Why do normative work in social policy?

In ideal terms, social policy can be understood as a combination of facts and values: the latter informing the aims of a policy, with evidence about ‘what works’ guiding how this should be achieved. In practice, it is acknowledged that there is a ‘complex interdependence’ between social norms, policy positions and social outcomes, which have a particular impact where value-laden issues such as poverty and inequality intersect with questions of state provision (Dean and Brady, Reference Dean and Brady2015; Farrugia and Gerrard, Reference Farrugia and Gerrard2015; Shutes, Reference Shutes, Dean and Platt2015; Vizard, Reference Vizard, Dean and Platt2015). Welfare and homelessness policy are highly value-laden areas, associated with emotive debates around who should receive help (Zufferey, Reference Zufferey2014; Gross and Wronski, Reference Gross and Wronski2019) and how (Watts, Reference Watts2014; Watts et al., Reference Watts, Fitzpatrick and Johnsen2018). These arguments often rest at the intersection of moral philosophy and policy in debates about distributive justice.

A normative approach pays explicit attention to the values at play in policy, offering opportunities to develop conceptual clarity and accurate description of the policy positions (Bengtsson, Reference Bengtsson1995; Wolff, Reference Wolff2011). It asks what values are being drawn upon in the construction of a policy, or – if this is not stated outright – with what normative perspectives it is aligned. Together with evidence of the policy’s effects, such clarity can inform a process of ethical scrutiny and evaluation (King, Reference King2011; Taylor, Reference Taylor2018), addressing questions such as the policy’s coherence with popularly held values and whether it achieves its stated aims (Gerring and Yesnowitz, Reference Gerring and Yesnowitz2006; Wolff, Reference Wolff2011; O’Leary and Simcock, Reference O’Leary and Simcock2020). Finally, the work of evaluation leads naturally to considerations of how to develop and justify policies in the future. These range from practical discussions about bringing their impact in line with stated aims or accepted norms, to developing theoretically- or ideologically-informed cases about what those aims should be (Bengtsson, Reference Bengtsson1995; King, Reference King2003; Stears, Reference Stears2005). As such, normatively-informed social policy can both build public acceptance by bringing policy in line with commonly-held values, and articulate visions of the future that inform new aspirations (Wolff, Reference Wolff2011).

Two approaches to normative work in social policy: ‘ideal’ and ‘non-ideal’ theory

Approaches to normative work in social policy can be differentiated according to their adoption of ‘ideal’ or ‘non-ideal’ theory. The former is concerned with articulating the best possible social structures, by applying ideas about what is right or good to some idealised assumptions about society. Normative goals are defined first and the shape of society follows, meaning that the value-based perspective is implemented ‘top down’. Richard Titmuss, a keen proponent of clarity about values in social policy who argued the discipline will “inevitably be concerned with … what we (as members of society) want (the ends); and how we get there (the means)” (Reference Titmuss1974: 132), has been described as an ideal theorist because he conceptualised social policy as a tool for bringing about pre-defined goals, such as reducing inequality and strengthening social solidarity (the ends), through the delivery of universal welfare rights (the means) (Rodger, Reference Rodger2003; Offer, Reference Offer2006).

Work tracing the history of normative thought in social policy suggests that the idealist approach dominated for much of the 20^th Century (Offer, Reference Offer2006), but such positions encounter several challenges when applied to real-world problems. First, different but equally valid foundational principles cannot be meaningfully weighed up or traded off against one another, and tensions between such ‘incommensurable’ values can create conflict or unclarity about the appropriate course of action in policy (Nagel, Reference Nagel1979; Raz, Reference Raz1986; Berlin, Reference Berlin2002; Lukes, Reference Lukes2008). Second, empirical evidence shows that people’s everyday moral reasoning, including in realms of social policy and distributive justice, draws on multiple values (Miller, Reference Miller1992; Haidt and Graham, Reference Haidt and Graham2007; Fletcher and Redman, Reference Fletcher and Redman2022; Gandenberger et al., Reference Gandenberger, Knotz, Fossati and Bonoli2022), and the prevalence of different value judgments in the public sphere may vary with socioeconomic conditions (McArthur and Reeves, Reference McArthur and Reeves2019; Noureddine and Gravelle, Reference Noureddine and Gravelle2020). This suggests that dogged application of ideal theory in a social policy context risks being “undemocratic, practically challenging, and strategically unwise” (Watts-Cobbe and Colliver, Reference Watts-Cobbe, Colliver and Adler2022: 193). The permanence and significance of pluralism in contemporary society is such that application of an ‘ideal theory’, or a single coherent normative position, is increasingly accepted as impractical in a policy setting (Raz, Reference Raz1986; Young, Reference Young1990; Mendus, Reference Mendus2002; Wolff, Reference Wolff2011; Valentini, Reference Valentini2012).

