Labour market protection across space and time: A revised typology and a taxonomy of countries’ trajectories of change

Abstract

We measure and interpret the evolution of labour market protection across 21 high-income countries over three decades, employing as conceptual foundations the ‘regime varieties’ and ‘trajectories of change’ developed by Esping-Andersen, Estevez-Abe, Hall and Soskice, and Thelen. We measure labour market protection considering four institutional dimensions – employment protection, unemployment protection, income maintenance and activation – and the evolution of the workforce composition. This measurement accounts for the joint evolution of labour market institutions, their complementarities and their relation to outcomes, and mitigate the unrealistic Average Production Worker assumption. We handle the multi-dimensional nature of labour market protection with Principal Component Analysis and capture the characteristics of countries’ trajectories of change with a composite score. We contribute to the literature in three ways. (1) We portray a revised typology that accounts for processes of change between 1990 and 2015, and that clusters regime varieties on the basis of coordination and solidarity levels, that is, Central/Northern European, Southern European, liberal. (2) We illustrate that, despite a persistent gap, a large majority of Coordinated Market Economies experiencing a decline in the level of labour market protection became more similar to Liberal Market Economies. (3) We develop a fivefold taxonomy of countries’ trajectories of change (liberalization, dualization, flexibility, de-dualization and higher protection), showing that these trajectories are not always path-dependent and consistent with regime varieties previously developed in the literature.

Keywords

labour market protection worlds of welfare varieties of capitalism liberalization dualization principal component analysis new social risks varieties of liberalization

Introduction

This article measures labour market protection across time and space and employs the conceptual foundations of ‘regime varieties’ and ‘trajectories of change’ to interpret its evolution in 21 high-income countries over three decades.

Our approach – taking stock of Varieties of Capitalism (VoC) and The Three Worlds of Welfare Capitalism (WoW) frameworks (Esping-Andersen, 1990; Estevez-Abe et al., 2001; Hall and Soskice, 2001; for an integration of these typologies, see Schröder, 2009) – considers labour market protection institutional arrangements and outcomes holistically. However, regime varieties focus on similarity and dissimilarity in specific points in time; consequently, clusters are built in relative terms to one another. For this reason, we complement this approach with an investigation of countries’ trajectories of change in absolute terms. Several streams of literature highlight how deregulation, re-commodification, recalibration of unemployment benefits and income maintenance, and the shift from compensatory to active forms of protection modified labour market protection levels. Thelen (2012, 2014) builds upon these insights and defines three typical trajectories of labour market liberalization related to regime varieties. These trajectories have been scrutinised only in a few countries and by using a limited number of indicators. Although bridging static and dynamic ways of looking at labour market protection and combining regime varieties and trajectories of change, we argue that this analysis can be systematised empirically in a Large-N setting and further refined conceptually.

Against this backdrop our research questions proceed as follows: have labour market protection varieties changed substantially across three decades? Have certain trajectories of change been overlooked? To what degree are countries’ trajectories path-dependent and consistent with regime varieties previously developed in the literature?

The investigation of these questions requires the formulation of a more complete picture of labour market protection than is customary. We classify labour market protection varieties and countries’ trajectories considering four institutional dimensions and the evolution of the workforce composition employing 13 indicators (Supplementary Appendix Table A1). The addition of the workforce composition to the institutional dimensions helps us to mitigate the unrealistic Average Production Worker (APW) assumption (Clasen and Siegel, 2007). VoC and WoW assume that labour market protection is homogenous across the population and measure it for the APW only. However, the APW assumption underplays the importance of significant societal changes, that is, the shift from the male-breadwinner to a variety of adult worker models (Daly, 2011), the advent of a service-based economy (Jessop, 1993), deep welfare state changes (Hay and Wincott, 2012; Hemerijck, 2013; Pierson, 1994), the growing divide between labour market ‘insiders’ and ‘outsiders’ (Emmenegger et al., 2012; Rueda, 2005) and the rise of new social risks (NSRs) (Bonoli, 2005; Taylor-Gooby, 2004).

We measure labour market protection varieties and countries’ trajectories of change using two methods. The first, a series of Principal Component Analyses (PCA), handles the multi-dimensional measure of varieties; the second, a composite score, captures the characteristics and depth of countries’ trajectories of change. The analysis of these trajectories focuses on Coordinated Market Economies (CMEs) because Liberal Market Economies (LMEs) before the 1990s had undertaken a liberalization trajectory; LMEs witnessed minor reductions of labour market protection levels in the period from 1990 to 2015 if compared to CMEs. The PCAs devise a Cartesian space that follows closely the conceptualization of regime varieties and trajectories of change previously developed in the literature. The horizontal axis distinguishes CMEs and LMEs, while the vertical axis underscores important differences in the levels of labour market protection across social democratic, Christian democratic, Mediterranean and liberal regimes. The space delineated by these axes retraces the conceptual map Thelen devised to interpret countries’ trajectories according to coordination and solidarity levels.

Our comparative analysis contributes to the literature in three ways:

(1) Developing a revised typology that accounts for processes of change between 1990 and 2015, and that clusters regime varieties on the basis of coordination and solidarity levels, that is, Central/Northern European, Southern European and liberal;

(2) Illustrating that, despite a persistent gap, a large majority of CMEs, experiencing a decline in the level of labour market protection, became more similar to LMEs;

(3) Proposing a fivefold taxonomy of countries’ trajectories of change – liberalization, dualization, flexibility, de-dualization and higher protection – which shows that these trajectories are not always path-dependent and consistent with regime varieties previously developed in the literature.

Our findings can be read at two interconnected levels. The first presents systematic empirical evidence about labour market protection generosity and change in high-income countries. Change is disentangled empirically (for more detail see the appendix material), then analytically recomposed with our PCAs and the composite score. The second goes beyond ‘mere description’ (Gerring, 2012) and provides scholars with a revised typology and a new taxonomy of change, which connect the rich empirical evidence for each country and indicator to our conceptual foundations.

The remainder of the article proceeds as follows. We first place our research questions within the literature and formulate our hypotheses. Then we describe our methodological approach and discuss our findings. We conclude by interpreting these findings in relation to regime varieties and trajectories of change.

Understanding varieties and trajectories of change

VoC and WoW classified high-income countries and compared cross-national similarities and differences in labour market protection (among other policy domains). However, both frameworks are insufficiently sensitive to population heterogeneity and, over time, trajectories of change. There is indeed a tension between the static notion of variety and the exploration of institutional change (Daly and Ferragina, 2018; Ferragina and Seeleib-Kaiser, 2011). Our work deals with this tension and provides a systematic analysis of labour market protection across space and time.

