Abstract
This paper studies the role of government partisanship in the transformation of two uniquely old care policy areas: the care of people with chronic and severe mental illness, and of children who cannot be cared for by their parents. While nineteenth-century ‘insane asylums’ and ‘orphanages’ have widely been understood as institutions of social control, they also served a social care function, which during the era of deinstitutionalisation was replaced by alternative forms of care. Studying mental health and child welfare policy decisions in 12 advanced capitalist countries between 1950 and 2015, I show that the types of care policies that replaced large, custodial institutions varied with government partisanship. I argue that partisan policy choices reflected parties’ core policy preferences shaped by trade-offs between their redistributive goals and individualist or familial ideals, and the lasting ideological effects of very old societal cleavages. The study contributes to theoretical debates in comparative politics about the role of partisanship in social policy making and the dimensionality of party competition over time.
Introduction
Care policies present decision-makers with trade-offs between their redistributive goals on the one hand, and their ideals about the role and nature of the family on the other. Conservatives seeking to scale back state expenditure might comfortably consign care to the home as unpaid labour, but for committed liberals, policies that promote gender equality by taking care out of the family can place a prohibitive burden on the public purse. This means care policy does not always divide political parties on purely left–right lines, and it can’t be explained by power resources theories of welfare politics (e.g. Esping-Andersen 1990; Korpi 1983). This paper studies the transformation of two uniquely old care policy areas, addressing whether and how party politics matters.
Most research on the relationship between partisanship and care policy has focussed on the expansion of care policies in the last three to four decades. This period of post-industrial change is characterised by the emergence of a new, more multidimensional politics as the salience of socio-economic conflict has declined while a second, non-economic or cultural dimension has become more important (Kitschelt, 1994). The vote-seeking behaviour of political parties in response to social change has opened up new space for cross-class alliances in support of social policy expansion (Gingrich and Häusermann, 2015). Care policy in particular is more multidimensional than other types of social policies because it concerns both economic and non-economic issues, shaping the cost of living, the labour market participation of women, and the role of the family (Schwander, 2018). It therefore attracts diverse groups of voters who may support the same policy for different reasons (Häusermann, 2018).
But politics has always been multidimensional. Both class and religious cleavages shaped the ideologies and policy preferences of social democrats, Christian democrats, conservatives, and liberals during the emergence of mass politics, as conflicts played out over not just how much the state should redistribute through public services, but who should provide those services: the state, the market, or voluntary providers, and religious or secular authorities (Ansell and Lindvall, 2021, p. 21).
In this paper I develop a theory of the partisan politics of care based on the ideological or policy-seeking behaviour of political parties, rather than the electoral strategies of the post-industrial age. I argue that the core policy preferences of the major parties that have led governments since the middle of the twentieth century reflect ideological positions shaped by historical conflicts and reinforced over time. I contribute to recent literatures emphasising the connection between contemporary party politics and historical social cleavages (Rovny and Polk, 2019), and calling for a rebalancing of partisanship theory in the wake of its ‘electoral turn’ to re-consider the role of parties’ own policy preferences (Wenzelburger and Zohlnhöfer, 2021).
I consider the relationship between partisanship and care policy in the remarkable transformation of mental health and child welfare policy since 1950, which has been largely neglected by the welfare politics literature – with the exception of Perera (2020, 2025). Large psychiatric institutions once known as ‘insane asylums’ and homes for children called ‘orphanages’ are very early examples of care policies. Although standards of living were brutal, asylums and orphanages were care policies because they served a social care function, to substitute for family care.
Large psychiatric institutions and orphanages were retrenched in the period during which state responsibility for welfare was expanding, and when informal care was becoming less available as women’s labour market participation grew. The secondary historical and sociological literature suggests partisanship was not an important cause of ‘deinstitionalisation’, compared with structural explanations like technological and ideational change (Jones 1993, pp. 179–80; Rochefort 1997, p. 124). The diverse forms of ‘community care’ that replaced asylums and orphanages included new, smaller institutions like group homes for both children and people with mental illness, psychiatric wards in general hospitals, the expansion of foster care and adoption services, and family welfare interventions.
