The politics of subnational social policy: Social consumption versus social investment in Austria

Abstract

Country comparisons, often suffering from unobserved heterogeneity and obscuring subnational variation, dominate the social policy literature. However, the subnational level is better suited to reduce the omitted variable bias. This article distinguishes between social consumption and social investment policies and investigates their determinants at the subnational level. Following the literature across countries, we test the role of incumbent parties’ ideology, but for within-country variation in social policy. Austria is a case in point because states have discretion in social policy (e.g., regarding public childcare and social assistance). Panel regressions covering all nine states in Austria for the years 1991 to 2019 reveal that the cabinet share of Social-Democrats increases social investment spending, while the Christian-Democratic party decreases it, and the populist radical right party reduces expenses for social consumption.

Keywords

Austria political parties social consumption social investment subnational governments

Introduction

Since social policy is a multi-dimensional concept resulting in different priorities and explanatory factors, recent studies distinguish between the passive consumption of social benefits substituting (lacking) income in the case of life course risks such as unemployment, and active social investments such as childcare that prepare people pre-emptively for the future.

Most studies analysing social policy compare countries as units of analysis. However, the national level is not always the main relevant level of analysis masking within-country variation (Ciccia and Javornik, 2019; Gleichen and Parolin, 2020; Greer et al., 2015) and comparisons of countries often suffer from unobserved heterogeneity due to omitted variables. Studying units below the national level allows us to control for many factors that are either difficult to observe and measure across countries (e.g., national history, culture) or for which data are not available (e.g., administrative records, details of spending categories). Ideas and policy legacies, for example, might explain differences in social policy between countries confounding the results for other alternative explanations. Comparing subnational units that belong to the same country and share several characteristics and regulations allows us to disentangle factors that may explain variation in social policy. Thanks to its lower degree of heterogeneity, the subnational level can be a better setting for the identification of explanatory factors such as party ideology than cross-country settings.

Only a few publications deal explicitly with social investment policies at the subnational level (e.g., Baines et al., 2019; Scalise and Hemerijck, 2024) and if they do, they are based on (qualitative) case studies. This article adds quantitative evidence for within-country variation in social consumption and social investment expenses.

Following the literature on country comparisons, we argue that incumbent parties’ ideology explains variation in social policy spending. Focusing explicitly on social investment, Kim and Choi (2020) as well as Kühner (2018) confirm a positive left-partisan effect on social investment expenditure. Similarly, Bürgisser (2022) finds that centre-left parties in Southern Europe are social investment protagonists, while centre-right parties are antagonists or consenters. Although populist radical right parties (PRRP) are more supportive of social consumption than social investment policies (Enggist and Pinggera, 2022), they have a negative influence on social consumption policies that target undeserving groups of the population such as the poor, unemployed and minorities (Chueri, 2021). Furthermore, Green parties in governments increase social investment expenses, while they show no effect on social consumption expenses (Röth and Schwander, 2021).

The existing literature on the subnational variation of social spending reveals substantial differences within countries (Kleider, 2018; Park, 2014; Vampa, 2016) and often explains them with party ideology. The review by Kleider and Toubeau (2022) confirms the important role of political conflicts in subnational policy choices. Left parties show a positive relation with social spending (Busemeyer and Seitzl, 2018; Toubeau and Vampa, 2021) and the provision of childcare (Andronescu and Carnes, 2015).

While it is true that there are studies on party effects on social spending at the country level as well as on the role of party politics at the subnational level, to our knowledge this is the first article that analyses with quantitative methods the association between party ideology and social consumption versus social investment policies at subnational level over time. The prerequisite for the analysis of subnational variation in social policy and the role of party ideology is that subnational levels have discretion over social investment and social consumption policies. This is the case for Austria and it therefore serves as a test case to investigate party effects on social policy at the subnational level.

We advance the literature by systematically analysing spending levels for social consumption and social investment policies at subnational instead of national levels and by investigating the partisan politics of both types of social policy over almost 30 years for four parties.

The main contribution to the literature is the finding that parties matter for subnational units in shaping social investment and social consumption policies. On the one hand, some of our results confirm findings from studies at national levels, such as the positive (negative) relation of Social Democrats (Christian Democrats) and social investment spending and the negative association of PRRP with social consumption spending. On the other hand, some results are new, such as the positive (lacking) relation of Christian Democrats (Social Democrats) and social consumption spending, which might be related to our analysis of the subnational and not national level, where incumbent parties have discretion over different types of social investment and social consumption policies compared to the national level.

In the following, we briefly describe the Austrian case before we introduce our arguments and hypotheses for the four most important parties in Austria based on the social structure of their constituency. After the introduction of our data, panel regressions covering all nine federal states in Austria for the years 1991 to 2019 test our hypotheses and the last section summarises the results.

Subnational social policy in Austria

Studying subnational social policy in Austria is a case in point for several reasons. Austria has a rather generous welfare state and the subnational level has high administrative responsibilities and autonomy (Fallend, 2003; Karlhofer and Pallaver, 2013) that has increased over the last decades (Hooghe et al., 2016). Social protection expenses accounted for 22% of the total expenditures of subnational governments in Austria in 2020, similar to the level of other federal states such as Belgium (23%), Germany (27%) and Switzerland (21%), and much more than in the federal states of the US (5%) and Australia (8%) (OECD subnational government structure and finance data).

Social policy responsibilities in Austria are shared between the central government, states, and municipalities. Although the central government sets the regulations, for example, for pension and unemployment benefits, subnational units have decision-making power in policies regarding long-term care, education, poverty prevention, health, housing, and family issues (Greer and Elliott, 2019). States set, for example, the regulation for childcare regarding group size, staff requirements and equipment of facilities (Baierl and Kaindl, 2011) and variation in social assistance policies is largely driven by party ideology (Heitzmann and Matzinger, 2021).

Although Austrian states only administer funding they receive from the federal level, based on strict criteria, for primary and secondary education and parts of research and science, they often overspend their budgets knowing that the national government will cover their deficits afterwards (bailout) (Kleider et al., 2018). This allows state governments (in small ways) to shape education in which they have no formal decision-making rights. Finally, the informal influence of subnational governments is even higher than the formal influence (Mätzke and Stöger, 2020).

