Practical implications and attitudes towards the Implementation of EU State Aid Rules in municipalities and provinces in The Netherlands

INSTITUTE of Public Administration/Instituut Bestuurskunde

Practical implications and attitudes towards the Implementation of EU State Aid Rules in municipalities and provinces in The Netherlands

Gheorghe Mogildea (2075180)

International and European Governance Track, Public Administration Thesis Supervisor: Bernard Steunenberg, Chair in Public Administration

i

This academic work is the result of two years of enjoyable and at times hard work, struggles, successes and motivation. In this journey I have been supported by my parents to whom I am very thankful for their love and advices.

Also, my gratitude goes to my supervisor prof. dr. Bernard Steunenberg who in the process of writing this academic work was very enthusiastic about the topic of state aid and provided the needed help and

motivation. Mister Steunenberg, you became like my academic father and you have my endless respect. The feedback given by my professor has enriched this work but also has made me a better personality. At critical moments your guiding helped in drafting the text, finding participants and most importantly getting me fascinated about the capstone. During the capstone the timely and precious planning made by the supervisor always allowed me to know what the next step is when doing my thesis and go ahead with confidence.

I would like to also share my thoughts with the participants of this research project. Without your input this final product would not be possible. The diverse perspective, experience and knowledge which I encountered is immense and your professional attitude and conduct is inspiring. This is the foundation on which my thesis is based.

Lastly, I want to express my gratitude to someone who has become a very good friend, Hans Groeneveld. His contribution to my life is very important because he is the person whom has influenced and

encouraged me to study the Master of Public Administration. His positivity and critical thinking were of great help for this thesis too. Due to his dedication coming every week to my place to teach me Dutch also has given me the chance to study Dutch sources and do the interviews with my participants. His help is invaluable. To all of you who have been part of this process I want to say that I am very grateful that I met such people in my life, thank you.

This thesis addresses the implementation of European state aid rules at the local and regional government level in The Netherlands. There are three discourses discovered in Dutch municipalities and provinces. The discourses have been discovered using the method of Q-methodology that allows identifying discourses that are built around statements that are similar in nature. These discourses are about when subsidies and loans are allowed, in which sectors the municipalities and provinces choose to subsidize etc. Also, the thesis discusses different factors such as the relationship between expertise and the application of rules, and the relationship between the European level (i.e. the EU Commission) and the subnational governments (regional and local). Public officials think that the EU Commission is still the main actor to decide and enforce the rules but at the same time they indicate that the regional and local governments should have their fair share of responsibility. This means that they are in search for a balance of power in which the Dutch government can decide themselves what to do with their local subsidies and loans without having to report every time to the European Commission. Most importantly public officials would like to see a change in the attitude towards this policy at the national level. On the other hand, in one of the discourses they still report that the rules on subsidies and loans as well as non-financial help are too complex for their colleagues.

There are two types of specialists that work with state aid in The Netherlands. Lawyers are people who check on the legality of the rules in accordance to the European criteria on state aid. Implementers are part of the public administration apparatus, specialists who actually provide and monitor the use of subsidies, check reports, organize the tender procedure etc. Often these two groups have a different outlook on which aspects are important in the implementation, for example the transparency of the rules and the transparency of the actual subsidies that are given.

In addition, this academic work makes a distinction between financial and non-financial types of help. This is important because practical implications for the work of civil servants are different. The manner in which subsidies and loans are given is also different depending on the size of the municipality, the project and the goals that the municipality or province wants to achieve.

An important aspect that this study aimed to find is the attitude of public officials towards their work. They view this policy area as very complicated but meriting attention for the future. Some of their colleagues are also afraid to get themselves busy with state aid precisely because it is complex. The rules on state aid are complex but known, while providing subsidies is seen as more acceptable in some cases

iii

than in others. Municipalities and provinces make use of the de-minimis amount of money that they are allowed to finance companies without having to notify to the Commission. This is an efficient instrument to boost the economy. In smaller municipalities and provinces there seems to be a shortage of expertise which also has an implication on the size of state aid departments and the amount of work that can be done. Civil servants are aware of the possibilities that subsidies and loans offer and would like the population to have a positive outlook on their daily work.

Table of Contents

1.Introduction ... 1

1.1. Principles of state aid ... 1

1.2. Forms of financial and non-financial aid ... 4

2.State aid issues in the Dutch and European literature and online sources ... 7

2.1. Theoretical ideas on state aid ... 7

2.2 State aid control in the Netherlands... 8

2.3. Sectors of State Aid in The Netherlands ... 11

2.4. Public Officials attitudes on state aid in The Netherlands and other EU MS ... 12

3.Research Design and Methodology ... 14

3.1. Presentation Q-methodology ... 14

3.2.Q-Methodology Use in this study ... 16

3.3.Q-set and the P-set ... 19

3.4 Risks of Q-methodology ... 22

4.Presenting Discourses found in PQM and Interviews ... 24

4.1 Q-sort Discourses ... 24

A. Legal Rules on State Aid ... 24

B. Awareness and Discovery of State aid ... 28

C. Levels of Responsibility of the Actors... 30

D. Statements of General Consensus ... 33

Results P-set and Interview Findings ... 35

5.Discourse Analysis ... 39

Discourse 1: Legal Rules on State aid ... 39

Discourse 2: Awareness of the Implementation, Discovery and Attitude ... 41

v

6.Conclusion ... 48

6.2. Limitations ... 52

7. APA References ... 53

8. Appendices ... 58

Appendix II. U- Shaped Table Distribution ... 62

1.Introduction

1.1. Principles of state aid

The Netherlands has been doing rather good in many policy areas and has been put forward as an example for the respect of the rule of law (The Dutch Government, 2019). State aid has been formally introduced as a European policy since the Treaty of Rome in 1957. (Treaty of Rome). Rules have been massively created in this area in the 1970’s and 1980’s. Previously the topic of state aid has been within the sovereign power of the nation-state in order to protect national strategic industries from competition and make them flourish. This protective attitude has changed with the development of the internal market and the transfer of supranational powers to the Commission. The Commission began to closely monitor rules when a new package of assistance was approved in 2005 and 2011 that permitted providing subsidies and loans if the public compensation principle was at stake in the Member states of the Community (William Fry Foundation, 2015, p. 4). This supranational institution, therefore extended its power in enforcing and monitoring state aid and the creation of rules in this domain. Lower level governments have the task to implement Commission directives in this policy as they wish.