Existing literature has focused on why deinstitutionalisation occurred, but there is also great variety in the types of providers that became responsible for care: new public services in the community, but also private for-profit or voluntary organisations, and the family. In this paper I focus on the question not of why deinstitutionalisation occurred, but how. I describe the types of care policies that replaced asylums and orphanages by categorising policy decisions based on the rich descriptions in the secondary literature. I examine whether governments led by social democrats, Christian democrats, conservatives, or liberals were more likely to pursue certain types of care policy. These policy decisions came in many cases much earlier than the expansion of child care or aged care, but involved the same types of trade-offs. I argue that partisanship relates to care policy in ways that are not new at all.
The partisan politics of care
Scholars agree that contemporary care policy is not well explained by left–right partisanship. In addition to more recent studies of multidimensional welfare politics, critics of power resources theory have long argued that factors such as gender ideologies have been more important for shaping policies targeted at families and dependents (e.g. Lewis 1992). I argue that we should expect partisan care policy preferences to be influenced by both left–right partisanship and a second dimension of partisan ideology dividing individualist and familial ideals.
I expect the ideological tendencies of major parties on these two dimensions to reflect their positions in historical class and religious conflicts. Just as the legacy of industrial class struggle shapes the ideology of the present-day left, historical religious cleavages influence partisan ideals about the role and nature of the family. The idea that much older societal cleavages can have lasting effects on party politics has been tested by Rovny and Polk (2019), who show that the non-economic dimension often styled as ‘GAL-TAN’ is associated with the nineteenth-century religious cleavage, even though explicitly religious content is often absent from contemporary politics. In studies of family policy, Morgan (2006) and Saraceno (1994) argue that partisan gender ideologies also relate to historical religious conflicts.
Figure 1 depicts the redistributive preferences and individualist or familial ideals of the four major party families that have led governments in advanced capitalist countries since the Second World War, drawing from the cleavage theory of Lipset and Rokkan (1967) and historical-sociological literature on political parties. Preferences for high or low levels of redistribution are shaped by class conflict. Social democratic parties emerging from workers’ movements are most distantly divided from conservative parties which originated in old regime elites defending the existing order against social upheaval (Ziblatt, 2017, p. 29). Dimensions of partisan conflict.
Both liberals and Christian democrats are placed toward the centre of the socio-economic dimension, neither as embracing of redistribution as social democrats, but for different reasons. Christian democratic parties became proponents of welfare state expansion both to attract working-class voters and out of moral obligation to assist the poor, while liberal parties overcame a reluctance towards collectivisation either to attract labour support or with secularising motives against a dominant religious opposition (Manow and Van Kersbergen, 2009, p. 4). With support among the middle classes, liberals were less threatened by redistribution than conservatives, and often formed alliances with labour parties (Luebbert, 1991, p. 7).
A second dimension of conflict divides Christian democrats and conservatives from social democrats and liberals. Partisan ideals of individualism and familism were shaped by late-nineteenth- and early-twentieth-century religious conflict. Struggles for control over public services between dominant churches and anticlerical forces generated opposing visions of whether economic and social rights should be attached to individuals or families (Van Kersbergen, 1995).
Both Christian democrats and conservatives privilege families. Christian democratic parties have their roots in catholic associations formed to defend against liberal attacks (Kalyvas, 1996, pp. 71–4). Christian democracy was distinctively shaped by catholic social doctrine and a philosophy of personalism which views people as flourishing in communities (Hanley, 1994, p. 4). The family played a more instrumental role for conservatives, to maintain the existing social order and curtail state intervention. Influenced by Christian morals or in opposition to commercial interests, conservatives aimed to preserve the value of kinship connections over status from industrial enterprise (Lipset and Rokkan, 1967, p. 19).