Theoretical framework

Social consumption versus social investments

The traditional concept of social policy, understood as a redistribution machine that is supported by the left and opposed by the right, has been replaced by multifaceted distributive conflicts. Social policy today is a multi-dimensional concept. As such, it is not about who is for or against social policy, but rather what type of social policy is supported. Over the last two decades, the social investment approach has gained in popularity (Esping-Andersen et al., 2002; Hemerijck, 2017; Morel et al., 2012; Nolan, 2013). Based on their functions rather than outcomes, social consumption policies provide compensatory cash benefits that substitute (lacking) income in the case of life course risks such as unemployment (unemployment benefits) and longevity (pension benefits). In contrast, the main goals of social investment policies are (a) education and training of the population from early childhood until retirement preventing unemployment and promoting economic growth; (b) high employment levels through education and training as well as work/family policies; and (c) prevention of individuals’ dependency on social benefits by enabling them to participate in labour markets (Choi et al., 2020).

Social policy discretion at the subnational level differs from the national level, however. Apart from policies set at the national level (e.g., pension and unemployment benefits) subnational units in federal states such as Austria often have discretion in both social investment (e.g., education and family services) and social consumption policies (e.g., social assistance and elderly care) (for details, see the section ‘Dependent variables’ below).

The role of parties

Despite the vanishing effects of partisan incumbency on social policy in times of austerity and fiscal constraints (Pierson, 1996), parties are still key actors for social policy reforms and spending decisions thanks to their legislative power (Choi et al., 2020). Although parties consider several factors when deciding on policy issues or the content of manifestos, such as vote-, office- or policy-related aspects, the preferences of their constituency have a major impact on their policy positions and spending priorities (Röth and Schwander, 2021).

Individual social policy preferences, in turn, are shaped by material self-interest (Meltzer and Richard, 1981), with low-income earners, people at risk of unemployment and from lower classes in favour of redistributive policies (Häusermann et al., 2016; Rehm et al., 2012), and ideological predispositions. Recent studies, however, suggest that the class cleavage lost in importance for the social base of large mainstream parties such as Social-Democratic and centre-right parties, while education largely structures the voter base of green parties and PRRPs (Marks et al., 2023).

Social-Democratic parties

Social-Democratic parties today rely on two major supporter groups: the traditional working class and the new urban middle class, consisting for example of socio-cultural professionals (Gingrich and Häusermann, 2015; Kitschelt, 1994; Oesch and Rennwald, 2018), both with rather contrasting social policy priorities.

Whereas lower-educated and low-income earners as well as those at high risk of unemployment, i.e., the traditional core constituency of left parties, are core supporters of social consumption policies, the highly-educated middle class as a new target group is in favour of social investment policies (Garritzmann et al., 2018; Han and Kwon, 2020).

Furthermore, people leaning towards the left on the economic dimension are supportive of both social investment and social consumption policies, and left-libertarians prefer social investments and oppose social consumption policies (Garritzmann et al., 2018). Therefore, it is difficult to identify clear social policy priorities for Social-Democratic parties based on the socio-structural characteristics of their supporters. Their core electoral constituencies consisting of the working class with lower education and income are, according to the discussed literature, more in favour of social consumption policies, while a new voter group in times of electoral competition is the highly-educated middle-class with left-libertarian views favouring social investments (Garritzmann et al., 2018; Kitschelt and Rehm, 2015). At the country level, left parties have a positive effect on social investment expenses (Kim and Choi, 2020; Kühner, 2018) and reforms in Southern Europe (except Greece) (Bürgisser, 2022).

Descriptive evidence from Busemeyer et al. (2022) further suggests that supporters of Social-Democratic parties tend to be in favour of both passive social transfers and social investments. Therefore, we assume that Social-Democratic parties try to speak to both core supporter groups for vote-seeking reasons and therefore increase spending for both types of social policy.

The cabinet share of the Social-Democratic Party (SPÖ) increases both social investment and social consumption expenses.

Centre-right parties

Centre-right parties draw mainly on the entrepreneurial class, i.e., large employers, self-employed professionals, small business owners and managers, as their core electoral constituency (Oesch and Rennwald, 2018). Their share of highly educated voters is around the average of the overall electorate (Marks et al., 2021) and income levels are rather high. They are leaning towards right-wing economic attitudes and are rather sceptical of generous welfare benefits for financial reasons. This also applies to the Austrian Christian-Democratic Party (Österreichische Volkspartei, ÖVP). Production and service workers are less likely to vote for them compared to clerks (Figure W.6 in Oesch and Rennwald, 2018) and the share of blue-collar supporters of ÖVP is substantially smaller than their share for the Social-Democrats (Sozialdemkratische Partei Österreich, SPÖ) and the PRRP (Freiheitliche Partei Österreichs, FPÖ) (Aichholzer et al., 2014).

Garritzmann et al. (2018) and Bürgisser (2022) confirm that centre-right parties are social investment antagonists or consenters. Their position on social consumption policies is more difficult to assess as high-income earners and right-wing economic values are associated with lower levels of social consumption policies, but people with traditional and authoritarian values are in favour of them. The results by Busemeyer et al. (2022) reveal that voters of Christian-Democratic parties are significantly less likely than voters of Green and Social-Democratic parties to support social consumption policies such as pension and unemployment benefits as well as assistance to the poor. Considering concerns about high labour costs due to generous welfare benefits, we would expect centre-right parties to reduce both social investment and social consumption expenses.

The cabinet share of the conservative party (ÖVP) reduces both social investment and social consumption expenses.

Apart from large catch-all parties, smaller parties such as PRRPs (Mudde, 2013; Oesch and Rennwald, 2018) and green parties (Grant and Tilley, 2019; Müller-Rommel and Poguntke, 2002) have become integral parts of party systems in parliamentary democracies such as Austria and are regularly coalition partners in governments.

Populist radical right parties

PRRPs heavily draw on the working class, i.e., production and service workers, as their core constituency (Marks et al., 2021; Oesch and Rennwald, 2018). Based on the results of Garritzmann et al. (2018) and given the social-structural composition of their electorate with above-average levels of low-income earners, people with lower education and traditional, authoritarian and national attitudes, we expect PRRPs to be against social investment policies, from which other groups profit more. Busemeyer et al. (2022) confirm the opposition of PRRP voters to social investment spending and also find a negative association with social consumption policies such as assistance to the poor and unemployment benefits. Poor and unemployed as well as non-native citizens are perceived as not deserving benefits (Ennser-Jedenastik, 2016). We, therefore, hypothesise a negative association of PRRPs for social consumption policies as state governments in Austria have discretion over measures that are directed at the poor. Furthermore, country-level studies confirm the negative association of PRRPs with spending for people regarded as undeserving of benefits such as poor, unemployed and non-native people (Chueri, 2021) and with social investment policies (Enggist and Pinggera, 2022).

