**Sozialstrukturanalyse**

# The Independent Variable Problem Katharina Kunißen

Welfare Stateness as an Explanatory Concept

## **Sozialstrukturanalyse**

#### **Reihe herausgegeben von**

Rasmus Hoffmann, Universität Bamberg, Bamberg, Germany

André Knabe, Rostocker Institut für Sozialforschung und gesellschaftliche Praxis e.V., Rostock, Germany

Christian Schmitt, Universität Rostock, Rostock, Germany

In der Reihe Sozialstrukturanalyse erscheinen Beiträge, die die Forschung zu sozialer Ungleichheit reflektieren oder neue Sichtweisen aufzeigen. Dazu zählen insbesondere qualitative sowie quantifizierende Analysen zur Verteilung von Ressourcen und Lebenschancen auf regionaler, nationaler und globaler Ebene sowie zur Wahrnehmung und Deutung von Ungleichheiten. Neben klassischen Dimensionen sozialer Ungleichheit werden dabei auch neuere Dimensionen wie Kultur, sozialräumliche Positionierung und die Einbindung in soziale Beziehungsnetzwerke betrachtet.

Die Reihe wurde von 2006 bis 2018 von Peter A. Berger (†) herausgegeben und 1994 von Stefan Hradil gegründet.

Katharina Kunißen

# The Independent Variable Problem

Welfare Stateness as an Explanatory Concept

Katharina Kunißen Mainz, Germany

Dissertation, Fachbereich 02, Sozialwissenschaften, Medien, Sport, Johannes Gutenberg University Mainz, 2021 (D77)

ISSN 2662-2947 ISSN 2662-2955 (electronic) Sozialstrukturanalyse ISBN 978-3-658-39421-9 ISBN 978-3-658-39422-6 (eBook) https://doi.org/10.1007/978-3-658-39422-6

© The Editor(s) (if applicable) and The Author(s) 2023. This book is an open access publication. **Open Access** This book is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this book are included in the book's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Responsible Editor: Stefanie Probst

This Springer VS imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH, part of Springer Nature.

The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

## **Acknowledgements**

I am deeply grateful for the great support and encouragement I have received from family, friends, colleagues, and institutions while working on this project.

First and foremost, I would like to thank my supervisors. Gunnar Otte, my main supervisor and head of the Social Stratification Research Group at Johannes Gutenberg University Mainz, has guided and supported me throughout the project. Patrick Sachweh from the University of Bremen has brought a valuable and enriching additional perspective as co-supervisor.

This project has benefited immensely from the academic exchange and financial support I have received as a member of the interdisciplinary Gutenberg Academy. In addition, research that led to these results was supported by the European Union's Horizon 2020 research and innovation programme under grant agreement No. 730998, InGRID-2—Integrating Research Infrastructure for European expertise on Inclusive Growth from data to policy. This has allowed me to spend time at the Swedish Institute for Social Research (SOFI) at Stockholm University. I would like to thank the team at SOFI for their great input. The Open Access publication of this manuscript received funding by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation; No. 499033181).

Since I started this project, many people have supported me greatly. I would like to thank my colleagues at the Institute of Sociology at Johannes Gutenberg University. In particular, Theresa Wieland, Dave Balzer, Debora Eicher, and Matthias Lehmann. I am deeply grateful to Lena Zimmer and Julian Hamann who have been by my side throughout the entire process—as wonderful friends and as academic peers. The same goes for Jasmin Fitzpatrick, who always had an open ear and great advice.

Finally, I would like to thank my family and friends for their unconditional support and my husband, Matthias, who was and is my rock.

Mainz 2021

Katharina Kunißen

## **Contents**




## **List of Figures**



## **List of Tables**


# **1 Introduction**

The welfare state is an integral part of western industrialised democracies. While its most obvious function is to provide a safety net for individuals, who are threatened by social risks, it also has far-reaching effects on a multitude of social phenomena, which seem only indirectly associated with social policy arrangements at first glance. Examples include the formation of attitudes and behaviours. Such relationships reveal that the welfare state is much more than a conglomerate of social rights: the arrangement of social policies actively shapes e.g. social stratification, incentivises behaviours, and conveys solidarity and justice principles. In this way, it bears the potential to influence almost all areas of social live. *Who is covered by policies? How is social security organised and financed? How generous is a welfare system? Does it aim at preserving status differences or does it promote egalitarian principles?* The answers to those and similar questions reveal important characteristics, which are quite consequential for individuals covered by a welfare state. Because of this strong tie between social policies, social inequality, and various other social phenomena, the welfare state is an important object of research in a variety of disciplines within the social sciences. Much research is focussed on the welfare state itself, but especially comparative approaches often highlight the *consequences* of different welfare state arrangements and ask *how far different social policies (as independent variables) lead to different outcomes*. Such outcomes can be found mainly on two analytical levels: the macroand micro-level. In this contribution, I focus in particular on outcomes on the micro-level and thus mainly argue within a *multilevel framework*. 1

<sup>1</sup> Many of the discussed issues should still be equally relevant for macro–macro analyses. This contribution focusses on the multilevel perspective, because the issues discussed in the course of this book are especially pronounced here.

<sup>©</sup> The Author(s) 2023

K. Kunißen, *The Independent Variable Problem*, Sozialstrukturanalyse, https://doi.org/10.1007/978-3-658-39422-6\_1

There is a great variety of outcomes examined in this regard: some research focusses on direct consequences of social policies like well-being (e.g. Cruz-Martínez, 2017; Schuck & Steiber, 2017) or the consequences of risk exposure (e.g. Angel & Heitzmann, 2015). Other scholars examine phenomena which are influenced in more subtle ways, such as attitude formation (e.g. Jordan, 2013; Eger & Breznau, 2017), political participation (e.g. Schneider & Makszin, 2014), political trust (e.g. Mattila & Rapeli, 2018), and much more. In such cases, an empirical operationalisation of the welfare state or of specific elements of social policy-making is always required. However, some substantial problems arise concerning prevalent operationalisation practices. Essentially, these problems all relate to one key issue: while there is a great number of contributions addressing the measurement of differences between welfare states per se and as a *dependent variable*, there is a distinct lack of feasible recommendations when it comes to the operationalisation of welfare stateness as an *explanatory variable* (or *independent variable*). In light of the great number of studies assuming an effect of welfare policies on other social phenomena, the lack of standardised proceeding we encounter in this context surprises. To this date, there is no systematic test of how such varying proceedings may affect results and their comparability. Similarly, a detailed conceptual discussion on which features of the welfare state are relevant for the explanation of specific outcomes is missing. This contribution intends to fill both gaps. It seeks to unravel the pitfalls and potentials of existing approaches and to propose a conceptual framework, which is intended to guide empirical operationalisations and may help to overcome some of the problematic issues that are identified.

#### **1.1 Aim of the Study and Research Questions**

At its core, this contribution takes up a very specific problem within a very broad field. While the treatment of the welfare state as an independent variable is indeed a small issue within the well-established field of comparative welfare state research, it is an important one in light of the great body of work that is affected. In the course of this book, I will argue that the lack of a standardised strategy in this case considerably impairs reliability and comparability of analytical approaches and empirical results, as the concepts behind the operationalisations are not sufficiently clear. This leads to one essential premise: we must be sure of the concepts we want to measure with an operationalisation if we are to measure them correctly.

In a way, this endeavour is related to a growing body of literature on *replications* in the social sciences and beyond. More specifically, it takes up issues that are raised by scholars who explore the reproducibility (or replicability) of results (Munafò et al., 2017; Breznau et al., 2019) and the transparency of scientific publications using cross-national survey research data (Damian et al., 2019). While those scholars mainly emphasise issues related more generally to the transparency of methodological choices and the rules that have to be followed when it comes to performing empirical analyses and interpreting results, my contribution focusses on one specific example: treating the welfare state as an *explanatory factor*. Furthermore, I approach issues from a more conceptual point of view. Thus, problematic practices such as p-hacking<sup>2</sup> and HARKing3 (Munafò et al., 2017: 1–2) are only brushed. Instead, the focus rests on the potential to avoid such sources of bias if an agreement on the process of operationalisation and the specification of underlying concepts is reached.

In order to narrow down the aim and research questions pursued in this book, it has to be discussed why the measurement of welfare states as an independent variable is an important issue and why its operationalisation is not (yet) standardised in this particular case. Starting with a very general observation, many comparative empirical studies dealing with individual-level outcomes implement a multilevel design4 in which properties of the welfare state serve as independent variables at the level of countries. As there is no agreed-upon way to operationalise 'welfare stateness' as an indicator in such studies, scholars usually borrow instruments from literature, which examines welfare policies as a dependent variable. These instruments include a broad spectrum of approaches such as a variety of different welfare regime typologies, single indicators, and composite measures. Even though the demand for treating the welfare state as an independent variable is high, the literature hitherto lacks comprehensive discussions of the extent to which the various existing measures *can* actually serve as suitable independent variables and the problems that may be associated with different operationalisations. In order to address these topics, it is necessary to have a brief look at the debate surrounding the general measurement of different welfare states.

<sup>2</sup> Modifying operationalisations until a significant value is obtained.

<sup>3</sup> Hypothesising after results are known.

<sup>4</sup> E.g. multilevel regression analysis or fixed effect models.

From early on, research on the welfare state has sparked lively and critical methodological debates.<sup>5</sup> More recently, the so-called *dependent variable problem* receives growing attention (e.g. Clasen & Siegel, 2007). This methodological debate emerged as a by-product of a discussion about welfare state change and retrenchment (e.g. Pierson, 1996). A key problem identified in this debate was the lack of a common understanding of what the object of research—the dependent variable—entails and how it should be measured (Green-Pedersen, 2004). Until today, there is an ongoing discussion based on the repeated observation that different conceptual and operational strategies produce different results (e.g. Kühner, 2007; Bolukbasi & Öktem, 2018).

In light of this existing debate on how to conceptualise and measure features of the welfare state best, one might wonder why we need an additional *independent variable* perspective instead of simply relying on the insights produced by literature addressing the dependent variable problem. Five arguments speak in favour of such an endeavour. First, there is no thorough account of how different conceptualisations affect explanatory power and informative scope when used as *independent* variables. Only recently, scholars start to voice concerns because existing measurements are treated as interchangeable options for the operationalisation of welfare policies as dependent as well as independent variables (Bolukbasi & Öktem, 2018). Second, the existing methodological discussions mainly remain on the *macro-level*. How far the proposed measures can be embedded in *macro-micro-*analyses remains unclear. Third, the exchange of *feasible recommendations* between general literature on the welfare state and research, which examines its outcomes, is highly underdeveloped. Systematic comparisons of varying strategies are rare and focus only on consequences of different ways to operationalise within one of the approaches and for singled out dependent variables (e.g. Howell & Rehm, 2009; Bergqvist et al., 2013). Fourth, difficulties in choosing an appropriate independent variable are frequently expressed in the literature and ultimate selections often entail *compromises*. 6 Fifth, it has never really been discussed or tested, whether the existing indicators adequately capture theoretically assumed *mechanisms* in multilevel analyses of the outcomes of welfare policies, even though concerns are voiced sporadically (e.g. Pfau-Effinger, 2005). Exploring in more detail, why features of the welfare state serve as independent variables, how they are empirically operationalised

<sup>5</sup> A more detailed discussion of the historical development of research on the welfare state follows in the second chapter of this book.

<sup>6</sup> For instance because indicators are not available for country samples covered by survey data. More specific examples for such cases are provided in the third chapter.

and whether different approaches influence results, should therefore be a helpful contribution to the state of research. The first research question that is pursued in this contribution addresses this issue:

*How comparable are the results that emerge from different approaches to operationalising the welfare state as an independent variable?*

The second research question is a logical consequence of the first. After the description and critical assessment of the status quo, it constructively focusses on how to potentially improve the inclusion of the welfare state as an explanatory factor:

*How can we derive a more standardised, transparent and comparable approach to operationalising the welfare state as an independent variable?*

Before exploring in more detail how this contribution approaches these issues, one restricting remark is necessary. Overall, the main aim of this study is not to provide a complete overview and final evaluation of existing and possible conceptualisations and empirical operationalisations of welfare stateness. Instead, it aims at exemplifying a possible approach to standardise how we conceptually and empirically include properties of welfare states in cross-cultural comparative analyses of their micro-level outcomes. In doing so, it takes up various important fields and functions of the welfare state, but it does not claim to be exhaustive.

#### **1.2 Remarks on the Analytical Perspective**

This book argues within one very popular analytical framework in cross-cultural research, which is the so-called *multilevel analysis*. It is thus necessary to clearly distinguish this approach from others within the field of welfare state research. Hence, this part of the introduction elaborates in more detail the analytical perspective that is pursued.

The choice of a methodological approach depends on two key questions: (1) *Which analytical levels do analyses incorporate?* In other words: does the study aim to explain an outcome on the macro-level (usually countries or regions) or on the micro-level (usually individuals). (2) *How many regional cases are analysed?* Here, studies comparing few cases in in-depth case studies or a large enough number to explore statistical variance between regional units are differentiated.

Since this book focusses on research, assuming the welfare state has an impact on *individuals*, I will restrict the following discussion to literature incorporating at least two levels (individuals and countries or regions). Regarding the second question, those research endeavours are emphasised, which explore differences between welfare states in a greater number of countries (such as all member states of the European Union). This means that two kinds of studies are treated only marginally. First, those focussing on only one level (usually the macro level). Examples for this are analyses of the impact of social policies on other policies (e.g. Guo & Gilbert, 2007; den Dulk et al., 2012) or on aggregated micro-level information (e.g. Mano-Negrin, 2004; Schneider & Makszin, 2014). Second, case studies (e.g. Alves, 2015; Oesch, 2015) are of secondary importance as the aim is not to confront a few examples in great detail but to explain variance within a sample of countries in a regression framework.

Testing research questions addressing at least two analytical levels requires a multilevel conceptualisation. Such macro-micro-models often draw on what is sometimes described as a transformed version of Coleman's (e.g. 1986) 'boat' (cf. also Lucas, 2016). Coleman himself distinguishes his approach from a limitation to the macro-level. His main objective is to formulate a theory, which incorporates micro- and macro-level phenomena and outlines transformations and processes between the two levels. Hence, he attempts to overcome limitations, which are especially prominent in macro–macro analyses following the logic of methodological holism. Coleman notes three shortcomings of such approaches: limited variation, limited insight about why a relationship exists, and the immanent premise that social orders exist7 (Coleman 1986: 1321–1322).

A graphical representation of his model is given in Figure 1.1. Instead of remaining on the macro-level, Coleman proposes to move down to the individual level, exploring how macro-level phenomena affect individuals on the one hand (type 2 relation) and how individual outcomes in turn shape macro-level processes (type 3 relation) on the other hand. In sum, such *methodological individualism* offers a framework in which relationships between the individual and the social context can be analysed in more detail than if one remains at only one level.

Even though Coleman was more interested in the opportunity to explore type 3 relations, this effect is usually not of key interest in the kind of multilevel approach addressed in this contribution. Instead, the dependent variable is usually situated on the micro-level. Thus, the last type 3 relation in the Coleman-inspired boat-model—the aggregation of individual outcomes—is excluded going forward.

<sup>7</sup> This premise does not allow to analyse for instance the Hobbesian problem asking who and why social orders exist in the first place (Coleman 1986: 1322).

**Figure 1.1** Coleman's boat. *(Figure based on Coleman (1986: 1322))*

Moreover, what Coleman describes as type 2 relation (the macro-micro effect) is frequently included in hypotheses but rarely modelled empirically as this would often require the exploration of a temporal sequences. Instead, the direct macro– micro link between social context and individual outcome is explored as well as the moderating effect of the social context on type 1 relations. The most common conceptual set up of multilevel explanations for individual outcomes is represented in Figure 1.2. 8

One may add general problems in causal assumptions, which such multilevel conceptualisations avoid. This especially includes several types of fallacies. So-called *ecological fallacies* can arise if macro–macro correlations are used to interpret macro-micro or micro-micro phenomena assuming a given association adequately reflects similar processes on the micro-level. An example for such ecological fallacies (e.g. Robinson, 1950) could for instance mean that a correlation between insurance coverage and aggregated health is interpreted as evidence that insured individuals are healthier. The latter could be true—but it does not have to be. In addition, there are also fallacies, which may arise if the context is simply not considered. They occur, if effects, which are dependent on a social context, are interpreted without any reference to it (*psychologistic fallacy*). The opposite,

<sup>8</sup> For a more detailed and critical discussion of the relationship between Coleman's boat and multilevel designs cf. Lucas (2016).

**Figure 1.2** The multilevel model

analysing and interpreting effects on the contextual level alone without including relevant individual level processes, is termed *sociologistic fallacy* (an overview is given by Diez-Roux, 1998; Loney & Nagelkerke, 2014). Such fallacies are potentially avoided if the presence of two (or more) levels is acknowledged in methodological approaches. Explicitly modelling both levels and the relationship between them represents the most accurate and most commonly used methodological approach to capture the effect of contextual features on outcomes on the level of individuals in cross-cultural comparisons.9 For the research endeavour at hand it is therefore especially relevant because it allows to model how and why welfare policies—as a contextual influence—account for differences in observed outcomes between individuals in different countries. Only if variance between individuals and countries can be conceptualised and empirically tested simultaneously, are we able to determine if differences between individuals are actually due to specific features of welfare states.

<sup>9</sup> Examples for the empirical implementation of such models will follow later in this book. Methodological prerequisites such a sufficiently high number of countries in order to achieve a reliable estimation and a sufficient amount of variance between countries will be discussed then.

#### **1.3 Structure of the Book**

This contribution addresses various relevant facets of the conceptualisation and operationalisation of welfare state policies as an explanatory factor and gradually approaches associated problems and possible solutions. After this introductory chapter, the second part (chapter 2) deals with fundamental issues, which constitute the groundwork for this contribution by addressing two smaller questions that are prerequisites for the discussion of the actual research questions: *how is the welfare state approached as an object of research and why is it an important independent variable?* This includes a general discussion of the evolution of comparative research on the welfare state and the debates that are relevant to this day. Such research usually treats the welfare state as a dependent variable. The chapter therefore also addresses general arguments for exploring social policy-making as an explanatory factor. This is pursued by discussing the functions performed by welfare states—either because they represent essential tasks (such as risk aversion) or because they are an unintentional side effect (such as conveying solidarity and justice principles).

The following part (chapter 3) explores the first main research question: *how comparable are the results that emerge from different approaches to operationalising the welfare state as an independent variable?* This raises various critical issues, which have already been brushed throughout this introductory chapter. Besides a more detailed discussion of these issues, this chapter also explores the implications for research findings, their comparability and transparency. As the conceptual as well as empirical confrontation in this chapter is going to show, there is a need for more standardisation and conceptual work, in order to more reliably include the welfare state as an independent variable in multilevel analyses.

Starting with the fourth chapter, such conceptual work on the welfare state as an explanatory factor commences as the second main research question is explored: *how can we derive a more standardised, transparent and comparable approach to operationalising the welfare state as an independent variable?* The discussion of this question is based on two steps. In a first step, popular objects of research, which are expected to be shaped by features of welfare states are reviewed. In each case, the focus rests on identifying *how and why* welfare states are assumed to have an impact on different individual-level outcomes between countries. By pinpointing such explanations in more detail, I deduce distinct conceptualisations of welfare stateness that can be found embedded in hypotheses. As argued in the second step (chapter 5), these distinct conceptualisations can be used not only to narrow down explanations but also to explicitly select empirical measurements. They thus present an intermediary step between theoretical discussions and empirical tests. In the sixth chapter of this contribution, the derived conceptualisations are put to the test in empirical analyses. This is done by applying them to several exemplary dependent variables.

A comprehensive discussion of main findings and open questions, which may spark critical debates and future research on the matter, concludes this contribution in the seventh and final chapter.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **2 Premises: Perspectives on the Welfare State**

What exactly do we deal with when exploring welfare states? Barr (2001: 4–5) notes several issues that complicate a definition: the different sources of measures generating welfare (employment, insurance, private financial contributions, et cetera), the way in which services are provided, and the issues assigned to the welfare state rather than another policy area. Obviously, the mere existence of some sort of social policy alone does not yet mean that we can speak of a welfare state (Titmuss 1974: 26) and the scope of the term has been discussed for decades now (among earlier contributions on how to conceptualise the welfare state is Briggs 1961). Regardless of the viewpoint, key elements of a welfare state include a strong emphasis governments place on the well-being of citizens (Esping-Andersen 1990) and the acknowledgement of social rights as a crucial part of democracies (Marshall 2000 [1950]; Garland 2014). Furthermore, Titmuss (1976: 14–15) highlights two different points of interest when researching social policies: the institutional organisation of social services and the perspective of those who receive benefits. While the first aspect is closely linked to the historical development of the structures in which a welfare state is embedded, the latter focusses on the needs of those who are protected by it. When studying the individual-level outcomes of welfare policies, both perspectives are equally important.

These different viewpoints give rise to very different definitions of the welfare state, ranging from a narrower emphasis on specific policy programmes to broad conceptions encompassing all mechanisms of social protection against market dynamics and outcomes (as summarised by Otto 2018b: 52). Even though a universal definition does not exist, a useful conception seems to be understanding the welfare state as "a mosaic, with diversity both in its sources and the manner of its delivery" (Barr 2001: 4). This is reflected in the literature, as scholars often focus on specific elements of the wide-ranging institutional net that can be attributed to the welfare state. For the purposes of this book, I will treat the term similarly, not focusing on the welfare state itself, but on its functions and consequences for specific social phenomena—in other words: the reasons why social policies are treated as independent variables in the first place.

In a very general sense, four terms are used in this book: welfare state, social policies, welfare regime, and welfare stateness. The term *welfare state* addresses the overall institutional set up, in which social protection is offered, organised, and financed. *Social policies* refer to specific measures, which make up welfare states and their legal basis. There can be social policies in countries that do not classify as welfare states, while the welfare state cannot exist without social policies (Esping-Andersen 1999: 33). *Welfare regimes* go beyond the welfare state and social policies, as they also incorporate what is sometimes called a 'welfare residual' or 'welfare culture'. Here, not only the welfare provided by the state is included but also the contribution of market and family (Esping-Andersen 1999: 34–35). Lastly, *welfare stateness* describes how comprehensive a welfare state is. Although this is sometimes used synonymously to the term welfare generosity (Otto 2018a: 765), it is more general than a ranking of more or less welfare provision. Instead, it is frequently used to capture what is meant by the German term *Wohlfahrtsstaatlichkeit* (e.g. Öktem 2016: 2). This term is actually a very broad expression capturing the general idea that a commitment to social policies is an immanent feature of welfare states. Furthermore, the term emphasises governmental aspects (Schnabel 2017: 212). Even though social policies and the face of welfare states may change, welfare stateness remains a key feature of modern western democracies.

Before turning to the empirical operationalisation of different social policy arrangements as independent variables, this chapter discusses the welfare state from a more general perspective by tracing the evolution of welfare state research. It begins with a brief overview of influential debates in the relevant literature. Those debates are summarised with particular emphasis on how they conceptualise and operationalise the welfare state. The review is kept compact and does not claim to be exhaustive. Instead, I highlight issues that contribute to a better understanding of the research objective of this book, focussing on those debates that have contributed most to the conceptual and empirical approaches we encounter today. This is followed by a discussion of selected functions of welfare states that help to understand the range of impacts that social policies can have on individuals.

#### **2.1 A Short History of Welfare State Research**

The welfare state has been part of policy-making for almost one and a half centuries.<sup>1</sup> Otto von Bismarck was the first to introduce comprehensive forms of social insurance in case of sickness, accident, and old age and invalidity between 1881 and 1889 in Germany. In the following decades, similar programmes quickly emerged throughout Europe and many other parts of the world. Even though the temporal sequence seems to suggest it, the assumption that the growing popularity of social insurance schemes was due to imitation of the German case (as proposed by Briggs 1961) is contested. Industrialisation, democratisation, and other decisive developments shaped politics in many parts of the world during the late nineteenth and early twentieth century. This spurred debates about social responsibilities of the state in various countries during a similar time period but with very different rationales (as discussed by Kuhnle & Sander, 2010).

After the end of World War II began what is often termed the 'Golden Age of the Welfare State' (e.g. Ferrera 2008). The post-war decades were characterised by fast expansion of welfare policies in the Western World. Benefits now covered entire populations and grew more generous. This development came to a halt in the 1970s when the two oil crises hit the industrialised countries. Further economic, structural, and demographic challenges eventually led the welfare state into crisis from the 1980s onwards, which caused cutbacks and long-lasting transformations in the 1990s and 2000s (the post-war developments of welfare states are discussed in more detail by Kuhnle & Sander 2010; cf. also Bonoli & Natali 2013b). Adding to these challenges, the labour market underwent severe changes in the last decades with growing atypical employment and inequality in the distribution of vulnerability among different population groups (Häusermann et al. 2016: 1047). In opposition to the 'Golden Age', such changes led scholars to refer to the contemporary state as a 'Silver Age' of austerity (e.g. Taylor-Gooby 2002).

There is a variety of possible approaches to classify sub-phases of this historical trajectory. For the purpose of this book, the most relevant ones are those that left a distinct mark on comparative welfare state research today. Three phases were especially influential in my perception of the literature. (1) The exploration and description of welfare states as a necessary condition of modern democracies shaped the literature until the late 1980s. (2) In the 1980s and 1990s, comparative

<sup>1</sup> A much more comprehensive overview of the historical development of welfare states is given among others by Briggs (1961), Kuhnle and Sander (2010), and Nullmeier and Kaufmann (2010).

welfare state research grew in importance as an academic discipline—addressing crisis and future of welfare states on the one hand and exploring patterns and shared paths on the other. (3) Contemporary research focusses especially on welfare state reforms and retrenchment.

These phases of research on the welfare state reflect the actual historical developments in western democracies outlined before but exhibit some delay. Thus, even if the historical perspective nowadays agrees that retrenchment started in the 1980s, research at that time was still focussed on other aspects. Actual and potential crises of the welfare state were explored but the systematic withdrawal of the welfare state was acknowledged more than a decade later (Pierson 1996 is one of the more influential publications in this line of research).

#### **2.1.1 First Phase: Exploration and Description**

The inseparable connection between democracy and the welfare state in the post-war years quickly spurred scientific interest in social policies. The focus of theoretical approaches was to explore the nature and importance of social policies (Marshall 1961; Titmuss 1974, 1976; Marshall 2000 [1950]). More empirical approaches aimed at describing specific welfare states in case studies and historical descriptions (e.g. Briggs 1961), while others focussed on conditions for the emergence of specific policies and correlates that drove these developments based on aggregated data (e.g. Cutright 1965; Tompkins 1975).

Two conceptualisations of the welfare state received particular attention. In the first, the welfare state is analysed in terms of *effort* (e.g. Tompkins 1975; Wilensky 1975). In order to describe and compare social policy arrangements, scholars thus focussed especially on *social expenditure*, which was perceived as the best indicator for the amount of work a welfare state puts into the wellbeing of its citizens. One of the key assumptions and empirical finding was that industrialisation leading to economic development would be a universal motor for the growth of social expenditure and population-coverage of social insurance schemes. Such interdependency between economic growth and the welfare state was supported by the Keynesian assumption that increasing public welfare would soften economic fluctuations (Quadagno 1987: 110). Since these processes affected all western capitalist democracies in the post-war decades, a certain degree of convergence between industrialising nations was expected. However, such hypotheses about the convergence of social policies did not prove to be true (as is discussed in more detail by Skocpol & Amenta 1986: 133–134). Instead, a variety of welfare states emerged following different principles and shifting the focus of scholarship on the matter from economic forces to political institutions (cf. Myles & Quadagno 2002) inspiring growing commitment to the power research perspective and new institutionalism (Pierson 1996).2

Such different theoretical viewpoints have manifested in discussions about the empirical conceptualisations of welfare stateness. From an empirical point of view, restricting the explanation for variations in welfare commitment to singular features (like economic development), has met criticism, since other political and social factors were viewed as equally important (e.g. Castles & McKinlay 1979; Esping-Andersen 1990: 19). In addition, social expenditure was criticised as being an inadequate and oversimplified proxy for the multidimensional nature of welfare state institutions (Korpi 1989: 310).

A second conceptualisation of the welfare state highlights *social rights*. This approach is based on Marshall's essay 'Citizenship and Social Class' (2000 [1950]). He describes social rights as key features of citizenship, parallel to civil and political rights and conceptualises the welfare state in terms of benefits it offers to citizens based on those rights. In this context, Marshall also highlights the link between benefits and social inequality. He identifies four factors influencing the degree of equality in modern welfare states of his time:

[…] whether the benefit is offered to all or to a limited class; whether it takes the form of money payment or service rendered; whether the minimum is high or low; and how the money to pay for the benefits is raised (Marshall 2000 [1950]: 39).

This differentiation introduces two important aspects, which influence the discussion about the nature and measurement of social policies until today. First, Marshall distinguishes *entitlement criteria*, *mode of benefits* (income replacement vs. services), *generosity of benefits*, and *financing of benefits*. This suggests a much more differentiated way of potentially capturing differences between welfare states compared to simply focussing on welfare effort. Second, by adding the link to equality, Marshall emphasises the relationship between welfare provision and social inequality based on social class. Thus, different modes of income replacement and other social services potentially shape inequalities in different

<sup>2</sup> Neither of these perspectives is restricted to the 'first phase' of welfare state research as they can still be found in contemporary approaches. While power resource theory focusses especially on how welfare provision is shaped by the distribution of political power, new institutionalism emphasises the role of institutions, which "establish the rules of the game for political struggles" (Pierson 1996: 152). Overall, especially historical analyses of such institutional developments have emerged as a very popular perspective in comparative welfare state research (Lynch & Rhodes 2016).

ways. This points to a function of welfare states that goes well beyond providing social security, as it frames it as a system of social stratification (this will be discussed in more detail later).

Empirically, the social rights perspective was first measured through social expenditure as well. However, even controlling for recipient population, this indicator was never able to adequately model the different internal logics of redistribution in different countries (Stephens 2010: 515). Thus, expenditure could not capture Marshall's conception of "what is the core of the modern welfare state the extent and quality of the social rights that constitute social citizenship" (Korpi 1989: 310). Perhaps the most influential contribution to a more elaborate operationalisation has been the Social Citizenship Indicator Program (SCIP), which Walter Korpi started at Stockholm University in the early 1980s. It provides detailed data on an array of different indicators of social rights in 18 OECD countries from 1930 until 2005 (Korpi & Palme 2008). This data constitutes the basis for a variety of publications, which have had a very strong impact on welfare state research until today (e.g. Korpi 1989; Esping-Andersen 1990). Other datasets and operationalisations followed this tradition (e.g. Scruggs 2006).

In addition to such attempts to pinpoint and operationalise key aspects of the welfare state (effort and social rights), there were also early contributions dealing with the exploration of different types of social policy arrangements in a crosscultural comparative perspective. Two ideas were especially influential: (1) the demarcation of different extreme poles of contribution and financial organisation and (2) the classification of multi-faceted ideal-typical social policy arrangements.

The first tradition most prominently distinguishes between what is called the *Bismarckian* and the *Beveridgean* social system.3 Named after Otto von Bismarck and William Beveridge as key figures in the establishment of very typical ways of organising and financing social policies, these terms are quite established in the literature until today. The key feature of Bismarckian welfare states is that they are based on social insurances. Tied strongly to employment, they are financed through contributions by employers and employees and benefits relate to previous income. Contrary to that, the principle of the Beveridge system is universal and independent of employment. Benefits consist of flat rates and—even though this may not be in line with Beveridge's initial ideas—are tax-financed and either universal or means-tested (Bonoli 1997: 357).

<sup>3</sup> This distinction was used from very early on (e.g. Sigerist 1943) and while a reference to Beveridge is not as common anymore, especially typological approaches often label continental welfare states Bismarckian (e.g. Ferrera 1996; Foubert et al. 2014).

In the second tradition, different types are identified on the basis of additional aspects, like the distinction between market, family and state. An early example for an approach of that kind is given by Titmuss (1974: 30–32). He distinguishes three different models of social policy-making. In the *Residual Welfare Model of Social Policy*, private markets and the family constitute the main providers of welfare. The state only intervenes if those main providers fail to meet individual needs and it does so only temporarily. In contrast, in the *Industrial Achievement-Performance Model of Social Policy* needs are compensated based on the individual performance and productivity at work. Lastly, the *Institutional Redistributive Model of Social Policy* grants universal services, which are independent of the market and needs testing. While these types incorporate some elements of the distinction between Bismarckian and Beveridgean policies, they go beyond this dichotomy by adding aspects (like the role of the family) and offering combinations of Bismarckian and Beveridgean elements. Such attempts to find similarities and shared paths in comparative analyses of the welfare state, which—as Titmuss puts it—"help us see some order in all the disorder and confusion of facts, systems and choices concerning certain areas of our economic and social life" (1974: 30) grew especially popular in the 1990s. Thus, they will be discussed in more detail in the next section of this chapter.

#### **2.1.2 Second Phase: Similarities and Classifications**

The results of economic, demographic and other societal crises during the 1970s and 1980s were reflected in the literature of the period. They inspired more academic interest in the subject and led to a considerable growth of comparative welfare state research as a discipline within the social sciences (Skocpol & Amenta 1986; Olsson 1987). At the same time, this increased the need for clearer conceptual work on how to explain and measure variations of social policies between countries and periods (Lockhart 1984). As Skocpol and Amenta conclude in a review article (1986), aggregated quantitative analyses reached a limit and new insights had to come—in their view—from more in-depth comparativehistorical analyses of specific cases or trajectories. Another influential impetus came from the aforementioned approaches to operationalise the welfare state as a multidimensional phenomenon by focussing on a variety of indicators instead of relying solely on social expenditure (e.g. Korpi 1989).

Indeed, comparative welfare state reseach in the 1990s focussed on more complex conceptualisation of welfare states and on historical analyses. In addition to case studies, this research intention was implemented in attempts to classify archetypical systems of welfare provision. Drawing on the above-mentioned ideas of differentiating distinct aspects of social policy arrangements (by for instance adding the social rights perspective), this period contributed a vast amount of proposals to capture different (ideal-) types of welfare states. While the focus of research on social policies had previously been on effort—the so-called "how much dimension" (Bonoli 1997: 352)—these attempts acknowledged that welfare states are multi-facetted and require more complex measurements than simply using expenditure-based indicators (Esping-Andersen 1990). However, in contrast to measuring the different facets through composite indices or a variety of single indicators, many of the studies aimed at proposing a typology of welfare states.

One of the most influential contributions to the literature on classifications of welfare systems is Gøsta Esping-Andersen's *Three Worlds of Welfare Capitalism (TWWC)* (1990). Based on a historical discussion and a very comprehensive database, he clusters 18 OECD countries into three welfare regimes. Being a former member of the team working on the Social Citizenship Indicator Program (Korpi & Palme 2008), he includes Marshall's (2000 [1950]) ideas regarding social rights of citizenship. In order to capture social rights and their connection to social inequality, he focusses on the degree of decommodification and social stratification in the examined countries. He defines decommodification as the extent to which individual dependency on labour market participation is reduced through social policies (Esping-Andersen 1990: 21) and the ability to uphold a "socially acceptable standard of living independent of market participation" (Esping-Andersen 1990: 37). Social stratification captures how far the organisation, distribution and financing of social policy reduces, reproduces or increases inequality. This again draws on Marshalls concept, in which—using Esping-Andersen's words—the "status as a citizen will compete with, or even replace, one's class position" (Esping-Andersen 1990: 21). These different dimensions are measured with the help of an array of different indicators, including a decommodification score<sup>4</sup> and three stratification indices.5 In addition, Esping-Andersen includes the role of state, market and family in the provision of welfare (although the latter is more prominently discussed in his later publications).

His theoretical and empirical analysis yields three types of welfare regimes, which show some resemblance to the one proposed Titmuss (1974). The first is the S*ocial Democratic* (or *Socialist*) regime, which is characterised by generous

<sup>4</sup> For the scoring procedure cf. Esping-Andersen (1990: 54).

<sup>5</sup> They measure the degree of conservatism, socialism and liberalism in different countries. For more details cf. Esping-Andersen Esping-Andersen (1990: 77–78).

benefits, a high degree of decommodification and a prominent role of the state as the main provider of welfare. A strong focus on universalism and equality as guiding principles means that this type of welfare state reduces inequality. Many Nordic countries resemble this ideal. In sharp contrast, the second regime emphasises the market. In this *Liberal* regime, decommodification is low, benefits and services are provided based on means-tested targeting following the objective of poverty-relief. Due to strict eligibility criteria and less generous benefits, inequalities are potentially even increased in this type of welfare regime. Anglo-Saxon countries show resemblance with this type of social security system. Lastly, the *Conservative* (or *Corporatist*) welfare state is based on social insurance tied to labour market participation. It follows the principle of status preservation and emphasises the role of the family. Germany serves as an example for a country closely resembling this type.

What followed was an era one might term *typology enthusiasm*, largely marked by the introduction of various typologies of welfare regimes in the 1990s and early 2000s, inspired by Esping-Andersen's seminal study. His intuitive threefold classification quickly inspired what Abrahamson (1999) memorably dubbed 'welfare modelling business'. The Three Worlds turned into a true classic over the last decades (Emmenegger et al. 2015), although the approach has also drawn criticism addressing both conceptual (Orloff 1993; Room 2000; Kasza 2002) and methodological (Bambra 2006; Scruggs & Allan 2006) issues. Adjustments led to even more classifications, introducing criteria like defamilialisation (Esping-Andersen 1999), additional types such as a Southern European type (Leibfried 1992; Ferrera 1996) or the extension of a typology to countries which were not previously classified or classified ambiguously. As a result, we are confronted with an enormous body of literature offering a great number of distinct typologies and various summarising contributions (e.g. Arts & Gelissen 2002; Bambra 2007; Ferragina & Seeleib-Kaiser 2011; van Kersbergen & Vis 2015). In light of the importance of welfare state typologies for research until today and their frequent use as an explanatory variable, a more elaborate discussion of problematic issues is necessary.

The theoretical foundation of most typologies focusses on broad similarities in the historical genesis of social security systems which are depicted as path-dependent—in line with Esping-Andersen (1990) and earlier historical institutionalist approaches. In this context, the terms *ideal type* and *real type* or *typology* are often distinguished (e.g. Kvist 1999; Aspalter 2011; Rice 2013) ideal type usually meaning an archetypical way of organising and financing welfare provisions that serves as a guideline for interpreting actual policy-making. In contrast, real types refer to clusters of countries that are based on similar patterns determined purely empirically based on indicators of actual policy-making. When considering how such types—regardless of being ideal or real—are translated into empirical classifications, a great variety concerning both indicators and methods can be found. These differences will be discussed in more detail in the next chapter. For now, suffice it to summarise that they concern a multitude of relevant aspects: the selection of indicators on which a typology (of real cases) is based, the methods applied to determine clusters of countries, the sample of countries that is analysed and the conceptual premises chosen to interpret results. In light of the severity of these differences in the theoretical conceptualisation and empirical operationalisation, it does not surprise that even though almost all scholars since the early 1990s base their work on Esping-Anderson's Three Worlds, the number, title, and composition of the regimes differ significantly. Even if we accept that comparable regimes are named differently,6 barely any country has been attributed to the same type in all studies.

Regardless of the critical debated surrounding welfare regime typologies (Aspalter 2011; van Kersbergen & Vis 2015), taking a meta-perspective on the literature reveals that there are also some common elements in the various different typological approaches. In an attempt to systematise the literature, Ferragina and Seeleib-Kaiser (2011) assess the overlap between Esping-Andersen's TWWC typology and 22 other prominent typologies published between 1990 and 2009. They find that although there are often more than the initial three types, most countries are attributed to either the Liberal, the Social-Democratic or the Conservative regime in more than 50 percent of the examined studies.7 The most frequent addition to the threefold classification is a Southern (or Mediterranean) type. The resulting fourfold classification seems to be a somewhat robust finding—at least until a decade ago. However, Central and Eastern European countries remain entirely ignored. Reviews including more recent typologies and post-socialist countries are rare and often focus on the country selection instead of the classification (Kim 2015). In general, there are not many new typologies introduced after 2010. The few attempts that were proposed, used very different perspectives and indicators—such as aggregated individual welfare attitudes and values (Vrooman 2013) and aggregated data on policy outcomes (Ferragina &

<sup>6</sup> For example, a cluster of nations comparable to Esping-Andersen's social-democratic welfare regime has been named Scandinavian (Leibfried 1992), Nordic (Bonoli 1997) and Encompassing (Korpi & Palme 1998).

<sup>7</sup> Only the Netherlands and Switzerland are confirmed to be hybrid cases as they exhibited no clear pattern in their regime-affiliation.

Seeleib-Kaiser 2015). Perhaps, the TWWC and related regime approaches have indeed lost some of their lustre.

This may be one of the reasons, why enthusiasm subsided recently. Instead of offering new classifications, the existing theoretical and empirical typologies are more intensively examined for their usefulness as analytical tools (Reinprecht et al. 2018). Still, even those scholars contesting the prevalent conceptual strategies agree that typologies remain a popular analytical tool and that Esping-Andersen's premises should be transformed but not rejected entirely (Rice 2013). Others argue in favour of concentrating on distinct areas of the welfare state, which means for instance separating healthcare systems and educational systems from overall typologies of welfare stateness, as they follow different objectives and are less related to decommodification and other key elements at the heart of ideal-typical welfare states (Ferragina & Seeleib-Kaiser 2011: 587). Following this line of arguments, these systems inspired their own distinct typologies (cf. Beckfield et al. 2013).

Only few authors take a more radical view, even suggesting that research which draws on Esping-Andersen's original typology starts to show signs of Kuhnian normal science since follow-up research always tests and contests previous findings within the existing paradigm (van Kersbergen & Vis 2015). Indeed, studies without any reference to Esping-Andersen are rare and usually address a very different perspective like Künzel's (2012) analysis of sub-national variations in welfare provisions. However, the sources of variation discussed and the resulting lack of comparability are quite consequential when typologies are used as independent variables. Thus, they will be quite important in the following chapters.

Overall, the old classifications distinguishing a Liberal, Social-Democratic, Southern and Conservative type are still seen as useful tools, albeit the fact that some of the foremost archetypical countries (such as Germany and Sweden) move continuously further away from the corresponding ideal types (Reinprecht et al. 2018).

