Discourse, framing and audiences

It would be naïve to assume that TFP was simply aimed at the particular families it identified, as if it were merely a neutral managerial response to problems whose definition was uncontroversial. Any social research that assumed this would be doomed to misunderstand its object, though it would make it easier for its authors to get a hearing from politicians. Social policies are not simply ways of dealing with alleged problems, but are always also – and sometimes more so – ways of advancing broader political strategies and worldviews, and gaining political support. Indeed, the electorate, reached via the news media, is arguably often the primary addressee.

The discourse of any particular policy or programme must therefore be understood not on its own but in its political context, both in terms of its origins and its intended audiences and effects. Such discourse overlaps with other kinds of discourses, including those of political speeches, broadsheet and tabloid newspapers, television news and the internet. As few members of the public read policy documents, their message is intended to be relayed through these other media, albeit it in their respective genres.Footnote ¹

Political discourse, especially regarding topics of popular concern, is predominantly eristic – that is, conducted in order to gain victory over opponents and persuade others by any means that works, regardless of whether the arguments are logical or illogical, empirically supported or not. It is driven by self-interest rather than a desire to find the best accounts, explanations, and courses of action, and it is characteristically impatient (Sayer, Reference Sayer2008). By contrast, in an ideal kind of discussion, or in an ‘ideal speech situation’, as Habermas termed it, the only motive would be that of finding the better argument, and participants would be willing to take time to consider others’ ideas and change their minds if they proved superior (Habermas, Reference Habermas1979). Academic discourse aspires to this ‘disinterested’ form of argument, though of course it often falls short of the ideal. Hence, whereas academic discourse generally seeks to remove ambiguity, political discourse often exploits it, as is unusually clear in the very term ‘troubled families’.

At the same time, we have to recognise that in everyday life, in contrast to (ideal) academic discourse, the persuasiveness of particular political discourses depends heavily on how they play into implicit but powerful systems of value and emotion (Lakoff, Reference Lakoff2004). Thus, social policies may sometimes be best understood as being partly intended to change or reinforce these wider feelings and prejudices as well as to address their target issue. I shall argue that this is the case for the TFP.

Finally, in assessing any social policy, it is always important to question the way in which it frames its topics: What is taken as given that might be questioned? What things are treated as parameters rather than variables? How does the definition and bounding of the problem limit diagnoses and prognoses?

Much ‘policy science’ takes the goals and framing of policies as given and merely assesses ways of achieving those goals that are allowed within that framing. Different framings not only suggest different accounts and explanations, but different ways of assigning responsibility for problems and hence different allocations of blame or credit. They lead not merely to technical judgements about causes and what should be done but moral judgements of the behaviour of individuals and institutions.

We can now look at the TFP with these points in mind.

The framing of the Troubled Families Programme

Recent policies such as TFP reflect the evolution of neoliberalism over the last four decades. Attacks on the welfare state, begun under the Thatcher and Major administrations and continued by New Labour, accused it of allowing its beneficiaries to avoid taking responsibility for their own actions and the consequences of these actions. Since the 2008 crash, the alleged necessity of austerity has been used to justify further welfare cuts.

One of the key characteristics of neoliberal political discourse is its selective refusal of social, and especially economic, explanations of social problems. Thatcher's hardline denial of the notion of society was later softened by communitarian influences under Blair and Cameron's notion of ‘the Big Society’, but the evasion or denial of the effects of economic processes on life chances and behaviour has remained a constant. Individuals are thus held largely or wholly responsible for their fortunes. If their pasts have been difficult, they can and must take steps to free themselves from their pasts from now on. This individualistic understanding of society fits well with commonsense thought. There are never job shortages, just under-motivated jobseekers who need to try harder to get work. There are never structural problems of poorly paid jobs, just individual workers who could be paid more if only they gained skills and worked harder. The fallacy of composition – that what is possible at certain times for some individuals must also be possible for all simultaneously – is a staple of neoliberal political discourse; zero-sum games cannot be acknowledged. Individuals are thus ‘responsibilised’ for dealing with problems that may originate elsewhere and which formerly might have been addressed by the welfare state (Shamir, Reference Shamir2008).

This framing fits with the idea of a meritocratic society and what psychologist Melvin Lerner called the belief in a ‘just world’ – a world in which people largely get what they deserve, so that effort and merit are rewarded and their opposites penalised: people are paid what they are worth, and worth what they are paid (Lerner, Reference Lerner1981). In political discourse, it is often unclear whether concepts like meritocracy are meant to describe current society, predict a new world that lies ahead or prescribe one. Such ambiguity is useful.