VoC classified high-income countries in CMEs and LMEs, and connected the national competitive advantage to the prevalent mode of coordination between firms and various institutions (Hall and Soskice, 2001). Accordingly, market forces did not lead countries to convergence, rather the interplay between firms and institutions produced two different varieties of capitalism: in CMEs there is a high degree of cooperation between firms and institutions, while in LMEs we observe low power-sharing relations and a lack of coordination between firms and institutional actors. Estevez-Abe et al. (2001) connected VoC to labour market protection reflecting on the skills formation process. This process, they argue, contributes to the national competitive advantage in the long-run, as employees’ decision of skills investment relates to levels of labour market protection (employment, unemployment and wage protection). In CMEs, where firms base mostly their competitive advantage on the presence of (industry/firm) specific skills, workers require higher employment and/or unemployment protection to safeguard their investment in specific skills. In contrast, in LMEs, where firms base mostly their economic competitive advantage on general skills, employment and unemployment protection are less generous. In sum, the interaction between the national competitive advantage, that is, the coordination between firms and institutions and individual incentives to invest in certain skills, shape labour market protection.

WoW suggests that welfare states in high-income countries cluster in three varieties – liberal, Christian-democratic and social democratic – grounded in the long-standing outcome of class conflict. Esping-Andersen employed the notion of decommodification – that is, the capacity of pension, sickness and unemployment benefits to guarantee individuals and families with ‘a socially acceptable standard of living independently of market participation’ (Esping-Andersen, 1990: 37) – and the ability of welfare states to reduce social stratification as measures of welfare state generosity. In the social democratic world, universalism is prevalent and countries within it are characterised by a high decommodification capacity and low social stratification. In the Christian democratic world, social insurance is prevalent and countries within it are characterised by a medium decommodification capacity and high social stratification. In the liberal world, social assistance is prevalent and countries within it are characterised by a low decommodification capacity and high social stratification. This typology has been debated, criticised, re-assessed and extended over time (for a review, Arts and Gelissen, 2002; Ferragina and Seeleib-Kaiser, 2011; for an empirical assessment, Scruggs and Allan, 2006; for a feminist critique, Lewis, 1992). For the sake of our analysis – which also includes Greece, Spain and Portugal – we note that scholars theorised the existence of a Mediterranean cluster (Ferrera, 1996) distinguished from the Christian democratic world because of lower solidarity levels, persistent clientelism and inadequate income assistance.

A debated assumption of both VoC and WoW is that cross-country differences are stable over time.¹ However, it has been noted that labour market protection in high-income countries underwent severe processes of reform. While these processes took place mostly across LMEs during the 1980s – Reagan and Thatcher’s reforms served as the blueprint – they have unfolded in CMEs only since the 1990s. Labour market protection reforms were influenced by the traction of liberalism, with globalization and the European integration process which fostered a ‘contingent convergence’ also among CMEs (Baccaro and Howell, 2017; Hay, 2004). In this context, the European Employment Strategy has been implemented to strike a balance between the objectives to increase competitiveness and maintain social cohesion (European Commission, 1997; Keune and Jepsen, 2007).

The Flexicurity strategy (Wilthagen, 1998; Wilthagen and Tros, 2004; for a review Viebrock and Clasen, 2009) became prominent as a way to handle these objectives simultaneously. On the one hand, increasingly deregulated labour markets were deemed to better respond to economic fluctuations, enhancing systemic performance and increasing employment levels (OECD, 1994). This fostered a decrease of employees’ protection and the decentralisation of wage bargaining systems (Esping-Andersen and Regini, 2000). On the other hand, this deregulation should have been compensated by a social security leg, to activate the labour force and reduce the cost of compensatory policies (Bonoli, 2010). However, while the reduction of compensatory policies translated into re-commodification and recalibration of unemployment benefits and income maintenance schemes (Neyer and Seeleib-Kaiser, 1995; Pintelon et al., 2013; Seeleib-Kaiser, 2002), activation policies remained mostly underdeveloped. Scholars also criticised the implementation of the flexicurity strategy, as reforms were often uncoordinated and unable to foster positive institutional complementarities (Burroni and Keune, 2011), and undermined the capacity of labour markets to face external shocks – the so-called ‘labour market resilience’ (Hijzen et al., 2017; OECD, 2012; for a review, Bigos et al., 2013). These issues became more salient in the aftermath of the Great Recession, when countries, after embracing initial short-term Keynesian-type stimulus, implemented a wave of austerity measures, by cutting social protection and further deregulating the labour market² (Blyth, 2013; Hemerijck, 2013; Heyes, 2013; Leschke, 2012).

Thelen (2012, 2014) associated VoC and WoW insights and these accounts of change by identifying three typical trajectories of labour market liberalization. According to her, countries’ trajectories of change are path-dependent and consistent with regime varieties previously identified in the literature. She argued that liberal countries (for example, the United States) followed a ‘deregulation’ strategy, which consisted in a classic form of liberalization that reduced social protection, dismantled the extent of residual coordination capacity between employers and the governments, and weakened trade unions. Christian democratic countries (for example, Germany) followed a ‘dualization’ strategy, which entailed the maintenance of strong protection and coordination levels for the core of the workforce and the unloading of liberalization costs onto the shoulders of labour market outsiders. Finally, social democratic countries (for example, Denmark) followed an ‘embedded flexibilization’ strategy, which led to an adaptation to liberalization based on ‘supply side solidarity’ and the collectivization of ‘risk by focusing resources on enabling society’s most vulnerable to get and keep a job’ (Thelen, 2014: 15).

We contribute to comparative social policy and political economy literature, empirically systematising and refining conceptually labour market protection varieties and countries’ trajectories of change. We envisage that VoC and WoW especially have lost some of their descriptive power over three decades. Following contingent convergence, Christian democratic and social democratic countries might be more similar in their level of labour market protection in 2015 than in the 1990s. In addition, the distinction between labour market protection levels in LMEs and CMEs should remain substantive, but following noticeable liberalization processes in CMEs during the 1990s and 2000s, we expect a reduction of the gap in protection levels. This approach grounded in regime varieties might underplay temporal change and hide absolute variations of labour market protection across countries. For example, a country might witness a substantial reduction of labour market protection generosity between 1990 and 2015, but might not move to a different regime variety because its labour market protection level remains higher than that of countries belonging to other clusters. For this reason, and differently to Thelen, we hypothesise that countries’ trajectories of labour market protection change could be heterogeneous and not necessarily be consistent to regime varieties previously developed in the literature.

Data and methods

Data

Our measure of labour market protection includes 13 indicators (Supplementary Appendix Table A1), and mixes insights developed in social policy and political economy to describe labour market protection change – deregulation, re-commodification, the shift from compensatory to active policies, dualization, and new social risks. To account for this change, we include four institutional dimensions of labour market protection – employment protection, unemployment protection, income maintenance and activation – and the main characteristics of workforce composition.³ From this, we provide a detailed account of the joint evolution of labour market institutions, their complementarities and their relation to outcomes, that is, ‘labour market resilience’.