#### **2.1.3 Third Phase: New Risks and Retrenchment**

Contemporary research on the welfare state is characterised by a strong emphasis on change. Such change manifests in two ways. First, economic and social structures are changing, leading to new challenges for welfare states, which in consequence have to adapt. As Gilbert puts it:

It is not demographic factors or tax ceilings, globalization, or the normative changes shaped by knowledge and experience with social policies or the rising faith in the market economy that by themselves account for the fundamental change in the character of social protection; rather, it is the combination of these forces. (Gilbert 2004: 42)

Second, adapting to such new challenges mainly manifests in two ways: cutbacks—in other words retrenchment of the welfare state (Pierson 1996)—and transformation such as reallocation of responsibility to the individual and other responses to new risks (Bonoli & Natali 2013a). Overall, there are two main perspectives on a similar subject: one focusses on external factors driving change and one on the way welfare states adapt internally.

Even though there are still attempts to classify differences between welfare states in typologies (cf. Ferragina et al. 2015), the perspective on the welfare state has shifted. While different trajectories of social policies have been viewed as path-dependent and thus somewhat stable during most of the twentieth century, a key focus of recent welfare state research is on *the politics of the new welfare state* as Bonoli and Natali (2013a) title an edited volume on the matter.

The first phenomenon receiving much attention is *retrenchment* of the welfare state (Pierson 1996, 1998). Retrenchment addresses cutbacks in welfare benefits and services. As Pierson argues, such cutbacks are unpopular and thus sometimes disguised within structural shifts including more means-testing, growing individual responsibility to provide one's own security through private insurance schemes and changes in eligibility rules (Pierson 1996: 157). Indeed, the majority of research agrees that cutbacks have taken place in most advanced welfare states, which do not primarily show up in a reduction of expenditure but in other areas such as replacement levels (for a review cf. Starke 2006).

However, cutbacks and changes in how the welfare state is financed and how resources are redistributed represent only one side of the politics of the new welfare state. There is also growing complexity in the risks and needs welfare states are confronted with. Male breadwinners not being able to generate income was the main risk in post war welfare states. This changed considerably. Social and economic transformations led to a wide array of new risks and needs which have to be met and partly appear regardless of employment (Bonoli 2005, 2007). As Zutavern and Kohli (2010: 175–176) summarise, new needs and risks include changes in the labour market, such as tertiarisation, feminisation and flexibilisation, as well as changes in life courses and life forms, including longevity, fertility and family stability. Overall, such changes shifted the focus of social policies from income replacement to the promotion of labour market participation (Bonoli & Natali 2013b: 5–6) and individual autonomy and responsibility (König 2017). This goes along with growing attention devoted to *new risk groups*. The compatibility of family and employment (especially for women) and the dualisation of labour markets, producing 'insiders' who continue to enjoy traditional social policy provision and 'outsiders' who face new insecurities (Bonoli & Natali 2013b: 8), are only two issues receiving growing attention in this field. In addition to changes in the 'old' agendas of social protection, the 'new' demands thus introduce efforts in other areas. This includes activating policies increasing labour market participation, which gain importance relative to passive policies focussed on income replacement. The changing objective of welfare policies embedded in this trend is sometimes referred to as moving from a welfare state towards an *enabling state* (cf. Gilbert 2004: 44).

Within approaches highlighting such activating policies, s*ocial investment* is a policy agenda, which received growing interest during the last decade. From a political point of view, it is seen as a promising way of grasping answers to the new challenges of welfare states (Kuitto 2016: 442). Broadly speaking, social investment refers to social policies, which target the above-mentioned new challenges and risks. Instead of focussing on income replacement, social investment incorporates policy measures, which enable individuals to take responsibility for their own welfare. Examples for such efforts being investments in human capital and education (cf. Andersson 2018: 109) as well as childcare provision (León 2017). A main agenda of social investment is to cater to the new rationales of social security, which promote individual over collective responsibility (cf. Ellison & Fenger 2013). As Vaalavuo (2013: 516) points out, the focus shifted from a redistribution of income to a redistribution of opportunities. However, it is important to emphasise, that this does not mean, that the former necessarily decreases or counteracts the latter. Instead, preventing risks (investment strategy) and compensating risks if they materialise (protection strategy) overlap and potentially aid each other (Vandenbroucke & Vleminckx 2011: 451).

A last field, which is strongly related to new risks and risk groups, can be found in research on the relationship between family and welfare state. These debates emerged out of criticism of old approaches in comparative welfare state research, which underemphasised the role of the family (cf. Orloff 1993). A key concept in this line of research is *defamilialisation* (and familialisation as its counterpart). Analogous to decommodification, this term captures the extent of individuals' (in-)dependence from the family (for an overview of the debate see Lohmann & Zagel 2016). Defined more specifically, defamilialisation refers to

the degree of women's autonomy from the family (e.g. the spouse) in achieving financial resources and the degree to which women's unpaid work in the family, particularly unpaid caregiving, is substituted by paid labour from outside the family by means of public, market or third-sector services (Keck & Saraceno 2012: 454).

Summarising the developments above, we can identify two dominant topics in recent literature on welfare states in Europe. One addresses retrenchment of the welfare state and asks whether austerity emerges as a stable characteristic. The other focusses on new strategies in social policy-making (such as social investment) and explores the potential of such new emphasis in social protection.

In line with the growing complexity of welfare states and the risks they respond to, research on the welfare state grew more complex as well. One of the biggest differences compared to previous strands of research on the welfare state is the detachment from historical path-dependence. Welfare states do not seem to expand further in their previous trajectories. A significant part of the literature identifies retrenchment as a stable and rather universal development in western capitalist democracies (Bonoli & Natali 2013b), which challenges the old conception of distinct welfare cultures or regimes.<sup>8</sup> The same holds true for a growing relevance of new risks and strategies to deal with them. Thus, the changing nature of social policy arrangements and objectives raises methodological questions as old operationalisations (like the identification of distinct welfare regimes) are challenged. In particular, the literature on welfare state change and retrenchment has not only introduced new conceptual perspectives, but also triggered a methodological debate about the selection of suitable measurements of social policy-making—the so-called *dependent variable problem* (e.g. Clasen & Siegel 2007). While researching change, scholars observed the lack of a common understanding of what the object of research—the welfare state as dependent variable—entails and how it should be measured (Green-Pedersen 2004). As a result, there is an ongoing discussion based on the repeated observation that different conceptual and operational strategies lead to different results (cf. Kühner 2007; Bolukbasi & Öktem 2018). Similar discussions have emerged in the literature on social investment. While this strand of literature initially relied dominantly on expenditure data, the demand for more nuanced operationalisations was raised more recently (cf. Andersson 2018).

For the purpose of this book, these strands of literature are relevant, as they provide and test indicators, which are taken up as independent variables. Within the retrenchment debate, especially policies related to old risks are highlighted. In

<sup>8</sup> This perspective is however contested, since some authors find that even though the growth of welfare states and the generosity of their benefits came to a halt, this does not mean that austerity is now the new reigning principle (Reinprecht et al. 2018).

particular, three perspectives on welfare states and social policy provision exist.9 The first focusses on expenditure (e.g. welfare effort), the second on social rights, and the third on benefit receipt. While the first two operationalisations represent well-established perspectives on social policy arrangements, the focus on benefit receipt is comparatively rare (van Oorschot 2013: 225; Otto 2018a). Regarding the operationalisation of effort, reducing the operationalisation to measures of expenditure is still criticised continuously (following the arguments discussed in the previous sections of this chapter). Focussing on social rights, including benefit entitlement and generosity, is thus a popular alternative.10

Figure 2.1 illustrates the relationship between the three components according to van Oorschot (2013: 230). Here, social rights define the access to benefit receipt in specific benefit areas, while social expenditure represent the cost associated with benefit provision and take-up. Using van Oorschots terminology, social rights thus represent *policy outputs*, while benefit receipt signals *social outcomes* and expenditure c*ost outcomes.*

The methodological debates emanating from the dependent variable problem have taken up arguments and perspectives—such as the differentiation between social rights and welfare effort—from earlier welfare state research. Nevertheless, they also offer important additional insights due to a very strong focus on the operationalisation and selection of very specific indicators and their treatment in empirical analyses. This leads to detailed discussions about the measurement of particular aspects of social policy-making and specific indicators. An example for this is the very detailed discussion about the measurement of net replacement rates (Scruggs 2013; Wenzelburger et al. 2013).

While the dependent variable problem refers mainly to perspectives related to old risks, the measurement of policies related to new risks and risk groups, is partly detached from these approaches. Especially the field addressing family policies and gendered outcomes works with their own sets of operationalisations (Keck & Saraceno 2012; Lohmann & Zagel 2016). There are several reasons, why research on family policies follows a somewhat different path than research on the old risks. First, this strand of literature emerged later and—as noted above out of criticism of the oversight of family care responsibilities in prominent

<sup>9</sup> Comprehensive and critical overviews are given e.g. by van Oorschot (2013) and Otto (2018b).

<sup>10</sup> This conceptualisation deviates slightly from Marshall's original approach. While Marshall saw social rights as based on citizenship alone and thus equally accessible and beneficial to all citizens, contemporary conceptions usually include earning-related benefits (Stephens 2010: 511).

**Figure 2.1** Social rights, benefit receipt and social spending. *(Slightly modified version of schema by van Oorschot 2013: 230)*

approaches in welfare state research. Second, enabling families to combine parenthood and employment is a comparatively new agenda in social policy-making and emerged partly diagonally to old paths within welfare states. This manifests among other things in contradictory policies within a welfare state—as discussed by Lohmann & Zagel (2016: 48–49) using the example of German family policy, which encourage female care responsibility through financial benefits and promote female labour market participation at the same time. Still, while indicators of (de-)familialisation are usually provided separately from indicators of the old welfare state functions, they show many similarities. Like in the case of old risks (esp. unemployment, sickness, and old age), indicators of eligibility, access and generosity exist (Keck & Saraceno 2012). Similarly to the already described methodological approaches, these indicators are used as single measures, as well as combined in indices (Lohmann & Zagel 2016).

In addition to a focus on the particular field of family policies, contributions on other responses to new risks such as social investment and active labour market policies also spawned methodological debates. This led to various types of measurement, including typologies of active labour market policies (Bonoli 2012) and spending on fields tied to investment, which emphasise activating components (Kuitto 2016: 447–448). However, as noted above, the discussion of how to operationalise social investment—especially going beyond expenditure—is ongoing (Kuitto 2016; Andersson 2018).

Summing up the main debates in more recent comparative welfare state research, the literature on welfare state change and retrenchment has led to new perspectives and empirical challenges. It has inspired a detachment from analyses of historical trajectories and led to a stronger focus on operationalising relevant features of welfare state change. Several different perspectives on the welfare state have been highlighted in more detail and provoked new debates about how to measure them. Thus, while they inspired many substantial contributions on the matter, recent debates in research on measuring welfare states are still far from agreeing on a best-possible solution to measuring the welfare state (cf. Otto 2018b). There is a great number of contributions addressing methodological issues and proposing specific ways of operationalising different welfare state arrangements. Nevertheless, if there is a trend in comparative welfare state research at the moment, it seems to be one towards complexity.

#### **2.1.4 Summary: Influential Debates**

This section briefly traced influential debates in research on the welfare state since the 1950s. Throughout the last 80 years, a variety of different perspectives can be identified (a summary of key premises, debates and conceptualisations is provided in Table 2.1). Each of the perspectives offers distinct proposals for the empirical operationalisation of the welfare state or specific dimensions of social policies. Since all of them are taken up by research that includes the welfare state as an explanatory variable, the origins and premises of the different methodological approaches are important for the following evaluation of their applicability as macro-level independent variables.

Overall, we can distinguish two different premises when it comes to the conceptualisation and empirical measurement of social policies. The first aims at grasping the welfare state as a whole, while the second focusses on specific dimensions of welfare provision. While the first conceptualises the welfare state as a multidimensional phenomenon (in a typology or composite index), the second highlights specific social polices either as proxies for overall welfare orientation or as singled out issues relevant for specific perspectives on the welfare state. The different focal points, debates and conceptualisations summarised in Table 2.1 are not restricted to one period even though they emerged roughly in


**Table 2.1** Summary of key premises, debates and conceptualisations

the described order. Until today, we find conceptualisations highlighting welfare effort, as well as typological approaches following the TWWC.

#### **2.2 Functions of Welfare States**

The debates outlined above have shaped the theoretical and empirical conceptualisation of welfare states to this day. While these debates help to understand what welfare states are, how they have evolved, and why there are differences and similarities, this still does not reveal how different welfare states affect societies and individuals within them. Exploring these processes is quite important to the research objective of this book, as the tasks and goals of welfare states and differences in how they are approached shape individual outcomes. As such, they are the reason why features of the welfare state are treated as independent variables in the first place. The literature offers various approaches to characterising welfare states' objectives and functions. While functionalist theories focus on the relationship between economy and social policies, institutionalist and political approaches highlight the importance of political and administrative institutions and actors (for a more comprehensive summary of these theoretical perspectives cf. Myles & Quadagno 2002). All of these approaches offer important perspectives. However, when asking how the welfare state interacts with individuals the last two are especially insightful. Here, the institutionalist perspective offers insights about processes within the welfare state, while the political approach helps contextualise political systems and channels of policy feedback.

In this section, I will briefly summarise a selection of important ways in which social policies influence individuals and highlight those that I believe are very relevant to contemporary research treating the welfare state as an explanatory concept in the social sciences.11 The term *function* is not meant here in the sense of a quasi-automatic reaction to policies, but as a mode of action. The discussion remains at a general level, focussing on how different welfare states are linked to individual outcomes before these links are gradually fleshed out in the course of this book.<sup>12</sup>

Overall, five key functions of the welfare state receive particular attention. This includes the functions of (1) providing security, (2) (re-)distributing resources, (3) shaping social stratification, (4) enabling and incentivising, and (5) socialising individuals. They incorporate processes that shape individual outcomes directly or indirectly and thereby provide the groundwork for the conceptualisation of the welfare state as an explanatory factor in following parts of this book (especially in chapters 4 and 5).

#### **2.2.1 Security: the Welfare State as a Safety Net**

The manner and efficiency of welfare states' responses to needs—such as a need for income or a job—and risks—such as the risk of poverty or unemployment tell us much about their nature. Needs are met foremost by protecting against risks (Zutavern & Kohli 2010: 169). In this sense, meeting needs by avoiding risks or moderating their outcomes is the key function of the welfare state—one may call it a 'minimum' function (Eger & Breznau 2017: 441). The main risk to be avoided is poverty. In its most basic form, this function is provided by social assistance schemes. Such—targeted or universal—basic transfer payments aim at avoiding the direst manifestations of need, such as poverty (Garland 2014: 342).

As more contemporary literature points out, many changes in social policies may be attributed to severe changes in needs and risks—as was outlined in the

<sup>11</sup> Other functions do of course exist. For instance functions relating to power and social control (cf. Higgins 1980) are not highlighted as they are less dominantly found in the type of multilevel literature focussed in this book.

<sup>12</sup> Those questions of how and why the welfare state is at work are often discussed too superficially in research treating the welfare state as an independent variable. It is quite curious, that important reference books such as 'The Oxford Handbook of The Welfare State' (Castles et al. 2010), 'The Welfare State Reader' (Pierson & Castles 2000) or the 'Handbuch Sozialpolitik' (Obinger & Schmidt 2019) only briefly discuss policy outcomes on the micro-level (especially those that go beyond risks and well-being).

previous part of this chapter. Furthermore, it is argued that addressing risks has become more important than responding to needs, as the latter are less profound in contemporary societies (Kemshall 2002). In addition, what actually counts as legitimate need is subject to social and political deliberation (Sachweh 2016). There is however evidence, that new risks are not covered as well as old ones by most welfare states. Following contributions on the dualisation of labour markets (e.g. Emmenegger et al. 2012), changing risks have created new risk groups (e.g. like single motherhood or youth unemployment), which slip through the net when it comes to the old way of social protection (Brady et al. 2017: 771). When asking how and why welfare states react to risks and needs from a contemporary perspective, it is thus important to bear in mind the old as well as the new risks.

**Figure 2.2** Links between the welfare state and risks and needs

In general, risks are distributed unequally amongst individuals, based on class, age, and gender (Esping-Andersen 1999: 32). The way in which the welfare state responds to them, thus actively shapes individual vulnerability as expressed by their socio-economic position and—as a result—patterns of social stratification on an aggregate level. The latter will be addressed in more detail in a proceeding section of this chapter (cf. Section 2.2.3). Here, the focus is on the aim of lowering individual risk. Pinpointing the role and function of the welfare state underlying its relationship to risks and needs can be best summarised as providing *security*. This objective of meeting risks and needs ties the welfare state to the individual in different ways, which are illustrated in Figure 2.2.


Its specific response to risks determines the nature of a welfare state. Risks and needs can be handled with varying efficiency and generosity, and lead to more or to less equality in society. Furthermore, how responsibility is divided amongst state, family and market reveals the underlying welfare regimes. In the terminology promoted by Esping-Andersen (1990, 1999), the stronger the role of the welfare state, the less responsibility is placed on the market (decommodification) and the family (defamilialisation).

Responding to risks and needs entails another noteworthy element, one might call a sub-function, which is the avoidance or at least reduction of uncertainty (Barr 2001: esp. chapter 2; Crouch & Keune 2013; Garland 2014). It is argued that this might even be one of the reasons social policies were implemented in the first place. As Sigerist (1943: 375–376) points out, uncertainties due to economic cycles and resulting constant risk of unemployment were among the "major grievances" workers in Germany (but this can be extended to other countries as well) had to face. The insurance schemes introduced by Bismarck were in parts a reaction to that—albeit not out of charity but in order to weaken socialist movements. By providing a safety net and increasing certainty regarding how risks and needs are met if ever occurring, welfare states potentially influence individual outcomes—especially those tied to cost-benefit considerations (Iversen & Soskice 2001).

As laid out in this part of the chapter, risks are at the heart of welfare state responsibilities. The function of lowering risks and moderating the outcome for individuals *at risk* is thus important to determine a welfare state's responsiveness in a comparative perspective. Evidently, varying responsiveness can be an important explanation when examining why different social policy arrangements have different effects on individual outcomes. Therefore, the different possible links between social policies and risks displayed in Figure 2.2 already suggest specific hypotheses, which will be discussed in detail later in this book.

#### **2.2.2 Redistribution: Robin Hood, Piggy Bank and More**

Another key function of the welfare state is the redistribution of resources. Such redistribution happens in different modes. On the one hand, it entails shifting income from stronger participants of the labour market to weaker ones. On the other, it relocates resources over the span of life and economic phases from times of strength to times of need. These two redistributive functions are sometimes referred to as the *Robin Hood* function and the *Piggy Bank* function (Barr 2001). Regardless of the mode of redistribution adopted by a specific welfare state, the instrument of shifting income from one group or time to another can in the most basic sense be described as the tool through which welfare states fulfil their most elementary obligation: social protection and thus the security function introduced before. While this shows that there is potential overlap between different functions, distinguishing the different components is important in order to understand the different facets in which welfare states operate.

Redistribution, which aims at relocating income from rich to poor individuals or families, falls under the *Robin Hood* type of resource allocation (Kvist et al. 2013: 322). The objective here is to reduce inequality between social groups by working towards vertical equity (Barr 1993: 10). This is found especially in targeted welfare state models, where means-tested benefits are provided to those in need. Such modes of equalising individuals seem to correspond to popular notions of fairness. However, evidence shows that there seems to be a so-called *paradox of redistribution*. This refers to the finding that the Robin Hood approach to redistribution appears to be less effective than a universal approach, which is funded by all and distributes to all. This paradox was most prominently discussed by Korpi and Palme, who conclude:

the more we target benefits at the poor only and the more concerned we are with creating equality via equal public transfers to all, the less likely we are to reduce poverty and inequality (Korpi & Palme 1998: 681–682).

This seemingly counterintuitive finding is explained by several arguments. Broadly summarised, universalism is expected to increase public support for redistributive policies—especially in the middle class—and thus strengthen political efforts implementing such policies, which increases the overall budget (Jacques & Noël 2018: 72). Recent research on the matter finds some changes in the link between logic of redistribution, public support and poverty. While the link between universalism and reduction of poverty proves to be stable, the relationship between universalism and preference for redistribution—as a bridge hypothesis—produces contradicting results (e.g. Brady & Bostic 2015). However, this seems to depend on the chosen operationalisation of universalism—considering the aim of this book, this is a small but important finding. Evidence suggests that capturing the institutional design of redistribution13 instead of its outcomes14 supports the link between universalism and policy support (Jacques & Noël 2018: 82).

Concluding, the Robin Hood logic of redistributing from one part of the society to another can follow different objectives. At one extreme, it can mean the targeted reallocation of resources from richer parts of the population to poorer ones. At the other extreme, we find universal redistribution which takes "from all to give to all" (Jacques & Noël 2018: 71). As the short summary in this chapter shows, such different faces of redistributive logic potentially influence individuals in a variety of ways. Among other things, they exhibit different efficiency in lowering risks, shape political support and redistributive preferences.

A second type of redistribution disperses resources throughout the life course. The welfare state is referred to here as a *piggy bank* (Barr 2001). Social risks are unevenly distributed throughout an individual's life, and obvious situations of high vulnerability occur in childhood and old age. However, as labour market participation is no longer just a matter of the male breadwinner, new risks such as motherhood gain importance, which arise at specific stages in the life course (Esping-Andersen 1999: 42). Redistribution, then, is not just an exchange

<sup>13</sup> Such as the share of means tested benefits.

<sup>14</sup> Such as the concentration of transfers in specific households.

of resources between individuals, but also within a life course from times of economic prosperity to times of need. Insurances against risks which may occur at later stages in life, result in *consumption smoothing* (e.g. Barr 2001: 5) by allowing individuals to decrease uncertainties and achieve more economic stability over their life course.

This perspective can also be found in areas outside of financial redistribution. For example, the social investment approach shows resemblance with the described logic. Like other policy measures, social investment follows a generational path in which—at least in an ideal-typical way—individuals switch from being recipients of investment (in childhood and youth) to being contributors (while participating in the labour market) and finally back to being recipients in old age (Andersson 2018: 110). In this sense, investments in human capital and other individual traits in younger years can be cashed in during times of need later in life. From the point of view of the individual, redistribution over the life course not only impacts risks (especially those arriving in later life), it also influences the perception of uncertainty and predictability of life.

#### **2.2.3 Social Stratification: the Welfare State and Social Inequality**

As pointed out at the beginning of this chapter, the link between the welfare state, social class and social inequality was already established by Marshall (2000 [1950]). It was later prominently taken up again by Esping-Andersen, who sums up pointedly: "The welfare state may provide services and income security, but it is also, and has always been, a system of social stratification" (Esping-Andersen 1990: 55). In this sense, the manner in which a welfare state is organised and financed, the kind of redistribution it is based on and the generosity of the provided benefits cause patterns of inequality in societies. In light of the present research question, it is not just important to establish *that* the welfare state has a stratifying function and that different regimes *differ* in how they stratify. The main interest should be *why* such a link exists in the first place, as the answer to this question might later reveal how to conceptualise the welfare state.

At a broad level, the three ideal-typical welfare regimes introduced by Esping-Andersen (1990) help to identify explanations for different stratification outcomes. Countries, which incorporate many features of the *Social-Democratic* ideal equalise the most. Because of universal programmes and high benefit generosity, they achieve lower income inequality and higher equality of opportunity. In contrast, countries approaching the *Liberal* ideal are characterised by social policies, which may even increase inequality. Targeted means-tested and ungenerous benefits, whose receipt is associated with stigmatisation potentially weaken the situation of those at the bottom of the social hierarchy even more and widen the gap between contributors and recipients. Lastly, welfare states approaching the *Conservative* spectrum tend to reproduce existing inequalities since benefits and services are closely linked to previous employment status and earnings. Furthermore, a strong focus on the male-breadwinner may even increase inequality as it marginalises women (Esping-Andersen 1990: 55–78; Sachweh & Olafsdottir 2012: 152–153; Esping-Andersen 2015).

Even though the primary aim of welfare states is not to equalise class structures, it can be argued that efforts towards more equality and equal opportunities have become increasingly important in political debates and agendas in the last decades of the twentieth century (Esping-Andersen 2015: 125). Thus, reducing social inequality can be seen as a more recent objective and function of welfare states. Lastly, we have to take into account that the stratifying role of welfare states not only manifests in objective indicators of social inequality. Since the latter is a highly controversial topic, it impacts attitudes about inequality and preferences for specific stratification patterns as well (Sachweh & Olafsdottir 2012).

Concretising such general explanations, a direct link between social policies and social stratification can manifest (for instance) in the generosity of income replacement in the sense that high income replacement can reduce the gap between income groups, while a minimal replacement can even widen existing gaps. Different patterns of stratification, manifesting in social inequality and mobility, are thus a side effect of the most basic function of lowering risks and meeting needs. They reveal how strong the impact of different principles behind welfare states can be on class structures, class struggles and the salience of class itself (e.g. Esping-Andersen 1990: 55). Therefore, the welfare state's role in shaping social stratification is still considered an independent function that goes beyond responsiveness and provision of security. The welfare state impacts social stratification directly by levelling out (or widening) income gaps.

#### **2.2.4 Activation: an Enabling and Incentivising Institution**

Another function of welfare states is to enable as well as incentivise specific behaviour. Overall, such activation is a relatively new agenda within welfare states, which has gained importance since the 1990s. A key trend, which has increased the promotion of activating policies, lies in the shift towards more individualisation of welfare provision. As discussed in the first part of this chapter, a variety of structural, demographic, and normative changes led to transformations of welfare stateness. Among the consequences of such processes is the rising importance of individual responsibility for one's own social security. Thus, behaviour preventing risks needs to be incentivised. This especially refers to labour market participation and privatisation of insurance schemes. In this context, the term *enabling state* gained relevance. Initially used to describe the Anglo-Saxon approach to welfare provision, this agenda is gaining momentum throughout Europe (Gilbert 2004: 42).

Activation is embedded foremost in the strategies of social investment that were discussed in a previous section of this chapter (cf. 2.1.3). Regardless of the policy field, the objective of such activating policies is always similar: to prevent risk by investing either in individuals or in the contextual structures surrounding them. In many cases, such investment is linked to the labour market. Since new risks often affect those, who were traditionally not part of the labour force, it is increasingly necessary to activate those particular groups. This applies especially to mothers (Kowalewska 2017; Dotti Sani & Scherer 2018), but also to other groups such as chronically ill or disabled individuals (Holland et al. 2011).

Within the field of social investment, the labour market plays an important role because the support of employment is seen as one way of securing individuals against risks such as poverty. Thus, active labour market policies (ALMP) include measures such as training, job-creation and more, which aim at increasing participation in the labour market and ending spells of unemployment by incentivising and enabling individuals to re-enter the labour market (Bonoli 2012). ALMPs are discussed critically in the literature and reviews confirm only partial effectiveness (Crépon & van den Berg 2016). While training and investments in human capital have an impact—at least in the long run—other measures such as public sector employment programs appear to be ineffective (Card et al. 2018), produce more costs than benefits or help one group at the expense of another (Crépon & van den Berg 2016). When it comes to increasing the chances of employment, ALMPs are more effective for women and long-term unemployed individuals and the impact of training manifests a few years after completing the programme (Card et al. 2018). Taking into account the reduction of risks, the relationship between investment in labour-market programs, reduction of unemployment and reduction of poverty varies considerably depending on wage structures (Cronert & Palme 2017).

Overall, while the importance of social investment within the labour market and beyond grows in policy-making, its actual effectiveness is not yet clear. Differences in the design and effectiveness of social investment between countries are the result of a number of issues. Here, not just the policies themselves differ, but since they are linked to more traditional social policy measures of the 'social protection' agenda (Ellison & Fenger 2013: 612), the combination of compensating policies (such as income replacement) and activating policies seems to determine their successfulness (Kuitto 2016: 445). In general, a growing emphasis on activating policies is in line with more general sociological contributions on social change and especially the individualisation of risk and responsibilities (Beck 1986; Beck & Beck-Gernsheim 2002), which speaks in favour of a continuous expansion. Furthermore, considering that the Europe 2020 strategy for smart, sustainable and inclusive growth includes social investment as a key element (European Commission 2013), convergence of social investment policies within the European Union is possible. However, there is evidence, that the continuous introduction of new policies in line with the social investment perspectives came to a halt with the Euro crisis (Ronchi 2018).

Regardless of its successfulness, *activation* is an important function of welfare states as it represents an increasingly relevant goal in policy-making and a distinct way of approaching social security. It represents a possible strategy to respond to new risks on the one hand and lower dependency on income replacement and other state-provided benefits on the other.

In sum, this active role of the welfare state is somewhat new, but it still goes along with the more classical function of providing income replacement and other benefits. Variations between countries can thus be attributed to how they combine an activating agenda with more conventional measures. For instance, what gained popularity under the term *flexicurity* describes the combination of extensive active labour market policies and generous benefits (Bonoli 2013: 15). However, old and new approaches to welfare provision can also undermine each other. Using the example of family policies, incentives in one field (such as female employment) can be combined with deactivating measures (such as monetary incentives for staying at home) in another. Other variation stems from the character of activating policies. As Kowalewska (2017: 4) summarises, a broad differentiation of activating agendas entails identifying the extent to which either employment or investments in human capital are emphasised. Moreover, the activating agenda of the welfare state is not restricted to classical labour market policies, but can also manifest in other areas such as family policies (Bonoli 2013: 26). For example, defamilialisation represents a key component of enabling parents in general and mothers in particular to participate in the labour market.

Lastly, while the activation of individuals is often an explicitly stated objective, which clearly manifests in policies, welfare states can also incentivise behaviours unintendedly. Again, the case of family policies can serve as an example, where generous benefits for stay at home care responsibilities can potentially deactivate women with regard to their labour market participation. This can be an unintended consequence of attempts to incentivise other behaviour such as child-bearing. Beyond this example, such unintended deactivation would be in line with neo-Marxist theory, which assumes that truly extensive social benefits (in any policy field) would disincentivise labour market participation and thus counteract the principles of capitalist market economy (Quadagno 1987: 114–115).

#### **2.2.5 Socialisation: the Welfare State, Endogenous Norms and Values**

The last objective discussed in this chapter, is to socialise individuals. While responding to risks, shaping inequality and redistributing resources are often politically steered objectives, socialisation is not a specifically targeted goal. Instead, the welfare state acts in an invisible and often unintended manner in this case.

There is a philosophical side to this issue. In a sense, welfare states can be understood as manifestations of certain ethical ideals. They represent principles of egalitarianism and they affect individual autonomy and responsibility (White 2010). If the nature of a welfare state is thus not just perceived as a system of security and inequality but also as a cultural authority, it can affect individuals much more than just through socioeconomic channels. The kind of equity promoted in different modes of redistribution on the one hand, and different ways of shaping social inequality on the other, dictate solidarity and justice principles. In this sense, welfare states can be seen as institutionalised ideas about social justice (Sachweh 2016). However, not only justice principles can be linked to social policies. As Gangl and Ziefle (2015: 511) summarise, they also legitimise and institutionalise aspects in other domains such as gender role beliefs, which in turn manifest in labour market participation. The welfare state is especially seen as a socialising institution when it comes to shaping outcomes associated with endogenous norms (Lindbeck 1995; Bisin & Verdier 2004).

Overall, key characteristics of the welfare state and the main principles underlying redistributive efforts (such as universalism or targeting) represent distinct manifestations of solidarity and justice principles (Arts & Gelissen 2001), gender equality (Bonoli & Natali 2013b: 3) and more. Living or growing up in a welfare state thus means being embedded in a cultural context, which represents and legitimises specific ideals (Pfau-Effinger 2005). This can influence individual perceptions of such ideals and lead to an increased probability of adapting underlying principles—such as egalitarian views. It may in turn even produce a feedback effect (e.g. through political participation) and increase egalitarian social policies (Lindbeck 1995: 488). In this line of argument, living in a welfare state during the formative years can—for instance—shape attitudes towards related issues such as redistribution (Neundorf & Soroka 2017). It also influences role-models related to gender and promotes and facilitates how such models manifest in reality—for example when it comes to division of housework (e.g. Fahlén 2016) or norms about the balance between labour market participation and care responsibilities for women (Barbieri & Bozzon 2016: 103). Especially in the field of gender inequality, social policies and particularly family policies are closely related to gender ideologies (Grunow et al. 2018).

**Figure 2.3** The welfare state, norms and principles. *(Modified version of theoretical model proposed by Arts & Gelissen (2001: 288))*

Arts and Gelissen (2001) propose a theoretical model capturing how welfare state regimes are linked to the formation of normative frames. Their model can be generalised in some respects in order to be applicable to areas beyond solidarity and justice principles (such as principles of gender equality). Figure 2.3 provides a display of such a generalised version of their approach and sketches the processes described by Arts and Gelissen. In short, the welfare state triggers cognitive processes, in which individuals (1) learn the dominant norms, (2) form habits and (3) frame situations. This is of course not a deterministic process and welfare states do not strictly impose their endogenous norms and reference frames. However, they possess the potential to coin individual perspectives in relevant areas. The three cognitive processes can thus manifest on the individual level as knowledge of rules and principles and their habituation as well as in adapting to normative frames with varying intensity, leading to variations in the distribution of such notions on the aggregate level. To elaborate using the above-mentioned examples, the socialising function of welfare states can manifest as varying degrees of adopting principles of gender equality or universalism underlying the nature of different welfare states.

In sum, the function of socialisation is tied to what is sometimes referred to as a 'welfare culture'. Here, social policies and ideologies are closely interrelated (Pfau-Effinger 2005). They present a context in which individuals are embedded and thus shape what is perceived as legitimate and dominant principles within policy arrangements.

#### **2.2.6 Summary: Functions and Variations of Welfare States**

The brief summary of the broad literature surrounding each of the functions highlighted in this chapter, gives various reasons why welfare stateness is an important independent variable. It also reveals a variety of potential dependent variables beyond the most obvious ones. This includes manifestations of inequality (and especially poverty), attitudes, behaviour and many other phenomena, which appear to be closely tied to the main functions of the welfare state. Furthermore, it shows that some of the functions are related. The role of the welfare state in addressing risks and needs for instance is strongly related to the redistributive function, as redistribution is one of the most important tools in financing social security. In this sense, redistribution is not only a function of welfare states, but also describes their modus operandi.

Furthermore, some functions represent explicit policy objectives (such as risk reduction), while others are characterised by rather subtle or unintended consequences. For example, the stratifying function has only recently become a more clearly defined objective in some European countries. As was argued, reducing inequality and increasing equal opportunities are no longer just a by-product of different social policy arrangements, but can be an explicit political goal. Activation can be such an explicit policy goal as well—for instance in the case of active labour market policies or social investment—it can however also be an unintended consequence, triggering individual behaviour unintentionally. The briefly discussed disincentivising effects are mostly due to such unintentional functions as well—in many instances one could even call them dysfunctions. Lastly, the socialising function represents another example for a process that is in most cases unintended. Albeit following different objectives and leading to different outcomes, the functions are related to one another. Figure 2.4 graphically summarises this.

**Figure 2.4** Summary of introduced functions of welfare states

#### **2.3 From Dependent to Independent Variable**

As a key element of modern democracies, the welfare state is an important object of research in the social sciences—especially in cross-cultural comparative research. Due to differences emanating from historical trajectories and extents to which the economic, societal and demographic changes since the second half of the twentieth century affected countries, the study of the welfare state is characterised by considerable complexity. Different conceptual and empirical perspectives on the welfare state discussed in the first part of this chapter reveal a variety of approaches and they outline distinct historical and theoretical conceptualisations and empirical operationalisations. However, the summary in this chapter also shows a considerable amount of heterogeneity in the approaches and the literature is far from reaching a consensus when it comes to conceptualising the welfare state. Furthermore, while the discussion of literature on the welfare state as a *dependent variable* may reveal popular positions and debates, we still know little about how scholars include the welfare state as an *independent variable*—theoretically as well as empirically.

Reasons for treating the welfare state as an independent variable were discussed in the second part of this chapter. While a great share of the literature focusses on how to grasp different social policy arrangements—conceptually and empirically—, this discussion shows that the welfare state also has important functions. These functions reveal why welfare states in general and social policies in particular are not only important dependent variables but also important independent ones. Depending on how welfare states fulfil the different tasks (either intended or unintended), a variety of other phenomena is impacted. Even though I focussed mainly on the description of functions so far, this already hints at important outcomes. Social security shapes individual risks and reduces uncertainty, redistribution shapes policy preferences, socialisation shapes attitudes, incentives trigger behaviour, and social stratification—from a sociologist's perspective potentially shapes almost everything. To explore the welfare state or features of it as an explanatory factor—not only when examining outcomes on the societal but especially those on the individual level—is more than evident. This leads us to essential questions, which are addressed in the following chapters of this book. How can the welfare state be operationalised if treated as an independent variable? Does this require different perspectives on the welfare state than the ones present in comparative welfare state analysis? And if so, what might such different perspectives look like?

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **3 The Welfare State as an Independent Variable: Debates, Pitfalls, Potentials**

As discussed in the previous chapter, there are many reasons for including the welfare state as an independent variable. It is an inseparable part of modern democracies, and its functions are far-reaching and consequential for individuals and societies alike. However, the very premise of this book is that there is a distinct lack of feasible recommendations and guidelines for scholars trying to include the welfare state as an *independent* variable in multilevel analyses. This chapter outlines the reasons for this.

So why is there a deficit at all? As mentioned in the preceding chapter, comparative research on the welfare state has always included a strong emphasis on modelling welfare policies empirically. There are comprehensive discussions about appropriate indicators—for instance surrounding the debate on spending vs. social rights (section 2.1.1) and the dependent variable problem (section 2.1.3). Equally extensive discussions address empirical classifications of welfare regime typologies (section 2.1.2). Thus, there is actually a great number of contributions dealing with methodological issues and proposing specific ways of operationalising different welfare state arrangements. However—as I will show in this chapter—the fact that these operationalisations aim at researching the welfare state its premises and functions per se and as a dependent variable is frequently passed over. As a result, measures are adapted under the implicit assumptions that (1) what is eligible as dependent variable should be suitable as explanatory variable as well and (2) that differences between operational approaches are negligible because they capture similar or at least strongly related elements of the same construct—the welfare state. I will argue that both assumptions are flawed and that they influence the comparability of results in a negative way.

This chapter critically examines different empirical approaches conceptually as well as empirically. Out of the operational strategies outlined briefly in the previous chapter, I discuss those in more detail that represent especially popular independent variables in the current state of research. More specifically, several examples for *single indicators*, *typologies*, and *composite indices* are inspected more closely in order to illustrate immanent problems. First, all three approaches are discussed conceptually with an emphasis on sources of dissent within each approach. In the next step, popular operationalisations are compared in empirical analyses of cross-national survey data from the International Social Survey Programme (ISSP) and the European Social Survey (ESS) in order to explore the consequences of different conceptual choices. This is followed by a discussion of possible points of departure for the development of more suitable and standardised operationalisations for the specific use as explanatory variables.1

#### **3.1 Approaches and Debates**

As the previous chapter revealed, there are many ways of approaching the welfare state and its functions. These different approaches are mirrored in the existing empirical operationalisations. While early research mainly focussed on welfare state effort—in most cases operationalised through social expenditure contemporary literature agrees that social policy arrangements are captured more adequately by focussing on social rights of citizenship (e.g. Esping-Andersen 1999; Stephens 2010). Both conceptualisations can still be found in empirical operationalisations of welfare stateness. Moreover, additional approaches and respective empirical measurements were introduced (such as social investment and benefit receipt). Regardless of their content, the operationalisation can result in three types of indicators, which will be discussed in more detail in the following sections: single indicators, typologies, and composite indices.

#### **3.1.1 The Single Indicator Approach**

A very popular way of operationalising different welfare policies is to utilise single indicators highlighting specific elements of the welfare state. They are used in

<sup>1</sup> For a more comprehensive discussion of the issues addressed in this chapter cp. Kunißen (2019).

literature on classifying regimes,<sup>2</sup> as well as in studies that treat characteristics of the welfare state as independent variables (e.g. Jæger 2006; Jordan 2013; Eger & Breznau 2017).

By far the most popular are expenditure-based indicators (Kvist 2011). Usually, this means including a variable on social spending as a percentage of GDP in one specific policy area (e.g. in the labour market, Schneider & Makszin 2014) or as an overarching measure (e.g. Steele 2015) of welfare effort. Such indicators receive much criticism, as discussed in the preceding chapter (cf. section 2.1). Main criticism includes that other areas of social policy-making—for instance entitlement criteria—are more important and that a focus on spending postulates a linearity of welfare *efforts* which is not given in reality (Esping-Andersen 1990) and disregards the multifaceted nature of welfare states (Bonoli 1997). Furthermore, high spending can be a signal of a generous system, but also a consequence of a higher number of people depending on social benefits (Bergqvist et al. 2013). From a comparative perspective, we are therefore confronted with the issues that equally high spending may not necessarily mean two countries actually provide similar benefits (Kvist 2011), and we cannot determine if higher or lower income groups profit more from redistribution (as already noted by Titmuss 1974). Such criticism led to a widespread consensus that social expenditure is a problematic operationalisation of welfare stateness (for a more differentiated discussion see Jensen 2011). Still, it has not been completely discarded, as some scholars point out that despite justified criticism, social expenditure is a good indicator of a country's commitment to social transfers and services (Reinprecht et al. 2018: 785).

An alternative is to use net replacement rates (NRR) for individuals in particular risk positions. In many cases, such replacement rates are seen as indicators of welfare *generosity*, which is an important part of the social rights perspective (cf. section 2.1). However, how to calculate replacement rates is still controversial and they vary depending on the source. This is discussed among others by Scruggs (2013), Wenzelburger and colleagues (2013) and Ferrarini and colleagues (2013), who explore differences between replacement rates calculated in the Comparative Welfare Entitlement Dataset (CWED2, Scruggs et al. 2014) and the Social Citizenship Program (SCIP, later included in the Social Insurance Entitlement Dataset (SIED)).<sup>3</sup> More recently, Bolukbasi and Öktem (2018) add that

<sup>2</sup> As indicators underlying the construction of typologies or composite indices.

