizing as a state-owned enterprise under government control (SOE’s). The second model is organized as privately owned regulated monopolies whereby the state took the role as regulator. At the time it was thought that these two models could best deal with the typical structure of the electricity market. This typical structure is described by a lot of scholars with the help of two characteristics that are highly related to each other; the existence of natural monopolies and economies of scale. (Holburn &Spiller, 2002; Pepermans, Driesen, Haeseldonckx, Belmans, & D’haeseleer, 2005).
The existence of natural monopolies in the electricity sector was a reason in favour of central organization and ownership. Natural monopolies are those monopolies that exist due to economies of scale and sunk costs. Economies of scale refer to the belief that electricity generation was only profitable for large vertically integrated companies to exploit (Pepermans et al., 2005). It was reasoned that producing units of electricity, in terms of average costs, would become cheaper if more units could be produced. To reach the point where electricity was produced efficiently, a large output was needed. In this case, there was no room to have many suppliers of electricity, each with their own transmission and distribution network. Government policy was aimed at allowing the number of vertically integrated companies that suited the optimal supply area for electricity production, transmission and distribution. This often resulted in only a few electricity companies in a country where no competition was or could be established. Each company had its own supply area which allowed it to keep its average costs low.
Specific or sunk investments are those, once undertaken, whose value in alternative uses is substantially below their investment cost (Holburn & Spiller, 2002). Or in other words cost that when they have been incurred cannot be recovered on exit and therefore having no alternative value. Sunk cost may create barriers by making it essential to undertake these sunk investments in order to enter the market (Niels, Jenkins & Kavanagh, 2016). These type of investments are typically for the transmission and distribution segments of the electricity market. Here, large investments in infrastructure
Page 12 of 66 are needed to transport electricity from generator to consumers’ homes, business and industry. Costly investments that cannot be recurred when exiting the market. This caused that other companies did not want to or couldn’t just join in due to the high degree of uncertainty and capital that was required to build this infrastructure.
Since transmission and distribution were viewed as natural monopolies, the industry as a whole was considered to be a natural monopoly, suggesting that an efficient regulatory framework would be a legal monopoly. On the other hand, a monopoly could also result in dead-weight losses when a profit-maximizing monopolist charges prices that exceed marginal cost. The presumption to introduce an SOE or a privately owned regulated monopoly was that such an enterprise did not maximize profits or was restricted to make profits. Therefore, public ownership or state control should lead to an increase in consumer welfare (Steiner, 2003). It, therefore, could be argued that if this fragile market was left alone, it would operate inefficiently. This is an important assumption for the public interest theory to hold, making it easy to argue that government interventions were simply responses of government to public demands for the rectification of the inefficiency in the operation of the free market (Posner, 1974).
In addition to these economic motives, there were also political aspects like the securing of an independent supply of energy and the use of the market as a way to perform socio-economic policy that determined that governments kept the two models for a long time. Countries did not want to depend on each other in the supply of energy. This was mainly due to the strong economic dependence on electricity to achieve growth and economic development. Therefore, protection of the sector was politically necessary (Talus, 2013). In the area of socio-economic policy, hierarchical coordination of the sector was also important. For example, national industries in Germany and the United Kingdom were supported in this system by burning domestic coal (a lot of domestic labourers worked in the coal/mining sector) or to fix the prices of electricity for national important industries as was done in The Netherlands (Arentsen & Künneke, 1996).
3.2 The changing status quo – towards the introduction of market restructuring
As indicated earlier, this hierarchical way of sector organization proved untenable. Serious problems began to arise in the ’80s when fuel prices, inflation, and interest rates rose quickly which affected the efficiency of the energy market negatively (Sioshansi, 2006). In response to these developments, retail prices rose significantly for the first time in the history of commercial electricity supply. There was variation in performances in the electricity industry, but the structure of the hierarchical market with a lot of government intervention limited opportunities for more efficient suppliers to expand and to price less-efficient suppliers out of the market (Joskow, 2005). Besides the economic argument, technological progress was an important factor in the process of market restructuring of the energy market. Over time the optimal scale of electricity generation, i.e. the scale at which the cost of electricity generation per unit is the lowest (McDowell, Thom, Pastine, Bernanke, & Frank, 2012), was decreased and small-scale electricity generation technologies were developed that could compete in terms of efficiency with
Page 13 of 66 big power stations, which made additional space free on the supply side of the market (Künneke, 1999; Borestein & Bushnell, 2000). At the same time, socioeconomic preferences changed as well creating momentum to start with market restructuring of the electricity sector (Arentsen & Künneke, 1996).