State aid is defined as “any advantage granted by public authorities through state resources on a selective basis to any organisations that could potentially distort competition and trade in the European Union” (United Kingdom Government Website, 2012). There are two aspects or so-called procedures where the EU Commission that is the official creator and enforcer of the EU state aid law has most powers in. These are the notification procedure and the infringement procedure. (Adriaanse & den Ouden 2008, p. 4) The European Commission is, however, not the only actor that decides on state aid rules. This thesis addresses the implementation at the local level gap with the aim of discovering the existing common and different discourses in a sample of Dutch municipalities and provinces. The central question that this Master

thesis aims to answer is: What are the discourses on state aid implementation and attitudes and their practical implications at the local level in Dutch municipalities and provinces”. The thesis will

address a few more specific questions: (1) what aspects do discourses relate to, (2) what is the attitude towards state aid of the people who are directly involved in this area but also (3) what are the successes and problems that Dutch municipalities face in their everyday work with state aid. Moreover, it should establish (4) how exactly is financial help given to companies within municipalities and provinces (5) and

2

different sectors where grants and subsidies are given in the Netherlands as well as different characteristics of municipalities in comparison to provinces that impact their work.

Discovering the discourses will also act as an evaluation of the Dutch implementation framework. While general state aid rules are important to get to know the topic in depth and to inform the public about current issues the focus of this Master thesis is twofold. In the Dutch scholarly literature, there is a description of the application of European rules in the Netherlands. Dutch scholars have less information about the implementation of state aid national and European rules at the local level. They mostly write about the national level and that the regional and local governments have the freedom to implement the rules, but do not go beyond how do they do that. This information is important in order to see how local public officials from municipalities and provinces react to European rules, how well do they understand them and what is their attitude on the existing policy. Mapping the discourses will also show what the challenges that municipalities and provinces face in their daily work.

The national governments also have sections on state aid in their national legislations. In the Netherlands subsidy rules have got a section under the General Administrative Legal Act (GALA), but this section according to analysts such as Adriaanse, Ouden and others is incomplete. (Adriaanse & den Ouden, 2009, p.8.) At regional or local level, the basis for offering subsidies can be a statutory law based on the General Act.

However, arguably the section on state aid in this law was too small and did not cover many issues. Here it is meant that the rules addressed only subsidies and no other financial instruments such as loans or the government partly holding shares in a company. According to Adriaanse often such laws would contain only minimum requirements. By the implementation, this would lead to a situation when it was not clear which rule applies in which context and some situations did not have rules. (Adriaanse, 2007). The point here is that local and regional authorities decide how to apply Commission rules according to their specific local context (Europa Decentraal, 2019). Although they are responsible for the correct application of the rules and healthy competition on the market. In practice, subnational governments issue their own acts that allow them to help companies that for example are close to market failure.

Famous law-making cases of course contributed to the mediatization of state aid and arguably, the EU Commission now has more of a say in enforcing a decision on national governments because of increased public attention to this topic on European level. One such case has been the lawsuit of the Commission started against Apple and the Irish government (News Sky; Irish Times, 2018). Tax arrangements with

governments have the potential to become state aid if the tax evasion gives one company a clear advantage. Therefore, tax evasion is a potential measure of state aid and not different in nature.

However, these are not the only instances when one can talk about giving a selective advantage. Examples include giving tender contracts on favourable terms, providing subsidies under incomplete applications or due to personal connections.

When the Commission suspects a case of unlawful state aid, then it can start a preliminary notification procedure. The preliminary notification procedure asks from the state in question to provide information on a case that would allow the Commission to make a decision whether it indeed was unlawful state aid or not. The main principles here are the principles of the advantage on the selective basis and the distortion of competition on the market.

The goal of these sections is to give the reader an overview when one can talk about state aid and when not. This explanation will help to understand the topic better but also the discourses and the language that public officials use to describe state aid. Without this prior knowledge I argue it is quite difficult to distinguish state aid from a simple measure of providing financial help (subsidy, loans or whatsoever). It is also to make the reader aware about the possibilities that municipalities have in order to boost economy and that they are not just subsidies and loans.

In addition to that, there are many certain exemptions when subnational governments allow subsidies and loans. These exemptions are also controlled by the Commission and refer to information and communication technologies, the cars industry, sports, social activities, and other industries that touch on the social life. Companies that apply for such projects indeed have to meet the criteria written in the application for state aid. The argument here is that if a measure contributes to enhancing the life of the public then it might not qualify as state aid. The conditions that a project must meet therefore have to be explicitly stated. Otherwise, the line between state aid and a compatible measure for instance a subsidy becomes somehow vague.

In the context of applying for financial help, it is crucial that the application and the way in which the finances are spent is transparent. This aspect merits further attention. Governments have an obligation to make the amounts transparent under the new state aid rules adopted in 2016 (Europa Decentraal, 2019) Before giving out a subsidy in accordance with the notification procedure the member state should obtain

4

There seems to be a difference between the way how lawyers and public administrators think about state aid in the Netherlands. For public officials it is more difficult to understand and apply the rules. “State aid is relevant for public administrators because it contributes to the nucleus of our work, the delivery of public value said one of the participants in this research (Participant 1, 2018).” This is why the implementation of this policy at this moment in The Netherlands gets much attention from the local governments, unlike in the past. However, there is difference in the forms of financial aid and each municipality needs a specialist that understands the difference between for example a grant and a loan. “For politicians and public administrators this is still an unknown, strange world”. (Participant 1, 2018). The knowledge on the topic is still underdeveloped. If you finance something with public money, you have to be responsible and explain every time beforehand what will be done with this money. There needs to be transparency and responsibility”, said the person interviewed (Participant 1, 2018).