<sup>3</sup> Among other issues, differences in the replacement rates are due to the calculation of taxes and the referenced period of time in which a benefit is received.

other non-replacement indicators—such as waiting days and qualification periods—are affected by the same problem and also differ depending on the data source because similar indictors are operationalised based on varying conceptual premises. Further ways to operationalise welfare state generosity using single indicators include the share of income that comes from welfare transfers (e.g. Brady et al. 2017).

The third perspective on the welfare state, which can be modelled in single indicators, focusses on *benefit receipt* (cf. section 2.1.3). Here, actual cash benefits are aggregated from survey data (e.g. Otto 2018a). Since this take on the operationalisation of welfare stateness is only starting to receive more attention (van Oorschot 2013) and is not particularly present in literature using the welfare state as independent variable, it is only partly relevant at this point. A similar argument can be made for *social investment* as an additional fourth perspective, which is only relevant for very specific research questions and, therefore, has not been used often so far.

Using single indicators for the two most common perspectives on welfare stateness—effort and generosity—as *independent variables* in comparative research has advantages and disadvantages, both of which are visible in the existing literature. The two main disadvantages address their limited informative value on the one hand and the above-mentioned deviations in the calculations on the other hand. In empirical studies, these disadvantages are often outweighed by the main advantage of this operationalisation: since a variety of international organisations such as the OECD and Eurostat offer extensive and regularly updated information on key indicators, data are easily accessible and available for a great number of countries.

A way to overcome the problem of limited informative value is to use more than one indicator. There are many studies that refer to a theoretically wellgrounded selection of several single indicators representing relevant areas of the welfare state (e.g. Jæger 2006), explain in detail why they prefer a single indicator to other operationalisation (e.g. Jakobsen 2010; Visser et al. 2018), or examine single indicators together with other operationalisations (e.g. Jakobsen 2011). However, there are also studies, which only briefly elaborate on their selection. This is problematic as there is an obvious conceptual difference between using, for example, replacement rates and social expenditure. Still, studies often refrain from justifying their selection, instead just arguing that they would have liked to use an alternative (e.g. a composite measure) that was not available for their sample of countries or time periods (e.g. Angel & Heitzmann 2015; Kulin & Meuleman 2015).

As for the second disadvantage, to my knowledge there is no study that analyses the consequences of deviations between data sources when using single indicators as independent variables. I therefore recommend that further research not only justifies the selection of each specific indicator, but also discusses the sources of the macro-level data in more detail and compares the selection with the referenced literature.

In conclusion, it can be stated that the use of single indicators as operationalisations of welfare stateness in multilevel frameworks has its limits. As there are no recommendations on which indicator to choose when modelling specific causal assumptions, the choice requires a well-founded justification. Given the different operationalisations, failing to do so can have consequences for the results and their comparability with other studies that use other measures or data sources.

#### **3.1.2 The Regime Typology Approach**

Using a regime typology is another popular approach to operationalise the welfare state. As outlined in the previous chapter, this is also surrounded by a broad debate. While the introduction of new typological approaches seems to have slowed down (cf. section 2.1), this is not the case when it comes to its application as an independent variable.

When exploring in more detail the potentials and pitfalls of using regime typologies as independent variables, the first observations is that the popularity of Esping-Andersen's (1990) *Three Worlds of Welfare Capitalism* (TWWC) is present here as well. As outlined before, his threefold classification, identifying a generous *Social-Democratic*, a status-oriented *Conservative*, and a market-oriented *Liberal* regime has turned into a true classic. Research following Esping-Andersen's initial typology has introduced a great number of varying classifications and it inspired a remarkable body of literature and a critical and ongoing discussion regarding the number, composition, and scope of regimes (comprehensive discussions are provided by Arts & Gelissen 2002; Ferragina & Seeleib-Kaiser 2011; Rice 2013; van Kersbergen & Vis 2015; Powell et al. 2019). Before discussing the applicability of typologies as independent variables, it is important to look closely at methodological sources of dissent between different classifications, which address conceptual as well as operational details.

In classifications of typical arrangements of social policies, scholars have focussed on different elements of the welfare state. While some focussed on *how much* a welfare state spends, others classified *how* social policies are organised and financed (Bambra 2007 and Bonoli 1997 discuss and combine both perspectives). Another lively debate surrounds the question *how many* welfare states exist. Popular additions to Esping-Andersen's typology include a Mediterranean (e.g. Ferrera 1996) and a post-socialist welfare regime (e.g. Castles & Obinger 2008).

As noted before (cf. section 2.1.2), such classifications are often referred to as *ideal types*. However, in empirical studies—and eventually most of the typologies are tested empirically—this bears potential for confusion. As soon as empirical evidence is interpreted and countries are clustered based on actual indicators of policy-making, the resulting classification actually captures *real types*. Two problematic issues are frequently raised in this context. First, the practice of historical argumentation is criticised. Rice (2013) for instance argues that in order to provide a sound groundwork, the idea of a historical deduction of types should be abandoned and replaced with a purely ideal–typical one which is detached from empirical evidence and focusses solely on overarching dimensions (she proposes welfare culture, welfare institutions, and socio-structural effects). Second, the misuse of the term *ideal type* is denounced. As van Kersbergen and Vis (2015) point out, most literature mainly offers typologies of real cases instead of an actual deduction based on ideal types. Even Esping-Andersen's final classification of countries bears more resemblance to an empirical typology than to an explicit operationalisation of approximation. Thus, the fact that the term is very present in literature describing how countries cluster into 'ideal–typical' groups gives a false sense of theoretical justification and distorts what ideal types should accomplish by definition.4 However, instead of using them as a point of reference, they often serve as a template to be reproduced in reality in studies on welfare state regimes. An actual empirical attribution of real cases to ideal types would model the proximity between a case and an archetype instead of a deterministic assignment. In practice, this would mean, for instance, that rather than using hierarchical cluster analysis to achieve a classification that is later labelled according to ideal types, one could look at how far each country departs from an ideal score on the examined variables. This of course raises the issue of what the 'ideal' would actually look like in terms of an empirically measurable point of reference. The role of ideal–typical welfare regimes, the insights they offer and their informative value are subject to an ongoing debate (Aspalter 2011; van Kersbergen & Vis 2015) and the fact that real types only approach ideal types

<sup>4</sup> According to Weber, they present an analytical tool – an exaggeration or even a utopia – reality can be measured up against. They neither do nor should coincide with reality Weber (1973: 190) and instead present a "middle ground between the uniqueness of historical events and the generality of laws" Ragin and Zaret (1983: 732).

is frequently voiced (Arts & Gelissen 2002; Kääriäinen & Lehtonen 2006) but rarely modelled. Furthermore, the debate mainly addresses the value of ideal and real types as classifications alone and not as concepts, which could serve as an explanatory variable.

There are numerous indicators and methods for the empirical operationalisation of such types that reflect the different conceptual considerations. While some studies base their classifications on expenditure (Kuitto 2011), others focus on benefit coverage and replacement rates (Ferrera 1996), or on a two-dimensional approach combining spending and funding of welfare provision (Bonoli 1997; Bambra 2007). Moreover there are those who add measures of economic insecurity (Menahem 2007) or stratification (Esping-Andersen 1990). These indicators are empirically merged into typologies through different analytical techniques and each methodological approach claims to shed light on aspects, which have been disregarded so far (e.g. certain indicators or countries).

Lastly, the country sample constitutes a considerable source of variation. Any classification is highly dependent on the sample upon which it is drawn. This insight, albeit having been voiced more prominently in the years following the boom of typologies, is not new and predates the TWWC (Uusitalo 1984). However, it is still common practice to develop typologies based on a sample that is neither random nor systematic. Even though consequences of case-selection have been addressed sporadically (Ebbinghaus 2011; Kim 2015), the most prevalent criterion appears to be data availability. Thus, countries are often chosen because they belong to an organisation, such as the OECD, which publishes a comprehensive amount of data on social policies of their members. In addition, belonging to any of these organisations means at least a minimal amount of similarity in the economic and political development is guaranteed which in turn often serves as a justification for comparability (Ebbinghaus 2011). Still, most studies only cover a selection of those countries and especially the Central and Eastern European (CEE) countries are highly underrepresented even though a meaningful complete survey for instance, could be achieved by examining *all* member states of the European Union. Bearing in mind possible applications as independent variable, the latter proceeding would be particularly fruitful since EU citizens are a commonly chosen population on the micro-level due to a multitude of research question, which includes attitudes, behaviour and living conditions in light of European Integration.

Apart from the oversight of countries, different samples may affect the classification itself because most approaches determine types based on proximity between cases. For instance, Esping-Andersen's (1990) classification is based on composite indices of decommodification and stratification where countries receive a score based on their deviation from the overall mean. However, mean and deviation vary depending on the included countries and are sensitive to slight changes or miscalculations. Ironically, Esping-Andersen himself serves as an example for this.5 A similar argument applies to cluster analysis (e.g. Castles & Obinger 2008; Kuitto 2011), which groups countries based on the proximity between them. In light of these differences in conceptualisations and operationalisations, it does not surprise that the number, title, and composition of regimes differ remarkably between typologies.

The lack of agreement on which typology suits best and which theoretical perspective is preferable is acknowledged in many studies using them as *independent variables*. Nonetheless, many of them still rely heavily on the regime approach sometimes even with an apologetic reference to the need to circumvent a more detailed discussion of the scientific debate (e.g. Motel-Klingebiel et al. 2009: 70). While regime typologies bear the advantage that they are easily operationalised as dummy variables, their main disadvantage is a practical one: the selection of countries in survey data (like the ESS) usually deviates from the countries covered by a typology. Hence, authors face a difficult conceptual choice having to either exclude unclassified countries or include them by combining classifications or extending them. Since cross-cultural analyses often aim at examining as many countries as possible, the second option is preferred. Such combination or extension often relies on instinct since the literature lacks consensus on what to do in this situation and there is a plethora of different typologies. As a result, a *buffet strategy* evolved in which authors pick a combination "from the vast array of welfare state typologies" (Arts & Gelissen 2001: 285) that seems helpful for the envisioned purpose. There are many examples for such buffet-approaches (more recently Deeming & Jones 2015; dem Knesebeck et al. 2016; Arundel & Lennartz 2017; Schuck & Steiber 2017). The proceeding often seems inspired more by practical considerations than by theoretical ones. As a result, we see many modifications adding countries that were not classified in whatever typology serves as a starting point, as well as uncommented reclassifications. In light of the existing debate on welfare state change and new risks (cf. section 2.1.3), it furthermore seems problematic that many of the buffet-type studies still rely heavily on typologies from the 1990s and assume that those classifications (very prominent are Esping-Andersen 1990 and Ferrera 1996) are still valid and only require some additions or slight modifications. A last and very general problem associated with regime typologies is that they represent strong reduction of

<sup>5</sup> A miscalculation in the decommodification score for several countries led to a misclassification of several countries (e.g. discussed by Bambra 2006).

complexity. This marginalises variation between countries by reducing variation between countries to a handful of types (Kvist et al. 2013: 331).

It has rarely been tested how different typologies affect results if treated as independent variables. Bergqvist and colleagues (2013) provide one of the few overviews using the example of health inequality as dependent variable. In their re-analysis of 34 studies employing regime typologies as independent variable they found not only considerable differences in the kind of typology used and the amendments made to classifications but also in the results. Since different associations with health were even found within identical typologies, they conclude that the main problem is not the theoretical and empirical conception but the general use of welfare regimes as an explanation for health inequality. However, they examined studies, which draw on different data sources and apply different methods of analysis. Thus, it should be tested if their finding holds true if these aspects were kept constant.

In summary, regime typologies can be an excellent tool for classifying different policy arrangements. However, they rarely fit the country sample in cross-national survey data, prompting scholars to resort to combinations and reclassifications. Given the strong conceptual and operational variations underlying different typologies, such an approach appears highly problematic. It is therefore important to examine the consequences of different classification more closely.

#### **3.1.3 The Composite Index Approach**

Composite indices and scores to measure welfare stateness represent a comparatively rare approach. Nevertheless, attempts to develop such indicators exist throughout the literature (e.g. Castles & McKinlay 1979). In particular, the two indices underlying Esping-Andersen's (1990) TWWC typology have had a major impact on more contemporary approaches. Especially his *decommodification index* has been replicated, updated and revised numerous times (e.g. Bambra 2005; Scruggs & Allan 2006; Scruggs 2014; Kuitto 2018). Noteworthy are furthermore works of Segura-Ubiergo (2007) and Cruz-Martinez (2014), who develop multidimensional measures of welfare state arrangements for Latin American countries. However, their proposals have not been adapted for European samples so far. Other composite measures in the literature either take a more specific perspective (e.g. on defamilialisation, Lohmann & Zagel 2016) or a more general one which goes beyond characteristics of social policies and includes overall features of governance (e.g. the Social Policy Index6). The main sources of dissent within the index approach include the operationalisation and country sample.

Hereinafter, some examples for differing operationalisations are named: Castles and McKinlay (1979) devise an index of welfare commitment based on educational expenditure, transfer payments, and infant mortality, Esping-Andersen's (1990) *decommodification index* includes replacement rates, extent, and duration of individual contribution, waiting periods and insurance coverage, and Menahem (2007) combines insurance coverage and replacement rates with disposable income. Besides these obvious differences in the choice of indicators, there are also differences in regard to weighting procedures and modes of standardisation. The Benefit Generosity Index in the Comparative Welfare Entitlement Dataset—an updated and slightly modified version of Esping-Andersen's decommodification index—z-standardises the underlying variables (Scruggs 2014). In contrast, Esping-Andersen's original version using data from the Social Citizenship Program gives countries a value between one and three for each underlying indicator representing levels of generosity and adds them up. Furthermore, Esping-Andersen only superficially justifies why some indicators are given more weight than others (discussed among others by Bambra 2006). However, as Wenzelburger and colleagues (2013) point out, not just the modes of combining indicators vary, the underlying indicators themselves may differ as well depending on the data source (as discussed in the preceding section on single indicators).

The second source of variation within the approach is closely linked to the first. The measures introduced above all rely on mean values and deviations from that mean and are thus very sensitive to the underlying country sample. If the composition of countries changes, these values will most likely change as well (as discussed in the case of typologies). This compromises the comparability of results and it impairs stretching composite measures to further countries. A way to overcome this problem, which I rarely encountered in the literature, is refraining from standardisations based on mean and deviation. An alternative is a benchmark approach, which standardises based on the highest existing occurrence of a given indicator in a meaningful population (as used by Sjöberg 2017). Such a population could for instance consist of the entire European Union or all OECD member states. In this case, the standardised score indicates how close a country

<sup>6</sup> Naren Prasad at the United Nations Research Institute proposed the SPI for Social Development in 2006. No final version of the index has been published apart from a research proposal, which has been picked up by other researchers (e.g. Garcés Ferrer et al. 2016).

is to an existing frontrunner (for instance the highest existing replacement rate) and they could be used independently of the country sample.

Composite indices are perhaps the most desired but least implemented independent variables (exceptions include e.g. Rothstein et al. 2012). They promise the multidimensionality of typologies while maintaining the metric scale and variation of single indicators. However, the number of existing measures is very limited and the most popular ones are only available for a limited selection of countries and points in time. This shortcoming is often stated as a reason for having to resort to a less desirable alternative (e.g. Angel & Heitzmann 2015; Kulin & Meuleman 2015).

Overall, composite measures represent very promising tools for capturing welfare stateness. However, since the most comprehensive ones cover only a small number of countries, their usefulness as independent variables is very limited at this point.

#### **3.2 An Empirical Confrontation**

In the following section, the discussed operationalisations are tested empirically with an emphasis on illustrating the advantages and disadvantages mentioned before. In this empirical test, welfare attitudes serve as an exemplary dependent variable on the individual level to illustrate the consequences of differing operationalisations. Welfare attitudes represent a very popular dependent variables in the relevant literature. In very broad terms, it is believed that attitudes towards social policies are shaped by the institutional context in which individuals are embedded—in this case the welfare state (Svallfors 1997; Arts & Gelissen 2001). A prominent hypothesis holds that generous and universal social policies, following social-democratic principles, generate political support, and positive attitudes towards the welfare state (Jaime-Castillo 2013; Roosma et al. 2014), while redistribution-based and targeted polices increase conflicts between beneficiaries and contributors and lead to disapproval (Jordan 2013). However, the empirical tests of this *policy feedback* hypothesis produce mixed results and several studies cannot confirm such a linear relationship between generosity and support (Jæger 2009; Jakobsen 2011). One reason for this may be that different operationalisations of welfare policies were tested—including different typologies and single indicators. While typologies may fail to grasp subtle differences between welfare states (Jordan 2013), single indicators could be correlated with other macroeconomic indicators and thus may have no independent effect once other variables are controlled (Jæger 2013 suspects this in the case of social expenditure). Due to these divergent findings and the ongoing discussion, welfare attitudes present a good example of a micro-level outcome, which may be explained differently depending on the conceptualisation of welfare stateness in an analysis. In this chapter, the focus rests on determining how sensitive results are to such different operationalisations, while the results themselves and their relation to hypotheses in the literature is of secondary importance.

#### **3.2.1 Data, Operationalisation and Method**

The following analyses are based on data from the fourth wave of the European Social Survey (ESS 2008) and the International Social Survey Programme (ISSP Research Group 2017). These two data sources are chosen for several reasons. First, they both include similar questions addressing attitudes towards the welfare state. Second, the data were collected over a similar period of time (mainly 2008 and 2009), which means that the same macro-level indicators can be used in both analyses. Third, both datasets are frequently used in comparative research on how welfare attitudes are shaped by different welfare state arrangements (more recently Kulin & Meuleman 2015; Steele 2015; Eger & Breznau 2017). Fourth, using ESS and ISSP data represents a common situation in which the researcher has no influence on the country selection. Lastly, the comparison between the two datasets will allow to determine—at least partly—the reliability of findings.

To ensure the examined population is suitable for the proposed analysis and covers comparable units of analysis, the sample is reduced to respondents from countries that are member states or have strong ties to the European Union.<sup>7</sup> As a result, 21 countries covered by both datasets are included.8

An item is chosen as the *dependent variable* that measures attitudes towards government responsibility for supporting the unemployed. This particular aspect of social policy attitudes is covered in both datasets in a comparable, if not identical, manner. The ESS includes the question "*how much responsibility do you think governments should have to ensure a reasonable standard of living for the unemployed?"* on an eleven-point scale ranging from "*should not be governments' responsibility at all"* to *"should be entirely governments' responsibility"*. In the ISSP, respondents indicated on a five-point scale to what extend they agreed with the statement *"the government should provide a decent standard of living for the unemployed"*.

<sup>7</sup> EU member states (in 2008) plus Norway and Switzerland.

<sup>8</sup> AT, BE, BG, CH, CZ, DE, DK, EE, ES, FI, FR, GB, HU, LT, LV, NO, PL, PT, SE, SI, SK.

The analyses focus on *independent variables* on the country-level. Since the main surveying period for both datasets was late 2008 and early 2009, these indicators are primarily based on 2008 data. The only exception is SCIP/SIED data, which is available at five-year intervals and is therefore from 2005. Furthermore, since the dependent variable addresses attitudes towards generosity in the field of unemployment, macro-level indicators, which relate to unemployment policies are chosen, whenever possible.

Four single indicators are tested as independent variables: overall social expenditure as percentage of GDP (Eurostat 2018a), social expenditure in the field of unemployment policies (Eurostat 2018b), and two versions of net replacement rates for unemployed average production workers, which stem from different data sources and are based on slightly varying operationalisations (CWED2 and SCIP/SIED). These two types of single indicators were chosen because they are especially popular in the relevant literature.

Since there are no typologies covering all analysed countries, two different *buffet-typologies* are included. The first version uses Esping-Andersen's classification as a starting point and adds a Southern type following Ferrera (1996). The CEE countries were all joined in an Eastern-European group by applying classifications used, amongst others, in analyses by Roosma and colleagues (2014) and Bambra and colleagues (2014). This leaves Cyprus (only included in additional analyses), which was classified as Southern following Castles and Obinger (2008). The second buffet-typology differs from the first in the classification of two countries, which represent ambiguous cases: Switzerland is classified as Liberal (instead of Conservative) following Obinger and Wagschal (1998) and Ferragina and colleagues (2013) and Austria is assigned to the Social-Democratic type instead of the Conservative one, which is supported by Arts and Gelissen (2001).

The *Welfare Generosity Score* provided in the CWED2 dataset is included as a composite measure. Since it only covers a small sample of countries and none of the CEE states, I added some missing indicators9 and updated the index following instructions by Scruggs' (2014) so that it now covers all 21 countries in the main analysis. The correlation of my version with the unemployment generosity score already provided in the dataset is very high (0.98) for the 12 countries shared by CWED2, ISSP, and ESS.

<sup>9</sup> Missing data on coverage of unemployment insurance was added from the SIED (SPIN 2015).

Furthermore *unemployment rate* is included as control variable in all models, as is often done in analyses of welfare attitudes (Jæger 2013; Arikan & Ben-Nun Bloom 2015; Eger & Breznau 2017).<sup>10</sup> In this specific case, it can also be seen as a proxy for benefit receipt, which is relevant because the dependent variable focusses on attitudes towards government responsibility for meeting the needs of the unemployed. Testing these different operationalisations within each of the two surveys should help illustrating differences while reducing potential bias stemming from varying survey periods and country samples.

The following empirical tests are based on *multilevel linear regression analyses* (MLA). During the last decades this method has become increasingly popular in comparative research because it takes into account the hierarchical structure of cross-cultural data in which individuals are nested in national contexts. Multilevel analysis is able to estimate variance components on the level of individuals and contexts (in this case countries) simultaneously. This leads to a more correct estimation of standard errors and reduces the risk of fallacies, which can arise when results on either level are translated to the other (cf. section 1.2). Moreover, it allows to estimate the effects of independent variables on the micro- and macrolevel in the same analysis (for a more detailed description see among others Hox 2010; Snijders & Bosker 2012; Marx et al. 2013). In order to estimate effects between countries in a regression framework, a sufficiently large country sample is required. With 21 countries, my analyses are at the lower end of what is recommended, but should produce reliable results as long as the model specification is not too complex (Stegmueller 2013).

In essence, the idea of applying multilevel linear modelling in cross-national comparative analyses is to extend regular regression analyses such as ordinary least squares (OLS) by explaining variance of a dependent variable on two levels: within and between countries. The resulting regression equation for the full model predicting attitudes among individuals *i* in country *j* and including all variables at the individual (*xi j*) and country level (*z*<sup>0</sup> *<sup>j</sup>*), is:

$$(\text{y}(attribute)\_{ij} = \text{y}\mathbf{0}) + \sum\_{h=1}^{r} \eta\_h \mathbf{x}\_{hij} + \sum\_{l=1}^{r} \eta\_l \mathbf{z}\_{li} \mathbf{u}\_l + \mathbf{u}\_{0j} + \mathbf{e}\_{ij}$$

Here, the intercept (γ00) represents a general mean and—in contrast to nonhierarchical regression models—two residual components are distinguished. One of them represents the residual effect on the level of countries (*u*<sup>0</sup> *<sup>j</sup>*) and the

<sup>10</sup> GDP would present another macro-level control variable. Since the effect of expenditure (as percentage of GDP) should not be confounded, it is not included in this analysis.

other on the level of individuals (*ei j*). This formulation of a multilevel regression model is a simple version of a hierarchical linear model, in which only the intercepts and the residual terms are assumed to vary randomly, while the slopes are fixed (*Random-Intercept-Fixed-Slope-*model*)*. This means that differences between countries are assumed when it comes to the general level of attitudes, while we do not expect the strength and direction of the effects caused by the independent variables to vary between countries. The latter (a random slope) is usually included if cross-level interaction effects are assumed because there is reason to believe that the slope of individual-level determinants varies between countries (cf. Snijders & Bosker 2012: 74–87). Since this is not the case and the limited number of countries calls for a slim model, random slopes are not examined in the following analysis.

In addition, information on the models and their explanatory power is included. In this analysis, changes in variance are especially informative. They are obtained by calculating the *intraclass correlation coefficient* (ICC), which represents the share of variance that can be attributed to differences between countries. A high enough value of the ICC can be considered a pragmatic prerequisite for a multilevel analysis as there should be variance in the first place in order to explain it. In addition, Bryk and Raudenbush's (2012) R-squared is included in order to evaluate the explanatory power (especially on the level of countries).11

#### **3.2.2 Results and Interpretation**

The following two tables (Table 3.1 and Table 3.2) present the results of the multilevel analyses for each data source (ESS and ISSP). Both versions show very similar intraclass correlation coefficients (ICC) in the random-interceptonly model (M0): in both datasets, about 10 percent of the variation in attitudes towards the role of government can be attributed to the country-level.

Looking at the coefficients, many similarities can be found in the ESS (Table 3.1) and ISSP (Table 3.2) data, which indicates a certain robustness of the findings. In both analyses, *overall social expenditure* is negatively associated with supporting a strong role of government in the field of unemployment policies and explains a considerable amount of variation between countries (M1). *Social expenditure in the field of unemployment policies* (M2) points in the same direction, although this effect is only significant in the ISSP analysis. Respondents

<sup>11</sup> Based on the *mlt ado* for Stata (Moehring & Schmidt-Catran 2012).




*Buffet-typology 1: Soc-Dem.: DK, FI, SE, NO; Cons.: AT, BE, CH, DE, FR; Lib.: GB; South.: ES, PT; East.: BG, CZ, EE, HU, LT, LV, PL; Buffet-typology 2: Soc-Dem: DK, FI, SE, NO, AT; Cons: BE, DE, FR; Lib: GB; CH; South.: ES, PT; East.: BG, CZ, EE, HU, LT, LV, PL, SI, SK; BGI* = *Benefit Generosity Index.*




*ICC in Model 0 (Random Intercept Only): 0.10;*

*Buffet-typology 1: Soc-Dem.: DK, FI, SE, NO; Cons.: AT, BE, CH, DE, FR; Lib.: GB; South.: ES, PT; East.: BG, CZ, EE, HU, LT, LV, PL; Buffet-typology 2: Soc-Dem: DK, FI, SE, NO, AT; Cons: BE, DE, FR; Lib: GB; CH; South.: ES, PT; East.: BG, CZ, EE, HU, LT, LV, PL, SI, SK; BGI* = *Benefit Generosity Index.*

from countries with higher social expenditure therefore want less government responsibility for ensuring a decent standard of living for the unemployed.

The two different *unemployment replacement rates* (models 3 and 4) lead to slightly different results. In the ESS analysis, only the version provided by the SIED data produces a significant and positive effect, while the CWED2 version is insignificant. In the ISSP analysis, neither of the rates exhibits significant effects. Still, this shows that varying data sources should at least be discussed—especially if results are compared with studies using indicators from a different data source. In this analysis, generous benefits in case of unemployment tend to lower support for government responsibility in the field but this effect does not appear to be very robust. Apart from this, the opposed directions of the effects compared to the spending indicators correspond to the expectation that welfare effort and welfare generosity represent different parts of the welfare state (cf. section 2.1)**.**

The two *buffet-typologies* (models 5 and 6) consistently show that people living in Liberal welfare states, which are assumed the least generous, are significantly less in favour of government responsibility than those in inclusive Social-Democratic welfare states. Furthermore, the first typology (model 5) also reveals a significantly lower preference for state responsibility in countries belonging to the Conservative type. This effect disappears in the second buffettypology (model 6) with the different classification of Austria and Switzerland, and it indicates that a potential bias due to slightly differing combinations and extensions of existing typologies should be taken seriously.

In interpreting these results, the two typologies seem to point in the direction of the policy feedback hypothesis: living in a Social-Democratic welfare state seems to increase support for government action – at least compared to Liberal regimes. On the other hand, the insignificant effect of the Generosity Index (model 7) undermines this finding. Since this index is based on many of the indicators Esping-Andersen used to construct his initial typology, it should at least roughly indicate patterns that correspond to the TWWC typology or one of the succeeding classifications. However, this is hardly the case (Figure 3.1). Instead, a ranking of generosity scores does not reveal clusters of countries that fit the typologies I used in the analyses, the TWWC, or in fact any other typology.

In addition to these findings, further analysis (Table 3.3) shows that when the same two buffet typologies are tested on a slightly larger sample of countries,12 the results turn out quite different. Suddenly, Liberal countries no longer differ significantly from Social-Democratic ones. Instead, Conservative welfare states now consistently show significantly less support for government action than the

<sup>12</sup> Based on ESS data: the Netherlands, Ireland, Cyprus, Greece and Romania are added.

**Figure 3.1** Unemployment generosity index by country. *(Colours indicate membership in regime according to buffet typology 1, data: CWED2, colouring of pillars indicates membership in regimes according to buffet-typology 1; stripes (horizontal): Social-Democratic, white: Conservative, stripes (diagonal): Southern, grey: Eastern, black: Liberal)*

latter, while respondents from countries belonging to the Southern type show significantly more support for state responsibility—however this effect is only found for the second typology.

This finding is quite problematic because although it may seem obvious that different country samples may produce different results, samples in secondary analyses of survey data like the ESS will always vary from wave to wave. Thus, even if scholars use the same typologies, the differing samples will still hinder the comparability of results with previous research. Of course, the same argument holds true for every kind of indicator and analysis. Still, typologies exhibit a sense of homogeneity among the members of a category, which may tempt to underestimate the problem. This problem of the country sample in connection with the large number of different classifications13 makes the application of welfare typologies very volatile.

Overall, the additional analysis again indicates that people in Conservative and Liberal welfare states prefer less government responsibility than in Social-Democratic regimes. However, as soon as a Southern and Eastern type

<sup>13</sup> For instance, the Dutch case is commonly acknowledged as being a hybrid welfare state as well (Arts & Gelissen 2002; Pennings 2005). Reclassifying the Netherlands as Liberal or Social Democratic would surely lead to different results again.


**Table 3.3** Comparison of regime typologies: Government responsibility for providing standard of living for unemployed (ESS 2008)

*Data: ESS (Round 4), multilevel lineal regressions (xtmixed), standard errors in parentheses,* + *p* < *0.1, \* p* < *0.05, \*\* p* < *0.01, \*\*\* p* < *0.001; ICC in Model 0 (Random Intercept Only): 0.11; unemployment rate controlled for in all models.*

are included, they relegate the Social-Democratic countries to an intermediary position. These results advice caution. The unexpected intermediate position of Scandinavian countries could be explained by the fact that citizens affected by crises or transitions in the South and East may wish for more social security. To postulate however that this is actually due to similar properties of the welfare state and not just to a geographical or developmental vicinity could be premature. If the welfare state is at work, it works in different ways: welfare generosity may inspire more confidence in Social-Democratic regimes than in Liberal ones, but if wanting more governmentally regulated welfare provision is due to less generous or malfunctioning welfare states in Southern and Eastern states, those countries follow a different logic.

Summarising all results, the negative effect of social expenditure (overall and in the field of unemployment) on attitudes opposes the policy feedback hypothesis at first glance while net replacement rates and typologies show a tendency to support it. However, the indicators produce very unstable results and small modifications influence the significance of effects severely. Moreover, the explanatory contribution of the different approaches varies considerably—reaching from close to zero (benefit generosity index and NRR) to moderate (social expenditure) and even very high values (regime typologies). Considering that the different operationalisations should at least be somewhat related, this is problematic.

Based on the findings in this analysis, it would be very difficult to answer *why* attitudes differ. Regardless—and in line with the aim of this project—the analysis reveals important sources of bias, such as sample selection and data source. Discussing these issues and finding ways to avoid them can help standardise the process.

#### **3.3 Summary: an Independent Variable Problem**

In this chapter, sources of dissent within each approach were identified and each of these issues was visible and consequential in the empirical test that followed.

Within the *single indicator* approach, limited informative value and differing data sources are critical issues. Although it may seem trivial to say that replacement rates and social expenditure address distinct and very different aspects of the welfare state, both are used as independent variables in analyses of welfare attitudes. The literature does not offer any guidelines that recommend a standardised selection of suitable indicators and sensible combinations as well as data sources. The latter leads over to the second issue. The analyses reveal variations in the results and their significance depending on the data source. This indicates a potential bias, which should be examined in more detail and—at the very least—should encourage taking sources into account when comparing results.

The *regime approach* is characterised by differences in the underlying conceptual and operational premises. As this chapter shows, different classifications can affect the results—and there are many other classifications in the literature that have not been tested in this contribution that could lead to even more different results. Furthermore, the differing country samples in survey data prove to be highly problematic. More research is needed to test how much combination and extension a typology can endure before the results are no longer comparable.

Lastly, it is very difficult to assess the *composite index approach*. Since comprehensive examples of this approach are only available for a small number of countries, they need to be extended to bigger country samples. However, the inclusion of other countries—especially CEE countries—proves to be quite difficult. There are many issues that are critical when attempting to include CEE states in existing measurements. For instance, de jure and de facto benefit generosity in these countries may not entirely match, labour market participation may differ systematically from older welfare states, atypical employment may be more common, and much more. A comprehensive discussion is given by Kuitto (2018) who extends Esping-Andersen's version of the decommodification index to CEE countries and raises these and more important issues.

Given the problems identified, several practical recommendations can be made at this point. First, different operationalisations should never be treated as interchangeable options – neither within nor between approaches. They have different conceptual premises and allow different interpretations. The selection of an indicator should, whenever possible, be based on the greatest possible comparability and should not be based solely on a lack of alternatives. Second, data sources should receive more attention. This directly applies to single indicators and indirectly to typologies and composite measures, because they are based on such single indicators. Third, combining and extending typologies should be avoided or follow clear theoretical justifications, as arbitrarily picking and blending classifications from the literature can severely affect the comparability of results. Fourth, the frequent exclusion of Central and Eastern European countries is dated and obstructive to comparative analyses of social phenomena in Europe and beyond. If the existing indicators do not fit the character of the welfare state in these nations, more attention should be paid to finding proxies that work equally for old and new welfare states. Finally, and this may be the most controversial finding, the disadvantages of using welfare regime typologies as independent variables far outweigh the advantages. Based on the conceptual and the empirical assessment in this chapter, I can only advise against using such typologies as independent variables until the problems discussed are solved.

Despite the many issues discussed, the differences between welfare states reflect very important features of modern democracies. They fulfil important functions (cf. section 2.2) and strongly influence individuals and their living conditions. The lack of a reliable, easily available and applicable instrument should lead neither to making unsatisfactory compromises nor to excluding the welfare state from the analysis. Thus, it is important to explore what kind of instrument is actually needed by scholars looking for an independent variable. Based on the previous discussion, I recommend two objectives that I believe can serve as starting points for a fruitful discussion. First, the requirements for a measurement that is to serve as an explanatory instrument must be examined in detail. Second, there is a need for a detailed theoretical and conceptual discussion of the mechanisms assumed when exploring the outcome of different welfare state arrangements.

The problems identified in this contribution already help to substantiate the first objective because they show obstacles that can be circumvented. Following the preceding discussion, the main problematic issues are lack of *clarity*, *availability*, and *comparability*. Those three aspects can be translated into criteria that can contribute to the development of a more standardised approach: it should be clear what information an indicator is based on, the indicator should be available for a sufficiently large sample to allow replications, and it must be comparable with other studies.

Strictly speaking, neither existing typologies nor composite measures fit these premises. In both cases, the lack of *availability* for a big enough population is rather obvious. Moreover, they also lack *clarity* because their operationalisation aims at capturing the multidimensionality of welfare states and are thus based on a variety of indicators. In the case of composite measures, this combination may level out and thereby mask important outliers (Kvist 2011), while the broad categories of typologies may represent much more than just welfare state policies (like political cultures, economic and democratic development et cetera). As a result, neither of the two operationalisations allow determining, which specific part of the operationalisation is at work if an effect is observed. This leaves single indicators as perhaps the most fruitful way to operationalise welfare policies as independent variables. Still, while *availability* is much better in this case, *clarity* and *comparability* are not a given. Social expenditure, for example, is anything but a precise indicator. As argued in section 2.1, high social spending can represent very different things. Furthermore, data sources have to be addressed. Still, single indicators bear one great advantage: by highlighting one specific aspect of welfare stateness, their interpretation is most unequivocal. Perhaps, the best way to include the welfare state as an independent variable is such focus on single issues instead of broad and multidimensional conceptualisations of this very complex institution.

Regarding the second objective, I suggest a closer look at potential *dependent* variables in order to get a clearer picture of the hypothesised mechanisms. It is not enough to assume that 'the welfare state' influences an outcome. A key question is *why* this should be the case and *how* the mechanism may work. The answer to both questions does not come from the independent variable, but from the dependent one. This suggests that different dependent variables may require different operationalisations of welfare stateness. Returning to the example of welfare attitudes helps to illustrate this argument. Here, the hypothesis highlighted attitude formation as a result of policy feedback. The underlying mechanism implies a process of evaluation. To test this assumed affect, we would thus require social policy indicators, which contribute to opinion-formation because individuals are likely aware of them. Indicators like waiting days and contribution periods, which are integral parts of composite measures and many typologies, do not fall under that category because only a small part of the population will know these details. However, respondents have at least a basic understanding of the generosity of benefits (e.g. replacement rates), potentially making this a much better indicator.

However, if another exemplary topic is chosen, the argumentation can be very different. When it comes to explaining the risk of poverty, for example, the individual perception and evaluation of social policy plays no role. Rather, the organisational principle and the effectiveness appear to be more important – regardless of whether or not the majority of individuals are actually aware of them (e.g. waiting periods or benefit duration). The aim should therefore be to collect such mechanisms, to systematise them and to offer suitable indicators for their testing that meet comprehensible criteria.

While this chapter has painted a very critical picture of approaches to operationalise the welfare state, the conclusions refer only to a very specific problem—the operationalisation as an *independent* variable. If the way in which the welfare state is included in cross-national analyses is so inconsistent in this case, how can we achieve more transparency in the future? I have already hinted at a possible approach: it might be worth asking whether the impact welfare states are hypothesised to have is adequately captured in an operationalisation. If this is not the case, how can the actually relevant characteristics of welfare stateness be identified? In the following, I am going to explore this issue in more detail. Based on the findings of this chapter, a catch-all approach to operationalising the welfare state as an independent variable is discarded. Instead, the focus rests on conceptualisations of welfare stateness that are embedded in the theoretical arguments that warrant its inclusion as independent variable in the first place. This means that the main objective going forward is to pinpoint *why* the welfare state is assumed to shape individual-level outcomes and which conceptual perspectives on the welfare state exist. Focussing on the dependent variables and hypothesised macro–micro mechanisms should be a good point of departure for determining what kinds of measures are actually needed by scholars who want to treat features of the welfare state as an independent variables in multilevel analyses.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **4 Literature Review: Mechanisms and Hypotheses**

The previous chapter highlighted a number of problems that result from a lack of standardisation when selecting operationalisations. There are two extremes when it comes to how we can possibly solve these problems. One would be to further improve the overall measurement of the welfare state in order to arrive at one single universally agreed-upon operationalisation. It seems unlikely that such an agreement can be reached in the near future and we might actually also question whether something like that could (or should) be realised at all (e.g. Spicker 2018). In fact, it seems doubtful that a catch-all approach is actually differentiated enough to capture all of the different relevant facets of welfare stateness. The opposed extreme would be to give up the search for a universal instrument and instead try to find fitting indicators for singled-out issues. Finally, a middle ground might be to propose clusters of conceptually related indicators that represent tailored independent variables for specific research questions, hypotheses, and dependent variables. This last solution appears to be the most promising, as it offers enough leeway for context-specific modelling without compromising comparability.

In the following part, I am going to develop these thoughts further. In order to explore which indicators are most suitable for specific research agendas, the causal relationships linking welfare state policies to individual-level outcomes are discussed in detail. For this purpose, theoretical premises, mechanisms and hypotheses are summarised and systematised. In order to capture the assumed causalities, I am proposing to approach the matter from two perspectives: on the one hand, I am taking up the *functions of welfare states* discussed previously (cf. section 2.2). On the other hand, I am systematising the individual-level outcomes that serve as *popular dependent variables* in the relevant literature. By linking specific functions of the welfare state to specific outcomes, I aim at pinpointing relevant mechanisms in more detail. This is intended to result in a proposal, which

K. Kunißen, *The Independent Variable Problem*, Sozialstrukturanalyse, https://doi.org/10.1007/978-3-658-39422-6\_4

describes how to derive theoretically grounded indicators by shifting the way we look at the welfare state when treating it as an independent variable. Instead of attempting to best capture social policy arrangements as a whole, the focus should be on modelling the key assumptions about the relationships between policies and micro-level outcomes. Such proceeding—and this is important to emphasise—relates foremost to the type of cross-cultural comparative multilevel research addressed in this book.

### **4.1 The Mechanisms Matter: a Short Introduction**

In the exemplary research objective chosen for illustration in the previous chapter, very basic theoretical assumptions about how welfare states influence individual attitudes towards social policies were briefly introduced. In this case, it has been hypothesised that generous and universal social policies following social-democratic principles generate political support and positive attitudes towards redistribution (Jaime-Castillo 2013; Roosma et al. 2014) while redistribution-based and targeted polices increase conflicts between beneficiaries and contributors, leading to disapproval of welfare policies (Jordan 2013). Since the aim of this chapter is to identify such hypotheses for a variety of popular research questions, a more comprehensive review of literature on welfare attitudes and many other research items is needed. Since the explicit measurement of cause and effect is the focus of this book, some general preliminary remarks on the nature and relevance of mechanism-based reasoning follow.

Research addressed in this book, usually fits into a framework shaped by theories of the middle range (Merton 1968). As the term suggests, such theories involve a perspective between very general, comprehensive approaches on the one hand and detailed, very specific ones on the other. As Merton puts it:

Middle-range theory involves abstractions, of course, but they are close enough to observed data to be incorporated in propositions that permit empirical testing" (Merton 1968: 39).

In this line of research, the identification of social mechanisms emerged as an important analytical task (Merton 1968: 45). Mechanisms have been defined in a variety of ways, which share a few important aspects. One of these shared features is that they reveal causal processes that show how and why a given phenomenon is observed (for an overview of definitions see Hedström & Ylikoski 2010: 51). Essentially:

[…] it is not enough to state that X leads to Y: a satisfactory explanation demands explication of the sequences and steps through which X and Y are causally linked, i.e. why and how X leads to Y (Tranow et al. 2016: 5).