The cabinet share of the Populist Radical Right party (FPÖ) reduces both social investment and social consumption expenses.

Green parties

In contrast to the role of left, centre-right and radical-right parties, the literature on green parties and social policy is rather scarce. The most encompassing study by Röth and Schwander (2021) reveals that green parties are associated with higher social investment expenses, while they show no effect on social consumption expenses. This fits with the social policy preferences of their core electorate of younger, highly educated and middle-class voters (Dolezal, 2010; Knutsen, 2004; Marks et al., 2021) representing the ‘urban “new” middle class’ (Röth and Schwander, 2021: 6). Most Green voters profit more from social investment than social consumption policies. However, we also expect Green parties to promote policies for people at risk of poverty or unemployment (social consumption policies) due to their middle-class voters with left-liberal attitudes valuing solidarity concepts. Green party supporters show similar levels of support for assistance to the poor and unemployment compared with voters of Social-Democratic parties and significantly different levels compared with supporters of PRRPs and Christian-Democratic parties (Busemeyer et al., 2022).

The cabinet share of the Green party increases both social investment and social consumption expenses.

Data and method

Our dataset contains yearly observations for all nine Austrian states. The availability of data limits our analysis to the years 1991 to 2019. Information for all independent and dependent variables are available only since 1990 (meaning a start in 1991 for our models since we measure the dependent variable as the first difference (see below)) and not yet after 2019. While the concept of ‘social investment’ was formulated only in the late 1990s, it does not preclude retrospective examination before the categorization of social policies under this term. Although political parties may not have explicitly touted increased spending on ‘social investments’ during that period, they may have allocated resources to specific social policies, such as childcare provision, which were later recognised as components of social investment. Generally, parties tend not to consider the academic and analytical distinctions between SI and SC when formulating social policies or presenting them to the public. This lack of awareness persists even today, decades after the introduction of the concept of social investment.

Dependent variables

The dependent variables measure social investment and social consumption spending for each federal state (a) in relation to their GDP and (b) per capita. Statistics Austria provides disaggregated regional expenses and revenues for nine different categories (Voranschlagsgruppen) (VA) that can be broken down into further sub-categories. Theoretical considerations and the literature on social investment and social consumption policies, as discussed above, guide the selection of individual (sub-)categories for our measures. Policies that are supposed to prepare people for the future, thanks to investments in human capital formation and by contributing to work/family reconciliation with the aim of high employment levels, are grouped into the social investment category. VA2 (education, sport, and science) entails spending on education (e.g., primary and secondary education, adult education), higher education and research, and public childcare. The subgroups VA43 (youth welfare services) and VA46 (family services) address families and children, e.g., by financing children’s (recreation) homes, by providing advice on educational and developmental issues and family problems and with income-related subsidies for childcare costs. Our measure of social consumption spending consists of five subgroups: spending for public welfare (e.g., social and disability assistance) (VA41), elderly care and benefits for refugees (VA42), promoting the repair of disaster damage (e.g., compensation for continued payment of wages) (VA44), compensatory allowances, commuter assistance and contributions to pension schemes (VA45), and housing (housing allowances, renovations, and subsidies) (VA48). We classify housing as social consumption in line with Röth and Schwander (2021), because it provides an ex-post social compensation covering basic housing provision needs rather than having a preventive (ex-ante) function (i.e., creation, preservation, or mobilization of human skills and capabilities according to Garritzmann et al., 2023). Moreover, the housing category, for which the data do not allow a breakdown into subcategories, consists not only of social housing, but also e.g., mortgage regulation, fiscal exemptions, and homeownership subsidies.

The decision to group individual social policy categories into social investment or social consumption is based on the idea that all categories in social investment possess a preventive (ex-ante) function (i.e., creation, preservation, or mobilization of human skills and capabilities), and all categories in social consumption provide compensatory cash benefits that substitute (lacking) income covering basic social needs. Analysing the categories individually without considering the others as part of packages would give an incomplete and biased picture of how states combine individual social policy categories to prioritise future-oriented preventive measures (social investment) or compensatory cash benefits (social consumption).

The descriptive statistics in Table A.1 in the appendix show considerable differences between states over time for the level of social investment expenses (ranging from 1.85 to 3.89% of GDP and €390 to €1,510 per capita) as well as for social consumption expenses (ranging from 1.22 to 4.09% of GDP and €310 to €1,274 per capita).

Explanatory variables

Evolving out of our research question and hypotheses, the main explanatory variable for our analysis is party ideology. We measure the cabinet share of the four most important parties during the analysed timespan (1991 to 2019) in states’ governments. The Austrian Freedom Party (FPÖ) can be characterised as a populist radical right party and the Austrian People’s Party (ÖVP), a centre-right Christian-Democratic party, shares a conservative worldview. The Social Democratic Party (SPÖ) represents a party with (formerly) strong ties to the working class, while the Greens (Die Grünen, GRÜNE) focus on environmental protection and are engaged in feminism and the protection of minority rights. The ÖVP dominates the analysed period with a mean cabinet share of 48%, followed by the SPÖ (36%), FPÖ (12%) and GRÜNE (3.5%) (Table A.1). We estimate all models for one party only and therefore assess their effect compared to all other parties.

Control variables

We consider several political and socio-economic control variables that might affect the results of our models. We integrate the share of women in cabinets and parliaments at the state level as this could influence the decision between cash benefits (usually social consumption policies) and social services (usually social investment policies) (Bolzendahl, 2011; Ennser-Jedenastik, 2017). The employment rate of women, the share of the population below 6 years and the unemployment rate represent the demand side for social policies (e.g., for family benefits/services and support of unemployed persons) which could bias our results. Similarly, fertility rates can determine specific policies aiming at their increase. We control for population density as some services or benefits prevail in urban or rural areas, and for deindustrialization (share of service sector jobs) as this shapes the structure of employment and specific social policy instruments. The level of social policy spending is influenced by the economic situation and we therefore include states’ GDP growth in our models. We add a year variable accounting for trending because downward or upward trending in spending levels might be influenced by time trends.