The European commission saw in countries such as the United States and the United Kingdom the electricity sector become increasingly competitive with the introduction of market restructuring, whereby the government took a step back in the electricity market. US electricity prices were on average 35% lower than in Europe and in the UK there was a reduction of 22% in prices (DTI, 2000). These successful stories in other countries made the market and competition approach a realistic alternative, free from scaremongering about the collapse of the energy system in Europe (Talus, 2013).
With the fear of collapsing markets being taken care of, Talus (2013) describes that the raison d’être of the electricity monopolies was gone. Consumers realized that these monopolies charged much higher tariffs than was necessary in order to cover their costs and that surpluses from these higher tariffs were not used for the public purpose, but was eaten up by privileged employees and managers of the electricity utilities or by politicians who provided the necessary support to maintain the electricity monopolies. Moral and political support for energy monopolies faded, resulting in the loss of persuasiveness of the public service through monopoly argument and the opening of market-based regimes (Karan & Kazdagli, 2011; Talus, 2013).
Page 14 of 66
4. Theoretical framework
This paper studies the effect of market restructuring on net prices of electricity for consumers and the industrial sector. As already defined in the introduction, with market restructuring in the electricity sector we mean separation (or unbundling) of vertically-integrated monopolies, privatisation of certain segments of the electricity market and incentive-based regulation for those parts of the industry not generally amenable to competition (Hyland, 2016). Therefore, market restructuring is built op from different components that together represent market restructuring. In this research, I use six components that together represent market restructuring by relying on earlier literature. In figure 1, these relationships are drawn. The indicators and the relationship with prices will be established here.
Figure 1: Schematic representation indicators of market restructuring and relationship on consumer and industrial prices
4.1 Vertical unbundling
As mentioned earlier, unbundling means separating business units that do and do not have the ability to become competitive and make market entry easier for new entrants. The addition vertical in vertical unbundling refers to the unbundling of different business process/parts that used to belong to one company (production, transport and sales)
Page 15 of 66 (Hyland, 2016). The literature speaks of diffuse effects of vertical unbundling on the efficiency of unbundled business units and the development of prices. Pollit (2008) provides a list of theoretical costs and benefits of vertical unbundling. This list is inspired by earlier studies performed by Mulder and Shestalova (2005) and Baarsma et al. (2007) on the Dutch energy market who provided the first “checklist” on the cost and benefits of vertical unbundling. The full table with theoretical relationships can be found in appendix 3. The most important advantages of unbundling provided by the list are improved cost transparency of the stand-alone business units; the possibility to introduce competition in those business units where it is economically viable but due to the integrated structure not possible before; and at last, the incentives of management of the stand-alone business units to improve efficiency and reduce cost, because cross-subsidization (i.e. subsidizing unviable business-units with money earned in viable business-units (Faulhaber, 1975)) is not possible anymore. The most important costs that are associated with vertical unbundling is the occurrence of double marginalization between the separated business units in comparison with the integrated company; giving up the internalization of spill overs between the different units; and the last, coordination possibilities within the integrated company that create efficiency advantages for the integrated company (Arocena, Saal & Coelli, 2009). A more general strand of economical literature outside the electricity sector confirms these cost and benefits of vertical unbundling. These studies are focussed on the measurement of general benefits and costs. They find that integrated economies have a lot of benefits which are lost when countries decide to unbundle utilities sectors. They make the theoretical claim that integrated economies are associated with vertical synergies – like coordination advantages, information sharing, using staff and inputs more efficiently - that are difficult to overcome when they are lost due to vertical unbundling (Triebs, Saal, Arocena, & Kumbhakar, 2016; Gugler, Liebensteiner, & Schmitt, 2017
Despite the theory indicating that there is a positive effect of vertical unbundling on prices, negative effects on efficiency and prices are reported by Gao and Van Biesebroeck (2014). They investigated the effect of sector unbundling in China. Here a reduction of between 5% and 20% in material and labour cost was established in comparison with electricity companies that were exempted from the regulation and therefore not be obliged to unbundle their business.
