Path dependence or policy change in the EU?: A case study on EU policy change and gender mainstreaming

(1)

Path dependence or policy

change in the EU?

A case study on EU policy change and gender

mainstreaming

Master thesis in Political Science

(2)

Abstract

This thesis aims to clarify how the Single market policy and the Social policy of the European Union has changed since the Treaty of Rome (1957) and what opportunities and/or constraints there might be for integrating gender mainstreaming within the policies.

In order to determine how the policy development affects the prospects of integrating gender mainstreaming, a case study supplemented with the method of content analysis is carried out. Furthermore, the thesis adopts a historical institutionalist perspective. Thereby, policy change is operationalised with the help of Peter Hall´s theory of the three orders of change. It is therefore assumed that the concept of path dependency will explain how policy changes opens up opportunities for or constrain the integration of gender mainstreaming in the policy-making process.

The main findings show that the Single market policy mostly changed in accordance with the first order of change and that the Social policy mostly changed in accordance with the second order of change. These findings imply that a paradigm shift of the fundamental ideas and objectives of the policies have not occurred since the signing of the Treaty of Rome which complicates future integration of gender mainstreaming in the policies.

Key words

Policy change, Gender mainstreaming, Historical institutionalism, Case study, Content analysis, The Single market policy, the Social policy, Path dependency

(3)

1 Introduction

The European Union (EU) has been committed to gender equality issues for more than 60 years. The first measure that was introduced was Article 119 in the Treaty establishing the European Economic Community (hereafter: The Treaty of Rome, 1957) that established equal pay between men and women (Art. 119, Treaty of Rome). According to Mazey (2001: 23) this article was, until the Treaty of Amsterdam, the only legal basis for gender equality measures. However, in 1995 after the Fourth world conference on women in Beijing, the United Nations and its member institutions decided to integrate gender perspectives into the policy-making (Rai, 2018: 17). The year after, the strategy was formally implemented in the EU which thereby had committed itself to the task of integrating a gender perspective in all policy areas and at

all levels of the policy making process (Mazey, 2001: 10).

Today, the European institute of gender equality (EIGE) has published suggestions of how gender mainstreaming should be implemented at four stages in the policy making process and have listed 19 policy areas that have integrated gender mainstreaming. These includes employment policy, education policy, youth policy and economic and financial affairs policy in which the first three are included in the Social policy and the latter is included in the Single market policy (EIGE, 2021a). All in all, by integrating a gender sensitive perspective in the policy making process, gender mainstreaming in the EU aims to prevent inequalities to be reproduced (EIGE, 2021b). However, according to Mazey (2001: 23) the lack of Treaty bases complicates the expansion of gender mainstreaming strategies. Therefore, in order for gender mainstreaming to become fully integrated in the policies, policy changes within the Treaties are required. Therefore, it is of relevance to question: how have two of the founding policy areas of the EU, namely the Single market

(5)

policy and the Social policy, changed over time and how has that development affected future integration of gender mainstreaming within them?

According to Locher and Prügl (2009: 185-186), the introduction of gender mainstreaming within the EU has been possible since the Union, in the Treaty of Rome, already had taken measures regarding gender equality and it was therefore included in the Social policy of the EU. The authors therefore stress the importance of previous decisions that has opened up the opportunities for further develop gender equality work in the EU (2009: 185-186). The ideas of Locher and Prügl (2009) derives from the theory of historical institutionalism. According to Guy Peters (2019: 80), a historical institutionalist perspective of policy change implies that previously determined policy decisions affects the future opportunities and/or constraints for policies to change.

Therefore, this thesis carry out an analysis of policy change with a historical institutionalist perspective to explore what opportunities and/or constraints there might be for integrating gender mainstreaming in the future. In order to do so, the two policies will be analysed over time to clarify how much they have changed since the Treaty of Rome to the Treaty of Lisbon. It is therefore assumed that the concept of path dependency can provide an understanding of the extent of change that has occurred within the policy areas and how it affects the prospects of integrating gender mainstreaming within them.

Moreover, Drachenberg and Brianson (2016: 203) states that policies can be perceived as policy chains in which the different stages in the policy-making process are connected with each other and constitute a chain of feedback-loops that affects the policy process from the definition to the implementation of it. Therefore, it is relevant to study how the policy-making process have changed concerning what competences they provide, the policy techniques that are being applied as well as what policy areas that are included since they all affect

(6)

the policy outcome. To exemplify, according to Drachenberg and Brianson (2016: 207-208), the Education and training policy developed substantially when the open method of coordination was implemented in the policy area since it acted as a window of opportunity to go around the requirement of unanimity between the member states and thereby facilitated policy decisions and cooperation. Owing to this example, with a historical institutionalist perspective on policy change, the changes in the policy-making process affects the policy outcomes. Furthermore, it is assumed that policy changes are most likely to be path dependent and thereby impede a fundamental shift in the underlying ideas and policy perceptions that is the main objective and a necessity for gender mainstreaming to become integrated in the policy-making process.

With this in mind, it is relevant to study how two of the fundamental policy areas of the European Union have developed over the years. An analysis of policy change within the European Union gives a hint of how the identity of the Union has changed over the years. According to Drachenberg and Brianson (2016: 209), the development of the policy-making process affects what possibilities there are for the member states manoeuvre the policy nationally. Studying how the policies have changed on the EU-level therefore indicates of how and to what extent one can expect the policies to be implemented in the member states. Furthermore, according to Corbett, Peterson & Kenealy (2018: 67), studying the European Union provides information about European politics in general since the organisation affects national politics on a daily basis. Analysing if, and if so how, the development of the EU policies is path dependent therefore contribute to the understanding of the future international cooperation as well as the future role of the nation state in Europe (2018: 7-8). Moreover, contributing to the general understanding of the role of the EU as well as the EU institutions could counterbalance the growing euroscepticism. According to Corbett, Peterson and Kenealy (2018: 9)

(7)

euroscepticism is believed to increase when knowledge and understanding of the EU is lacking. Analysing the EU and providing information of it´s temporal development therefore contributes to more accessible definitions of the complex body of the EU.