We capture trends towards employment protection deregulation using the Employment Protection Legislation (EPL) index for permanent and temporary contracts, the percentage of workers covered by collective bargaining agreements and union density. We measure re-commodification of unemployment protection with unemployment benefit replacement rates (as a share of the APW’s previous wage) and public spending for unemployment benefits and severance pay. We suggest that re-commodification relates also to the income maintenance dimension; a measure targeted to those who might not be eligible for unemployment benefits (for example, atypical workers, long-term unemployed and those below the poverty line). We measure this dimension through the adequacy of guaranteed minimum income benefits and public spending for income maintenance schemes. To capture the shift from compensatory to active forms of protection, we employ a measure of public spending for Active Labour Market Programmes (ALMPs). Finally, taking stock of the critique to the APW assumption developed within NSRs and dualization literatures – and to account for the interdependence between labour market protection and labour market outcomes – we include in the analysis the unemployment rates, the share of workers holding a permanent or temporary contract, and the share of workers involuntarily employed with a part-time contract (a proxy for bad quality jobs). Other scholars have employed outcome indicators in the regime debate (Ferragina et al., 2015). Supplementary Appendix Table A1 lists all indicators and data sources.

Methods

We employ PCA because it appropriately handles our multi-dimensional measurement of labour market protection (Ferragina et al., 2013; Jolliffe, 2002). In order to investigate the inter-temporal validity of classic varieties and observe if countries depart from their cluster, we assume a priori that countries belong to WoW typologies.⁴ Hence, Denmark, Finland, Norway and Sweden are assigned to the social democratic group; Austria, Belgium, France, Germany, the Netherlands and Switzerland to the Christian democratic group; Greece, Italy, Portugal and Spain to the Mediterranean group; and Australia, Canada, Ireland, Japan,⁵ New Zealand, the United Kingdom and the United States to the liberal group. PCA creates uncorrelated dimensions, the Principal Components (PCs), that are linear combinations of the indicators employed for the analyses. The PCs are sorted in descending order according to the share of information they explain. The first PC provides more information than the second PC, and the portion of information decreases with each successive PC until it becomes negligible. Our bi-plots are based on the first two PCs. Each indicator contributes to the definition of the two axes of our bi-plots (the two first PCs) according to a correlation index that ranges from −1 to +1. Countries with high (positive or negative) values for the indicators that are strongly correlated with the two PCs are positioned at the extremities of the bi-plot, while those with values close to the average of the distribution are placed nearby the origin. The interpretation of the Cartesian space is based on a holistic reading of our indicators⁶ and helps us to evaluate the persistency of the descriptive power of previous conceptualizations.

We also construct a composite score to evaluate the depth and characteristics of countries’ trajectories independently from classic varieties. We focus on CMEs only because LMEs mostly liberalised their labour market protection before the 1990s. We calculate the distance between the country values in 1990 and 2015 for each indicator, and the corresponding value of the United States in 1990. We employ labour market protection in the United States as the benchmark because it was among the first countries that liberalised labour market protection and, for this reason, it displayed the lowest level of protection in 1990. This benchmark allows us to standardize the variation of different indicators. Then we calculate the difference between each relative distance in 2015 and in 1990, obtaining a series of sub-scores. If the variation of the relative distance is positive, the country observed an increase in the labour market protection dimension in 2015. Conversely, if the country displays a negative score, the generosity of the labour market protection dimension diminished. Since most of our dimensions include multiple indicators, we average the values for the indicators contained in each dimension to calculate the sub-scores. However, when defining the trajectories, we also consider every indicator (Supplementary Appendix Table A6). For example, comparing the change in the EPL for permanent and temporary workers allows us to distinguish between liberalization and dualization. We then construct the sub-scores for our outcome indicators, computing the relative deviation for each indicator between 1990 and 2015. Finally, we averaged each sub-score to compute the Labour Market Protection Change⁷ (LMPC) score and defined a series of trajectories accordingly.

The limitations of our measurement relate to data availability and their characteristics. In some cases, we had to manually impute missing values for specific years (using the score of the closest year, Supplementary Appendix Table A5). Also we had to exclude Eastern European countries, as data series are incomplete. Moreover, comparative data cannot cover all characteristics of labour market protection. For example, extensive comparative evidence exists on unemployment protection conditionality (Scruggs et al., 2017), but this is not available for income maintenance. Hence, to construct a harmonised measurement, we employed data on generosity and spending only. The measurement of activation relies solely on spending data, as there are not comparative data on coverage and training-to-job transition. Individual-level data might have helped to partially deal with these issues, for example, EU-SILC survey data have been used to complement macro institutional indicators (Otto and Van Oorschot, 2019). Unfortunately, these data have only been available since the early 2000s and exclude non-EU countries. Finally, we acknowledge that there is a trade-off between ‘quantity’ – the geographical and temporal extension of the comparison – and ‘quality’ – the accuracy in the study of policy change that can be reached studying reform processes in a Small-N setting. Taylor-Gooby (2002) illustrated that qualitative and historical case studies are more sensitive to policy change than Large-N quantitative comparisons. This means our analysis might underplay the depth of labour market protection change.

Scrutinizing institutional varieties across space and time

The horizontal axis of our PCAs⁸ is structured around the positive correlation between employment protection, unemployment protection, activation, the percentage of temporary contracts⁹ and the first PC (Supplementary Appendix Table A2). The horizontal axis echoes VoC: CMEs – which display the highest levels of employment protection, unemployment protection, activation and the highest share of temporary workers – are located at the right-hand side of the bi-plot, while LMEs – with lower values for these dimensions – stand on the left-hand side. The vertical axis is structured around the positive correlations between income maintenance, activation and the second PC¹⁰ (Supplementary Appendix Table A2). Therefore, countries with higher levels of solidarity, better-financed activation programmes and labour market outcomes (especially in 2015) are close to the top, while those with lower levels of solidarity, cheaper activation programmes, and more deteriorated labour market outcomes are nearer the bottom. The vertical axis echoes WoW, distinguishing different levels of protection across countries. Overall, the bi-plot is similar to the space Thelen (2014: 7) devised ideally to connect VoC with the liberalization processes. In the first quadrant, we find countries with high levels of coordination and solidarity; in the second, those with low levels of coordination and high solidarity; in the third, those with low levels of coordination and low solidarity; and in the fourth, those with low levels of solidarity and high levels of coordination (Supplementary Appendix Figure A1).

Clusters in the 1990s

Within the social democratic cluster, positioned in the first quadrant (with the exception of Finland), countries had a strong labour market protection in all dimensions (Supplementary Appendix Figure A2 and Table A3) with high levels of solidarity and coordination (Figure 1). Finland and Sweden were close to the horizontal axis, while Denmark and Norway were placed near the vertical axis. The position of Finland and Sweden is due to the strong employment protection for permanent contracts¹¹ and high shares of temporary workers; Sweden had also higher scores than the social democratic average for employment protection, unemployment protection and activation. Denmark had the highest level of income maintenance and, like Norway, was close to the Christian democratic average for most other indicators.