Fioro and Florio (2009) investigated the effects of vertical unbundling for consumer welfare and prices in the European Union between 1975 and 2008 whereby countries from the EU-15 group were studied. The conclusion was that the consumer welfare as was perceived by EU-citizens themselves, measured by Eurobarometer data between 2000 and 2006, decreased. Thereby, empirical evidence of OECD market indicators between 1975 and 2007 resulted in mixed-evidence on the price effects. It should be noted that unbundling, in the form of ownership unbundling, mainly took place after the third EU directive (2009), so it did not include the effects of unbundling in its entirety. Another conclusion can be found in the study of Arocena, Saal & Coelli (2009) whereby the effect of vertical integration and unbundling in the US electricity market was investigated. They concluded that vertical integration resulted in 13,5% cost savings,
Page 16 of 66 because of the benefits described above. Fabrizio, Rose & Wolfram (2007) concluded that decoupling of the US energy sector resulted in a reduction of costs due to competition and efficiency of 7%, which is substantively below the earlier indicated 13,5% to offset the loss of benefits due to vertical integration.
All together they confirm the literature that in a western world setting the positive effects of vertical integration that are lost are not offset by the benefits of vertical unbundling. I, therefore, expect that H1: vertical unbundling leads to an increase in prices for both consumers and the industrial sector.
4.2 Introduction of a wholesale market
Establishing a wholesale market in the context of the energy sector means that there is a market where the generator of electricity can offer his product (electricity) and where retailers can buy the product and sell it to their customers (Erbach, 2016). Before market restructuring was implemented in the European energy sector, prices were regulated by national regulators which determined what prices, according to the cost, integrated companies were allowed to charge their customers (Shioshansi, 2006). After market restructuring, when wholesale markets for electricity where implemented, a trading floor was introduced were bilateral contracts (contracts freely negotiated between seller and buyer), as well as spot contracts (which acts as a clearinghouse), were closed. When competition on the market is established, prices of electricity are not determined by the system operator (national regulator) anymore. Instead, a bid-based system is used where each individual generators submit its own price bids every day to the system operator to meet the expected customer demand for the following day (the day-ahead market). When all bids are made, the system operator selects the generators whose bid are lowest (Cain & Lesser, 2007). This wholesale market ensures that under the right competitive pressure generators offer their prices as low as possible on this wholesale market. Because without low prices they never will be chosen by the system's operator to deliver electricity to the net which causes losses because generator plants come to a standstill. This mechanism, therefore, transforms these low energy bids by the generator into lower prices for both consumers and producers end users.
However, the literature finds mixed-results for this causal claim. Steiner (2003) shows in here analyses of the energy market in 19 OECD countries between 1987 and 1999 that there is no evidence of lower electricity prices due to the introduction of a wholesale market. In contrary, she claims that the introduction increased prices for both consumers and industry. She also shows that the ratio between industrial and consumer prices is increasing which means that due to this regulation prices are growing apart (Steiner, 2003). Nagayama (2007) and Erdogu (2011) replicate the findings of Steiner (2003), however, in a different time period and in other non-OECD countries. The data on which Steiner's (2003) findings are based date back to a time when many, particularly European countries, had not yet introduced a liberalised wholesale market. This may mean that the estimated effect does not correspond to the current data where many more countries have liberalized their whole sale market.
Page 17 of 66 In contrast to the above authors and in line with theoretical mechanism Hyland (2016) finds a significant lowering effect of the introduction of a wholesale market on electricity prices for industrial buyers. This effect on industrial prices may have been achieved because Hyland (2006) uses data that is complete and has a focus on European economies. Hyland (2016) used the same variable for the wholesale market as Erdogu (2011) but looked at EU countries and the industry prices specifically with a time period that, in contrast to Steiner (2003) was affected by EU reforms (2001-2011). The fact that Erdogu and Nagayama present opposite findings can be explained by the use of developing countries in their datasets. These countries which are not comparable to western economies can possibly create biased outcomes causing the real effect to be blurred .