Method

We test our hypotheses with time-series cross-sectional models (TSCS). The fact that Austria only has nine states is a natural limitation of our sample. However, several authors and studies apply TSCS models to data with 5–12 units (Bolzendahl, 2009, 2011; Chang, 2020; Dreher et al., 2008; Jensen and Wenzelburger, 2020; Kim, 2007; McManus, 2018; Onaran and Boesch, 2014). Although the preferred choice of most studies are fixed effects models, for theoretical reasons they are of limited value in our case. The composition of states’ governments (especially the party of the head of government) varies little over time. In six out of nine states, one party had been dominating and appointed the head of government over the whole time (1991–2019). The variance of incumbent parties’ ideology is much larger across than within states. Hence, we apply random effects models.

In contrast to random models with countries as units of analysis, subnational units do suffer less from the omitted variables bias, because they belong to the same country (rather small size in Austria) sharing similarities such as the national welfare state and the cultural heritage. Some studies suggest the use of mixed-effects models (multilevel models with three levels) allowing the effects of variables at different hierarchical levels to be combined (Ennser-Jedenastik, 2022; Garritzmann and Seng, 2020). However, the methodological literature for multilevel models advises against such models when the number of units at one level is very low, as in our case with nine states (Hox et al., 2017; Lazega and Snijders, 2016).

To examine the temporal properties, we introduce a one year time lag for all independent variables. Panel-corrected standard errors account for heteroscedasticity and autocorrelation (Bailey and Katz, 2011). Furthermore, we transform all dependent variables to first differences and include their lagged level as control variables accounting for the dependence of the dependent variable on its previous year’s level. Table 1 shows the results of 16 models as we run separate models for each of the four dependent variables and each of the four parties.

Table 1.

Time-series cross-sectional models (random effects) for social consumption and social investment spending.

	Δ Social consumption % GDP	Δ Social investment % GDP	Cross-model difference	Δ Social consumption per capita	Δ Social investment per capita	Cross-model difference
	(1)	(2)	(1–2)	(3)	(4)	(3–4)
SPÖ cabinet share (t-1)	0.0002	0.001 ^**	−0.0008	−0.342	0.168 ^*	−0.51
SPÖ cabinet share (t-1)	(0.001)	(0.0004)		(0.373)	(0.098)	−0.51
ÖVP cabinet share (t-1)	0.001	−0.001 ^**	0.002^*	0.415 ^*	−0.184 ^**	0.599^**
ÖVP cabinet share (t-1)	(0.001)	(0.0003)		(0.230)	(0.074)	0.599^**
FPÖ cabinet share (t-1)	−0.003 ^**	0.0001	−0.003^**	−0.766 ^**	0.096	−0.862^**
FPÖ cabinet share (t-1)	(0.001)	(0.0004)		(0.329)	(0.109)	−0.862^**
GRÜNE cabinet share (t-1)	0.0001	−0.001	0.001	0.679	0.085	0.594
GRÜNE cabinet share (t-1)	(0.002)	(0.001)		(0.655)	(0.219)	0.594
Share of women cabinet ministers (t-1)	0.0004	−0.00005		−0.046	−0.094
Share of women cabinet ministers (t-1)	(0.002)	(0.001)		(0.433)	(0.149)
Share of women in parliament (t-1)	−0.009^***	−0.004^***		−0.887	−0.553^***
Share of women in parliament (t-1)	(0.003)	(0.001)		(0.665)	(0.209)
Female employment rate (t-1)	0.001	−0.005^**		1.040	−1.119^*
Female employment rate (t-1)	(0.006)	(0.002)		(1.570)	(0.622)
Fertility rate (t-1)	0.247	0.077		91.548	45.579^**
Fertility rate (t-1)	(0.252)	(0.081)		(70.735)	(20.810)
Share of population younger than 6 years (t-1)	−0.116^***	−0.014		−26.109^**	1.948
Share of population younger than 6 years (t-1)	(0.041)	(0.016)		(12.670)	(3.883)
Unemployment rate (t-1)	−0.023	−0.021^**		−10.602^**	−7.113^***
Unemployment rate (t-1)	(0.018)	(0.010)		(5.397)	(2.021)
Regional GDP growth (t-1)	−0.002	−0.001		−0.125	0.248
Regional GDP growth (t-1)	(0.006)	(0.003)		(1.840)	(0.674)
Level of Deindustrialisation (t-1)	−0.003	−0.005^**		2.046^*	−0.568
Level of Deindustrialisation (t-1)	(0.004)	(0.002)		(1.054)	(0.361)
Population density (t-1)	0.00002	0.00004^***		0.005	0.018^***
Population density (t-1)	(0.00003)	(0.00002)		(0.008)	(0.004)
Social consumption % GDP (t-1)	−0.213^***
Social consumption % GDP (t-1)	(0.036)
Social investment % GDP (t-1)		−0.091^***
Social investment % GDP (t-1)		(0.025)
Social consumption per capita (t-1)				−0.166^***
Social consumption per capita (t-1)				(0.040)
Social investment per capita (t-1)					−0.110^***
Social investment per capita (t-1)					(0.025)
year	−0.007	0.007^***		−0.311	5.051^***
year	(0.005)	(0.003)		(1.724)	(1.061)
Constant	1.454^***	0.860^***		55.080	91.228
Constant	(0.553)	(0.316)		(128.523)	(56.575)
Observations	252	252		252	252
R²	0.154	0.114		0.099	0.135
Adjusted R²	0.111	0.070		0.054	0.092
F Statistic	43.490^***	30.836^***		26.261^***	37.309^***

Note. Panel-corrected standard errors in parentheses. We run separate models for each party family. For ÖVP, FPÖ and GRÜNE, only their party coefficients and PCSE are shown, all other coefficients are from the SPÖ model (for full models, see Table A.2 and A.3 in the appendix). We use the R package ‘systemfit’ (Henningsen and Hamann, 2007) for the seemingly unrelated regressions and compare the system fit of the restriction to the party variable for both models with the ‘linear.hypothesis’ command (Wald chi-square test) in the R package ‘car’ (Fox and Weisberg, 2019).

^*p < .1; ^**p < .05; ^***p < .01.

Two effects are not necessarily statistically and substantively different, independent of their significance (Gelman and Stern, 2006). To be able to make a statement regarding the social policy priorities of parties, we compute 84% confidence intervals for parties’ point estimates allowing us to assess whether their difference is significant, as suggested by Payton et al. (2003). In addition to this visual inspection, we estimate seemingly unrelated regressions and compare the effect of each party on social consumption and social investment spending with a linear hypothesis test (Mize et al., 2019; Radean, 2023).