Therefore, the main Treaties of the EU are analysed in order to explore to what extent the Single market policy and the Social policy have changed over time. According to historical institutionalism, policy changes tend to be path dependent and are affected by reinforcing mechanisms and previous policy decisions. In order for a fundamental change to occur, a paradigm shift that changes the perception and the idea of the policy of it is required (Peters, 2016: 81). However, to what extent has such policy change occurred in practice?

1.1 Research purpose and research questions

The aim of this study is to explore how changes in the Single market policy and the Social policy of the EU are affected by opportunities and/or constraints established by previous policy decisions, starting from the Treaty of Rome and finishing with the Treaty of Lisbon, and how these might affect future integration of Gender mainstreaming within the policy areas. This is done by adopting a historical institutionalist perspective and applying Peter Hall's perspective of the three orders of change.

* To what extent have the Single market policy and the Social policy of the European Union changed over time?

* In accordance with historical institutionalism, what opportunities and/or constraints can be identified for gender mainstreaming to become integrated in the policies?

(8)

2 Theoretical framework

This section will present the two theoretical orientations that will constitute the theoretical framework of the thesis. First, historical institutionalism and Peter Hall's three orders of change will be presented, along with other scholars on the subject, which will be used to operationalise policy change within the European Union. Secondly, the theory and strategy of gender mainstreaming will be presented. Furthermore, the chapter includes previous research, definition of relevant concepts and outline the theoretical connection between policy change and gender mainstreaming.

2.1 Historical institutionalism

A historical institutionalist perspective on policy change is used in this thesis with the ambition to explore if and, if so, to what extent Single market policy and Social policy have changed over time. According to Guy Peters (2019: 80) the core assumption of historical institutionalism is that the changes in policies are affected and constrained by the policy choices that have been taken previously. Taken the other way around, policy decisions that are being made today will have long-term influence and determine the future of an institution. This is the concept of “path dependency” which implies that institutions are likely to follow a previously formed path that constitute the possibilities and limitations of policy change. Significant changes of an institution therefore include changes in the underlying ideas of the institution that was relevant at the time when the institution was created. (Peters, 2019: 81). In this thesis the development of the Single market policy and the Social policy are being studied over time and the historical institutionalist concept of path dependency is suitable to clarify whether considerable changes have occurred or if the policies are being path dependent. Moreover, it is a suitable theoretical

(9)

framework to determine if, and if so, how todays application of gender mainstreaming is affected by policies being path dependent.

Furthermore, Paul Pierson (2000) problematized the concept of increasing return-mechanisms within institutions, also known as positive feedback-mechanisms. According to Pierson (2000: 252), political actors within political institutions are prone to act in accordance with the previously established path because of the high costs that a deviation from the path infers. In the context of extending gender equality within the Union, it is a reasonable assumption that it comes to a halt when in deviates from the previous established policies. The further the actors follow the path, the more likely they are to take more steps in the same direction. The costs of exiting the path and breaking the pattern of the development increases over time. Pierson describes this process of behaviour among political institutions as self-reinforcing mechanisms and increasing returns (2000: 252). Furthermore, considering the nature of politics is a collective action, political actors within political contexts tend to adapt their behaviour in relation to the existing expectations of the institution. It would therefore be arguable to claim that taking political initiatives that deviate from what has been the standard implies considerable effort, both for the individual but also for the institution and would thereby explain why developments of policies tends to come to a halt (2000: 257). This can also explain why power asymmetries remain within an institution as well as why policy learning is subject to path dependency (2000: 259).

Kathleen Thelen (1999: 386) is another historical institutional theorist and she defines institutions as socially constructed that are embodying shared ideas, culture and interpretative frameworks. Considering this, Thelen (1999: 384) states that an institution can not be understood unless one understands the political context that the institution is created within. According to Thelen

(10)

(1999: 385), ideas become locked in the institution and constitute a common path that affects human behaviour. The political actors therefore adapt their behaviour by following a predetermined path which results in the institutional development being path dependent. Furthermore, the author states that understanding the mechanisms of reproduction and positive feedback are crucial in order to understand policy change (1999: 400). As well as Pierson (2000: 259), Thelen also emphasises the process of positive feedback mechanisms as a reason of why power privileges of certain groups reinforces within institutions and also the other way around, other groups continue to be marginalised (Thelen,1999: 394). According to Thelen (1999: 397) institutions possess different reinforcing mechanisms that affect the way that they react to external trends. She illustrates this with gender equality being most likely to occur in a political context where liberal values are more embedded in the institutional structure, than in more conservative settings. Moreover, Thelen states that this is due to the reinforcing elements that in a conservative setting will impede that type of change since the foundation of the institution rests upon ideas about, for example, a conservative perception of gender roles (1999: 397).

Peter Hall published in 1993 an article about three orders of change within institutions and specifically within policy making. These constitute the different stages of change that occurs when, and to what extent a policy is changing. According to Hall (1993: 280) the first order of change applies to incrementalism and small changes that are common in policy making such as small adjustments. The second order of change goes deeper than the first and applies to new policy instruments and changes in the policy techniques. The third order of change is more of a fundamental change of the institution and a paradigm shift occurs (Hall, 1993: 280). Despite presenting the three orders of change, Hall also emphasises the importance of ideas within an institution. According to Hall (1993: 289) institutions are founded upon a set of shared

(11)

ideas and considering ideas are floating and changing in a society, the institution that is connected to the society will be affected by these ideas.

Furthermore, Guy Peters presents a classification scheme with four types of change, that could be said to correspond to the ideas of Peter Hall (2019: 91-92). The concept of layering relates to the first type of change in which the idea of the policy persists but new actors and ideas enter that only adjust the institution in a limited manner (Peters, 2019: 91). According to Peters (2019: 92) drift and displacement is two other types of changes where new interpretations and extension of working areas are entering as well as institutional mechanisms replacing others. According to Peters (2019: 92), actors within the institutions might attempt to make changes to fit with the previous institutional path, however at this stage changes will nevertheless occur. Furthermore, Peters illustrates this type of change with the EU institutions which replaces some of the national institutions. Finally, Peters presents conversion as the final type of change where the fundamental idea and meaning of the institution is being changed.