In contrast to a values-led approach, ‘non-ideal’ theory uses social realities as its starting point. This approach attempts to engage with the array of normative and non-normative factors (e.g. public opinion, political interests and the status quo) from the ‘bottom-up’. It draws upon tools from political theory and moral philosophy to help make sense of the moral difficulties associated with social issues and move the policy conversation forward in a more normatively-informed manner (King, Reference King2011; Wolff, Reference Wolff2011; Valentini, Reference Valentini2012). A degree of pragmatism is involved in finding policy solutions that can meet with public and political support. Amongst other advantages, a non-ideal approach has been praised for its use of empirical information to gain an accurate understanding of a policy problem ‘as it is’ within an imperfect social context, and taking people ‘as they are’, meaning that principles and solutions can be tailored to their capacities (Anderson, Reference Anderson2010). As such, it responds to diverse values held within the population and seeks areas of consensus that may guide policy actions, even where underlying beliefs about what is right or good continue to differ (Wolff, 2011; Reference Wolff, Lever and Poama2019).

The framework presented below is informed by this ‘non-ideal’ approach. It assesses what values are at play within a given policy sphere, asking what justification they might offer for a particular policy position. This forms a basis for exploring whether the policy implementation is a defensible expression of such values.

Seeking defensible coherence

This paper introduces the idea of ‘defensible coherence’ as a simple test or rule of thumb for assessing the ethical defensibility of a policy landscape. It starts with the assumption that a policy’s actions and associated outcomes ought to reflect its normative commitments i.e. it should achieve its aims. But influences on policy development and implementation are diverse, including potentially hidden objectives and conflicting public and private goals. For instance, factors such as local structures, budgets, availability of affordable accommodation, service attitudes and working culture are known to affect the way that welfare and homelessness policies are interpreted and implemented (e.g. Dwyer et al., Reference Dwyer, Bowpitt, Sundin and Weinstein2014; Alden, Reference Alden2015; Midgley, Reference Midgley2016; Dobson, Reference Dobson2018; Ahmed et al., Reference Ahmed, Madoc-Jones, Gibbons, Jones, Rogers and Wilding2019). As such, policy must be understood as the product of multiple compromises rather than an expression of ‘ideal’ solutions (O’Leary and Simcock, Reference O’Leary and Simcock2020). Policy success cannot, therefore, be assessed solely according to the fulfilment of stated goals.

The test of defensible coherence is a strategy of normative evaluation that aims to account for the ‘messiness’ of policy in practice. It seeks to establish conceptual clarity around policy values, looks at a policy’s inward- and outward-facing normative coherence including potential explanations for deviances or discrepancies, and makes an evaluation based on the balance of these factors. A policy’s internal coherence refers to how far its stated values or normative alignments are reflected in its practices and outcomes. Coherence is high when the values and other influential factors are ‘pulling in the same direction’ and undermined when they conflict (May et al., Reference May, Sapotichne and Workman2006). A policy landscape is likely to be defensible as a whole where this coherence is apparent between an array of related policies.

As it is not realistic to expect an exact correspondence between policy and practice, it can be informative to look for signs of ‘progressive realisation’ of normative commitments (Lynch, Reference Lynch2005; Brems, Reference Brems2009). This involves considering whether values are being partially or incrementally expressed in a policy area and whether its overall trajectory is consistent with its stated principles. An apparently incoherent policy may be considered defensible if a ‘consistent’ normative position is factored into decision-making but subordinated to other policy influences i.e. there is evidence of an attempt to strike an appropriate balance between multiple influences. Nonetheless, deviations from stated values should be explicable with reference to overriding alternative priorities and, where they are not, this will raise questions about the defensibility of the position and could indicate a need for greater scrutiny and accountability in that area.

Desert

‘Desert’ captures the principle that the distribution of social goods or benefits should reflect something about the recipient’s behaviour or ‘performance’. Assessments of desert can be influenced by factors including effort (dedication to the performance), contribution (merit of the performance), motivation (degree to which the performance was intentional) and luck (degree to which the performance was – or was not – in the agent’s control) (Miller, Reference Miller1976; Olsaretti, Reference Olsaretti, Dryzek, Honig and Phillips2008). As such, the agent’s responsibility for their actions is key to the assessment, and a person who acts ‘deservingly’ is understood to earn a fitting form of treatment.

