by Paul Collier
Running a café was judged by the mental health team’s managerial hierarchy in the NHS as too peripheral to justify the continued claim on the budget: the core activity of the team was treatment. Hospital stays were reduced, but that was a different budget. As people got jobs, they came off benefits, but that was the Social Security budget. As to Social Services, why should they divert money from core activities to fund something that the NHS wanted to discontinue? Better parenting helped the children in school, but the priority for the Education budget was its core activity of teaching. Hierarchies removed from the coalface, managing fragmented specialisms, killed an initiative that addressed the core of the problem, instead of learning from it and scaling it up. For each, the priority was the symptom that it was treating. As the despairing psychotherapist remarked: ‘without better interventions this will be perpetuated down the generations with only a relatively few individuals escaping the cycle.’
That is where social maternalism begins; it continues. Young parents struggling with an unplanned child face pressures for which they are unprepared. Most parents feel the duty of care instinctively most of the time, but parenting young children can be immensely stressful: moments inevitably occur in which couples get angry with their children, and angry with each other. It requires skill, self-discipline and forgiveness to prevent such moments escalating into enduring damage. Teenagers scarcely out of childhood are plunged into a situation in which they need to sacrifice their own desires, control their tempers, and plan ahead. Young parents need money, relief and non-judgemental mentoring. This is the core of social maternalism: how can these things be provided?
Households fit their lifestyle to their income: with a little planning and prudence a large majority are able to meet the basic needs of their children. Paternalist largesse can be a two-edged sword. Britain provides free housing for single mothers; Italy and Spain don’t. Britain has one of the highest rates of teenage pregnancy in Europe; Italy and Spain among the lowest. In 1999 Britain introduced increased benefits for those low-income families with children. Modern statistical methods enable us to tease out the consequences of this policy change: low-income families responded with a massive increase in births, estimated at an extra 45,000 children each year.7 So, as a result of this free housing and enhanced benefits, many children are being raised in households that have a bit more money. But many women were encouraged to bear children who will not be raised well. These are hugely expensive benefit programmes with ambiguous effects, whereas other uses of public money are unambiguously beneficial, yet under-provided. Here is an example.
Young couples have not had the time to build up a cushion of savings, so they are vulnerable if hit by an adverse shock. Cushioning such shocks is therefore a valuable use of public money. The most evident shock is unemployment. In the USA, the financial crisis of 2008 caused a large and prolonged surge in the unemployment rate. New research by one of my doctoral students shows convincingly that this increased the neglect of young children.8 The effect was large and it was causal. For each 1 per cent increase in the unemployment rate in a county, the incidence of child neglect rose by 20 per cent, affecting young children the most. But public policy can help mitigate the damage caused by unemployment. Counties vary in their rules for the duration of unemployment benefits; in those counties where benefits lasted longer, the impact of unemployment on neglect was substantially reduced.
So much for the money to raise children; now for relief in managing tasks that, done properly, are hugely demanding. Relief starts with the extended family: there is an obligation on other family members to rally round, but the extended family has shrunk. My father was one of seven siblings, my mother one of four, and so there was an army of aunts and uncles to support them in rearing me. Now, parents have fewer siblings, so the obligations of those that remain have correspondingly increased. But parents such as myself are only-children, and in such situations the extended family needs to be revived. Norms need to change; offsetting the horizontal shrinkage, greater longevity is expanding the family vertically. In response to the new need, people are indeed changing their norms appropriately: grandparents are far more engaged with their grandchildren than in the past.
Governments can do much more as well. Most governments have the sense to provide financial support for parents with young children, but increasingly this has been merged with the objective of encouraging people to get a job. In stressed young families, the period while the adults are parenting young children is not the right time to do this. Those people who never have children receive a huge benefit from those who do: the retired are only able to live off their savings because the subsequent generation is putting those savings to work. While parents are struggling to rear young children is the key time for the state to make the transfer payments that reflect this contribution to society.
But the state can do more than give money: it can provide in-kind support both within the household and beyond it. Parenting is difficult for every new parent, but some couples are in such inauspicious circumstances that trouble is all too likely. Where trouble can be predicted, it can also be averted by intensive pre-emptive intervention.
Just as there is a limit to what the market can do, there is a limit to what the state can do through public services of support. However, we are not yet at that limit. There are a few examples of intensive publicly provided support, and to the extent that they have been evaluated they show signs of success. One example is the Dundee Project, a modest experiment in unconditional support for stressed families. Practical, day-to-day support for a young family is expensive, but far cheaper than the consequences of family breakdown.