One could argue that the first type of change that Peters presents, Layering, corresponds to Hall's first and second order of change because of the small amount of changes and them being routinized and common in nature. Furthermore, the second and third type of change that Peters presents, Drift and Displacement, corresponds to Hall's second order of change. These are related in terms of the creation of new policy instruments and a deeper change of the techniques and interpretation of them. Furthermore, Pollack (2009: 136) states that so-called decision making” in the Union leads to a “joint-decision trap” which results in policy outcomes being path dependent. This could be illustrated in the Treaties as rules on decision making require unanimity which would indicate that member states remain in control over decisions due to their opportunity to veto decisions. A change in the decision

(12)

making process towards qualified majority would imply decreased control for the member states, and it would thereby provide a window of opportunity to further integrate gender mainstreaming into the policies. Finally, Peter and Hall present a third type of change which derives from what Hall states as a paradigm shift where the fundamental idea of the institution is being changed. Considering that the ambition for this thesis is to explore how previous policy decisions provide opportunities and/or constraints for integrating Gender mainstreaming in the policies, the historical institutionalist is helpful when the ambition is to understand haltered change in specific institutional settings. Guy Peters and Peter Hall's categorisation of political change is suitable to apply since it can provide an analytical tool to structure the material in order to carry out the analysis. On the one hand, historical institutionalism has been criticised for not explaining policy change but rather the lack of change. Guy Peters (2019: 90) states that with historical institutionalism it is difficult to predict change and that the conceptualisation of policy change within the theory lacks concrete explanations of why institutions do change. On the other hand, this thesis aims to study how policies have remained more or less static over time and whether the fundamental assumption within Gender mainstreaming, that the ideas and perceptions of a policy needs to be changed in order for gender mainstreaming to become integrated and fulfil its purpose, has occurred. Therefore, historical institutionalism is applicable and suitable for the purpose of this thesis. According to Hall and Taylor (1996: 9), the historical institutionalist approach is more suitable to analyse the maintenance of policy legacies and the difficulties for policies to change rather than predicting how they might change in the future.

(13)

2.2 Gender mainstreaming

Gender Mainstreaming was first acknowledged in Mexico at the World conference on Women in 1975 but received further international attention at the Fourth world conference on women in Beijing in 1995. According to Sonia Mazey (2001: 10), gender mainstreaming is a theory inspired by constructivist feminist theory which emphasise that gender and gender roles are socially constructed and are affecting the way the society perceive politics. Furthermore, according to Mazey (2001: 10), the overrepresentation of men in power-positions and decision making bodies results in: “…laws and policies,

which perceive women as men perceive women rather than acknowledging the reality of women's lives.” Mazey (2001: 10) also states that: “Defining equality in this way results in the world of men (and the public sphere) becoming the norm to which women must fit.”

According to Mazey (2001: 13-14) policies are created in a political context that sets out a “policy frame”. A policy frame is socially constructed and sets out the frame in which a group of people or an institution perceive phenomenon such as the prevailing gender roles. Furthermore, a policy frame affects how a group perceives a problem, how the responsibility of the problem is distributed and also how it could be solved. Mazey states that policy makers need to problematize the current policy frames they are operating within and with a gender perspective becoming aware of how policy outcomes are affecting women and men unequally. To fulfil this aim, the author states that policy learning, training and an increased participation of women are needed. Furthermore, the author emphasises the importance of working with gender mainstreaming constantly and in parallel with the policy making and not working with it separately. Since the aim of the study is to analyse how policies have changed over time and how the change constrains or provide with opportunities for gender mainstreaming to become integrated in the policies, this theoretical perspective of gender mainstreaming is of relevance and could

(14)

be said to complement the theoretical framework of historical institutionalism well. Mazey emphasises policy change as a component that affects policy makers in their decision making which could relate to the historical institutionalist perspective of the institutional setting that affects how policy decisions are being made in relation to previous choices.

Furthermore, Goetz (2018: 69) states that the complexity of an institution and the many different actors that are operating within it require gender mainstreaming to be integrated in all policy areas and at all stages. Specifically, the author problematizes the ways that institutions are affecting each other and therefore the importance of gender mainstreaming being implemented in all policy areas. Goetz emphasises that even though women are included in the administration and the administration is gender sensitive, gender mainstreaming is at risk of being too isolated from various policy areas which thereby undermines gender mainstreaming to integrate the institution to the fullest (2018: 86). Goetz exemplifies this with a situation where gender mainstreaming is not integrated within the institutions of budgetary and how it affects other policy areas in terms of not getting enough funding to promote and practice gender equality. The author underlines one of the main reasons for this as bureaucratic norms that are operating and affecting the way that policies are being made (2018: 69). According to Goetz (2018: 79) there are bureaucratic interests of protecting what the authors present as “ministerial territory”, that impedes the process of including other aspects of policy and cross-cutting policy areas.

The mentioned perspectives of gender mainstreaming could be said to relate to what Pierson (2000: 259) presents as “self-reinforcing mechanisms”. As stated previously, Pierson claims that power asymmetries remain within institutions since the institutions themselves are founded upon them and political actors tend to adapt their behaviour after what is expected (2000:

(15)

257). Pierson also states that the learning process within institutions is path dependent since they develop within a certain political context (2000: 259). This is something that Goetz (2018: 69) expresses in terms of bureaucratic norms. Furthermore, Pierson also claim that the costs of breaking an already determined path increases over time which makes changes of norms and policies more difficult (2000: 252), which is something that is expressed in this theoretical framework of gender mainstreaming in terms of an institutional resistance towards policy learning and expansion of policy areas (Goetz, 2018: 79). Moreover, Pierson (2000: 259) and Thelen (1999: 394) emphasises the shared ideas and interpretative frameworks within institutions as sources for reinforcing power asymmetries. Thelen states that ideas become locked in within an institution when it is being created and due to the reinforcing mechanisms, changes of the interpretative frames that gender mainstreaming aims for is difficult.

However, Locher and Prügl (2009: 183) state that when studying European integration with a feminist or gender perspective it enables an understanding of how gender is structuring integration and provides a clarification of the development of gender equality within the policies of the EU. According to the authors (2009: 184) gender mainstreaming is a third and most recent step towards gender equality within the union and illustrates how gender equality has expanded in the EU over the years. The authors claim that the process of integrating gender perspectives in the policy making processes have been explained by historical institutionalist as a result of the first attempts toward gender equality that have affected the gender equality work of the Union throughout the years. Moreover, the authors claim that the implementation of gender mainstreaming is a result of changing and developing policy frames that activists and supranational bodies within the Union have succeeded in implementing their efforts of new gender norms (Locher & Prügl, 2009: 185-186).