Liberals and social democrats support policies that confer rights upon individuals. Early liberal parties comprised periphery cultures, dissident religions and either industrialists or agrarians, allied in opposition to the powerful classes and dominant religious groups (Lipset and Rokkan, 1967, pp. 41–6). The role of individualism in liberal thought long predates the emergence of liberal parties, but for these diverse coalitions, individual rights ensured both freedom of enterprise and freedom from religious authority (Gould, 1999, pp. 3–4). For social democrats, individual rights underpinned an active role for the state to increase both freedom and equality (Keman, 2017, pp. 54–60). A view of individuals as the core social unit enabled more active family policy from the early twentieth century, and the incorporation of gender equality into political platforms from the 1970s (Morgan, 2006).
Partisan care policy preferences.
Comments:
I expect parties of the centre and right to prefer other policy options based on the interaction of their ideological positions on the two dimensions of conflict. Reflecting less support for redistribution, Christian democrats prefer corporatist models of welfare based on the principle of subsidiarity. Their position in religious conflicts reinforces a preferences for voluntary provision, based on the direct role played by religious providers and the intrinsic value of voluntary organisations. I expect Christian democrats to implement policies that expand not-for-profit care services.
Liberals’ commitment to free markets and non-intervention underpins a preference for a smaller public sector. Liberals may support voluntary welfare providers based on the principles of freedom of choice and association, and as an alternative to socialism (Rothstein, 1992). But, concerned not to empower religious providers, liberals may not expand voluntary provision to the same extent as Christian democrats.
Both conservatives and Christian democrats have familial values, placing stronger emphasis on the male-breadwinner model and informal care, so they should promote family responsibility to a greater extent than social democrats or liberals. I expect conservatives and Christian democrats to implement policies that incentivise or assist family members to take on primary caring responsibilities.
Conservatives’ familism reinforces their preference for low levels of redistribution, favouring private forms of welfare provision. Based on the socio-economic dimension of conflict, both conservatives and liberals should be more likely to implement policies that authorise or subsidise for-profit providers.
Asylums and orphanages as care policies
We might think of care policies as relatively new in the history of the welfare state, but asylums and orphanages are care policies that are very old. They first developed alongside poor law institutions, but expanded during the nineteenth century as public services that cared for specific groups of so-called ‘deserving’ poor, with a reformist purpose distinct from the punitive principle of indoor poor relief. In this section I argue their social function to substitute for family care made them care policies.
The first asylums and orphanages emerged in the fifteenth to eighteenth century, mostly operated by private and religious organisations. Prior to the establishment of laws on ‘lunacy’, people with what would later be understood as mental illness were governed by the laws on poor relief (Ansell and Lindvall, 2021, p. 184). Early orphanages cared not mainly for genuine orphans, but for children conceived out of wedlock or from single-parent families affected by poverty (Boswell, 1998, pp. 271–6). During the nineteenth century, asylums and orphanages were increasingly publicly funded and administered, and their expansion was coupled with more public goals. Addressing the problems of infanticide and neglect was a means of increasing population and national strength (Ulbricht, 1985). Asylums and orphanages took on the character of other emerging public services, aspiring to the mental and moral improvement of the population (Ansell and Lindvall, 2021, p. 16).
This reformist quality, and the provision of indoor relief on the basis of desert and not punishment, separates asylums and orphanages from general poor law institutions. This does not contradict the fact that living conditions were often very poor, and treatment at times abusive. The social care function of asylums and orphanages is not incompatible with their other function, of social control, which is central in so much of the sociological literature (e.g. Cohen 1985; Scull 1984). The coercive or controlling aspects of institutional care existed alongside a nurturing, protecting, and improving function – a tension that is also present in the programmes and services of the wider welfare state (Morgan and Orloff, 2017).
Historical evidence from around the world suggests most asylum admissions were by family members. Finnane (1985) shows that temporary patients in Ireland and Australia were often admitted by relatives through involuntary committal procedures. In nineteeth-century New York, requests by families dominated forced committals (Dwyer, 1987, p. 87). Similar patterns have been revealed in France, the Netherlands, and New Zealand (e.g. Coleborne 2009; Prestwich 1994).
Peak institutional populations (by date of peak asylum population).
Sources: See online appendix.
