1
Camiel Colsen 291192 Leiden University MA Thesis Economics and Governance, Public Administration
Social Convergence in the Southern EU-4 in a Post-Crisis
Europe: An Explorative Study on Various Welfare State
2
Abstract: Reducing social disparities within the European Union has been an objective of EU
policymakers since its establishment. Within the welfare state literature however little consensus exists on the occurrence and direction of welfare state convergence. Some scholars argue welfare convergence does occur over time and point to increased economic integration and Europeanization as causal factors behind this. Others say welfare states are resistant to such international forces and argue welfare states remain nationally unique phenomena. Domestic political structures, and path-dependent factors are behind this singularity the latter group argues.
The disagreement within the literature is largely attributable to ‘the dependent variable problem,’ the issue of a lack of consensus on what indicators to use for measuring welfare state development. The study performed here attempts to contribute to the debate by exploring social convergence in the Southern EU-4 countries over the period 2006-2016. With regard to quantitative studies on welfare state convergence, the past decade is a rather uncovered period. 2006-2016 is, however, a highly relevant period to study because of the occurrence of the financial- and euro crises. This study shows that in most social areas covered in this study convergence has occurred in the Southern EU-4. There are indications that these patterns of welfare convergence are brought about by pressure of similar problems the Southern EU-4 experienced. Thereby, trajectories of Europeanization imposed in reaction to the financial- and euro crises may have instigated patterns of welfare convergence in the Southern EU-4 Member States as well.
As this study serves more an explorative than explanatory purpose, I would advise future scholars to look more in depth to the causal factors behind Southern European welfare convergence in the wake of the biggest crises Europe has witnessed in decades. Such information could be highly valuable for policymakers in making the right decisions in their pursuit to reducing social disparities within the European Union. This research shows that the Southern Member States showed patterns of convergence amongst each other, but in some crucial social areas diverged substantially from the European Union at large. In order to respect this diversity within unity, I reckon it would be useful for EU policymakers to consider the development of a more regionally focused approach in stimulating the social dimension in Europe.
3
Table of contents
1. Introduction 6
2. Theoretical considerations on convergence and divergence 11
2.1 Theories on convergence 11
2.2 Theories on persistence and divergence 14
2.3 Measurement considerations 17
2.3.1 Social expenditure indicators 18
2.3.2 Social outcome indicators 21
2.4 Theoretical overview 23
3. Europeanization and economic developments 25
3.1 Pre-crisis developments 25 3.1.1 Economic developments 25 3.1.2 Europeanization 26 3.2 Post-crisis developments 27 3.2.1 Economic developments 27 3.2.2 Europeanization 31 4. Research design 34 4.1 Social indicators 34 4.2 Hypothesis 38 4.3 Methodology 39 4.4 Data collection 39
5. Results and analysis 41
5.1 Social expenditure indicators 41
5.2 Social outcome indicators 50
5.3 Convergence or divergence? 52
6. Conclusion and Recommendations 54
Bibliography
4
List of figures and tables
Figures
Figure 1: Development of gross government debt in the Southern EU-4 compared with the
EU-27, 2006-2016
Figure 2: Development of unemployment rates in the Southern 4 compared with the
EU-27, 2006-2016
Figure 3: Development of GDP per capita in the Southern EU-4 compared with the EU-27,
2006-2016
Figure 4: Openness of Southern EU-4 economies compared with the EU-27, measured as sum
of imports and exports in percentage of GDP, 2006-2016
Figure 5: Coefficient of variation for total public social expenditure as percentage of GDP in
the Southern EU-4, 2006-2016
Figure 6: Coefficients of variation of replacement rate indicators in the Southern EU-4, 2006-2015/16
Figure 7: Coefficients of variation of expenditure on ALMP’s and tertiary education in
Southern EU-4, 2006-2015
Figure 8: Coefficients of variation of family and household benefits in cash and in kind as
percentage of GDP in the Southern EU-4, 2006-2016
Figure 9: Coefficients of variation of the at-risk-of-poverty rate and Gini-index, 2006-2016
Tables
Table 1: An overview of scenarios and causes of welfare convergence and divergence
5 Table 3: Gross total public social expenditure as percentage of GDP, 2006-2016
Table 4: Averages of unemployment net replacement rates for a single person household
without children, and two-person household with children as a percentage of previous earnings, 2006-2015
Table 5: Pension aggregate replacement ratio in percentage of former income, 2006-2016.
Table 6: Public expenditure on active labour market policies in percentage of GDP,
2006-2016.
Table 7: Educational expenditure on tertiary education as percentage of GDP, 2008-2016
Table 8: Public expenditure on family benefits in cash and in kind as percentage of GDP,
2006-2015.
Table 9: Public expenditure on household benefits averages in cash and in kind as percentage
of GDP, 2006-2016.
Table 10: At-risk-of-poverty-rate, 2006-2016
6
1. Introduction
Ever since the offset of the European project, policymakers agreed that in order to attain balanced and sustainable integration and development within the European Union (EU), social systems of Member States (MS) should converge (European Parliament 2018). In the founding treaties of the Union this is expressed as follows: “the EU aims at reducing disparities between the levels of development of the various regions and the backwardness of the least favoured regions” (Treaty on the European Union, 2007). In line with this European ambition, welfare convergence in this study is defined as ‘’the reduction of disparities between the levels and types of social protection and social outcomes within countries.’’ The literature distinguishes two main types of convergence: growing together (decrease in variation over time) and catching-up (least favoured regions approximating steady state level of more favoured regions). In this study the focus will be on the first type of convergence. Furthermore, it is important to mention that in the literature welfare- and social convergence are viewed as interchangeable concepts, and both terms are used in this study.
With the signing of the Maastricht Treaty in 1992, a significant stimulus to European integration by among others the introduction of the single currency, the importance of social harmonization was emphasized even more and got a prominent place on the EU’s agenda. In the wake of the Maastricht Treaty, a number of policy frameworks have been set up that aimed for social harmonization in the EU. The European Employment Strategy, the Lisbon agenda and the subsequent Europe 2020 goals, and the European program for Employment and Social Innovation (EaSl) are examples of such policies.