(16)

Nevertheless, Kardam and Acunter (2018: 105) argues that one needs to study the objectives and procedures of an institution in order to understand how the institution is practicing gender mainstreaming. It is therefore applicable to study two policy areas over time to understand how they correspond to gender mainstreaming. Furthermore, similarities between historical institutionalism and gender mainstreaming have been identified and they therefore contribute to a coherent theoretical framework that corresponds to the purpose of the thesis to clarify how the Union is developing towards integration and how gender mainstreaming is related to the development.

Roberta Guerrina and Kathrine A.M Wright (2016) have written an article which scrutinises Europe as a normative power in terms of promoting equality in transnational settings. In accordance with feminist institutionalism they analyse how and why gender mainstreaming is excluded in external policies. Guerrina and Wright (2016: 295) criticize the EU for committing to gender equality in rhetoric and communicational terms, but not when it comes to operationalising the values in practice. Furthermore, they state that gender mainstreaming is more likely to be integrated in the policies that traditionally are associated with gender equality issues and might therefore be more integrated in social policies in comparison to issues relating to the Single market policy. Furthermore, Guerrina and Wright (2016: 304-305) emphasise that the member states play a crucial role in impeding gender mainstreaming of external policies. Therefore, it has traditionally been the supranational institutions of the EU, such as the European commission, that have promoted a gender perspective and the expanded of policy areas that includes gender mainstreaming. The authors conclude the article by stating that the EU shows limited engagement to incorporate gender mainstreaming as a core principle and policy strategy.

(17)

In summary, it is plausible to assume that a third order of change is required in order for gender mainstreaming to become integrated and fulfil its purpose within the policies. This theoretical connection between policy change and gender mainstreaming emerges since various scholars of gender mainstreaming, for example Goetz (2018: 69) and Mazey (2001: 13-14), stresses that changes in the policy frames are necessary for the integration of gender mainstreaming. Furthermore, Kardam and Acunter´s (2018: 105) also argue that it is relevant to focus on the objectives and procedures when one analyses application of gender mainstreaming. These argumentations go hand in hand with the theoretical assumptions of historical institutionalism and the two theories therefore supplement each other suitably.

All in all, gender mainstreaming aims to change the objectives and ideas of a policy which corresponds to Peter Hall´s third order of change. Furthermore, the two theoretical orientations stress the importance of reinforcing mechanisms as reasons for why policies are difficult to change and thereby are path dependent. One the one hand, the objective of gender mainstreaming is not to change the fundamental idea of the policy and one could therefore think that the connection to Peter Hall´s third order of change is exaggerated. However, on the other hand gender mainstreaming aims to affect the gender awareness of the policies and argues the importance to change the reinforcing mechanisms of power asymmetries that are embedded in the institutions and that results in policies having unequal consequences for men and women. Moreover, it is arguable to state that the more changes that are visible in the policies where supranational tendencies are to be shown, the better opportunities there are for gender mainstreaming to become integrated since it historically has been member states that impedes the integration of gender mainstreaming and supranational institutions that has opened up the possibilities (Guerrina & Wright, 2016: 304- 305).

(18)

3 Methodological framework

The following section will outline the methodology of the thesis. Initially the research design will be presented, followed by the methodological framework of case study and how it is supplemented with content analysis. Finally, the material of the study will be presented as well as the delimitations of the study.

3.1 Research design

The research design of this study will throughout have a qualitative and comparative approach with the ambition to explore policy change in two policy areas. According to Esaiasson et al. (2017: 211), a qualitative approach is meaningful to apply when the ambition is to structure texts and bring forward the relevant parts to analyse them both individually as well as in a greater context. In this thesis the main Treaties of the European Union are of focus to clarify how the Single market policy and the Social policy have changed over time. Considering that the theoretical framework of policy change includes an analysis of the ideas within the policy areas, a qualitative approach is suitable since it signifies that ideas and meanings are created in the relationship between individuals and therefore are expressions of how a group of people or an institution perceive a phenomenon (Esaiasson et al., 2017: 211-212). Furthermore, according to Esaiasson et al. (2017: 213) systematically structuring of texts with the help of themes and categories enables clarification of the underlying ideational framework which is suitable for this thesis.

Therefore, the case study method is chosen and is supplemented with content analysis. According to George and Bennet (2005: 77), conducting a case study means to explore a delimited historical event in detail. It is therefore of

(19)

importance to clearly define if and how the phenomenon that is being analysed is a subclass of a more general phenomenon. In this thesis, institutional change is narrowed down to policy change and delimited to two specific areas of study. Furthermore, Drachenberg and Brianson (2016: 203) states that the different stages of the policy-making process are interconnected with each other. Therefore, since the research purpose of this thesis aims to clarify to what extent the policies have changed over time, it is of relevance to include the policy-making process in the analysis to understand whether the policy development facilitates and/or impedes the opportunities for gender mainstreaming to become integrated.

3.2 Case studies

The methodological framework of this thesis is a case study. Two policy areas will constitute two cases that will be analysed in order to answer the research question. According to George and Bennet (2005: 5), the purpose of case studies is to conduct detailed information of a historical event that can or can not be generalised. Furthermore, the authors present three phases of case studies that the researcher should follow. Phase one intends to formulate a clear and focused research objective and a research design. Moreover, in phase two, the researcher uses the analysis scheme to structure the material of the cases to make them comparable. The result of the material structuring in phase two will thereafter be used in the analysis of phase three with the aim of providing answers to the research questions and fulfil the research purpose (2005: 89). In this thesis, the structuring of the texts which George and Bennet (2005: 89) presents as asking questions to the materials, is carried out in accordance with the method of content analysis. In other words, phase two of this thesis will systematically structure the content of the chosen material with the help of Peter Hall´s three orders of change. Phase three consists of the analysis which will determine to what extent the policy areas have developed

(20)

and what conclusions that can be drawn in accordance with historical institutionalism.