The Christian democratic cluster was positioned near the origin of the axes. France, Belgium and Netherlands were in the first quadrant close to the social democratic countries. France was the closest to the origin, with levels of employment protection, unemployment protection and activation higher or similar to the Christian democratic average; it had also the most deteriorated labour market outcomes within the cluster. Belgium, positioned in the middle of the quadrant, had a stronger employment protection,¹² income maintenance and activation; it had the lowest share of temporary contracts in the cluster. The Netherlands had similar characteristics to Belgium but noticeably had the highest level of income maintenance and share of involuntary part-time contracts in the sample;¹³ Goodin (2001) referred to this feature as a ‘post-productivist possible utopia’ (see also, Visser and Hemerijck, 1997). Switzerland and Austria occupied the third quadrant, and revealed some liberal traits. Switzerland had scores for employment protection, income maintenance, spending for unemployment benefits¹⁴ and activation lower than the cluster average (Armingeon et al., 2004; Obinger, 1999); it had also a low unemployment rate, a low share of permanent and involuntary part-time contracts compared to the other Christian democratic countries. Austria, despite a strong level of employment protection for permanent contracts, had a level of income maintenance and spending for activation lower than the group average. Like liberal countries, Austria had a low unemployment rate, a low share of involuntary part-time contracts, and a high incidence of permanent contracts. Germany, positioned in the fourth quadrant, had strong employment protection, but unemployment protection and income maintenance were below the Christian democratic average.

The Mediterranean cluster was characterized by strong employment protection, low levels of income maintenance, and high unemployment rates (Ferrera, 1996); therefore, high levels of coordination – especially in Spain and Portugal – but low solidarity. Accordingly, Mediterranean countries were scattered at the bottom of the Cartesian space, with the exception of Spain. Spain was placed at the extreme right of the first quadrant, because of a strong employment protection and the highest share of temporary contracts in the sample (Dolado et al., 2002). Greece and Portugal occupied the middle of the fourth quadrant. However, a comparison between these two countries reveals that Greece had higher EPL for temporary contracts, unionization rate, coverage of collective bargaining agreements and spending for unemployment benefits (see Petmesidou, 1996), whereas Portugal had higher EPL for permanent contracts, unemployment benefits and minimum income replacement rates, and spending for activation (see Pereirinha, 1996). Moreover, the share of permanent contracts was lower in Portugal, which explains its position closest to the first quadrant in comparison to Greece. Italy was positioned at the bottom of the third quadrant close to the vertical axis, and had less EPL for permanent contracts than the Mediterranean average, the strongest EPL for temporary contracts of the whole sample, and high unionization rates. Italy, as with liberal countries, also had low unemployment protection, a high share of permanent contracts and low spending for activation; its income maintenance dimension, as with all Mediterranean countries, was underdeveloped.

The Liberal cluster occupied the second and third quadrants (Australia, Canada, Ireland, Japan and New Zealand in the second; United Kingdom and United States in the third). Low employment protection and a high share of permanent contracts are common features of this cluster, and this confirms that countries liberalised their labour market protection prior to our 1990 measurement. Countries positioned in the third quadrant had a lower unemployment protection, income maintenance and activation spending than those placed in the second. Australia, positioned at the far left of the second quadrant, had the highest share of permanent contracts in the sample, and had employment protection and income maintenance higher than the liberal average (Castles, 1996); its unemployment protection and activation spending were below the liberal average. Canada, at the top of the second quadrant, had low levels of employment protection, a smaller share of permanent contracts than the liberal average (Cooke and Zeytinoglu, 2004); it had high unemployment protection, income assistance and activation spending. Ireland, positioned in the middle of the second quadrant, had strong EPL for permanent contracts and union density, but low EPL for temporary workers and non-inclusive collective bargaining agreements. Due to high unemployment rates, spending for unemployment benefits for Canada was almost twice that of the liberal average, even if replacement rates were below the liberal average; the minimum income replacement rate was above the liberal average and spending was two times lower. New Zealand had similar levels of employment protection and unemployment protection to Ireland, and a similar minimum income replacement too, but had lower spending than Ireland for income assistance and activation policies. Japan, positioned in the middle of the second quadrant, had the highest EPL among liberal countries, but lower unionization rates and less inclusive collective bargaining agreements. Replacement rates for unemployment benefits and minimum income were higher than the liberal average, but the associated spending was lower (also spending for activation). The low unemployment rate was accompanied by a low share of permanent contracts. The United Kingdom and the United States occupy the extreme left of the third quadrant, and had the lowest levels of solidarity and coordination in the sample. The United Kingdom is positioned above the United States, and had slightly higher protection levels than the United States in all dimensions; moreover, it displayed a higher share of involuntary part-time contracts.

Inter-temporal change of clusters (1990–2015)

The varieties of labour market protection have changed significantly during the last three decades (Figure 1; Supplementary Appendix Table A4). By 2015, social and Christian democratic countries have grouped mostly in the first quadrant; liberal and Mediterranean countries have grouped more closely together and occupy the third and fourth quadrants respectively. The social and Christian democratic clusters overlap, but Denmark is an exception, for it occupies the top of the vertical axis due to the most developed flexicurity strategy among high-income countries. Belgium, Sweden, Finland and the Netherlands are positioned in the middle of the first quadrant and not far from them and close to the vertical axis are Norway and Austria. Austria, Switzerland¹⁵ and Germany have moved towards the centre of the bi-plot, compared to their position in 1990. The most important change within the Mediterranean cluster is the deterioration of labour market outcomes,¹⁶ with an increase of the unemployment rate (especially in Greece and Spain) and the share of involuntary part-time contracts (especially in Italy). Spain, Italy and France (that joined the Mediterranean cluster because of poor labour market outcomes) are positioned closer to Greece and Portugal than in 1990. All Liberal countries – Australia, Canada, Ireland, Japan and New Zealand – now cluster in the same space¹⁷ with the United States and United Kingdom. These countries continued to deregulate employment protection marginally, with a decline of union density and collective bargaining agreements, and witness a deterioration of labour market outcomes (with an increase in the share of temporary and involuntary part-time contracts).

Countries^’ trajectories

Overall the composite score portrays a shift towards liberalization in CMEs between 1990 and 2015¹⁸ (Table 1; Supplementary Appendix Table A6). However, there are important cross-national differences in the intensity and trajectory of change. CMEs are closer to the United States in 2015 than 1990, largely because they deregulated their employment protection. EPL declined in most countries, a drop that is three times higher for temporary contracts. The decline of collective bargaining agreements and the decrease in union density is less severe than EPL, although – especially in the case of union density – highly generalised. In addition, labour market outcomes worsened almost everywhere because the share of unemployment, temporary and (especially) involuntary part-time contracts have increased; in contrast, unemployment protection, income maintenance and activation have been slightly expanded. However, all that glitters is not gold. The scores for unemployment protection and income maintenance are positive because spending has increased, while replacement rates have remained stable. Hence, generosity levels did not averagely expand, but rather the deterioration of labour outcomes have translated into higher spending levels. But if we exclude Denmark – an outlier with a significant spending increase – activation spending averagely has declined across CMEs.