Recently, the issue of social cohesion in the EU has witnessed a revival in attention. In the ‘Five President’s Report’ on the future of the European Monetary Union (EMU), launched in 2015 by the presidents of the EU’s main institutions, an effective and EU wide social protection system to protect the most vulnerable members of society was advocated. A 2016 report of the International Labour Organization (ILO) argued that the process of economic and social convergence in the EU has stalled, mainly due to the financial and euro crisis. In 2017, the European Commission proposed the creation of an EU Pillar for Social Rights. Commission President Juncker argued that ‘convergence towards better socio-economic outcomes, underpinned by a Social Pillar, could be the foundation for a more integrated and stable Europe and a fully functioning EMU,’ (European Commission 2018 & ILO 2016). In June 2018, German finance minister Olaf Scholz said he is a supporter of an EU wide unemployment scheme that should complement national protection mechanisms (Huggler 2018). Even more recently, during the European Council of December 2018, the ministers of
7 finance of the euro area agreed on the establishment of a ‘budgetary instrument for convergence and competitiveness’ (European Commission 2018).
This renewed attention for social governance in the EU can for a large part be attributed to the financial and euro crises that laid bare tender asymmetries in the EU’s economies. These imbalances resulted in disproportionate negative (social) effects throughout the continent. A belief exists that in order to make the Eurozone and the EU as a whole more ‘shock proof’ for the challenges of the future, the disparities between Member States’ economies and their social systems should disappear (European Parliament 2018). An important step towards reducing these disparities could be the provision of social protection at the EU level, policymakers argue (European Commission 2018). However, politicians in Northern member states often temper these plans by arguing that the EU is not a transfer union, and social protection should be provided for by national governments (Chazan 2018). In this light, it is interesting that a recent survey revealed that a major share of EU citizens does want a more social EU that is regulated and provided for at the European level (Vandenbroucke et al. 2019).
Partly because of this attributed importance, social convergence in the EU is an often-studied topic within the welfare state literature (Starke, Obinger and Castles 2008, Caminada, Goudswaard and van Vliet 2010, Paetzold 2012, Dvorokova 2014). Academics are particularly interested in the direction and drivers of social development in Western Europe. Does social convergence occur at all, to what extent are national institutional traditions determinative for the development of social policies, and how is social convergence affected by EU policies and international economic integration, are examples of questions that are studied intensively. However, despite the thorough attention welfare state development in the EU receives from academics, little consensus exists on the evolution of social policies in the EU.
Broadly speaking there are two opposing views within the literature. Supporters of the convergence hypothesis argue that welfare states show signs of convergence over time. This group points to increased economic integration and an accumulation of EU policy as being the main drivers behind trends of convergence. According to this group, convergence can occur in a positive and a negative direction. In the case of positive convergence, social policies expanse similarly to higher levels, whereas negative convergence indicates comparable cutbacks of social protection over time. A popular theory that explains retrenchment of social services in the Western world is the social-race-to-the-bottom hypothesis. Due to increasing globalization and more open borders, countries experience higher levels of regulatory
8 competition when it comes to attracting investments and corporations. Therefore, governments are likely to design an investment climate that is as favourable as possible for entrepreneurs but may go at the expense of tax-revenue that was first reserved for social policy: a general cutback in social provision lurks.
The other group, defenders of the path-dependency thesis, argue that welfare states do not significantly converge over time. National institutional design and traditions are much stronger determinants of social development than international factors, they argue. These scholars do recognize the potential influence of international developments on welfare states but find that domestic political forces are much stronger and foresee in stabilizing the status quo of the welfare state (Pierson 1996; Taylor-Gooby 2004). Welfare states are to a large extent resilient to international developments, they argue. The disagreement in the welfare debate finds its origin partly in the fact that welfare states are such enormous and dynamic phenomena that they are intrinsically challenging objects to study. This makes that no consensus exists on what indicators are best and most indicative for determining welfare state development. Some scholars advocate studying aggregate expenditure data, whereas others say studies should focus on the programme and individual level. This controversy on measurement techniques was named the ‘dependent variable problem’ by welfare state authority Green-Pedersen (2004).
This study aims to contribute to this debate on European welfare state convergence by analysing the development of welfare systems in the Southern EU-4 countries -Greece, Italy, Portugal and Spain- over the period 2006-2016. By analysing a broad spectrum of social areas, I attempt to give an exploratory description of social development and convergence in the Southern EU-4 during and after the crisis. I selected the Southern-EU 4 for two reasons: (1) the Southern EU-4 were particularly negatively affected by the crisis, and from a policy
and an academic perspective it is highly interesting to give an insight into how this has
affected welfare convergence within the region. (2) Within the literature on welfare state development the Southern EU-4 are often categorized into one cluster: the Mediterranean welfare states. In the literature this club is often compared to other EU regions, but this study aims to give insights into convergence amongst the Southern EU-4. The Mediterranean taxonomy is not limited to academia, also in politics and media reports the Southern EU-4 states are often regarded as one group (NRC Handelsblad 2018; Ministerie van Buitenlandse Zaken 2014). The Mediterranean qualification is -among others- based on the idea that the role of the family, the Catholic church, and universal benefits, like for example general, non-means tested child support, are important factors in the provision of social protection in these
9 countries (Ferrera 1996 & Gal 2010). Next to these traditional domestic similarities, external factors exist that are said to affect social development in the EU. Two main concepts can be distinguished: similar problem pressure and Europeanization. By the former, the literature understands circumstances that arise due to increased economic integration, and economic developments that demand a similar response. In the time span of this study the most striking events that fall under this category are the events that followed on the crisis.
Europeanization knows various definitions in the literature. In this study Radaelli’s 2003 definition will be used as guideline: ’Europeanization consists of processes of a) construction, b) diffusion and c) institutionalization of formal and informal rules, procedures, policy paradigms, styles, 'ways of doing things' and shared beliefs and norms which are first defined and consolidated in the EU policy process and then incorporated in the logic of domestic (national and subnational) discourse, political structures and public policies (2003).’ The attempt of legislatively harmonizing social dimensions, or the fiscal reform conditions that accompanied financial support during the crisis can be considered examples of Europeanization.