With the theory in mind and the evidence presented by Hyland (2016) I expect that: H2: the introduction of a wholesale market leads to an decrease in consumer and industrial electricity prices.
4.3 Openess of the market
One of the big steps in the market restructuring process of the European Union was to open the retail market for both consumers and the industrial sector. With the opening of the retail market, consumers could choose freely which retailer they opt for delivering energy and it could provide social benefits, when the competition was well established. Defeuilley (2009)noted in his paper the following societal benefits. (1) Reducing market imperfections due to the cost reduction behaviour of entrepreneurs entering the retail market and by products that where historically niche markets where extra-normal profit was made. In a competitive market, those consumers who become engaged in the market process will force suppliers to become more efficient with the benefits being passed on to all consumers, even those who are not actively engaged (NAO, 2008). (2) Stimulating the alertness of consumers with the existence and potential merits of alternative offers. They provide accurate price information to consumers helping consumers to become active participants in the market process. (3) Stimulating competition in the generation process by inciting that efficient investments choice are made and that entry on the generation markets will be easier (Defeuilley, 2009). Which in the end will result that lower prices on the wholesale established by the retail market are passed on to final consumers.
However, empirical evidence so far is not sure about the existence of this relationship yet. Olsen, Jophnsen and Lewis (2006) provide evidence that retail competition so far has not resulted in the theoretical plausible cost reduction. To explain why they focus on consumer behaviour. Very few consumers are participating actively in the retail market to search for lower electricity prices. It appears that between 2013 and 2015, a few years after liberalization, less than 5% of households, on average, switched their electricity contract in the roughly 20 EU countries that had liberalized their electricity markets within the previous 10 years (Schleich, Faure, & Gassmann, 2018, p. 2). These low switching rates are mainly on the account of price information which is not correctly passed on to customers causing difficulties to arise when consumers want to
Page 18 of 66 evaluate the “best-fit” supplier on the energy market. They do not know what they in the end pay for the energy supplied by the intransparent energy bill, and they also do not know their real-time energy consumption (Joskow & Tirole, 2006). Information needed to make an informed decision and to actively participate as a consumer in the retail market. Thereby, ACER, the EU Agency for Cooperation of Energy Regulators, identified the perceived (insufficient) monetary gain, the lack of trust in new suppliers, the perceived complexity of the switching process, as well as satisfaction with the current supplier as the most influential ‘preventers’ of consumer switching behaviour (ACER, 2015).
This mechanism is not working for the industry sector. This is explained by the usually high saving potential of industrial consumers when switched to another supplier (ACER, 2015). The industrial sector uses more delivered energy than any other end-use sector, consuming about 54% of the world’s total delivered energy (EIA, 2016). Because they are large buyers of energy, a small saving on price per unit can lead to large savings overall for the industry. Thereby, the industry sector has a higher price sensitivity compared to consumers (ACER, 2015). This is because the price of their products is determined by the input of energy. Electricity usually represents a significant proportion of total energy costs for the industry with percentages standing between 48% and 54% of the total (Moreno, García-Álvarez, Ramos, & Fernández-Vázquez, 2014, p. 816). When this input reduces production costs this will lead to competitive advantages (Moreno et al., 2014). To attain this advantage the industry sector will decide to switch more quickly, even if the savings are initially small. This mechanism ensured that switching rates in 2015 were 2.5 times higher (12.5%) compared to consumer switching rates (ACER, 2015). Despite previous theoretical substantiation Su (2014) reports positive effects of retail market opening on consumer electricity prices in the U.S. market. However, this effect disappears, in the long run, resulting in the conclusion that overall retail competition did not deliver significant lower electricity prices to consumers over time. Due to the non-competitive behavior of consumers which is represented by the low switching rates of consumers between different energy suppliers and the theoretical opposite effect for industry behaviour I expect that H3: opening of the market leads to an electricity prices decrease for the industrial sector, but no effect is expected on consumer prices.