A vital feature of the Dundee Project was that it was completely separated from the service that scrutinized the family. Scrutiny is necessary: in extremis, a child should be removed from its parents. But without such absolute separation of functions, the basic conditions for building a trusting relationship between the parents and the workers providing support are not met. In Britain, the Dundee Project inspired a massive scale-up into the Troubled Families Programme (TFP), but while well motivated, this was contaminated both by the additional objective of getting young mothers into work and by being run by the existing social services, with their role of scrutiny. This overload blunted the TFP’s efficacy.
Although integrating support with scrutiny undermines each service, integrating physical support with mental support could be reinforcing. Often, the parents in those families that are predictably liable to become troubled have incipient mental health problems. Mental health interventions, such as cognitive behavioural therapy and anger management programmes, have been rigorously evaluated and show impressive success rates. Such pre-emptive support costs money, but it may avert behaviours that are much costlier to society in the long run. While the provision of child support, mental health care and scrutiny should be joined up, their functions need to be kept sharply separate.
Teenage couples expecting a baby are parents with L-plates and need non-threatening guidance. The occasional evening class is unlikely to be enough. Grandparents can help, but quite often the couples most liable to become dysfunctional parents are from families that are themselves dysfunctional. Young couples need some source of mentoring and informal support beyond the family. One way to supplement dwindling or dysfunctional extended families is to create a new resource: a modern analogue for our own societies of the Peace Corps or Voluntary Service Overseas that once inspired many thousands of American and British youth. Then, the new social resource was a growing pool of educated young people looking for a sense of purpose beyond their own enrichment. Today’s equivalent is the growing pool of fit and savvy retired, financially comfortable with their pensions but with a hole in their lives left by an empty nest. These people have been equipped by life with the non-cognitive skills with which to become the unthreatening helpers for the stressed young couples who are in need of support. Rising to a duty of rescue can bring a deeply satisfying sense of purpo
se into life at a stage when otherwise it can become wistful or complacent. As with all support, the role would have to be sharply delineated, and participants trained, to ensure that it could not deteriorate into a patronize–blame–scrutinize–report relationship. Maybe it should be paid; if so, payment should depend on the authorization of the young parents, so that they would feel empowered. Maybe young parents could be given a budget that they could draw upon for such a purpose. Rather than being government-organized, a new crop of NGOs could recruit the capable-with-time to help the thousands of young families that are not coping with their obligations. Whereas governments are terrified of failure, and so ill-equipped to experiment, NGOs are ideally equipped to try new approaches.
There is good reason for the phrase ‘the terrible twos’: young children are periodically impossible, stressing even experienced parents to the limits of endurance. From that age, children benefit from being socialized into groups beyond the family – kindergartens. There is a strong case for these to be provided by the state and open to all without charge. All states provide school-age education, and the case for state provision of kindergartens is stronger than for any other level of education. In general, as children get older their educational needs become more complex and differentiated. The main advantage of states over other forms of provision is in those activities that lend themselves to standardization and are more cheaply done at scale. Kindergartens are not complex: the key feature that society should want them to offer is to provide a standardized forum in which young children meet others drawn from a wide cross-section of society. Standardization and free provision has the vital advantage that by making the parental decision to send the child to the kindergarten normal for the entire society, those parents who are least equipped to take good decisions will be more likely to do so. The universal provision of free public kindergartens thus achieves two highly desirable outcomes: they are socially mixed at a time when children are at their most readily shaped by social influence, and the children most in need of pre-schooling are likely to attend them. Instead of public kindergartens, however, many countries have a complex plethora of subsidy schemes for private provision that have accumulated incrementally with each new ministerial initiative to meet an evident need. For example, the British Sure Start programme prioritized getting mothers into work, and was readily gamed by targeting recruitment on the easiest ‘successes’ who just ticked the criteria: complexity almost guarantees that schemes will tend to be used by those who least need them, and private provision guarantees differentiation of intake. The exemplar of free, state public provision of kindergartens is France, with its écoles maternelles. We experienced it first hand while living in a low-income Breton town; in neither Washington nor Oxford were we able to find its market-provided equal.