1.2. Forms of financial and non-financial aid

In this subsection, I will discuss different forms of help that government can give in order to make a

distinction and a clear divide between financial and non-financial help. This information is useful for the technicalities of practical implementation and to know which rule applies when giving different forms of help from the government. This is the actual practical implication of each type of state aid and in later chapters the reader will see how civil servants look at these rules. Consequently, this will help with understanding what do municipalities and provinces are allowed or not do, when they are dealing with different types of financial and non-financial help and also how are they distinguished in discourses. This description will also introduce the reader deeper into the topic of understanding what possibilities municipalities and provinces have.

Grants- A grant is a sum of money or a gift given by the government for a particular purpose (Business

Dictionary, 2019). Grants are non-reimbursable which means that they do not have to be given back unless they have been given illegally in comparison for example with loans. (Business Dictionary, 2019) This type of governmental help is usually given for public purposes. In the context of the EU grants are often given to the Common Agricultural Policy (CAP) because governments believe in the support of their farmers and want to keep them competitive with their European counterparts.

Grants might also be given when there are natural disasters or to save national industries. Former Commissioner on competition policy Almunia said in one of his speeches that it is perfectly fine to save a

company that is in need (Almunia, 2010). However, grants are not given easily; there should be always an application process within the municipality or province. The receival of a grant is usually conditional on the achievement of a public goal and sometimes on an own contribution of the grantee to the project.

Loans- “A loan is money, property or other material goods given to another party in exchange for future

repayment of the loan value amount, along with interest or other finance charges. A loan may be for a specific, one-time amount or can be available as an open-ended line of credit up to a specified limit or ceiling amount.”. (Investopedia, 2019). A loan is, therefore, very different from a grant. The Commission gives much attention to governmental loans in the member states. The Commission even updated the notice on state aid to give governments clearer guidelines on how to implement EU provisions. A large section of these provisions contains information on granting loans. The loans have to meet specific criteria. One of these criteria is that the borrower is not in financial difficulty. (EUR-Lex, 2016). This is a clear criterion because the borrower has to be able to repay the loan but not all criteria are that clear-cut. The duration of granting, a loan and the period of its effects are issues that the Commission and national state aid authorities are very careful. Many state aid issues arise because the duration of receiving a loan is too long. (E.g. Alitalia, Starbucks, IKEA.)

The loan should further not cover more than 80 per cent of the financial activity for which the loan was asked. Thus, the company asking for the loan also has to make a financial contribution. Again, when the activity concerns the provision of a service of economic interest the loan might be provided in full amount. This rule also has practical implications for public officials. Imagine that the lender gets the loan in full then the lender is not very motivated to look at the risks hindering the repayment and minimize the risks because they did not suffer any cost. Therefore, the Commission has introduced the risk assessment instrument in the notice on state aid. The risk assessment is carried by banks that lend the money to the creditor.

In order to receive the money, the company has to pass a market investor test that will establish what equivalent the company has to pay back to the state. The rule here is that the amount provided to the company should not exceed the market prices that are paid in that industry.

Tax reliefs- “Tax relief is any program or entity that reduces the amount of tax owed by an individual or a

6

the industry. When the tax relief takes place, it gives the recipient company an advantage over the other companies because the company can keep more profits in the internal budget, which gives them the advantage of better salaries, better technology and equipment, stronger shares etc. By putting the other players in a disadvantageous position, it distorts competition because the other firms do not have the same resources anymore. However, public officials in this study report that this has not happened within their municipalities and provinces but nonetheless this has happened in the Netherlands.

Government holdings- The government can also hold assets in a private company operating in a business

environment. If the company is a hundred percent government owned such as in telecommunications industries, then it is not state aid because it is viewed as a service of general economic interest for the public. However, when the government owns only part of the shares of a private entity, they are prone to use the state budget in order to give the company a selective advantage as has happened in the case of Alitalia. Even more complicated is when the government sets up a parent company in order to move assets or control assets of one entity or a corporation. Via the parent company, the government can have influence on the management and the decision making of the company. State aid in this case is considered when an investment of fresh capital is not applicable and would result in giving advantages. In municipalities and provinces companies might be created or asked to merge in order to achieve a public goal (Participant 7; Participant 10, 2019). This is also a practical implication of this type of non-financial help.

Preferential treatment

Preferential treatment can occur in all cases described above but it is not limited to them. Preferential tax treatment is very often the case. However, preferential treatment is visible through preferential contracts or personal ties with the governmental officials in high or low ranks. Preferential contracts can be signed during the tender procedure or there might be special clauses that the company will get paid more for a certain service than its competitors might. Preferential treatment is also considered state aid. A case has been when a company received a very large amount of land for a lower price than the normal Dutch market but this measure was not deemed state aid. According to the Dutch state the fact that the company received the land did not distort competition because the other companies participating in the auction were on equal terms. The market thus could operate on the normal conditions and the price of the land did not affect the conduct of business.

2.State aid issues in the Dutch and European literature and online sources

2.1. Theoretical ideas on state aid

The main principle of European integration is the free market. According to the values of the Founding Fathers of the European Community there should be no trade barriers between countries of the European Union (Schimmelfennig, Leuffen, & Rittberger, 2015, p. 6). Non-tariff trade barriers are also inadmissible and state aid is considered one of them. As said earlier, state aid is any form of aid financial or not that distorts competition and gives one firm a comparative advantage over another in an economic sector and threatens the free market (Sciskalova & Munster, 2014). States, however, justify public aid to achieve universal access to public services such as housing or simply do not let a national industry for example the car industry lose ground in the face of international competitors (Rubini, 2009). The difficulty when giving a subsidy is establishing who the receiver is and motivating this to other companies who are the competitors. (Participant 1, 2018). To qualify as state aid an intervention must meet the following conditions set out by the European Commission, which is the body responsible for European wide state aid control. The conditions that make an intervention state aid are:

 “1. The state aid measure is given to a company or organization that does an economic activity

 2. There has been an intervention by the State or through State resources which can take a variety of forms (e.g. grants, interest and tax reliefs, guarantees, government holdings of all or part of a company, or providing goods and services on preferential terms, etc.);

 3.The intervention gives the recipient an advantage on a selective basis, for example to specific companies or industry sectors, or to companies located in specific regions.