4.4 Level of public ownership
The impact of private versus public ownership in the utility sector is a heavily discussed topic in the literature. On the issue of the welfare impact of privatisation, a paper written by Sappington and Stiglitz (1987) established the primer conditions for indifference between public and private ownership. With their analysis, they showed that under different information structures a policy-maker who is able to write complete contracts achieves the same welfare outcomes for the economy in comparison with a private owner. However, Schmidt (1996) indicates that a government is unable to gather enough information that is needed to achieve this Sappington-Stiglitz condition.
Page 19 of 66 Therefore the key point here that matter is information asymmetries and the way they interact with welfare outcomes. Roland (2008) writes in his book privatization: success and failures that there are two major ideas in relation to information asymmetries under public and private ownership:
“The first one is that government ownership reduces access to firm information for outsiders, namely, stock market participants. This, in turn, reduces opportunities for monitoring the managers of firms. Thus, monitoring in government firms is poorer, and incentives for efficiency are potentially lower. The second idea emphasizes the difference in the information that government has about private and public firms. The main idea here is that the government is less informed about private firms than about public firms because ownership of the latter gives government privileged access to the accounts of the firm. Because of this difference in information, privatization lowers the government’s ability to extract rents from firms.” (Roland, 2008, pp. 13-14)
Economical privatization theory points to the first idea as the most plausible and take in to account cost efficiency and quality of products produced. Hart, Shleifer and Vishny (1997) consider two types of investments: those to reduce costs and those to improve quality or innovate. When assets or contracts are publicly owned, the civil servant has relatively weak incentives to make one of these investments. This is because the civil servant is not the owner of the company and hence gets his pay check regardless of the company’s return. On the other hand, private managers or owners have much stronger incentives because they are directly paid by those returns on investments. The weak incentives of civil servants with respect to both cost reduction and quality innovation underlie the basic case for the superiority of private ownership, a case that has been confirmed by the variety of empirical studies and general observations (Shleifer, 1998, pp. 137-138; Armstrong & Sappington, 2006). It, therefore, doesn’t make sense to have publicly owned companies when private companies can deliver the same, or even better, quality of products at lower costs. Yes, with some exceptions in the economy, the empirical evidence on privatization processes by governments points to both cost reductions and improvements in quality in privately owned relative to publicly owned companies (Ehrlich, Gallais-Hamonno, Liu, & Lutter, 1994). I, therefore, expect that: H4: an increase in Public ownership leads to an increase in prices for both consumers and the industrial sector .
4.5 Degree of wholesale market concentration
Utility services like electricity generation, transmission and distribution are historically seen as essential for the society. Before 1980 the dominant view was that public ownership could provide a safe harbour and, when needed, subsidized it in order to provide these necessary services to society (Niels et al., 2016). These national electricity companies were monopolies, each serving its own geographical area. When countries
Page 20 of 66 decided to liberalise and restructure the energy market one of the core focusses was to do something about these monopolies and especially the market dominance resulted from this monopoly position. Market dominance in the electricity generation market means that a company has a significant degree of market power, meaning that a dominant company is large and powerful enough to have a substantial effect on the market and therefore is able to offer prices above the cost of production, causing consumer welfare to deteriorate (Niels et al., 2016). The EU electricity directives, therefore, took several entry regulation measures, like a permit system for entering the generation market, prohibiting of a single buyer, and introduced regulatory organisations like a system’s operator to cope with these monopolies and the risks of having market dominance (European Parliament and Council, 1996, 2003).
According to economic theory, having market power is not a bad thing necessarily. The reason for this stems from the Pareto improvement which is derived from to presence of a dominant firm. This is explained by the Schumpeter-visits-Chicago argumentation. This argumentation takes a positive view of dominant firms in a market. It states that they are generally good for consumers, create a lot of jobs in the sector and also innovate and make the best use of scale economies (Canoy, Rey & Van Damme, 2004). The innovation here is key. Innovation is a strong driver of long-term economic growth (Vatiero, 2006). The literature found an inverted U-shape relationship between competition and innovation: “oligopolistic markets tend to be more innovative than either monopoly or highly competitive markets. Too much competition will dissipate the post-innovation profits and hence reduce the rewards from trying to escape intense competition by innovation” (Niels et al., 2016, p. 147). Too little competition causes a wait-and-see attitude of the monopolist (Aghion, Bloom, Blundell, Griffith, & Howitt, 2005).