Results

Public expenses for social consumption and social investment policies

The descriptive results for the level of social investments and social consumption expenses across states and over time in Figure 1 reveal that there is no homogenous pattern. Social consumption expenses decreased between 1990 and 2019 in five out of nine states, while social investment expenses are rather stable except for increases in three states. Furthermore, spending levels of social investment and social consumption policies are rather similar in Salzburg and Tyrol, while we see major gaps in Carinthia, Upper Austria, Vorarlberg and Vienna. In the latter group, social investment expenses are substantially higher than social consumption spending. This variation confirms our assumption of subnational differences regarding the type of social policy expenses across states and over time.

Figure 1.

Development of social investment (dotted line) and social consumption (solid line) spending (% regional GDP) between 1990 and 2019 for the Austrian states.

Multiple regression results

The regression models presented in Table 1 indicate that there is a significant positive association between SPÖ’s cabinet share and both social investment spending measures (models 2 and 4). However, there is no significant relation between the cabinet share and social consumption spending (models 1 and 3). This only partially confirms Hypothesis 1, which assumed a positive association between both social consumption and social investment spending. It could be that voters of the highly educated middle class who favour social investments over social consumption policies already have been established as SPÖ’s new core constituency. Moreover, the national and the subnational levels have discretion over different types of social consumption policy which might explain why Social Democrats show a positive relation with social consumption policies at the national (with discretion, for example, over pension benefits), but not necessarily at the subnational level. However, although social investment spending at the subnational level is higher when the cabinet share of Social-Democrats increases – compared to all other parties – we cannot draw a conclusion regarding Social-Democrats’ social policy priority. The coefficients of SPÖ for SI and SC are not significantly different as column 3 in Table 1 shows. Furthermore, the 84% confidence intervals in Figure A.2 overlap, confirming that there is no difference and hence Social-Democrats do not prioritise SI over SC.

In contrast to the result for the SPÖ, the cabinet share of the ÖVP is significantly negatively associated with both measures of social investment spending. This confirms a part of Hypothesis 2. However, we do not find a similar negative relationship with social consumption spending as hypothesised. Moreover, ÖVP’s cabinet share shows a significant positive association with social consumption spending per capita. The difference between ÖVP’s effect on SI and SC is significant and confidence intervals in Figure A.3 do not overlap. This means that the Christian Democrats seem to have a clear priority against (in favour of) SI (SC) at the subnational level. It could be that the ÖVP at the states level behaves differently to the federal level, where lower social consumption benefits are the major discourse. Furthermore, the ÖVP has a rather important workers’ fraction (ÖAAB) with strong ties to trade unions which might add evidence for Christian Democrats being pro-welfare state parties (as suggested by van Kersbergen, 1995).

We also have to reject a part of Hypothesis 3, because FPÖ’s cabinet share reduces social consumption expenses, but not social investments. The ‘deservingness of welfare benefits’ argument seems to apply here since the responsibility of Austrian states regarding social consumption policies lies mainly in the area of assistance to the poor and unemployed. We cannot confirm the negative relationship of PRRPs and their electorate with social investments that previous studies found at the national level (Busemeyer et al., 2022; Enggist and Pinggera, 2022) and the significant cross-model difference for both spending measures (as well as non-overlaps in Figure A.4) indicates a preference of the FPÖ against social consumption spending.

Finally, we do not find any significant relationship between the Green party and social consumption or social investment expenses. It might be that policy areas other than social policy such as environmental policies are of higher importance for Green parties and their constituency at the state level. The very low average cabinet share of Green parties of 3.5% across all states and years might also serve as an explanation for lacking significant results.

Computing 84% confidence intervals for all spending measures across parties reveals that the negative association between FPÖ and SC is significantly different (and more negative) compared with the same association for SPÖ and ÖVP (Figure A.5). In contrast, there is no significant difference for SI spending between SPÖ and FPÖ (Figures A.7 and A.8), but – again – between FPÖ and ÖVP. This last result confirms the need to analyse the Christian Democrats and PRRP separately for the analysis of social policy priorities at the subnational level.

Regarding the robustness of our results, measuring the dependent variables in levels instead of differences yields the same results (Table A.4). Models that exclude housing from social consumption (Table A.5), because the housing category in some states also encompasses instruments that do not have a clear social policy aim (e.g., mortgage regulation), confirm the main results, only the significant positive association between the ÖVP and social consumption spending disappears. This signals that housing benefits matter in particular for ÖVP’s cabinet share.

We further calculated models with a different measure of party ideology. Grouping parties to centre-left (SPÖ, GRÜNE) and parties right to centre (ÖVP, FPÖ) reveals similar results (Table A.6). The positive relationship of ÖVP and the negative one of FPÖ with social consumption in Table 1 level out resulting in non-significant coefficients, confirming our approach of measuring parties individually. Accordingly, it is not sufficient to distinguish between centre-left and centre-right parties, because the association of the PRRP FPÖ with social consumption spending differs substantially compared with the centre-right party ÖVP.

However, we also estimated one model including both left-wing parties (SPÖ and GRÜNE) as separate variables within one model (Table A.7) and one model including both right-wing parties (ÖVP and FPÖ) as separate variables within one model (Table A.8). The results show that the negative effect of right-wing parties’ cabinet party share on SI is caused by the Christian Democratic and not by the Populist Radical Right Party. Furthermore, the positive effect of left-wing parties’ cabinet party share on SI is caused by the Social Democrats and not by the Greens. This corroborates the results of our main models.

Asymmetries regarding power and ideology between national and state governments and dependencies of regional parties/governments on national parties/governments (alignment effects) might bias the results (Golden and Picci, 2008; Kleider et al., 2018). Models considering whether a party is part of the subnational and national government (Table A.9) still reveal results similar to our main models. Interestingly, national governments with SPÖ participation show significant positive associations for all four dependent variables suggesting the SPÖ is the dominating social policy party in Austria. In contrast, when the FPÖ belongs to the national government the relationship is significantly negative for all dependent variables (and to a lesser extent for ÖVP participation).

It could be that the share of foreigners, especially refugees, drives the results of social consumption expenses since benefits for refugees are part of it. Integrating a measure for the share of foreigners from Africa, Asia, Europe (non-EU) and stateless as a proxy reveals slightly different results, although the directions of coefficients for all parties are the same (Table A.10). The relationship of the SPÖ with social consumption spending per capita turns significantly negative (instead of non-significant in the main model), the results for the ÖVP and social investment spending are not significant any more, and the Green party shows a significant instead of non-significant positive relationship with social consumption spending. However, data for foreigners across states have been available only since 2002 and restricting the main model to the same period reveals the same results indicating that the restricted period and not the share of foreigners drives the slightly different results.