Table 1.

Labour market protection change (LMPC) score.

Country	Employment protection	Unemployment protection	Income maintenance	Activation	Workforce composition	LMPC score	Trajectory
Sweden	−3.73	−0.70	−0.52	−1.37	−0.89	−1.44	Liberalization
Spain	−2.78	−1.34	0.79	−0.81	−1.82	−1.19	Liberalization
Greece	−3.95	0.12	0.21	0.37	−1.68	−0.99	Liberalization
Germany	−2.66	0.10	0.89	−1.10	−1.12	−0.78	Dualization
Norway	−0.27	−0.73	−1.28	−1.74	0.31	−0.74	Liberalization
Belgium	−2.20	0.57	0.23	−1.71	−0.17	−0.65	Dualization
Italy	−3.01	0.99	0.11	1.39	−1.84	−0.47	Dualization
Portugal	−3.48	0.88	0.94	0.43	−1.10	−0.47	Liberalization
Netherlands	−0.67	−1.07	2.11	−2.04	−0.36	−0.41	Liberalization
Japan	−1.40	−0.15	1.12	−0.84	−0.20	−0.29	Liberalization
Switzerland	−0.10	0.87	1.54	1.86	−2.64	0.31	Flexicurity
Austria	−0.72	0.40	1.61	2.01	−0.87	0.49	De-dualization
Finland	−0.32	2.12	1.30	0.83	−0.79	0.63	De-dualization
Denmark	−1.82	−0.11	0.70	6.35	0.02	1.03	Flexicurity
France	0.62	0.47	3.12	1.41	−0.34	1.06	Higher protection
Mean	−1.77	0.16	0.86	0.34	−0.90	−0.26

Note: in bold the indicators displaying negative (sub-)scores. Values for the four labour market protection dimensions are expressed as number of standard deviations from the value of the United States in 1990; values for the workforce composition are expressed as the relative deviation of each country from 1990; data for minimum income spending for Greece is missing, the LMPC score for the income maintenance dimension is computed on the replacement rate only.

We observe five trajectories of change in labour market protection, characterised by negative (liberalization and dualization) and positive (flexicurity, de-dualization and higher protection) composite scores. Ten CMEs followed liberalization and dualization trajectories. In these countries, labour market protection in 2015 was lower than in 1990. The three other trajectories – flexicurity, de-dualization and higher protection – concerned only five CMEs; these countries strengthened their labour market protection during the last three decades (Table 1).

A liberalization trajectory entails a deregulation in employment protection and a reduction in protection due to a decline of some other dimension. We employ the notion of liberalization instead of deregulation (Thelen, 2014) because we observe a deeper process of change than the simple reduction of EPL. A dualization trajectory indicates a deregulation of temporary contracts stronger than the deregulation for permanent contracts. Dualization is associated also with an increase of temporary and involuntary part-time contracts (Rueda, 2005). A flexicurity trajectory encompasses a deregulation of employment protection, stability for unemployment protection and income maintenance; or when these two dimensions are in decline, a symmetric increase in spending for activation can be seen. We do not use the notion of ‘embedded flexicurity’ (Thelen, 2014) because the flexicurity trajectory can also be applied to a country that does not belong to the social democratic tradition. To account for the observed increase in labour market protection, we add two new trajectories to the three previously defined in the literature – de-dualization (the opposite to dualization) and higher protection (the opposite to liberalization). A de-dualization trajectory is similar to flexicurity but the reduction in EPL is stronger for permanent than temporary contracts. A higher protection trajectory entails a re-regulation of employment protection, and an increase of decommodification and higher spending for activation.

Sweden, Spain, Greece, Norway, Portugal, Netherlands and Japan followed a liberalization trajectory. Among them, we can identify two slightly different liberalization modalities. Spain, Portugal and Norway underwent a linear liberalization trajectory that took place without substantial difference for temporary and permanent contracts (confirming previous analyses for Spain, Picot and Tassinari, 2017 and Prosser, 2016). In Portugal and Spain, liberalization intervened in previously highly dualized labour market contexts (Cárdenas and Villanueva, 2021; Cardoso and Branco, 2018). Portugal decreased employment protection, and expanded unemployment protection and activation spending. Spain witnessed a decrease in all labour market protection dimensions,¹⁹ and especially employment protection (with a decline of coverage for collective bargaining agreements). Although to a lesser extent than Portugal and Spain, Norway pursued a classic liberalization trajectory; moreover, differently from all other countries in the sample – and with Denmark – it experienced an improvement of labour market outcomes.

Sweden, Greece, Netherlands and Japan followed a liberalization trajectory with a tendency towards dualization.²⁰ Sweden underwent a generalized retrenchment of labour market protection (Anderson, 2001; Lindbom, 2001). Our data indicate a light deregulation of EPL for permanent contracts, a reduction of unemployment protection, income maintenance and activation accompanied by harsh deregulation of temporary contracts (the highest decrease in the sample). Sweden is the CME that liberalised the most, confirming a decline in its coordinating capacity and social solidarity (Fleckenstein and Lee, 2017). The Great Recession and subsequent austerity measures boosted the Greek liberalization trajectory (Matsaganis, 2012; Prosser, 2016). The deregulation of EPL is the deepest of all CMEs and is driven mostly by a reduction of the employment protection for temporary contracts. Greece also experienced a steep increase of unemployment rate and share of involuntary part-time contracts. In apparent countertendency to the liberalization trajectory, we also observe an increase in unemployment protection, income maintenance and activation, though in reality this moderate increase is due to the low levels (close to zero) of protection displayed by the country in the 1990s. Labour market protection remains very weak overall (Matsaganis, 2018). In the Netherlands, liberalization took place through a reduction of all labour market protection dimensions, with the exception of income maintenance. We witness a deregulation of temporary contracts and also their exceptional numerical increase (the second highest in the sample after Italy). In Japan, the liberalization trajectory hit mostly labour market outsiders in a context of worsening labour market outcomes (Coe et al., 2011). The liberalization trajectory is underpinned by a deregulation of EPL, a decentralization of wage bargaining agreements, and a reduction of unemployment protection and spending for activation policies.

Germany, Italy and Belgium followed a dualization trajectory. This trajectory is punctuated by a strong deregulation of EPL for temporary contracts coupled with a slight increase (Belgium and Germany) or decrease (Italy) of protection for permanent contracts. Unemployment protection and income maintenance have increased slightly²¹ (mostly because of a spending increase related to worsening labour market outcomes). Spending on activation policies decreased both in Germany and Belgium, yet it increased in Italy. However, spending for activation in Italy remains much lower than in the other two countries. Our findings confirm that Germany is a prototypical case of dualization (Eichhorst and Marx, 2011; Emmenegger et al., 2012; Seeleib-Kaiser and Fleckenstein, 2007) and that Italy followed a similar path (Berton et al., 2012; Emmenegger, 2014). However, it is important to note that our analysis does not capture the 2015 labour market reform (the so-called ‘Jobs Act’) which, according to some scholars, transformed dualization into liberalization (Ferragina and Arrigoni, n.d.).