In recent years, exploring mechanisms has become increasingly popular in the social sciences and the term is widely used—especially in research attributed to the field of analytical sociology (for an earlier overview see Manzo 2010). While it important to exame in more detail how and why "nuts and bolts, cogs and wheels" (Elster 1989: 3) exist and help explain social processes and phenomena, the term is sometimes used prematurely. As Kalter and Kroneberg (2014: 100– 103) point out, the popularity of the term led to it being used even in cases where it is not appropriate. This includes (1) tautological ('mechanism' and explanandum are equal), (2) speculative (ad hoc explanations without empirical proof), (3) elliptical ('mechanisms' refer only to concepts) and (4) partial (incomplete) assumptions or evidences. Out of these four, the first two are especially problematic. Overall, the adequacy of a mechanism is determined by its ability to lead to specific and informative hypotheses (Kalter & Kroneberg 2014: 102).

A popular approach to conceptualising mechanisms is the macro-micro-macro scheme of sociological explanation. Since this approach is well compatible with the multilevel logic pursued in this book (cf. section 1.2), this analytical framework is chosen as conceptual representation of mechanisms (and other types of relationships) that are relevant for the research questions at hand. A very basic illustration of different kinds of mechanisms within the macro-micro-macro scheme is provided in Figure 4.1. Three types of mechanisms are usually explored when applying this analytical framework to research addressing how and why an outcome on the societal level (S2) results from an initial situation (S1). Following Hedström and Ylikoski (2010: 59), the link between macro- and micro-level has to be taken into account in the first step. Such *situational mechanisms* (1) address how and why societal contexts shape individual actors by affecting their position in the social structure, their desires and beliefs et cetera. The second type of mechanism grasps processes leading individuals from an initial situation (A1) to a specific outcome (A2). It is labelled *formative mechanism* (2) in the figure. This deviates slightly from popular versions of such models in the literature, which usually work in an action-based framework. Since, however, the relevant hypotheses regarding the impact of welfare states on individual-level outcomes refer to a variety of phenomena, going well beyond action-based approaches, this more general label is chosen to incorporate other processes and outcomes, such as attitude formation. Lastly, *transformational mechanisms* (3) refer to how and why individuals generate intended or unintended outcomes on the societal level.<sup>1</sup> As mentioned earlier, this last step is not typically part of the research covered in this book, as the outcome to be explained is usually at the micro- rather than the macro-level.

**Figure 4.1** Types of mechanisms. *(Figure based on Hedström & Ylikoski (2010: 59) and Tranow et al. (2016: 8))*

Although research exploring how and why the characteristics of welfare states affect individuals seems to fit well within the analytical framework, detailed references to social mechanisms, causal processes, and the key issues involved are indeed rare in the relevant literature. This problem is sometimes acknowledged. For instance, O'Campo and colleagues (2015: 89) refer to it in the case of the impact of welfare policies on health and poverty.<sup>2</sup> In addition, many hypotheses in relevant contributions do not actually contain mechanisms or even causal relationships at all—even if some still refer to those terms (often in the problematic sense observed by Kalter and Kroneberg (2014) that was mentioned before). Instead, studies frequently focus on mere effects and statistical relationships, without much specification of causal links, thus often lacking clear-cut theoretical

<sup>1</sup> These three steps correspond to what Coleman (1986) calls type 1, 2 and 3 relations (cf. section 1.2).

<sup>2</sup> Since their contribution is a rare example of systematic attempts to pinpoint mechanisms, it will be taken up again in more detail later in this chapter.

elaborations on how and why a correlation is expected. In a more recent publication Spicker (2018) takes a very critical perspective on the treatment of causality in the specific example of assumed effects of welfare policies on individuals:

[…] how can we say which is the cause, and which is effect? Wherever there is a complex, multi-faceted set of phenomena, it is almost impossible to distinguish generative mechanisms or to demonstrate genuinely effective processes (Spicker 2018: 225).

Even though there is some ground to this argument, I disagree with his conclusion that this 'evidential' issue—combined with methodological concerns about the reduction of complexity in statistical approaches and theoretical concerns about the equivalence of countries—should lead to questioning the worth of quantitative approaches in general (Spicker 2018: 226).3 If we want to understand at large why individuals differ not only within a country but also between countries, turning away from statistical approaches means abandoning entire research questions. A middle ground between giving up on mechanisms and giving up on statistical models, could be to put more effort into the theoretical specification of the assumed effects and the selection of measures. In addition, it is helpful to know the limits of the actual modeling of possible mechanisms. This includes acknowledging not all complex mechanisms *can* be modelled empirically. This can result from a lack of data on the policy-level or individual-level as well as from statistical limitations.

In contrast, some hypotheses simply do not entail complex mechanisms, as they describe in fact almost tautological relationships. This can be because they highlight one particular path in a potentially more complex model, or the outcome is a necessary consequence of the condition—sometimes even trivially so (e.g. income replacement decreases poverty). Still, regardless of whether we are dealing with complex mechanisms or singled-out relationships between welfare state and individuals: it it important to distinguish between the various epistemological interests motivating hypotheses and to ask how and why differences in social policy-making impact individuals differently.

To summarise the main arguments in this brief excursion into the field of mechanism-based explanations: its main advantage for the task at hand is its focus on breaking down assumed relationships between social contexts and individuals and exploring in more depth how and why such links exist. In the case of the specific problem at hand this means asking *why* an outcome is influenced by the welfare state—in the assumed way and *how* the underlying causal process works.

<sup>3</sup> Many of Spicker's arguments resemble my own criticism of different operational approaches in multilevel analyses (cf. chapter 3), even though he emphasises other aspects.

If we can answer these questions—or at least approach possible answers –, it will considerably help to select appropriate and meaningful operationalisations of welfare stateness for the empirical test of theoretical models.

For this purpose, the following section summarises and systematises popular hypotheses. The term 'mechanism' will be used to refer to the explanation of a causal relationship between welfare stateness and specific outcomes. In some cases, it can be discussed very critically if using the term overstates the complexity of a given explanation in the sense described before. For the sake of consistency, I will use the term regardless, albeit cautiously. By linking the hypotheses to the functions of welfare states discussed before,4 I hope to further concretise the key causal assumptions—especially in cases where the relationship between welfare state and micro-level outcomes lacks strong theoretical foundation.

#### **4.2 Summary of Popular Hypotheses**

While it is easier to research relevant literature on a specific dependent variable, it is more complicated to obtain a comprehensive review of literature on a specific *independent* variable. There are various social phenomena, which are influenced by social policies and therefore receive much attention in the literature. Among the most popular dependent variables are attitudes towards the welfare state (e.g. Roosma et al. 2014; Steele 2015), health (e.g. Bambra et al. 2014; Foubert et al. 2014; Deeming & Jones 2015), poverty (e.g. van Lancker & van Mechelen 2015; Brady et al. 2017) and work-family balance (e.g. Beham et al. 2014; Lunau et al. 2014). However, it is important not to overlook other notable research that suggests a significant impact of the welfare state. For this purpose, the pool of examined literature is enhanced by a review of relevant journal publications in a five-year-period (mid-2013 to mid-2018) listed in the Social Science Citation Index (SSCI). Contributions are included in this review if they meet four criteria. First, the welfare state must be treated not just as a control variable, but as an important explanatory variables. For this purpose a reference to the welfare state, public policies or social policies has to be included in the abstract, title or keywords. Second, the examined dependent variable must be on the microlevel. Third, a cross-cultural comparison must be carried out. Thus, comparisons in general or multilevel analyses in particular have to be mentioned. Lastly, since

<sup>4</sup> The highlighted functions were: 1) provision of security 2) redistribution, 3) stratification,

<sup>4)</sup> socialisation, and 5) activation.

this project focusses on western democracies and especially the European countries, a reference to this geographical region is required.5 These search parameters resulted in just above 530 contributions of which about 15 percent perfectly fit these criteria. This is by far not an exhaustive stocktaking of all relevant publications. Yet, since the aim is to identify popular topics in the body of literature, it is not necessary to include every study into the sample. The configuration of the search parameters should provide a sufficient picture of contemporary research in which the welfare state serves as an important independent variable, and thus key causal assumptions should be represented in the selection.

Even though the identified publications treat the welfare state as a central explanatory variable in their hypotheses and empirical models, many studies lack a comprehensive discussion of why and how features of welfare states are at work. Thus, the explicit mention, conceptualisation and operationalisation of social mechanisms is rare. This somewhat sobering finding makes it difficult to identify causalities beyond loose assumptions or statistical correlations. However, the mere fact that mechanisms are often not explicitly mentioned and modelled does not necessarily mean that the hypotheses posed are not based on causal explanations. Furthermore, while the majority of relevant studies relies only on very short theoretical explanations, there are also contributions, which set an example for well-grounded and differentiated conceptualisations. These studies will be referenced more extensively in the course of this chapter as they provide much-needed material to bridge some existing gaps in the argumentations.

The following sections provide a detailed overview of key objects of research and the hypotheses tying them to social policies. The majority of dependent variables can be subsumed under at least one out of four broader categories: (1) well-being, (2) risks and needs, (3) attitudes and (4) behaviours. Even though there are overlaps between them, hypotheses within each of these topics exhibit similarities. These similarities are discussed and systematised in the following four sections of this chapter, which all follow the same structure: first, main debates are summarised. Second, one especially characteristic topic is selected for an in-depth discussion within each of the four broader topics. This is followed by exploring how and why the welfare state is assumed to affect the specific outcome highlighted in this example. Lastly, hypotheses are summarised in an attempt to deduce general conceptualisations and patterns from the specific examples. While conceptual issues receive much attention in the following

<sup>5</sup> Search string (Web of Science): (TS <sup>=</sup> (("welfare state\*" OR "public polic\*" OR "social polic\*") AND ("multilevel" OR "compar\*") AND ("EU" OR "Europe\*"))) OR (TI = (("welfare state\*" OR "public polic\*" OR "social polic\*") AND ("multilevel" OR "compar\*") AND ("EU" OR "Europe\*"))).

review, actual empirical findings are not going to be discussed in much detail. The reason is that this contribution is interested in the theoretical and conceptual arguments. Since the results are assumed to be biased, or at least not fully comparable, including them in the discussion could distort this view. Thus, I am interested in the main hypotheses and explanations, not in the question if they are confirmed (although—as they are the most popular ones—they usually are).

Before turning to the dependent variables, one final comment is required. Comprehensive bodies of work deal with each of the topics covered in the next part of this chapter. There are many complex contributions highlighting important theoretical and methodological issues in each of the four examples. My summaries of how and why welfare states affect each dependent variable cannot and do not do justice to each one of them. Instead, this chapter aims to summarise very general assumptions on a general level—in the sense of a minimal consensus. It will not permeate all facets of the subjects but substantiate the connection between them and the welfare state for the purpose of this contribution.

#### **4.2.1 The Impact of the Welfare State on Well-Being**

Well-being is an outcome of social policies, which is closely related to risks. However, there are several reasons why literature on the impact of welfare states on well-being deserve a separate inspection. One reason for this is the wide variety of manifestations of well-being, ranging from health outcomes to life satisfaction and other forms of subjective well-being, beyond the more 'classical' risks associated with work and social security. Another reason is that literature dealing with the most apparent expression of well-being—individual health—offers some elaborate debates about the theoretical conceptualisation and empirical operationalisation of welfare stateness as a determinant of health outcomes. As I will show in the course of this chapter, such debates can be applied to the relationship between welfare state and other outcomes as well.

When examining well-being, several different perspectives have to be taken into account. Two areas are especially prominent: subjective well-being (SWB) and health. While subjective well-being is based on individual perception, health outcomes are usually related to objective risks. Even though a broader focus on well-being covers a variety of topics and incorporates various conceptualisations of well-being (ranging from economic to philosophical aspects),6 the different

<sup>6</sup> Research on well-being incorporates a variety of perspectives on well-being, which go beyond the ones that are important for the purpose of this book as their relation to the welfare

dependent variables share many similarities when it comes to the assumed causalities tying the welfare state to a manifestation of well-being. In the following section, I focus on health outcomes. This dependent variable is selected for various reasons. First, health outcomes are—by their very nature—closely linked to the welfare state (particularly the healthcare system) and therefore receive much attention in the literature (for a review see among others Muntaner et al. 2011; Bergqvist et al. 2013). Second, literature on the relationship between welfare state arrangements and health outcomes is quite sensitised to the issue at hand—especially compared to other fields of research. Thus, conceptualisation and operationalisation of social policies as an independent variable and problems arising from inconsistent approaches are explicitly discussed in this line of research (Brennenstuhl et al. 2012; Bergqvist et al. 2013). While the problems of unstandardised proceedings were discussed from a more general perspective in the previous chapter of this book (cf. chapter 3), the different perspectives on social policies pinpointed in the critical literature on including social policies as determinants of health are a helpful addition. This includes theoretical approaches conceptualising mechanisms linking the welfare state to health outcomes (such as Beckfield et al. 2015), as well as methodological reviews (such as Bergqvist et al. 2013) and studies selecting their indicators as explicit representatives of distinct perspectives on the welfare state (e.g. Ferrarini et al. 2014a).

Empirically, health outcomes can be explored based on objective measures of health as well as self-rated ones. Since comparative research on health usually relies on survey data, self-perceived health is much easier to measure than objective health.<sup>7</sup> Therefore, the following discussion mainly relates to this kind of health measurement. Furthermore, it is important to distinguish between two different perspectives on health in the literature. One focusses on health outcomes in general, the other on health inequalities. An increase in overall health in a population does not necessarily go hand in hand with a decline in socioeconomic health inequalities (e.g. Bambra 2013; Bergqvist et al. 2013). I will take up both perspectives when summarising relevant hypotheses later in this chapter as they

state is less pronounced. Such perspectives include more market-oriented and philosophical approaches (for a more comprehensive overview see Gasper 2005, 2010).

<sup>7</sup> Yet, the measurements produce comparable results (as pointed out by Wu et al. 2013). While subjective health usually measures how healthy respondents perceive themselves to be, objective health can be surveyed by asking about the occurrence of specific manifestations of health. This includes long-standing or chronic illness (e.g. Maskileyson 2014), dental health (e.g. Guarnizo-Herreño et al. 2014), or determining the occurrence of psychological illnesses through diagnostic questions (e.g. Hansen et al. 2017). Lastly, mortality is a related dependent variable as well (e.g. Mackenbach et al. 2017).

are almost inseparable. The social gradient of health outcomes on the micro-level will be especially important when differentiating between direct effects of social policies on health and those that moderate the impact of social inequality (or are mediated by social inequality).8

I proceed in three steps to identify different conceptualisations of welfare stateness. First, I determine possible relationships between welfare state arrangements and health outcomes and the mechanisms responsible for such relationships. Second, I summarise conceptual views on the welfare state that emphasise different features. Third, I explore specific hypotheses in the literature to determine which mechanisms and perspectives are assumed in explanations. Overall, this process is used to identify shared aspects, which help to distil distinct perspectives on the welfare state.

Turning to the first step, the welfare state is linked to health outcomes in several ways. The relationship can take the form of a *moderation* of the severity of other determinants of poor health as well as *indirect* and *direct* effects. When looking at the first two, we have to take into account determinants of health on the micro-level, which are part of moderated or mediated effects. Key determinants of health outcomes on the micro-level are manifestations of social inequality: higher education, social status and income increase health (Beckfield et al. 2015: 228), while especially unemployment and poverty reduce health considerably (Lundberg et al. 2010; O'Campo et al. 2015). The welfare state is intimately linked to such inequalities. Therefore, it is clear that *moderating* effects of social policies on health outcomes exist because they shape the risk of ill health associated with individual vulnerability in different ways. The relationship between social inequality and health is sometimes also described as being *mediated* by social policies (Beckfield et al. 2015: 228; O'Campo et al. 2015: 88). It is my understanding that the described effects refer rather to a moderating function of the welfare state, instead of a mediating one.9 Instead of the welfare state, social

<sup>8</sup> Very particular perspectives such as gender inequalities in health (Palència et al. 2014; Borrell et al. 2014) are noteworthy contributions but will not receive particular attention in this chapter.

<sup>9</sup> The authors state that welfare regimes "[…] have mediated the impact of social determinants of health and also of socioeconomic class on health to varying degrees" (Beckfield et al. 2015: 229). In my opinion, such an impact of welfare states is not necessarily mediating. Since the mediator is logically situated after the occurrence of another variable, it seems to make more sense to conceptualise the welfare state as a moderator of the impact of social determinants on health. However, if an analyses is carried out solely on the macro-level, we can expect that welfare states mediate the effect of social inequality on health inequality at the aggregate-level, as there might actually be an effect of aggregate social inequality on the

determinants may be mediators, as they are shaped by social policies (*stratification function*) and in turn impact health outcomes. In addition to moderating and mediating effects, there can also be a *direct* impact of social policies in general and health policies in particular on individual health outcomes. However, such effects are not as prominent in the literature. Still, features such as access to medical services (e.g. Kim et al. 2017) can account for cross-national differences in health outcomes.

While links between features of welfare states and health outcomes appear in empirical correlations, they still do not reveal how and why a relationship exists. For the purpose of identifying such possible mechanisms behind social policies on health outcomes, a contribution by Beckfield and colleagues (2015) is noteworthy. The authors propose an institutional theory of welfare state effects on health because—as they point out—theories of health inequalities have focussed mainly on micro-level explanations of health outcomes,10 while an institutional perspective is missing. Overall, they highlight four "mechanisms" (Beckfield et al. 2015: 232), which have to be distinguished when examining the impact of welfare states on individual health and health inequalities: (1) redistribution, (2) compression, (3) mediation and (4) imbrication.11

*Redistribution* refers to different modes of shifting resources to those in need. While the authors only discuss this in relation to the redistribution of resources between differently situated groups, this could also refer to the redistribution of resources over the life-course. This mechanism clearly corresponds to the *redistributive function* of welfare states described in section 2.2.

The mechanism of *compression* describes upper and lower bounds in the social gradient of health set by institutional efforts (such as a minimum wage). In addition, welfare states regulate the minimum for the provision of healthcare (or other) benefits, which is manifested in health-care access, benefit levels, and generosity, among other things. This process appears to be closely related to the responsiveness of a welfare state. In this sense, it seems to be largely linked to what was previously termed the *security function* of minimising risks and meeting needs.

nature of social policies (for a comprehensive discussion of moderating and mediating effects (cf. Hayes 2018).

<sup>10</sup> As micro-level explanations, they highlight materialist explanations (income influences the access to goods and services, which increase health), cultural-behavioural explanations (health behaviour varies in different milieus) and psychosocial explanations (inequality and exclusion lead to stress) (Beckfield et al. 2015: 230).

<sup>11</sup> The authors explicitly refer to these processes as 'mechanisms'. Whether all of them are actual mechanisms in the sense described in the beginning of this chapter, will not be judged here. Since the authors use the term and for the sake of consistency, I will retain it.

As Högberg and colleagues (2018: 314) add, benefit levels can even be beneficial if they are not used. They represent potential resources that are understood in the subjective perception as a protection against possible risks in the future, which reduces uncertainty. However, this should relate more strongly to subjective (psychosocial) well-being (Sjöberg 2010) than health outcomes. Still, as discussed in the following section, psychosocial stress is a predictor of morbidity and reducing such stress may therefore be a relevant moderating effect.

*Mediation* means a process in which institutional aspects shape a third variable, which in turn has an impact on health outcomes. More specifically, Beckfield and colleagues (2015) refer to the welfare state as a mediator in an indirect effect of socio-economic indicators on health. As noted before, it seems only partly convincing to refer to the welfare state as a mediator from a macro-micro perspective since it seems implausible that social policies are directly shaped by individual-level social inequality. It can of course be a *moderator*, shaping the impact of another variable (such as socio-economic status or unemployment) on the outcome. Such a moderating effect is often associated with the reduction of psychosocial stress as a morbidity-increasing factor (O'Campo et al. 2015: 89). Lowering stress by reducing uncertainty and compensating for (e.g. financial) deficits again corresponds to the *security function*. While such moderating effects of welfare states are more prominent in the literature, mediating effects can exist. Indicators of social inequality on the individual level might in fact be mediators in the sense that the impact of social policies on health is an indirect effect which is mediated by vulnerability of individuals. The link between welfare state and individual social status would thus represent the function of shaping patterns of *social stratification* and inequality, while the moderating role of social policies is a manifestation of the *security function*. Concluding, I suggest to add moderation to the proposal of Beckfield and colleagues, and in the following, I will discuss the welfare state as a moderator on the one hand and its impact on the outcome as a possible indirect effect, which is mediated by other features (especially related to social inequality) on the other hand.

Finally, the concept of *imbrication* refers to the simultaneous and overlapping effects of different institutions. While it seems evident that policies in the field of health care have an impact on health outcomes, other areas within the welfare state (and beyond) can be influential as well. In this sense, institutions can amplify each other's impact; they can moderate it or undermine it. Taking this argument further, it seems prudent to consider other areas outside of the healthcare system as well, simply because they may be more relevant. To take examples from the literature: if social inequality is conceptualised as a mediator between welfare state and health outcomes, the health system may have a different and perhaps even less influential effect than—for instance—pension policies (Högberg et al. 2018), unemployment insurance (Renahy et al. 2018) or active labour market policies (Voßemer et al. 2018). Moreover, it may be worth discussing, whether imbrication does not only addresses overlapping policy areas, but also overlapping policy manifestations. An example of this could be empirical evidence finding an interplay of indicators such as insurance coverage and benefit generosity, suggesting that high replacement rates indicate well-functioning security only when combined with high insurance coverage (Ferrarini et al. 2014b). Imbrication thus raises many conceptual questions. As a mechanism, it seems most closely tied to the *security function*, as it addresses how closely knit a securing net is.

**Figure 4.2** The impact of the welfare state on health outcomes

Figure 4.2 summarises possible impacts of the welfare state on health outcome and the interaction with social determinants of health. The first part of the graph summarises all possible effects (A). Here, a transformational process (final micromacro link) is included. Since I focus on micro-level manifestations of health, it will not receive further attention. The second figure (B) illustrates direct effects of the welfare state on health outcomes. Such effects include what was described as welfare state's primary function of providing security (cf. section 2.2.1). C and D in Figure 4.2 illustrate the moderating effect of social policies and the indirect effect in which social determinants serve as mediators.

Summarising potential mechanisms based on the proposal by Beckfield and colleagues as well as the functions of welfare states discussed in the second chapter of this book, several mechanisms can be at play when explaining the impact of welfare states on health outcomes and health inequalities. Health outcomes are shaped directly by the welfare state in its function of *securing* individuals against the risk of bad health by *redistributing* resources (e.g. in order to improve access to services of insurance coverage). In contrast, health inequality is shaped mainly in a non-direct way through mediation or moderation. Different outcomes can be explained by different patterns of social inequality (*stratification*) and different effectiveness of securing against the consequences of vulnerability (*security*). Regulations of benefits (*compression*) and the interplay between different areas of social policy-making (*imbrication*) are underlying those other mechanisms.

The second step in identifying different conceptualisations of the welfare state and how it works requires a closer look at the components of welfare states on which research is focussed. In the relevant literature, several perspectives are distinguishable. Following a proposal by Dahl and van der Wel (2013), three approaches to grasping welfare states can be identified that are especially relevant for empirical operationalisations. The first focusses on *regime* affiliation, the second on *expenditure* and the third on the *institutional* organisation (also Bergqvist et al. 2013; Ferrarini et al. 2014a). These three approaches already hint at specific operationalisations such as typologies, indicators of effort, generosity and more (as discussed in previous chapters). Apart from offering a guideline for the selection of indicators, these different approaches reveal different perspectives on the welfare state. While regime typologies highlight broad policy principles, expenditure addresses the size of the budget and institutional arrangements direct the focus on eligibility criteria, generosity and population coverage.12 The expenditure and regime perspectives both suffer from the problems discussed in chapter 3. Both approaches lack clarity, as they are too broad in the case of regimes and too

<sup>12</sup> This corresponds particularly to the social rights perspective.

fuzzy in the case of expenditure. Depending on the conceptualisation, the expenditure approach is sometimes used to capture the "concept of 'welfare resources'" (Dahl & van der Wel 2013: 61). Even though differentiating between areas of spending helps against the ambiguity of an overall spending indicator, it is still not entirely clear, whether those resources result from a generous welfare agenda or a high number of recipients. As a result, out of the three options the institutional approach is viewed as best suited to explore health outcomes and especially the social gradient of health inequality (Ferrarini et al. 2014a: 635–636; Lundberg et al. 2015: S. 32). Pinpointing these three fundamental perspectives on welfare states illustrates the importance of choosing an explicit conceptualisation of welfare stateness for a specific research question. The differentiation between expenditure, institutional and regime perspectives reveals distinct operationalisations, which are by no means identical in nature and therefore allow very different interpretations.

After discussing through which mechanisms the welfare state can influence health outcomes and how different perspectives on the welfare state may reveal distinct conceptualisations, the third and last step explores why outcomes are expected to differ based on particular policies. Even though the term mechanism has been adopted from the literature discussed, the actual causal processes that explain which social policies produce which outcomes and why this might be the case often remain somewhat elusive. In order to substantiate the processes, direct, moderating and indirect effects are now discussed with the help of key hypotheses on how and why the welfare state shapes health outcomes. Summarising prominent hypotheses is complicated by the fact that different studies highlight different aspects within the relationship between social policies and health. To illustrate the diversity of perspectives and arguments in the literature, Table 4.1 gives an overview of hypotheses in the papers sampled for this review. I mainly includes empirical studies and do not list hypotheses twice if several authors assumed the same exact effect. Evidently, this is only a selection and there are most likely many other hypotheses, which are not acknowledged. However, it serves to make an important point: even though the literature on health outcomes and health inequalities discusses the identification of relevant mechanisms and the standardisation of indicators more elaborately than other fields, the tested assumptions still cover a variety of different perspectives on the welfare state.

The selection of hypotheses presented in Table 4.1, illustrates a variety of issues. First, hypotheses often include an implicit (and sometimes explicit) reference to specific perspectives on the welfare state (e.g. expenditure or generosity). What is not represented in the table but visible in the publications is that the reasons for choosing a specific perspective are not always entirely clear. While


**Table 4.1** Welfare states and health—summary of hypotheses

(continued)

#### **Table 4.1** (continued)


*Emphasis (italics) by KK*

some contributions elaborate in much detail why they choose a specific area of the welfare state (e.g. Ferrarini et al. 2014a), others do not. Second, hypotheses frequently go beyond a focus on health policies. Instead, all areas of social policy-making are addressed. Thus, health is assumed to be shaped by welfare state activities in the field of health care as well as unemployment and pensions either (in the sense of imbrication) because policies are intertwined or because they are assumed to have their own distinct effects on the outcome. Third, while moderations are addressed frequently, indirect effects seem to play a subordinate role. Fourth, while what I call the security function receives the most attention in empirical studies, social stratification is only of secondary importance and redistribution is not addressed in the selected hypotheses. A reason for this may be that identifying a distinct effect of redistribution (as cause) and disentangling it from effects of income inequality (as outcome) in empirical analyses is complicated and so far rather unsuccessful (Pickett & Wilkinson 2015: 322).

Lastly, the list of hypotheses—no matter how anecdotal—highlights how diverse assumptions about the relationship between welfare state, health outcomes and health inequalities are. Considering the discussed contributions share a very comparable conceptualisation of the dependent variable (either self-reported good health or bad health), this diversity is remarkable. In order to narrow down the hypotheses and combine them with mechanisms identified before, we can summarise key explanations. Those explanations particularly highlight the *institutional* perspective and the *expenditure* approach. Furthermore, the first explanation highlights the direct impact of social policies on health, while the moderating function is covered by the second explanation. Lastly, for the sake of completeness, the third explanation addresses an effect that is mediated through social inequality, although this one is not as prominently pursued in the literature so far.


In addition, two recommendations regarding the empirical operationalisations can be distilled from the literature: regime approaches are not well equipped to explore general health and health inequalities, since they are too crude to determine whether an empirical result actually captures an assumed causal effect or mechanism. In a similar manner, social expenditure is an ambiguous indicator. Its prevailing presence and importance is most likely a result of data availability. This should still inspire caution. Both recommendations relate to the critical discussion in the previous chapter (cf. chapter 3) and they should be transferable to other instances where welfare stateness is assumed to shape individual-level outcomes.

Although, literature on the link between welfare states and health outcomes and inequality is quite heterogeneous, several aspects can be systematically deduced. In terms of the mechanisms behind the relationship between social policy and health, there is significant overlap between the functions of welfare states and the mechanisms introduced in the relevant literature. A synthesis is quite possible by combining different attempts of grasping how and why the welfare state shapes health outcomes on the individual level. Based on the discussion in this section, the functions of welfare states introduced in the second chapter of this contribution, actually capture relevant mechanisms quite well. In the case of health inequality different ways of providing *security* (including compression and mediation), shaping *social stratification* (including moderation) and organising *redistribution* explain differences in the impact of welfare states on individual health. Overlapping effects of institutions (imbrication) can be subsumed under security as well. If the functions of welfare states actually do represent mechanisms explaining why welfare states shape individual outcomes, they should fit other dependent variables as well.

All of these mechanisms relate to a similar meta-perspective on the welfare state. Studies investigating well-being in general and health in particular mostly conceptualise the welfare state from a *top-down perspective*: it is assumed that the nature of social policies is responsible for an outcome or the moderation of another effect, instead of—for instance—individual perceptions of such policies. This common feature can be helpful in identifying which operationalisations of welfare stateness better fit the mechanisms discussed in this chapter.

Before turning to other dependent variables, it is important to note that while measures of health receive much attention in the literature, other manifestations of well-being are equally important—in particular subjective well-being. There is however some disagreement regarding their measurement.13 Regardless of the operationalisation, many assumptions about the relationship between social policies and well-being are similar to those addressing health outcomes and they are often examined following the same theoretical premises (Voßemer et al. 2018). Even though I focus on hypotheses and mechanisms referring to the link between welfare state and health outcomes, many considerations are considered transferable to the field of subjective well-being as well. Others however may not. If subjective well-being is more strongly linked to the individual perception of the welfare state—for example psychosocial stress is reduced if the welfare state is perceived as a good provider of security—this does not necessarily correspond to the actual responsiveness of welfare states. Because individual perceptions can

<sup>13</sup> More recent contributions on the matter suggest that it is problematic that subjective wellbeing is sometimes operationalised as happiness and other times as life satisfaction (e.g. Ngamaba et al. 2018).

differ from actual policy-making, a focus on subjective manifestations of wellbeing can conceptualise the welfare state not only from a top-down but also from a *bottom-up perspective*, in which it is not the main actor, but merely subject to an individual assessment. This may fall under mechanisms, labelled 'psycho-social' by Molnar and colleagues (2015: 4). Here, emphasis lies on psychological and cognitive processes, interaction with social environments and more. While these thoughts are not explored further in the case of well-being, similar arguments will be taken up again in the course of this chapter, when addressing the relationship between welfare states and attitude formation and behaviour.

#### **4.2.2 The Impact of the Welfare State on Risks and Needs**

The most obvious areas affected by the welfare state are related to risks and needs. Considering that lowering risks and responding to needs is one of the key functions of social policies, it does not surprise that a considerable amount of research is devoted to exploring the relationship between the welfare state and individual risk. Especially comparative studies on the matter usually ask in what way—and often also how well—different social policies lower risks and respond to needs.

As noted before (section 2.2.1), poverty represents the most elementary risk, the welfare state responds to (Saunders 2010). Therefore, it is one of the most common dependent variables in this line of literature. From a comparative perspective, the basic research interest is to explore the extent to which social policies account for cross-national variations in individual poverty (for a review focussing on unemployment policies see O'Campo et al. 2015). Albeit very popular and of course evident, poverty is by far not the only risk explored in the relevant literature. Other studies for instance investigate indebtedness (e.g. Angel & Heitzmann 2015) and (self-reported) economic deprivation<sup>14</sup> (e.g. Visser et al. 2014). While the focus is often on risks and thus the likelihood of an emergency (such as poverty) occuring, deprivation is an example of a research topic in which needs are focussed. Still, both aspects are strongly intertwined as the need that the welfare state responds to originates from risks an individual is exposed to. As in the previous part of this chapter, I will focus on one manifestation of

<sup>14</sup> Research on self-reported economic deprivation shows some overlaps with studies exploring (self-reported) well-being. When further pinning down and systematising key assumptions at the end of this section, this is taken into account.

risk in particular. Since it is a major variable in this line of research, I highlight the relationship between social policies and poverty.

Empirical analyses conceptualise poverty in several ways. We can first distinguish between short-term and long-term poverty as well as absolute and relative poverty (Molnar et al. 2015: 4).15 In addition, poverty among individuals, which are not part of the labour force is often distinguished from in-work poverty (cf. Andress & Lohmann 2008). In most cases, poverty is conceptualised as relative poverty and measured through income. A person is considered poor if a household falls below a threshold of either 50 (e.g. Brady et al. 2017) or 60 percent (e.g. Polin & Raitano 2014) of a country's median equivalised disposable income.

The previous section of this chapter already hinted at some overlap between explanations for morbidity and other risks. Thus, testing whether the mechanisms and conceptual perspectives identified in the case of health outcomes are transferrable to poverty as an explained outcome seems valuable and is further encouraged by a variety of contributions addressing the link between social policies and poverty and health simultaneously (Molnar et al. 2015; O'Campo et al. 2015). Therefore, the following discussion is structured analogously to the previous section of this chapter. I begin by describing the ways in which welfare states can affect individual poverty and discuss the underlying mechanisms. This is followed by a brief summary of the various features of social policies highlighted in the literature, before summarising related hypotheses.

Again, the welfare state is linked to poverty through direct as well as indirect and moderating effects. Similarly to the case of health outcomes, non-direct effects address how social policies interact with micro-level determinants of poverty either by moderating their influence or by reducing (or increasing) their occurrence in case of mediated effects. Overall, there is a number of relevant individual-level predictors of poverty, which are often referred to as penalties. Brady and colleagues (2017: 742) highlight four major socioeconomic determinants increasing the risk of poverty: low education, single motherhood, young household headship and unemployment. In case of mediated effects, welfare policies thus reduce the occurrence of such (and more) risk situations, while moderating effects shape how strongly they increase the risk of poverty. An example for a moderating effect of social policies is given by studies examining the risk of entering poverty as a consequence of childbirth (e.g. Barbieri & Bozzon 2016). In addition to such mediated or moderating effects, there can also be

<sup>15</sup> Absolute poverty refers to an income that is so low, that individuals are not able to meet basic needs such as food and water, while relative poverty means that individuals are below a defined income threshold, which is based on the distribution of income within a society (Eskelinen 2011; Waglé 2014).

direct effects of welfare policies on poverty risk. Since poverty is closely related to income and most commonly measured with reference to income, such direct links seem almost tautological—especially when focussing on policies regulating income replacement (e.g. Lohmann 2009) and generosity of redistributive effort (e.g. Saltkjel & Malmberg-Heimonen 2017).

**Figure 4.3** The impact of the welfare state on the risk of poverty

Turning to mechanisms behind such direct, mediated and moderating effects, again reveals many similarities to the previous discussion of welfare states' impact on health (summarised in Figure 4.3). What Molnar and colleagues (2015: 3–4) summarise as 'materialist' mechanisms is of particular importance. The main explanation here is that individuals are *secured* against risk by replacing deficits through the *redistribution* of resources between individuals and across lifecycles. Furthermore, the provision of security is tied to determinants of poverty at the individual level, as it potentially moderates the impact of being at risk. Following the redistribution paradox (cf. section 2.2.2), targeted welfare states may respond more effectively to occurring risks and needs for the lower social classes, but universal welfare states generate more public support and thus a greater budget for redistribution. They are therefore more advantageous for middle and higher income classes but the increased redistributive budget leads to more effective overall poverty reduction (Korpi & Palme 1998).

Like in the case of health outcomes, *compression* can be seen as a subordinate mechanism underlying the security function. It is setting the upper and lower bounds in the provision of benefits and services that secure against poverty, while *mediation* (and moderation) relates to the relationship between welfare states and *social stratification*. Turning to the mechanism of *imbrication*, the relationship between welfare state and poverty again suggests that many fields of policy-making are relevant and amplifying and overlapping effects may occur. For instance, while unemployment policies are considered very important predictors of poverty (Molnar et al. 2015), they are still embedded in a context in which such effects are shaped by other policy measures as well. An example for this is the suggestion that generous unemployment benefits may even increase poverty as they can be a disincentive to find a job among unemployed individuals in some contexts (O'Campo et al. 2015: 92). Besides hinting at potential overlap and interaction between different policies, this reveals a potential additional mechanism to the ones described before. This additional mechanism relates to what was termed the function of *activation* in chapter 2. Especially active labour market policies (e.g. Clasen et al. 2016), which help and incentivise re-entering employment, aim at reducing risks. While their overall effectiveness is contested (cf. section 2.2.4), evidence suggests they are indeed able to reduce poverty levels (e.g. Cronert & Palme 2017). Thus, investment strategies of welfare states potentially bear a relevant explanation for different poverty outcomes and their relationship to unemployment and other situations of vulnerability—especially as a moderating institution. Differentiating between security and activation in such moderations, represents the difference between a protection strategy and an investment strategy (cf. Vandenbroucke & Vleminckx 2011).

After exploring mechanisms, which serve as explanations for the relationship between social policies and cross-national differences in prevalence of individual poverty on the one hand and penalties associated with risk situations on the other, the next step entails exploring which components of the welfare state are highlighted in relevant contributions on the matter. Here too, a distinction must be made between contributions that highlight welfare state effort (expenditure) and those focussing on the overarching nature of welfare states either by using regime typologies (e.g. Whelan & Maitre 2010) or by emphasising targeted versus universal premises (e.g. Brady & Bostic 2015). These perspectives overlap considerably, for example in studies using social expenditure as an indicator of universalism in welfare states (e.g. Zwiers & Koster 2015). While these two perspectives emphasise what was labelled *expenditure* and *regime* approach before, we also find *institutional* perspectives on the welfare state. Again, the latter include a focus on eligibility criteria (as discussed by O'Campo et al. 2015), benefit generosity (e.g. Brady et al. 2017), and logic of redistribution (e.g. Jacques & Noël 2018).

One last thing should be noted when it comes to perspectives on the welfare state and its impact on individual poverty. A focus on how welfare states meet old risks associated with the labour market and economy provides only a partial explanation for individual poverty. While it lowers penalties resulting from unemployment and low education, individuals affected by the 'new risks' (cf. section 2.1.3) can slip through the net (Brady et al. 2017: 771). Hence, the effectiveness of overall poverty reduction should be a function of how well welfare states adapt (Brady & Bostic 2015: 272). Therefore, examining the role of social policies in poverty related to new risks (e.g. single motherhood or youth unemployment) may require additional perspectives on the welfare state. This could include, for instance, a focus on social investment and active labour market policies.

To take a closer look at the application of the summarised mechanisms and perspectives, a brief summary of hypotheses helps to systematise how and why social policy and individual poverty are related. Analogously to the previous section on health outcomes, Table 4.2 provides a fragmentary listing of hypotheses. Again, this is not intended as an exhaustive review of all existing hypothesis, but as an anecdotic illustration of heterogeneity. Indeed, it shows that the assumptions regarding explanations for differing poverty risks are divers, as it is assumed that very different policy areas influence poverty risks, with particular attention being paid to unemployment-related policies (the link between unemployment insurance and poverty is covered in a comprehensive review by O'Campo et al. 2015).

We can summarise several main explanations for differing poverty levels. While the expectations regarding moderating effects and methodological issues are similar to those formulated in the case of health inequality (section 4.2.1), the first explanation—highlighting the direct impact of social policies on poverty appears to be especially popular. This is to be expected considering the close link between income deficits and measures replacing income. Still, a moderating effect (2.2) and a mediated one (2.3) are present. In addition, a fourth explanation


**Table 4.2** Welfare states and the risk of poverty—summary of hypotheses

(continued)



*\* Review article, Emphasis (italics) by KK*

(2.4) captures the activating mechanisms expressed in those policies, which aim at enabling individuals to leave vulnerable situations such as unemployment.


The main purpose of this chapter is to identify different perspectives on the welfare state, which underlie the central assumptions about its functioning and explain the different distribution of outcomes between countries. In the case of poverty, this perspective bears much resemblance to the one explaining different health outcomes and health inequality (cf. section 4.2.1). The welfare state is again conceptualised from a *top-down* perspective. As such, it is seen as an institution, which actively shapes individual outcomes through several intertwined mechanisms. A main mechanism is to secure individuals from the occurrence of poverty in the first place (*security*) through income replacement or other benefits and services. This is achieved through the *redistribution* of resources on the one hand, and through enabling individuals to overcome situations in which they are at risk (*activation*) on the other hand. The latter can obviously also be funded and organised by shifting resources towards active labour market policies and other activating measures. In many of these cases, the welfare state does not directly affect individual risks, but indirectly through its impact on patterns of social inequality (*social stratification*) and its potential to moderate the outcomes of social determinants of risk. Again, this shows that we cannot identify one singular mechanism responsible for the welfare state's impact on individual risks. Instead, various mechanisms are at work simultaneously. Each of the explanations described can be singled out and examined in more detail—as can the combination of different mechanism.

#### **4.2.3 The Impact of the Welfare State on Attitudes**

Numerous publications deal with the effects of social policy on attitudes.16 Most of the contributions in this field deal with attitudes directly related to the welfare state, such as evaluations of social justice (e.g. Arts & Gelissen 2001), support for social policies (e.g. Brady & Finnigan 2014), general social policy preferences (e.g. Steele 2015; Breznau et al. 2019), welfare chauvinism (e.g. Cappelen & Peters 2018) or preferred spending in the field of social policy-making (e.g. van de Walle & Jilke 2014). I will focus on such welfare attitudes in this section, as they are not only the most extensively discussed, but also stimulate the more sophisticated hypotheses about how and why different welfare states are responsible for the different distribution of attitudes between countries.

