Value pluralism and social policy

Value pluralists endorse a view of ethics that recognises that there are a “plurality of morally significant values” (Wolf, Reference Wolf1992, p.785). It is important to note that value pluralism does not entail moral relativism – that is, the view that there are no universal moral truths (Wolf, Reference Wolf1992; Galston, Reference Galston2002), far less ‘subjectivism’, the notion that morality represents merely the expression of emotional attitudes (Wolf, Reference Wolf1992). On the contrary, it represents a commitment to the objectivity of a multiple, but finite, set of morally relevant fundamental values (Berlin, Reference Berlin2002). Crucially for our purposes here, value pluralism stands in contradistinction to ‘monist’ theories that see all forms of goodness or rightness as reducible to a singular measure or meta-value (Galston, Reference Galston2002, p.4). Utilitarians, for instance, view ethics as fully explained by attention to the level of collective wellbeing different courses of action achieve. A rights-based deontological perspective, on the other hand, interprets the source of fundamental ethical value as lying in people’s basic rights and “morally protected sphere of individual autonomy” being respected at all times (Nagel, Reference Nagel2021, p.5).

Pluralists hold that these and other monist approaches can only ever hope to capture part of the ‘moral truth’ (Wolf, Reference Wolf1992; Wolff, Reference Wolff2015). This proposition is linked, at the social policy level, to the insight that the ‘good society’ has to be judged on multiple criteria, not by a single standard, with a balance struck between a range of (often competing) social objectives, such as the protection of personal liberty, the advancement of social justice, the nurturing of social cohesion, the pursuit of economic prosperity, and the establishment and institutionalisation of democracy (Miller, Reference Miller1999; Wolff, Reference Wolff2015). Reductionist, monist approaches can thus be obfuscatory as well as partial in their intellectual impact, as was famously argued by Isaiah Berlin:

It is surprising, in this light, that rights-based perspectives have come to carry such sway within social policy scholarship as the predominant normative lens via which to defend the welfare state (see below and e.g. Dwyer, Reference Dwyer2000; Dean, Reference Dean2015), and part of our goal in this paper is to explicate alternative perspectives to it.

While value pluralism is a philosophical concept, psychologists have bolstered its empirical foundations by mapping the ‘moral foundations’ that people draw upon when thinking about ethics in practice. Haidt and Graham (Reference Haidt and Graham2007), for example, have highlighted that people’s moral foundations extend beyond philosophers’ – and we would add social policy scholars’ – traditional concerns with justice, rights and fairness, to “group-level moral concerns” such as authority, loyalty and tradition (Graham et al., Reference Graham, Nosek, Haidt, Iyer, Koleva and Ditto2011, p. 367).

This evidence that, as a matter of political and psychological fact, people are attracted to a plurality of moral values, bolsters the democratic argument for social policy underpinned by value pluralist-style analysis (Miller, Reference Miller1999). Berlin was pessimistic regarding the incommensurability of such divergent values in policy and practice (see above). However, our pluralist approach is also inspired by Nagel (Reference Nagel1979), who more optimistically argued that “Radical disagreement about the basis of ethics is compatible with substantial agreement about what the important factors are in real life” (p.141). Galston (Reference Galston2002) also saw potential for reasonable and constructive debate about policy dilemmas in the face of plural values. “Deliberative argument”, he argued “can provide reasons for choices among qualitatively different claims even when no common measure of value is available” (Galston, Reference Galston2002, p.35). In other words, people may agree on what is ethically permissible, or required, in concrete cases, even though they arrive at this judgement for quite different reasons. One pertinent example is Deacon’s (Reference Deacon2004) analysis of the then UK Labour Government’s policies on anti-social behaviour targeting social tenants. Deacon found that three distinct normative frameworks – contractualist, paternalist and mutualist – could be engaged to justify these policies, despite many in the academic social policy community viewing them as oppressive interventions.