Denmark and Switzerland followed a flexicurity trajectory (Andersen, 2012; Fossati, 2018). The Danish flexicurity strategy was put in practice by deregulating EPL for temporary contracts, safeguarding unemployment protection and income maintenance, and expanding activation spending at the highest level among OECD countries (Green-Pedersen et al., 2001). Switzerland had an already deregulated EPL in 1990 and increased its active and compensatory forms of labour market protection. If we compare the two countries, we can say that Denmark followed a trajectory grounded in ‘embedded flexibility’ (Thelen, 2014), reforming labour market protection in continuity with the social democratic model; Switzerland followed a more liberal pathway, starting to expand a previously meagre generosity level. Austria and Finland followed a de-dualization trajectory, deregulating EPL for permanent contracts while maintaining (Austria, see Rathgeb, 2017 on ‘smoothed dualization’) or increasing (Finland) EPL for temporary contracts. Moreover, unemployment, income maintenance and activation were expanded to counter the deterioration of labour market outcomes. France, despite a further deterioration of labour market outcomes (especially for young people, see Chevalier, 2016), is the only CME which displayed positive sub-scores for all institutional dimensions.

Discussion and conclusion

The article investigates systematically and refines conceptually labour market protection varieties and countries’ trajectories of change. Specifically, we adjudicate how labour market protection varieties changed over three decades in 21 countries, and propose a revised typology. We explore also to what extent labour market protection trajectories are path-dependent and consistent with regime varieties previously identified in the literature, and establish a new fivefold taxonomy. In addition, from a theoretical perspective, we suggest that WoW and VoC frameworks are useful departing points to analyse labour market protection varieties; however, they need to be considered cum grano salis. Their heuristic quality does not imply a perpetual ability to capture the varieties of labour market protection.

Our PCAs show that the WoW framework has lost some of its descriptive power recently; this provides support to the literature that identifies globalization, Europeanization and trends of liberalization as factors explaining the reduction of labour market protection, and illustrates that we need a revised typology. While in 1990 the varieties of labour market protection broadly conformed to WoW – with a clear differentiation between social democratic, Christian democratic, liberal and Mediterranean regimes – in 2015 the distinction between social democratic (with the exception of Denmark) and Christian democratic countries disappears. The worlds of labour market protection in 2015 can be portrayed as a well-defined typology including three varieties. A first variety places Northern and Central European countries, characterised by the highest levels of coordination and solidarity (although both elements are in decline), in one cluster together; a second variety includes Southern European countries (with France moving to this cluster); a third variety includes liberal countries. In both cases, countries are closer to each other in 2015 than in the 1990s, with regime varieties that are internally more consistent. In the Southern European cluster, we observe liberalization and further deterioration of coordination, solidarity levels and labour market outcomes; in the liberal cluster, labour market protection has declined only slightly as an advanced process of liberalization was already underway in the 1980s.

VoC seems to have stood the test of time better than WoW – perhaps also a reflection of the more parsimonious dual categorization, but with one important caveat. Our PCA shows in 2015 a persistent difference between the two clusters of CMEs (the unified cluster of social democratic and Christian democratic countries, plus the Mediterranean one) and the LMEs (all liberal countries). However, our composite indicator reveals that the level of protection has significantly declined since the 1990s in 10 out of 15 CMEs. These ten countries are following liberalization (Greece, Japan, Netherlands, Norway, Portugal, Spain and Sweden) and dualization (Belgium, Germany and Italy) trajectories, and consequently are currently more similar to LMEs.

Thelen’s work helps to conceptualise the space where countries move and correctly define the trajectories of certain countries (that is, Denmark, Belgium, Germany and all liberal countries). However, trajectories of change appear to be more numerous than the three she formulated, suggesting the development of a fivefold taxonomy. Three of these trajectories closely resemble those Thelen has conceptualised, that is, liberalization, dualization and flexicurity. It is important to note that we replace deregulation with liberalization, to indicate a more encompassing process of labour market protection decline; and we substitute embedded flexicurity with flexicurity, as the trajectory also involves a country outside the social democratic regime. In addition, we have conceptualised two further trajectories to refine our taxonomy. The first is de-dualization – characterised by a reduction of EPL, though stronger for permanent rather than temporary contracts; the second is higher protection – a generalised re-regulation of labour market protection across all dimensions. Further, the trajectories of CMEs are not always path-dependent and consistent with classic regime varieties. Among Social democratic countries, we find flexicurity in Denmark only; in Sweden and Norway our indicator reveals a trajectory of liberalization, and in Finland a trajectory of de-dualization. Among Christian democratic countries, we see dualization in Belgium and Germany, but we see liberalization in the Netherlands, de-dualization in Austria, and higher protection in France. Finally, Mediterranean countries seem to have undergone a trajectory of liberalization (with the exception of Italy, but see Ferragina and Arrigoni, n.d.).

In conclusion, we have shown that the WoW framework is a valid heuristic tool to conceptualise regime varieties, but it is losing some of its explanatory power in the face of widespread processes of change. This consideration reinforces the rationale of this article: it is important to systematically scrutinise and conceptually refine how labour market protection varieties and countries’ trajectories of change evolve over time, and this consideration should also be applied to other policy domains (on family policy see Daly and Ferragina, 2018). Moreover, despite the persistent distinction between CMEs and LMEs in levels of labour market protection, CMEs are becoming more similar to LMEs. It is for scholars of the future to determine whether this process in the long run will undermine the validity of VoC, as some commentators have argued. Substantial institutional change can be obscured from sight if we take for granted the immutability of typologies or consider that new trajectories are always inscribed within classic varieties. For the time being, in the face of the stability bias of which both WoW and VoC seem to suffer, we rehearse the words declared allegedly by Galileo during his trial: Eppur si muove.

Supplemental Material

sj-pdf-1-esp-10.1177_09589287211056222 – Supplemental Material for Labour market protection across space and time: A revised typology and a taxonomy of countries’ trajectories of change

Supplemental Material, sj-pdf-1-esp-10.1177_09589287211056222 for Labour market protection across space and time: A revised typology and a taxonomy of countries’ trajectories of change by Emanuele Ferragina and Federico Danilo Filetti in Journal of European Social Policy

Footnotes

Acknowledgements

We thank Quentin Ramond and the participants of the OSC scientific seminars for the useful comments to earlier drafts of this work.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research was funded by a public grant overseen by the French National Research Agency (ANR) as part of the “Investissements d'Avenir” program LIEPP (ANR-11-LABX-0091, ANR-11-IDEX-0005-02) and the Université de Paris IdEx (ANR-18-IDEX-0001).