The empirical measurement of welfare attitudes is diverse and allows for different foci besides general welfare support. They range from principles of justice to ideal redistribution principles to welfare chauvinism. Furthermore, studies of attitudes towards the overall welfare state have to be distinguished from those focussing on attitudes towards specific policy areas (such as attitudes towards public childcare provision, Chung & Meuleman 2017). This lack of agreement on how to measure welfare attitudes is attracting increasing attention (e.g. Svallfors

<sup>16</sup> The newly-founded Social Policy Preferences Network (SPPN) already lists close to 250 relevant publications: https://sites.google.com/view/sppn/bibliography (accessed 24 July 24, 2018).

2012b: 9). In many cases, different aspects are emphasised in the dependent variable, although general ideas about social justice, for example, differ considerably from the assessment of how well the welfare state functions in one's own country. There are several contributions that are sensitive to the issue of operationalising welfare attitudes, such as the ones bundled in Svallfors edited volume "Contested Welfare States" (Svallfors 2012a). Although my focus is on the macro-micro mechanisms linking the welfare state to attitude formation, this issue is at least noteworthy. Since this macro-micro-link produced mixed results (van Oorschot & Meuleman 2012: 27; Deeming 2018), the question arises whether this is related to different operationalisations of the dependent as well as of independent variables. The latter will receive more attention in this book, while keeping the former in mind. Since the aim of this section is the same as in the previous ones to identify mechanisms, shared conceptualisations and common hypotheses—the discussion follows the accustomed structure.

In general, welfare attitudes are formed "reflecting a mix of ideology and interest" (Brady & Finnigan 2014: 21). They are shaped both by individual-level characteristics such as socioeconomic status and by contextual influences of the welfare state. The welfare state can logically be linked to attitudes *directly*, *indirectly* and as a *moderator*—although not all of these links are examined with equal prominence. At the micro-level, socio-demographic indicators (such as age, class position and gender) as well as political orientation and party preferences (e.g. Koster 2014) and value preferences (e.g. Kulin & Meuleman 2015) are considered relevant influences. Based on these characteristics, it is assumed that individuals have different perceptions of deservingness on the one hand and different self-interests on the other (as summarised by Deeming 2018: 1107). Here the welfare state serves as a potential *moderator*, shaping the effect of socioeconomic position on attitude formation. Alternatively, the effect of social policies can also be *mediated*—for example by individual perceptions of performance (van Oorschot & Meuleman 2012: 26). Another mediated effect could result from the fact that welfare states influence individual-level variations in social status, which in turn explains different levels of satisfaction with democracy (e.g. Sirovátka et al. 2018). Finally, *direct* links between welfare state and welfare attitudes can be found in the cultural imprint welfare states leave on citizens. Following the argument that welfare states represent institutionalised ideas about social justice (e.g. Sachweh 2016), they have the power to shape collective notions of norms and principles (Arts & Gelissen 2001). These processes have been described as the function of *socialisation* before (section 2.2.5).

The latter already leads to the question of mechanisms. In the case of attitude formation, the direct effect of welfare stateness can be explained, on the one hand, by the responsiveness (*security*) and the redistributive logic (*redistribution*) that generate political support for the welfare state. On the other hand, the function of *socialisation* plays an important part when explaining different attitudes. In the sense described above, the socialising function results from the welfare state as a cultural context in which individuals are embedded. Following the previously discussed processes (cf. section 2.2.5), welfare states cultivate dominant normative frames, which are adopted by individuals or at least shape them (Arts & Gelissen 2001; Pfau-Effinger 2005). The distinction between socialisation and the other mentioned mechanisms corresponds to what is differentiated by Sirovátka and colleagues as the "level of the protection and redistributive effect" and the "formative effect" (2018: 3). In addition to such direct effects, the welfare state and especially the socialising mechanism can also be seen as a moderator, shaping how social status is expressed in attitudes. Although plausible, contributions modelling such a relationship are rare. The same is true for the argument that welfare states shape socioeconomic status, which in turn serves as a mediator between social policies and attitudes. Although such effects, which highlight the mechanism of *social stratification*, are discussed (Sirovátka et al. 2018: 4), usually only one of the paths (welfare state—stratification or stratification—attitude) is considered further.

While these explanations for the relationship between welfare state and welfare attitudes follow the same top-down perspective encountered in the case of well-being and risks, this perspective is not sufficient in the case of attitudes. In addition to focussing on how welfare states impact individuals, an important question here is, how individuals *perceive* the welfare state. These perceptions are relevant for explaining indirect effects in which individuals' views on social policies mediate an effect of actual policies as well as direct effects of social policies on the process of attitude formation. It is important to distinguish such a *bottom-up perspective* from the top-down perspective pursued so far. While the mechanisms explaining how and why welfare states shape individual outcomes correspond to the functions of the welfare state discussed before, the question how and why individuals perceive welfare state performance in a particular way requires additional insights. Especially *self-interest* offers a potential answer for the causal mechanism behind the relationship between welfare state and attitude. As Sachweh and Olafsdottir (2012: 151) point out, individuals evaluate, how likely and how strongly they profit personally from welfare state efforts and base their preferences for redistribution on such evaluations. In addition, evaluating how likely one could be *at risk* may influence such considerations as well (Iversen & Soskice 2001). However, self-interest is not the only driver of policy-driven attitude-formation. Instead, individual evaluations of welfare state performance can be inspired by other aspects such as ideology (Petersen et al. 2011; Svallfors et al. 2012; Chung & Meuleman 2017). If a welfare state perceived as guided by egalitarian principles meets egalitarian values at the individual level, this coherence may even amplify support (Calzada et al. 2014: 187). I will refer to such processes of perceiving and assessing the welfare state without necessarily doing so based on self-interest as *evaluation*.

The two different perspectives on the relationship between welfare state and welfare attitudes are illustrated in Figure 4.4. The top-down perspective (A) is characterised by similar pathways as in the previous sections: the welfare state is the main actor and several mechanisms are highlighted. The link between social policies and individual determinants is explained mainly by the stratifying effect and could to some extent also include securing against the occurrence of risk. Direct effects can be explained by socialisation and responsiveness (as indicated by how well welfare states provide security and how they redistribute). Lastly, the moderating effect could be characterised by similar mechanisms as the direct effect. Since it is not as prominently examined, it does not receive as much attention as the other paths (as indicated by the dotted lines). Evidently, the mechanisms overlap and this summary highlights relevant explanations (while others can be plausible as well). Most importantly, these mechanisms illustrate the described top-down influence of welfare states. Thus, impact of policies on attitudes does not necessarily have to be salient.

**Figure 4.4** The impact of the welfare state on welfare attitudes

In contrast, where the focus rests on the individual perception and evaluation of social policies (B), the individual is assumed to have at least partial information about actual policy-making. Such information is used when evaluating social policies. Hence, the impact of the welfare state is filtered through individual perception. While both perspectives are equally plausible and coincide, by distinguishing between the two, I would like to emphasise that each corresponds to a specific theoretical perspective on the matter. Following the discussion in chapter 3, this distinction is highly relevant for the question, which areas and components of the welfare state should be highlighted conceptually as well as empirically.17

Following these arguments, identifying relevant areas of welfare stateness in this case requires considering both perspectives. Starting with the *top-down perspective*, relevant areas of welfare stateness can be found in literature on the formation of public preferences. As Breznau (2017: 584–585) summarises, one perspective (the thermostatic model) focusses on *expenditure*, while a second perspective (the increasing returns model) emphasises entitlements and generosity, and thus what falls under the *institutional* component of the welfare state mentioned earlier. Studies examining the extent to which welfare states have a socialising effect on individuals sometimes use a *regime* approach (Arts & Gelissen 2001). However, more recent contributions of this kind produce mixed findings (van Oorschot & Meuleman 2012: 27). Overall, using a selection of distinct characteristics, representing either spending or generosity is preferred (Jæger 2006). This is supported by my own analyses conducted in chapter 3. A further component highlighted in this context is the degree of egalitarianism in a welfare state. This can for instance manifest in egalitarian gender norms (Chung & Meuleman 2017: 54). Since gender ideologies and family policies are losely linked (Grunow et al. 2018), such cultural perspective on the welfare state are highly relevant.

While the above-summarised perspective on welfare stateness is familiar, as it corresponds to what has been discussed in the two previous parts of this chapter, capturing the *bottom-up perspective* adds an entirely new challenge. The literature agrees that individual perception of welfare state performance (Roosma et al. 2014) and the likeliness of ever being at risk and in need of welfare services are important aspects when explaining why and how individual attitude-formation takes place (Jordan 2013: 136). However, little is known about which components

<sup>17</sup> I leave out the question, whether public opinion may feed back and potentially alter policy-making. Within literature on welfare attitudes, this is certainly an important issue, covered e.g. by Breznau (2017). Since I am concerned with the initial macro-micro link, such reciprocal processes are of secondary importance to this project.

of social policies are relevant in this case. Instead, the above-discussed ways of approaching the welfare state (und ultimately measuring policy differences) are used. Following the discussion in chapter 3, too little seems to be known about which elements of welfare states are actually perceived—and if so correctly. After all, it is unlikely that individuals are fully and accurately informed about all aspects of social policy-making (Breznau 2013: 3). How problematic this can be, is illustrated in literature on welfare knowledge. This strand of research reveals that citizens perceive the reality of welfare provision in a distorted way. For the UK Taylor-Gooby and colleagues (2003) show that in particular spending for unemployed individuals is usually overestimated quite severely. Evidence from the US furthermore shows that the cost of replacement payments is overrated and that benefits are perceived to be more generous than they actually are (a good summary of main findings is given by Geiger 2018). Hence, it seems insufficient to rely on established approaches of capturing characteristics of welfare states if the bottom-up perspective is emphasised. Evaluation of policies in general and self-interest in particular, were identified as key mechanisms explaining how individuals look at the welfare state and why they form different opinions based on what they perceive. For this purpose, it seems highly relevant that the emphasised components within the welfare state represent salient features. As (Sachweh 2018: 50–51) points out, especially when confronted with crisis, it is likely to expect that indicators more closely related to security are more salient than those representing patterns of inequality. Furthermore, it seems plausible to assume that benefit generosity is related to perceptions guided by self-interest, as they represent potential benefits. Other indicators of eligibility such as benefit duration, waiting periods and overall insurance coverage may be of secondary importance, as they require very specific knowledge about benefit provision. These considerations, which go beyond the approaches highlighted in the literature, will be revisited in the next chapter when it comes to elaborating different conceptualisations and their empirical operationalisation. For now, suffice it to summarise that the link between welfare states and individual attitudes can and should be approached from two sides—a bottom-up perspective highlighting the individual process of attitude formation and a top-down perspective emphasising the welfare states impact on shaping the context in which attitudes are formed. Both perspectives can be combined well in one explanatory model, if both kinds of processes are examined.

Hypotheses in the literature on welfare attitudes are characterised by a variety of different approaches to modelling the relationship between welfare state and attitudes. Summarising hypotheses in this case is complicated because not only do perspectives on the welfare state vary, but so do operationalisation of welfare attitudes. Therefore, I refrain from presenting hypotheses anecdotally and instead turn directly to summarising the relevant assumptions. Overall, universalism is expected to be among the main determinants of public support for redistribution (cf. section 2.2.2 and the paradox of redistribution) as well as support for social policies in general (e.g. Jordan 2013). The same can be said for the generosity and size of welfare states, which one may call their performance. In both cases, the explanation for such relationships may lie in both the socialising effect and the active evaluation by the individual. Thus, these relationships can be approached from the top-down as well as the bottom-up perspective. Summarising the literature presented in this section of the chapter, the following explanations for how social policy influences attitudes can be derived from the discussion.


The impact of social policies on attitudes towards the welfare state raises more complex issues than those that arise in the case of poverty and well-being. While the latter are closely related to responsiveness of welfare states and mostly address top-down perspectives, the process of attitude formation is less straightforward in many ways. Not only does it conceptualise individuals as actors of their own, but it also highlights more intricate processes. Conceptualising und ultimately operationalising mechanisms related to socialisation or self-interest is much more challenging than in the case of responsiveness in situations of risk.

While attitudes towards welfare state policies are covered by the majority of literature assuming an impact of social policies on attitudes, there are notable further dependent variables addressing attitudinal phenomena. These include general political attitudes, such as satisfaction with democracy (e.g. Sirovátka et al. 2018) and support for Europe (e.g. Beaudonnet 2015). In addition, there are miscellaneous other attitudes explored, including attitudes towards migration (e.g. Rapp 2017) and self-employment (e.g. Rapp et al. 2017). Essentially, hypotheses tying welfare stateness to all three groups of dependent variables—welfare attitudes, political attitudes and miscellaneous other attitudes—are quite similar at their core. They all can focus on a bottom-up perspective or a top-down perspective (or both). Therefore, methodological issues arising from this distinction can be considered transferrable.

#### **4.2.4 The Impact of the Welfare State on Behaviour**

Various behaviours are shaped by welfare states. The most evident ones are those tied to the labour market and the balance between paid employment and household labour (including care responsibilities). This includes objects of research such as absenteeism (Sjöberg 2017), the likelihood of entering risky endeavours such as self-employment (Rapp et al. 2017) as well as gendered division of household labour (Hook 2010) and labour market participation of fathers (Bünning & Pollmann-Schult 2016) and mothers (Gangl & Ziefle 2015). In order to discuss mechanisms and perspectives on the welfare state, I choose the latter as an example for two reasons. First, mothers' labour market participation is a popular and widely covered topic. Second, it allows for the discussion of very different areas of welfare stateness than those explored so far, as it is more closely linked to family policies and thus contains aspects that have not received much attention in the last three sections of this chapter. Following the same structure as before, I will first turn to paths and mechanisms, before elaborating on relevant facets of welfare stateness and summarising main hypotheses and explanations in the literature.

Mothers' labour market participation can be categorised in the field of new risks and risk groups (cf. section 2.1.3). As Chung & Meuleman (2017: 52) point out, responses to such new risks and new risk groups may not fit mechanisms developed with reference to the old risks. So far, I have indeed focussed on classical risks tied to benefits and services that are provided if unemployment, sickness or old age threaten income. In explaining behaviour related to the labour market and gender issues in particular, it is necessary to consider not only the classical functions of the welfare state but also more recent policy agendas such as the increasing promotion of female employment in general and maternal employment in particular. When researching labour market participation of mothers empirically, the existence of employment and its scope in terms of working hours are often differentiated (Erhel & Guergoat-Larivière 2013: 78). Welfare states are assumed to shape both outcomes in a similar manner.

The links between social policies and maternal labour market participation partly overlap with those discussed in the previous sections. Again, we are confronted with direct effects, as well as indirect and moderated ones. Furthermore, this type of behaviour as an outcome is tied to both essential perspectives on the welfare state as an explanatory factor: *top-down* and *bottom-up*. When the research question is why and how individuals respond in a particular way to social policies, a *bottom-up perspective* is taken. In contrast, studies exploring the impact of welfare policies on behaviour follow a *top-down perspective*.

Within the top-down perspective, direct effects can especially be found in policies that target a specific behaviour—in this case labour market participation of mothers. This is especially relevant in the case of *activation*. The design of workfamily reconciliation policies enables equal participation in the labour market to varying degrees (Grunow et al. 2018: 42–43). Such activation can be directly tied to labour market participation—for instance in case of a general tendency to enable all women. In addition, it can also have a moderating effect if activation especially affects women in a particular situation (e.g. unemployment or single motherhood) or shapes the impact of other determinants such as the gendered division of household labour. It can also be argued that there is a link to the provision of *security* because increasing labour market participation of parents in general and mothers in particular also leads to economic benefits (Morrissey 2017: 3). Since this happens through activation, provision of security seems to be of secondary importance in this case. However, family-policies can also have an adverse effect. Expanding policies related to the length and generosity of parental leave may reduce maternal labour market participation (Gangl & Ziefle 2015: 520). For instance, generosity of financial support is therefore assumed to have a negative impact on mothers' employment (Nieuwenhuis et al. 2012: 615). In sharp contrast to an activating impact, this can be referred to as a *deactivation* of mothers (cf. section 2.2.4).

Besides an activating component, the welfare state is tied to gendered outcomes through its impact on role models and gender ideologies (Grunow et al. 2018). In particular, work-family policies are shaped by both welfare culture and gender culture (Pfau-Effinger 2018: 170) and provide legitimised reference points for preference formation (cf. Pollmann-Schult 2016) as well as incentives for the dissemination of specific gender roles. On the one hand, they set certain norms about the relationship between employment and motherhood and, on the other hand, they determine how widespread caregiving is in a society. While the first mechanism is referred to as norm setting, the second is related to role exposure (Gangl & Ziefle 2015). It is my understanding that both explanations can be subsumed under *socialisation* because shaping what is perceived as a legitimised cultural frame is the driving mechanisms behind both. In this case, socialisation exhibits direct effects, as the dominant cultural norms regarding gender equality influence how natural, supported and even socially desirable female labour is. However, it can also be incorporated as moderating influence, which shapes the severity of individual-level determinants of female employment such as education, partnership status, age, but also preferences and more (Gangl & Ziefle 2015). Since at least the determinants related to the individual-level manifestations of social inequality are potentially shaped by social policies, indirect effects are conceivable as well. Such effects, where the impact of family policies is mediated by mothers' characteristics, are not examined in the sampled studies (a similar conclusion is reached by Pollmann-Schult 2016: 24).

Turning to the bottom-up perspective, individual perception again represents an important link between policy-making and behaviour. In the case of behaviour, *self-interest* is a commonly assumed explanation for observed outcomes. Like in the case of attitudes, individual perceptions of social policy-making are assessed in terms of costs and benefits, resulting in a rational behavioural choice. Regarding labour market participation, this is based on the premise that having children negatively affects female employment. From an economic point of view, a major reason for this is higher relative income among men (Erhel & Guergoat-Larivière 2013: 77–78), which promotes the male breadwinner model. Thus, the more welfare states focus on compensating for income differences or incentives for parental leave among fathers (cf. Geisler & Kreyenfeld 2019), the higher maternal labour market participations is assumed to be.

Another reason for employment can simply be individual preferences although literature confirms a prevailing mismatch between preferred and actual working hours—especially among women and parents of children below school age (Steiber & Haas 2018). Welfare cultures may be responsible for the formation of such preferences, which are then weighted against the costs and benefits of labour market participation or are *evaluated* according to other principles (as in the case of attitude-formation). This is related to the effects of *socialisation* mentioned. In this bottom-up perspective, however, the difference is that normative reference frames not only leave an imprint on individuals, but are perceived and actively incorporated in a behavioural choice. Still, it is difficult to attribute findings like the fact that women in countries that are closer to the conservative ideal appear to prefer less working hours (Pollmann-Schult 2016) to either socialisation or an active assessment. It seems advisable for further research to explore which of the perspectives on the interplay between policies and behaviour is more accurate or if both are similarly at work.

A summary of the paths and mechanisms introduced briefly in this section is provided in Figure 4.5. From a top-down perspective (A), social policies in general and particularly family policies can serve as direct and as moderating effects. Here, activating mothers by incentivising them to work and relieving them of care responsibilities can increase labour market participation and it can shape the impact of social determinants. Similarly, the prevailing welfare culture may foster a social climate that can account for variations in maternal labour market participation across countries. In contrast, the bottom-up perspective includes individual perception as an intermediary premise (B). Here, behaviour is based on an assessment of social policies in a similar manner as in the case of attitude-formation. However, especially in the example of mothers' employment, cost-benefit considerations play a more important role.

**Figure 4.5** The impact of the welfare state on employment of mothers

Similarly to the three previous exemplary dependent variables, different areas of welfare stateness are highlighted in literature on the impact of policy-making on maternal employment. The main difference is that in the previous examples the focus was on managing old risks and the classical policy areas (mainly unemployment, sickness, disability and old age), whereas here the focus is on family policies. Accordingly, in sharp contrast to the previous examples, maternal labour market participation is a phenomenon that is assumed to be shaped mainly by very specific components of welfare states. Evidently, work-family policies are highly relevant and an impact of other policy areas (such as unemployment or health) is unlikely or marginal. While this is a very different emphasis, the way in which these policies are operationalised is quite similar to the other examples discussed in this chapter. Here as well, *expenditure* is among the approaches capturing differences between countries. In this case, expenses on childcare are highlighted (e.g. Andringa et al. 2015). Moreover, literature frequently focusses on *institutional* characteristics, which—again analogously to the previous examples—include work-family reconciliation policies; in particular childcare provision (Erhel & Guergoat-Larivière 2013; Morrissey 2017). In addition, eligibility criteria and other aspects that emphasise the level of defamilisation represent institutional characteristics (Keck & Saraceno 2012). Lastly, *regime* approaches do exist in this field of research. Usually, they take a more general perspective and examine patterns within policy-making (Saraceno & Keck 2011). As independent variables, they receive criticism in line with previously stated arguments in this book such as that they are too broad to determine, which specific aspects are at play (Nieuwenhuis et al. 2012: 615–616).

In summary, hypotheses about how and why family-policies shape female employment are strongly related to incentives provided top-down and cost-benefit considerations by the individual (bottom-up). An additional perspective is provided by contributions that emphasise cultural explanations and focus on the potential to foster egalitarian behaviour through policy-making that embodies an egalitarian principles. The following three explanations summarise the most popular hypotheses. Again, this list does not claim to be exhaustive and focusses on the most basic assumptions about the impact of policy-making on labour market participation of mothers.


The first two explanations could also be reformulated to include a moderating effect in which certain groups of mother (e.g. based on education or number of children) are particularly influenced by family-policies. Since such analyses are not prominent in the sampled studies, they are not listed here. Furthermore, the discussed deactivating effect that may result from policies that support women's responsibility for childcare is not explicitly mentioned, as it is the counterpart to Explanation 4.1.

While labour market participation of mothers in particular and of women in general is a popular issue, other behaviours are shaped by welfare states as well. Several examples have been mentioned in the beginning of this section. In the examples chosen so far in this chapter—well-being, risks and attitudes—it was assumed that the summary of mechanisms and hypotheses for the exemplary dependent variables is at least partially transferable to many other research topics. As far as the employment of mothers is concerned, this is not so obvious, since in this case a very specific part of the welfare state is emphasised. Nevertheless, there is reason to believe that explanation such as activation and self-interest are also relevant to other research questions that examine the impact of policy-making on behaviour.

### **4.3 Summary: Analytical Perspectives and Mechanisms**

The summary of dependent variables and hypotheses shows that there is a small number of distinct mechanisms and perspectives on the welfare state embedded in literature on the impact of social policies on individual-level outcome. However, these different perspectives are not necessarily tied to a specific dependent variable—we find similar hypotheses about the way in which the welfare state works across different outcomes.

The short review in this chapter reveals another noteworthy issue. Although theoretical assumptions are formulated in many fields about why and how the welfare state influences individual-level outcomes, this does not usually lead to a recommendation for specific operationalisations. For example, when responsiveness is identified as a key feature in explaining differences in poverty risk across countries, the empirical operationalisation of responsiveness still varies. This is true even in areas that seem to be sensitive to the issue, such as in the literature on health outcomes. While processes and mechanisms are laid out in much detail, they are still not operationalised in a standardised manner—even though proposals for such standardisation exist. This seems rather unsatisfactory and raises the question why it seems to be so difficult to agree on a set of operationalisations that fit certain hypotheses better than others. Following this thought, I would like to concretise this missing link.

Essentially, we can take away three things from this literature review. The first is a set of mechanisms on which the hypotheses discussed are based. These mechanisms are consistent with the main features and functions of welfare states (cf. chapter 2) and emphasise different modes of functioning. The second point we can take away is that there are two overarching perspectives on how the welfare state is addressed. One of them sees it as a top-down institution that actively shapes processes and outcomes at the individual level. The other approach is based on a bottom-up perspective, where the welfare state is a passive element perceived and judged by the individual. Lastly, these two elements—the general perspective on the welfare state and the mechanisms emphasised—are linked by theoretical and empirical conceptualisations that highlight specific areas that are assumed to be at work. This includes *regime* approaches as well as those that focus on *expenditure* (welfare effort) or *institutional* aspects (social rights in general and/or benefit generosity in particular). Benefit receipt as an additional perspective, does not appear to be a popular choice as an independent variable.

Depending on how one approaches these concepts, this can involve a variety of different indicators and operationalisations, which brings us back to the main objective and the second research question of this study: *How can we derive a more standardised, transparent and comparable approach to operationalising the welfare state as an independent variable?* I believe that the identified mechanisms and the two meta-perspectives on the welfare state can contribute very much to answering this question. The first step in finding the right indicators that capture those features of the welfare state that are essential for particular research questions and hypotheses should be to state clearly which mechanisms are assumed to be at work and which perspective on the welfare state is chosen. Before however we can approach explicit measurements, the welfare state has to be conceptualised as an explanans. This intermediate step could be the missing link between the existing hypotheses and measurements and has the potential to standardise the approach considerably, as explicitly modelling different perspectives on the welfare state should narrow down the choice of indicators in a meaningful way.

Table 4.3 summarises the main perspectives, mechanisms and associated processes identified in the literature review. This includes the distinction between top-down and bottom-up conceptualisations. Underlying these two higher order perspectives are at least18 seven mechanisms. The mechanisms framing the welfare state from a *top-down perspective* correspond to the functions of the welfare state discussed in the beginning of this book. Here hypotheses posit that social policies have a distinct influence on the examined outcomes. These effects can be intentional (as in the case of *securing* against risks), unintentional (as often

<sup>18</sup> I would like to reiterate that I do not claim exhaustiveness. I intend to illustrate a systematisation of the main approaches and hope that the applicability of the idea to other research topics, questions and hypotheses will be critically tested in the future.

in the case of *socialisation* of individuals), or both (in the case of *activation*). Empirically, this presupposes that the features of the welfare state are measured against objectively relevant criteria and that its effects can be invisible to the individual. Furthermore, two mechanisms—*stratification* and *redistribution*—are theoretically relevant, but it proved difficult to include them in the discussion. This is most likely due to the exemplary dependent variables chosen, where the influence of stratification and redistribution is only relevant as part of the provision of security. At this point, it is not possible to determine whether this is always the case or if the two provide important explanations for other dependent variables. They therefore remain in the concluding overview (in italics in Table 4.3), but should receive more attention in future research.

In contrast, mechanisms falling under the *bottom-up perspective* assume a more passive role of the welfare state. What is most important here is not the objective features of a welfare state, but the individual perception thereof. Such subjective perspectives on social policies, which are mostly found in studies exploring individual *evaluations* in general and the role of *self*-*interest* in particular (rational choice), can vary considerably from objective characteristics and performance.


**Table 4.3** Summary of key perspectives, mechanisms and processes

Three processes that were also introduced as mechanisms are missing from this table: compression, imbrication and mediation (all inspired by Beckfield et al. 2015). Throughout this chapter, compression (regulating upper and lower bounds of health inequality) and mediation (or moderation) were always subsumed under either security or stratification. I believe that they relate to important processes within these two mechanisms, but should be seen as components rather than independent mechanisms. Since I am concerned with minimising the complexity of the theoretical premises, this seems to be a viable option for my purposes. However, if these specific elements are to be studied separately, there is no reason not to consider them as isolated mechanisms. In contrast, imbrication seems to be situated on an even higher level than the mechanisms in Table 4.3. Overlap and interaction between policy areas (and perhaps also mechanisms themselves) should always be taken into account, but as a downstream step.

This chapter has summarised and systematised the main hypotheses on the impact of social policies on various individual-level outcomes. The literature review has revealed distinct perspectives on the nature of welfare states. It now remains to specify which concepts of welfare stateness emerge and how empirical operationalisations can best capture them.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **5 Welfare Stateness as an Explanatory Concept**

After systematising main assumptions about the influence of the welfare state on individual-level outcomes, I turn to deriving specific measurements of welfare stateness that capture the mechanisms underlying the explanations and the corresponding hypotheses. According to Adcock and Collier, "[v]alid measurement is achieved when scores […] meaningfully capture the ideas contained in the corresponding concept" (Adcock & Collier 2001: 530). Achieving such meaningfulness is a considerable challenge and, to my knowledge, there has not yet been a comprehensive attempt to derive measurements for specific concepts of the welfare state that serve as explanans.

A very good and recent example for a similar undertaking is given in a publication by Andersson (2018). She proposes a theoretically grounded way to operationalise the specific examples of social investment policies. Relying on Sartori's (1970) ladder of abstraction, she first deduces relevant dimensions of the broad concept of social investment, before suggesting specific variables capturing these dimensions. Although her aim is to provide a conceptual framework leading to the proposal of a typology of social investment approaches, the main objective of her study—to improve and standardise the measurement—is related to the aim of my project. In this chapter, I pursue a similar approach of isolating indicators by gradually narrowing down the underlying concepts. To this end, I draw on an analytical framework proposed by Adcock and Collier (2001), which shows resemblance to Sartori's ladder but focuses more on deriving indicators for empirical operationalisations.

Adcock and Collier's framework starts at an abstract level. The *background concept* is very broad and includes all meanings associated with a research object. This most general way of looking at a concept is usually not the one that is actually addressed in specific research questions. Instead, the focus is on only one specific meaning or understanding. This selection of specific parts within a

**Figure 5.1** Conceptualisation and Measurement. *(Own figure based on Adcock & Collier (2001: 531))*

broader background concept is guided by the research question and it involves the task of conceptualising a given phenomenon in depth. This leads to the second step in the framework, which is the formulation of a *systematised concept*. Up to this point, the main objective has been a conceptual one, aimed at identifying the relevant understanding of a given phenomenon. As soon as the systematised concept is deduced and defined, the question of empirical operationalisation arises. The third step within the framework is thus to select indicators that allow the systematised concept to be captured empirically. The last step consists of the actual attribution of *scores for cases* based on these indicators. After implementing the measurement and putting it to the test, the model also includes the possibility to revise the measurement based on empirical observations (Adcock & Collier 2001: 530). Figure 5.1 summarises these steps and tasks.

Their framework fits my objective for several reasons. First, it is possible to combine their proposal with the criteria derived at the end of chapter 3. *Comparability,* c*larity*, and *availability* have been identified as key factors to consider when selecting an operationalisation. These three criteria can be well integrated into the argumentation when it comes to selecting indicators (third step in Figure 5.1). Second, their framework allows for continuous improvement as it implies a circular process of approximation. Since I hope that my systematisation of mechanisms, concepts and indicators will be critically discussed, applied and expanded in future research, it seems very useful to review each step and modify it if necessary. Third, it enables a differentiation of perspectives within a broader phenomenon through a specific distinction between a background concept and systematised concepts. In this case, this refers to different conceptualisations of welfare stateness depending on the respective dependent variable, research question, pursued explanation and hypotheses. Fourth, and perhaps most importantly, it emphasises the importance of separating conceptual issues from issues related to measurement. Since I argue that the conceptualisation of welfare stateness is lacking prior to its operationalisation as an independent variable, it is particularly important to distinguish between the concept and the resulting possible measurements. This additional step holds the potential to narrow down and standardise distinct perspectives on welfare stateness as an intermediary step.

### **5.1 From Background Concept to Systematised Concepts**

As a background concept, the welfare state embodies various different meanings, influences different societal levels and affects all the functions and mechanisms discussed in relation to social policy-making. As such, the welfare state is tied to institutions and institutional developments on the macro-level and to individuals as beneficiaries, contributors and/or judges on the micro-level. Likewise, it can be an explanatory factor and/or the phenomenon to be explained as it is interrelated with structures and dynamics on both levels. It causes outcomes and is shaped by feedback processes. Industrialisation and democratisation were certainly relevant drivers of welfare state development and change (chapter 2), but conversely, social policy arrangements have also shaped macroeconomic and societal outcomes. Not least, because they are strongly linked to economic cycles and demographic developments. Moreover, there are also reciprocal relationships with the micro-level. Since welfare states are shaped by governments, individuals do in fact possess the opportunity to influence the direction of social policy-making in elections (this idea of simultaneous feedback is discussed in more detail by Breznau 2017, 2018).

With the elaborate conclusions from previous chapter in mind, it is apparent that conceptualising 'the welfare state' as one all-encompassing construct is not conducive if it is to serve as an independent variable. Therefore, if we want to approach a more precise, standardised and comparable manner of looking at distinct facets of welfare stateness, the idea of deriving systematised concepts is quite promising.

In the broadest sense, the previous chapter revealed two ways of approaching the welfare state: (1) top-down, as a responsive and not necessarily visible actor, and (2) bottom-up, perceived as a visible and salient institution. These two perspectives frame the welfare state in different ways. Within these two main perspectives, several different mechanisms are at work, depending on the research question, explanation and hypotheses.

The *top-down perspective* takes up mechanisms that explain why and how different welfare states in general and different social policies in particular affect individuals. As illustrated in the review of hypotheses in the previous chapter, those mechanisms correspond to the functions of the welfare state highlighted in the second chapter of this book. In this sense, the welfare state can constitute a source of security against risks. Moreover, it redistributes resources, shapes patterns of social inequality, enables or incentivises specific behaviours and shapes a normative reference frame. Some of these functions overlap quite a lot, while others offer a more isolated view.

In contrast, the *bottom-up perspective* explores why and how individuals perceive and evaluate welfare states and why and how these assessments shape outcomes—especially related to behaviours and attitudes. Two mechanisms may be in effect, which both entail an evaluation of social policies by the individual. On the one hand, the evaluation may follow self-interest. On the other hand, other explanations—such as an activation of individual values—are possible. Since both processes are based on the same premise, the systematised concept of welfare stateness is similar: a welfare state that is being assessed by individuals. Since mechanisms do not yet reveal systematised conceptualisations of the welfare state, I would like to use and further distil the insights gained so far on theoretical premises, mechanisms and hypotheses.

Looking at the welfare state from the broadest perspective, all seven discussed mechanisms are related to each other—even transcending the two main perspectives. Modes of redistribution may mould political support for the welfare state and public support—if reaching a critical mass—can change the logic behind redistribution of resources. Equally, the way in which the welfare state stratifies societies is linked to how it socialises respondents and this may again be subject to individual evaluation. Incentives may be shaped by patterns of inequality instead of intended (or unintended) activation through specific policies, which in turn can determine how effective a welfare state responds to risks. One could further combine these different mechanisms and it would become clear that all these different processes related to social policy-making are interconnected in one way or another. Considering that they are tied to the same background concept the welfare state—this is no surprise. In deriving systematised concepts based on these mechanisms, the question is therefore not which mechanisms are at play, but which mechanisms are *most characteristic*. Here, an important distinction lies in the key actor within an explanation– welfare state or individual—and the mechanism (or mechanisms) highlighted.

Following this approach, there are at least four systematised concepts that emerge from the discussion so far. Three of these concepts refer to the top-down perspective; one addresses the bottom-up perspective. They are summarised in Figure 5.2. This figure also highlights the most relevant mechanism(s) for each concept. It should be reiterated that other combinations are possible.

*conceptualisation*

**Figure 5.2** Conceptualisations of the welfare state

*The Responsive Welfare State* emphasises those mechanisms that determine how effective a welfare state operates. Especially characteristic mechanisms are the efficiency of reducing risks and uncertainties (security) and reducing social inequality (stratification). This happens particularly through redistributive processes. Evidently, the responsiveness of welfare states could also be determined by its potential to activate individuals and promote specific norms and justice principles (socialisation). These mechanisms are however of secondary importance in this case.

*The Enabling Welfare State* is characterised above all by the mechanism of activation. The name is based on the term *enabling state* used in the literature, which is introduced in the second chapter of this book. This systematised concept highlights processes in which the welfare state incentivises and enables individuals to take responsibility for their own welfare. As discussed, one way this is achieved is through labour market integration through investment in human capital. Again, there is overlap with other mechanisms, but the focus of the conceptualisation is on activation.

*The Normative Welfare State* represents the last systematised concept within the top-down perspective. Socialisation is particularly characteristic here, while other mechanisms are possibly at play but are only relevant in terms of their socialisation potential (e.g. when it comes to justice principles underlying redistributive logic).

In contrast to these three conceptualisations, *the Assessed Welfare State* represents a bottom-up perspective. Here, individual evaluation in general and activation of self-interest in particular represent the most relevant mechanisms. Even if individuals form attitudes and behaviours based on an assessment of the five top-down mechanisms, it is their perception of those processes that matters. Moreover, these individual perceptions can deviate substantially from how policies are designed and performing in reality.

At this point, it is important to reiterate that this summary does not claim exhaustiveness. Its purpose is to illustrate a possible approach to conceptualising the welfare state in a way that is explicitly intended to treat it as an explanans. In this contribution, a selection of dependent variables is explored and adding others may reveal further conceptualisations or adjustments of my proposal. Following the selected measurement framework, revisiting the proposal critically is not just possible but highly encouraged. While this is important to recognise, the central message is that when adressing any of the concepts of welfare stateness in a hypothesis, it is important to clarify which perspective and mechanism is chosen (and why). Committing as clearly as possible to a systematised concept, will allow the most targeted selection of indicators. Moreover, by stating, which perspective is adopted, comparability should be increased considerably. Finally, a combination of different concepts can be beneficial to many research objectives. This includes combinations of top-down and bottom-up perspectives. The strength of adding systematised concepts of welfare stateness as an intermediate step to empirical operationalisations lies in the limitation to specific perspectives. By not attempting to cover all facets and meanings of the welfare state, theoretical arguments on the one hand and empirical decisions on the other are sharpened.

Overall, this proposal resembles a typology. Unlike the typologies discussed so far, however, this is not a typology of welfare state regimes, but of conceptual perspectives that we encounter when we assume that social policies affect individual outcomes—be they risks or attitudes or behaviour or more. Intended as an auxiliary tool, it still has to prove its usefulness in this contribution and hopefully further research on the matter.

### **5.2 Indicators and Scores for Cases: Criteria and Measurement Validity**

While the previous section sketched a possible framework for conceptualising distinct perspectives on the welfare state, the following part discusses how to get from concept to operationalisation aiming at ultimately deriving specific indicators. The analysis of the problems associated with common operational approaches in the third chapter of this book led to the identification of three criteria that are often not satisfactorily met: (1) comparability, (2) clarity, and (3) availability. These three criteria are discussed in detail in this section, with emphasis on how systematised concepts of welfare stateness may help to solve underlying issues at least partly. This is followed by a discussion of how measurement validity can be determined.

#### **5.2.1 Criteria for the Selection of Indicators**

The first criterion, c*omparability*, relates to four issues: (1) different kinds of indicators, (2) different sources for the same indicator, (3) different points in time, and (4) different countries. Turning to the first, it seems almost trivial to say that comparable indicators need to be used. However, as explained in chapter three, this is often not the case. Instead, different types of indicators are compared in the review of literature as well as in the selection of empirical operationalisations. This happens for a number of reasons already outlined—not least because of the limited availability of many indicators.

Comparability can be affected not only by the use of different indicators but also by using the same indicator based on different data sources. In the third chapter, I referred to the case of net replacement rates and illustrated different effects obtained by using two different versions of this indicator. Such deviance in results seems highly problematic—especially since someone less familiar with the surrounding debates would probably not expect it. As discussed in detail by Ferrarini and colleagues (2013) and others (such as Scruggs 2013; Wenzelburger et al. 2013), different values result from different theoretical premises and operational decisions. The discussion shows that comparability is not just a matter of comparing different indicators but also different operationalisations within one indicator. Both issues can be solved at least partly by being more aware of the issue as well as discussing and explicating the selection. Thus, differing results in literature using different indicators for testing similar hypotheses, should not come as a surprise but should instead be expected. In my opinion, much can be achieved by providing an overview of possible indicators and results obtained from different approaches.

The situation is more difficult with the other two issues that could affect comparability. The same score for an indicator can have a different meaning depending on its context (Adcock & Collier 2001: 534). This can refer to different times, in which specific indicators have different meanings (e.g. due to external shocks or reforms). Furthermore, it can refer to different countries—the latter being of particular importance. As the case of CEE countries frequently shows, there may be severe differences between de jure and de facto benefits and other issues, which may complicate the use of established indicators in these cases. Recent contributions dealing with the greater inclusion of CEE countries in comparative welfare state research are a first step towards addressing the issue (e.g. Kuitto 2018).

So, how can comparability be maximised in the process of operationalising distinct systematised concepts of welfare stateness? First of all, explaining which concept is adopted and why draws attention to how the concept should be operationalised. By reducing the welfare state to narrowed and specific concepts instead of a comprehensive and multidimensional background concept, it should also be easier to determine the comparability of different operational decisions as they are limited to a clearly defined part of the welfare state. This especially helps solving issues associated with the selection of indicators but it should also help to assess the context-spanning comparability of operationalisations.

The criterion of *clarity* is related to comparability but emphasises a particular concern. Indicators should not only be comparable; they also have to be sufficiently meaningful and interpretable. In other words, it should be as obvious and clear as possible what an indicator is measuring. In the case of social policy indicators, this is often not the case. Several examples for this were used repeatedly throughout this book. Indicators on social expenditure and regime approaches are the most affected. Both kinds of operationalisations give too little indication of what they actually measure. As the case of social expenditure as an indicator of welfare effort shows, it cannot be ruled out that high spending in one country is a sign of high generosity, while equally high spending in another country may indicate a high number of beneficiaries. Moreover, even controlling for benefit receipt (e.g. by adding unemployment rates) still does not reveal redistributional principles, eligibility rules and many other important aspects. Regime approaches, on the other hand, incorporate too many facets of welfare stateness making it impossible to determine, which one is responsible for an observed effect. In addition, cross-cultural differences are marginalised by reducing variation between countries to a handful of types. Such approaches, which aim at grasping either distinct types of welfare states or overall measures of redistributive budget in a broad manner, have their place in comparative welfare state research. However, they are rarely suitable for testing hypotheses. Too little is known about what they capture to test whether specific mechanisms behind causal assumptions prove to be valid. The same can be said about composite indicators (Kvist et al. 2013: 332).

Much of the literature and my own analysis in chapter 3 are consistent with this assessment of the regime and expenditure approaches,<sup>1</sup> but other operationalisations receive less criticism. However, popular indicators such as replacement rates may be affected by similar problems—albeit to a lesser extent. For instance, addressing average benefits (e.g. in the case of replacement rates) may not adequately reflect the situations of different income groups (Kvist et al. 2013: 329).

