by Paul Collier
Similarly, mutual obligations eroded beyond the nuclear family. Extended families shrivelled under the pressures of smaller family size and the geographic mobility of the skilled. Again, I exemplify the change at its most extreme. I grew up with twelve aunts and uncles within five miles of my home; my children are growing up with none. The extended ethical family gave way to the nuclear dynastic family.
As the highly educated became a class, they developed a new form of family in which some of the reciprocal obligations were restored and even reinforced. We see this pattern in the data. Extramarital births in this class were rare in 1965: only 5 per cent; they are still only 5 per cent today.4 After its initial surge, divorce declined; by 2010 its incidence was down to one-in-six marriages. With few extramarital births, and few divorces, the number of young children reared in single-parent well-educated families has also reverted to very low levels: they are now back to less than one-in-ten.
The new ethics of self-fulfilment through personal achievement came with some downsides, but these were as of nothing compared to the consequences of the shocks that hit the class of the less educated.
SHOCKS AT THE BOTTOM
Just as the Silicon Valley technocrats predicted that the new social connectivity would reduce hatreds, so the Pill and abortion were predicted to reduce unwanted children. We see the resulting surge in the sexual activity of the less-educated half of teenage girls in the data. In the 1960s only 5 per cent had intercourse before they were sixteen; by 2000 it had reached 23 per cent. In contrast, even by 2000, only 11 per cent of the girls who went on to become graduates had had underage sex.5
But the Pill only stopped conception if combined with prudent foresight, and this favoured the educated. Aborting a foetus turned out to be a decision that, while comfortable within the new ethical belief system of personal fulfilment, was fraught within the old one of family obligations, and this again favoured the educated. The result was an explosion in teenage pregnancies among the less educated due to couplings never intended to be enduring. Such a teenage mother had four potential options. One was the old model of marriage to the father – the shotgun wedding has a long tradition. Another old model was that she and her baby would continue to live with her parents; my great-grandmother had relied on that without dire consequence within her village. A third option was to ape the new model of individual fulfilment of some educated women and branch out as a single mother, and for this the paternalistic state offered financial support and social housing. A final option was to pioneer a new model of cohabitation: fathers were often less wary of cohabitation than of public commitment. Of course, a relationship can be stable without people being married, but most cohabitations do not lead to lasting relationships; the average one lasts only fourteen months.6
The final shock at the bottom was economic. With the decline in manufacturing, middle-aged men lost their jobs. Many less-educated households had never bought into the new ethics of self-fulfilment, and many couples had held on to the norms of the ethical family in which the husband was the head, his authority underpinned by his role as the breadwinner. This role had an awful implication: redundancy at work implied redundancy at home. Such a marriage went from being a tight network of mutual esteem to an asymmetric one; the wife kept her esteem but her presence amplified the husband’s loss of esteem. Sometimes the husband sought to reassert authority through violence, sometimes he sank into depression. These were the wellsprings of divorce.7
Again, we see this in the data. As among the educated, there was an initial surge in divorce. But in contrast to the educated, among the less educated divorce kept on rising. By 2010 its incidence had reached one-in-three marriages, double that of the educated.
In place of the obligations towards children provided by the ethical family, the paternalist state stepped in with the ‘rights of the child’. These new rights did not extend to the right to be reared from birth to adulthood by the two parents from whom the child was genetically descended. On the contrary, the ‘rights of the child’ obliged the state to remove children from birth parents if there were grounds for thinking that the child was being abused. In response to highly publicized cases in which children had died at the hands of their parents, the obligation was progressively tightened. For example, in the USA if a doctor saw that a child had an injury, unless they could satisfy themselves beyond reasonable doubt that it had not been caused by the parents, they were obliged to report it to the authorities, who were in turn obliged to remove the child from the parents. But, correspondingly, the ‘rights of the child’ required the highest standards to be met before these removed children could be adopted into another family, and a correspondingly exhaustive bureaucratic checking process to ensure that any placement decision by the authorities would be above public criticism. The inevitable consequence of a high rate of removal from birth parents and a low rate of placement with new ones was that increasing numbers of children found themselves in limbo: in Britain there are now 70,000 of them. In practical terms, limbo meant that the state paid couples to foster children on a temporary basis, often circulating them from one fostering couple to another. Quite evidently, fostering fails on all the important metrics of child rearing: the relationship is quasi-commercial whereas children need manifest love; it is explicitly temporary whereas children need permanence; and it cannot evoke a sense of belonging.
CONSEQUENCES OF SOCIAL DIVERGENCE
The consequences of this selective breakdown in family obligations were most profound for children. In the USA, where these effects are most pronounced, and which may become the cultural future of Europe, over half of all children are now likely to spend time in a single-parent family before reaching eighteen.8 As implied by the preceding analysis, this is highly class-selective. Among the educated class, the upper-half of American households, family obligations to children have largely been restored and enhanced. In contrast, among the half with less education, single-parent – or no-parent – children have become the norm, accounting for two-thirds of all children in this group.