By systematically examining contentious matters of social policy via multiple ethical lenses, such analysis lends weight to the pluralist case by illuminating the powerful moral intuitions that each perspective taps into (Nagel, Reference Nagel2021). Moreover, in demonstrating that similar moral conclusions on specific matters of policy can be reached via very different ethical routes, they reinforce the pragmatic value of the pluralist approach. Even for monists, convinced of the reducibility of all normative considerations to one fundamental value, there may therefore be good strategic reasons to ‘reach across the aisle’ beyond their preferred ethical repertoire, in order to engage those with different normative orientations but with whom common ground may be found in particular cases.

Consistent with Berlin (Reference Berlin2002), and other key pluralist thinkers including Miller (Reference Miller1999) and Ignatieff (Reference Ignatieff2000), our proposed approach to social policy analysis lays strong emphasis on the interdependence of moral reflection, logical deduction, and empirical evidence in progressing contentious social policy debates. In particular, some apparent ethical disagreements reflect errors of logic, which can be corrected, or conflicting factual assumptions, which are apt for empirical testing. Close attention to these non-normative matters as part of more systematic ethical analysis of social policy allows us to identify the narrowed territory where genuine values-based trade-offs, or ‘tragic choices’ (Berlin, Reference Berlin2002), might be required. In the sections that follow, we seek to pull apart these strands of the highly polarised debate on welfare conditionality, foregrounding the presence (or absence) of fundamental value-disagreements that undermine effective policy debate in this area.

Welfare conditionality: the view through distinctive ethical lenses

Welfare conditionality, which links access to welfare goods and services to behavioural requirements, has become a mainstream policy tool globally. A central component of social security policy across advanced welfare states, from the so-called ‘liberal’ regimes of the UK, Australia and the US, to the social democratic welfare regimes of the Nordic countries (Immervoll et al, Reference Immervoll, Knotz and Otmani2020), conditionality also plays a core role in the evolving social protection systems of the Global South, most often in the form of Conditional Cash Transfers (CCTs) (Fiszbein et al., Reference Fiszbein, Schady, Ferreira, Grosh, Keleher, Olinto and Skoufias2009). While conditionality attached to cash benefits usually requires recipients to seek work, CCTs, and some conditional policies in Australia, focus instead on health or education-related behaviours, such as ensuring that children are appropriately immunised or attending school (Bastagli, Reference Bastagli2009; Taylor et al., Reference Taylor, Gray and Stanton2016).

Though all conditional welfare approaches share a core logic – involving the imposition of behavioural requirements, the monitoring and verification of required actions, and the use of sanctions in instances of non-compliance (or sometimes incentives for compliance) – there is wide variation in the design and specification of these components in practice. Some conditional welfare programmes have stringent behavioural requirements (e.g. 35 hours a week job search activity, learning a new language, overcoming an addiction), harsh sanctions (e.g. permanent ‘full family’ sanctions in parts of the US), and quite intrusive levels of surveillance (e.g. compulsory daily visits by a support worker) (Watts and Fitzpatrick, Reference Watts and Fitzpatrick2018). In many other instances, however, conduct requirements are much more modest (two job search activities per month in Korea, for example), and sanctions relatively mild (a first warning followed by a loss of five days benefit in Sweden) (Immervoll and Knotz, Reference Immervoll and Knotz2018).

Despite the ubiquity of, and immense variation in, programmes captured under the banner of welfare conditionality, debates in this field tend to be highly polarised, even inflammatory, in style and content. Proponents see conditionality as self-evidently justified: as an “integral” component of welfare systems (Miscampbell, Reference Miscampbell2014, p.6), and as a “key element of the mutual obligation that underpins both the effectiveness and fairness of the social security system” (Oakley, Reference Oakley2014, p.4). Opponents, on the other hand, have characterised conditionality as “institutional barbarity against the most vulnerable people in the system” (Corbyn, Reference Corbyn2016), and even described the underlying logic as reflective of “a government of the grotesque” (Ryan, Reference Ryan2014).