The use of specific concepts rather than broad conceptualisations of welfare stateness helps to provide a frame for discussions on clarity. Like in the case of comparability, it reduces the complexity of potential operationalisations by limiting them to those relevant for the addressed mechanisms. The main agenda

<sup>1</sup> Here, I would like to refer to the previous chapter. In several strands of literature, such as literature on social policies and health inequality, regime and expenditure approaches are considered inadequate (Ferrarini et al. 2014a; Lundberg et al. 2015).

here is to find indicators, which meaningfully and comprehensibly capture the specific aspect of social policy-making relevant to test respective hypotheses.

*Availability* is a rather pragmatic criterion and it is closely tied to data sources. As such, it is the most strongly affected by a need for compromise. As previously discussed, availability of indicators is often limited to country samples and periods. When selecting indicators, it is thus important to make sure that operationalisations can be replicated as easily as possible. This entails compromise, because focussing on accessible indicators may restrict the selection. While it cannot be ruled out that even better measurements than the ones available exist, the advantage of relying on easily accessible data sources is its high comparability with other research using the same sources and indicators. This issue will be discussed in more detail in a later section of this chapter, where data sources are explored (cf. Section 5.4). For now, besides comparability and clarity, the availability of indicators ought to guide the selection of meaningful operationalisations of welfare stateness in a given systematised concept.

The three criteria discussed so far, highlight important issues, which cannot be solved entirely at this point. Instead, they require compromise at first and more extensive conceptual and empirical work in the long run. Hereinafter, I am going to substantiate the required proceeding further and I will illustrate its potential application in the next chapter. The fact that compromises are necessary (to meet the three criteria) should in no way discourage further research into improved measurements to separate the wheat from the chaff (in the words of Bambra 2007). After all, at this point of the discussion there seems to be evidence for quite a bit of chaff out there. Applying these three criteria, the use of typologies and overarching composite measures should be ruled out in many cases. The same could be said for measures of social expenditure. What remains and calls for a more detailed discussion are other single indicators that—precisely selected can do exactly what is intended: addressing distinct perspectives on the welfare state and clear-cut concepts within the broad spectrum of welfare stateness to test hypotheses about the impact of specific social policies on individuals.

#### **5.2.2 Determining Measurement Validity**

Once indicators have been selected, it must be clarified whether they meet the criteria for measurement validity. Adcock & Collier (2001: 538–543) highlight three types of measurement validation: content validation, convergent (or discriminant) validation and nomological (or construct) validation, which are summarised in Table 5.1. The table also points out the main disadvantages of each approach that need to be taken into account.


**Table 5.1** Types of measurement validity

*Summary based on Adcock & Collier (2001: 538–543)*

The first type of validation (content) occurs on a conceptual level. It requires a certain amount of agreement on the nature of the examined phenomenon and involves conceptual reasoning. In contrast, the other two types (convergent and nomological) address empirical aspects and applications. In line with Adcock and Collier, I believe that distinguishing between these types of validation helps to assess which indicators are a better or worse fit for a concept. Moreover, the distinction between conceptual and empirical issues is particularly important in the specific case of welfare state indicators.

After discussing criteria that potential indicators should meet and the types of validity against which they should be measured, the following section is devoted to identifying concrete indicators for each of the four systematised concepts.

#### **5.3 Measurement of the Systematised Concepts**

While the dependent variables assumed to be influenced by social policies are diverse, the underlying assumptions are not. In the course of this chapter, four systematised concepts of welfare stateness have been highlighted: the *Responsive*, *Enabling*, *Normative* and *Assessed* Welfare State. While the first three capture the top-down perspective, the latter represents the bottom-up perspective. These concepts help to narrow down the role of the welfare state for specific research objectives and hypotheses, thereby making its operationalisation simpler and more standardised. After the rather abstract discussion so far, concrete indicators for the individual concepts are mentioned below. Since I am only approaching a solution for the independent variable problem discussed in this book, these recommendations are phrased relative to possible alternatives. A deterministic proposal of operationalisations seems inadvisable for two reasons. First, I do not expect my proposal to be an exhaustive solution to all problems related to the use of the welfare state as an independent variable. Rather, it is intended as a blueprint for a possible procedure. Secondly, some of the recommendations cannot (yet) be based on empirical evidence. Rather, they represent expectations arising from the preceding discussions. I will empirically test some of them in the next chapter to determine their validity.

Even before going into the individual concepts in more detail and proposing indicators, two very specific recommendations can already be made at this point, as they relate to issues that have arisen repeatedly throughout this book. One addresses the shortcomings of composite approaches, the other those of expenditure approaches. Neither seems to be well suited as independent variable.

Recommendation 1: The use of single indicators instead of composite measures such as regime typologies and indices allows a more targeted operationalisation of systematised concepts.

This recommendation is supported by the discussed lack of clarity, comparability and applicability of composite measures. If one assumes that a certain area of the welfare state is effective and follows a certain mechanism, composite measures cannot be used to determine which part of the multidimensional operationalisation is responsible for the observed effects. Drawing conclusions about mechanisms from such measures is therefore highly problematic in most cases. This view is supported by several voices in the relevant literature, highlighting the importance of distinguishing between important aspects of welfare stateness in order to accurately assess effects (Palme 2006: 400; Kvist et al. 2013). It is crucial to add that there can be cases in which regime approaches are a feasible option, since the conceptualisation does not imply mechanisms addressing distinct policies. This can occur if a somewhat latent welfare culture is addressed—for instance in the case of the Normative Welfare State, which will be discussed later in this section.

The second recommendation addresses expenditure measures. Even though they possess the great advantage of being easily available and frequently updated, they are no panacea to the problem of operationalising welfare states as independent variables. While they represent the overall redistributive budget or the "concept of 'welfare resources'" according to Dahl & van der Wel (2013: 61), they only partly reveal information about which of their components is responsible for an observed effect. Therefore, their applicability and potential to reveal in detail how and why welfare states shape outcomes is limited. This view is supported by various scholars, raising concerns about the validity of expenditure data (e.g. Kvist et al. 2013).

Recommendation 2: Social expenditure should not be chosen as a targeted operationalisation of systematised concepts. Only hypotheses addressing the redistributive budget can be more adequately modelled by expenditure indicators than by others (e.g. institutional).

While these two recommendations address general issues, the following sections elaborate on a possible measurement for each of the four systematised concepts. Two general questions will guide the selection of indicators for each of the four concepts. (1) Which policy field is addressed in the hypotheses? (2) Is an immediate effect of policy arrangement assumed or one that entails some delay? These two issues provide an overarching frame, relevant to all concepts and they address important aspects of the operationalisation. The first issue relates foremost to the *policy area* covered by an indicator. While the general nature of an indicator is the same within each concept (e.g. eligibility, expenditure, universalism et cetera), the policy area determines which risk is highlighted (such as old age or sickness). The second issue adds a *temporal perspective*. In some cases, recent policies are relevant, while in others, long-term policy developments or policies at a particular point in the respondent's life are important. Since these two aspects apply in all cases, they will not be discussed separately for each concept of the welfare state. Instead, they are highlighted only when they are particularly relevant and are taken up again in the summary discussion at the end of this chapter.

#### **5.3.1 The Responsive Welfare State**

The Responsive Welfare State is arguably the most obvious systematised concept within the welfare state and is characterised by how it fulfils its essential task of preventing risk and meeting need. Furthermore, since reducing social inequality receives growing attention in political agendas, shaping social stratification can be partly attributed to such main tasks as well. As the preceding literature review and discussion reveals, the Responsive Welfare State is embedded most frequently in research examining how and why different welfare states show varying degrees of success in securing individuals and reducing social inequalities and inequalities in outcomes. Exemplary dependent variables, which are closely related to the Responsive Welfare State are various risks (such as poverty), as well as manifestations of well-being (such as health).

As discussed in the previous chapter, the literature dealing with methodological issues related to the selection of an independent variable in this context emphasises the role of the institutional approach (e.g. Ferrarini et al. 2014a). Particular attention is paid to indicators measuring eligibility, generosity of benefits and population coverage. These indicators capture the social rights perspective and address different areas within it. Indicators of eligibility include who is entitled to benefits, while generosity refers to the type of benefits and coverage measures how widespread access to such benefits is in a society. All of these issues refer to immediate responses to risk: who is covered, who is entitled and how much does a person at risk receive? In contrast to other aspects of welfare stateness such as redistributive budget or overall welfare culture, indicators of social rights represent features of welfare states that capture characteristic responses to risk.

Recommendation 3: The *Responsive Welfare State* should be operationalised using indicators for social rights (e.g. benefit generosity, insurance coverage, eligibility criteria).

It seems plausible to expect that, depending on the dependent variable and the research question, not all indicators of the social rights perspective are equally suitable. The previous chapters have also discussed evidence that some indicators produce unexpected results when others are not controlled for in the same model (e.g. in case of net replacement rates and benefit coverage). This will be taken into account in the empirical test of various potential indicators in the next chapter.

#### **5.3.2 The Enabling Welfare State**

Highlighting the mechanism of activation, the Enabling Welfare State is characterised by policies, which incentivise a specific behaviour. This can happen in an intended as well as in an unintended way—though the former is more relevant in this case than the latter. Indicators capturing this conceptualisation of welfare stateness should thus highlight policies, which are intended to incentivise a specific outcome. Operationalising the Enabling Welfare State in the field of family policies presents a very tangible example for such intended incentives. Here the main question is how strongly equality is emphasised when it comes to the relationship between care responsibilities, housework and employment. In this context, the Enabling Welfare State can thus be operationalised by features tied to incentives for equal division of labour and female labour market participation. For instance, this includes public childcare provision (Brewer & Shaw 2018).

The case of enabling in the field of employment is similarly palpable. Here, active labour market policies are the most obvious choice. However, their measurement is not as obvious. Apart from expenditure in the area of active labour market policy, there do not seem to be many alternative indicators. The same can be said about other indicators of social investment. As mentioned before, the empirical operationalisation of social investment is an ongoing issue.

Regardless of the emphasised policy field, indicators capturing the Enabling Welfare State should model an *activating* component. This can relate to a specific field (such as childcare provision or ALMP) or can refer to social investment in a more general sense.

Recommendation 4: The *Enabling Welfare State* should be operationalised with the help of indicators for efforts to activate and provide incentives (e.g. active labour market policies).

As long as it is assumed that the indicators capture enabling components of the welfare state, they are suitable for measuring the Enabling Welfare State.

#### **5.3.3 The Normative Welfare State**

While responsiveness and activation are linked to specific policy goals such as risk reduction and labour market integration, the normative welfare state is characterised by its potential to socialise the individual, which is more subtle. There are only a few examples where conveying justice principles and egalitarianism are seen as actual policy agenda. Indeed, welfare states can create normative environmental pressure on companies, leading (for instance) to more work-family support within them (Beham et al. 2014: 34–35). Such pressure can be politically stirred and such active shaping of normative frames by political actors may be found in other examples as well. Still, in the majority of cases—including those related to attitudes formation—the socialising influence of welfare states is more likely to be unintentional. This makes the proposal of indicators more difficult. The Normative Welfare State has to be captured through policy measures, which are tied to dominant justice principles, like egalitarianism—without this being necessarily part of political agendas. In principle, various kinds of social policies can thus have a socialising effect, since they all shape what is perceived as legitimised and institutionalised norm and serves as a reference frame for individuals. Narrowing down specific indicators therefore means identifying which policies are particularly tied to cultural principles.

According to the literature discussed, policies that determine the principles of stratification and eligibility are particularly relevant. Thus, the degree of equality and universalism in benefit access and provision appears to be especially important. This excludes, for example, the use of social expenditure, as it may reflect the welfare commitment, but does not reveal principles guiding social policy-making. In contrast, it is however worth discussing, whether the regime approach, which has received criticism so far, is somewhat suited in this particular case. Given that many typologies capture the overall design of welfare states and their dominant characteristics, they come closest to what is sometimes referred to as different "welfare cultures" (e.g. Vrooman 2013). While this does not help to overcome the problem of the approach's lack of clarity, this fuzziness does not appear to be as problematic in this case as in other cases where mechanisms address clear implications of policy-making (such as lowering specific risks). This should be kept in mind and it seems worth discussing, whether indicators of universal access (e.g. benefit coverage) are better suited as proxies for a normative framework than typologies. Regardless, the recommendation at this point is to focus more on indicators of organisational principles than expenditure or other indicators of budget size.

Recommendation 5: The *Normative Welfare State* should be operationalised using indicators linked to principles that guide the provision of social services, such as universalism or egalitarianism (e.g. coverage, eligibility criteria).

In this recommendation, welfare regime approaches are intentionally left aside. As mentioned, it may be worth discussing, whether they could aid in capturing the Normative Welfare State. Since however, I focus particularly on defining characteristics of each systematised concept of welfare stateness and intend to compare different operational choices, using typologies does not fit the agenda of this project.

#### **5.3.4 The Assessed Welfare State**

While the three conceptualisations discussed so far follow a top-down perspective, the Assessed Welfare State represents a bottom-up approach. It particularly emphasises individual *evaluation* and *self*-*interest* as mechanisms. Even though self-interest implies a (cost-benefit) evaluation of what is perceived to be the social policy arrangement in a relevant context, not all hypotheses addressing evaluation, necessarily include a rational-choice argument for the link between evaluation and outcome. As discussed previously, other determinants—such as value orientations—can be more relevant than cost-benefit considerations. Thus, evaluation and self-interest represent different mechanisms within the Assessed Welfare State.

Various dependent variables are closely related to the Assessed Welfare State. Attitudes and behaviours are the most obvious ones discussed in the previous chapters. In addition, subjective manifestations of well-being (such as happiness) may entail this kind of conceptualisation as well.

The operationalisation of the Assessed Welfare State is more difficult than the previous types and the literature offers less guidance in this case. The most challenging aspect is that evaluation requires a certain amount of knowledge about relevant processes. However, as previously discussed (cf. chapter 4), citizens' knowledge about spending and complex regulations appears to be rather unreliable. It thus seems very important to choose indicators that represent aspects that are likely to be familiar to most citizens. Literature treating the welfare state as an independent variable does not offer much insight in this case. We can still derive some hints from the mechanisms assumed to be in effect. For instance, if self-interest (or even rational choice) is assumed to be responsible for a given outcome (e.g. a specific attitude or behaviour), indicators that capture potential benefits appear more relevant, as it seems likely that there is at least vague knowledge about the average amount of replacement benefit one is entitled to and about the eligibility criteria. The same applies to explanations that do not rely on selfinterest but on other aspects guiding evaluation of social policies. Salient issues, which are measured up against individual ideology, value orientation and other determinants of the formation of attitudes and behaviours, are likely those that are the most tangible for people who are not themselves beneficiaries of social security measures. It seems plausible to expect that this again relates to replacement rates and eligibility criteria, as most adult citizens should have at least some knowledge about what they could expect in case of need.

In contrast, indicators of procedural aspects—such as waiting days or benefit duration—are unlikely to be salient for individuals who have never been in any contact with social security measures. These considerations are derived from the previous discussions, but they lack empirical evidence.Testing citizens' knowledge about specific policy aspects beyond expenditure or very particular examples (such as unemployment) and on a comparative scale seems to be a desideratum, which is worth exploring in detail in the future. Determining to what extent the indicators commonly used in comparative welfare state research correspond to individual knowledge would greatly facilitate the operationalisation of the Assessed Welfare State. For now, two recommendations can be stated:

Recommendation 6: If the mechanism highlights the individual evaluation of performance, the *Assessed Welfare State* should be operationalised using indicators representing salient evidence of the performance of a welfare state.

Recommendation 7: If the mechanism entails an activation of self-interest, the *Assessed Welfare State* should be operationalised using indicators representing salient potential benefits for the individual and their generosity.

While both recommendations take up the premise of salience, the latter relates especially to those research questions and hypotheses that highlight the mechanism of self-interest. Here, potential benefit seems particularly relevant. Additionally, it may also be worth exploring whether an operationalisation of costs could also provide valuable insights. Again, this goes beyond the issues examined in this book—yet individual (private) contributions to insurance schemes, for example, could be a measure of the perceived costs of the welfare state.

#### **5.4 Data Sources**

Four different conceptualisations of the welfare state emerge from the discussion in the previous sections of this chapter, which are used to identify possible empirical operationalisations for each concept. In addition, several criteria are reviewed that should be met by such operationalisations. These criteria include *clarity*, *availability* and *comparability*. Before turning to an empirical test and illustration of the proposed framework, I would like to take a closer look at the possible data sources that can be used for the indicators discussed. Although there are many sources of comparative data on social policies, many of them require some kind of compromise because at least one of the above criteria is not met. Therefore, this section briefly presents a selection of particularly useful data sources, including a brief assessment of their respective advantages and disadvantages.

Among the most commonly used datasets are those provided by international organisations such as the OECD and the World Bank, or by official statistical agencies such as Eurostat. These sources provide information on popular variables such as social spending (e.g. European Commission 2016; Adema & Fron 2019).

In addition to such sources of specific indicators, there are also comprehensive comparative datasets that contain a variety of variables on welfare states. The Comparative Welfare Entitlement Dataset (CWED, Scruggs et al. 2014), the Social Citizenship Indicator Project (SCIP, SPIN 2017) and the Social Insurance Entitlement Dataset (SIED, SPIN 2019a) were mentioned and/or used previously in this contribution. Other datasets with very general information on welfare policies include the Quality of Government—Social Policy Dataset (QoG, Teorell et al. 2019). These different sources are discussed and compared in detail for instance by Grünewald (2014), and they all have advantages and disadvantages. Some are limited in the size of their country sample, others in terms of years for which data are provided. Furthermore, the selection of indicators varies. Still, the listed examples share that they all emphasise the "old risks". Thus, they are focussed mainly on decommodification and unemployment, pensions, sickness and disability policies. Family policies or social investment are not widely covered. Information on such 'newer' risks and policies are included in more specified datasets. For instance, the Multilinks Database (Keck & Saraceno 2012) offers information on family policies; so does the Child Benefit Dataset (CBD, SPIN 2019b).

One could add many other examples to this list but I would like to refer to existing comprehensive overviews instead (e.g. Grünewald 2014; Lohmann & Zagel 2018). For the purpose of this contribution, it is mainly important that information on many aspects of welfare stateness does exist—even though it is scattered over different sources and not always available for all countries and periods. One additional deficit has to be noted: information on social investment is indeed scarce (Andersson 2018)—at least when aiming at an operationalisation that goes beyond social expenditure.

At least two aspects need to be considered when choosing data. First, there are a number of sources from which to choose. The available indicators, the country sample, and the years for which data are available should guide the selection. Second, especially the last aspect (the reference year or period) requires additional attention. Here we may have to take into account a certain time lag in the impact of social policy on a variety of issues. Especially drastic changes in social policy-making—for instance as a result of external shocks such as the 'Eurocrisis' or reforms such as the 'Agenda 2010' in Germany—will not immediately impact outcomes. This refers especially to hypotheses referring to individual perception and evaluation of social policies (*the Assessed Welfare State*) as well as socialisation (*the Normative Welfare State*), where policy changes have to sink into the fabric of societies first before they will shape individual outcomes. This raises entirely different issues regarding the age (difference) of macro and micro data depending on hypotheses. Although this problem is not the focus of this project, it will be kept in mind in the next chapter, where the conceptual considerations of this chapter will be discussed and empirically implemented.

#### **5.5 Discussion**

In this chapter, I set out to explore how to improve and standardise the measurement of social policy arrangements as independent variables in multilevel analytical frameworks. This is done by deriving different conceptualisations of the welfare state based on the existing literature and its demand. This introduces an intermediate step to narrow down different perspectives before an empirical operationalisation takes place. This proposal is supposed to serve as an input to a—hopefully—larger debate on how to standardise the selection of operationalisations of welfare stateness in multilevel frameworks. At this point it is important to point out again that there are probably other conceptualisations in the literature that are not covered in this book. Should the approach I have outlined in this project prove useful, it would be necessary to further test the proposed conceptualisations to explore possible sub-variants and additional types.

In summary, I argue that in order to standardise the selection of indicators, to improve the comparability of the methods and to increase the robustness of the empirical results, the explanatory concept—the welfare state—must first be considered in more detail. By determining conceptualisations of welfare stateness embedded in the mechanisms highlighted by research objectives and hypotheses, we fill in a piece missing so far. The main concern of this project is to show that the lack of specification of the perspective on the welfare state when it is studied as an independent variable results in empirical operationalisations that lack theoretical guidance. As a result, approaches used in the relevant literature are too heterogeneous (cf. chapter 3) and lack comparability. By narrowing down the perspective on the welfare state, the selection of indicators can be standardised within explicitly defined research objectives. Furthermore, combining several conceptualisations may even enrich debates by adding new perspectives.


**Table 5.2** Guiding questions

In the course of this chapter, four systematised concepts of welfare stateness were discussed in detail, derived from different strands of literature. They highlight specific perspectives on the welfare state—all of which are supported by previous work on the matter, albeit with varying specificity.<sup>2</sup> For each of these

<sup>2</sup> For instance, while the *enabling state* is a long-standing term, which is used in a related manner in this contribution, other concepts such as the normative welfare state were addressed in many contributions—especially those dealing with distinct welfare cultures but (at least to my knowledge) never explicitly conceptualised in the sense proposed here.

concepts, I have discussed possible sets of indicators which—from a theoretical point of view—seem more suitable for an empirical test than others. In the selection, I tried to meet the criteria of comparability, clarity and availability as much and with as little compromise as possible. Still, there is certainly room for improvement in many cases. For the different steps and considerations necessary to move from the research question to the systematisation of concepts to specific indicators, it is helpful to ask several questions (cf. Table 5.2). The entire selection process is summarised in Figure 5.3.

**Figure 5.3** Conceptualising and operationalising welfare stateness (proposal)

In order to test the applicability of the proposal in detail, two exemplary analyses are carried out in the following chapter, which on the one hand illustrate the practical implementation of the proposed framework for the selection of indicators and on the other hand help to assess the measurement validity.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **6 Welfare Stateness as an Explanatory Variable: Empirical Illustration**

While the previous chapters explored ways to conceptualise the welfare state as an independent variable, the following part will deal with the practical applications of the proposed framework. Several objectives are pursued. First, I am going to illustrate how embedding research questions and hypotheses in the discussed systematised concepts of welfare stateness may be used in order to choose conceptual as well as empirical approaches for a given research objective. Second, I use the following exemplary analyses to determine—as far as possible—the validity of the proposed measurement. Third, I hope to contribute to the state of research in two subject areas where the welfare state is considered an important explanatory factor. In this case, *welfare attitudes* and the *risk of poverty* serve as exemplary dependent variables. These two topics tie in with the previously discussed strands of literature that have proven to be important and popular in the relevant literature. Furthermore, they represent both elementary perspectives that were identified in the previous chapters: while explanations for differing poverty levels conceptualise the welfare state as an institution influencing individuals topdown, individual attitudes are explained using both perspectives. Thus, they are expected to result from bottom-up perceptions of welfare stateness and to be shaped by welfare state arrangements top-down at the same time. They should therefore serve as adequate examples for the differences between the two perspectives and—in a more fundamental sense—provide a test of their usefulness. Even though these two examples only cover singled-out issues and it seems an overstatement to expect an exhaustive test of the proposed framework from the following analyses, I expect to be able to test the possibilites and limits of my

**Supplementary Information** The online version contains supplementary material available at https://doi.org/10.1007/978-3-658-39422-6\_6.

proposal, at least in part. This refers to the proposed selection framework in general and the concrete recommendations for the operationalisation of specific systematised concepts of welfare stateness in particular (cf. chapter 5).

The chapter presents analyses of the two exemplary topics that follow all steps usually taken when working on a research question—albeit in some passages in a slightly shortened way suitable for this demonstration because many general theoretical and conceptual premises for both topics were already discussed previously. Apart from that, hypotheses, mechanisms, conceptualisation, and operationalisation follow the proposed path and are tested in analyses using standard methodological approaches for the analysis of multilevel comparative data.

#### **6.1 Remarks on the Following Analyses**

Some remarks have to be made before starting with the analysis of the first exemplary dependent variable. This concerns general premises of the following analyses. In both cases, I chose to rely on the newest available data—when it comes to macro- as well as micro-level. Furthermore, I made sure that both analyses rest on data collected during a similar period—even though they stem from different datasets. This is intended to allow the best possible comparison of the applicability of indicators as it ensures that the country-level data on welfare stateness are identical—even though the individual survey data differ. Since the dependent variables require different methodological approaches, I will elaborate on the method of analysis applied in the individual chapters for each outcome. For now, I will only introduce the different data sources used in the analyses as well as the indicators of welfare stateness that will be tested.

#### **6.1.1 Data Sources for Individual-Level Survey Data**

The following analyses are based mainly on two data sources for individuallevel data. The selection follows the objective to model both exemplary research questions as adequately as possible. This means that recent data, covering most countries in the relevant sample1 and allowing a satisfying operationalisation of the dependent variables, are selected.

<sup>1</sup> Analogously to chapter 3, EU-27 plus Norway and Switzerland are chosen. Croatia is left aside in the analyses because macro-data availability for the relevant period is unfortunately limited.

The micro-level data in the analyses of the risk of poverty (section 6.2) stem from the European Union Statistics on Income and Living Conditions (EU-SILC) provided by the European Commission. EU-SILC is the main reference for comparative information on income distribution and social exclusion in the 28 European member states2 and 5 additional countries (Republic of Macedonia, Iceland, Turkey, Norway, and Switzerland) (European Commission 2017: 13). The analysis focusses on the EU-27 countries as well as Norway and Switzerland. In order to maximise comparability with the other analyses, data collected in 2016 were chosen. The dataset has a variety of advantages, which justify why it was preferred over other comparative survey data including similar information on income such as the European Social Survey (ESS) or the International Social Survey Programme (ISSP) used in chapter 3. First, it covers an exhaustive sample of all member states of the European Union. The number of observed countries is thus much higher and does not rely on an arbitrary selection, which makes it more meaningful compared to other datasets. Second, it offers high-quality data as it is an important source of information and policy-making within the EU. Third, the information on personal and household income—which is normally affected by non-response or other biases—is quite comprehensive. Since poverty is determined based on income, the high coverage of income information in EU-SILC is advantageous.3

In the case of the analysis of welfare attitudes (section 6.3), the eighth wave of the European Social Survey (ESS 2016) is very fitting. This dataset was collected in 2016 and early 2017. It covers 20 countries from the relevant sample4 and includes a comprehensive module on attitudes towards the welfare state, covering various facets of the subject, including welfare chauvinism, opinions about state responsibility for the provision of social services and benefits, and general attitudes about the welfare state, its purpose, and effectiveness. While the relatively small number of relevant countries is surely a downside of the dataset especially for applying multilevel methods—the differentiated measurement of welfare attitudes is a great advantage.

In both analyses, I am going to refer sporadically to additional analyses that were performed in instances where I felt that it was necessary to explore the robustness of results. Furthermore, in the case of welfare attitudes, results are

<sup>2</sup> This includes the United Kingdom, which was still part of the European Union in 2019 when these analyses were conducted.

<sup>3</sup> This high coverage is not only due to very thorough sampling and surveying but also to statistical imputation performed prior to the data release (cf. European Commission 2017: 55–56).

<sup>4</sup> 18 EU member states plus Norway and Switzerland.

compared to those obtained in chapter 3 using earlier data from the ESS and ISSP.5

### **6.1.2 Data Sources for Macro-Level Data and Country Sample**

Several sources are used in order to obtain data on welfare stateness and other features of the observed countries. Following the criterion of *availability* and the agenda of increasing the reproducibility of results, only publicly available data are chosen. The datasets correspond to those used in previous analyses in this book (cf. chapter 3) and are among those discussed at the end of the previous chapter.

The majority of indicators stem from the Social Insurance Entitlement Dataset (SIED). This source, which represents the continuation of the SCIP (Korpi & Palme 2008), was referred to in detail throughout this book. In short, it offers very comprehensive information on social rights in 34 countries, including all member states of the European Union (with the exception of Croatia) and some adjacent countries. While my previous analyses relied on the 2005 version of this dataset, the following analyses will be based on the newest available version at this time, which was published in early 2019 and includes data from 2015 (SPIN 2019a).

Additional information is obtained from Eurostat and in particular the European system of integrated social protection statistics (ESSPROS) dataset (cf. European Commission 2016) and the Organisation for Economic Co-operation and Development (OECD 2019). These sources offer detailed information on social expenditure, as well as additional variables on the level of countries such as gross domestic product (GDP) and unemployment rates. Again, data from 2015 were chosen to match the indicators of welfare stateness best.

Due to the missing data on many indicators of welfare stateness in Croatia, it is excluded from the analysis. Apart from that, the macro-level data allow the analysis of a very comprehensive sample of countries consisting of a full coverage of the former EU-27 countries, Switzerland, and Norway as adjacent countries with strong political and economic ties to the European Union.

<sup>5</sup> Those analyses were performed with data from 2008 (cf. chapter 3). I chose not to update the analyses when new macro-level data was published in 2019, because I believe that comparing the periods allows to draw some interesting additional insights and will be valuable to determine the robustness of some of the findings over time.

#### **6.1.3 Indicators of Welfare Stateness**

The following analyses will take up and compare a wide bundle of indicators. While the operationalisations chosen in each example will follow theoretical and conceptual premises as outlined in the conceptual framework, I will introduce the full set of indicators at this point already to illustrate the range of available information.

Furthermore—in the sense of *comparability*—the selection of indicators is guided by their previous application as independent variable. I hope that the following analyses not only lead to new insights but also measure up to existing evidence that has been generated using similar or even identical indicators.

As mentioned when introducing the datasets, I rely on information from 2015 in case of all macro-level variables. Thus, the individual-level survey data (which stems mostly from 2016) is delayed by one year. This is done intentionally, as it seems unlikely that policies manifest in individual reality instantly. Rather, it is to be expected that the outcomes of a specific policy reach the individual with some delay. It may be debatable whether one year is too short an interval or whether the average value for several previous years should be chosen to ensure that no outliers disturb the information. Such considerations seem useful, but they will not be at the core of my analyses, as I focus foremost on the indicators themselves.

Overall, welfare stateness is covered by five blocks of indicators, which stem from the above-mentioned sources and should allow to model at least the majority of systematised concepts. In the following, the distribution of the chosen indicators is briefly introduced. For a comprehensive list and additional information, the appendix can be consulted.6

The first block of indicators includes *net replacement rates* (NRR) in four scenarios of need: unemployment, old age, sickness, and accident. It would have been possible to add an income replacement rate for parental leave as well (cf. Saraceno & Keck 2011). However, as the chosen dependent variables are more closely related to the old risks, I remain within the more traditional areas of social policy-making for the time being. In all four cases, the replacement rates stem from the Social Insurance Entitlement Dataset (SIED) and represent combined information. The net replacement rate in case of unemployment refers to the average production worker and incorporates the mean benefit obtained by a single person a) during the first week and b) after 26 weeks of receipt. A single person is chosen instead of another type of model family because a family may receive

<sup>6</sup> Cf. Table A6.1–1.

other benefits that interfere with the clarity of the measure (for a similar reasoning cf. Ferrarini et al. 2014b: 658). The net replacement rates for those unable to work because of sickness or accident are based on the same information. Since pensions do not vary depending on how long they are obtained, the net replacement rate again refers to the average production worker and is calculated for a single person but without the combination of different times of benefit receipt.

**Figure 6.1** Net replacement rates 2015. *(Data: SPIN (2019a))*

A descriptive overview of the differences between the social policy areas and the 29 countries in the analysis is provided in Figure 6.1. It reveals strong variation not only between policy areas but also between countries. Overall, replacement rates in case of sickness or accident tend to be higher in most countries than replacement rates for unemployed or retired citizens. In addition, the Anglo-Saxon countries (Ireland and the United Kingdom) are clearly among the less generous ones, while the more generous countries—such as Luxembourg and Portugal—do not seem to reveal a clear pattern. Since the aim of this contribution is not to cluster countries, this is not consequential for my analyses. The strong heterogeneity in replacement rates remains remarkable, as it seems to support the argument that welfare state typologies may unduly reduce the existing complexity of the matter.

As argued before, net replacement rates are often considered a measure of benefit generosity. As such, they may be relevant for the operationalisation of several systematised concepts. By signalling how much an average worker is entitled to, they seem suitable for operationalising the *Responsive Welfare State* (cf. section 5.3.1). However, it is also plausible to expect that the generosity of potential benefits one is entitled to is a very salient feature of social policy-making. As previously argued (cf. section 5.3.4) this would mean that net replacement rates could also be suitable to operationalise the *Assessed Welfare State.* Lastly, it can also be argued that comprehensive benefits signal an underlying welfare culture and thus fit the concept of the *Normative Welfare State*. It is not clear at this point, which of the concepts can best be operationalised by NRRs, and it needs to be observed if the results obtained in the exemplary analyses contribute to shedding light on the issue.

The second block of indicators addresses *insurance coverage*. As before, the four old risks are differentiated. Both, accident insurance and unemployment insurance, refer to social policy instruments that are clearly linked to employment. Therefore, both indicators refer to coverage as a proportion of the labour force. In contrast, since referring to insurances needed by every citizen, pension and sickness insurance coverage is measured as the insurance coverage as a proportion of the population.<sup>7</sup> An overview of insurance coverage is provided in Figure 6.2.

Overall, insurance coverage—regardless of the insurance scheme—is above 50 percent in most cases (exceptions can be found in Spain, Italy, Greece, and Romania). Since insurance coverage is very important for how well citizens in a country are potentially protected against risks, this indicator seems to suit the operationalisation of the *Responsive Welfare State*. As mentioned before (cf. section 5.3.3), it may also be worth discussing, whether insurance coverage also signals universal

<sup>7</sup> There is one important addition necessary, which relates to Greece. In the SIED data, there are no up-to-date information on health and accident insurance coverage in Greece. This may be due to the fact that the Greek healthcare system was in some turmoil in the wake of the financial crisis. As a result of the crisis, there has been a loss of health insurance coverage especially among long-term unemployed and self-employed citizens (OECD 2015: 10). This was met with several attempts to reform the healthcare system (WHO 2019, 2018). Since insurance coverage in 2015 is thus most likely distorted by a temporary sharp increase in the number of uninsured individuals, health and accident insurance coverage is missing in this case.

**Figure 6.2** Insurance coverage 2015. *(Data: SPIN (2019a))*

access. If so, it may also be a suitable operationalisation of the *Normative Welfare State*.

The third block of variables addresses the *contribution period* required to be eligible for benefit receipt. The contribution period indicates how quickly welfare states step in when needed and how small the barriers to obtaining benefits are. As such, they are closely related to the *Responsive Welfare State*. In addition, one could argue that the contribution period signals the strictness of eligibility criteria and thus the equality and universality of benefit access. In this sense, it may also fit the *Normative Welfare State.* Like before, different policy fields are differentiated. However, the contribution period required to be eligible for accident insurance is excluded from the analysis, as it has almost no variation.8 For the remaining benefit areas—unemployment, pension, and sickness—the contribution required is measured in weeks (cf. Figure 6.3). In several cases, countries

<sup>8</sup> Apart from Malta (1 week) and Estonia (2 weeks), none of the countries has any waiting period for being eligible for accident insurance.

exhibit very similar regulations. For instance, in almost half of the examined countries, there is immediate eligibility for sickness benefits. Similarly, countries seem to agree on how long contributions must be made in order to qualify for unemployment and pension benefits: in most cases, benefits can be obtained after one year of contributions. Only a few outliers (Lithuania and Slovakia) require a longer contribution period—and that is only in the case of unemployment benefits. Because of the uneven distribution of the variables, they are turned into dichotomous variables. This results in indicators capturing, whether the required contribution is zero in the case of sickness and lower than the median for unemployment and pensions.<sup>9</sup>

**Figure 6.3** Contribution period 2015. *(Data: SPIN (2019a))*

Another indicator that is closely linked to social rights and thus to the concept of the *Responsive Welfare State* is the *duration of benefit receipt*. In this fourth block of indicators, too, I distinguish between different types of benefits, but omit

<sup>9</sup> More information can be found in the appendix.

the duration of pensions, as there is no time limit on their receipt. Like in the case of the contribution period, benefit duration is measured in weeks. There is considerable variation in how long benefits can be obtained across schemes and countries.

**Figure 6.4** Benefit duration 2015. *(Data: SPIN (2019))*

Overall, the duration of unemployment benefits is the shortest with a median value of 51 weeks, while the median duration of accident benefits is 104 weeks (cf. Figure 6.4). In several countries, benefits can be obtained for 10 years or longer (especially in the case of accident insurance). However, values higher than that were all fixed at an upper threshold of 520 weeks in the SIED data. As a result of the extreme range of the variable and its uneven distribution with most benefit schemes being obtainable for up to 2 years and just a few exceptions with substantially longer benefit duration, the variables were again recoded into dichotomous indicators. The resulting variables capture whether a country offers benefit duration, which is higher than the median.

All of the above-mentioned indicators were taken from the SIED Dataset (SPIN 2019a). In addition, the fifth and last block of variables includes four indicators capturing *social expenditure* on Eurostat data (ESSPROS and European Commission 2018). Analogously to the other indicators, the different areas of social policy-making are distinguished by including spending in the fields of old age, health, and unemployment (cf. Figure 6.5).

**Figure 6.5** Social expenditure 2015. *(Data: European Commission (2016, 2018).*10*)*

Furthermore, total social expenditure is examined, which incorporates the sum of spending in the three above-mentioned areas and additional policy fields, which are of secondary importance for my analyses, such as child and family benefits and survivor benefits (European Commission 2016). While these four indicators represent the overall effort, an additional spending indicator is added to capture

<sup>10</sup> Since the data on social expenditure stem from another dataset than the information about spending on active labour market policies, some overlap between unemployment spending and labour market spending may exist. Interpreting the stacked graph as a sum of expenditure might thus be biased. Furthermore, data on ALMP spending in Switzerland are missing.

the commitment to labour market activation. More specifically, it captures expenditure in the field of training, employment incentives, supported employment and rehabilitation, direct job creation, and start-up incentives. All expenditure measures represent spending as percentage of the gross domestic product.

Throughout this book, concerns about using social expenditure have been voiced. Such indicators were described as unintelligible and fuzzy. For two reasons they are still included in the following analyses. First, they are by far the most popularly chosen single indicator in the literature. As such, it seems commendable to use them as a reference and to compare them to other strategies of operationalisation. Second, knowledge about expenditure may be biased, but it still represents a better known feature of welfare stateness compared to other alternatives. As such, indicators of expenditure may offer an operationalisation of the *Assessed Welfare State*. Third, since comprehensive operationalisations of social investment and active labour market policies are scarce, spending on ALMP is among the only available options of capturing the *Enabling Welfare State.*

Turning to the types of measurement validity discussed in the previous chapter, the only one that can be tested at this point is the convergent (or discriminant) validity. According to this criterion, indictors for the same systematised concept should correlate. A glance at the correlation matrix does reveal a mixed picture.11 While policy areas within one type of indicator tend to be correlated, there is also evidence for an association of different indicators within the same policy field (e.g. unemployment expenditure and unemployment replacement rates). Since there is, however, considerable overlap in possible indicators for different conceptualisations, clearly delineating groups of indicators based on correlations is impossible. This overlap has implications for the entire endeavour of this chapter, which will be discussed shortly.

The introduced five blocks of variables (cf. Table 6.1) were selected with the aim of best fulfilling the three criteria set out in the previous chapter. They are *comparable* in that they originate from transparent sources and for the most part were already used in the relevant literature.<sup>12</sup> They are *available*, as information is easily and freely accessible for scientific use. Lastly, they at least approach *clarity* as*,* apart from social spending (and to some extent also net replacement rates), the selected variables can be assigned to one or a few systematised concepts. In

<sup>11</sup> Due to the low number of countries (N <sup>=</sup> 29), strength of effects is interpreted as a tendency even if statistical significance is not given (cf. Table A6.1–2).

<sup>12</sup> Benefit duration and contribution period have been used as descriptive indicators (Kuitto 2018) and in composite measures and typologies (Esping-Andersen 1990), but—to the best of my knowledge—they are not commonly used as single indicators.

addition, they capture clearly distinguishable elements of social policy-making and differentiate the various areas within the old risks.


**Table 6.1** Indicators of welfare stateness—summary

However, the chosen selection also entails limitations. As briefly mentioned, information on the whole area of child and family policies is not taken into account. The only indicator including such benefits and services is the total social expenditure. By omitting measures in the areas of child and family policies, I rely on a selection that matches the chosen exemplary dependent variables. Another restriction is that only one operationalisation of active labour market policies and social investment is included, which means that the entire concept of the *Enabling Welfare State* is only scarcely covered in the following analyses. For several reasons, both restrictions seem acceptable for the purpose of this project. On the one hand, the analyses are reduced to the old risks, as they are easiest to compare with previous studies. Moreover, the availability of indicators, especially for activating measures (ALMP and social investment), is very low and the discussion on how to operationalise these measures is still ongoing and comparatively new (as repeatedly noted in the previous chapters). This by no means implies that the policies that were left out are incompatible or of subordinate importance. Rather, they may provide valuable insights depending on the chosen explanandum. However, for the two dependent variables this chapter focusses on, the old risks promise to be sufficiently fitting for an illustration of the proposed framework.

While these limitations do not directly affect the results and the way in which the goal of this book is achieved, there is one issue that could have serious implications: there is considerable overlap when it comes to which indicators are suitable for which conceptualisation. This even happens between the two main perspectives (top-down and bottom-up). As a result, a clear-cut empirical distinction between the concepts is impossible. This is very regrettable, as the main advantage of using systematised conceptualisations is supposed to be their resulting in distinct operationalisations—at least if one envisions an ideal empirical test of the framework. However, arguing based on the existing indicators of welfare stateness (the ones used here and potential others), this ideal clear allocation of the one perfect indicator (or set of indicators) to *only one* concept is not feasible. If indicators are selected based on criteria such as the ones chosen in this chapter, some amount of overlap has to be taken into account—especially during this first application of the framework. Furthermore, the very nature of some of the conceptualisations—in particular the *Responsive Welfare State*—is to span different policy fields. For that reason, they simply cannot be reduced to one facet of policy-making.