 4.Competition has been or may be distorted;

 5.The intervention is likely to affect trade between Member States” (European Commission,

2019)

Generally, if something is seen as offering an advantage in the economic field, then in accordance with EU regulations that measure is not allowed and must be repaid (Tasan-Kok, Groetelaers, Haffner, Heijden van der, Korthals Altes, 2014). State aid measures are never that simple and often involve conflicting interests. EU law may allow certain forms of help for example paying less tax in areas like cohesion

8

policy, infrastructure etc. if those are parts of an EU Commission project. The difficulty in these regulations lies in the fact that general conditions allowing state aid are decided by the national authorities and sometimes can be very broad and go in contrast with the EU legislation.

Before talking about situations when receiving money from the government is allowed an important distinction has to be made. State aid is not the same as a government subsidy. Therefore, it is somehow controversial that in the Dutch scholarly literature it is only subsidies and loans that are discussed when there are many other instances of financial and non-financial help. A government subsidy means just giving an entity (public or private) a sum of money to run its activities. In contrast state aid is aimed at reducing the entrepreneur’s costs and have an effect on its profits. They are often very similar because a subsidy that results in having more profit than initially planned becomes state aid. For a subsidy to become state aid it is important the money is distributed by a public body (government, school, hospital) in order to qualify as preferential treatment to another undertaking. It is sufficient if the government exercises decision-making power and via the transaction reduces the costs or increases the profits of a firm.

For example, a criterion in the housing market is that the value of a grant or subsidy does not exceed the actual value of the property on the market. (Tasan-Kok et al, 2013, p. 5) Moreover, if there is an open public auction through which the contract is given to a housing agency than the grant does not constitute state aid.

Other instances when boosting the economy is allowed are cases of preservation and stimulation of national culture such as public broadcasting and film production. Member states usually broadly define this principle of national culture that derives from the treaties in order to finance public broadcasting and film production. Funding is allowed if it fulfills public objectives and the Commission cannot intervene in national jurisdiction. Here again the definition of public objectives is left at the local level. (Donders, 2011).

2.2 State aid control in the Netherlands

The general principles of state aid are laid down in the national Administrative law. The Netherlands has not met the last deadline from the Commission on harmonizing state aid rules (Nouhuys, 2018, p.49). Late implementation was one of the reasons why the Dutch system of rules was not completely harmonized and why the Dutch government had received a deadline from the Commission to adjust its procedures (Nouhuys, 2018). However, another problem is aligning the European rules with the national legislation

that are often in contradiction. The national law often goes in contradiction with state aid rules that make implementation really difficult as argued by one of the interviewees.

Several times The Netherlands has been brought to the European Court of Justice (ECJ) for giving government funding that was not allowed in various sectors of the national economy. In particular the Commission has identified that The Netherlands has problems with the recovery of the funding offered through state aid. (Adriaanse & den Ouden, 2008, p. 9). The recovery of state aid can be done only through a domestic law procedure and that has complicated the situation for the country because such a procedure was absent.

The recovery is further complicated by the difficulty of the rules and the technical calculations of how much of the state aid has to be paid back due to the inclusion of the profit in the sum that was initially received (Participant, 2, 2018). Precisely because it is a challenging task to calculate the effect of state, aid, it is sometimes impossible to recover the exact amount of state aid. It is expected that this concerns other areas of implementation as well. The specific problems concern mostly the recovery of the interest payments.

The situation is complicated by the fact that the country has different levels of government and rules apply equally to all of them while decision-making power is different for all three. The national court may refuse the order to recover state aid if it deems that impossible. (Adriaanse, 2007). The national government still makes some amounts transparent when they submit their budgets to the EU Commission or a notification from the Commission arrives (Adriaanse, 2007).

Another issue is the reduced transparency of these public expenses. One way of improving the current affairs is to conduct an audit carried out by a professional accountant keeping all state aid records for a state agency.

However, European rules from the Commission oblige the Netherlands to notify the amounts of state aid to the EU institutions. The records are kept by different state aid coordinators that work in different ministries. The Dutch government created a central point of coordination for state aid that is part of the Ministry of Interior that monitors all written reports on this topic from the regional local and national level. The purpose of such a mechanism of notification is for the Commission to identify if everything goes well with a measure and report instances of unlawful state aid if they have been discovered. The requirement from the Commission was to harmonise the Dutch national law with that of other member

10

In establishing its state aid rules, the country always gets some room for maneuvering from the Commission (Adriaanse, 2007).

In terms of the implementation, it is left to the Commission to qualify a measure as state aid and as a consequence court judges issue a national decision for recovery of the money it is suspected that the manner in which this process happens depends on the form of state aid.

In the Netherlands subsidies are mostly given through the provisions of the General Administrative law. Indeed, often it happens that subsidies are given by the government using a private contract agreement. Under Dutch law there is no rule that prohibits such contract. It is these agreements that are difficult to govern and penalize. The Dutch government enters into public as well as private contracts. The power to grant subsidies lies within the municipality or provincial government but must have some legal basis. This legal basis contains only minimum requirements for an entity to receive a grant. For example, if the grant is for a short period of time there is no need to have a legal basis. A multitude of aid applications falls under these loose conditions, which makes it difficult to monitor the use of the money.

The decision to allow a subsidy is with the administrative body, the decision maker in our case provinces and municipalities to fix the amount of the grant. A filed application that meets all the conditions for a grant is conditional for receiving the amount of money if the application contains a description of activities and the obligations of the beneficiary in relation to the giver, thus the government. Once the obligations have been fulfilled, from that moment the receiver has the right for the subsidy including the obligation of the state to name the fixed sum of money provided

The notification procedure serves the purpose to give the Commission time to study the case of state aid in detail and see if the case complies with the provisions of the Treaty. Further, this is done to make sure that a measure of aid cannot become operational before the Commission issued its opinion. The Dutch state does not have an authority that makes sure the notification procedure is complied with (a sort of checkup authority here is meant). (Adriaanse & den Ouden, 2008, p. 6). I argue this is needed in order to qualitatively assess if a measure has effectively become state aid and make sure the government payments are stopped before the case is resolved completely. This change would also allow for better reporting to the Commission and could reduce the number of cases where the Netherlands is getting an infringement procedure. As noted before the ministries concerned deal with state aid and then notify the Dutch Permanent Representation in Brussels. This is a major lack of the Dutch state aid system. When several

ministries are involved in granting state aid then the Ministry of Economic Affairs has the task of conferring a joint meeting. (Adriaanse, 2007).