Academic perspectives are also divided, with a small number of scholars robustly defending the essential role of welfare conditionality (Mead, Reference Mead1986; Dunn, Reference Dunn2014), while others feel it can only be justified once onerous background ‘conditions of fair reciprocity’ (White, Reference White2003) are met. The majority view in the social policy community, however, is firmly set against welfare conditionality, which has been variously described as “neither effective nor ethical… [and] unfit for purpose” (Wright et al., Reference Wright, Dwyer, Jones, McNeill, Scullion and Stewart2018, p.10), “morally dubious” (Standing, Reference Standing2011, p.33), “illiberal” (Curchin, Reference Curchin2019, p.2) and “superfluous, pernicious, atrocious and abominable” (Freeland, Reference Freeland2007, p.75). Recent contributions have interpreted sanction-based reforms through the lenses of ‘social abuse’, ‘institutional violence’ and even ‘social murder’ (see Wright et al., Reference Wright, Fletcher and Stewart2020), and characterised sanctions as primarily a “weapon” (Redman, Reference Redman2020, p.84) deployed to “degrade” (McNeill, Reference McNeill2020, p.295) welfare recipients.

Stepping back from these impassioned and antagonistic, positions, and inspired by Deacon’s (Reference Deacon2004) normative analysis of anti-social behaviour interventions, we now map the contours of six key perspectives which our review of the international evidence indicates are the primary ethical lenses through which welfare conditionality has been debated to date (Watts and Fitzpatrick, Reference Watts and Fitzpatrick2018; see also Curchin et al., Reference Curchin, Weight and Ritter2021). While some of these perspectives have historical roots and specific points of substance in common, they are distinct in foregrounding particular rationales via which to defend or object to conditionality. In keeping with our pluralist analysis, we seek to draw out the most compelling aspects of each of these moral frameworks (‘steel-personning’ rather than ‘straw-manning’ them), while also noting the key objections they face. Crucially, we explicate their implications with regard to the moral permissibility of, or even requirement for, conditions to be attached to receipt of welfare provisions. Our desire to ‘steel-person’ people’s positions also informs our choice of normative frameworks. We focus upon established normative traditions that a reasonable person of “general goodwill and human sympathy” (Harsanyi, Reference Harsanyi, Sen and Williams1982, p.56) could plausibly endorse, rather than assume malign intentions on the part of any stakeholders.

A pluralist assessment of the ethics of conditional welfare

The normative lenses highlighted above foreground different ethical priorities, and in many cases propose alternative decision-making procedures. How persuasive one finds each of these specific ethical perspectives, and particular combinations of them, will depend on one’s own personal underpinning ‘moral foundations’ (Haidt and Graham, Reference Haidt and Graham2007).

However, looping back to our introductory comments inspired by Galston (Reference Galston2002) and other value pluralists, we would contend that there is more scope for working across these competing normative outlooks in dealing with real-world policy dilemmas than is sometimes assumed, with people often agreeing on what is ethically justified in specific cases via very different (even opposing) processes of moral reasoning. We would make the case that policy choices are most defensible where their aims and consequences can be considered as legitimate from multiple normative perspectives (Carens, Reference Carens1987; Deacon, Reference Deacon2004). For value pluralists, such an ‘overlapping consensus’ (Rawls, Reference Rawls1999[1971]) strengthens the ethical weight of any given normative conclusions as a matter of philosophical principle (Galston, Reference Galston2002). But even for non-pluralists, such ‘moral triangulation’, or ‘incompletely theorized agreement’ (Sunstein, Reference Sunstein1995), has profound democratic and strategic benefits.

Nonetheless, there will be occasions where ethical perspectives clash rather than triangulate on concrete cases of concern, pointing in different directions with regard to the justifiability or otherwise of underlying policy aims and impacts. A value pluralist conceptual analysis will not always reveal ‘hidden consensus’ on relevant matters, but will also surface irreconcilable divergence. Even in these more difficult cases, however, pluralist approaches have something important to offer in the way of constructive tools for moving debate forward.

Critical here is the pluralist readiness to engage protagonists on their own terms, testing the logical and empirical foundations of their arguments, rather than getting mired in irresolvable disagreements about fundamental values. For example, opponents of conditional welfare are on indisputably firm ground challenging utilitarian advocates to evidence that these measures do in fact maximise societal utility, while a sincere paternalist can fairly be held to account on whether conditional policies do in reality benefit those targeted. Similar empirical claims or contingencies can be identified within all plausible normative frames, and establishing an agreed evidence base on these is a key first step in moving past an unproductive debate.