Finally, gridlock effects due to coalition partners as veto players might bias parties’ pure effects on social policy spending. Controlling for the number of coalition partners does not change the direction of party coefficients of the main models (Table A.11). However, the FPÖ shows no significant relationship with any spending measure signalling that it affects social policy spending only with no or few coalition partners. In contrast, the positive association of the Green party with social consumption spending per capita turns significant, partially corroborating Hypotheses 4.

Conclusion

This study aims to investigate the role of the incumbent’s party ideology on social consumption and social investment spending at the subnational level. Unlike most other contributions in the field, we use subnational data, which allows us to control for variables that are difficult to observe and/or measure, like national history and culture. The main contribution to the literature is the finding that parties matter for subnational units in shaping social investment and social consumption policies.

Our models, covering four parties between 1991 and 2019, indeed show the expected positive association of the Social Democrats (SPÖ) with social investment spending, but not with social consumption which might indicate the increasing importance and reliance of the SPÖ on the highly educated (urban) middle-class voters who favour social investments. In line with our hypothesis, the cabinet share of the Christian Democratic party (ÖVP) reduces social investment spending, but in contrast to our expectations, it increases social consumption spending (only for the per capita measure). However, this positive association is largely driven by housing which also entails instruments without a clear social policy aim such as mortgage regulation. The negative relationship with social consumption but not with social investment spending only partially confirms our hypothesis about the Populist Radical Right party (FPÖ) and the argument that PRRPs cut benefits for people regarded as undeserving of benefits such as poor, unemployed and non-native people. Surprisingly, we do not find any significant relationship between the Green party and social policy spending, which might be related to their rather low average cabinet share and policy priorities that distinguish them from other parties in different policy fields (e.g., environmental policy or immigration).

With regard to social policy priorities, the associations of the Christian Democrats and the PRRP with SC spending are significantly different to their associations with SI spending revealing that the former prefer lower (higher) SI (SC) spending and the latter lower SC spending. On the contrary, the effect of the Social Democrats on SI and SC is not significantly different hinting at a more balanced social policy approach.

In sum, some of our results confirm findings from studies at national levels (e.g., the positive (negative) relation of SPÖ (ÖVP) and social investment spending and the negative association of FPÖ with social consumption spending) and some are new (e.g., the positive (lacking) relation of ÖVP (SPÖ) and social consumption spending). The latter might be related to our analysis of the subnational and not national level, where incumbent parties have discretion over different types of social investment and social consumption policies. Although serving the same aim and function, our groups of social consumption and social investment policies consist of (slightly) different individual policies as compared to similar groups at the national level.

It seems, however, that parties show similar effects for social investment policies at both national and subnational levels, despite decision-making power over different types of social investment policies.

As the major limitation, we could analyse the role of party ideology on different groups of social policy spending for only nine subnational units, but the existing states in Austria leave no room for increasing the number of units. Although fixed effects models are usually more appropriate to establish causality, the limited variance of the cabinet composition in Austrian states together with the small number of states call for random effects models. Thanks to the rather small size of Austria and the fact that all states share important characteristics such as a common cultural heritage and the national welfare state, we are confident of keeping the omitted variable bias limited. Whether or not our results for subnational party effects hold for other countries might be worth investigating further.

Distinguishing several dimensions of social policy and specifically social consumption from social investment policies seems to be a fruitful endeavour not only for (comparative) analyses at the national level but also at subnational levels, although the policy content and level of discretion might vary.

Supplemental Material

Supplemental Material - “The politics of subnational social policy: Social consumption versus social investment in Austria

Supplemental Material for The politics of subnational social policy: Social consumption versus social investment in Austria in Carmen Walenta-Bergmann and Tobias Wiß in Journal of European Social Policy

Footnotes

Acknowledgements

We are grateful to the participants of the 2022 annual conference of the Council of European Studies, the 2022 annual conference of ESPAnet and Bruno Arpino for their feedback on previous versions of this study.

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Carmen Walenta-Bergmann

Tobias Wiß

Supplemental Material

Supplemental material for this article is available online.

References

Aichholzer

Kritzinger

Wagner

, et al. (2014) How has radical right support transformed established political conflicts? The case of Austria. West European Politics 37(1): 113–137.

Andronescu

Carnes

(2015) Value coalitions and policy change: the impact of gendered patterns of work, religion and partisanship on childcare policy across German states. Journal of European Social Policy 25(2): 159–174. DOI: 10.1177/0958928715573480.

Baierl

Kaindl

(2011) Kinderbetreuung in Österreich. Rechtliche Bestimmungen und die reale Betreuungssituation. Working Paper Nr. 77: 91.

Bailey

Katz

(2011) Implementing panel-corrected standard errors in R: the pcse package. Journal of Statistical Software 42(1): 1–11. DOI: 10.18637/jss.v042.c01.

Baines

Bassi

Csoba

, et al. (eds) (2019) Implementing Innovative Social Investment: Strategic Lessons from Europe. Bristol: Policy Press.

Bolzendahl

(2009) Making the implicit explicit: gender influences on social spending in twelve industrialized democracies, 1980–99. Social Politics: International Studies in Gender, State & Society 16(1): 40–81. DOI: 10.1093/sp/jxp002.

Bolzendahl

(2011) Beyond the big picture: gender influences on disaggregated and domain-specific measures of social spending, 1980–1999. Politics and Gender 7(1): 35–70. DOI: 10.1017/S1743923X10000553.

Bürgisser

(2022) The partisan politics of family and labor market policy reforms in Southern Europe. In: Garritzmann

Häusermann

Palier

(eds) The World Politics of Social Investment (Volume II): The Politics of Varying Social Investment Strategies. Oxford: Oxford University Press, 86–107.

Busemeyer

Seitzl

(2018) The partisan politics of early childhood education in the German Länder. Journal of Public Policy 38(2): 243–274. DOI: 10.1017/S0143814X16000313.

10.

Busemeyer

Rathgeb

Sahm

AHJ

(2022) Authoritarian values and the welfare state: the social policy preferences of radical right voters. West European Politics 45(1): 77–101. DOI: 10.1080/01402382.2021.1886497.