Furthermore, George and Bennet (2005: 6-7) focus on the method of “process-tracing”, where the researcher aims to understand the relationship between historical events and how they cause certain outcomes. In this thesis the historical development of the policy areas is analysed with the ambition to clarify how the policy change affect the opportunities and/or constraints for integrating gender mainstreaming in the policies. Therefore, two of the founding policy areas of the EU are chosen as units of analysis: The Single market policy and the Social policy. The Single market policy is chosen on the grounds of being a fundamental institutional bloc of the Union since the creation of the organisation. Egan (2016: 256) describes the Single market policy as the “core element of European integration” and the European Commission (2020) stresses that the single market is one of the greatest achievements that has been done in the Union and is the beating heart of the EU. Furthermore, although the EU historically has presented itself as a free market union, elements from the Social policy have been included since the Treaty of Rome. Since the aim of the study is to explore policy change these two policy areas are the cases of study since they have been included from the Treaty of Rome and are therefore assumed to have been subject to policy development over the years.

According to Esaiasson et al (2017: 102), case studies with fewer number of cases require a strategic selection process. In this thesis the most similar-design is chosen as case selection-strategy. Esaiasson et al (2017: 102-103), states that, in accordance with the most similar-design strategy, cases are selected on the basis of having similar value on most of the relevant, independent variables except from the one that is assumed to explain variation

(21)

in the dependent variable. In this study, the Single market policy and the Social policy are similar in being policy areas of the EU and, to varying extent, have been included and developed since the Treaty of Rome.

Nevertheless, one could argue that case studies provide limited information of a phenomenon considering the limited amount of cases and therefore it is difficult to generalise the studies to a greater context. However, this thesis aims to explore how the two policy areas have changed in detail and the case study method is helpful in doing so. There are therefore no ambitions to generalise the findings to other policy areas, but rather to provide information of how the specific policies have changed over time. Furthermore, the analysis includes two cases that are being compared which, according to Esaiasson et al. (2017: 109), provides further information of the causes and effects of a phenomenon and thereby facilitates the opportunities for drawing correct conclusions. Furthermore, another way of carrying out the analysis could be to go deeper into the textual expressions of the treaties, by for example applying an ideal type analysis, and analyse the underlying ideas and ideologies of the treaties (Bergström & Svärd, 2018: 257). Nevertheless, the aim of this thesis is to first and foremost study what competences, techniques and policy areas that are included and provided in the policies and how these have changed over time and to thereafter draw conclusions if there are indications of changes in the objectives and ideas of the policies. This thesis therefore does not attempt to clarify what ideologies or ideas that are expressed in the narrow way of an ideal type analysis, but rather to clarify more concrete changes that the methodological framework of case study and content analysis is more suitable for.

(22)

3.3 Content analysis

According to Esaiasson et al (2017: 213) one can distinguish between two different orientations when it comes to qualitative textual analysis: systemising and critically examining. In order to conduct this thesis, the systemising approach is the most relevant since the aim is to clarify to what extent policy change has taken place and, if so, how that change has been expressed in the treaties. Therefore, content analysis is chosen as a method to systemise and structure the treaties to understand the content of it (Esaiasson et al, 2017: 213).

Nevertheless, according to Boréus and Kohl (2018: 52) content analysis is a suitable method to use when the ambition is to structure a larger amount of material and compare textual documents over time. Therefore, the method is relevant to this thesis considering it will bring forward an overview of the material and make an analysis of the potential policy change possible. Boréus and Kohl (2018: 58-59) emphasises the importance of an analytical instrument to structure the texts as precisely as possible. In accordance the theoretical framework of the thesis, an analysis scheme is developed with the purpose of operationalising policy change. This instrument is applied to the material with the ambition to categorise the content of the material and enable an analysis of identified policy change. According to Boréus and Kohl (2018: 58), the precision of the analysis scheme depends on the codification. The material in this thesis will be codified manually since the ambition is not to cover a large amount of material but rather to carry out a deeper and more narrowed analysis of a delimited amount of material. In order to create a coding scheme that corresponds to the material the researcher reads the material to collect comprehensive information about the characteristics of the Treaties and thereafter create a detailed codification scheme. This is done by incorporating

(23)

parameters to Peter Hall´s three orders of change that are suitable for the chosen material.

3.4 Operationalization of policy change

According to Esiasson et al (2017: 64-65), high validity and high reliability are reached when the design of the analysis instrument is measuring what it is intended to measure as well as being applied correctly to the material. The decision to delimit the policy areas to two decreases the risk of applying the analysis instrument inaccurately since it facilitates a narrower focus on the material and therefore aims to reach high reliability (Esaiasson et al., 2017: 64). Furthermore, the indicators of the analysis scheme are adapted to fit with the material in the treaties as well as being theoretically connected to policy change in order to reach high validity. Since this thesis adopts a historical institutionalist perspective, the analysis scheme derives from Peter Hall´s three orders of change. However, the ambition is to clarify what opportunities and/or constraints the development of the policies might bring forward, further parameters are included that derives from the theoretical framework consisting of historical institutionalism and gender mainstreaming.

An example of similar operationalization and composition of theory and method is to be found in Brigitte Pircher´s article about EU public procurement policy. In this article, Pircher (2020) operationalise Peter Hall´s three orders of change with the purpose of studying and measure policy change. Furthermore, Pircher (2020: 511-512) emphasises deepening treaty basis, changing policy instruments and techniques as well as increasing supranational elements as indicators of policy change, which are perspectives that are included in this thesis as well. It is therefore reasonable to state that the chosen method for the purpose of this thesis is applicable in relation to the

(24)

theoretical framework. Moreover, according to Pircher (2020: 512) historical institutionalism is applicable to policies that can be regarded as institutions when they imply binding rules. This applies to the chosen cases in this study since the policies are part of the EU Treaties and thereby implies competences and binding rules to a varying degree.

The first and second order of change are similar in the way that they both include aspects of “layering” in policy change. Unlike the third order of change, these two types of change do not infer any fundamental changes in the policy goals or the underlying idea or meaning of the policy. However, what separates two is that the first order of change implies changes in the policies that are in accordance with the already routinized decision making processes (Hall, 1993: 280). This should indicate that the policies are changing in accordance with policy legacies where the changes are of the nature that could be expected at a certain time. Peters (2019: 91-92) exemplifies these types of changes in policies as new individuals and ideas are included in the policy, but the functioning and the techniques of the policy remain. In the treaties, it could constitute that new aspects corresponding to the main goal are included without changing the fundamental setting of the policy. Furthermore, examples of a first order of change implies amendments of already established policy techniques rather that the replacement of them.