ORCID iD

Emanuele Ferragina

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Andersen

(2012) A flexicurity labour market in the great recession: The case of Denmark. De Economist 160(2): 117–140.

Anderson

(2001) The politics of retrenchment in a social democratic welfare state. Comparative Political Studies 34(9): 1063–1091.

Armingeon

Bertozzi

Bonoli

(2004) Swiss worlds of welfare. West European Politics 27(1): 20–44.

Arts

Gelissen

(2002) Three worlds of welfare capitalism or more? A state-of-the-art report. Journal of European Social Policy 12(2): 137–158.

Baccaro

Howell

(2017) Trajectories of Neoliberal Transformation. Cambridge: Cambridge University Press.

Berton

Richiardi

Sacchi

(2012) The Political Economy of Work Security and Flexibility: Italy in Comparative Perspective. Bristol: Policy Press.

Bigos

Qaran

Fenger

, et al. (2013) Review Essay on Labour Market Resilience. Netherlands: INSPIRES Working Paper Series.

Blyth

(2013) Austerity. Oxford: Oxford University Press.

Bonoli

(2005) The politics of the new social policies: Providing coverage against new social risks in mature welfare states. Policy & Politics 33(3): 431–449.

10.

Bonoli

(2010) The political economy of active labour-market policy. Politics & Society 38(4): 435–457.

11.

Bruff

Horn

(2012) Varieties of capitalism in crisis? Competition & Change 16(3): 161–168.

12.

Burroni

Keune

(2011) Flexicurity: A conceptual critique. European Journal of Industrial Relations 17(1): 75–91.

13.

Cárdenas

Villanueva

(2021) Flexibilization at the core to reduce labour market dualism: Evidence from the Spanish case. British Journal of Industrial Relations 59(1): 214–235.

14.

Cardoso

Branco

(2018) Liberalised dualisation. Labour market reforms and the crisis in Portugal. European Journal of Social Security 20(1): 31–48.

15.

Castles

(1996) Needs-based strategies of social protection in Australia and New Zealand. In: Esping-Andersen

(ed) Welfare States in Transition. London: Sage, pp. 88–115.

16.

Chevalier

(2016) Varieties of youth welfare citizenship: Towards a two-dimension typology. Journal of European Social Policy 26(1): 3–19.

17.

Clasen

Siegel

(2007) Investigating Welfare State Change. Cheltenham: Edward Elgar.

18.

Coe

Johns

Ward

(2011) Transforming the Japanese labour market: Deregulation and the rise of temporary staffing. Regional Studies 45(8): 1091–1106.

19.

Cooke

Zeytinoglu

(2004) Temporary employment: The situation in Canada. In: Burgess

Connell

(eds) International Perspectives on Temporary Work. London: Routledge, pp. 113–133.

20.

Crouch

(2005) Capitalist Diversity and Change. Oxford: Oxford University Press.

21.

Daly

(2011) What adult worker model? A critical look at recent social policy reform in Europe from a gender and family perspective. Social Politics 18(1): 1–23.

22.

Daly

Ferragina

(2018) Family policy in high-income countries: Five decades of development. Journal of European Social Policy 28(3): 255–270.

23.

Dolado

García‐Serrano

Jimeno

(2002) Drawing lessons from the boom of temporary jobs in Spain. The Economic Journal 112(480): F270–F295.

24.

Eichhorst

Marx

(2011) Reforming German labour market institutions: A dual path to flexibility. Journal of European Social Policy 21(1): 73–87.

25.

Emmenegger

(2014) The Power to Dismiss. New York: Oxford University Press.

26.

Emmenegger

Häusermann

Palier

, et al. (eds) (2012) The Age of Dualization. New York: Oxford University Press.

27.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

28.

Esping-Andersen

(1997) Hybrid or unique? The Japanese welfare state between Europe and America. Journal of European Social Policy 7(3): 179–189.

29.

Esping-Andersen

Regini

(2000) Why Deregulate Labour Markets? Oxford: Oxford University Press.

30.

Estevez-Abe

Iversen

Soskice

(2001) Social protection and the formation of skills. In: Hall

Soskice

(eds) Varieties of Capitalism. Oxford: Oxford University Press, pp. 145–183.

31.

European Commission (1997) Green Paper Partnership for a New Organisation of Work. Brussels: European Commission.

32.

Ferragina

Arrigoni

(Forthcoming) Selective neoliberalism: How Italy went from dualization to liberalization in labour market and pension reforms. New Political Economy. https://doi.org/10.1080/13563467.2020.1865898.

33.

Ferragina

Seeleib-Kaiser

(2011) Thematic review. Welfare regime debate: Past, present, futures? Policy & Politics 39(4): 583–611.

34.

Ferragina

Seeleib-Kaiser

Tomlinson

(2013) Unemployment protection and family policy at the turn of the 21st century: A dynamic approach to welfare regime theory. Social Policy & Administration 47(7): 783–805.

35.

Ferragina

Seeleib-Kaiser

Spreckelsen

(2015) The four worlds of welfare reality–Social risks and outcomes in Europe. Social Policy & Society 14(2): 287–307.

36.

Ferrera

(1996) The ‘Southern model’ of welfare in social Europe. Journal of European Social Policy 6(1): 17–37.

37.

Fleckenstein

Lee

(2017) The politics of labor market reform in coordinated welfare capitalism: Comparing Sweden, Germany, and South Korea. World Politics 69(1): 144–183.

38.

Fossati

(2018) How regimes shape preferences: A study of political actors’ labour market policy preferences in flexicurity and dualizing countries. Socio-Economic Review 16(3): 523–544.

39.

Gerring

(2012) Mere description. British Journal of Political Science 42(4): 721–746.

40.

Goodin

(2001) Work and welfare: Towards a post-productivist welfare regime. British Journal of Political Science 31(01): 13–39.

41.

Green-Pedersen

Kersbergen

Hemerijck

(2001) Neo-liberalism, the ‘third way’ or what? Recent social democratic welfare policies in Denmark and the Netherlands. Journal of European Public Policy 8(2): 307–325.

42.

Hall

Soskice

(eds) (2001) Varieties of Capitalism. Oxford: Oxford University Press.

43.

Hall

Thelen

(2008) Institutional change in varieties of capitalism. Socio-Economic Review 7(1): 7–34.

44.

Hancké

Rhodes

Thatcher

(eds) (2007) Beyond Varieties of Capitalism. Oxford: Oxford University Press.

45.

Hay

(2004) Common trajectories, variable paces, divergent outcomes? Models of European capitalism under conditions of complex economic interdependence. Review of International Political Economy 11(2): 231–262.

46.

Hay

Wincott

(2012) The Political Economy of European Welfare Capitalism. London: Palgrave.

47.