Now ten years after the onset of the financial crisis it is relevant to find out how the Southern EU-4 welfare systems developed, what drove this development, and what theoretical framework can best explain it. Some academics argue that the financial and debt crisis have pushed Europe on a path of convergence, whereas other say the crisis has led to increased heterogeneity (Bongardt and Torres 2013; International Labour Organization 2016; European Parliament 2018). This study contributes to the debate by providing a quantitative analysis of various indicators of social development in the Southern EU-4 countries over the period 2006-2016, and thereby give more insights into the possible differences or similarities in their development. I use a number of indicators that cover different areas of the social dimension. By combining data from the aggregate, program and individual levels of social protection, a first exploratory indication will be given on the development of welfare systems in the Southern EU-4. The indicators are: total public social expenditure, the aggregate pension replacement ratio and net unemployment replacement rates, active labour market policy (ALMP) expenditure, tertiary educational expenditure, family- and household benefits expenditure, people at-risk-of-poverty rates, and an economic inequality indicator: the Gini-coefficient index. By looking at the annual changes of these parameters in the Southern EU-4 countries, an exploratory insight into the process of social convergence in the Southern EU-4 can be provided. By employing a widely recognized method for studying convergence,
-10 convergence, I will measure trends of convergence for the various indicators. (Barro and Sala-i-Martin 1992).
The study performed here contributes to the literature on social convergence in two ways: (1) the recency of the time period: the period 2006-2016 is highly suitable for shedding light on social development before and after the financial and euro crises. The crisis affected the Southern EU-4 in comparable ways, and such similar problems can according to the literature be a driver of convergence (Iversen and Cusack 2000; Starke, Obinger and Castles 2008; European Commission 2015). (2) This work contributes to the scope of the European welfare debate because most studies look at convergence between the South and the North, the East and the West, or between old and new EU member states, where this study sheds light on convergence within the European South. In light of the similar problems that were forced upon the Southern EU-4 by the crisis in combination with the Europeanization that took place in the aftermath of the crises, I hypothesize that social convergence has occurred within the Southern EU-4.
This study consists of five sections. First, the position of the academic debate is considered by reviewing the most important authors and their theoretical considerations in the field of European welfare state convergence. This section will also cover considerations on measurement indicators and their linkage with theory. Secondly, an overview is given of notable Europeanization that aimed for harmonizing social policy in the EU since 2000. In addition to this, I describe how the financial- and euro crisis affected European economies and social systems, and how this resulted in similar external problems for the Southern EU-4. In the subsequent section, my selection of variables as well as the economic concept of -convergence is explained more in depth. Hereafter the results are presented, followed by an analysis on whether or not convergence has occurred within the Southern EU-4 countries over the period 2006-2016. In the final part concluding remarks are made, accompanied by some thoughts on policy implications and future research.
11
2. Theoretical considerations on welfare state convergence and divergence
Within the welfare state literature, a number of theoretical frameworks exist that guide academics when addressing welfare state development. These frameworks help authors with selecting indicators for measuring welfare state development and explaining welfare state behaviour in general. Although the interconnectedness between theory and the measurement considerations that flow out from it is clear, I decided to discuss these concepts in two different paragraphs. Measurement considerations can be viewed as ‘smaller’ units of analysis that fall under the umbrella of theory. Therefore, I find it relevant to first discuss the overarching theories before I explain how mine and other’s measurement choices fit into these ideas. As mentioned in the introduction two opposing perspectives prevail within the debate on welfare state development: those who advocate convergence occurs over time, and those who believe national trajectories preserve uniqueness of welfare systems. I will first discuss theories that stress convergence occurs over time.
2.1 Theories on convergence
Robert Goodin, an authority among welfare scholars, puts autonomy, social stability, and poverty alleviation forward as the central motives for welfare. He argues that welfare is invoked to protect the vulnerable members of society against those in control over a country’s resources, with these three values as guiding principles (1988). Besides these intrinsic values, the literature says external factors may decide how welfare states are structured: international developments like increased economic integration, and legislation stemming from international treaties or organizations may influence the development of social policy as well (Iversen and Cusack 2000; Scharpf 1999).
In his widely acclaimed book ‘Has Globalization Gone Too Far?’ Dani Rodrik argues that internationalization has put increasing pressure on the functioning of the welfare state (1997). In light of globalization, the mantra among politicians seems to be that preserving a country’s competitiveness in the global economy is of the highest priority. Deregulation, liberalization, privatization, tax-cuts and other neoliberal tools are viewed as imperative for the preservation of a country’s level of competitiveness. These instruments may go at the expense of social provision since governments are likely to receive less tax-revenue when pursuing such policies, creating less fiscal room for extensive social policy. It is thus argued that national governments react to the level of welfare benefits in neighbouring countries. According to the literature the above described struggle for international competitiveness may lead to a so-called ‘social race-to-the-bottom’ (Meyer 2000; Dahlberg and Edmark 2004). By
12 this the literature understands a general retrenchment of social services that is motivated by an international encounter for competitiveness. In addition to this idea of retrenchment, Fritz Scharpf argues that a social race-to-the-bottom is more likely to occur in the Economic and Monetary Union (EMU). The line of reasoning behind this is that in a single currency area, countries can only increase their competitiveness with supply side policies as devaluation of the national currency is not on the menu (Scharpf 2011). More than their Northern counterparts, the Southern EU members -in particular Italy- used to regularly devaluate their currency to the Deutschmark in the pre-euro era (Demertzis, Efstathiou and Matera 2017). In light of the Southern EU-4 the social race-to-the-bottom theory should thus be considered when addressing welfare development in the Southern EU-4.
In the literature it is often argued that globalization and high levels of international trade result in winners and losers (Rodrik 1997). The losers of internationalization may demand compensation for their worsening situation in society and this creates a political incentive to offer this compensation (Sener, Bayrakdar and Hacioglu 2015). In the literature, empirical evidence shows that the large welfare states of countries like the Netherlands, Denmark, and Sweden, are correlated with these countries’ high openness to the global economy (Rodrik 1998). The logic behind these large welfare states is that losers of globalisation receive social compensation for their deteriorated position stemming from globalization. This empirical evidence for the compensation of losers of globalization has led to the construction of the compensation theory (Sener, Bayrakdar and Hacioglu 2015).
The social race-to-the-bottom- and compensation theory thus attempt to explain choices national governments make with regard to their welfare systems in light of international economic integration. In the literature, the social-race-to-the-bottom- and compensation theory are often accounted for by incorporating data on the openness of a country’s economy. A popular indicator for interpreting the openness of an economy is the sum of annual exports and imports in a given country (Paetzold 2012). In the fourth paragraph I elaborate more in depth about the openness of the economies of the Southern EU-4. In case of a social-race-to-the-bottom, convergence towards lower social levels is expected, whereas the process of social compensation may lead to upward convergence or preservation of the status quo.