Grounds, on which an application could be refused, indeed have an optional character and again a wide margin of discretion. However, in practice more and more subsidy acts have to be approved by the Commission before given by the regional or local government (den Ouden, 2009).

By a decision the Commission informs the state to recover the aid from the beneficiary public body or firm, including the rent acquired on it. Often this is not needed since after the notification procedure the company already starts paying back state aid. Only the national government and not regional or local government are held responsible if state aid rules are not followed.

2.3. Economic sectors of subsides in The Netherlands

Often grants are provided to the medical system companies in social welfare states. The reasoning for this is to provide better health care. In July 2005 the European Commission sent the Dutch government a letter concerning competition and the non-admissibility of certain forms of aid for housing associations (Tasan-Kok, et. al, 2013). The forms of financial aid listed in the letter were: loan guarantees from the Dutch Social Housing Fund, systematic aid to the social housing sector, exemption from corporation tax and cheap loans from other private banks such as the Bank of Dutch Municipalities (BNG).

The association of Dutch property investors has complained to the European Commission that the Dutch municipalities offer land to social housing companies at an unfair price. The question here is whether this action in fact contributed to the provision of affordable social housing and how the European state aid rules affect the social housing market. This is a matter of debate and using this argument the state may be able to provide subsidies to social housing companies. The issue has been ultimately resolved because the Commission regarded the number of houses built not having a significant effect on the price of the land.

Currently (since January 2019) the Commission is studying a tax payment scheme between two Nike franchises that paid too little tax too each other, fact that basically allowed the two franchises to keep too much of the profit to themselves. The Commission is concerned about the fact that one of the companies had zero employees and the other one has more than one thousand employees and suffers marketing and sales costs, therefore one entity should have paid more tax than the other. (Loyens & Loeff, 2019).

Worrying is also the fact that five out of eight ministries which are giving money from the budget do not have internal rules on how to deal with state aid. Based on previous research it was recommended that

12

such procedures should be issued and subsequently used by all ministries. Initiatives have been taken by the Coordination Point for State aid from the Ministry of Interior and Kingdom Relations to inform the regional and local governments about the EU rules on state aid.

In exceptional circumstances the concerned ministers might give advice to the regional and local governments on how to deal with state aid. This is stipulated in the General Administrative law that in cases where a court decision is unclear the concerned minister has the final say on a decision. (den Ouden, 2009). Having internal rules could improve identifying cases that are suspect to the notification procedure but also serve as a point of reference for public officials who when advice is needed could approach the ministry.

The existence of additional powers from the national government that would financially penalize the regional and local government for giving unlawful subsidies is likely to reduce instances of unlawful state aid. According to the literature it is still a question if there are incentives for private companies not to accept subsidies when they suspect it can lead to an advantage over the other competitors. In case such incentives are absent the state or subnational governments should create this regulation.

2.4. Public Officials attitudes on state aid in The Netherlands and other EU MS

The regulations on state aid will gain prominence in light of the upcoming Brexit. It is expected that in the UK after Brexit the state will give more subsidies to rescue national companies that are in financial danger. One of the officials of the Consumer Market Agency (CMA) has said that “they are not pitching for extra work”. From this statement one can conclude a rather passive attitude of UK authorities towards subsidies. By contrast in France the government is very active in giving grants to key sectors of the economy such as the viniculture (wine) and agriculture as well as money to invest in research and development. The EU Commission also has a positive attitude on state aid and motivates this by saying that grants are good to save companies all over the EU. The change in attitude can be described from cautious to active in monitoring state aid activities. Officials from the Commission see state-aid policy to be in their interest. (Aydin, 2009).

Pieter de Zwaan motivates in the introduction of his paper about expertise and attitudes on state aid rules in the Netherlands that it is the difficulty of these rules that makes the implementation at the decentralized level rather complicated (de Zwaan, 2018, p. 5). According to him the quality of implementation depends on the size of the municipality and the level of expertise in it. (de Zwaan, 2018). Municipalities often use the de-minimis rule of providing an acceptable amount of state aid (200,000 Euros) in order to avoid an

infringement procedure. This allows them to say that a certain measure does not constitute state aid because they are not obliged to notify de-minimis (Participant 1, 2018). The costs of controlling whether a measure is state aid, including the political costs, have an influence on how municipalities respond to it (de Zwaan. 2018, p. 11). It may be that in some cases the municipalities will be more careful than in others because of reputational damage. (de Zwaan, 2018, p.11) The initial expectation of this thesis based on the findings of de Zwaan is that the public officials in the Netherlands have a neutral attitude towards providing state aid that is rather careful. Therefore, a significant account of the attitudes of public officials in The Netherlands could reveal trends of factors that influence the decision-making of those who make the regulations on this topic in The Netherlands but also on a supranational level.

Most European citizens are not aware themselves of rules on state aid and think that information on state aid is difficult to find. The Eurobarometer has revealed that many public officials working for the Commission have not seen any information about a company or public body receiving a subsidy. The source of information from which people hear about state aid is the TV. Information on this topic is not really spread in public newspapers or magazines but in the specialized ones (Tijdschrift voor Staatsteun,2019).

More than half of the respondents to the survey organized by the Commission report that they do not feel like information about this sector is made publicly available. From this we can conclude that those responsible on state aid at least on the European level are not very transparent. The answers dependent on the level of education and the position of the surveyed people. From this one can conclude that people with a higher level of education might be better experts on state aid (Aydin, 2009). In fact, all participants in this research are highly educated which confirms this expectation. Young people are more likely to be concerned about state aid than older people. However, it should be noted that information about public attitudes on state aid is not widely spread in information sources, from which this report gains even more relevance.