A paradigmatic shift toward deinstitutionalisation occurred in most countries from around the 1980s. One prominent explanation is that the fiscal crisis of the 1970s promoted a shift toward cheaper policy alternatives (Scull, 1984, p. 135). While adequate community care requires no fewer resources, the belief among policy-makers that it would be cheaper contributed to deinstitutionalisation as a cost-cutting strategy (Lamb, 1993).
Another key driver was changing scientific knowledge. Critiques of asylums came from both psychiatrists working within the system and the ‘anti-psychiatry’ movement. Experiments with more open forms of care were publicised in Western Europe during the 1950s and 1960s. Experts increasingly influenced policy-making directly through government inquiries from the 1970s, and new ideas also shaped public opinion, including through popular culture (Rochefort, 1997).
Experts in child development and child psychiatry also contributed to the deinstitutionalisation movement. John Bowlby’s work on the importance of parent–child attachment gained acceptance in policy circles from the 1960s, promoted through the World Health Organisation. Formal networks of child welfare professionals expanded during the 1980s and 1990s through the UN programme on child welfare and the reporting process for the Convention on the Rights of the Child, adopted in 1989 (Mccall, 1999). Because experts espoused family-based care specifically, I expect more policy decisions in this policy area to promote family responsibility regardless of partisanship.
Patterns of care policy change
In the rest of the paper I describe the relationship between government partisanship and care policy in a diverse selection of countries, presenting descriptive statistics and the results of a logistic regression. I begin by identifying policy decisions that affected the use of institutional care. For asylums, this includes the regulation of services for people with mental illness, but also more general long-term-care services if people with mental illness were eligible. For orphanages, the scope is broader, because children in institutional care were more diverse: children from poor or single-parent households, whose parents were absent or incapacitated, and children accused or convicted of crimes. I therefore include the regulation of adoption and foster care, child protection laws about the removal of children, laws regulating what happens to children accused or convicted of crimes, and support services targeted at families perceived as vulnerable. I exclude minor changes to admissions, discharge, or child-removal procedures if they did not affect the type of provider responsible for care.
I select the countries in Table 2 because they have different providers of institutional care and different experiences of deinstitutionalisation: some early movers and others laggard, some that made sweeping changes and others favouring incrementalism. They also have a variety of constitutional arrangements and party competition environments. Half are effectively two-party systems, mostly majoritarian, while the others are proportional with multi-party competition and typically coalition governments. Among the two-party systems, half are characterised by competition between social democrats and conservatives, and half by competition between liberals and conservative or religious parties. Half the multi-party systems have a history of Christian democratic leadership.
The study begins in 1950, before institutional populations peaked in most countries, but after institutions in central Europe returned to their pre-war capacity, having been affected by the terrible loss of life during the Second World War. I identify 136 policy changes between 1950 and 2015. I do not count public statements, commission reports, government strategies, or non-binding ministerial directives as policy changes unless they are associated with new legislative or funding decisions, or formal agreements with responsible levels of government. For example, the famous ‘water tower’ speech by the British Health Minister Enoch Powell in 1961 was associated with grants to local government the following year. However, the Progetto 80 report published in 1969 by the Italian Ministry of Budget and Economic Planning was not a policy decision, because it lacked associated funding or regulatory change.
Indicators of responsibility for care.
Policies that promote market responsibility subsidise for-profit providers or promote their expansion. This includes increasing the scope of public subsidies to include for-profit providers of psychiatric hospitals where they had been excluded before, as in Ireland in 1957, or creating new funding streams to subsidise private group home providers, as in England in 1979.
Policies that promote family responsibility include payments to family carers, such as those introduced in Italy in 1980. They also include regulatory changes that prioritise family care, for instance by only granting access to residential services once family-based support options have been exhausted, as in England in 1990, or by increasing barriers for the removal of children. This category also includes new family services to prevent child removal.
Policies that promote voluntary responsibility include new subsidies for voluntary providers to establish shared residences, as in the Netherlands in 1985, or to expand existing voluntary domiciliary services, as in France in 2005. This category also includes investments in existing institutional care delivered by voluntary providers, or shifting long-term care beds from public mental hospitals into voluntary-run institutions or general hospitals.