Within the UK welfare and homelessness systems, desert-based reasoning is reflected where entitlements are conditional upon an applicant’s actions. Examples include the DWP’s Claimant Commitment, which requires recipients of Universal Credit to engage in a range of job search activities to receive out-of-work benefits (UK Government, 2021), and the ‘intentionality test’ which ends local authorities’ duty to provide settled accommodation to homelessness applicants if they are assessed as responsible for their circumstances (Robson and Poustie, Reference Robson and Poustie1996). The contributory principle is visible in the different rates of unemployment benefit paid according to an individual’s history of national insurance contributions (White, Reference White2003; Watts and Fitzpatrick, Reference Watts and Fitzpatrick2018). It is an influential form of desert-based reasoning amongst welfare recipients and policymakers alike: for example, benefit claimants may justify drawing welfare entitlements with reference to their own actions, such as making tax contributions (Dwyer, Reference Dwyer and Dean2004).

This principle of contribution offers a potential explanation for young people’s lower rates of minimum wage and benefit entitlements. Their employment entitlements could be linked to a ‘lower offer’ – both in terms of the quality of their contribution when at work and the quantity of their contribution over time.

The rationale for lower benefits follows from the lower minimum wage: paying young people the same rate of unemployment benefit as older adults risks creating a ‘perverse incentive’, potentially discouraging them from entering work.

But despite this prima facie plausibility, there are reasons to be concerned about such an expression of desert-based policy. Key to the principle of desert is that it responds to individual performance (Miller, Reference Miller1976). Treatment based on aggregate performance across the category of young people crucially misses this element, inappropriately determining one person’s treatment according to anticipated performance across the group. This means that individual young people may experience an unfairly low wage that fails to capture their true contribution: “…young people would say they work just as hard, if not harder, than people who are older than them,” (Third-Sector Organisation, England). Such an outcome is illegitimate from a desert-based perspective and, as such, it fails to offer a satisfactory normative explanation for the welfare system’s differential treatment of young people compared with older adults.

In relation to homelessness, the intentionality test can be understood as a desert-based policy because it links applicants’ acts or omissions to their housing entitlements. Whereas the test applies in full to young people presenting as homeless in England, policy developments in Scotland and Wales have attempted to reduce its impact upon this group (Housing (Wales) Act, 2014; Scottish Government, 2019).Footnote ¹ The appropriateness of doing so hinges upon whether there is a meaningful difference in the bases of desert between younger and older adults – do effort, contribution, motivation or luck affect their actions to different degrees?

Interviewees, including those attached to national government, drew on common conceptualisations of young people as ‘inexperienced’ or ‘immature’ (Harding, Reference Harding2004) to imply that they were less responsible for their actions than older comparators: “I do not think you can be expected to take that level of responsibility if you’re sixteen. And that’s to do with physiology and maturity and everything,” (National Government, England). If this is the case, then assessments of desert must account for this discrepancy, perhaps combined with less extensive life experience, in considerations of motivation and intent. To the degree that developmental stage inhibits reasoned control over action, then it would be consistent to reduce the impact of desert-based policies on young people.

Need

Concepts of ‘meeting need’ range from preconditions for avoiding serious harm (‘thin’ views, e.g. night shelter homelessness provision), to everything required to live a good life (‘thick’ views, e.g. access to long-term settled accommodation) (Doyal and Gough, Reference Doyal and Gough1991; Dean, 2010; Reference Dean2013). Reflecting the reality of constrained resources, the UK welfare system takes a thin ‘sufficientarian’ approach to distribution – the needier a person is, the more they receive, up to a minimally sufficient threshold (Crisp, Reference Crisp2003; Brock and Miller, Reference Brock and Miller2019). For example, although its sufficiency is contested (Fitzpatrick et al., Reference Fitzpatrick, Bramley, Sosenko, Blenkinsopp, Johnsen, Littlewood, Netto and Watts2016), the DWP’s non-contribution rate of unemployment benefit represents a minimum income ‘safety net’ below which no one should fall (Wolff et al., Reference Wolff, Lamb and Zur-Szpiro2015). The system also recognises that the level and type of resources required to bring people up to the same sufficiency threshold can vary (Dworkin, Reference Dworkin1981): for instance, the Personal Independence Payment (PIP) benefit provides additional resources to support disabled people with living and mobility costs.