Although many social scientists emphasise structural causes of social problems, there is also a long history of social researchers lending support to what might be termed anti-poor policies, for example through notions of a ‘culture of poverty’, ‘welfare dependency’ or an ‘underclass’ (Murray, Reference Murray1994; Welshman, Reference Welshman2006). Under neoliberalism, the influence of these latter researchers has grown, while those social scientific explanations which give considerable weight to social causes of individual characteristics and behaviour are viewed with suspicion both by neoliberals and the public as excusing individuals of any responsibility for their actions. But then, as we shall see, reconciling individual responsibility and social influences is no easy matter, and sociology in particular has often left little or no room for individual reflexivity and responsibility.

Over the last twenty years we have seen a gradual shift in social policy discourse from explanations for social problems that refer to history, backgrounds and social contexts, to ones which discount these, replacing understanding with moral judgement of individuals and correction of their behaviour. As former Prime Minister John Major once suggested, we should ‘condemn a little more and understand a little less’ (interview with the Mail on Sunday, 21 February 1993). This of course is emblematic of a shift in worldviews from Left to Right.

The ‘responsibilisation’ strategy was clear in New Labour policy, though there are many historical precedents (Welshman, Reference Welshman2006). It involves a shift away from a rights-based view of welfare benefits to a more disciplinary, contractual view in which benefits are conditional on evidence of ‘responsible’ conduct, particularly being employed – hence ‘workfare’, and a corresponding shift of focus from needs to behaviour. TFP itself has its precedents in family policies pioneered by New Labour. In linking social policy to criminal and anti-social behaviour, it represents a criminalisation of social policy (Burney, Reference Burney2005; Rodger, Reference Rodger2008). This was a risky strategy insofar as many Labour voters still regarded structural features as significant causes of social problems, and supported the notion of a welfare state that offsets these and provides a safety net. This is presumably why, even though New Labour's social policies increasingly attempted to discipline individuals, this was described in policy documents as ‘support’.

Like Brown, only in stronger language, Cameron attributes an alleged decline in responsibility to the welfare state.

But then the families are held responsible for a variety of social problems:

In the TFP, the shift from social explanations to moral condemnation is matched by a heavier emphasis on discipline and punishment:

On the face of it, the term ‘troubled families’ suggests they are beset by external problems, and are troubled in the sense of being anxious about or disturbed by certain things. While this could be construed as compassionate, it is clear from the above that it is nothing of the sort (Levitas, Reference Levitas2012a). Presumably, the government did not want to risk using a more inflammatory word like ‘troublesome’ in official documents, and it could signal its intention in more informal and direct terms elsewhere, while relying on the tabloid press to elaborate the narrative and provide graphic illustrations of ‘benefit cheats’ and the like for popular consumption. Further resonance was provided by a string of television programmes such as ‘Benefits Street’, ‘We pay your benefits’; ‘On benefits and proud’ (Jensen, Reference Jensen2014; Tyler, Reference Tyler2014) that invited audiences to condemn and despise its subjects. Yet, it has to be acknowledged that such programmes – political and televisual – were popular, including among many who themselves benefitted from the support of the welfare state (Shildrick and MacDonald, Reference Shildrick and MacDonald2013). The cheats, scroungers and skivers, were always other people, less deserving and supported by the long-suffering deserving majority.

Explanations in terms of external circumstances such as local economic decline and lack of adequate jobs were conspicuously absent. A report published by Louise Casey, Director General of the TFP, called Listening to Troubled Families, contained not one mention of poverty. The title turned out to be another misnomer, in that the narratives of the selected families themselves were discounted, and peremptorily glossed with Casey's own explanations:

As MacLehose comments, the families were defined by their behaviour, not by their experiences and situations (2014: 42–3).

Yet while the misuse of evidence appealed to in TFP policy discourse to suit the political climate and strategy has been pointed out by experts on social policy (Levitas, Reference Levitas2012a; MacLehose, Reference MacLehose2014), it does not follow from this that none of the pathologies that it claims to identify exists. One can analyse how the allegedly troubled families have been ‘constructed’ by the government and policy makers, but some of the anti-social behaviours – theft, physical and verbal abuse, child neglect – are real and serious; they cannot all be dismissed as tabloid inventions. Consequently, opposing the stigmatisation and symbolic violence of the policy and media discourses by denying any culpability at the level of the individuals is problematic, and offensive to those who are victims of anti-social behaviour. Note too, that the denial of responsibility is typically selective: sociologists do not generally deny or ignore individual responsibility in cases of racism, rape or murder, nor do they treat them as products of (mis)labelling or negative social construction by the dominant. Further, while it is indeed important to stress social causes, analyses that leave no space for any individual responsibility, whether causal or moral, are problematic because they render people as helpless victims, having no agency.