In order to maximise the output from the following trial, several aspects are considered. First, the results are linked to the theoretical assumptions, which should help to assign the indicators and the results they produce to a specific conception. Second, if overlap cannot be resolved, it will be discussed if indicators should be ruled out as potential operationalisations. Third, it is acknowledged that some overlap is actually theoretically plausible, as—for instance—indicators representing responsiveness may be the same ones as those responsible for cost-benefit considerations. The fact that they can be interpreted in two ways may actually bear important additional insights. These and more issues will be discussed at the end of each analysis and in the discussion at the end of this chapter.

#### **6.2 The Welfare State and the Risk of Poverty**

Since responding to needs and risks is among the most basic functions of welfare states, the literature often asks whether social policies are responsible for differences in individual poverty risk across countries (cf. section 4.2.2). For this reason, the risk of poverty was selected as an exemplary topic for the following exemplary application of the proposed framework. In this section of the chapter, I will illustrate all proposed steps that lead from hypotheses and mechanisms to conceptualisations and lastly distinct operationalisations of welfare stateness.

#### **6.2.1 Conceptualisation of the Welfare State**

The previous chapter concluded with a list of questions that should guide the theoretical conceptualisation and empirical operationalisations of welfare stateness. The first question (w*hich perspective on the welfare state is chosen?*) can be answered quickly when explaining cross-national differences in the risk of poverty. Here, only the *top-down perspective* is relevant, since individual perception and evaluation of social policies in the sense of the *bottom-up perspective* are inconsequential. This is reflected in the mechanisms outlined in the literature and the hypotheses deduced from them. Turning to the second question (*which mechanisms are addressed?*) several mechanisms are relevant. As one of the main objectives of social policy is to protect individuals from risks and provide support though benefits and services when needed, *security* is a prominent mechanism that explains how and why welfare states achieve different levels of poverty. This happens through the *redistribution* of resources, where targeted and universal premises represent opposed logics (cf. the paradox of redistribution). In chapter 4, this was summarised in the following explanation:

• **Explanation 2.1:** Comprehensive social rights (including benefit coverage, generosity, and eligibility criteria) and redistributive budget decrease poverty because these actions *secure* against risks by *redistributing* resources to those who require them.

Phrased as hypotheses and distinguishing between the institutional perspective (social rights) and the expenditure perspective (redistributive budget) this means:

• *Hypothesis Pov1: The more comprehensive the social rights, the lower the prevalence of poverty in a country.*

• *Hypothesis Pov2: The higher the redistributive budget, the lower the prevalence of poverty in a country.*

In addition, moderating effects were discussed. Thus, *security* not only has a direct impact on the risk of poverty but also moderates the effects of individual determinants such as low educational attainment or unemployment. Such moderating effects are more pronounced, yet chapter 4 addresses additional mediated effects in which the welfare state shapes those individual determinants (*social stratification*), which in turn are responsible for the risk of poverty.<sup>13</sup> Since the latter is of secondary importance, the moderating effects of social policies are highlighted in the following exemplary analyses:


Lastly, *activation* is emphasised. Even though activating policies may have a direct effect on risks, the more relevant influence is a moderating one. The last explanation and hypothesis thus capture the potential to lower risk by equipping individuals in vulnerable situations with tools needed to avoid being at risk of poverty. This is again expected to manifest especially as a moderating effect:


At this point, it is irrelevant whether these five hypotheses fully cover all theoretical assumptions in the literature, as they cover a sufficiently broad spectrum of

<sup>13</sup> This was captured in **Explanation 2.3**: Comprehensive universal social rights decrease poverty, because these actions shape patterns of social *stratifications* in a way that increases equal opportunities, thereby lowering risks among those who are especially vulnerable.

the topic to allow for the discussion of several systematised concepts that might be relevant to the empirical investigation.

The third question (*which concepts of welfare stateness are addressed?*) is of particular importance. Based on the assumed mechanisms, two systematised concepts of welfare stateness are highlighted. The *Responsive Welfare State* captures those explanations emphasising security, social stratification, and redistribution, while the *Enabling Welfare State* captures the idea of activation and incentive. The other systematised concepts—the *Normative* and *Assessed Welfare State*—are not relevant in this case.

Two more questions have to be addressed before turning from theoretical conceptualisation to empirical operationalisation: *which policy fields are relevant* and *which temporal perspective has to be chosen?* Since the risk of poverty emanates from very different situations, a restriction to specific policy fields is not necessary. Even though the labour market is closely tied to poverty, other fields such as health and pension policies are equally important depending on which social groups (e.g. with regards to employment, educational status, or age) are highlighted in an analysis. In terms of the temporal perspective, it seems plausible that the risk of poverty is determined at all times by the prevailing contextual influences. The conceptualisation and operationalisation of the welfare state can therefore be based on a reference period similar to the survey data.

#### **6.2.2 Operationalisation of the Welfare State**

The previous section of this chapter reveals that two conceptualisations of welfare stateness are especially important when exploring the impact of social policies on the risk of poverty. The aim of this contribution is to explore how well such different conceptualisations can be operationalised empirically and whether this improves the informative value of results and discussions on the one hand and the transparency and comparability of operational approaches on the other hand. Following the recommendations in chapter 2, the two systematised concepts of welfare stateness should be grasped by the following sets of indicators:

The *Responsive Welfare State* should be operationalised using indicators of social rights (e.g. benefit generosity, coverage, eligibility criteria).

The *Enabling Welfare State* should be operationalised using indicators of efforts to activate and incentivise (e.g. active labour market policies).

At the beginning of this sixth chapter (cf. 6.1.3), specific indicators for each systematised concept were introduced. Following those recommendations, all but one set of indicators were discussed as potentially suitable for the operationalisation of the *Responsive Welfare State*. Only social expenditure was not added to the list of potential operationalisations, because of its lack of clarity. However, as spending continues to be widely used (e.g. capturing the redistributive budget) and spending on active labour market policy is the only available operationalisation of the *Enabling Welfare State,* this set of indicators is still included. The following analysis should help to find out which of these operationalisations depicts the conceptual premises more accurately and tests the hypotheses more precisely.

#### **6.2.3 Additional Variables and Analytical Strategy**

While poverty is the dependent variable in the following analyses, several independent variables are also added at the individual level. The inclusion of such micro-level determinants is particularly important as the hypotheses suggest that social policies not only directly influence poverty by reducing or increasing risk but also act as moderators. In this sense, the welfare state is expected to mitigate the risk posed by situations in which the individual is vulnerable—such as unemployment or low education. In the following, the operationalisation of the dependent and independent variables is briefly outlined as well as the analytical strategy.

#### *Dependent variable*

The dependent variable—being at risk of poverty—can be operationalised in various ways. In the literature, such different approaches have in common that they are usually based on information about the disposable household income, which is then equivalised in order to control for different household sizes and compositions. However, they differ when it comes to details. For instance, when constructing equivalised income, some authors rely on the so-called "old" OECD scale, while others choose the modified one or divide income by the square root of household members (e.g. Haughton & Khandker 2009; UNECE 2017). Furthermore, different cut-off points for determining whether a person is at risk of poverty are used. While some define it as having less than 50 percent of a country's median equivalised disposable income (e.g. Brady et al. 2017), others use 60 percent as a threshold (e.g. Polin & Raitano 2014). For my analyses, I will rely on the latter version, as it corresponds best to official approaches. This measure of poverty (using the modified OECD scale) is already included as a variable in the EU-SILC dataset.

#### *Independent variables*

Individual characteristics are needed to model moderated effects and to serve as control variables in those models that include a cross-level interaction term. Therefore, several socio-economic indicators are added as control variables.<sup>14</sup> These control variables include *employment status* as a predicator of economic well-being. Employment status is derived based on a variable on the self-defined economic status. The recoded variable captures whether a person is *unemployed*, *employed*, *self-employed*, *retired*, or *inactive for other reasons* such as being in school, military service, disabled, or fulfilling domestic or care responsibilities. In addition to indicators of economic status, *age*, *sex*, and *education* are included as sociodemographic control variables. Since EU-SILC only provides detailed information on year and month of the birth of respondents until 1930 and all older respondents are treated as if they were born in 1930, using age as a continuous variable seems inadvisable. Instead, seven age cohorts are constructed starting with respondents aged 29 and younger, continuing in 10-year steps and ending with respondents aged 80 and older. Sex is included as a binary predictor with women being the reference category.15 Lastly, information on the highest level of attained education is provided in the dataset following the ISCED-97 classification. This information is recoded into three groups of educational attainment. The first encompasses individuals who attained primary or lower secondary education or less (ISCED 0, 1, & 2). The second group includes individuals with (upper) secondary or post-secondary non-tertiary education (ISCED 3 & 4) and the highest group possesses tertiary education (ISCED 5 & 6).

#### *Analytical strategy*

The following analyses are based on a number of multilevel logistic regression models. This procedure deviates from the linear approach used in the third chapter, because the dependent variable (poverty) is dichotomous. The main idea behind the method is still the same. Following the approach described in chapter 1 from a conceptual point of view and later in chapter 3 in the case of (pseudo-) continuous dependent variables, the main advantage of multilevel modelling is

<sup>14</sup> A summary of all independent variables used in the analysis of the EU-SILC dataset is provided in Table A6.2–1 in the appendix.

<sup>15</sup> Since there were no missing values or additional categories in the original variable, there was no need to distinguish a third (or diverse) sex.

that it is able to estimate effects of individual- and country-level effects simultaneously. In the case of dichotomous variables, the outcome *yi j* for an individual *i* in a country *j* is either zero or one. Thus, the analytical approach focusses on the probability (*Pi j*) of the occurrence of an outcome (for a detailed description cp. Snijders & Bosker 2012: 293–295). In the full multilevel logistic regression model, the logarithmised probability of poverty occurring *logit*- *Pi j* is based on the average probability γ<sup>0</sup> (much like a grand mean in linear models), the effects of all individual-level variables *xi j* , country-level variables *z*<sup>0</sup> *<sup>j</sup>* , and a random group-dependent deviation in the intercept *u*<sup>0</sup> *<sup>j</sup>* . Since moderating effects of social policies are assumed to exist from a theoretical point of view, there is also the need to explore random slopes. Thus, the impact of at least one independent variable on the individual level is assumed to have different slopes in different countries, which adds the random effect *u*<sup>1</sup> *<sup>j</sup> x*<sup>1</sup> *<sup>j</sup>* . It is furthermore assumed that features of the welfare state can at least partly explain those different slopes, which means that a cross-level interaction effect is included (γ1*x*1*i jz*<sup>10</sup> *<sup>j</sup>*). This results in the following *Random-Intercept-Random-Slope-*model*:*

$$\begin{aligned} \operatorname{logit}\left(P(\text{risk of property})\_{ij}\right) &= \wp\_{00} + \sum\_{h=1}^{r} \wp\_{h} \chi\_{hij} + \sum\_{l=1}^{r} \wp\_{l} \underline{\chi}\_{l0j} \\ &+ \wp\_{1} \chi\_{1ij} \underline{\chi}\_{10j} + \mu\_{0j} + \mu\_{1j} \chi\_{1ij} \end{aligned}$$

Following the advice given by Heisig and Schaeffer (2019), the slope of the lower level variable in the cross-level interaction is explicitly defined as random in all models.16,17

Turning to the analysed population sample in the EU-SILC data, three different strategies can be found in cross-cultural analyses of poverty. The first focusses on in-work poverty and thus reduces the sample to employed individuals (e.g. Lohmann 2009; Halleröd et al. 2015). The second highlights poverty among the working-age population (e.g. Saltkjel & Malmberg-Heimonen 2017) and the third strategy explores poverty among the entire population (e.g. Watson et al. 2018). In this contribution, I focus on the second option—the working-age population

<sup>16</sup> Since cross-level interactions are included, it can furthermore be advisable to use grandmean centring to reduce multicollinearity (Dedrick et al. 2009). However, since there are no continuous micro-level variables in the models and centring of macro-level variables is not necessary, centring is not applied.

<sup>17</sup> A side note to support the caution advised by Heisig and Schaeffer (2019): if the slope is not explicitly set to random, the significances of the lower level variable and the interaction effect itself are severely overestimated.

(16–64)—for several reasons: first, reducing the sample to the "working poor" (e.g. Andress & Lohmann 2008) would exclude the effect of unemployment and thus leave out key situations of risk in which welfare states potentially intervene. Secondly, old-age poverty is excluded because it is based on different determinants than the poverty of those who could potentially participate in the labour market (for a review, see Kwan & Walsh 2018). Moreover, poverty among the retired population is presently much less pronounced than among the workingage population (Watson et al. 2018: 8–9). There is reason to believe that this will change in the future as pensions are expected to continue to decrease (Kwan & Walsh 2018: 1–2) and the threat of old-age poverty to rise (e.g. Ebbinghaus 2015). However, at the time of this analysis, poverty among the older population is still significantly lower. There is thus reason to suspect, that analysing the entire population would interfere with results.<sup>18</sup> Still, pension policies are included but they are only expected to be relevant as indicators of overall responsiveness of welfare systems. Apart from that, all other operationalisations of welfare stateness refer to the decommodifying effects of social policies. Besides this focus on the working-age population, the sample is further reduced to those cases, which have valid responses in all relevant variables in the analysis.

Lastly, weighting of data can be necessary. Several weights are provided in the EU-SILC data. They correct for non-response patterns, data shortcomings and adjust to household and population distribution in the target population. Depending on the unit of analysis, weights are given at household level and at person level (European Commission 2017: 33–45). Since the following analyses focus on the cross-sectional information for selected respondents, a weight for this specific population is chosen in descriptive analyses. In the multivariate analyses, weighted and unweighted analyses were tested. The unweighted results will be presented for three reasons: adjusting for the differences in the size of European populations (1) strongly reduces the presence of smaller countries, (2) severely complicated the estimation procedure, and (3) did not lead to noteworthy differences in the results.

<sup>18</sup> Still, additional analyses of the full population sample were conducted. They will only be discussed if they offer other or complementary findings.

#### **6.2.4 Results of Bi- and Multivariate Analyses**

It does not come as a surprise that poverty varies between the European countries and the share of population at risk of poverty by country (cf. Figure 6.6) corresponds to the numbers reported by the EU.<sup>19</sup>

**Figure 6.6** Risk of poverty in Europe. *(Data: EU-SILC (2016), weighted data)*

In addition to this visual confirmation of the existence of variation, the intraclass correlation coefficient (ICC) is an important indicator of how much of the variance between individuals can be attributed to the contextual level—in this case, the country.<sup>20</sup> The resulting ICC of roughly six percent may appear low at first sight, however, it is in line with the numbers obtained in comparable analyses of poverty in Europe (e.g. Lohmann 2009; Saltkjel & Malmberg-Heimonen 2017). Before turning to the question of whether welfare states explain some of this

<sup>19</sup> Deviations from official reports (cf. European Commission 2019) are due to the listwise deletion of missing values.

<sup>20</sup> The reported ICC stems from the *Random-Intercept-Only-Model* before predictors are added.

variance between countries, and whether differentiating between conceptualisations of welfare stateness helps to shed some light on the results, the micro-level determinants of poverty are briefly examined (cf. Figure 6.7).<sup>21</sup>

**Figure 6.7** Individual-level determinants of poverty. *(Data: EU-SILC (2016), coefplot based on multilevel logistic regression (melogit), odds-ratios, subsample (working age))*

The risk of poverty is distributed unequally among the population. Age exhibits a nonlinear effect. It increases the risk of poverty among those aged under 50 but decreases the risk among the highest age group compared to the youngest cohort. Furthermore, men do not exhibit a different risk than women, while lower educational attainment strongly increases the risk of poverty. The same effect can be found for all other employment status compared to regular employment. Unemployment increases the risk severely. This result—that poverty manifests especially among disadvantaged social groups—is in line with previous findings

<sup>21</sup> All analyses in this chapter are visualised in coefplots (Jann 2014), which allow a comparatively easy interpretation of results. More comprehensive tables of all estimated models and information on the model fit are provided in the appendix.

using similar data and a similar country-sample (e.g. Lohmann 2011; Ingensiep 2016; Brady et al. 2017).

In the following part of this chapter, the empirical application of distinct conceptualisations of welfare stateness will be explored. For this purpose, only the effects of social policy indicators will be reported. Still, all models include the micro-level determinants of poverty discussed above. Furthermore, all coefficients will be reported in two versions: with and without macro-level control variables. Those control variables are the GDP and the unemployment rate. While the results within sets of indicators are reported in one figure, they were all analysed in separate models.22

All five sets of indicators are tested: net replacement rates, benefit coverage, contribution period, benefit duration, and expenditure. As explained before, they are expected to represent operationalisations of two different analytical perspectives on the welfare state. The first is the *Responsive Welfare State*. According to the hypotheses, generous social rights compensate for deficits and combat poverty more effectively. The most relevant mechanism underlying this assumption is the provision of *security*. As discussed previously, measuring the Responsive Welfare State through social rights means that all sets of indicators represent potentially relevant operationalisations. The second relevant conceptualisation is the *Enabling Welfare State.* Here, *activation* was highlighted as an important mechanism. However, the operationalisation offers less alternatives. Therefore, only spending on active labour market policies represents a distinct operationalisation of this systematised concept.

In order to achieve a lean presentation of results, the analyses are reported in pairs of sets of indicators. Within the *Responsive Welfare State,* the first two sets—*replacement rates* and *insurance coverage*—relate to generosity of benefits*.*

The results reported in Figure 6.8 reveal that net replacement rates do not appear to reduce the risk of poverty significantly in my analyses. This partly contradicts the notion that generous benefits should decrease poverty (cf. section 4.2.2) and will be explored in the discussion, which succeeds this short description of main results. Insurance coverage, on the other hand, appears to reduce the risk of poverty in all policy fields with the exception of accident insurance. One may argue that especially the effect of pension insurance coverage is somewhat implausible considering that the analyses conducted are based on the working-age population. However, as the variables are expected to indicate the responsiveness of a welfare state in general, pension coverage may be a

<sup>22</sup> The individual models are reported in the appendix to this chapter.

general characteristic of the welfare state, that affects the working-age population as much as the retired population.23

**Figure 6.8** Poverty on replacement rates and insurance coverage. *(Data: EU-SILC (2016), coefplots based on multilevel logistic regression (melogit), odds-ratios, subsample (working age), analyses of sickness and accident insurance coverage (b) exclude Greece)*

The second set of indicators representing the *Responsive Welfare State* includes the *contribution period* and the *duration of benefit receipt* (Figure 6.9). While it seems highly plausible that such criteria of eligibility and the temporal comprehensiveness of benefits and services reduce risks, neither of the variables appears to reduce poverty systematically in this analysis. I will discuss why this might be the case in more detail later. For now, suffice it to say that both sets of indicators do not appear to contribute to an explanation for differing risks of poverty

<sup>23</sup> Tentative analyses of the full sample reveal comparable effect directions. However, they have a much greater error. This supports the argument that poverty in old age follows a different logic than poverty among the working-age population.

between individuals in different European countries—at least not if their impact is examined separately.<sup>24</sup>

**Figure 6.9** Poverty on contribution period and benefit duration. *(Data: EU-SILC (2016), coefplots based on multilevel logistic regression (melogit), odds-ratios, subsample (working age))*

As a last set of indicators, several measures of social expenditure are included. It was argued repeatedly throughout this book, that social expenditure is a potentially ambiguous indicator. Its lack of clarity renders it an undesirable candidate for a clear and comprehensive measurement of any of the theorised conceptualisations of welfare stateness. They are, however, frequently referred to in the relevant literature (cf. section 4.2.2) and as such were even considered with their own hypothesis (**Pov2**). Indeed, all expenditure items produce negative results—with the exception of unemployment expenditure, which is negative but insignificantly so (cf. Figure 6.10). Thus, the redistributive budget (or "welfare effort") reduces

<sup>24</sup> Both indicators are frequently used for the construction of composite indices and typologies (most prominently by Esping-Andersen 1990). However, they have—to the best of my knowledge—not been used as single indicators in the literature so far.

the risk of poverty in a country—however, it is not possible to attribute this effect clearly to one of the systematised conceptualisations of welfare stateness. It seems worth discussing whether other effects may be underlying the impact of expenditure on poverty. Furthermore, it is curious that of all policy fields, unemployment expenditure is the one with an insignificant impact on poverty—after all, individual unemployment is a strong predictor of being at risk of poverty. While unemployment expenditure tends to decrease poverty, this effect does not appear to be as systematic as the impact of other measures of social expenditure. I return to these issues later in the more detailed discussion of findings.

**Figure 6.10** Poverty on expenditure. *(Data: EU-SILC (2016), coefplot based on multilevel logistic regression (melogit), odds-ratios, subsample (working age))*

In contrast to the four expenditure indicators tied to more classical policy fields, one of the indicators is expected to capture a distinct conceptualisation: expenditure on active labour market policies is the only available indicator for measuring the *Enabling Welfare State* in this contribution. ALMP should reduce poverty by enabling individuals to re-enter the labour market or continue to participate even if they are in situations of vulnerability—such as unemployment or low educational attainment. Indeed, expenditure on ALMP decreases the risk of poverty significantly. This is generally in line with expectations. However, the mechanism of activation should be especially relevant when it comes to a moderating influence of welfare states on risks. Thus, the next step is to explore, whether ALMP spending actually captures the assumed reduction of risk emanating from situations of vulnerability.

**Figure 6.11** Moderating effects of ALMP spending. *(Data: EU-SILC (2016), marginsplot based on multilevel logistic regression (melogit) with cross-level interaction, subsample (working age), controlled for GDP and unemployment rate)*

For this purpose, several cross-level interaction effects are tested. If the effect of ALMP actually captures the assumed causality behind the *Enabling Welfare State*, it should significantly reduce the risk of poverty among vulnerable individuals—in particular, when they are unemployed or have a low educational status. Thus, cross-level interactions between ALMP spending on the national level and unemployment and low educational status (ISCED levels 0–2) on the individual level are tested (cf. Figure 6.11). In both cases, the risk of poverty is higher in those two vulnerable groups, but—in line with the expectation—ALMP spending reduces the risk more strongly in these groups than in the reference groups.<sup>25</sup>

When explaining different poverty risks between countries, the *Enabling Welfare State* is not the only conceptualisation assumed to be a relevant candidate for such a moderating effect. As discussed, especially the mechanism of providing *security* is considered to not only directly influence risks, but also mitigate the risk posed by vulnerable situations. Hence, the *Responsive Welfare State* can also manifest as a moderator here. Again, this is tested using cross-level interactions. While all sets of indicators with the exception of social expenditure were introduced as plausible candidates for this systematised conceptualisation, I only tested those combinations empirically that are theoretically plausible. Considering the sample is restricted to respondents in working-age, indicators representing pension policies are excluded from the reported interaction models.26 Furthermore, since individual health is not included in these models either, sickness and accident policies are equally implausible candidates for an interaction. This leaves especially the unemployment policy indicators.

As Figure 6.12 reveals, unemployment insurance coverage tends to lower the risk in both groups, but there is not much evidence for a moderating effect. In case of education, a slightly steeper reduction of risk among individuals with low educational attainment appears (even though this effect is not statistically significant). In case of individual unemployment, however, no notable moderating effect can be detected.

The same result can be found for the interaction effects of net replacement rates, benefit duration, and length of the contribution period required to qualify for unemployment benefits (all included only in the appendix).<sup>27</sup> All three indicators of welfare stateness produced no significant effect by themselves and the crosslevel interactions reveal no noteworthy pattern.

The only other—albeit insignificant—hint towards a moderating effect of unemployment policies is revealed by unemployment expenditure (cf. Figure 6.13). It shows a tendency to decrease risk, especially among those with low educational attainment and tends to lower risk more effectively among the unemployed. This is curious because unemployment expenditure does not produce a significant effect by itself (cf. Figure 6.10). Again, this should inspire

<sup>25</sup> The effects only brush statistical significance (p < 0.10). In the light of the small number of macro-level cases, they are still interpreted as indicating a moderating effect.

<sup>26</sup> As discussed before, the only reason pension policies are not omitted all together is that they may signal overall responsiveness.

<sup>27</sup> Cf. Figures A6.2–1, A6.2–2, and A6.2–3 in the appendix.

**Figure 6.12** Moderating effects of unemployment insurance coverage. *(Data: EU-SILC (2016), marginsplot based on multilevel logistic regression (melogit) with cross-level interaction, subsample (working age) controlled for GDP and unemployment rate)*

caution when using indicators of social expenditure. Spending in the field of unemployment should decrease poverty considerably if one follows theoretical expectations. The fact that this is not the case in this empirical test underscores the concerns about whether expenditure correctly measures a generous redistributive budget.

Overall, the preceding analyses generated several insights regarding the link between the welfare state and the risk of poverty. Social expenditure and insurance coverage tend to decrease the risk of poverty, while benefit generosity, contribution period, and duration of benefit receipt all produced insignificant results. Furthermore, activating policies appear to mitigate risks emanating from vulnerable situations as expected. So does unemployment expenditure. Roughly, these results are in line with existing literature (cf. section 4.2.2), although some of the insignificant results—especially in case of the net replacement rates and unemployment expenditure are unexpected.

**Figure 6.13** Moderating effects of unemployment spending. *(Data: EU-SILC (2016), marginsplot based on multilevel logistic regression (melogit) with cross-level interaction, subsample (working age) controlled for GDP and unemployment rate)*

#### **6.2.5 Summary and Discussion**

The results of the preceding analyses generated insights regarding the five hypotheses derived from three explanations for differing levels of poverty in the literature. Benefit generosity did not appear to have an effect on the risk of poverty, which partly rebuts **Hypothesis Pov1** (*The more comprehensive the social rights, the lower the prevalence of poverty in a country*). In contrast, benefit coverage did reduce risk, which can be interpreted as a partial confirmation of this hypothesis. Similarly, the weak effects of unemployment expenditure as the main and moderating effect partly contradict **Pov2** (*The higher the redistributive budget, the lower the prevalence of poverty in a country*), while the significant effects of all other spending indicators tend to confirm it. However, there is not much support for **Pov4** (*The higher the redistributive budget, the lower the risk emanating from social determinants of poverty*). Furthermore, the evidence for a moderating effect of social rights (**Pov3**: *The more comprehensive social rights, the lower the risk* *emanating from social determinants of poverty*) is weak at best. Lastly, activating policies leads to the expected reduction of risk among vulnerable individuals and therefore can be seen as evidence supporting **Pov5** (*The higher the effort directed towards activating policies, lower the risk emanating from social determinants of poverty*).

While these are interesting results, which add to research on the matter and are partly controversial, the interest of this contribution does not rest on the verification of hypotheses. Instead, it explores whether or not conceptualising and operationalising welfare stateness following the proposed framework, helps to achieve a more standardised, transparent, and comparable process. The following discussion is therefore divided into two steps. First, I discuss how well the systematised concepts could be applied to the object of research. Second, the empirical measurement and the results are critically discussed.

Turning to the first step, two mechanisms were highlighted in the explanations and hypotheses guiding this exemplary empirical test. Those mechanisms are provision of *security* and *activation*. Among those two, security is assumed to influence the risk of poverty directly (accounting for a lower prevalence of poverty) and as a moderator (reducing the risk emanating from situations of vulnerability). Activation, on the other hand, is only expected to moderate the impact of being at risk. While it may also signal a particularly involved welfare state in a direct effect, it would do so only because it serves as a proxy for a more general perspective not pursued in this contribution.

I argued that underlying these two mechanisms are two systematised concepts of welfare stateness. The *Responsive Welfare State* combines those mechanisms and hypothetical effects where the welfare state is assumed to directly reduce the risk of poverty by providing security (e.g. through income replacement and insurance coverage). Moreover, the Responsive Welfare State is in line with the explanations for poverty where social policy is expected to have a moderating effect. The second relevant systematised concept is the *Enabling Welfare State*, which is mainly assumed to shape the risk of poverty as a moderator. Since these two conceptualisations represent very different perspectives on the welfare state, distinguishing between the two appears to be very helpful. Furthermore, applying this kind of differentiation not only helps to conceptualise and operationalise more systematically but to interpret and differentiate results. Overall, embedding the framework into the research process seems quite unproblematic and will help to standardise the structure of argumentations.

The second step of this discussion is more complex as it entails discussing how successfully the concepts were operationalised in order to explain the exemplary outcome. The essential question is how success can be determined. A confirmed hypothesis might partly help to assess this in the sense of nomological validity. However, this alone is somewhat tautological (cf. section 5.2.2). Since the initial selection of indicators was already guided by the premise that they should fit the nature of the systematised concept they were selected for, the effects they produce in analyses do not change this initial assessment. Such difficulty of determining measurement validity will be discussed in more detail at the end of the chapter (cf. 6.4). For now, I will focus on the results and their interpretation.

The *Responsive Welfare State* embodies what it is at the heart of the welfare state: securing against risks and meeting needs. As such, it incorporates the primary perspective we choose when we examine the link between social policies and poverty. In this exemplary analysis, I chose several indicators of social rights, which capture the institutional set-up of welfare states: eligibility criteria (contribution period), generosity (duration of receipt, replacement rate), and insurance coverage. These elements of social policy-making should embody responsiveness as they signal how quickly, how easily and how comprehensively security is provided. In theory, they should therefore all account for cross-national variations in the level of poverty (direct effect) and moderate the consequence of vulnerability. However, as the results show, this can only be observed for insurance coverage and in this case only when it comes to a direct effect. There is only a small tendency for coverage to benefit particularly those with low educational attainment, but it is not statistically significant. Various factors could explain this result. First, it could of course mean that the chosen indicators are not suitable. Perhaps, other indicators capture the essence of responsiveness better than the ones selected. Second, it could also mean that the selected operationalisation or source of the indicators is not suitable or that some countries distort the results (such as the CEE countries). Third, I cannot rule out that the impact of those mechanisms tied to the *Responsive Welfare State* is not as clear or as strong as expected. Since only a comparatively small part of variance can be attributed to the country level (ICC = 5.5%), differences in the risk of poverty between European countries are perhaps not that pronounced. All of these aspects require further attention in the future. Nevertheless, speaking very strictly, the indicators in this analysis were chosen following clear criteria. Thus, the fact that some of them fail to produce significant effects does not yet mean that they are unsuitable candidates for the Responsive Welfare State. Instead, they should be tested again—in other analyses of outcomes that are associated with this specific perspective on welfare stateness. Furthermore, the result that social expenditure *does* decrease poverty should only encourage further efforts to flesh out the nature of distinct conceptualisations of welfare stateness. Since we do not know *what exactly* is responsible for the observed effects of spending—redistributive budget, comprehensive benefits, etc.—this should discourage from using expenditure indicators if specific mechanisms and conceptualisations are tested. Since, however, there is a notable effect of expenditure and welfare effort (or redistributive budget) highlighted prominently in the relevant literature, it is only logical to try to figure out *why*.

In contrast to those mixed findings, the operationalisation of the *Enabling Welfare State* proved successful—at least in terms of confirming the hypothesis. Both—a direct and a moderating influence of ALMP spending on the risk of poverty were found. Since this indicator was reduced to spending on very specific and explicitly activating measures (training, employment incentive, supported employment and rehabilitation, direct job creation, and start-up incentives), it is more clear-cut than the heavily criticised other measures of spending. Still, it is only one indicator and it is highly recommendable to try other operationalisation of social investment in the future.

Some additional remarks have to be made about the preceding analyses. While I focused especially on the impact of macro-level indicators of welfare stateness and thus on the question if social policies account for some of the variances in the individual risk of poverty between countries, it predicts poverty with some limitations. A stronger focus on explaining poverty instead of exploring the impact of welfare stateness would be more informative if a longitudinal instead of a cross-sectional strategy was implemented. This especially refers to one particular aspect of the *Responsive Welfare State*: the moderating effect of securing individuals once they enter vulnerable situations. Thus, my cross-sectional analyses reveal how far welfare states potentially reduce (or increase) poverty among the unemployed or low educated but they do not tell us if welfare states lower the risk of poverty at the moment of entering unemployment.28 Similarly, a comprehensive analysis of the *Enabling Welfare State* would also spell out the paths in more detail: is a reduction of risk among the low educated and unemployed actually due to their active participation in ALMP measures? These things have to be kept in mind when interpreting the causality behind the security and activation mechanism in the cross-level interactions that have been performed.

Further restrictions relate to the test of the introduced hypotheses and in particular to the first one (*the more comprehensive the social rights, the lower the prevalence of poverty in a country*). Here, the direct effect of generous social rights may be included, but its link to *redistribution*—as expected based on the redistribution paradox—was not explicitly tested. Neither was the support

<sup>28</sup> Obviously, low educational attainment is not effected by this.

for redistribution, which may serve as an intermediary factor in this case (cf. section 2.2.2). A comprehensive test of these different paths, leading to a comprehensive explanation of poverty is not the main objective of this contribution. Therefore, the results should be discussed in terms of policy measures, but should not be understood as detailed contributions to the literature on poverty and the redistribution paradox.

Overall, the findings of this first test of the proposed framework are mixed. Using distinct conceptualisations of welfare stateness as an analytical tool that guides the operationalisation is certainly helpful. This is not just the case for the selection of indicators but also for the interpretation of results. Furthermore, using single indicators forces to ask whether we actually test the mechanisms we assume. Again, this seems very advisable. Still, the selection should be expanded in the future, tested on other dependent variables, and be critically discussed. The following section takes up at least the latter two recommendations, as it presents a second exemplary application of the framework using attitudes towards the welfare state as a dependent variable.

#### **6.3 The Welfare State and Welfare State Attitudes**

Analysing how far welfare state policies influence attitudes towards the welfare state is an evident undertaking. As such, it is also a very popular one. The literature on the matter was summarised in brief in chapter 4 and it revealed a great diversity in theoretical and conceptual as well as empirical approaches and results. The state of research will not be repeated at this point, but it is important to recapitulate the main explanations for how and why social policies are assumed to shape attitude formation in this case.

#### **6.3.1 Conceptualisation of the Welfare State**

Like in the previous example, the first question to be addressed is: w*hich perspective on the welfare state is chosen?* As was argued during the literature review in chapter 4, bottom-up and top-down perspectives on the welfare state can both be found in hypotheses about how it influences individual attitude formation. Thus, it is possible to ask how the welfare state influences the individual as well as how the individual perceives the welfare state and to what end. This is reflected in the mechanisms highlighted in the literature and the hypotheses that can be deduced from them. This relates to the second question that should guide conceptual considerations: *which mechanisms are addressed?* Within the top-down perspective, the welfare state is assumed to shape attitudes by conveying solidarity and justice principles through the mechanisms of *socialisation*. In addition, it is also argued that responsiveness—as represented by the comprehensiveness of the provision of *security*—leads to political support and positive attitudes towards the welfare state. The following explanations and corresponding hypotheses sum up these assumptions:


Turning to the bottom-up perspective the focal point is the individual perception of welfare stateness and the perceived (potential or actual) individual benefit. Here, two principal explanations were deduced from the literature, which highlight the mechanisms of *evaluation* and *self-interest*. Underlying both explanations is the premise that social policies have to be known to a certain degree in order to be included in the formation of attitudes. The corresponding hypothesis highlights support for welfare states stemming from the assessment of performance in general (*Att3*) or specific considerations of individual costs and benefits (*Att4*) that are tied to social policies. This emphasises individual perception over responsiveness. However, it comes with the restriction that individual perception, which is assumed as a bridging hypothesis, cannot easily be included in empirical models. In essence, the link between welfare state and attitudes in both cases is therefore similar to *Hypothesis Att2*.

• **Explanation 3.3:** Perceived fairness and good performance lead to more positive attitudes towards the welfare state because individuals *evaluate* these actions positively (bottom-up perspective).


Again, it is irrelevant whether these four hypotheses exhaustively cover all theoretical assumptions in the literature. Instead, they serve as a minimal consensus on how and why welfare states may shape attitudes. For this contribution, it is important that they reveal different perspectives, highlight different mechanisms and help to answer the third question: *which concepts of welfare stateness are addressed?* When exploring this question, we have to take into account that the impact of social policies on attitude formation can be approached from the *topdown* as well as the *bottom-up* perspective. In accordance with the proposed framework, the chosen perspective and highlighted mechanisms guide the conceptualisation. Within the top-down perspective, two mechanisms are highlighted in particular: provision of *security* and *socialisation.* One could argue that *redistribution* and *social stratification* may provide explanations for differing attitudes as well—they are, however, of less importance as their influence is mostly covered by security and socialisation.29 Thus, the way welfare states redistribute and shape stratification might influence solidarity and justice principles—but this happens in the sense covered by socialisation. Similarly, redistribution and stratification are also integral parts of responsiveness. Again, their influence can be seen as embedded in the security function. Since the way in which security is provided (redistribution) and its impact on social stratification is not explicitly emphasised in the tested hypotheses, the security mechanism seems to be the most relevant one. Thus, when the security mechanism is emphasised in examining the relationship between welfare stateness and welfare attitudes, the welfare state is conceptualised as the *Responsive Welfare State*. If socialisation is highlighted, the *Normative Welfare State* is at the core of the analysis.

<sup>29</sup> They would be of more importance in analyses of attitudes towards specific modes of redistribution or if hypotheses assumed a specific way of stratifying societies is responsible for attitude formation. Both research objectives are plausible but represent very specific examples, which are not pursued further in the following analysis.

Turning to the bottom-up perspective, *evaluation* and *self-interest* are tied to the concept of the *Assessed Welfare State*. As previously argued, this conceptualisation emphasises the individual perspective on the welfare state, its performance and potential individual benefits gained from it. Regardless of whether the process leading to the formation of a certain attitude is guided by self-interest or other considerations, it is always based on an individual assessment and is therefore summarised in the same systematised concept.

None of the three concepts is a better fit than the others and nothing speaks against conceptualising, operationalising, and testing all three. However, the distinction is of additional informative value and a means of achieving more targeted and comparable measurements, as the concepts underlying the welfare state are more clearly distinguished.

Two more questions have to be addressed before turning from theoretical conceptualisation to empirical operationalisation: *which policy fields are relevant* and *which temporal perspective has to be chosen.* Regarding policy fields, there is no reason to limit the selection to just one area. The *Responsive Welfare State* manifests regardless of the policy field and—contrary to the analyses performed in chapter 3—welfare attitudes will not be measured in just one specific policy area in the following analyses. It could be argued that there is reason to believe that some issues are more salient than others (e.g. unemployment, cf. section 4.2.3) and might therefore be more fitting for the *Assessed Welfare State*. However, since this cannot be generalised to all examined countries with certainty, reducing the analysis to only one policy field seems inadvisable. Furthermore, distinguishing between as many issues as possible could bear important insights.

Grasping the temporal perspective is a bit more difficult. Regardless of the perspective, attitudes can be formed with reference to the current state of social policies. However, it might be more than just the status quo that is responsible for attitude formation. This especially relates to the *Normative Welfare State*, which potentially shapes individual attitude formation over a long period and varies between individuals depending on their age. Likewise, the *Assessed Welfare State* may be affected by a certain delay, as it is unlikely that all citizens are up to date on political issues—especially in areas, such as accident policy, which may be far removed from individual reality. Such considerations are important and are—to the best of my knowledge—not prominent in the relevant literature.<sup>30</sup> Still, in my analyses, the selected datasets and indicators (cf. section 6.1.3) only cover features of welfare states in 2015 and therefore only allow a short-term perspective.

<sup>30</sup> Perhaps, with the exception of studies using welfare state typologies based on historical trajectories—although the broader time horizon is seldom a clearly stated objective.

Since the primary objective is to test and compare the different indicators and assess their fit with the systematised concepts, this restriction seems acceptable. The issue does however deserve much more attention in the future.

#### **6.3.2 Operationalisation of the Welfare State**

The previous discussion reveals three conceptualisations of welfare stateness, which help to explain different attitudes towards the welfare state. More specifically, they narrow down particular characteristics of social policy-making, which offer starting points for targeted operationalisations of these characteristics. In the previous chapter (cf. chapter 5), the following recommendations were introduced:

The *Responsive Welfare State* should be operationalised using indicators of social rights (e.g. benefit generosity, coverage, eligibility criteria).

The *Normative Welfare State* should be operationalised using indicators linked to principles that guide the provision of social services, such as universalism or egalitarianism (e.g. coverage, eligibility criteria).

Since they both refer to indicators stemming from the social rights perspective, there is some overlap in the operationalisation of these two conceptualisations within the top-down perspective. This is important to note and makes it difficult to distinguish between the two.

In case of the *Assessed Welfare State,* the expectations include that salient features of welfare stateness and those representing individual benefits represent the most fitting indicators.

This recommendation remains a bit more abstract than in the case of the other two concepts. Benefit generosity is expected to be more salient than other features and there is reason to expect that some policy areas are better known than others (e.g. unemployment policies might be more salient than accident insurance policies).

Consulting the list of indicators described at the beginning of this chapter (cf. 6.1.3) shows that almost all sets of variables promise an explanatory contribution as they relate to at least one of the three relevant conceptualisations of welfare stateness. Social expenditure is the only exception. It is included only because it can be assumed to be somewhat salient—albeit knowledge about actual expenditure may be biased (cf. chapter 4). The following analyses might therefore reveal important insights that can be added to the conclusions drawn from the analyses in the previous section of this chapter.

#### **6.3.3 Additional Variables and Analytical Strategy**

Several dependent and independent variables are included to test the conceptualisation of welfare stateness. The following part describes the operationalisation of these variables as well as the analytical strategy for testing the welfare state's impact on attitudes towards the welfare state.

#### *Dependent variables*

Welfare attitudes are measured through various items in the following analysis. While chapter 3 relied on government responsibility alone—arguably a common indicator in the field—the ESS 2016 offers a more nuanced operationalisation of such attitudes. As noted previously (section 4.2.3), the operationalisation of welfare attitudes is far from consistent and surrounded by debate (e.g. Svallfors 2012b). As pointed out earlier, the preferred amount of government responsibility presents a conceptually different issue than more general welfare attitudes or other related phenomena such as welfare chauvinism.