Proponents of conditional welfare approaches, from whatever normative perspective, must also demonstrate a logical connection between the policy aims they seek and the use of behavioural conditions as a means to these ends. In other words, are there good grounds for believing that the behaviour change sought (e.g. more job search activity) is genuinely a solution to the social problem that the intervention seeks to address (e.g. unemployment)? They must have a convincing answer to the criticism made by many, for example, that an emphasis on welfare conditionality in out-of-work benefits implies a misdiagnosis of a fundamental structural problem (weak labour demand) for a behavioural problem (unwillingness to work) (Beatty and Fothergill, Reference Beatty and Fothergill2016).

If the social problem at issue can be empirically defended as in significant part behavioural, the ethical justifiability of welfare conditionality is also dependent on there being plausible grounds for thinking that sanction-backed requirements will in fact lead to the behaviour change desired. Relevant here is the challenge laid down by behavioural economists who have highlighted the cognitive biases and boundaries that limit expected rational responses to behavioural cues such as sanctions (Darnton, Reference Darnton2008). These social-psychological complexities require policy designers to take seriously the worldview, context and actual decision-making processes of the individuals whose behaviour they seek to influence (Curchin, Reference Curchin2017). At the very least, it requires concerted efforts to ensure that welfare recipients understand the relevant behavioural demands, and have the resources and capacity to fulfil them, and evidence that they have fulfilled them. It also requires recognising, for instance, the impact of trauma, addiction and mental health conditions on people’s ability to respond predictably when exposed to stressors or triggers. Existing evidence suggests that conditional welfare interventions, in the UK at least, often fail these basic tests of efficacy (Griggs and Evans, Reference Griggs and Evans2010; Oakley, Reference Oakley2014; Wright and Stewart, Reference Wright and Stewart2016).

The international evidence base on the impacts of welfare conditionality is limited in important respects, and is also mixed. For example, it is clear from the US case that harsh forms of conditionality attached to out-of-work benefits can effectively drive down the claimant count (Ray et al., Reference Ray, Sissons, Jones and Vegeris2014). But if no regard is had to the destination or well-being of those ‘hassled off’ welfare programmes, this is difficult to justify within most normative perspectives, except perhaps those versions of communitarianism which give overriding priority to the enforcement of key social rules regardless of the consequences for social exclusion, or in a utilitarian calculus that weighs public approval of such rules much more heavily than their impacts on those subject to them. With regard to programme goals that might be considered respectable from a wider range of moral standpoints, such as improved labour market outcomes and poverty reduction, the evidence base is highly variable. Some studies across the US and Europe point to positive impacts of conditionality on levels of work search, employment entry and earnings, while others identify no or negative impacts (Griggs and Evans, Reference Griggs and Evans2010; Ray et al., Reference Ray, Sissons, Jones and Vegeris2014), or only short-term positive effects (Hofmann, Reference Hofmann2008; Müller and Steiner, Reference Müller and Steiner2008; Arni et al., Reference Arni, Lalive and van Ours2009).

In the case of CCTs in the Global South, a stronger evidence base, often involving randomised control trials, suggests that well designed CCTs can be effective in achieving programme goals – around, for example, school enrolment and attendance, and reductions in child labour (Bastagli, Reference Bastagli2009; Fiszbein et al., Reference Fiszbein, Schady, Ferreira, Grosh, Keleher, Olinto and Skoufias2009; Kabeer et al., Reference Kabeer, Piza and Taylor2012; Pellerano and Barca, Reference Pellerano and Barca2014). Even here, however, there is ongoing debate regarding whether the conditionality of the cash transfers is key to these positive impacts, or whether they could be achieved by alternative less coercive means, including the provision of ‘unconditional’ benefits (Baird et al., Reference Baird, Ferreira, Özler and Woolcock2014; Kidd, Reference Kidd2016).