14

3.Research Design and Methodology

3.1. Presentation Q-methodology

During this thesis project the Q-methodology technique was used to collect data. Q-methodology is a tool used mostly in psychology and political science (Dryzek & Berejikian, 1993, p.48). This method is useful when uncovering the opinion of participants around a subject. Q-methodology is distinguished from a survey in more distinct ways. Q-methodology, despite being an old method started enjoying popularity recently because it allows grouping intersubjective points of view together (de Graaf & Exel van, 2005, p. 1). Therefore, the discourse of the participant becomes crucial. This characteristic leaves little room to the researcher themselves to decide the categories given to the participants. Scholars such as Dryzek and Berejikian argue that giving the participants the opportunity to decide on what is important for them introduces more subjectivity and validity to the research because it lets subjects speak for themselves about aspects of their work and categorize them individually. If findings are replicable across participants, then the research gains more validity.

This research method allows to group people in similar or dissimilar clusters of meaning rather than presenting them with a reality of the researcher beforehand. In this sense Q-methodology is a discourse-centered approach because it is interested in what people have to say rather than mechanisms. Therefore, using this tool proves very fruitful when uncovering the discourses around a topic. Because the topic of this dissertation is the discourses around state aid at the local level in Dutch municipalities and provinces, this method is ideal in discovering what professionals have to say about this topic especially when studying this rather vaguely explored area of state aid practical implications and attitudes in The Netherlands.

A discourse can be anything said around a theme that is widely shared in society or by experts (de Graaf & Exel van, 2005). This method gains validity when discourses are replicable across participants because they tend to stand very close to viewpoints in society. Thus, the issue of the number of participants is not of interest to researchers but rather that the participants come from a diversity of backgrounds (Dryzek & Berejikian, J. 1993, p.50). Selection of participants should be different on age, gender, race, nationality, function etc. However, the researcher should minimize any bias towards selecting a participant because of his or her viewpoint.

In terms of the steps when conducting Q-methodology, the first step is done by the researcher and namely the collection of statements on the topic. This step can be considered a pitfall of this method because it is

up to the researcher to select the statements that the participant will be grouping in the scoring scale. A statement can be anything found in the literature on state aid for instance, an opinion of an expert, an example from a newspaper, something said in a verbal conversation etc. However, the researcher should be careful to literally copy or record a statement as it has been said or written without changing the meaning because this could have consequences at a later stage.

This process results in what is termed a concourse of statements, or more simply all statements found on a topic. The concourse usually has around 250-300 statements, but it depends on the popularity of the topic in the discipline (Dryzek & Berejikian, 1993, p. 51). A concourse about democracy might be larger than a concourse on network governance.

After a concourse has been established, it has to be reduced to a plausible set of statements that are believed to be most important in the field and for the purpose of the study. This step is done in order to establish a manageable set of statements to be presented to participants that should cover most issues around a topic be representative of the views around state aid. Also, this selection ensures that we do not overload our discourses with statements that make it harder to distinguish between discourses Statements have to be different in nature and selection takes place in two dimensions. The first dimension on which statements are selected is the political discourse dimension. In this dimension one has statements referring to the ontology, agency, the motivations of actors and natural general statements that can be placed in more than one category. Ontology statements refer to existing facts about the topic in my case state aid implementation. Agency are the actors involved in the implementation of the policy, thus public officials working on state aid in Dutch municipalities and provinces. Motivations are the choices they make in their daily work and the intrinsic and extrinsic motivations for those choices. Natural statements are concerned with general aspects of policy implementation. Further there is one more dimension on which statements are selected. The second dimension is about the type of claim that has been made and the strength of that claim (Dryzek & Berejikian, 1993, p. 51).

For this, statements are categorized into definitive, designative, evaluative and advocative. Definitive statements for instance refer to the ontology of a phenomenon or its core as described in the literature. In the case of state aid this is a common affirmation such as “state aid distorts competition”. An advocative statement then refers to the opinion of the subject participant about aspects of state aid. An example can be “state aid rules in the Netherlands are underdeveloped”. This is an advocative statement because it makes a claim. Using this approach, the concourse is brought down to the most important statements in

16

each category. This is still done by the researcher. Therefore, there are more moments in the research process when the researcher is central, but this creates balance in selecting the Q-set that the participant later forms a discourse around.

The Q-set should contain a diverse selection of statements. This selection is ideally 64 statements chosen from each category of definitive, designative, evaluative and advocative statements. During the selection the researcher should focus on the clarity of the statements and that the statements represent diverse views. The statements do not have to say the same things. There is always a tradeoff that the project undergoes between the time and availability of participants and the number of statements in a Q-set. On the one hand the more statements we have the more causal relationships and discourses that are different to each other we can discover, but one has to be careful that more statements might also mean things that are practically the same. This might result in distorting the meaning of an important factor or missing out on other explanations and therefore, arriving to the wrong conclusions. Therefore, the aim is not always to look for a large number of statements because this is also time consuming.

The respondent then, is asked to categorize the resulting set of statements according to his preference that is usually shown on a scale from “mostly disagree to mostly agree” (de Graaf & Exel, van 2005, p. 4). This is termed the Q-set sorting. The established practice is that the Q-set has not more than 60 to 80 statements and takes around one hour to complete.

3.2.Q-Methodology Use in this study

For this dissertation study the first methodological step was to search for literature sources that present the situation with the implementation of state aid in the Netherlands. Online search engines such as Google Scholar have brought up in the list of results the European wide state aid rules and the project went in this direction. This direction was chosen because the EU law is the primary source of rules in the Member States and the watchdog of the implementation of state aid rules in the country.

Those statements about the EU rules however were selected in relation to the national legislation. Therefore, it is more logical to first define what state aid is, study the European rules on state aid and then study the implementation of these rules in the Netherlands. This dissertation followed this path. This choice was also justified by the fact that during desk research for scholarly articles a limited number of sources discussed the public administration around this topic in The Netherlands. Keywords used for the search were: “state aid Netherlands”, implementation state aid Netherlands”, “public officials and state aid”, public official attitudes on state aid” municipality state aid”, municipality subsidies Netherlands” etc.