In the second order of change, Hall (1993: 280) presents new policy instruments as indicators. Changes in instruments such as voting procedures are indicators of a second order of change. Peters (2019: 92) emphasises changes in the institutional character as indicators, for example when one institution replaces another one. Changes in accordance with the second order does not necessarily indicate a “deeper” change than the first order at first glance, but a change in voting procedures for example, could result in more

(25)

radical changes that affects policy outcomes. Actions that correspond to harmonisation processes, furthering supranational tendencies or changes in the policy techniques related to furthering and deepening European integration could be said to be indicators of a second order of change. Furthermore, if the policy has expanded in terms of further policy areas being included in the Treaties, it also categorises as a second order of change.

Third order of change, is the policy change with the distinguishing

characteristics of a paradigm shift (Hall, 1993: 280). Changes in the overall policy goals and objectives and overturning changes in the competences of the EU indicates of changes that a third order of change takes place. According to Peter Hall (1993: 278), a third order of change implies that policy makers have experienced social learning and therefore attempts to change the ideas and goals of a policy, rather than only changing the policy instruments and techniques which indicates that previous policy decisions are still influencing the policy making process. In this stage of change, the policy development is less path dependent and thereby decreasing the risk for previous policy obstacles to impeding the process.

3.5 Table 1: The operationalization of Hall’s order of change

Order of change Policy changes Indicators First order of

change

Adjustments in basic policy instruments. Layering.

Visible policy legacies. Expected and normal change over time.

Second order of change Change in policy techniques. Institutional change. Layering. Harmonisation ambitions. Supranational tendencies. Changes in voting systems (from unanimity to qualified majority voting).

(26)

Expansion of policy areas and competences.

Third order of change

Paradigm shift Changes in the policy objectives and meaning of the policy.

Competences being changed.

3.6 Materials

Considering that the policies are perceived as institutions in this thesis, the legal basis of the policies constitute the main material. In order to answer the research questions and explore how policy change has occurred over time the Treaties of the European Union will constitute the main material. The collection of data therefore covers the time period from when the Treaty of Rome (EEC Treaty) came into force in the year 1957 to year 2009 when the Treaty of Lisbon (2007) became effective. Nevertheless, the original version of the Treaty of Rome was not available in English on the EU official website for legal documents (EUR-Lex), therefore this treaty is collected from the official website of the European Commission instead. One could argue that this decision affects the coherence of the material gathering. However, the treaty is collected from a primary EU source and is therefore perceived as legitimate. Furthermore, other secondary material from primary sources from the EU will be included to supplement the main material and to provide with other information that is relevant for understanding how policies have changed over the years. These includes fact sheets from the European Commission and the European Parliament, memos and reflection papers from the European Commission, information from the EU agency for the improvement of living and working conditions “Eurofound” as well as Treaty summaries and information pages from the official website for EU legislation EUR-Lex. In summary, the material chosen for this thesis is primary EU-sources and original documents by EU institutions.

(27)

However, in order to clarify policy change, other literature is included as secondary source of material in the analysis. This source of material is included to isolate the policy areas from each other and alternative policies, as well as a supplement to the information from the EU-documents to extend the understanding of the temporal change of the EU. Finally, the information of both the primary and secondary sources of material are evaluated in accordance with source criticism. According to Esaiasson et al. (2017: 288), sources of material needs to be authentic, independent, contemporary and objective. In this thesis the primary material is gathered from EU official websites. Nevertheless, in order to fulfil the criteria of authentic sources, the information from the European Commission and the European Parliament have been compared to the treaties and therefore two sources of material supplement each other. Therefore, the material also is perceived to be independent. Furthermore, updated sources have been used to collect data that are of relevance at time of the treaties becoming effective and therefore provide with the primary information that was brought forward at the time. Moreover, the literature that has been used as supplementary and secondary sources have been controlled to come from independent authors and sciences.

3.7 Delimitations

Firstly, this thesis is delimited the to two policy areas Single market policy and the Social policy. These are chosen on the grounds that they constitute the most relevant in relation to the research purpose and to study policy change during the historical development of the EU since they have been included since the Treaty of Rome. Furthermore, they are suitable when the ambition is to explore the opportunities and/or constraints of gender mainstreaming becoming integrated in the policies. According to Goetz (2018: 69), gender mainstreaming is required to be integrated in all policy areas and in all stages

(28)

of the policy making process. Therefore, considering that the scope of this thesis can not include all of the policy areas, two of the most founding policies of the EU are chosen. Especially since the Social policy have historically included the treaty base for gender equality. Furthermore, Goetz (2018: 69) emphasises that institutions affects each other and it is therefore necessary for gender mainstreaming to be integrated in all policies. Thereby, the Single market policy constitute a relevant policy area since other policy areas, including Social policy, are closely attached to the Single market policy and development of them tends to be legitimised by its consequences for the Single market. In summary, the two policy areas are at the core of the EU development in general. Although, other policies include gender mainstreaming and are interesting in a historical institutionalist perspective, the development of the Single market policy and the Social policy are closely connected to each other and are therefore two suitable policies to compare. Secondly, the thesis is delimited to an elected amount of material. On the one hand, the following treaties of the EU are chosen for the thesis are: Treaty of Rome, Single European Act, Maastricht Treaty/Treaty on the European Union, Treaty of Amsterdam, Treaty of Nice, Treaty of Lisbon and its amendments to the newest versions of the Treaty on the European Union and the Treaty of the Functioning of the European Union. The Merger Treaty is excluded since the main focus of this thesis is on policy change and the Merger Treaty mainly includes amendments to the institutional setting of the EU and is therefore not as relevant.

(29)

4 Analysis

This chapter will outline the identified policy changes in both cases, the Single market policy and the Social policy, and analyse how they correspond to historical institutionalism.

4.1 The Single market policy:

According to Ina Sokolska (2020a) the Treaty of Rome (1957), was initiated in the aftermath of a fragmented Europe as a result of the world wars. The idea was to unite Europe by creating a common market and, thereby, the first steps of European integration where taken. Consequently, in Treaty of Rome Article 1, the European Economic Community is established. Moreover, in Article 2 Treaty express the Community´s task of creating a common market between the member states (Treaty of Rome, 1957).