Any evidence demonstrating the effectiveness of conditional interventions in achieving their intended aims must be placed in the context of wider unintended impacts (both positive and negative). Here it is apt to note that, in both the US and UK, harsh forms of benefit sanctions have been associated with material hardships including food poverty, debt, rent arrears, homelessness, and destitution (Lindhorst and Mancoske, Reference Lindhorst and Mancoske2006; Loopstra et al., Reference Loopstra, Reeves, Taylor-Robinson, Barr, McKee and Stuckler2015; Wright and Stewart, Reference Wright and Stewart2016; Reeve, Reference Reeve2017; Fitzpatrick et al., Reference Fitzpatrick, Bramley, Sosenko, Blenkinsopp, Wood, Johnsen, Littlewood and Watts2018), and negative physical and mental health outcomes (Griggs and Evans, Reference Griggs and Evans2010; Dwyer, Reference Dwyer2018). Such findings seriously challenge the utilitarian and paternalistic case for these strong sanctions, and probably the contractualist case too, at least as regards those theorists concerned with the fairness of the imagined social contract (White, Reference White2003). These negative impacts may even be seen to undermine the potential communitarian argument for strict sanctioning regimes, given their exclusionary consequences (Gray, Reference Gray, Askonas and Stewart2000); they certainly seem fatal to any plausible social justice-based defence of these measures.

Concluding remarks

The core aim of this paper has been to advocate for the application of a value pluralist perspective to ethical questions in social policy. By offering a means to map systematically the diversity of moral registers within which social policy concerns might legitimately be considered, we would argue that this pragmatic pluralist approach can aid constructive policy dialogue and decision making. There are practical, democratic and strategic grounds to adopt a value pluralist analytical approach, even if in philosophical terms one remains a ‘monist’ wedded to a singular ethical measure or meta-value. A pluralist approach helps provide tools for the development of robust policies that are ethically permissible from multiple normative perspectives. Such moral triangulation, where possible, offers the surest means to pursue policy options that gain the widest public endorsement. Moreover, serious attempts at moral triangulation, even where not fully realised, shine a bright analytical light on genuine areas of irreconcilable difference, as well as on areas of (hidden) agreement, in order to maximise clarity and mutual comprehension on both. Confronting relevant matters from multiple plausible moral perspectives increases the rigour, thoroughness and scope of the ethical evaluation task.

For case study purposes, we have explored the potential contribution of six key normative lenses – rights, utilitarianism, contractualism, communitarianism, paternalism and social justice – in considering the ethical legitimacy of welfare conditionality. We have elucidated the implications of each of these perspectives for assessing the ethics of conditional welfare approaches, and sought to pull out the empirical claims and logical assumptions on which their support for or critique of such approaches depends. Rigorous testing of these non-normative elements is an invaluable first step in bringing diverse starting positions closer together when applied to concrete policy choices.

One key conclusion of our substantive case study analysis is that there is unlikely to be a single, consistent answer to the question of whether welfare conditionality is ethically ‘justified’. Much turns on the context, focus, rationale and, most crucially, impacts of the specific instance of conditionality under discussion. As argued above, the harshest forms of benefit sanctions found in the US and UK are difficult to justify under most (though not necessarily all) plausible normative frameworks. However, this is far from ruling out the possibility of a broader ‘overlapping consensus’ of support for milder forms of conditionality that can be empirically demonstrated to benefit welfare recipients and/or wider society. Some of the CCT programmes in the Global South reviewed above may come into this category, for example, and one can envisage utilitarian, contractualist, communitiarian, paternalist and social justice arguments triangulating in support of at least some such programmes. Only those rights-based theorists, currently so dominant in UK social policy, that reject conditional welfare approaches as a matter of principle, regardless of their practical consequences, may stand outside such an overlapping consensus.

We hope to have demonstrated the utility of the value pluralist approach in contentious areas of social policy, and also that a great deal of the consensus-building work in a field like welfare conditionality can take a non-normative form, by establishing the relevant factual basis for decision making and expounding the logical consequences that flow from particular value commitments. In many cases, protagonists coming from very different moral orientations may find themselves drawing similar substantive conclusions on concrete policy questions. It is true that “pockets of indeterminacy” (Wolf, Reference Wolf1992, p.788) will remain even after all these avenues for reasoned debate have been exhausted. Berlin’s ‘tragic choices’ thus cannot be avoided altogether but they can be greatly narrowed in scope.