The limited availability of articles can be explained by the novelty of this topic in Dutch and English-speaking literature. Therefore, my choice was to acquire general information on European rules in order to understand better what state aid is and the criteria that constitute state aid. The five criteria outlined by the European Union provided a better focus for this study and an understanding of what is a measure of state aid. Then, I studied through the Dutch sources that discuss various aspects and problems of state aid policy. The Dutch sources were selected based on the topic that they discuss for instance implementation of the subsidies, general Dutch rules on state, relationship between the Dutch and the European framework. During the search more recent information has been prioritized. Some academic and internet sources had been known beforehand such as the website of Europa Decentraal or the academic Journal Tijdschrift voor Staatsteun.

In addition to that, the desk research has provided general ontological statements for the Q-concourse such as “” state aid distorts competition” and other natural statements. This general information was found using online sources such as the EU Commission website, the legal database EUR-Lex or the European wide journal State Aid Quarterly. This media selection was chosen because the European Commission, EUR-Lex and the State Aid Quarterly as well as other European Union websites retain the most information on this topic because of the expertise that these organizations have and the responsibility that they carry in this policy area.

After this, the specific situation with the implementation at the national level is analyzed using academic articles, newspapers, websites of policy bureaus such as Europa Decentraal and other governmental internet sources. These sources provided a good starting picture of what issues there are in the implementation of state aid in The Netherlands and how are these issues tackled. Some of the issues discussed are the transparency of state aid amounts in the Netherlands, requirements for granting a subsidy and the responsible body etc.

When collecting statements, I tried to focus on the different aspects of this policy such as the legality of it, the opinion of public officials on it, the lawsuits and critical cases of state aid when the Netherlands has been notified or an infringement procedure was opened by the Commission. A general remark about the collection of statements in the articles is that priority was given to statements that compare the national level with the subnational one. Also, statements were chosen based on whether they talk about an actual practical problem (e.g. the process of giving subsidies) rather than other sentences that do not need clarification. This was done in order to subsequently find out the opinion of participants towards the

18

issues that are of most interest for this thesis. In the selection statements that hint to the practical work of public officials were prioritized because they allow to evaluate what happens in municipalities (for example a statement such as “refusal conditions refer to EU state aid rules) allows us to see whether local officials follow EU wide rules when refusing a subsidy or loan. This statement tells us something about a practical implication and therefore fits perfectly the selection. All statements thus have been selected based on an interesting or controversial rule but also on the attitude level since this is a very important influencer of the policy.

More specific information about state aid implementation in the Netherlands has been extracted from two academic articles of Paul Adriaanse and Willemien den Ouden and one study about the attitudes and expertise of public officials written by Pieter de Zwaan. From the article of de Zwaan I selected the statement that has identified that there is a shortage of expertise in this policy area because that could be a serious obstacle in the quality of the implementation at the local level. When this niche was sufficiently explored, I thought about looking at different sectors in which subsidies are given in the Netherlands and gather statements about the implementation in these sectors for example the social housing sector. This is regarded useful when presenting participants with different sectors of their work and see what their meaning is about financial help in different sectors. An important aspect when selecting the statement was to be clear, simple and diverse in the state aid discourse. Moreover, the criterion was for the statements to be different to each other and focus on both positive and negative sides of state aid policy.

A series of statements about sectors of state aid had been included in the concourse, because there are specific cases when subsidies are allowed to protect national culture, the public values of a society or enhance the quality of life in a city (Almunia, 2010) Establishing sectors in which subsidies are given in the municipalities and provinces gives a very good picture of the implementation that is why statements such as “in the housing market the amount of money given should not exceed the value of the property” have been included in the Q-set The public is not always informed that state is per se not illegal if it meets certain conditions and it can bring benefits to companies and the government because the state budget would grow. Despite of that, the desk research on available literature did not reveal much about the attitude of public officials towards this topic at the national or local level.

A known, respected scholarly source in the Netherlands is De Tijdschrijft voor Staatsteun (The Journal for State-Aid) (2018). A remark about the quality of this journal is that the language and the explanation of terms about state aid is very technical and difficult to comprehend for someone who is not an expert. Also,

the state aid rules laid down in the journal do not necessarily refer to the implementation at the local level but mostly to the translation of European rules and their importance for the Dutch economy and when one is confronted with state aid.

Another point to take into account when discussing state aid in The Netherlands is the limited availability of literature on this topic that discusses implementation of state aid by municipalities. One known study is that of Pieter de Zwaan conducted at Radboud University in Nijmegen. There is still a perception that state aid may lead to something negative (i.e. a judiciary procedure started or an obligation to re-pay the costs of the aid given) but the counselling bureau Europa Decentraal lets the municipalities see that state aid can offer many positive possibilities (Europa Decentraal, 2018). The focus of the literature and policy making according to this bureau should be on what kind of rules are there and why clients should respect the rules. The Netherlands arguably is still lagging on establishing this link. At the same time the questions from the municipalities are getting more and more complex

3.3.Q-set and the P-set

Around 60 people were contacted by phone or by email and asked to participate in the research. The response rate is about 20 percent, resulting in 15 participants actively taking part in my research project. Ideally, I wanted to have between 15 to 20 participants from a diverse set of municipalities which would likely allow me to discover more discourses. Also, I aimed for lawyers as well as other public officials directly dealing with financial help offered to companies. This choice was made in order to see if there is a difference in how these two groups think about and act on state aid. Moreover, I selected municipalities that are different in terms of population size, size of the department working on state aid, experience with this topic, geographical location etc. In terms of sampling the goal of this research was to gather a diverse set of participants from different units (P-set). The units in this case would be different municipalities and provinces around the Netherlands. The P-set is diverse in terms of gender, background, size of municipality, age, level of government (municipality or province), and years of expertise. As pointed out by Pieter de Zwaan the level of expertise in a municipality can influence greatly how they respond to state-aid requests. (de Zwaan, 2018, p. 7). Therefore, the units (the municipalities and provinces) were sampled in terms of their size (large, small, medium), the size of the state-aid department, the area of the country and the budget available for state aid. This is done to ensure variation in potential explanatory variables. The logic behind this selection is that the municipalities that have fewer resources on state aid such as employees or budget are expected to perform different than larger municipalities. Also due to

20

established practices (e.g. the tender procedure) the manner in which state-aid policy is implemented might be different in the North than in the East of the country. Naturally such characteristics will influence the discourse in a municipality. The participants fitted in these characteristics because they allowed for discovering enough discourses, had different functions, age groups etc.