One limitation of my approach is that I don’t capture hidden policy change, through policy drift in the face of changing social conditions, or the redirection of existing policy settings towards new goals (Hacker, 2004; Streeck and Thelen, 2005). But alternative measures of policy change, like spending patterns or the changing service mix over time, are affected by changes in demand as well as time lags that can distance an observed change from the government that made it. While not sensitive to hidden policy change, formal policy decisions are a direct and timely measure of the choices made by governments.
Government partisanship is measured by the proportion of cabinet posts held. I follow the widely accepted party family classifications listed in Armingeon et al. (2020, Appendix D), where all parties holding the largest proportion of cabinet posts in my study are classified as social democratic, conservative, liberal, or religious.
The top two panels in Figure 2 describe patterns in care policy changes by government partisanship, displaying the frequency of policy changes promoting state, market, family, or voluntary responsibility for care in each policy area by leading party of government. There are 71 policy changes relating to mental health, and 65 relating to child welfare. All parties make policy changes at a similar rate, ranging from 9.2% of country-years led by Christian democrats to 7.7% of country-years led by Conservatives. Policy changes in each category, by policy area and government partisanship, and over time.
Policy changes promoting state responsibility are mostly made by social democrats, and those promoting voluntary responsibility mostly by Christian democrats. Most changes promoting market responsibility are made by conservatives and liberals, while policy changes that promote the responsibility of the family are more common for child welfare regardless of partisanship. The widely accepted science of child development gave all types of governments reason to promote family-based care for children.
The drivers of deinstitutionalisation discussed above played a role over time. If policy change was motivated by cost savings, then we would expect fewer policies promoting state responsibility and more promoting market responsibility over time. We might also observe an increase in policies promoting family responsibility as the science of child development is gradually transmitted through professional and policy networks, and because family responsibility could be pursued as a form of cost cutting.
The third panel in Figure 2 plots patterns in care policy decisions over time. The frequency of policy changes promoting state responsibility peaks around 1970 and declines sharply from the 1970s, while other types of policy changes gradually increase. In particular, policy changes promoting family responsibility are most common in the 1980s and 1990s, and those promoting market responsibility from the 1980s to 2000s. These patterns are consistent with the idea that changing scientific knowledge and fiscal pressure played a role over time, but we still observe all types of policy change throughout the period.
Care policy changes may also be affected by welfare regime type. Welfare regimes are shaped by dominant political parties and their ideologies over time, but once established these institutions are sticky, and can perpetuate existing policy settings or facilitate certain types of change. To disentangle the effect of partisanship from the institutional effects of dominant political ideologies, we need to move beyond descriptive statistics.
I run a multinomial logistic regression, controlling for time-variant factors as well as welfare regime to account for the effect of historically dominant political parties on welfare policy. Standard errors are clustered by country to mitigate the risk of falsely inferring statistical significance on the basis of country-specific factors or patterns in a limited number of countries with dominant political parties.
I also control for other factors that could affect both the composition of government and the propensity for state, market, voluntary providers, or families to take responsibility for care. A policy area dummy captures differences between the policy areas. I expect all parties to promote more family responsibility for children, but because I don’t expect heterogeneous effects of partisanship across policy areas, I don’t interact these variables. Because pre-existing asylum and orphanage providers could shape demand for particular types of policy change, I control for the dominant type of provider in each country and policy area, and whether a policy change has been made before.
I control for economic performance using GDP per capita, because this is known to affect who governs and could constrain capacity to invest in public services. I control for women’s labour market participation, because this shapes both voting behaviour and the availability of family care. I also include controls for coalition and federal governments, whose capacity to pass preferred policies is likely constrained. In robustness checks to address coalition governments, I recode partisanship in line with the head of government or responsible minister, and it does not change the direction or statistical significance of any results.
Multinomial logistic regressions on type of policy change.
Standard errors in parentheses.
*p < 0.05, **p < 0.01, ***p < 0.001
Many of the results align with expectations. Even after introducing controls and decade fixed effects, both conservative- and liberal-led governments are more likely to make policy changes that promote market responsibility. The three significant partisanship coefficients for ‘Family’ suggest that social democratic governments prefer state responsibility over family responsibility to a greater extent than all other types of government. Also in line with expectations, Christian democrats appear more likely to promote voluntary care, even controlling for welfare regime characteristics that make voluntary provision more likely.