If young people were shown to have fewer or less significant needs, such as lower outgoings or “less responsibilities I guess – they’re less likely to have children” (Third-Sector Organisation, England), then their lower benefit entitlements might be justified because they would require fewer resources to meet the sufficiency threshold. But interviewees contested this premise, arguing young people have the same basic needs and require the same level of resources as older adults:

Instead of being consistent with a needs-based response, young people’s structural disadvantages within the welfare system leave them with fewer resources to meet their basic needs. This is borne out by evidence that young people receiving the SAR experience severe difficulties finding affordable accommodation (Watts et al., Reference Watts, Johnsen and Sosenko2015) – their lower benefit entitlement and minimum wage make it more difficult to top up housing benefit to meet rental costs (Fitzpatrick et al., Reference Fitzpatrick, Pawson, Bramley, Wilcox and Watts2017; Greaves, Reference Greaves2019) – with rates of hidden homelessness increasing as a consequence (Simcock, Reference Simcock2022).

An alternative explanation for their lower benefits is that young people could be more likely to receive familial support in meeting their needs, meaning they require less state support. This reflects a normative position advanced by former Prime Minister David Cameron during the introduction of the SAR, that “the only time a young adult should leave the family home is once they have reached financial independence” (Wilkinson and Ortega-AlcÁzar, Reference Wilkinson and Ortega-AlcÁzar2017: 334). Family plays an important role in supporting some young people: growing exposure to social risks and inaccessibility of the housing market has resulted in increasing numbers of young people living at home (Wong, Reference Wong2019), and returning to the family home is one of the main routes out of homelessness for young people (Mayock and Parker, Reference Mayock and Parker2019). But not all young people have recourse to such support, and for many – particularly women and LGBT people – ‘home’ does not represent a place of safety (Wilkinson and Ortega-AlcÁzar, Reference Wilkinson and Ortega-AlcÁzar2017). Basing entitlements on aggregate expectations is at odds with the individual means-testing rationale operating elsewhere in the DWP’s social security system. Young people who are eligible for benefits have, by definition, been assessed as requiring state support. In this context, their lesser entitlements appear to contradict the principle of responding to instantiated need.

Although England’s statutory homelessness system represents an over-arching response to acute housing need, needs-based reasoning is not an obvious driver of young people’s treatment within that domain. The groups picked out for settled accommodation through the priority need test are identified in terms of risk characteristics rather than instantiated need, meaning it is more accurately conceptualised as an instance of vulnerability-based policy (discussed below).

Outside of the statutory system, interviewees suggested that a needs-based response to young people experiencing homelessness should look different to that for older adults. Accommodation options were a particular focus, with calls for local authorities to “make sure your supply, your commissioned supply in particular, meets [their] needs,” (Third-Sector Organisation, England). This was often couched in terms of providing additional on-site support:

Supported communal accommodation is a popular service model for young people experiencing homelessness (Steen and MacKenzie, Reference Steen and MacKenzie2016), and some evidence suggests stays in this kind of temporary accommodation contribute to more successful transitions to independent tenancies (Crane et al., Reference Crane, Warnes, Barnes and Coward2013). Although these ‘housing readiness’ models have met with some criticism (Stewart, Reference Stewart2019; Hoolachan, Reference Hoolachan2020), the provision of support represents an attempt to address more than young people’s thin need for shelter, contributing to thicker need-fulfilment through developing a ‘plan of life’ (Doyal and Gough, Reference Doyal and Gough1991; Dean, Reference Dean2010). As such, whilst the statutory system does not formally respond to acuity of need, elements of third-sector practice are sensitive to the types of needs young people have compared to older adults, aiming to tailor services accordingly.

Vulnerability

The concept of vulnerability is prominent – even fashionable – in everyday discourse around welfare, but its meaning is not always clearly articulated. Kate Brown (Reference Brown2015) describes vulnerability as a ‘shadow concept’ of risk, and I distinguish ‘vulnerability’ from ‘need’ as the difference between the potential for harm and the instantiation of harm. It is also useful to distinguish two types of risk that might render an individual or group ‘more vulnerable’: a) higher likelihood of incurring harm, and b) likelihood of incurring more severe harm. Some people may be vulnerable in both respects. For example, young people with experiences of multiple disadvantage are more likely to become homeless (Fitzpatrick et al., Reference Fitzpatrick, Pawson, Bramley, Wilcox and Watts2017) and, compared to their peers, young people experiencing homelessness are more likely to incur harm due to mental ill-health, violence and sexual exploitation (Hodgson et al., Reference Hodgson, Shelton, van den Bree and Los2013; Heerde et al., Reference Heerde, Scholes-Balog and Hemphill2015; Heerde and Patton, Reference Heerde and Patton2020). These factors can also intersect with other sources of oppression such as class, gender and sexuality that exacerbate the severity of these harms (Fraser et al., Reference Fraser, Pierse, Chisholm and Cook2019; Norris and Quilty, Reference Norris and Quilty2020; Watt, Reference Watt2020).