On the other hand, accounts of behavioural problems that allow a qualified role for individual responsibility are at risk of being claimed by the Right as support for their own unqualified, one-sidedly individualistic views, and painted by the Left as conceding the blaming and stigmatisation of the dominant discourse. Even to appear to step partly within the dominant frame is to risk being captured by it, or at least of being accused of it, though the risk may be lower in patient academic discourse than within more eristic, impatient debates. So strategically, within eristic debates, it may often be best simply ‘not to go there’, and talk instead about social causes of problems. This has been common amongst social researchers, and indeed generally I would recommend it on such occasions. But sooner or later one is likely to have to confront the individualistic accounts and acknowledge that some behaviour is seriously anti-social. It could take decades to change the broader social causes, and even then, given that dispositions acquired from early life are hard to change, one might not expect problematic behaviours to disappear quickly.

A strategy of denying problematic behaviours underestimates the extent of injuries of class and other forms of inequality and confuses partisanship with critique. The case against it has been well made by Martha Nussbaum through the example of Joseph Steinbeck's famous book, The Grapes of Wrath. In that story, the impoverished migrants face a succession of disasters, injustices and indignities, and yet retain their dignity and generosity. Their moral qualities are undiminished, even strengthened by their experience, so the story easily wins our sympathy and co-opts our sense of injustice. But it is also a romantic conceit:

Inequalities tend to injure people – that is mainly why they are so problematic (Sennett and Cobb, Reference Sennett and Cobb1972; Sennett, Reference Sennett1998; Sayer, Reference Sayer2005). Poverty, insecurity, exclusion, stigmatisation and a depressed/ing environment make daily life extremely difficult and are not ennobling. Those difficulties do not necessarily lead to the pathologies highlighted in the TFP discourse: the relation is probabilistic rather than deterministic. And as Shildrick et al.’s empirical studies of families living in poor areas shows, they are not found in most such families (Shildrick et al., Reference Shildrick, Macdonald, Furlong, Roden and Crow2012).Footnote ² Nevertheless, poverty and inequality can harm people in many ways, including causing stress and ill-health, making conflict more likely and reducing faith in legitimate ways of making a living (Charlesworth, Reference Charlesworth2000; Wacquant, Reference Wacquant and Bourdieu1999). In turn, anxiety and susceptibility to depression can reduce the ability to make rational decisions. Further, psychological research on altruistic behaviour shows how it is encouraged (discouraged) by good (bad) treatment and conditions that people experience (Appiah, Reference Appiah2008). Research on attachment in early life shows that neglect and abuse of children can make it hard for them to cope well in later life, and can cause behavioural problems that are difficult to change (Cassidy and Shriver, Reference Cassidy and Shriver1999). Note again the qualifiers – ‘are likely to’, ‘can’: there are often over-riding factors that inhibit these pathologies, such as strong support networks of family and friends. There may also be strengths fostered by living with poverty and exclusion, such as resilience, toughness and experience of dealing with heavy responsibilities.

A common response from the Right to claims that behavioural problems are substantially caused by poverty and inequality is that these do not justify anti-social behaviour and criminality. Hence, in eristic discourse, those who point to such conditions can easily be put on the defensive: ‘are you saying poverty justifies violence?’. The challenge confuses explanations of behaviour with justifications. Much behaviour lacks moral justification, but is nevertheless made more or less likely by particular circumstances. An under-regulated financial system does not justify irresponsible actions that risk crashing the economy, but causally it makes them likely.

The intergenerational transmission of advantages and disadvantages is a further difficult issue to address in debates on social policy. The Right either ignores it as an inconvenient fact for narratives of social mobility and meritocracy, or selectively exaggerates it through references to a culture of poverty or alleged genetic inferiority or, more specifically, through claims that there are families that have not worked for three generations. Empirical research shows the latter not to be true, and that the typical pattern for the most disadvantaged families is to move in and out of work (Shildrick et al., Reference Shildrick, Macdonald, Furlong, Roden and Crow2012). In an eristic context, it can be difficult both to acknowledge that children are likely to take on the disadvantages of their parental context, and to counter politically motivated attempts to blame their parents’ approach to child-rearing. As Welshman shows, the history of social policy is littered with cases of folk beliefs about the poor that keep returning – usually in new guises – no matter how many times social research shows them not to be true (Welshman, Reference Welshman2006).