The main driver to invest in innovative technologies are the potential cost reductions that can be achieved. Empirical literature supports this relation between innovation and cost reductions by stressing the importance of innovation by achieving and maintaining high learning rates. Learning form production process creates in the end better understanding of these processes resulting in the technologies needed to become more efficient (MacGillivray, Jeffrey, Winskel, & Bryden, 2014). When competition and
regulatory institutions are available, this increase in efficiency will, in the end, result in lower prices for consumer and producers. I, therefore, expect that :H5: wholesale market concentration leads to an increase in prices for both consumers and the industrial sector.
4.6 Degree of electricity imports
Another variable that is often cited in the scientific literature, despite the possible relation with the dependent variables, is the degree of electricity imports (Fiorio and Florio 2013; Hyland, 2016). The three electricity directives issued by the EU also aimed to increase electricity trade between member states. Before the interconnection of electricity networks was pushed by the EU, large price differences could be observed between the different member states. To give an example, in the UK, fossil fuel-fired generators were mainly used for electricity. Sweden has a different way of meeting its energy demand.
Page 21 of 66 They choose to get their energy from hydropower plants. According to Hyland (2016), energy prices are generally lower in countries using hydro or nuclear sources. This is also reflected in the prices. Where in 2000 the price of electricity in the UK was around € 20 per 100 kWh, for Sweden it was only € 16 per 100 kWh (Eurostat, 2003). A substantial price difference. By enabling interconnectivity between networks, these large price differences between the Member States can come to an end. It will then become possible to import cheap hydropower into a market where otherwise only expensive fossil energy is generated, which has, in the end, positive effects for consumers in the EU (DECC, 2013).
Mulder and Giesbertz (2008) note that there are more benefits involving consumer welfare, benefits from enhanced competition may result from the interconnection. Competition in some member states, like in the Dutch power market, is stagnating due to the limited number of competitors. Therefore In those hours where demand for electricity is high, prices increase due to this limited competition. If the available interconnection capacity increases, prompting other providers to enter the wholesale market, the current players will be exposed to other more competitors entering the market. As a result, the wholesale price (particularly during peak and super-peak hours) will according to Mulder and Giesbertz (2008) decrease. When competition in the end-user market is high, the price benefit will be largely passed on to the consumer.
Giving the above theoretical mechanism and empirical evidence I, therefore, expect that: H6: electricity imports leads to an price decrease for both consumer and industrial electricity prices.
4.7 Hypotheses:
Based on the above theories and empirical evidence, various expectations are presented for this research in the form of a testable hypothesis. Below in table 1 is a brief summary of the discussed hypotheses given.
Table 1: Overview of hypotheses
Hypothesis Variable Consumer Price Industrial Price
H1 Vertical Unbundling Pc Pi
H2 Wholesale Market Pc Pi
H3 Market Openess Pc Pi
H4 Public Ownership Pc Pi
H5 Wholesale Market Concentration Pc Pi
Page 22 of 66
5. Research Design
This chapter will consider the methodology to assess the effects of market restructuring on electricity prices for both consumers and the industrial sector. I start with the operationalisation of the different variables used in the model and give some explanation on the origin of the data underlying these variables. Here I also present the control variables needed to control for country-specific effects. This is followed up by the methodology used to estimate the models. Further, a specification of cases used in the model is given.
5.1 Variables
In the introduction to this study, it was already indicated that the researchers in this field contradict each other and that established literature is disputed. The fact that researchers do not agree with each other can be explained on the basis of the shaped variables that are included in statistical models. It is therefore important to properly define the variables used and to make use of variables that have already proven themselves in other studies.
Some variables are transformed into natural logarithms. This has been done for several reasons. First of all, this technique is applied often within the research field of electricity markets as can be seen in the recent study by Hyland (2016). In addition, it transforms a number of variables from a very skewed distribution to a more or less normal distribution, as further explained in the next chapter. As a third point, it makes the interpretation of coefficients somewhat easier because the effect variables with a very large scale such as GDP per capita can now be expressed as a percentage change. This also applies to the other variables, which allows a uniform interpretation in percentages across the whole model which can be seen in the analysis.