The size of the coefficients in this table should not be directly compared to assess the differences between parties. Likewise, the positive association between liberal governments and family responsibility only indicates that liberals prefer family over state responsibility to a greater extent than social democrats – not surprising if social democrats privilege state responsibility above all else.
The most direct test of the relationship between partisanship and care policy choice is to compare a change from one type of government to another. Figure 3 plots the average marginal effect of a change in partisanship on the probability of each outcome category. The top three panels compare Christian democrats, conservatives, and liberals to social democrats, now directly testing the statistically significance of this change in partisanship. The other panels compare parties of the centre and right with one another. Average marginal effect of a change in government partisanship. Comments: Average marginal effect reported with 90% confidence intervals. A confidence interval which excludes zero (does not cross the dotted line) suggests that the change has a statistically significant effect.
The clearest result, in the top three panels, confirms that social democrats prefer state responsibility to a greater extent than all other types of parties. Social democrats are most clearly differentiated from the party family that is most distant on both dimensions of conflict. Conservatives are more likely to promote market or family responsibility, and less likely to promote voluntary responsibility.
Social democrats and conservatives compete as major parties in half the countries in this study, in all three welfare regime types and in both majoritarian and proportion systems. These two party families exhibit expected patterns of care policy making in very different contexts. Consider Sweden, a proportional social democratic welfare regime where the Social Democratic Party has been extremely dominant, and Britain, a majoritarian liberal welfare regime where Labour has controlled parliament about half as often as the Conservatives. In both cases, the majority of changes made by social democrats promote state responsibility and the majority of changes made by conservatives promote care by the family or private, for-profit providers.
The patterns in Figure 3 for Christian democrats also align with expectations. Christian democrats have a higher probability of making a policy change promoting family or voluntary responsibility than social democrats, but are no more likely to promote market responsibility. Christian democrats also have distinct care policy preferences from other parties of the centre and right. They are more likely than conservatives to promote voluntary responsibility, and less likely than liberals to promote market responsibility.
But Christian democratic parties often govern in conservative welfare states with a wider structure of voluntary and familial policies. Consider Germany, a conservative welfare state where the Christian Democrats (CDU/CSU) are the largest party in two-thirds of the years in this study. Psychiatric hospitals had been state-operated since 1871, and in the mid-1970s the government led by the Social Democrats (SPD) began to expand public outpatient care, excluding the voluntary providers that operated most general hospitals. The major parties agreed that institutional care should be reduced, but disagreed about who should deliver care in its place, with CDU/CSU-led states refusing to implement the SPD policy (Cooper, 1987). After the 1983 election, the Kohl administration suspended the programme and funded new voluntary services in 1986, aiming to integrate psychiatry into the wider health system. Conflict between the SPD and CDU/CSU resulted in different types of care policy decisions when each party controlled government, and perpetuated relatively high levels of inpatient care (Goodwin, 1997). Even though wider welfare institutions constrained policy making, partisanship also shaped care policy outcomes.
Liberals also exhibit distinct preferences, but not all results in Figure 3 align with expectations. As expected, liberals have a higher probability of promoting market responsibility than both social democrats and Christian democrats, and a higher probability of promoting voluntary responsibility than social democrats or conservatives. The most unexpected result is the liberal preference for family responsibility. Surprisingly given their individualism, liberals are no less likely than Christian democrats to promote family responsibility for care. Comparing liberals and social democrats directly, there is a significant difference in the probability of promoting family responsibility.
One plausible explanation is that liberals expand family responsibility to cut costs, reflecting an underlying preference against state responsibility. Liberal-led governments make care policy changes in half the countries in this study, in all three welfare regime types and both majoritarian and proportional systems. In half these cases, liberals lead governments that are on the right in redistributive conflicts, with social democratic parties in opposition. Family responsibility is preferred by liberal governments in Ireland, where Fianna Fáil favour lower levels of public spending than Fine Gael, who often govern with Labour. Ireland’s catholic history and relatively low female labour market participation make familial policies more likely, but that Fianna Fáil promote family-based care to a greater extent suggests redistributive conflict could also play a role.