Next to regulatory competition countries experience in a globalized economy, the literature suggests other mechanisms that may bring about welfare state convergence. Iversen and Cusack stress that similar problem pressure can lead to convergence of social systems among countries (2000). By similar problems the literature understands challenges like
13 increasing public debt, social risks that accompany changing labour markets, declining economic growth, and demographic changes like the ageing of societies (Starke, Obinger and Castles 2008). In an internationally integrated economy, such challenges often have a transboundary character and can therefore affect social systems in neighbouring countries in a similar way. This may in turn result in similar policy responses triggering patterns of convergence. Within the literature social risks are increasingly divided into old and new social risks. By old social risks the literature understands unemployment, sickness and old age, whereas child- and youth poverty, low educational participation and low female employment rates are examples of new social risks (Taylor-Gooby 2004). This distinction between new and old social risks fits in a trend in the welfare research community wherein is argued that governments states are gradually shifting their focus from poverty alleviation to investing in social capital. In order to respect this distinction, I incorporate indicators on new and old social risks, so that I can find out whether these dimensions of social policy show varying patterns of convergence. The fourth paragraph discusses the more technical aspects of the indicators used.
Besides challenges arising from increasing international economic integration, it is argued in the literature that social convergence may also come about by intergovernmental or supranational factors. The most compelling example of this dynamic I already mentioned in the introduction: Europeanization; the process of rules, beliefs and norms created at the EU level incorporating into the logic of domestic discourse, political structures and public policies (Radaelli 2003). Following the literature, Europeanization can be distinguished into binding
hard regulations, and non-binding soft rules (Milotay 2017). Although the history of the EU
has witnessed numerous policies aiming for harmonizing the social dimensions of its member states, these are often regarded as having minimal enforcing power (de la Porte and Heins 2014). The Europe 2020 strategy, the Social Investment Package and the Youth Guarantee are examples of social policy frameworks that are created to enhance social cohesion in the EU but work mainly on a voluntary basis via the Open Method of Coordination (OMC) (de La Porte and Heins 2014). Within the OMC, social harmonization is aimed for by sharing best practices, and mutual learning, mechanisms generally known for having little enforcing power.
However, in the aftermath of the financial- and euro crisis, a series of so-called ‘bail-out policies’ imposed highly restrictive austerity measures on debtor states within the Eurozone in exchange for financial support (de la Porte and Heins 2014). These acutely created fiscal instruments influenced the social dimensions of the debtor states as well (de la
14 Porte and Heins 2014). Within the sample of this study, Greece, Portugal and Spain faced such compelling reforms between 2006 and 2016. The pressure of austerity that accompanied these bail-outs should be considered when addressing the evolution of social systems in these countries. In chapter three I discuss more in depth how Europeanization took place in practice in the countries of my sample. For now, it is important to stress what theory says about Europeanization. Earlier scholars argued that Europeanization can lead to social convergence
as well as divergence (Radaelli 2000; Van Vliet and Goudswaard 2010). As mentioned above,
Europeanization in social areas is not always of a compulsory nature, and this may instigate divergence when some Member States follow-up on EU social policy more than others. Thereby, Europeanization is not always EU wide, but can also be targeted at specific countries. This may lead to regional convergence but can have diverging effects on the EU as a whole. Since the onset of the crisis the EU has witnessed various policies that were targeted at specific member states, with the bail-out policies probably being the most compelling. Since the Southern EU-4 were in many cases subject to these targeted policies, there is reason to believe this could have led to regional convergence of the Southern EU-4. It is therefore a scenario to consider when interpreting the results further on in this study.
To summarize, following the literature welfare state convergence can come about through different means: a social-race to the bottom may herald downwards convergence, social compensation may lead to upward convergence or preservation of the status quo, similar problem pressure and economic integration may generate upward or downward convergence, and Europeanization can lead to upwards convergence as well as divergence. The next section discusses theoretical ideas that advocate the preservation of national uniqueness of welfare states.
2.2 Theories on persistence and divergence
Besides studies that argue welfare states alter and converge over time, an extensive literature argues there are traditional and preserving variances between welfare states. In 1992 Gøsta Esping-Andersen made a sizable contribution to the controversy in the welfare debate by proposing the idea of welfare state regimes. He argued that Western welfare states can be divided into three categories: corporatist, liberal, and social democratic welfare regimes. Esping-Andersen based his classification on levels of decommodification, social stratification and the state-market relationship. Corporatist welfare states are characterized by attaching great value to the role of the family. Benefits tend to be family-focused, and states only step in those areas where and when family structures are not able to provide social protection.
15 Examples of corporatist regimes are France, Austria and Germany. In liberal welfare states, like the United States (US), United Kingdom (UK) and Australia, the state tends to have a small role in the provision of social protection, and the assistance that is provided by the state is often means-tested. In these welfare systems -like one would expect from a liberal system- self-reliance is a guiding principle. Social democratic welfare regimes, which are seen in Northern European countries like Denmark, the Netherlands, Sweden and Norway, are said to be pursuing the highest social standards. These regimes are known for providing a mix of universal and means-tested benefits, and besides solving poverty, creating equal societies is a high priority in these countries (Esping-Andersen 1992).
In the years following Esping-Andersen’s seminal work, his welfare regime taxonomy became a widely acclaimed and useful analytical tool for studying and addressing welfare state development in the Western world (Ferragina, Seeleib-Keiser and Spreckelsen, 2014). However, some scholars criticized Esping-Andersen’s contribution for having some essential shortcomings. Wil Arts and John Gelissen stressed for example that pure welfare regimes hardly exist (2002). They argue welfare systems are more likely to manifest themselves in the form of hybrids, combining elements of the three regimes proposed by Esping-Andersen (2002). Maurizio Ferrera argued a distinctive Mediterranean regime should be incorporated into the model of welfare regimes, and this has found support in the welfare state literature (Ferrera 1996; Gal 2010; Paetzold, 2012; World Bank 2015). The argument for the distinctiveness of the Mediterranean welfare states is based on typical features these countries are said to share: a large role for the family in providing social assistance in combination with universalistic programs, and highly protective policies for the labour market insiders: once employed, workers tend to be hard to replace (Ferrera 1996). Additionally, when categorizing the Mediterranean cluster, scholars often point to the influence the Catholic church -and the Greek Orthodox Church equivalently- exerts on the social doctrine. It is argued that the emphasis on the responsibility for the family in social provision leads to political inaction in some social policy areas in the Southern EU-4. (Marí-Klose and Moreno-Fuentes 2013). John Gal distinguishes more characteristics for the Mediterranean cluster. He finds dualism, a clientelist political culture, late industrialization and overall ineffectiveness of the welfare state as typical features for the Southern EU-4 welfare regime (2010).