11.

Chang

Y-L

(2020) Unequal social protection under the federalist system: three unemployment insurance approaches in the United States, 2007–2015. Journal of Social Policy 49(1): 189–211. DOI: 10.1017/S0047279419000217.

12.

Choi

Huber

Kim

, et al. (2020) Social investment in the knowledge-based economy: new politics and policies. Policy and Society 39(2): 147–170. DOI: 10.1080/14494035.2020.1782577.

13.

Chueri

(2021) Social policy outcomes of government participation by radical right parties. Party Politics 27(6): 1092–1104. DOI: 10.1177/1354068820923496.

14.

Ciccia

Javornik

(2019) Methodological challenges for comparative welfare state research: capturing intra-country variation in cross-national analyses. Journal of Comparative Policy Analysis: Research and Practice 21(1): 1–8. DOI: 10.1080/13876988.2018.1551598.

15.

Dolezal

(2010) Exploring the stabilization of a political force: the social and attitudinal basis of green parties in the age of globalization. West European Politics 33(3): 534–552. DOI: 10.1080/01402381003654569.

16.

Dreher

Sturm

J-E

Ursprung

(2008) The impact of globalization on the composition of government expenditures: evidence from panel data. Public Choice 134(3–4): 263–292. DOI: 10.1007/s11127-007-9223-4.

17.

Enggist

Pinggera

(2022) Radical right parties and their welfare state stances – not so blurry after all? West European Politics 45(1): 102–128. DOI: 10.1080/01402382.2021.1902115.

18.

Ennser-Jedenastik

(2016) A welfare state for whom? A group-based account of the Austrian Freedom Party’s social policy profile. Swiss Political Science Review 22(3): 409–427. DOI: 10.1111/spsr.12218.

19.

Ennser-Jedenastik

(2017) How women’s political representation affects spending on family benefits. Journal of Social Policy 46(3): 563–581. DOI: 10.1017/S0047279416000933.

20.

Ennser-Jedenastik

(2022) The impact of radical right parties on family benefits. West European Politics 45(1): 154–176. DOI: 10.1080/01402382.2021.1936944.

21.

Esping-Andersen

Gallie

Hemerijk

, et al. (2002) Why We Need a New Welfare State. New York: Oxford University Press.

22.

Fallend

(2003) Föderalismus – eine Domäne der Exekutiven? Die Bund-Länder-Verhandlungen über die Länderforderungsprogramme und die Bundesstaatsreform in Österreich seit 1945. In: Dachs

(ed) Der Bund und die Länder (‘Geschichte der österreichischen Bundesländer von 1945 bis 1995’, Sonderband 10). Vienna: Böhlau, 17–68.

23.

Fox

Weisberg

(2019) An R Companion to Applied Regression. 3rd edition. Thousand Oaks, CA: Sage. https://socialsciences.mcmaster.ca/jfox/Books/Companion/.

24.

Garritzmann

Seng

(2020) Party effects on total and disaggregated welfare spending: a mixed-effects approach. European Journal of Political Research 59(3): 624–645. DOI: 10.1111/1475-6765.12371.

25.

Garritzmann

Busemeyer

Neimanns

(2018) Public demand for social investment: new supporting coalitions for welfare state reform in Western Europe? Journal of European Public Policy 25(6): 844–861. DOI: 10.1080/13501763.2017.1401107.

26.

Garritzmann

Häusermann

Palier

(2023) Social investments in the knowledge economy: the politics of inclusive, stratified, and targeted reforms across the globe. Social Policy and Administration 57(1): 87–101. DOI: 10.1111/spol.12874.

27.

Gelman

Stern

(2006) The difference between ‘significant’ and ‘not significant’ is not itself statistically significant. The American Statistician 60(4): 328–331.

28.

Gingrich

Häusermann

(2015) The decline of the working-class vote, the reconfiguration of the welfare support coalition and consequences for the welfare state. Journal of European Social Policy 25(1): 50–75. DOI: 10.1177/0958928714556970.

29.

Daiger von Gleichen

Parolin

(2020) Varieties of liberalism: a comparative analysis of family policy and poverty outcomes across the 50 United States. Social Policy and Administration 54(6): 933–951. DOI: 10.1111/spol.12617.

30.

Golden

Picci

(2008) Pork-barrel politics in postwar Italy, 1953–94. American Journal of Political Science 52(2): 268–289. DOI: 10.1111/j.1540-5907.2007.00312.x.

31.

Grant

Tilley

(2019) Fertile soil: explaining variation in the success of Green parties. West European Politics 42(3): 495–516. DOI: 10.1080/01402382.2018.1521673.

32.

Greer

Elliott

(2019) Federalism and Social Policy: Patterns of Redistribution in 11 Democracies. Ann Arbor, MI: University of Michigan Press. DOI: 10.3998/mpub.9993201.

33.

Greer

Elliott

Oliver

(2015) Differences that matter: overcoming methodological nationalism in comparative social policy research. Journal of Comparative Policy Analysis: Research and Practice 17(4): 408–429. DOI: 10.1080/13876988.2015.1060713.

34.

Han

Kwon

(2020) Employment insecurity and social policy: preferences for investment vis-à-vis consumption. Policy and Society 39(2): 247–265. DOI: 10.1080/14494035.2019.1699005.

35.

Häusermann

Kurer

Schwander

(2016) Sharing the risk? Households, labor market vulnerability, and social policy preferences in Western Europe. The Journal of Politics 78(4): 1045–1060. DOI: 10.1086/686972.

36.

Heitzmann

Matzinger

(2021) The (changing) consideration of social investment in the design of welfare benefits: the case of poverty relief in Austria. Social Policy and Administration 55(4): 573–588. DOI: 10.1111/spol.12643.

37.

Hemerijck

(2017) The Uses of Social Investment. Oxford: Oxford University Press.

38.

Henningsen

Hamann

(2007) systemfit: a package for estimating systems of simultaneous equations in R. Journal of Statistical Software 23(4): 1–40. https://www.jstatsoft.org/v23/i04/

39.

Hooghe

Schakel

Marks

and, et al. (2016) Measuring regional authority: a postfunctionalist theory of governance. In: Transformations in Governance. 1st edition. Oxford and New York: Oxford University Press.

40.

Hox

Moerbeek

Schoot

R van de

(2017) Multilevel Analysis: Techniques and Applications. Third edition. Quantitative Methodology Series. New York, NY: Routledge.