It is reasonable to state that these articles are the founding ground of what later will become the Single market policy of the European Union. Furthermore, the Treaty outlines which procedures that are to be applied regarding the decision making of the economic policy. It is plausible to argue that unanimity is the primary decision making rule of the Treaty of Rome. To illustrate, the requirement of unanimity is expressed in for example Article 14(7) as well as in Article 149 (Treaty of Rome, 1957), which follows:

“Where, in pursuance of this Treaty, the Council acts on a proposal from the

Commission, unanimity shall be required for an act constituting and amendment to that proposal.” (Treaty of Rome, 1957: Article 149)

(30)

However, qualified majority voting is also embedded in the policy making process to a certain extent. To illustrate, in Article 14(5) and in Article 21(1 and 2) (Treaty of Rome, 1957) the Council is to act by qualified majority on proposals by the commission (Treaty of Rome, 1957: art14 & 21). However, there are also articles where both unanimity and qualified majority is required for decision making. For example, Article 20 requires unanimity in the Council when it is to act in the previous stages and thereafter voting in accordance with the qualified majority rule is applied. The inclusion of qualified majority voting in specific articles indicates that the Union initially takes a few supranational steps regarding the Single market policy since it implies that the member states partly delegate their sovereignty to the EU. Furthermore, the aim of establishing a European community and a common market was to change the trade and production conditions as well as bringing the member states closer to each other and to integrate them politically. These where the first steps of the development of the EU and the upcoming watershed regarding an intergovernmental or supranational role of the EU (EUR-Lex, 2017).

Moreover, according to Michelle Egan (2016: 256), the first initiatives taken regarding the Single market policy was the attempt to create a customs union and the attempt to liberalise the internal market of the Union. The latter applies to the establishment of the free movements and to remove trade barriers of goods, capital, services and labour. The main objectives of the Treaty were to harmonise and integrate the member states both in economic and in political terms (EUR-Lex, 2017). However, Egan (2016: 257) states that the creation of a single market also required that all the member states harmonised their national economic policies and unified their economic ideologies. Egan (2016: 258) stresses that the different perceptions of capitalism and economic ideologies between the member states impeded the process of integrating the

(31)

Single market and was specifically distinct due to the requirements of unanimity in the policy which haltered the decision making process.

Moreover, according to Sokolska (2020b) the Single European Act (SEA) that was signed in 1986 in Luxembourg. The treaty formulized a time frame for the establishment of the single market which stated that the internal market was to be finished in 1993. Creating an internal market was one of the main objectives of the Treaty and therefore the competences of the Union where extended and the voting procedures where changed. Qualified majority voting was applied to new policy areas including the internal market. Unanimity where thereby abolished in decisions regarding the single market, except from taxation, free movement of persons and the rights and interests of employed persons (EUR-Lex, 2018a). According to Egan (2016: 261), by changing the decision-making processes from unanimity to qualified majority voting, the Single European Act accelerated the work of creating a single market and thereby deepened the European integration within the policy area. In Section 2, subsection 1 (SEA, 1969) outlines the policy area of the internal market. In Article 16 (2 &5) it is explicitly stated that the requirement of unanimity in the voting procedures is replaced by qualified majority voting (SEA, 1986).

Although, the policy area contains sections where unanimity is still required, it appears that qualified majority is in the Single European Act the primary voting procedure to apply. According to Egan (2016: 261), the Single European Act modified the conditions for the internal market by changing the decision making procedures, in the majority of the policy areas, to qualified majority voting. The changes opened up opportunities for furthering the development towards a common market since the control of the member states was undermined when their possibilities to veto against decisions within the

(32)

that the policy changes that occur with the Single European Act are results of policy learning from the previous experiences of national interests haltering the integration of Single market policy. However, the changes mainly amended and altered policy procedures that where already embedded in the policy area to certain extant and that are in line with overall objective of harmonising and integrating the member states in order to create a common market, which where expressed in Article 2 in the Treaty of Rome (1957). Furthermore, the amendments where not only layered on top of the previous policy techniques but replaced them and are therefore to be classified as second order of changes (Hall, 1993: 280). It is reasonable to state that the changes of policy techniques strengthen the objectives of the Treaty of Rome.

Further examples of second order of changes are expressed in Article 18(4&5) (SEA,1986) which explicitly emphasises the ambition to harmonise economic policies of the member states that is required in order to create an internal market which is an indicator of a second order of change. Moreover, according to Phinnemore (2016: 19) the Single European Act facilitated harmonisation to occur within the policy area since a deadline for complete the internal market was presented. Furthermore, Ralf Drachenberg and Alex Brianson (2016: 200) states that the member states came to the conclusion that integrating the Single market policy required further competences for the EU in the policy area and therefore they abandoned their national sovereignty as qualified majority voting where implemented to larger extent. Moreover, according to Drachenberg and Brianson (2016: 200) the implementation of qualified majority voting and the change in the policy instrument that it implied, resulted in a historic development since it facilitated policy making that would otherwise have been impeded due to national interests. All in all, no indicators of a third order of change are to be identified in the Single European Act concerning the Single market policy. According to Thelen

(33)

(1999: 384), ideas becomes locked in institutions when they are created and the changes that occur thereafter are likely to follow the same path. Such changes of second order are visible with implementation of the Single European Act that are in line with the overall objectives that where expressed in the Treaty of Rome.

The Maastricht Treaty (Treaty on European Union, 1992) is characterised by the introduction of the three pillar structure. According to David Phinnemore (2016: 20-21), these where of importance to the single market policy since it was included in the first pillar. However, the changes that where made with the Maastricht treaty where perceived as determinant for the establishment of the EU although the ideas of them can be traced back fifty years (Phinnemore, 2016: 12). Furthermore, Phinnemore (2016: 12) states that the idea of supranational cooperation and integration of a common market was far from new when the Maastricht treaty entered into force in 1993. The introduction of the three pillar structure and what it signified for Single market policy did not change the Union´s competences regarding the Single market, which is outlined in in Title 6, Chapter one in the Maastricht Treaty (1992).