In light of the debate on Esping-Andersen’s welfare regimes Moreno-Fuentes and Mari-Klose questioned whether it is still justified to place the Southern EU-4 countries in one cluster. By studying political documents, and providing insights in political decisions and institutional design, they found numerous differences in the Southern welfare states and
16 concluded the Mediterranean mapping is outdated (Moreno-Fuentes and Marí-Klose 2014). Since the new millennium Greece, Italy, Portugal and Spain developed asymmetrically because of different responses to global pressures and increased European integration the authors argue. An example they offer is the development of Portugal’s labour market. By Eastern European EU enlargement Portugal’s high share of labour-intensive manufacturing faced an unforeseen decrease in demand due to low-wage competition from the East. This development drove its economic and social status quo away from its three ‘counterparts’, the authors argue (2014). In contrast to the authors I discussed earlier, Marí-Klose and Moreno-Fuentes thus find that increased economic integration and external challenges can also lead to social divergence. This conclusion is symbolic for the disagreement that exists among scholars on the developments of the welfare state. Although Marí-Klose and Moreno-Fuentes bring up plausible arguments, they supported their conclusions with minimal quantitative data. The study conducted here will build on the findings of Moreno-Fuentes and Mari-Klose and author authors by providing quantitative data on the development of the welfare states in Southern Europe and attempts to give an exploratory description of social development in the Southern EU-4 up to 2016.
Furthermore, two other influential schools of thought exist that advocate national uniqueness of welfare states: the ‘old politics’ theory, and the ‘new politics’ theory. The former theory is structured around the idea that welfare states may indeed face similar problems, but these are mediated differently due to unique national political situations, which can lead to patterns of divergence. Partisan composition of governments and the interplay between the state and interest groups (e.g. trade- and labour unions) are concepts this school of thought points to as main causes for persistence and/or divergence (Garrett 1998; Huber and Stephens 2001).
The new politics theory advocates that existing theoretical ideas on welfare state development are outdated: welfare states have fundamentally changed since the existing theories were established, it is argued (Starke, Obinger and Castles 2008). Paul Pierson, in the literature generally viewed as the most influential proponent of the new politics school, argues welfare states are largely unaffected by developments brought about by internationalization. Instead of pointing to national political factors like ideological positioning or partisan composition -as the old politics theorists do- Pierson argues welfare state persistence is caused by the fact that policymakers are pursuing blame avoidance strategies. Cutting down on social provision is in general an unpopular move, and for electoral reasons politicians are reluctant to pursue such strategies. Additionally, Pierson points the firmness of the institutional status
17 quo. The political-institutional architecture is in many Western societies so complex and interconnected with numerous stakeholders that radical welfare change is hard to achieve (Pierson 1996).
To summarize, within the literature a number of theoretical ideas exist that advocate the preservation of national uniqueness of welfare systems. Esping-Andersen’s welfare regime taxonomy states that institutional and cultural traditions of welfare states preserve their national uniqueness over time. Additionally, the literature states that international challenges and developments may affect national socioeconomic structures in different ways, which in combination with the uniqueness of national political structures can lead to patterns of divergence. The final form of internationally imposed policy may thus vary between countries, what potentially can lead to social divergence.
The above part has showed that theories on welfare state development in the Western world are conceptually helpful but cannot fully address welfare state development. Of course, the theoretical considerations discussed above create expectations on the directions and levels of social convergence, but the study of welfare state development remains to a large extent an empirical issue, as Caminada, Goudswaard and van Vliet argue as well (2010). The next section will discuss the findings of earlier studies on welfare state convergence, what choices on measurement techniques authors make to come to their conclusions, and how these choices often lead to discussion.
2.3 Measurement considerations
As mentioned earlier, convergence in the EU is an often-studied topic within the welfare state research community. Within the welfare literature, however, little consensus exists on whether social convergence has occurred or is occurring in the EU, what the driving forces behind welfare state development are, and in what direction welfare states are developing. Authors in support of the convergence hypothesis, argue that welfare states do show signs of convergence over time, whereas other authors argue that welfare states tend to remain idiosyncratic due to domestic institutionalism and political traditions and legacy. The latter group’s arguments are in line with the path-dependency theory. The disagreement in the welfare literature is partly attributable to the vast size of welfare states. Because of the dynamic and enormous character of welfare states determining how to measure their development is a daunting task, logically creating plenty room for discussion. As mentioned earlier, Christian Green-Pedersen called this the ‘dependent-variable problem’ (2007). He stated that disagreement on welfare state development is a problem of conceptualization
18 rather than of data: different measurement techniques yield different conclusions, he argues. Below I discuss earlier quantitative studies on welfare state convergence in the Western world, and review some of the vast number of indicators these studies used. It becomes clear that in the literature plenty different conclusions were reached on whether convergence occurs or not. For reasons of clarity and oversight I make a distinction between social expenditure indicators and social outcome indicators. The former category is informative of the effort and level of generosity a government puts in its welfare systems, whereas the latter says something about the actual social situation in a country. Besides discussion on what indicators to use, the literature proposes various ideas on what statistical techniques are best for indicating welfare convergence. The most prominent of these will be discussed below as well.
2.3.1 Social expenditure indicators
Within the welfare state literature, expenditures on social benefits are popular indicators for addressing social convergence, and social development in general. I first discuss earlier findings on social expenditure indicators, before I review findings on social outcome indicators.