In the hypotheses just presented, electricity prices for both consumers and the industrial sector are not affected by the indicators of market restructuring. They have therefore been identified as dependent variables.
5.1.1 Dependent variables
- Consumer electricity price. This is the average first half yearly price paid without levies or taxes in euro’s per kWh, price level of 2005 corrected, in a specific country and in a specific year by household consumers. The data is extracted from the Eurostat Energy Database (Eurostat, 2020a). Where small data gaps arise in observations, linear interpolation is used to estimate the missing value. Where the data gap is too large - larger than two missing observations - then these are left empty. In line with earlier research, I transform the observation to logarithms so that interpretation of coefficients in the estimation model will become more clear. - Industry electricity price. This is the average first half year price paid without levies or taxes in euro’s per kWh in a specific country and in a specific year by industrial consumers with an annual consumption of 2000 MWh. This specific annual consumption pattern is chosen due to data availability and use in other empirical literature (Erdogu, 2011; Hyland, 2016). The data is extracted from the Eurostat
Page 23 of 66 Energy Database (Eurostat, 2020a). In accordance with the former variable, I transform these observations into a logarithm scale.
Six indicators of market restructuring have been presented in the hypotheses that may have an effect on electricity prices. These six indicators have been translated into the following explanatory variables.
5.1.2. Explanatory variables
- Vertical unbundling. This variable indicates the extent to which the activities of an electricity company are separated in a country. A distinction is made between the generation market, the transmission/distribution market and the retail market. This variable is displayed on a scale of 0 to 6, where 6 represents fully separate business activities and 0 represents fully integrated business activities. This data comes from the Product Market regulation (PMR) indicator of the OECD (OECD, 2020).
- Introduction of a wholesale market. An important step that many countries took in order to restructure the market was to establish a wholesale market for energy. This variable is added as a dummy where a 1 indicates the existence of an energy wholesale market for a country in a given year. The data for this variable comes from the dataset of Erdogu (2011) and will be completed with data from Hyland (2016) and individual country reports of the European Commission.
- Openness of the market. This variable is constructed from the PMR database of the OECD. This variable benchmarks the degree of market openness in a national market on a given year. The scale ranges from 0 to 6 whereby 6 represents a fully open retail market that is open for competition and 0 represent one national monopoly. The limitation to use this constructed OECD indicator is that it also includes information on general market openness. The results, therefore, can be a bit blurred but it gives a first valuable insight into the direction of the effect market openess has on electricity prices (OECD, 2020).
- Level of public ownership. In addition to market liberalization being an important step in market restructuring, privatization of former national monopolies is also an important step to further restructure the market, making the introduction of competition possible. This variable is subtracted from the OECD PMR database. This scale ranges again from 0 to 6 whereby 0 represents a market that is fully in private ownership. 6 represents a market that is fully dominated by public ownership (OECD, 2020).
- Degree of wholesale market concentration. In order to see if competition really takes place and has its influence (besides the policy measures to make competition possible) I add the market share of the largest generator. Market share is measured as percentage generated by the main generators of the total generated electricity in a national market in a given year. It would be better if a concentration measure like the Herfindahl-Hirschman index would be used or to take the market share of the three largest generators in a country. Due to data limitations, this is unfortunately not possible. Therefore it is accepted in the literature to use the
Page 24 of 66 market share of the largest generator (Hyland, 2016). Data is subtracted from Eurostat Energy Database (Eurostat, 2020c).
- Degree of electricity imports. The EU policy of market restructuring in the electricity sector is also aimed at increasing electricity imports from surrounding member states. Therefore, the model also includes information of net electricity imports given as a percentage (% imports - % exports) in each country per given year as is mentioned in the study by Fiorio and Florio (2013) and Hyland (2016). Data is subtracted from the International Energy Statistics Database compiled by the United States Energy Information Administration (EIA, 2020).
In order to estimate the correct effect of market restructuring, it is necessary to check for country-specific effects that may result in a lower energy price by nature.