However, considering that the EU where provided with further competences regarding the Single market policy as well as the application of qualified majority voting in most of the policy areas in the Single European Act, there were limited changes in the policies and policy techniques when the Maastricht treaty was signed. Therefore, considering that the changes that were made in the Single market policy mainly amended the already established policy-making process, they are categorised as first order of change (Hall, 1993: 280). Nevertheless, according to Phinnemore (2016: 21), the Maastricht treaty was a setback for the European Integration process since the three pillar structure

(34)

implemented further instruments of intergovernmental character rather than supranational which impeded opportunities and effectiveness of the Union. Therefore, Phinnemore (2016: 21) states that the mix between supranational and intergovernmental measures subverted effectiveness of the union and its opportunities for further integration. However, the Single market policy belonged to the first pillar and although adjustments where made in the EU in general, they did not apply to the Single market policy to a significant extent. According to Pierson (2000: 252), the longer an institution or political actor follow the same path the more difficult it becomes to divide from it. At this stage of the development of the Single market policy, the it is therefore plausible to assume that it was more difficult than with the Single European Act, to establish considerable changes to the character of the EU. Furthermore, regarding the Single market policy the EU was already provided with competences. The amendments that where made in the Maastricht Treaty are therefore examples of policy changes in which previous policy legacies are visible and when the changes are merely results of a temporal process and external trends rather than intentional changes or replacements of the policy techniques (Hall, 1993: 280).

According to the European Commission (1999: 1) the Treaty of Amsterdam was signed in 1997 and was first and foremost a Treaty emphasising social rights, developing the common foreign and security policy and prepared the community for future enlargements. Furthermore, according to the Commission (1999: 1), several questions of importance where not included or negotiated in the development of the Treaty of Amsterdam. One of them being the future application of qualified majority voting (1999: 3). Nevertheless, in Article K.12 in the Treaty of Amsterdam (1997), the opportunity closer cooperation between member was provided in areas where joint decision making had not been possible. According to Phinnemore (2016: 23), the

(35)

procedure, also known as enhanced cooperation, applied to the EC framework which belonged to the first pillar of the Union and therefore included the economic policy area of the Union. According to Phinnemore (2016: 24) enhanced cooperation did enable collaboration on an intergovernmental level which fragmentized the harmonisation of the EU and instead fostered differentiation between the member states.

As stated above, experiences had shown that harmonisation and integration of policies opened up opportunities for further development of the policies. Therefore, taken the other way around, fragmentation and differentiation constrained the integration process of the policies. Similar to the Maastricht treaty, the Treaty of Amsterdam did not provide considerable policy changes concerning the Single market policy. It is thereby plausible to assume that the previous embedded intergovernmental mechanisms continued to undermine the opportunities for integrating supranational mechanisms and resulted in the policy development coming to a halt. According to Kathleen Thelen (1999: 385) and Paul Pierson (2000: 252) this process can be perceived as affected by positive feedback mechanisms and thereby reinforce previous policy decisions that impedes the implementation of new policy techniques.

To conclude, even though the ambition with the Treaty of Amsterdam was to bring the Union closer together, previous policy decisions continued to determine the developmental path of the policy and the Treaty did not extend neither competences or policy areas. However, the procedure of enhanced cooperation was introduced which affected the decision making process and development of the Single market policy. On the one hand, one could argue that the implementation of enhanced cooperation implies a second order of change since it changes the policy techniques and settings. On the other hand,

(36)

the implementation of enhanced cooperation differentiated the member states and constrained the harmonisation of the Single market policy which, considering the parameters of Peter Hall´s orders of change, do not indicate of second order of changes. However, both the first and second order of change imply the concept of “layering” that signify changes to the policy setting but not in the overall objective and idea of the policy (Hall, 1993: 280 & Peters, 2019: 91). Therefore, it is plausible to state that the changes did not affect the policy area considering supranational elements, harmonisation of policies or a paradigm shift in the ideas and the changes are therefore categorised in accordance with Peter Hall´s (1993: 280) first order of change.

According to the European Commission (2003) the Treaty of Nice aimed to prepare the Union for future enlargement. Regarding the Single market policy, the opportunity for the member states to engage in enhanced cooperation between each other was simplified that previously was delimited to cooperation between the majority of the member states. With the Treaty of Nice, the delimitation did instead concern one third of the member states. Furthermore, the opportunity for the member states to veto against enhanced cooperation regarding the first pillar was abolished. Instead, dissatisfaction regarding an initiated process of enhanced cooperation was to be brought to the Council which would decide by a qualified majority. Moreover, if the enhanced cooperation concerns issue where the co-decision procedure applies, the decision requires consent from the European Parliament (European Commission, 2003). These changes affect the the Single market policy since it belongs to the first pillar.

Furthermore, Phinnemore (2016: 26-27) states that when the Union simplified the opportunity of engaging in enhanced cooperation, the integration process

(37)

of the community became more difficult. As presented previously, further integration and harmonisation between the member states had previously provided opportunities for integrating Single market policy. On the one hand, the purpose with simplifying enhanced cooperation was to get around the problem of member states veto and impediment of policy development. On the other hand, it affected the uniformity of the Union. The amendments to the enhanced cooperation procedure are, similar to the Treaty of Amsterdam (1997), examples of what Peter Hall identifies as the concept of layering and first order of change. Furthermore, with the Treaty of Nice, new policy areas or competences where absent and it is therefore arguable to state that the changes that where made regards adjustments, rather than replaced or new policy instrument which would indicate of second order of changes (Hall, 1993: 280). Furthermore, it is also plausible to assume that the decision was a result of previous policy experiences and positive feedback mechanisms in accordance with previous policy decisions of intergovernmental character rather than supranational. Therefore, according to Pierson (2000: 257), the costs of breaking the already established path increases and the political actors are less prone to deviate from it which results in the development of the Single market policy coming to a halt.

Finally, the Treaty on the European Union (TEU) and the Treaty on the Functioning of the European Union (TFEU), previously named the Treaty of Rome, are summarised as the Treaty of Lisbon (2007) (Sokolska & Pavy, 2020). The Single market policy is included in Title 1 in the Treaty on the functioning of the European Union. According to Church and Phinnemore (2016: 40), the Treaty of Lisbon developed the decision making process and applied the community method to all policy areas with the exception of Common foreign and security policy. According to the European Parliament (2020c), the Community method applies to the first pillar and thereby the

Path dependence or policy change in the EU?: A case study on EU policy change and gender mainstreaming