In 2008, Starke, Obinger and Castles published a study on welfare convergence amongst OECD countries between 1980 and 2002. The aim of their study was to discover the broad trajectories of welfare state development in the OECD, and therefore the authors incorporated a large set of variables. In order to measure convergence, Starke, Obinger and Castles employed two widely acknowledged tools for measuring convergence: β- and σ-convergence (2008). The former measures σ-convergence by looking at the pace by which lagging entities (countries) catch-up with more advanced entities. σ-convergence, on the other hand, measures convergence by determining the reduction in disparities between entities over time - growing together (2008). These concepts were first introduced by Barro and Sala-I-Martin in 1992 in a paper on the sources of economic growth and found great resonance in the literature on welfare state convergence (Starke, Obinger and Castles 2008; Caminada, Goudswaard and van Vliet 2010; Schmitt and Starke 2011; Paetzold 2012). In their selection of variables Starke, Obinger and Castles included a number of expenditure indicators: total social expenditure in cash and in kind, pensions, health programmes, and unemployment- and family benefits. For total social expenditure the authors find evidence for decreasing disparities between countries over time (σ-convergence). Furthermore, the authors also find evidence for welfare states with initially lower levels (e.g. Portugal, Greece, Japan) of social
19 expenditure catching up (β-convergence) with high initial spenders (e.g. Belgium and the Netherlands). However, the authors recognize that total public social expenditure does not provide a nuanced indication of welfare expenditure. For this reason, Starke, Obinger and Castles broke total social expenditure down into health, family- and unemployment benefits, and pension expenditure. For the first three indicators the authors found σ- as well as β-convergence, but for pension expenditures the authors surprisingly found patterns of divergence over the period 1980-2003 (Starke, Obinger and Castles).
Caminada, Goudswaard and van Vliet published a comparable study on European welfare state convergence in 2010. Different to the study performed by Starke, Obinger and Castles, these authors invigorated their argument by studying welfare state development on three different levels: macro, programme and individual. The authors studied social convergence in the OECD countries over the period 1980-2002. Non-EU OECD-countries were included in their study for the aim of distinguishing European and global trends. Goudswaard and van Vliet employed β- and σ-convergence as well to measure patterns of convergence. On the macro-level they studied total public- and total private social expenditure as a percentage of GDP. Convergence on the programme-level was determined by incorporating data on expenditure on various social programmes, like active labour market policies (ALMP), disability, pension, health, and family policies. Developments on the individual level were measured by net pension- and unemployment replacement rates, and minimum social assistance levels. Although incorporating numerous variables on the macro, programme and individual certainly adds up to the academic credibility and completeness of a study, it may at the same time go at the expense of a study’s generalizability. In the case of Caminada, Goudswaard and van Vliet’s contribution this is to some extent the case. The authors reach a multifaceted conclusion wherein they argue that in some countries or areas strong signs of convergence were observable, whereas other countries and areas showed signs of divergence. Old age expenditures showed for instance strong signs of convergence, and the same goes for expenditure on active labour market policies. The former finding is especially interesting considering the fact that Starke et al. found patterns of divergence for old age expenditure (2008).
Another interesting finding is that the countries within the EU showed stronger signs of convergence than the OECD countries as a whole. This is indicative of an effect of Europeanization, the authors argue. The authors also find a quite strong convergence of social expenditure in EU member states over a lengthy period of time, but this development tends to have stalled in the final period of the time span (2010). Net unemployment rates showed clear
20 signs of convergence, a trend observable as well for poverty rates and poverty gaps. On the contrary, social assistance benefits moved to substantial lower levels in some countries than in others, logically hindering convergence. These contradictory findings in a single study underline the complexity of the dependent variable problem and depict the need for thoughtful decisions on what indicators to use. In the literature on welfare state convergence there can be observed that some authors choose to study a broad spectrum of the social dimension in order to get the big picture (Starke, Obinger and Castles 2008) whereas others are more after the causal mechanisms that underlie welfare development. The study performed here can be placed in the former category. An example of the latter is discussed below.
Jorg Paetzold examined welfare state convergence in 14 EU countries (EU-15 minus Luxemburg) by studying three social expenditure indicators over the period 1980-2005. The indicators Paetzold used were the percentage of GDP spent on total public social expenditure (SE), public pension net replacement rates, and unemployment net replacement rates. Paetzold chose to use a relatively narrow set of social indicators but included a large set of independent variables to determine what causal mechanisms were behind social development in the EU-14. Total public social expenditure can be viewed as an aggregate indicator that says something about all spheres of public welfare (Paetzold 2012). Paetzold uses pension net replacement rates because of the large share of total social expenditure they make up. Unemployment net replacement rates are selected because Paetzold regards those as being sensitive to policy changes, making them suitable for determining convergence (2012). Paetzold found that for total public social expenditure convergence has occurred between 1980 and 2005. For the replacement rate indicators, the results were less clear cut: unemployment net replacement rates barely showed signs of convergence whereas pension rates did. Paetzold theoretically explained this variation in results by pointing to Southern member states ‘catching up’ with Northern member states in the field of pension benefits. Unemployment replacement rates on the other hand are much more sensitive to domestic institutional arrangements, Paetzold claims, and do therefore not behave in line with the other variables (2012). According to Paetzold’s findings, pension and total benefits are thus more sensitive for convergence than unemployment benefits. In his search for causality, Paetzold included a large set of independent variables he incorporated into an econometric model.
Another study, done by a research team of the International Labour Organization (ILO) over the period 2000-2015, presents a more unilateral conclusion on socioeconomic development in the EU. The researchers argue that EU social standards have in general seen a cutback in generosity over the past decades (ILO 2016). Thereby, the authors argue that the
21 economic crisis has highlighted the heterogeneity of the social systems amongst member states which led to patterns of divergence (ILO 2016). To reach these conclusions, the authors studied a broad set of social indicators during the period 2000-2015. Instead of employing both β- and σ-convergence tests, the ILO researchers determined patterns of convergence by doing σ-convergence tests and studying the variation in the ratio between the lowest and highest performing member states over time. The authors found for example that in 2007 the lowest to highest unemployment ratio was 1 to 3 (Denmark and Slovenia), whereas this ratio increased substantially to 1 to 5 in 2015 (Germany and Greece), indicating divergence in this area. For social expenditure, the authors studied expenditure on ALMP’s, and came to the conclusion that patterns of convergence were observable in this area.
Besides a distinction between macro, programme and individual levels of social benefits, there is also a growing group of scholars that emphasizes the difference between social protection, like poverty reduction, and social investment, like educational policies or ALMP’s (Kersbergen, Vis and Hemerijck 2014; van Vliet and Wang 2015). The literature suggests that Western welfare states are transforming from a traditional- to a social investment model, which means that welfare states are increasingly focused on investing in citizen’s competences rather than just relieving poverty (van Vliet and Wang 2015). A wedge in redistributive and active policy seems to be on the rise in the Western world, and it is argued that this is related to the growing attention for new social risks (Taylor-Gooby 2004). In line with this proposition the World Bank warns for a growing gap between labour market demands and educational programs and insists therefore on educational investments to overcome this gap (2015). Currently, large differences exist in how much European welfare states spend on social investment, and this thus indicates a difference in poverty reduction and reducing inequality as aim, which is interesting in the light of Esping-Andersen’s welfare regimes. Sweden and Germany are for example strongly committed to both, whereas France prioritizes poverty reduction over battling inequality. The Southern states are in general not fairly known for this (Ferragina, Seeleib-Keiser and Spreckelsen; 2014). The study performed here recognizes the social investment dimension and incorporates data on ALMP- and educational expenditure.