When liberals govern in coalition with conservatives, as in Sweden in the late 1970s and early 1980s, and Denmark during the 1980s and 2000s, they expand market and family responsibility for care. Successive liberal–conservative coalitions in Denmark during the early 2000s aimed to cut rising costs in child welfare and mental health services by trebling the use of kinship care and incentivising municipalities to subcontract for-profit providers of residential services (Fersch and Jensen, 2011). When liberals govern with conservatives, an expansion of family responsibility could reflect a compromise between the two parties’ positions across both conflict dimensions, with liberals trading off individualist values to reduce public spending. Coalition dynamics also constrain the relationship between partisanship and care policy.
The association between liberal governments and family responsibility is also present in North America, where liberal parties are on the left in redistributive conflicts. In the United States, Democratic governments only expand family responsibility in the area of child welfare. In Canada, the 2004 introduction of ‘home care’ for acute mental illness was part of a larger Health Accord designed to address cost pressures, based on a view that community-based care care could save money (Forbes and Edge, 2009).
Even if a preference for family responsibility can be explained by a desire to reduce public spending, partisan care policy choices cannot be explained by redistributive conflict alone. Familial policy changes made by conservatives and Christian democrats actively promoted family care, even when it came at a cost. Consider the introduction of new, ongoing funding for family carers, introduced by conservatives and Christian democrats in Australia, Germany, and Italy. Other policy changes made by conservatives combined the direct promotion of family care with cost-saving initiatives. In the United States, the Omnibus Budget Reconciliation Act 1981 slashed federal funding to community mental health centres by 30% and established a new grant for family-based initiatives (Rochefort, 1997, pp. 63–72). In 1983 and 1990, the Thatcher government combined a cost-saving agenda with an emphasis on families as the primary source of care for people with chronic mental illness (Jones, 1993). Policy changes that promote family responsibility are most common among conservative governments, who have ideological reasons both to reduce the size of the state and promote the role and independence of the family.
Conclusion
During the period of deinstitutionalisation since 1950, large, custodial institutions for people with chronic and severe mental illness and children who could not be cared for by their families were replaced by a variety of care policy arrangements. While changing scientific knowledge and ideas and growing fiscal pressure were important drivers of deinstitutionalisation, the types of care policies that replaced asylums and orphanages also varied depending on government partisanship.
My analysis of the association between partisanship and care policy suggests, in line with wider theories of welfare politics, that social democrats prefer state responsibility for care over all other types of governments, while conservatives prefer private responsibility by the family or market. But parties of the centre and right also make distinct policy choices about provision by non-state actors. When compared with conservatives, Christian democrats are more likely to promote voluntary responsibility, and liberals are less likely to promote family responsibility. But liberals do promote family responsibility to a greater extent than social democrats, especially in the area of child welfare policy.
This suggests that the motivations for promoting family responsibility are complex, reflecting either a preference for family responsibility or against state responsibility for care. It is among conservative parties, where both motivations are present, that the relative preference for family responsibility is strongest. These patterns are consistent with ideological trade-offs being made across two lines of partisan conflict, reflecting redistributive preferences on the one hand and individualist or familial ideals on the other. Welfare politics has long been multidimensional, and patterns of care policy making in these two uniquely old care policy areas appear to reflect parties’ core policy preferences, shaped by the lasting ideological effects of the much older societal cleavages of class and religion, not just an electoral response to the social conditions of the post-industrial age.
Supplemental Material
Supplemental Material - Party ideology and care policy: The decline of institutional care since 1950
Supplemental Material for Party ideology and care policy: The decline of institutional care since 1950 by Katren Rogers in Journal of European Social Policy
Footnotes
Acknowledgments
Thanks to Lund University Department of Political Science, where the research for this article was conducted, and to Johannes Lindvall and Dalia Mukhtar-Landgren for their invaluable guidance.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Vetenskapsrådet (2016-00783).
Supplemental Material
Supplemental material for this article is available online.
References
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.