5.1.3. Country specific effects
- Real per-capita GDP. Above mentioned explanatory variables are not the main price determinators of electricity. Therefore we should look at the demand side of the electricity market. The most used method in the literature to control for demand is to add a variable that indicates the real per-capita GDP. Data for this variable comes from the World Bank National Accounts Dataset and is expressed in 2005 US dollars (World Bank, 2020a).
- Cost of electricity generation. To control for cost advantages in electricity generation due to national resources, I add the price of gas for generators in a member state into the estimation model. Because no relevant dataset exists, I use industrial gas prices as a proxy The study by Hyland (2016) finds that the industrial gas price and the price of gas used in power generation has a correlation of 0.96, indicating that is suitable as a precise proxy. Data of gas prices for the industrial sector can be found in the Eurostat Energy Database (Eurostat, 2020b). To make interpretation of the coefficients easier I decided to transform the latter two variables – Real per-capita GDP and cost of electricity generation – into a logarithm.
- Hydro share. The energy mix that is used in electricity generation has also an important impact on prices paid by consumers and the industrial sector. To control for this country-specific advantage we add the share of energy in percentage that is produced by hydro-electricity generators. In the literature, it is suggested that a country with high percentages of hydro generation in its total generation has a comparative advantage (Hyland 2016). Hydro generation is a relatively cheap way to produce energy constraining the energy price. I subtract data for this variable from the EIA International Energy Statistics Database (EIA, 2020).
- Nuclear Share. In line with the above reasoning, we include nuclear share as a percentage of total production in a given country and year. I subtract data for this variable from the EIA International Energy Statistics Database (EIA, 2020).
- Urbanisation. In the paper written by (Gutiérrez, 2003) it is suggested that countries with a high urbanisation measure have per definition lower electricity
Page 25 of 66 prices for consumer and industrial. This variable is constructed from the World Bank Development Indicators (World Bank, 2020b).
- Renewable energy policies. During the time period analyst, the EU did invest in significant effort to increase the production of renewable energy sources into the energy mix. These policies could possibly influence market prices for consumers and industry by increasing the production prices of electricity (Mulder & Scholtens, 2013). Additional investment in infrastructure and generation facilities should be made which increases the cost price of production. This increase is then passed on to the end-user. I use data from the International Energy Agency (IEA/IRENA) Global Renewable Energy Policies and Measures Database which gives information on policies in place to stimulate renewable energy in IEA countries (IEA, 2020). I use policies that are classified by IEA as economic instruments. These policies include measures like Financial/fiscal incentives and direct investments. I create a dummy variable that gets a 1 when such an instrument is in place.
5.2 Estimation methodology
The literature represents different statistical methods to test the overall impact of market restructuring as is pursued by the three different EU energy packages (Pollitt, 2019). These methods range from a social-cost benefit analysis as represented by Jones et al. (1990) to performance metric regression on panel data as is represented by Steiner (2003) in her OECD-paper about performances in the electricity supply industry. Because this research topic is focused on change in different country's over a time period of 20 years, the panel data method will be the most suited. In addition, the previously presented variables all have the structure of panel data, which makes this methodology a logical choice.
The regression model used by Steiner (2003) can be described by the following model: 𝑦𝑖𝑡 = 𝑐 + 𝑓𝑖+ 𝑍′𝛽 + 𝑅′𝛿 + 𝜀
𝑖𝑡
Here we can describe consumer and industrial sector price measure yit, as a function of
country-specific effects 𝒇𝒊, a set of variables which influence performance independently of regulation Zit, and set of regulatory indicators Rit, for each country i and year t (Steiner
2003). Here, the random effects, 𝒇𝒊, account for unobservable (and, therefore, unmeasurable), time-persistent differences across countries and therefore unobserved heterogeneity. A form that is not included in standard regressions, but is important to check for in the models to be estimated. Examples of unobserved heterogeneity are the cultural differences between countries. By compensating for these differences with fixed effects in the regression analysis, the results become more valid. The observed differences in economic and technological conditions that characterize countries over time represent
Zit. 𝜺𝒊𝒕 is the models’ error term which incorporates the residuals produced by the
statistical method of analysing. This model will be estimated by using panel data techniques across countries and years and will be tested with a regression analysis in