Above, I described how welfare scholars used social expenditure indicators as a means to address welfare state development. Next to expenditure, scholars also address welfare development by studying how countries score on indicators that are indicative of social performance. The incorporation of social outcome indicators is relevant as it says something
22 about the relation between policy input and output, and thus about the efficiency of the welfare state. Below I discuss how such social outcome indicators are used in the literature.
2.3.2 Social outcome indicators
When studying welfare states, a distinction should be madebetween input and outcomes, the literature argues (European Parliament 2018; Pavolini et al 2014). The relation between input and outcome namely says something about the efficiency of redistribution of the state. With input, scholars mean the amount of tax money and efforts spent on unemployment, health, or labour policies. By outcomes, academics mean the actual performance of societies at large expressed in indicators such as poverty and unemployment rates. The importance of social outcomes is derived from the fact that they are regarded as essential for the legitimacy of the welfare state (Ferregina, Seeleib-Kaiser and Spreckelsen 2015).
Lopez-Bazo et al. delivered an interesting contribution to the quantitative welfare state literature in 1999. The academics argued that country averages are not a suitable indicator for measuring convergence because of the large disparities within countries (Lopez-Bazo et al. 1999). A well-known example provided by the authors is the large differences in wealth between the rural South and industrial Northern regions of Italy. They advocate a more regionally focused perspective instead of a nation-state perspective, comparable to the EU NUTS regions. Although the added value of a regional perspective speaks for itself, this study focuses on nation states. The reason I bring up Lopez-Bazo et al. is of a different nature: I want to shed light on their selection of variables. Their study is interesting because the authors study GDP per worker and GDP per capita. Comparing these two indicators may say something about the distribution of wealth and income within countries’ borders. In the light of Esping-Andersen’s welfare regime typology this is an interesting consideration, since he stressed reducing inequality as a distinguishing feature for welfare in his welfare typology: a characteristic not typical for the Mediterranean cluster. By incorporating the Ginicoefficient -an often used indicator to measure a country’s economic inequality- in my study I try to find out how the Southern EU-4 countries score in this area, as this may thus say something about the justification of clustering the Southern EU into a single Mediterranean group.
I already mentioned the 2010 study by Caminada, Goudswaard and van Vliet in the above section on social expenditure indicators. In that same study the authors also incorporated indicators on social outcomes: the at-risk-of poverty rate after social transfers, the poverty gap and the poverty rate. The former is an official EU indicator for social cohesion and is defined as the rate of people that have a disposable income of 60% of the
23 national median disposable income (Eurostat). The latter two are official OECD indicators to measure poverty. The poverty gap is defined as ’the ratio by which the mean income of the poor falls below the poverty line. The poverty line is defined as half the median household income of the total population (OECD).’ The OECD defines the poverty rate as ‘the ratio of the number of people (in a given age group) whose income falls below the poverty line; taken as half the median household income of the total population.’ These indicators thus attempt to say something about the extent to which welfare states can actually foresee in combating poverty. The authors found convergence for all three poverty indicators (Caminada, Goudswaard and van Vliet 2010).
The 2016 ILO report mentioned earlier also studies indicators on social outcomes in order to measure socioeconomic convergence in the EU. Just like Caminada, Goudswaard and van Vliet, the researchers incorporate the at-risk-of-poverty rate. In addition to this official EU indicator, the authors also include the Gini-index and the median income as indicators for social outcomes. In contrast to Caminada, Goudswaard and van Vliet, the ILO researchers found patterns of divergence for all three social indicators in the EU-28 (2016). This can be due to the different timespan, and different methodology the ILO team used.
Measuring social convergence is not limited to academia. The European Commission uses an own set of indicators to measure convergence within the EU. This set of indicators falls under the scope of the in 2011 established Macroeconomic Imbalances Procedure (MIB). With regard to the social dimension the Commission uses the following ‘scoreboard indicators:’ 3-year change in the youth unemployment rate, 3-year change in the long-term unemployment rate, and 3-year change in the activity rate (European Commission 2019). With respect to the MIB scoreboard indicators, the Commission does thus not include indicators on social expenditure of Member States. This can be seen as a limitation, since the relation between social input (levels of expenditure) and social outcomes can say something about the efficiency of the welfare state within a Member State.
In the case of welfare state development studies in the EU, not only numeric indicators draw the attention of scholars. There is an extensive qualitative literature that attempts to address how EU policy affects the social dimension of Member States. De la Porte and Heins (2014) and Pavolini et al. (2014) studied the institutional dimensions behind social integration of EU welfare states and created a typology of how to address three different forms of infringement: objectives, surveillance, and enforcement. In light of these concepts chapter 3 will discuss how Europeanization may have altered the European social dimension and what
24 international economic developments may have affected welfare state development in the Southern EU-4 in the period 2006-2016.
2.4 Theoretical summary
Before I discuss the for this study most important trajectories of Europeanization and economic integration, I give a brief overview of the above discussed theoretical considerations and indicators. According to the literature, welfare state convergence can come about in different ways: (1) similar external problem pressure, (2) economic integration, and (3) Europeanization. Authors that advocate idiosyncrasy of welfare states point to national political-institutional theories like (1) welfare regimes, and (2) theories of old- and new politics as being responsible for national persistence.
In order to find evidence for these theoretical considerations, scholars use a large variety of indicators to measure welfare state development. Social expenditure indicators are the most popular and can be broken down into various programs and benefits that represent different dimensions and levels of analysis of the welfare state. Scholars regularly incorporate data on old age, health, unemployment, educational, and labour market expenditures. These different expenditure indicators are informative for the different levels and dimensions of social development, like generosity, social investment, and aggregate, programme and individual. Besides social expenditure indicators, scholars include data on social outcomes, as this may say something about the efficiency of a welfare system.
An adequate selection of indicators is dependent on the type of study a researcher wants to perform: explorative or explanatory. Explorative studies are often of a large-N type, whereas explanatory studies tend to be of the small-N type. In order to address the complex puzzle welfare state development is, both types of research are necessary. The study performed here belongs to the first category.