The selection of participants happened using random sampling and convenience sampling focusing on the availability and expertise of subjects. Another issue could be that participants did not foresee what the Q-sorting will entail and therefore were more passive in accepting an offer to participate.

The starting point when looking for participants was to approach an organization that works with municipalities at the local level and gather general information. Then contacts were gathered using an expert network of public officials and academics. Also, municipalities and provinces were contacted by phone using a general number and asking the municipality client service for a person in the state-aid department. It is interesting to note that many small municipalities did not have a department or section on state aid and directed me to larger cities. Further, the P-set has been complemented with snowball sampling. Snowball sampling is a technique in which current participants refer you to other potential respondents during the research.

Following the development of the original concourse consisting of 144 statements, the councourse had to be reduced to 48 statements to be distributed and sorted to the participants of this research. The method used by Dryzek and Berejikian was employed for designing the final Q-set. Four statements were selected for each cell of the matrix using the dimensions of Dryzek and Berejikian. When selecting statements, the criteria was that they either speak about the ontology (origin of state aid), agency (how state aid is being implemented), motivations of state aid policy and the public officials themselves, or natural statements concerning other aspects of state aid. Priority is given to statements that are simple, clear to understand and are diverse in their characteristics (Steunenberg, B., 2018). At the same time, they had to be either definitive, designative, advocative or natural. Statements that met these criteria were drawn from the concourse and given a cell in the matrix. Statements about the characteristics of state aid that give a judgement fall into the category of ontology evaluative, for example such a sentence is “state aid is legal”. The selection criteria for the Q-set is that even though a statement falls into the same category with the other three from the same matrix cell, they had to cover different issues. Priority is given to statements that are shorter because of the ease of understanding.

Another tradeoff hinted at before is between the number of statements and the time availability of each participant. For this study 48 statements were selected using the matrix of Dryzek and Berejikian from the total concourse of 144. This means that many important statements might have been omitted because I had to focus on only 48 statements due to time constraints and because the total concourse on state aid is not very large. Also, for participants it is easier to form clearer discourses with 48 statements than they would do with 144. For the researcher this means getting more participants to take part in the research which ensures diversity and reliability. This process is done in order to reduce the Q-set and make it more manageable and understandable for the participant while keeping the diversity of statements and categories. A benefit of shorter Q-sets is that they also allow for conducting the Q-sorting with more participants which is beneficial for the method because each participant is treated as a potential source for a factor. Conducting more Q-sorts will in turn lead to better validity of the results across the wider population of municipalities.

Immediately after the Q-sorting participants were asked to do a short clarification interview to explain better the scoring of statements and elements of the discourse that are important for them, interesting or unclear. These clarifications allow to form a better picture of the discourse. Doing additional interviews is an important aspect of Q-methodology because they can give insights into new issues or complement the knowledge on existent issues. (de Graaf & Exel van, 2005, p.10). The interview with Europa Decentraal gave insights that the size of the municipality and the availability of experts working on state aid greatly influences the ability to implement policies in the municipalities. Thus, the interview was a good addition to fill in the gap about the views of public officials working in this area.

Because the context of this research is Dutch municipalities and provinces and the people working there are Dutch, subsequently the statements had to be translated into Dutch as well. This was done for the 48 statements in the Q-set that participants engaged directly with. Statements from the preparatory interviews and Dutch literature were in Dutch while statements from online sources were mostly in English. They were translated by the researcher using Google translate and one editor whose mother language is Dutch. It was crucial not to distort the meaning of statements when translating in order to allow for a genuine, valid subjective interpretation.

After all participants have done the Q-sorting the data has been inputted in the PQM Method program that is able to cluster statements together and give each statement a coefficient score in relation to a factor.

22

Factors are decided by the researcher depending on what statements are related to each other and what their scores are they get a coefficient score for their main factor and in relation to other factors. Then each respondent also gets a score in relation to each of the factors. Depending if some scores are larger than the other participants with similar views also can be clustered together. When a series of statements is grouped and connected to a factor an interpretation is then possible. The Q-sort centroid Factor analysis is a programmed step in the process where the researcher chooses how many factors can be left for an analysis. Choosing less factors to be left for analysis results in more statements being grouped together in a certain discourse. For this thesis I did an analysis with 4 factors. One important step in the analysis is flagging participants whose coefficient score is too high for a factor. The average score used in Q-methodology is 0.37 and everything with six decimal points above that should be flagged. This is a Z-score (coeffiecient of the nominal value Z-score) that is decided by a specialized formula used in Q-methodology. The flagging allows grouping more statements together for a better interpretation and disregarding factors that do not have a strong interpretation. (PQM Manual, 2019). It should be noted that wrong flagging is going to show statements that might not be connected to a discourse and therefore this step should be done carefully for a better validity of results. After flagging, factor coefficients should be rotated. In the final output file, I had three factors out of the four that I started with after the flagging has been done. Elements of the output file are participants coefficients for a factor, statement scores with corresponding ranks and quasi- Q scores and Z-scores for each statement that is clustered together around a given factor and in difference with other factors. Q-scores represent the range of agreement or disagreement value that each statement gets (it can be +4 or -4) while the Z-scores is the factor loading (a sort of importance score) that a statement receives in relation to a discourse. (PQM Manual, 2019). This Z-score can have a positive or a negative value. At the end of the output file there is a list with statements that are of general consensus and do not belong specifically to any factor. Based on the output file ways of talking about a topic are formed (i.e. discourses).

3.4 Risks of Q-methodology With Q-methodology there is always a risk that the meaning of found

statements are misinterpreted by either the researcher or the participant which might influence their response. Misinterpreted responses in turn, will drastically change the scores and the importance given to each sentence and will result in establishing the wrong factors and discourses. Most participants that took part in the research found the Q-methodology process enjoyable and learning. However, they also report that scoring statements in relation to each other requires some intellectual effort but that is good according

to the participants for a deep analysis of state aid. Clarification on the meaning was sometimes asked depending on the participant. but this is a normal step for Q-methodology.