41.

Jensen

Wenzelburger

(2020) New evidence on the effect of partisanship on the welfare state. In: Careja

Emmenegger

Giger

(eds) The European Social Model under Pressure: Liber Amicorum in Honour of Klaus Armingeon. Wiesbaden: Springer Fachmedien, 123–137. DOI: 10.1007/978-3-658-27043-8_8.

42.

Karlhofer

Pallaver

(2013) Strength through weakness: state executive power and federal reform in Austria. Swiss Political Science Review 19(1): 41–59.

43.

Kim

(2007) Social risk and social insurance: political demand for unemployment insurance. Rationality and Society 19(2): 229–254. DOI: 10.1177/1043463107077388.

44.

Kim

Choi

(2020) Does social protection crowd out social investment? Policy and Society 39(2): 208–225. DOI: 10.1080/14494035.2019.1677377.

45.

Kitschelt

(1994) The Transformation of European Social Democracy. Cambridge: Cambridge University Press.

46.

Kitschelt

Rehm

(2015) Party Alignments: Change and Continuity. In: Beramendi

Häusermann

Kitschelt

(eds). The Politics of Advanced Capitalism. Cambridge: Cambridge University Press, 179–201.

47.

Kleider

(2018) Redistributive policies in decentralised systems: the effect of decentralisation on subnational social spending. European Journal of Political Research 57(2): 355–377. DOI: 10.1111/1475-6765.12229.

48.

Kleider

Toubeau

(2022) Public policy in multi-level systems: a new research agenda for the study of regional-level policy. Regional & Federal Studies 32(3): 277–305. DOI: 10.1080/13597566.2021.2018681.

49.

Kleider

Röth

Garritzmann

(2018) Ideological alignment and the distribution of public expenditures. West European Politics 41(3): 779–802. DOI: 10.1080/01402382.2017.1395634.

50.

Knutsen

(2004) Social Structure and Party Choice in Western Europe: A Comparative Longitudinal Study. Basingstoke and New York: Palgrave Macmillan.

51.

Kühner

(2018) Partisan ideology, veto players and their combined effect on productive and protective welfare policy: how the left still matters after all? Policy Studies 39(2): 165–187. DOI: 10.1080/01442872.2018.1451499.

52.

Lazega

Snijders

TAB

(eds) (2016) Multilevel network analysis for the social sciences: theory, methods and applications. Methodos series. Cham: Springer International Publishing, Vol. 12.

53.

Marks

Attewell

Rovny

, et al. (2021) Cleavage theory. In: Riddervold

Trondal

Newsome

(eds) The Palgrave Handbook of EU Crises. Basingstoke: Palgrave Macmillan, 173–193.

54.

Marks

Attewell

Hooghe

, et al. (2023) The social bases of political parties: a new measure and survey. British Journal of Political Science 53(1): 249–260. DOI: 10.1017/S0007123421000740.

55.

Mätzke

Stöger

(2020) Health care federalism in Austria. In: Fierlbeck

Palley

(eds) Comparative Health Care Federalism. Federalism Studies. London and New York: Routledge, 15–28.

56.

McManus

(2018) Political parties as drivers of post-crisis social spending in liberal welfare states. Comparative European Politics 16(5): 843–870. DOI: 10.1057/s41295-017-0105-y.

57.

Meltzer

Richard

(1981) A rational theory of the size of government. Journal of Political Economy 89(5): 914–927.

58.

Mize

Doan

Long

(2019) A general framework for comparing predictions and marginal effects across models. Sociological Methodology 49(1): 152–189.

59.

Morel

Palier

Palme

(eds) (2012) Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: Policy Press.

60.

Mudde

(2013) Three decades of populist radical right parties in Western Europe: so what? European Journal of Political Research 52(1): 1–19. DOI: 10.1111/j.1475-6765.2012.02065.x.

61.

Müller-Rommel

Poguntke

(2002) Green Parties in National Governments. Basingstoke: Routledge. DOI: 10.4324/9781315039169.

62.

Nolan

(2013) What use is ‘social investment’? Journal of European Social Policy 23(5): 459–468. DOI: 10.1177/0958928713499177.

63.

Oesch

Rennwald

(2018) Electoral competition in Europe’s new tripolar political space: class voting for the left, centre-right and radical right. European Journal of Political Research 57(4): 783–807. DOI: 10.1111/1475-6765.12259.

64.

Onaran

Boesch

(2014) The effect of globalization on the distribution of taxes and social expenditures in Europe: do welfare state regimes matter? Environment and Planning A: Economy and Space 46(2): 373–397. DOI: 10.1068/a45370.

65.

Park

(2014) The politics of redistribution in local governments: the effect of gender representation on welfare spending in California counties. Journal of Public Policy 34(2): 269–301. DOI: 10.1017/S0143814X14000075.

66.

Payton

Greenstone

Schenker

(2003) Overlapping confidence intervals or standard error intervals: what do they mean in terms of statistical significance? Journal of Insect Science 3(1): 34–36. DOI: 10.1093/jis/3.1.34.

67.

Pierson

(1996) The new politics of the welfare state. World Politics 48(2): 143–179. DOI: 10.1353/wp.1996.0004.

68.

Radean

(2023) The significance of differences interval: assessing the statistical and substantive difference between two quantities of interest. The Journal of Politics 85(3): 969–983.

69.

Rehm

Hacker

Schlesinger

(2012) Insecure alliances: risk, inequality, and support for the welfare state. American Political Science Review 106(2): 386–406. DOI: 10.1017/S0003055412000147.

70.

Röth

Schwander

(2021) Greens in government: the distributive policies of a culturally progressive force. West European Politics 44(3): 661–689. DOI: 10.1080/01402382.2019.1702792.

71.

Scalise

Hemerijck

(2024) Subnational social investment in three European cities: an exploratory comparison. Journal of Social Policy 53(2): 530–550. DOI: 10.1017/S0047279422000496.

72.

Toubeau

Vampa

(2021) Adjusting to austerity: the public spending responses of regional governments to the budget constraint in Spain and Italy. Journal of Public Policy 41(3): 462–488. DOI: 10.1017/S0143814X20000094.

73.

Vampa

(2016) The Regional Politics of Welfare in Italy, Spain and Great Britain. Basingstoke: Palgrave Macmillan.

74.

van Kersbergen

(1995) Social Capitalism: A Study of Christian Democracy and the Welfare State. London: Routledge.

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

0.70 MB