25
3. Europeanization and economic developments
This chapter gives an overview of noteworthy courses of (social) Europeanization that have taken place before and after 2006 in the EU, and in particular in the Southern EU-4. Within the literature it is acknowledged that convergence can be a result of Europeanization and it is therefore relevant to discuss the most important policies the EU has witnessed in the timeframe used in this study (Radaelli 2000). Next to giving an overview of Europeanization, trajectories of European economic integration, and significant economic developments will be discussed, as the literature views these as potential drivers of social convergence as well (Iversen and Cusack 2000; Taylor-Gooby 2004). The purpose of this chapter is thus shedding light on the contextual developments in which the Southern welfare states developed from 2006 to 2016. I will first elaborate on the most important pre-crisis developments before discussing the events that occurred in the wake of the crisis. As dividing point, I take the October 2009 discovery of Greece’s fiscal imbalances that preluded the Eurozone debt crisis.
3.1 Pre-crisis developments
In this paragraph I discuss economic developments and trajectories of Europeanization that may have contributed to reducing social disparities within Europe in the period before the financial- and euro crises. Portugal and Spain only joined the European Community in 1986, so for reasons of demarcation I take this year as starting point of my comments.
3.1.1 Economic developments
In 1986, the year Spain and Portugal accessed the Union, the Single European Act (SEA) was launched. The SEA, by many viewed as the first major alteration of the founding Treaty of Rome, paved the way for the completion of the internal market. The SEA made it easier for the European institutions to pass laws, emphasized the ambition for integration, and advocated a path of deregulation, that was fostered by – amongst others- Margaret Thatcher (James 2015). These developments indirectly enhanced a more open, and freer European economy, which eventually resulted in new challenges for the social systems of the Member States.
The biggest game changer for European economic integration was the launch of the single currency, the euro, in 1999. Among EU policymakers a belief existed that this significant step towards more economic integration would spontaneously lead to more integration in other areas via spill-over effects (Busch et al. 2013). However, the events that followed on the 2008 crisis generated widespread scepticism on this idea, as I will discuss
26 more in depth in the next section. In 2004, the EU saw the largest enlargement in its history when ten, mostly Eastern European countries, joined the Union. This sudden increase in size presented new but similar challenges for the Southern EU-4. Increased wage competition with Eastern low-wage countries, potential capital flight, and the integration of financial markets are examples of these challenges (Marí-Klose and Moreno Fuentes 2013). Following the literature such challenges can lead to patterns of convergence as well as divergence (Iversen and Cusack 2000; Marí-Klose and Moreno Fuentes 2013).
3.1.2 Europeanization
Besides compelling economic developments that may have exerted influence on social development in the EU, the years before 2009 witnessed the creation of numerous EU policy frameworks that affected the social dimensions of the Member States. With the establishment of the EMU in 1992 the EU acquired mandate to intervene indirectly in fiscal and social policy of member states that embraced the euro. Instruments like the Eurozone’s Stability and Growth Pact (SGP) gave the EU far-reaching capabilities to measure and steer national fiscal policy by using compliance tests and EU wide benchmarks. For social policy, the capability to influence was of a lower level. However, fiscal as well as social policies are worth considering in this study since the former affects the budgetary space for national governments to set out their social policy. Since social expenditure makes up the biggest share of government spending, the pressure the SGP exerts on welfare states is considerable, De la Porte and Heins argue (2014).
The desire for social convergence in the EU was reflected in particular by the 92/442/EEC Recommendation of 1992 wherein convergence of social protection objectives was explicitly mentioned (Caminada, Goudswaard and van Vliet 2010). In 2000, the EU adopted the Lisbon Strategy in which the improvement of social cohesion was an important topic. The Lisbon Strategy also established the Open Method of Coordination (OMC). This policy framework respects the national sovereignty of Member States but attempts to inspire social cohesion by sharing best practices in the social dimension. By ‘policy mimicking’ and ‘contextual learning’ governments of EU Member States are encouraged to learn from and adopt aspects of each other’s social systems. The non-binding character of the method however, especially in comparison with the much more enforceable SGP, spurred the debate on the OMC’s impact on social convergence (Caminada, Goudswaard and van Vliet 2010).
In 2001, the Treaty of Nice was signed and came into effect in 2003. With the signing of this treaty the Member States agreed to actively take steps towards more social cohesion by
27 mapping common objectives, based on a then created set of social indicators (European Commission 2003). This set contained indicators on health, long-term unemployment, income inequality, life expectancy, and financial poverty (Caminada, Goudswaard and van Vliet 2010). Following theory, the above-mentioned Europeanization developments in the first decade of the 21st century may have set in motion a process of social convergence that still has an effect today.
Although I surely recognize the developments discussed here took place before 2006, they are still of relevance for this study because socioeconomic convergence is a long-term process (Starke, Obinger and Castles 2008). The next session will elaborate on the most important events that happened in the EU after the onset of the financial and euro crises.
3.2 Post-crisis developments
After the outbreak of the financial crisis in 2008, and the eurozone debt crisis in October 2009, the EU as a whole faced an economic downturn unwitnessed since the 1930’s great depression. In particular the Southern EU countries faced severe consequences of the euro crisis, which according to Zamora-Kapoor and Coller initiated economic as well as political crises in Greece, Italy, Portugal and Spain (2013). This section first gives a chronological overview of the most striking economic developments the EU, and in particular the Southern EU-4 faced in the wake of the crisis. Additionally, I address EU policy responses aimed at enhancing economic and financial recovery within a crisis-torn Union.
3.2.1 Economic developments
When the American investment Lehman Brothers filed bankruptcy in September 2008, and Greece’s fiscal deception was discovered in 2009, the EU was confronted by a deep and lasting economic recession. Skyrocketing public debts, unemployment rates, and budgetary deficits, and stagnating economic growth became the new reality for Europe in late 2009. Starke, Obinger and Castles underline the similar effects crises can exert on countries and argue that increasing public debt can create new social risks that put welfare states under strain (2008). According to the literature the events following the financial- and euro crises tremendously impacted the social dimension of the Member States, and in particular the Southern European Member States (De la Porte and Heins 2015; León and Pavolini 2014; World Bank 2015). The similar pressure such drastic events exert on welfare systems may trigger patterns of convergence (Iversen and Cusack 2000; Taylor-Gooby 2004; Marí-Klose and Moreno-Fuentes 2013).