Causes of vulnerability span multiple dimensions, ranging from those ‘innate’ to an individual, to situational and ‘structural’ factors like exposure to social risks (Fineman, Reference Fineman2008; Brown, Reference Brown2019). Reflecting the ‘thin’ and ‘thick’ concepts of need described above (Dean, 2010; Reference Dean2013), these risks range from physical to social and psychological forms of harm (Caraher and Reuter, Reference Caraher and Reuter2017). Individual-level sources of vulnerability might include a person’s characteristics, experiences or behaviours – for instance, young people’s ongoing neurodevelopment and lack of life experience (Brown, Reference Brown2015; Emmel, Reference Emmel2017). Structural vulnerability from social risks includes material deprivation – for instance, through poor nutrition and shelter – and is shaped by environmental factors including access to key resources such as welfare state provision (Caraher and Reuter, Reference Caraher and Reuter2017; Emmel, Reference Emmel2017). As such, a person or group’s vulnerability can be understood as a product of internal and external influences.

Within the English homelessness system, the term ‘vulnerability’ is commonly deployed as grounds for targeted exceptional treatment: in particular, the priority need test gives some young people an entitlement to settled housing (Homelessness Act, 2002; Brown, Reference Brown2015). But the level of protection afforded to individual young people is determined by an uneven policy landscape, involving complex interactions between statutory requirements: “What I see at the moment is there’s kind of inequality depending on what route a young person goes down” (Local Authority, England). For example, whilst English homelessness policy makes long-lasting provision for care leavers, including them in priority need up to at least age 21, there is a drop-off in protection for most other young people after the age of 18 (MHCLG, 2018). Multiple interviewees reported that these differences in treatment feel unfair to young people who are not care-experienced and that the current system can fail to recognise or respond to vulnerabilities in that group:

There was a widespread sense that vulnerability-based considerations for young people in policy were failing to translate into the expected protections in practice. Whilst the English Code of Guidance for implementing homelessness legislation states there must be special consideration about youth when assessing vulnerability (MHCLG, 2018), interviewees suggested this may count for little when resources are constrained:

As young people age, a gradual drop-off in protections could be consistent with vulnerability-based policy because risks stemming from immaturity and inexperience may diminish over time. But interviewees suggested that expectations and practice in the statutory homelessness system are contradictory, with young people aged 18 and over facing higher risks of harm compared with 16-17 year-olds because of barriers to accessing priority need. In this sense, whilst vulnerability-based reasoning appears to provide a sound normative basis for the treatment of young people in English homelessness policy, current practice falls short of delivering on this goal.

Rather than providing vulnerability-based protection, the welfare system was highlighted as a source of vulnerability for young people: “At a policy level it’s quite clear that young people, by their very nature, are vulnerable. Because of… their vulnerability in terms of the benefits system,” (Third-Sector Organisation, Wales). Their experiences echoed evidence that young people’s lower rates of benefit entitlement place them below the destitution line for single people living alone (Watts et al., Reference Watts, Johnsen and Sosenko2015; Fitzpatrick et al., Reference Fitzpatrick, Bramley, Sosenko, Blenkinsopp, Johnsen, Littlewood, Netto and Watts2016; Homeless Link, 2018):

Young people’s disadvantaged position in DWP employment and unemployment policy makes it more difficult to top up housing benefit to meet rental costs, issues which were exacerbated by the 2016-2020 Local Housing Allowance freeze (Fitzpatrick et al., Reference Fitzpatrick, Pawson, Bramley, Wilcox and Watts2017). The cumulative impact of their position has made young people more susceptible to the rising cost of living and less likely to be able to afford rent in their area. As such, the realities of young people’s position in the welfare system are highly inconsistent with arguments that they should be treated as ‘more vulnerable’ due to their comparative immaturity and inexperience (Brown, Reference Brown2015).