Hemerijck

(2013) Changing Welfare States. Oxford: Oxford University Press.

48.

Heyes

(2013) Flexicurity in crisis. European Journal of Industrial Relations 19(3): 71–86.

49.

Hijzen

Kappeler

Pak

, et al. (2017) Labour Market Resilience: The Role of Structural and Macroeconomic Policies. Paris: OECD Economics Department WP NoOECD Publishing, pp. 1406.

50.

Jessop

(1993) Towards a schumpeterian workfare state? Preliminary remarks on post-fordist political economy. Studies in Political Economy 40(1): 7–39.

51.

Jolliffe

(2002) Principal Component Analysis. New York: Springer.

52.

Keune

Jepsen

(2007) Not balanced and hardly new: The European commission’s quest for flexicurity. In: Jørgensen

Madsen

(eds) Flexicurity and Beyond. Copenaghen: DJØF, pp. 189–211.

53.

Leschke

(2012) Has the Economic Crisis Contributed to More Segmentation in Labour Market and Welfare Outcomes? ETUI Working Paper. Brussels: European Trade Union Institute.

54.

Lewis

(1992) Gender and the development of welfare regimes. Journal of European Social Policy 2(3): 159–173.

55.

Lindbom

(2001) Dismantling the social democratic welfare model? Has the Swedish welfare state lost its defining characteristics? Scandinavian Political Studies 24(3): 171–193.

56.

Matsaganis

(2012) Social Policy in hard times: The case of Greece. Critical Social Policy 32(3): 406–421.

57.

Matsaganis

(2018) Income support policies and labour market reforms under austerity in Greece. In: Theodoropoulou

(ed) Labour Market Policies in the Era of Pervasive Austerity. Bristol: Policy Press, pp. 43–68.

58.

Natili

(2018) Explaining different trajectories of minimum income schemes: Groups, parties and political exchange in Italy and Spain. Journal of European Social Policy 28(2): 116–129.

59.

Neyer

Seeleib-Kaiser

(1995) Bringing the Economy Back In. Zentrum für Sozialpolitik WP No.16. Bremen: Zentrum für Sozialpolitik.

60.

Obinger

(1999) Minimum income in Switzerland. Journal of European Social Policy 9(1): 29–47.

61.

OECD (1994) The OECD Jobs Study. Paris: OECD.

62.

OECD (2012) OECD Employment Outlook. Paris: OECD.

63.

OECD (2019a) Employment and labour statistics. https://www.oecd.org/employment/labour-stats/.

64.

OECD (2019b) Social expenditure statistics (SOCX). https://www.oecd.org/social/expenditure.htm.

65.

Otto

van Oorschot

(2019) Welfare reform by stealth? Cash benefit recipiency data and its additional value to the understanding of welfare state change in Europe. Journal of European Social Policy 29(3): 307–324.

66.

Pereirinha

(1996) Welfare states and anti-poverty regimes: The case of Portugal. South European Society and Politics 1(3): 198–218.

67.

Petmesidou

(1996) Social protection in Greece: A brief glimpse of a welfare state. Social Policy & Administration 30(4): 324–347.

68.

Picot

Tassinari

(2017) All of one kind? Labour market reforms under Austerity in Italy and Spain. Socio-Economic Review 15(2): 461–482.

69.

Pierson

(1994) Dismantling the Welfare State? Cambridge: Cambridge University Press.

70.

Pintelon

Cantillon

Van den Bosch

, et al. (2013) The social stratification of social risks: The relevance of class for social investment strategies. Journal of European Social Policy 23(1): 52–67.

71.

Prosser

(2016) Dualization or liberalization? Investigating precarious work in eight European countries. Work, Employment and Society 30(6): 949–965.

72.

Rathgeb

(2017) Relying on weak governments: Austrian trade unions and the politics of smoothed dualization. OZP: Austrian Journal of Political Science 45(3): 45–55.

73.

Rueda

(2005) Insider–outsider politics in industrialized democracies: The challenge to social democratic parties. American Political Science Review 99(1): 61–74.

74.

Schneider

Paunescu

(2012) Changing varieties of capitalism and revealed comparative advantages from 1990 to 2005: A test of the Hall and Soskice claims. Socio-Economic Review 10(4): 731–753.

75.

Schröder

(2009) Integrating welfare and production typologies. Journal of Social Policy 38(1): 19–43.

76.

Scruggs

Allan

(2006) Welfare-state decommodification in 18 OECD countries. Journal of European Social Policy 16(1): 55–72.

77.

Scruggs

Detlef

Kati

(2017) Comparative Welfare Entitlements Dataset 2. Version 2017-09. Mansfield, CT: Universities of Connecticut and Greifswald.

78.

Seeleib-Kaiser

(2002) A dual transformation of the German welfare state? West European Politics 25(4): 25–48.

79.

Seeleib-Kaiser

Fleckenstein

(2007) Discourse, Learning and welfare state change: The case of German labour market reforms. Social Policy & Administration 41(5): 427–448.

80.

Taylor-Gooby

(2002) The silver age of the welfare state. Journal of Social Policy 31(4): 597–621.

81.

Taylor-Gooby

(2004) New Risks, New Welfare. Oxford: Oxford University Press.

82.

Thelen

(2012) Varieties of capitalism: Trajectories of liberalization and the new politics of social solidarity. Annual Review of Political Science 15(1): 137–159.

83.

Thelen

(2014) Varieties of Liberalization and the New Politics of Social Solidarity. Cambridge: Cambridge University Press.

84.

Van Rie

Marx

Horemans

(2011) Ghent revisited: Unemployment insurance and union membership in Belgium and the Nordic Countries. European Journal of Industrial Relations 17(2): 125–139.

85.

Viebrock

Clasen

(2009) Flexicurity and welfare reform: A review. Socio-Economic Review 7(2): 305–331.

86.

Visser

Hemerijck

(1997) A Dutch Miracle. Amsterdam: Amsterdam University Press.

87.

Wilthagen

(1998) Flexicurity: A New Paradigm for Labour Market Policy Reform?. Berlin: WZB Discussion Paper FS I, pp. 98–202.

88.

Wilthagen

Tros

(2004) The concept of “flexicurity”. Transfer: European Review of Labour and Research 10(2): 166–186.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.27 MB

Labour market protection across space and time: A revised typology and a taxonomy of countries’ trajectories of change

Abstract

Keywords

Introduction

Understanding varieties and trajectories of change

Data and methods

Data

Methods

Scrutinizing institutional varieties across space and time

Clusters in the 1990s

Inter-temporal change of clusters (1990–2015)

Countries’ trajectories

Discussion and conclusion

Supplemental Material

sj-pdf-1-esp-10.1177_09589287211056222 – Supplemental Material for Labour market protection across space and time: A revised typology and a taxonomy of countries’ trajectories of change

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

ORCID iD

Supplemental Material

Notes

References

Supplementary Material

Countries^’ trajectories