Schools as sites of support
Recall that the most important activity that happens in a school is not the teaching but the interactions within the peer group; the differences that begin in the family are replicated and amplified by the differences in the social composition of schools. Silicon Valley think that their technology has opened the world of knowledge to the children of the less educated. But the evidence is quite contrary to their hopes: the internet has widened rather than narrowed differences in opportunities. Everyone now has access, but recent research shows that the children of the educated learn to use the internet for expanding their knowledge, whereas the children of the less educated use it for distraction.9
The most valuable change that could happen to schools would be to make them more socially mixed. The key impediment to social mixing is school catchment areas. Because where people choose to live has become so socially stratified, catchment areas have the effect of mirroring this stratification in schools. One way of breaking free of this trap for post-primary schooling is to create publicly funded schools that have city-wide catchment areas, differentiated by purpose rather than location. One school might promote itself as the best place for aspiring sports professionals; another for aspiring actors; another for the children of parents who value discipline. Drawing on the concepts introduced in Chapter 2, the idea here is that heads and school governors would be trying to build schools with somewhat distinctive belief systems: they become networked groups in which distinctive narratives circulate. The schools would know they had to be good at what they did, otherwise parents living in affluent catchment areas would continue to send their children to the local, affluent-only school in preference. New rules have now enabled such schools to be established in Britain, and I was part of a team that tried to start one in Oxford, a city whose catchment areas are grotesquely skewed. Our plan for city-wide, lottery-based access met with a predictable response: a wall of vested interests and ideology. The outraged local educational elite, led by the school in the most affluent catchment area, rose up in fury. They succeeded in blocking us; perhaps you might have better luck.
Schools as organizations
The teaching activity of schools could be improved. This is a heavily researched topic with a vast literature, but the dominant motif is that teacher quality is far more important than money. Four simple things can raise teacher quality: attract a better intake; base training on the pragmatism of evaluated experiments; assign the best teachers to the most difficult settings; and weed out the weakest teachers.
In Britain, the Teach First programme has had a dramatic impact. Its aim is simple: to induce good students graduating from universities to spend their first few years teaching before switching to another career. The approach has potential for analogous targeted recruitment: how about Teach Last? Upon retiring from his chair in Amsterdam, Professor Jan Willem Gunning, my co-author for many an article, became a maths teacher in a local school. He tells me it has been the most rewarding experience of his life. But the Teach First programme was restricted to teaching in London, the area of the country that least needed it. The schools that need Teach First are in the provincial towns and cities, where good teachers are often wary of taking jobs lest they become marooned. Precisely because those who plan to teach for life are deterred by fear of getting stuck, those who don’t plan to remain a teacher should be the easiest to recruit. The London-bias in Teach First is compounded by a wage premium currently paid to teachers in London, where schools receive far more funding per pupil than those elsewhere. London has the best school results in the country. Teach First, the wage premium and the funding premium per pupil should all be closed in London and be switched to the places that need them. Teach First was precisely the right programme, targeted at precisely the wrong place.
Deciding between teaching methods is well-suited to learning from randomized trials. But politicians and the educational establishment are wary of such experimentation. Pragmatism is an admission of ignorance, and the confidence that comes with an ideology is much more satisfying. However, the wide variations in the PISA scores between countries and schools suggest that there is much still to learn, and that will only come from evaluated experiments. Teacher training should be built around this evolving evidence, and students should be taught how to keep learning from it.
Weeding out the weakest teachers can have a dramatic effect.10 While it takes some very fancy social science technology to establish that the worst teachers are doing massive damage, it does not require much research to understand why nothing is done about it. The vested interests of the teaching profession, represented by various unions, threaten annihilation for any politician who dares to suggest it. Understandable? Yes. Ethical? No.
There are a few classroom policies that appear to help address attainment problems, although fashions change and ideology again impedes analysis. In addition to teaching, student effort is crucial: the question is how best to induce it among those least inclined to try. Economists at the University of Chicago are using lab experiments to test different approaches,11 and have found that quite simple techniques can have substantial effects. One is that, in order to be effective, any rewards need to be delivered
almost immediately after the effort – within minutes, not months. As to the sorts of reward, esteem works better than money (once again, we are revealed to be more social animals than greedy ones). But rewards turn out not to be the best motivator. People are much more motivated to avoid losses than to acquire gains – the technical term is ‘loss aversion’ – so swift, esteem-related losses for low effort may pack the biggest punch. Yet this is not a message prominent in teacher training colleges.