In the following analysis, I take a more differentiated perspective on the matter than before and capture different components of welfare attitudes. Differentiating between distinct perspectives aids in determining, how the different welfare state indicators relate to different manifestations of welfare attitudes. The European Social Survey 2016 includes a comprehensive module on welfare attitudes called "Welfare Attitudes in a Changing Europe", which suits this purpose well. It includes a battery consisting of six items measuring general attitudes towards the welfare state. Here, respondents indicate how much they agree with the following statement: social benefits (1) place a great strain on the economy, (2) cost businesses too much in taxes, (3) make people lazy, (4) make people less willing to care for one another, (5) prevent widespread poverty, and (6) lead to a more equal society. A principal components analysis reveals that two dimensions underlie these statements. The first includes the four negative attitudes towards social benefits, the second the two positive ones. The items are bundled in two mean indices, which are labelled *welfare state scepticism* and *welfare state support*. 31

While preferred role of government (*government responsibility for welfare provision*) was covered by six items in the 2008 version of the ESS used in chapter 3,

<sup>31</sup> Cronbach's alpha is 0.73 for welfare state scepticism and 0.67 for welfare state support.

the 2016 wave only includes three of those items. In this case, respondents were asked to indicate how much responsibility they think governments should have for ensuring (1) a reasonable standard of living for the old, (2) the unemployed and (3) providing childcare services for working parents. The three items are combined in a mean-index.<sup>32</sup>

The last of the four dependent variables is *welfare chauvinism* and it is covered by one item asking respondents how long immigrants should have to wait until they are eligible for social benefits and services—ranging from immediately on arrival to never. An overview of these four dependent variables and univariate descriptive statistics are provided in the appendix.

#### *Independent variables*

The analyses include several independent variables, which address individualas well as country-level features. Even though the focus rests on the explanatory contribution of different aspects of welfare stateness, this is explored in models representing comprehensive analyses of welfare attitudes best. Therefore, I added several socioeconomic and sociodemographic variables on the level of individuals. This includes information on respondents' *age* (also age squared in order to account for non-linear effects), *sex*, *educational level*, *employment status* and *financial insecurity.* Most of these variables and their operationalisation correspond to those used previously in the analyses of the risk of poverty. Even though they stem from different data sources, the operationalisation of the different response categories follows the previous proceeding as accurately as possible. The only noteworthy difference lies in the operationalisation of poverty. While EU-SILC allowed to actually model whether a person falls under the poverty threshold, the ESS only includes broad income categories with many missing values and no correction for household size and composition. Therefore, a proxy is included: subjective financial insecurity (finding it very difficult to live with present income). In addition to the individual-level variables, the *gross domestic product* (GDP) and *unemployment rate* are included on the level of countries. This selection is guided by established literature on the matter, and was discussed in more detail in my previous analyses (cf. chapter 3 and section 6.2).<sup>33</sup>

#### *Analytical strategy*

In accordance with the theoretical conceptualisation and analogous to the analyses performed previously (cf. chapter 3) and the majority of similar research,

<sup>32</sup> Cronbach's alpha is satisfactory (0.68).

<sup>33</sup> For more information and univariate descriptive statistics cf. Table A6.3–1 in the appendix.

multilevel analyses are performed (MLA). More specifically, since all four dependent variables are surveyed using a common response scale, hierarchical linear models are applied. It can be argued that dichotomising the items and performing logistic or linear probability analyses would be a viable option as well. In previous research, comparisons of both strategies did not produce gravely deviating results if the ordinal scale was maintained (in case of the responsibility items cf. Brady & Finnigan 2014: 25). Furthermore, it could be argued that since the original response scales were ordinal, multilevel ordered logistic models may be more appropriate than linear ones. Since, however, the initial scales were further broken apart and differentiated by constructing composite indices, the ordinal nature of the original items is softened up (with the exception of welfare chauvinism). Still, to avoid bias, ordered logistic models were tested in addition to the linear multilevel models. They did not lead to differing conclusions.

The advantages of multilevel linear models were already introduced in an earlier chapter of this book (cf. chapter 3). Like before, the analyses are realised in several successive models. Since it is plausible to expect that GDP and unemployment rate are not independent of features of the welfare state, they are added stepwise. The resulting regression equation for the full model predicting attitudes among individuals *i* in country *j* including all variables at the individual (*xi j*) and country level (*zj*), is identical to the one in chapter 3:

$$\text{y(attribute)}\_{ij} = \eta\_{00} + \sum\_{h=1}^{r} \eta\_h \chi\_{hij} + \sum\_{l=1}^{r} \eta\_l z\_{l0j} + \mu\_{0j} + e\_{ij}$$

The explanatory contribution and fit of the models is determined based on various indicators. This includes tracing several information criteria such as AIC, BIC and Loglikelihood. Concerning changes in variance, the linear estimation allows obtaining information on R-squared in addition to the ICC. Like in chapter 3, Bryk and Raudenbush's (2012) R-squared is obtained for the individual and the country level.<sup>34</sup>

<sup>34</sup> In order to avoid bias stemming from design issues or the under- or oversampling of population groups, the ESS offers several weights (ESS 2014). In descriptive analyses (in the appendix), a combination of population and design weight is chosen. All descriptive statistics in the appendix use weighted data; the multivariate analyses are calculated with and without active weights. Since there is no difference in effect direction and significance, the unweighted version is reported.

#### **6.3.4 Results of Bi- and Multivariate Analyses**

A first glance at the distribution of the dependent variables (Figure 6.14) reveals noteworthy variations between countries. The most positive general attitudes towards the welfare state (*welfare state support*) are observed in Norway, Austria and Belgium, while Hungary, Lithuania and Estonia range at the lower end of welfare support. Interestingly, this does not seem to correspond directly to the index of more *sceptical attitudes* towards the welfare state confirming its status as an independent latent dimension. Here, Portugal, France and the UK exhibit the highest average values. Welfare provision is especially seen as *government responsibility* in Lithuania, Italy, Portugal and Spain while the Netherlands, the UK and Switzerland exhibit the lowest preference for government responsibility for the provision of welfare benefits and services.

Lastly, *welfare chauvinism* is strongest in Hungary, Czech Republic and Lithuania and lowest in Spain, Sweden and Portugal. Overall, there seems to be a tendency for less positive and generous attitudes in CEE countries and the UK. However, there is no obvious pattern for the rest of the European regions.

While this descriptive information already reveals some variation, it is important to determine how much variance can actually be attributed to differences between countries. Like before, the ICC is used for this purpose. Here, welfare chauvinism, preferred government responsibility and general attitudes towards the welfare state exhibit very similar ICC values. In all three cases, between eight and nine percent of variance can be attributed to the country-level. While these figures are not overwhelmingly great, they still reveal a noteworthy amount of variance that can potentially be explained by differences in welfare stateness. Considering that attitude formation is at its core an intra-individual process, a tenth of variance actually seems to be quite a lot and it corresponds to previous research and my own analyses of 2008 data (cf. chapter 3). In contrast to these three variables, scepticism towards the welfare state produces an ICC of only four percent. The potential to explain differences through country-specific features is thus considerably lower for this dependent variable. Nevertheless, the ICC is high enough in all four cases to justify performing multilevel analyses.

The next step entails exploring how much of the observed variance can be explained by indicators of welfare stateness and—more importantly—whether the indicators produce results, which are in line with the described expectations. While the focus of the analyses rests on the macro-level, a short glance at the effects produced by key micro-level variables is still important to make

**Figure 6.14** Mean and standard deviation of welfare attitudes. *(Data: ESS (round 8))*

sure that main premises are fulfilled.35 In order to provide a short overview of the individual-level determinants of the different welfare attitudes, Figure 6.15 summarises the result of the purely individual-level models. They reveal that general support for welfare states decreases with age (a), while age increases chauvinism (d) and demand for government responsibility (c) and has no significant impact on scepticism (b). Sex shapes at least some of the attitudes with male respondents being less in favour of strong government responsibility for welfare provision and more chauvinistic when it comes to extending benefits to immigrants. Furthermore, high education (tertiary, ISCED 5) leads to more positive general attitudes towards the welfare state (a), less scepticism (b), and less chauvinism (d) compared to the lower educational groups. The only exception is preferred government responsibility (c), where the lowest educational group is in favour of more responsibility. In contrast, being employed appears to lower support and increase scepticism. Lastly, respondents expressing difficulties when it comes to living on present income show less general support for welfare states, more chauvinism, but also a stronger preference for government responsibility and less welfare state scepticism.

Overall, the individual-level results reveal some interesting insights about the different dependent variables, which are noteworthy even though the focus of this contribution rests on the country-level. First, the patterns of effects for the four dependent variables are quite different—even between the two pairs of positive (support, responsibility) and negative (scepticism, chauvinism) attitudes. This supports the approach of distinguishing different facets of welfare attitudes. For instance, the fact that high education fosters general support for welfare states but decreases the preference for government responsibility seems worth more systematic exploration in future research. Similarly, it should be explored further why while those who contribute (the employed) overall tend to have more negative attitudes towards the welfare state than those who benefit (unemployed or retired individuals), this mainly manifests in scepticism and preferred role of government, not in chauvinism and only partly in general support. Second, the individual-level indicators only explain a small share of variance. The highest explanatory contribution is achieved in the case of chauvinism (roughly two percent).36 Thus, the control variables may account for different socioeconomic and sociodemographic backgrounds of respondents but do not deliver comprehensive

<sup>35</sup> A comprehensive documentation of all estimated models and the effects of further control variables is provided in the appendix.

<sup>36</sup> Information on explained variance and other fitness indicators is provided in the full regression tables in the appendix.

**Figure 6.15** Individual-level determinants of welfare attitudes. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

explanations for varying attitudes between individuals. While my findings do not contradict the state of research for the selected individual-level characteristics and the respective dependent variables (e.g. Häusermann et al. 2016; Eger & Breznau 2017; Kölln 2018), other explanatory factors should be included if a comprehensive explanation of micro-level processes is targeted.

After exploring individual-level determinants of welfare attitudes, the following section focusses on indicators of welfare stateness. Like in the analysis of the risk of poverty, the coefficients are reported in two versions: before and after controlling for GDP and unemployment rate.

*Net replacement rates* (NRR) represent the first set of indicators. According to the conceptual expectations, they should relate to various systematised concepts of welfare stateness: the *Responsive*, *Normative* and *Assessed Welfare State*. Since all three conceptualisations are assumed to explain differences in attitudes, net replacement rates should be strong indicators—albeit being too undifferentiated to be attributed to either one of the three concepts. The results reported in Figure 6.16 support the expectation that NRRs indeed deliver an explanation for varying attitudes. Respondents in countries with more generous benefit replacement exhibit more positive general attitudes towards the welfare state (a), are in favour of more government responsibility (b) and are less sceptical. In case of unemployment replacement rates, they also tend to be less chauvinistic. There are some exceptions—for instance, the NRR in case of pensions does not appear to predict welfare state scepticism and chauvinism. Furthermore, it loses significance for the explanation of preferred government responsibility once the control variables are included, while it is only significant as a predictor for welfare state support with controls. The latter happens in several instances, which confirms the importance of controlling for GDP and unemployment rate, which are clearly cofounded with welfare stateness. Overall, the majority of indicators show a clear relationship between benefit generosity and attitudes, which is supported by the post-estimates.37 When evaluating these results and their fit with the three potentially relevant conceptualisations, there seems to be support for using this indicator, since generosity—in most policy fields—fosters support for the welfare state. However, whether this is due to responsiveness, the socialisation of respondents or their perception and evaluation cannot be determined at this point.

While NRRs may suit several conceptualisations of welfare stateness, the *contribution period* required to qualify for a benefit, was selected as an indicator fitting only two conceptualisations: the *Responsive Welfare State* and partly also the *Normative Welfare State.* Turning to the results, the picture seems to be a bit ambivalent (cf. Figure 6.17). While being eligible for sickness benefits without having to fulfil any contribution period tends to increase positive attitudes (a) and decrease scepticism (b), it also increases chauvinism (d). In contrast, a short contribution period required to be eligible for pension benefits, tends to lower support and increases scepticism (d), while the contribution period is insignificant in the case of unemployment benefits. There seems to be no clear pattern and without exploring in detail why the contribution period in the case of sickness tends to foster positive attitudes and chauvinism at the same time, it is not possible to evaluate whether the indicator adequately captures either the *Normative* or the *Responsive Welfare State*.

The next set of indicators *– benefit duration*—was proposed as mainly providing a measurement of the *Responsive Welfare State*. As such, it is assumed

<sup>37</sup> For instance, Bryk and Raudenbush's R-squared indicates a value between .1 and .3 for all significant indicators (before adding control variables). These and more fitness indicators can be found in the corresponding regression tables in the appendix.

**Figure 6.16** Welfare attitudes on benefit replacement rates. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

to shape attitudes only as so far as factual performance automatically generates support. As the results (Figure 6.18) show, benefit duration only explains varying welfare attitudes between countries in some policy fields. The only significant effects are produced by unemployment benefit duration, which increases general support (a) and decreases chauvinism (d) and by accident benefit duration, which increases scepticism (c). While the effects of unemployment benefit duration are in so far plausible, as unemployment policies are assumed to be comparatively salient features, the effect of accident benefit is counterintuitive and may signal some other underlying or confounding effect, which cannot be explored in more detail at this point. In addition, as salience is no prerequisite of the *Responsive Welfare State*, it is possible that the observed effect of unemployment benefit duration also expresses another confounding effect*.* Overall, the results seem to suggest that the *Responsive Welfare State* is of secondary importance for the explanation of attitude formation—if (and this has to be analysed very critically later on) the selected indicators allow an adequate operationalisation of this concept.

**Figure 6.17** Welfare attitudes on contribution period. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

Contrary to benefit duration, *insurance coverage* was selected as an indicator of universalism of benefits, which may not only represent the *Responsive Welfare State*, but also the *Normative Welfare State*. However, the indicator produces barely any significant effects—especially not after controlling for GDP and unemployment rate (Figure 6.19). At first glance, insurance coverage therefore does not seem to be a fitting indicator for either conceptualisation, which is supported by a neglectable share of explained variance.38 Again, one could also critically add that it may signal a subordinated relevance of the conceptualisations in case of welfare attitudes.

The last set of indicators consists of several measures of spending (Figure 6.20). Again, the disadvantages of this operationalisation have to be kept in mind. As argued before, expenditure was included in this analysis because it is by far the most commonly used among the single indicators. Furthermore, there is some reason to believe that it is among the more salient features of

<sup>38</sup> Cf. tables in the appendix.

**Figure 6.18** Welfare attitudes on benefit duration. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

social policy-making and as such may be considered when trying to capture the *Assessed Welfare State*. Even though perhaps salient, it may still be biased by a lack of accurate knowledge about actual spending (as discussed in section 4.2.3). A first glance at the relevant coefficients reveals several insights. First, social expenditure increases a positive general attitude towards the welfare state (a). This happens regardless of the addressed policy field and remains a robust finding even after controlling for GDP and unemployment rate. Second, neither of the expenditure measures delivers significant explanations for variations in welfare state scepticism and preferred role of government—even though there seems to be a small (but insignificant) tendency to support more government responsibility if health care spending is low. Interestingly, health expenditure and overall expenditure also decrease welfare state chauvinism—this time significantly so. Thus, high spending in those fields does not appear to activate a need to secure benefits against outsiders, but instead fosters generous attitudes. This could be interpreted

**Figure 6.19** Welfare attitudes on insurance coverage. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

in the sense of the *Normative Welfare State*. However, this has to happen in combination with other indicators, more closely tied to this conceptualisation and suffering from less fuzziness.

Lastly, ALMP spending, which was only included for the sake of exhaustiveness, does not produce significant effects. Since there is no reason to assume the *Enabling Welfare State* shapes attitudes, this finding is in line with the expectations.

To sum up the main findings: benefit generosity leads to more positive attitudes towards the welfare state in most models. A similar tendency is observed regarding social expenditure, although not in the case of scepticism and only partly for government responsibility. Benefit coverage in contrast, does not appear to notably shape attitudes. Lastly, the duration of benefits only fosters positive attitudes in the case of unemployment duration and contribution period—as an indicator of eligibility criteria—exhibits a tendency to increase scepticism in case of contribution to pension schemes and chauvinism in the field of sickness

**Figure 6.20** Welfare attitudes on social expenditure. *(Data: ESS (round 8), coefplots based on multilevel linear regressions (xtmixed))*

polices. The question how well the distinction between different conceptualisations of welfare stateness was embedded in this analysis and how accurately the different perspectives on the welfare state were modelled will be addressed in the following section.

#### **6.3.5 Summary and Discussion**

Do social policies explain differing attitudes towards the welfare state between countries? The results obtained in this chapter support the same assumptions as the literature on the matter (cf. section 4.2) and as my previous analyses (cf. chapter 3). Indeed, welfare policies account for cross-national variation and very broadly speaking, comprehensive and generous social policies foster positive views towards the welfare state. This finding can be integrated in the state of research and even the effects pointing in other directions (such as the fact that some pension and sickness policies increase scepticism and chauvinism) could be interpreted in line with existing argumentations. In fact, one can interpret this result as support for all hypotheses stated previously. However, there seems to be more support for those hypotheses highlighting the bottom-up perspective (*Hypothesis Att3 and Hypothesis Att4*) since indicators that should capture the top-down perspective perticularly well, mostly produced ambiguous, unexpected, or insignificant results.

However, the aim of this contribution is to offer more than that. By conceptualising and testing narrower perspectives on welfare states, the results offer additional insights and help to differentiate to a higher extent. Based on the argument that adequately modelling the different hypotheses in the literature requires not only adequate measurement of the dependent variable but also of the independent variable, it introduces a more nuanced view on welfare states. In contrast to the previous analysis of the relationship between social policies and the risk of poverty, this approach is put to the test more extensively in the case of welfare attitudes because here the top-down *and* the bottom-up perspective applies. Potentials and limitations of the proposed approach should therefore be revealed especially in this exemplary case. The usefulness of the analytical framework is assessed in two steps. First, I discuss the applicability of the systematised concepts to the object of research. Second, I discuss the empirical measurement, bearing in mind validity criteria.

Three systematised concepts were applied in order to capture different perspectives on welfare stateness behind mechanisms and hypotheses. The first conceptualisation is the *Responsive Welfare State*, which underlies hypotheses assuming that factual performance creates general support for welfare states (independently of individual perception). The second concept is the *Normative Welfare State*. This conceptualisation can be found if hypotheses focus on the way in which welfare states convey solidarity and justice principles and socialise citizens. Finally, the *Assessed Welfare State* can be seen as a counter perspective to the Responsive Welfare State as it manifests in individuals' perspective on welfare states and their assessment of policy-making. Especially these two closely related conceptualisations reveal the potential usefulness of distinguishing different facets of welfare stateness early in the research process. Even if they capture very similar things, the Responsive Welfare State may perform "well" without individuals explicitly noticing why, while the Assessed Welfare State requires a certain amount of knowledge about policy-making. Discussing this difference may already enrich debates and the attempt to operationalise the difference empirically may contribute even more insights.

Evaluating how well this empirical operationalisation of the concepts worked, is however again difficult. And,—as discussed before—it is complicated even more by the fact that the indicators selected for this empirical test can be suitable for more than one conceptualisation.

The operationalisation of the *Responsive Welfare State* is the same as in the analysis of poverty. Indicators are attributed to this conceptualisation if they signal how comprehensively security is provided (benefit coverage, replacement rates, duration of receipt) and how quickly benefits can be obtained (contribution period). The fact that these indicators for the most part fail to deliver clear results is conspicuous. It could signal that responsiveness is not adequately captured through these indicators. However, it could also mean that the mechanisms tied to the Responsive Welfare State are not as influential for the explanation of attitudes as expected. It is difficult to discuss whether this might be because another conceptualisation such as the Assessed Welfare State (as a counter perspective) is actually more suitable for the research subject. Rejecting an entire conceptualisation requires more detailed analyses and more attempts to capture the Responsive Welfare State—perhaps from another perspective. If a more clearcut operationalisation of the Responsive Welfare States does not succeed, this finding may bear insights that go beyond matters of operationalisation. If responsiveness does not shape attitudes as much as expected, hypotheses that highlight mechanisms tied to the provision of *security* might require a critical examination. This renders the exploration of nomological validity all the more difficult.

Findings are similarly unsatisfying when it comes to the measurement of the *Normative Welfare State*. Here, benefit coverage and benefit generosity were highlighted as operationalisations. Benefit coverage does not contribute much to the explanation of different attitudes. Even though replacement rates overall produce results in line with expectations, this cannot be taken as clear confirmation for their usefulness as operationalisations of the Normative Welfare State because they may also suit other conceptualisations—especially the *Assessed Welfare State*. Turning to the latter, indicators were selected if they are assumed to be salient and represent either individually perceived performance or potential benefits. Out of the sets of indicators, this referred to net replacement rates and social expenditure.

The aim of this project is not to assess which perspective on the welfare state is theoretically more appropriate, but how to measure different conceptualisations more accurately. Still, at this point one wonders, whether attitude formation is actually related as much to a top-down perspective as to a bottom-up perspective. All conceptualisations but the Assessed Welfare State were only partially confirmed empirically. Still, the indicators chosen to operationalise the Assessed Welfare State—social expenditure and net replacement rates—are not only used in a very general way, but also lack clarity. I will thus refrain from detailed assessments of theoretical premises; exploring much more, whether we can actually assume that the welfare state shapes attitudes independently of individual evaluations still seems highly valuable.

Again, there are restrictions to the performed analyses and thus limitations when it comes to their interpretation. One is the mentioned incapability of promoting one conceptualisation over another based on the results. Even more serious, however, is the fact that formulating substantial recommendations for the empirical operationalisation of each of the concepts is almost impossible because similar indicators could suit all conceptualisations. Thus, there is only a tendency to say that at least benefit generosity (as measured through replacement rates) is more closely related to the Assessed Welfare State because other more specific indicators of either the Responsive or the Normative Welfare State fail to produce results that are in line with the hypotheses. However, this definitely requires more research.

Another restriction is that actually testing the Normative Welfare State and the mechanism of socialisation is difficult, as it requires detailed data on the contextual and the individual level. It is a strong postulate to say attitudes are formed because respondents grow up under certain political conditions—or at least live under them long enough. Strictly speaking, this would require longitudinal or retrospective data for respondents and information about social polices during their formative years. Overall, the Normative Welfare State is perhaps the most difficult concept to grasp.

A final limitation is that the hypotheses described all involve a bridging assumption that has not been tested in detail. Thus, assuming the welfare state shapes attitude formation, because it raises support, or that individual preferences entail weighting up social policies with one's own values is full of prerequisites. And for the sake of completeness, I must also point out that the analyses carried out cannot do justice to the complex contributions on policy feedback and the interdependency between public opinion and policy-making (Breznau 2017, 2018). Like in the analysis of the risk of poverty (cf. section 6.2), this exemplary test of the analytical framework was intended to compare operationalisations, not to contribute substantially to the complex literature on welfare attitudes.

There are however also important insights to be gained for the objective of this book. In particular, the distinction between the bottom-up and the top-down perspective on the welfare state is fruitful in this exemplary analysis—not necessarily in terms of operationalisation, but in terms of discussing theoretical premises and interpreting results. Furthermore, the results at least hint towards confirming the expectation that unemployment policies constitute especially salient features of welfare stateness and therefore shape attitude formation—in the sense of the Assessed Welfare State—the most.

Even though this is not the focal point of this book, it should also be noted that my results support the need for a more differentiated perspective on welfare attitudes. The different measurements offered in the 2016 ESS show potential in capturing various facets of the issue. I would like to encourage such endeavours based on my results.

#### **6.4 Concluding Remarks**

Aiming at operationalising distinct conceptualisations of welfare stateness bears great potential to improve the transparency and comparability of measurements. It does so for several reasons. It demands that we look more closely at the mechanisms behind theoretical premises, justify the choice of indicators, discuss the scope of the results they offer, and it guides the interpretation of results. The conceptualisations are a tool not unlike others used in scientific endeavours when it comes to the conceptualisation of complex phenomena. However, as argued in the second chapter of this book, welfare states—as very complex arrangements of policy-making—have so far mainly received attention when it comes to their operationalisation as *dependent variable* and it does require a different perspective when their features are included as an *independent variable*. The analytical framework proposed in this book and tested in this chapter aims at filling this gap. And it succeeded in doing so at least partially by presenting a guideline for important analytical steps and also delivering some more specific insights—such as the fact that distinguishing between those operationalisations tied to assessment and those representing the mode and effectiveness in which welfare states function, does indeed narrow down operational choices. This refers to the indicators themselves (benefit coverage seems to be a promising indicator of responsiveness and benefit generosity appears closely tied to individual assessments) but also to analytical strategies (oftentimes moderating effects are more adequate than direct ones—especially in case of responsiveness).

While evidence for the usefulness of distinguishing conceptualisations of welfare stateness exists, discussing the validity of the proposed operationalisations from a broader perspective is difficult. Here, several things have to be noted critically. First, only a selection of possible indicators was tested. As argued in the beginning of this chapter, I prioritised indicators used previously in the literature that fulfil the criteria of clarity, comparability, and availability best. However, it is not possible to rule out that other indicators that were not tested are potentially more suitable. Thus, in the sense of *content validity*, it is not possible to determine with confidence that the selection perfectly fits the concepts. Second, most chosen indicators fit more than one conceptualisation of welfare stateness and therefore judging, which fit is better or worse should be done with caution. Again, this restricts the discussion of *content validity*. Third, the initial introduction of the indicators of welfare stateness already revealed that correlations among potential candidates for a similar systematised concept are difficult to interpret. This hampers determining *convergent validity*. Lastly, concluding that those indicators are suitable (in the sense of *construct validity*) that confirm established hypotheses is potentially tautological. As discussed in section 5.2.2, operationalisations validated through hypotheses cannot be used to test the very same hypotheses.

All of these issues were foreseeable and they were mentioned before in this book. They are the reason why more data sources were discussed than tested, why more than one dependent variable was selected, and why the discussion of selection criteria and measurement validity criteria pervades throughout almost all sections of this book. Even though my analyses only shed light on some of the possible candidates for distinct operationalisations, I believe that following these and similar guidelines to test and interpret operationalisations of welfare stateness in follow-up research has great potential to improve measurement even more. The findings in this chapter should be used to deduce more indicators fitting the different conceptualisations. Potential additions include transfer share39 for the *Responsive Welfare State*, private-public mix of benefit financing for the A*ssessed Welfare State*, and many more.

Furthermore, the temporal perspective deserves much more attention—especially when it comes to the operationalisation of the *Normative Welfare State.* If the focus rests on the socialising potential of social policies, the set-up of welfare states during the formative years of respondents might bear valuable insights.

Regardless of these restrictions, this chapter demonstrated that taking a step back to conceptualise the chosen analytical perspective on the welfare state before turning to its empirical operationalisation is quite fruitful. It can be integrated

<sup>39</sup> Meaning the share of individuals who are entitled to benefits but do not receive them (e.g. Brady & Bostic 2015).

easily into argumentations, it helps to frame hypotheses, justify operationalisations, and interpret results. At a minimum, it forces to spell out the assumed mechanisms in much more detail. Thus, even though only a part of the proposed indicators proved to contribute to explanations for the selected exemplary outcomes, the potential to standardise the selection by explicitly committing to a systematised conceptualisation is great.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **7 Towards Solving the Independent Variable Problem**

This book started with a reference to literature dealing with transparency and replicability of research. Such literature for the most part highlights issues related to the transparent documentation of methodological proceedings (Damian et al. 2019) and the replicability of existing results (Munafò et al. 2017; Breznau et al. 2019). This book relates in so far to these contributions, as it shows that achieving a transparent, comparable and ultimately replicable result not only means disclosing how an empirical analysis is designed but also spelling out the theoretical conceptualisations that are later included in an empirical test.

The operationalisation of the welfare state as an independent variable proved to be a suitable example to show why it is highly advisable to look in detail at the conceptual aspects behind the operationalisations. It relates to both issues—transparency and replicability: if it is not clearly stated why one operationalisation is preferred over another, the process is not transparent; at the same time, replication of results is not possible if the selection of measurements is not based on comparable conceptual premises. While the operationalisation of the welfare state seems to be particularly affected by these issues, greater attention to the conceptual premises behind the independent variables is likely to be beneficial in other research areas as well.

### **7.1 Summary and Main Findings**

The objective of this book was twofold: to evaluate common approaches to including the welfare state as an independent variable and to develop a proposal for standardising measurement. Two research questions guided this endeavour:

*How comparable are the results that emerge from different approaches to operationalising the welfare state as an independent variable?*

*How can we derive a more standardised, transparent and comparable approach to operationalising the welfare state as an independent variable?*

In order to answer these questions, I provided an overview of how the welfare state is measured in existing research, and why—from a more theoretical point of view—it is not only an important dependent variable but also an essential independent variable in comparative analyses of various objects of research on the individual level. Both aspects are discussed in the second chapter of this book. In the brief introduction to comparative welfare state research and the functions welfare states fulfil—intentionally or unintentionally—the high relevance of the indicator is confirmed.

In the third chapter of this book, this led to the question of how the welfare state is operationalised in current research that treats it as an independent variable and thus to the exploration of the first research question. A review of existing approaches reveals considerable heterogeneity in the existing operationalisations. This heterogeneity is reflected in the results of a brief empirical comparison, which confirms that empirical results are very susceptible even to minor changes in the operationalisation.

Based on this, the answer to the first research questions can be stated very clearly. Comparability of results is frequently impaired on several levels if different operationalisations are treated as interchangeable options. First, the type of selected operationalisation matters. The *single indicator approach*, the r*egime approach* and the *composite index approach* manifest entirely different perspectives on welfare stateness. They conceptualise based on different premises and neither of the approaches can substitute another. Second, within each approach the choice of distinct indicators is crucial. Whether they highlight, e.g. *social rights*, *welfare effort*, or *social investment* represents a commitment to only one perspective that cannot be compared unconditionally to others. Therefore, combining approaches is problematic. Third, multidimensional operationalisations (regime typologies and composite indices) are highly dependent on the respective country sample. The extent of a possible bias is most evident in the *buffet approach* in case of regime typologies where small changes between typological approaches due to mixing or extending classifications already produce severely differing results (cf. section 3.2). This realisation represents an important finding of this contribution: it suggests that the very frequent practice of using welfare regime typologies as independent variables is highly problematic and should be abandoned in the majority of cases. Fourth, comparability can also be distorted by subtler issues such as differing data sources for similar indicators.

All of these issues have in common that they appear almost trivial. At the same time, they are neglected frequently as soon as welfare stateness is treated as an *explanatory variable*. A lack of available macro-level data, a lack of discussion of *how* and *why* welfare stateness is conceptualised in a specific way and the lack of a guideline clearly spelling out possible proceedings for specific research questions are among the explanations for this deficit. All of this raises the importance of discussing not only a dependent variable problem but also an *independent variable problem*, which arises when operationalising features of welfare states in cross-cultural multilevel analyses. Furthermore, this independent variable problem requires its own distinct debate. While we can borrow some insights from macro-sociological comparative welfare state research (such as the distinction between a focus on social rights and expenditure), the mechanisms at work when social policies shape individual-level outcomes represent unique puzzles and require their own reflection.

This answer to the first research question leads directly to asking the second question (*how can we derive a more standardised, transparent and comparable approach to operationalising the welfare state as an independent variable?*). Clearly, what is needed in order to increase comparability is a guideline specifically targeting research endeavours in which welfare stateness serves as an explanatory factor. If existing approaches to including the welfare state as an independent variable are too heterogeneous, produce too deviating results and are not comparable enough—how can we solve this problem? Fathoming this question, I took a step back in the fourth chapter and looked at the research questions we are actually dealing with: what are frequently examined outcomes assumed to be impacted by social policies? What are the reasons for including the welfare state as an independent variable? And, more specifically: which mechanisms are underlying popular hypotheses and how do these mechanisms relate to the functions of welfare states, which warrant their inclusion as an independent variable in the first place? This review resulted in a list of mechanisms and explanations that are either implicitly or explicitly addressed in theoretical argumentations and hypotheses. More importantly, it also revealed two distinct perspectives on the welfare state as an explanatory factor. The first is a *top-down perspective* tied to mechanisms assuming that explanations originate from the factual nature of welfare states (provision of *security*, *redistribution*, *stratification*, *activation*, and *socialisation*). The second—a *bottom*-*up perspective*—captures mechanisms and explanations tying welfare states to outcomes based on the individual perception of social policies—regardless of the accuracy of such perception (*evaluation* and *self*-*interest*). I have discussed in detail whether this list of mechanisms and perspectives is exhaustive or not. In the end, I reach the conclusion that this question is not important for the purpose that they serve in this contribution. This purpose is to be a starting point for a differentiation of analytical perspectives on the welfare state that are embedded in popular research objectives.

This differentiation was the focal point of the fifth chapter where distinct conceptualisations of welfare stateness were explored in more detail. Within the top-down perspective, this revealed at least three systematised concepts: The *Responsive Welfare State,* the *Enabling Welfare State*, and the *Normative Welfare State.* In addition, the *Assessed Welfare State* captures the bottom-up perspective*.* In the course of the chapter, several suitable sets of indicators for each systematised concept of welfare stateness were proposed. The selection was made on the basis of three criteria (*clarity*, *comparability*, and *availability*) and it highlights single indicators rather than composite measures, as this is the only way one can be sure what is actually captured by an indicator. I argued that distinguishing such conceptualisations should considerably help to achieve operationalisations that are more comparable. Instead of trying to capture "the welfare state" in an empirical measurement, it narrows down the operational choices to indicators that represent a distinct concept of welfare stateness. Ultimately, this presents a possibility to standardise the operationalisation in a theoretically meaningful way.

The usefulness of distinguishing between such conceptualisations of welfare stateness was put to the test in the sixth chapter. Here, two popular objects of research—risk of poverty among the working-age population and attitudes towards the welfare state—served as exemplary dependent variables. Indeed, the different systematised concepts proved useful for the theoretical deduction of relevant analytical perspectives on the welfare state and the interpretation of results. They narrow down the broad spectrum of relevant facets of welfare stateness and their functions and thus serve as guidelines for the conceptual discussion. However, it remains somewhat unclear, which indicators capture these concepts best.

The performed analyses revealed some helpful insights. First, some conceptualisations could be captured more successfully than others. Especially the *Enabling Welfare State* could be differentiated clearly from other conceptualisations. Second, the importance of distinguishing the top-down perspective from the bottom-up perspective was clearly confirmed. The performed analyses suggest that in the field of attitude formation the latter—represented by the *Assessed Welfare State*—is of particular importance. Third, capturing the *Responsive Welfare State* and the *Normative Welfare State* proved to be more difficult. In case of the former, this was because almost all indicators of social policy-making potentially represent responsiveness. Still, only benefit coverage allowed a clear interpretation in the sense of securing against risks, while the rest of the indicators (benefit duration, contribution period) failed to produce results that could be allocated clearly to the Responsive Welfare State in either of the two analyses. Regarding the Normative Welfare State, the operationalisation was complicated because measuring a socialising impact of social policy-making is difficult, may require a more nuanced approach and possibly another temporal perspective (for instance, by including features of welfare stateness during formative years of a respondent).

The encountered difficulties in pinning down a definitive list of indicators per conceptualisation should by no means discourage from pursuing the proposed approach. At its very core, this contribution showed that taking a step back to narrow down the concepts behind hypotheses about the impact of social policies on individual-level outcomes is a fruitful and highly valuable endeavour. Against the backdrop of the very problematic inconsistency of approaches in the literature, a major goal must be to achieve better comparability and standardisation. I believe the proposal of using systematised concepts of welfare stateness as an intermediary step before the empirical operationalisation, is able to increase such comparability and standardisation substantially. In short, this is because we need to know which areas within the welfare state are highlighted in our theoretical arguments and hypotheses *before* turning to any kind of measurement. Using a shared conceptual framework will reduce the operational choices to those that are eligible for a specific concept. Moreover, it will force to elaborate in more detail the reasons for choosing one operationalisation over another and it will present a shared reference for discussions reviewing previous research on a given explanandum.

While all of these concluding remarks are promising and worth more attention in the future, they remain fuzzy when it comes to specific operationalisations and concrete practical recommendations. The open questions emanating from the empirical analyses in the last chapter are in so far unsatisfactory, as they still leave us without a clear recipe spelling out which specific indicators to select for which research question. While such a recipe would have been attractive, it may also be premature if only based on the two analyses performed in this book. In order to agree on a distinct set of indicators for specific research questions and hypotheses, much more research is needed. Similarly, the proposed framework including the four deduced conceptualisations of welfare stateness is anything but set in stone. Here, as well, further critical research is needed.

Finally, embedded in the issue discussed in this contribution is a much bigger question: what does the welfare state do? This question is not meant in the sense of mere arrangements of policies, the manner of resource allocation and similar organisational details. It aims at the functions of welfare states, which go well beyond those intended when social policies were first introduced in the late nineteenth century and which are not necessarily intentional at all. Behind the proposed conceptualisations is the implicit assumption that today's welfare states shape individuals in much more complex and subtle ways than they did in the past. It suggests that relying on institutional aspects alone does not fully cover the functions contemporary welfare states serve for individuals—the evidence for the importance of the *Assessed Welfare State* supports this quite well. In this contribution, I treated this issue as being relevant to empirical choices. It may, however, also be a more general theoretical problem, which deserves much more attention. It is my hope that the discussion in this book inspires more research on the matter.

### **7.2 Implications for Further Research**

This contribution highlights a problematic issue, which has not received detailed and systematic attention before. As argued in the second and third chapters of this book, empirical operationalisations of welfare stateness are discussed in comparative welfare state research—but usually not with an emphasis on the consequences of different methodological choices if features of the welfare state are analysed as *independent* variables. Because this specific issue presents a *blackbox* in many ways, even the more detailed discussion in this contribution leaves some aspects unanswered.

From a *conceptual* perspective, the proposed framework and especially the four conceptualisations of welfare stateness have to be explored in more detail. Throughout this book, I always reduced the discussion to exemplary research questions, dependent variables and hypotheses. In many cases, it seems likely that arguments can be translated to other subjects as well, but this remains a postulate until explicitly explored in more detail. Moreover, the addition of other conceptualisations, besides the four discussed here, seems quite possible. From a more general perspective, the universal usefulness of the proposed framework and the intermediary step of narrowing down discussions to distinct conceptualisations of welfare stateness needs to be tested. The mere fact that they proved useful in the examples chosen in this contribution does not yet mean that this is always the case. In the preceding section of this chapter, I have compared the proposed approach to a typology of analytical perspectives that may serve as templates during the research process. Such templates can be very positive and fruitful but they can also present a limitation for research processes if they narrow down proceedings too much. Only further research can show, which of these two options proves to be true.

Room for further research exists especially from an *empirical* perspective. The proposed indicators have to be tested on other data sources, perhaps using other country samples and other analytical approaches. More importantly, however, other indicators representing further targeted operationalisations of specific concepts should be explored. As this contribution shows, this will require looking beyond the more established indicators. Lastly, the proposed framework has to be tested empirically on other outcomes—for instance, behaviour or physical well-being (as discussed in chapter 4). Only if the proposed conceptualisations are translated successfully into empirical operationalisations that fit all research endeavours highlighting similar mechanisms, can they amount to a universally applicable tool.

Besides such points of departure for further research, this contribution also produced several implications in the sense of clear recommendations. The first is an appeal: discussing operationalisations of welfare stateness is not just a dependent variable problem. Instead, it is equally relevant in the case of independent variables—especially because welfare states are assumed to be responsible for outcomes of great socio-political relevance. However, the assumption that the same issues discussed in literature on the dependent variable problem are also relevant in the case of operationalisations as independent variable is wrong and potentially dangerous for the reliability of results and their comparability. This is because established operationalisations—such as typologies—are often unsuitable as independent variables and because the role of social policies in explanations of individual-level outcomes as diverse as well-being, risks, behaviours, attitudes, and more is entirely different compared to *the welfare state* in comparative welfare state research.

The second implication is one that may be relevant for other research objects besides the welfare state as well: the operationalisation of central independent variables is as important as that of the dependent variable. Justifying the selection of a measurement only by its previous use in the current state of research is not sufficiently accurate and it may lead to severe bias in results—again, using welfare regime typologies and the *buffet-approach* comes to mind.

The third implication addresses the general approach towards the welfare state as an independent variable. Evidently, instances in which the entirety of social policy-making is emphasised when explaining why outcomes differ between countries are very rare. Hypotheses highlight specific perspectives on the welfare state (e.g. top-down versus bottom-up), specific functions of welfare states (e.g. activation versus socialisation) and specific areas of policy-making (e.g. unemployment versus family policies). In light of this, it should be obvious that there cannot be one universal operationalisation and that it is almost never "the welfare state" as a whole that is relevant for testing hypotheses. Regardless if it is the framework proposed in this book or another—we must narrow down which parts of social policy-making are relevant in hypotheses and model these specific parts. Only then can empirical analyses be comparable between studies exploring similar dependent variables.

The fourth implication is a very specific one: only those operationalisations seem suitable as independent variables that are *clear*, *comparable*, and *available*. Considering these criteria (in addition to the various other problematic issues), typologies appear to be very problematic candidates for independent variables. Other composite measures are similarly problematic as testing distinct explanations and capturing specific mechanism is almost impossible as long as effects cannot be attributed to clearly definable features of welfare states. Some single indicators—such as social expenditure—are partly affected by similar problems. Overall, single indicators present the most recommendable choice as they allow narrowing down results to specific aspects.

This book approached an issue, which has received too little attention so far and this oversight is indeed problematic. The welfare state is an important independent variable for the explanation of a multitude of objects of research. Such research is only comparable if it follows a transparent and replicable operationalisation of welfare stateness as an explanatory factor. After it became evident that such a standardised approach could not be found in existing methodological contributions within comparable welfare state research, this contribution addressed the issue from an entirely different perspective. By focussing in much detail on the conceptualisation of welfare stateness as an explanatory concept, it proposes a new intermediary step, which serves as a tool for narrowing down distinct perspectives on welfare stateness. This intermediary step entails a strong focus on pinpointing distinct conceptualisations of welfare stateness before an operationalisation takes place. Explicitly committing to a conceptualisation and explicitly modelling it empirically will increase the transparency of approaches as well as the comparability of results. While this contribution leaves open several questions and raises several more, it still shows the importance of discussions on this issue. Such discussions will hopefully continue in further research, including a critical assessment of the framework proposed in this book and fruitful additions to the list of empirical choices for the operationalisation of distinct conceptualisations of welfare stateness.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **References**


© The Editor(s) (if applicable) and The Author(s) 2023

K. Kunißen, *The Independent Variable Problem*, Sozialstrukturanalyse, https://doi.org/10.1007/978-3-658-39422-6

