Bringing in the third pillar: protective legislation and the welfare state

ABSTRACT

Welfare state research over the past five decades has largely focused on two broad social policy pillars: (1) arrays of income programmes, such as social assistance, unemployment insurance, family allowances and pensions, and (2) a range of social services, including childcare, healthcare, social housing and education. However, an important cluster of policy measures, constituting a third policy pillar, has been largely neglected in the cross-national welfare state research to date, protective legislation. It comprises the networks of protective and regulatory laws and constitutional rights emplaced to secure and promote our welfare. Central to all social policy areas, these measures, including rent control policies, workplace health and safety laws, legislation preventing discrimination against people with disabilities and child corporal punishment bans, are distinct from laws that determine access to income programmes and social services or set out the terms of their provision. Rather, they constitute a standalone third category of policy measures, are key components of all welfare states and, depending upon how they are actualised, can be just as central to our well-being and the promotion of our welfare. They can also have compounding or synergistic interactions with policies in the other two policy pillars.

KEYWORDS:

• Third pillar
• protective legislation
• regulatory legislation
• rights
• policy domains
• cross-pillar coordination

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. This task often proved challenging because the character of different policy domains (e.g. labour market policy or family policy) and the specific policy instruments within them (e.g. unemployment insurance or childcare) could differ markedly within the welfare state of a single nation. Nations with developed, universal healthcare systems, like the UK and Canada, for example, were typically slotted in the residual/social safety net category because targetted or ‘selective’ programmes requiring some form of means test or income test to access them were prominent components of their welfare states. Canada also provided a universal family allowance programme relatively early (1944) – three years before Sweden – but its benefit level was much lower (Olsen 1994).

2. This tripolar typology underscored the impact of the varied interactions among the state, market and family on the character of social support in each of its three models or regimes: liberal, social democratic and conservative. It also highlighted the different orientation and goals of the policy approaches in each regime, reflecting the distinct constellations of political forces that shaped them – an aspect not as fully addressed in earlier research proposing three models of welfare (e.g. Titmuss 1974, Furniss and Tilton 1979).

3. Research on this second pillar could also include consideration of a range of ‘in kind’ benefits associated with the provision of some social services, such as healthcare or old age care (e.g. prescribed drugs and medications, wheelchairs, walkers and canes, dental implants, eyeglasses and hearing aids) or education (e.g. books, school supplies and nutritional meals provided in school cafeterias).

4. Norway’s social democratic welfare state, for example, was notably less service-heavy than those in Sweden and Denmark; France’s conservative welfare state was more service-oriented than that in Germany; and the universal healthcare systems in Canada and the UK distinguished them from the US, the closest approximation of the liberal regime ideal type.

5. Although they interact with the legislation that established first and second pillar provisions, determines their character, and sets out and enforces their conditions for receipt, third pillar legislation constitutes a separate form of social support. Thus, third pillar measures, such as laws prohibiting discrimination against groups, or banning corporal punishment of children, are distinct from laws and regulations which qualify someone for unemployment insurance benefits or social assistance (pillar 1), or set out details concerning the character and provision of childcare or social housing (pillar 2). Within the workplace health and safety policy area, for example, laws that limit the length of the workday, or protect workers from exposure to hazardous substances and dangerous conditions (pillar 3) are notably different from laws that determine eligibility for claiming workers’ compensation payments (pillar 1) or ensure access to treatment and healthcare when someone is injured in the workplace (pillar 2). In the housing policy domain, housing allowances and tax credits for renters (pillar 1), the provision of social housing (pillar 2) and rent controls and eviction protection laws (pillar 3) can be distinguished.

6. A notable partial exception here is feminist research identifying the gendered character of welfare states and proposing alternative typologies highlighting legislation related to violence against women, gender discrimination and divisions of labour, affirmative action, caring work and body rights, (e.g. Julia et al. 1999). Although this research was specifically targetted at women and gender policies and politics, it provides insightful discussions of how income programmes, social servicers and protective legislation have been deployed across nations, and their varied impacts.

7. This includes the literature addressing the ‘regulatory state’ (e.g. Jessop 1990, Majone 1997, Leisering and Mabbett 2011, Levi-Faur 2014).

8. Designed to render peasants and workers increasingly dependent upon markets, employers and capitalists, many of the laws that were introduced as capitalism emerged and took root served to enclose the commons and repeal longstanding rights to resources that were previously accessible, shared and collectively managed (Linebaugh 2008, Angus 2023).

9. This orientation would be largely replicated by many social assistance programmes (‘welfare’) introduced across Europe and North America in the 20^th century, regulating, disciplining and punishing the poor (Olsen 2021, pp. 122–131, Piven and Cloward 1972, Joe et al. 2011, Wacquant 2009).

10. It is important to remember that progressive legislation, such as child labour laws, for example, can routinely violated or reversed, and that there can be marked variation in their character and impact across jurisdictions within a single nation (Gerstein 2024).

11. In 2024, France amended its constitution to enshrine women’s right to an abortion – the first nation to do so – as rights have been rolled back in the US.

12. A recent bibliometric study highlighted the extensive involvement of the alcohol industry in scientific research on alcohol consumption (Golder et al. (2020).

13. On the challenges and limitations of third pillar legislation and other welfare state interventions within the context of capitalism and corporate power see, for example, Glantz (2020), Gough (2016), Madden and Marcuse (2016), Olsen (2021, chapters 6 & 7); Tulleken (2023) and Whyte David (2020).

14. However, Houston also upholds an ordinance outlawing the provision of food to people who are homeless but who are not located within the one area that has been designated for this – a third pillar regulation that undermines their welfare and criminalises the actions of individuals and volunteer groups providing such support (Holpuch 2023).

15. Despite the existence of tax-financed, universal public transportation (pillar 2) in some nations (such as Luxembourg) or regions or cities within them (e.g. Kansas City, Missouri) and/or the availability of transportation allowances and subsidies for particular demographic groups (e.g. students, children, the elderly) transportation policy is virtually never directly included in discussions of inequality, well-being and welfare states. However, Layard’s (2022) discussion of the bus project in Bristol provides another example of the kind of research that could be further explored cross-nationally and integrated with the welfare state literature.

16. The ‘bio-medical’ approach is also referred to as a ‘client-service’ model because its primary form of redress is through the support of cadres of professionals in medicine (e.g. doctors, nurses, physical therapists, occupational therapists, physiotherapists, and speech pathologists) and in other allied fields such as social welfare and education (e.g. clinical social workers and special education teachers) whose central focus is to treat, fix, cure or rehabilitate people.

17. Section 504 of the 1973 Rehabilitation Act marked an initial change of direction for disability policy in the US, away from the medical model but the new approach was most evident with the American Disabilities Act (ADA) introduced in 1990. The ADA was at odds with earlier disability policies, such as Social Security Disability Insurance (SSDI). SSDI was shaped by medical model ideas and principles that provided support based on a presumed inability to work, fostering greater segregation. The explicit goal of the ADA, in stark contrast, was to bring people with disabilities into the labour force so they could become’ taxpayers instead of tax burdens’ (Hahn 1996, Pettinicchio 2019). For a thorough discussion of ‘policy layering’ see Burke and Barnes (2018), Hacker (2004) and Thelen 2003).