While TFP is clearly an example of ‘responsibilisation’, this concept also needs qualification. As is typical in the Foucauldian tradition, it is ‘crypto-normative’ (Sayer, Reference Sayer2011); that is, the description sounds ominous and suggests something problematic, yet fails to explain why. Are all forms of responsibilisation problematic or only certain forms? After all, in their daily lives, Foucauldians, as much as anyone else, hold people responsible for certain actions. Responsibilisation is problematic where it expects individuals to resolve problems for which they cannot reasonably be held responsible, or correct effects of formative influences which are difficult to change. But it is not easy to distinguish what people can and cannot be expected to do, given their backgrounds and current situations, including existing opportunities and constraints. Just what in each case is reasonable to expect and what is too much is partly a practical question of what is feasible given constraints and available resources, and partly a normative matter of what people should be willing to do and what lengths they should go to in overcoming constraints and limitations.Footnote ³ Punitive observers, such as Casey, attach little importance to such constraints. What can be said is that, other things equal, the greater the incidence of poverty, neglect, inequality, marginalisation and stigma, the less we should be surprised that behavioural problems will arise. However we might choose to judge the behaviours, their incidence is heavily influenced by social circumstances, and available ‘cultural toolkits’ for dealing with them (Swidler, Reference Swidler1986), so for practical as well as academic purposes, we need to understand more and condemn less. Neoliberal discourse systematically ignores these structural forces, but given the massive impact these have on individuals and families, any adequate response to policies like TFP must, above all, highlight them rather than remain with the narrow frame set by the official discourse or indeed that of the academic discipline of social policy.

Value systems and framing in political discourse

It is easy for academics, accustomed to prioritising reasoned analysis, to underestimate the extent to which the effectiveness of political discourse depends not on the adequacy of its arguments and evidence, but on how it plays into major systems of values. To varying degrees, its authors know this. George Lakoff, the US cognitive linguist, argues that elections are decided primarily on values rather than policies, and that if progressives are to win against the Right, they need to realise this and stop arguing about policies within value frames set by their opponents, and argue instead from and for their own values (Lakoff, Reference Lakoff2004). The Right, by contrast, seem to be aware that communicating general values favourable to their policies is more effective than explaining the policies themselves.

Lakoff (Reference Lakoff2004) argues that the Republicans mobilise a frame of values involving a ‘strict father model’. This involves obedience to authority, self-discipline and self-reliance, patriarchal families and heteronormativity, respect for the wealthy as deserving and contempt for the dependent as undeserving, as in the just world ideal. Outsiders – anyone who is different – are regarded with suspicion and fear: conformity connotes morality. Empathy and compassion are associated with gullibility and weakness. In this model, conviction trumps reason: intellect is seen as redundant and dangerous, and those who need to appeal to evidence, arguments and experts are regarded as weak, ‘flip-flop’ people who do not know what they think.Footnote ⁴ Explanations are already to hand, commonsense is authoritative and what is needed is firm action; the strong father always already knows what is right and will protect all those who put their trust in him. The more populist political discourse on TFP exemplifies this value framing. Insofar as official documents refer to empirical evidence, they do so in ways that support the policy and its value frame.

Lakoff (Reference Lakoff2004) recommended that in political debate, rather than respond within this frame, ‘progressives’ (American left liberals) should use their own, ‘nurturant parent model’ as an alternative; he claimed this would appeal to voters if characterised effectively. It has a more compassionate and generous view of others, is open to difference, values dignity and respect for all, and values reason and evidence over simple authority per se. Lakoff argues the two value frames involve different neural systems, and that while most people may operate with a mix of each, the balance can be influenced by discourse and actions which reinforce one rather than the other.

He also argues that conservatives tend to think in terms of simple, direct causation; if a child is neglected or abused by an alcoholic parent, then we need look no further for an explanation of this outcome and can proceed immediately to the solution: discipline to change behaviour. Progressives, by contrast, appeal to more inclusive, complex, systemic explanations, similar to those of sociology, which imply a wide range of measures to resolve the problem. While this means that in eristic discourse they have a more difficult task in having to make complex, qualified arguments persuasive, Lakoff (Reference Lakoff2004) argues that by mobilising vocabularies that chime with the nurturant parent system of values, there are ways of succeeding in such contexts.