Does this matter? Unfortunately, it does. Despite the powerful and understandable taboo on stigmatizing one-parent families, social science has now demonstrated rigorously, and causally, that children do better if they are reared from birth to adulthood by two parents from whom they are genetically descended.9 For many children, even one-parent families have ceased to be an option. The responsibility for child-rearing has increasingly shifted from parents to the state. Yet social paternalism has a poor record of success. This is unsurprising; state provision, whether in children’s homes or foster care, suffers from the drawbacks implied by ‘what money can’t buy’, as described in other contexts by Michael Sandel. Paying people to care for children can supplement parental care, but not substitute for parents.
Whereas in the less-educated half of the population many families are disintegrating into empty shells, among the more-educated half we are seeing a proliferation of dynasties. The new hothousing model adopted by educated households has dramatically increased parental input. As never before, the children of the educated are exposed to intensive and purposive interaction with their educated parents.
Cumulatively, hothousing makes a difference. It starts early. Indeed, the child’s pre-school experience is now recognized to be decisive: by the age of six the differences in performance that appear after a decade of schooling can already be predicted. In short, what the family does in the few years before school is more important than what schools do in the twelve during which they are responsible.
The differences start with objectives and are then implemented through techniques. Parents who are single and poor are far more likely to be stressed – their priority is not hothousing but the more mundane one of containing chaos. Among those parents who dropped out of school, obedience is rated above self-reliance by almost four-to-one; among those with postgraduate education, this is reversed. Such stress-induced parental behaviour has been found to impair children’s non-co
gnitive development, which we now know to be at least as important as cognitive skills.10 But cognitive skills also start to diverge early on. The earliest measured divergence is in language: hothousing involves talking to young children. A celebrated study found a class difference of 13 million words by kindergarten. The words themselves differ: the children of professionals hear eight times more encouraging words than discouraging ones; the children of those on welfare hear only half as many encouraging words compared with discouraging ones. Then comes reading. Parental reading fosters child development and is the biggest single factor explaining differences in readiness for school. And then, of course, there’s money. The shift to hothousing has massively increased spending. But since the 1980s, whereas such spending by an American household in the top tenth of incomes has doubled to $6,600, among those in the bottom tenth it has fallen to $750; the biggest divergence has been for the decisive period of pre-school.
The same pattern of wide and widening divergence continues during the school years. In the USA, by 2001 the gap between the income classes for maths and reading levels was around a third larger than a generation earlier. Not only is the same pattern continuing, but it is being driven by the same process: the underlying differences between families.
The most dramatic consequence of this divergence between the educated and uneducated classes is a recent discovery about American children made by Robert Putnam, whose work has been seminal. Grouping children according to their cognitive abilities, he analysed their chances of getting into college. Of course, we would expect the children of the educated class to stand a better chance of getting into college since they are likely to inherit higher cognitive abilities. But Putnam found that those children of the educated class who are in the bottom national group of cognitive ability have a higher chance of getting into college than those children from less-educated families who are in the top group. The new hothousing rears not only trophy children, but camouflaged clots.
The trends towards rising social inequality and stagnant or falling social mobility are recent, and the numbers essentially track the change from my own generation to the next. But the most worrying news is that these observed changes are liable to understate, by a considerable margin, the true persistence of social inequality. In a remarkable recent book, wittily entitled The Son Also Rises, Gregory Clark studied the transmission of family inequalities over many generations.11 Usually, social mobility is measured only by comparing one generation with the next, but he hit on the clever technique of using rare surnames, which can be more easily traced over the centuries. Evidently, what he was usually tracing here was the male line, which for most of history implies that he was tracing the role of the head of the household. What he found is that success is highly persistent, often over centuries. Clark shows that the conventional estimates of social mobility, based only on the transmission from one generation to the next, are radically inconsistent with such enduring inequality, and he provides a plausible explanation for the bias. Some asset is being passed down the generations without being dissipated. What can it be? It is unlikely that financial wealth can cascade in such a way: it only takes one rogue to dissipate a fortune and the cliché of rags-to-rags in three generations rests on this insight. He comes down to two assets that cannot be dissipated. One is genetic, but even though genetic inheritance is important, over several generations exceptionally useful genes are liable to be diluted by mating. The other possibility is what Clark refers to as family culture. This is shorthand for the norms and narratives of the belief system that shapes behaviour in the networked group of the family. Being at its hub, the head of the household is well placed to induce continuity. We know that elite parents put considerable effort into the transmission of their culture,12 and perhaps particularly to those attributes conducive to success, even though the specific attributes will change over time.
The same technique of tracing rare surnames can be used to measure the other end of the social spectrum, families that get stuck around the bottom of society from one generation to the next. Clark found the same pattern of persistence over many generations: failure is being transmitted down the generations. Since indebtedness cannot be inherited, cascading lack of financial wealth is an implausible explanation. Indeed, for most of history most people have had no such wealth so most people have had the same monetary inheritance – nothing.
Clark explains why the conventional measures of social mobility based on adjacent generations are likely to exaggerate it. Simplifying to bring the point out, suppose that success in each generation was due only to family culture and luck. Each generation inherits a family culture and draws a ticket out of the hat called ‘the wheel of fortune’. If family cultures pass down the generations intact, the only source of social mobility is luck. But the change in luck between the first generation and any of the subsequent generations is the same whether we take the adjacent generation or one that is distant. In this deliberately exaggerated example, the social mobility we observe between the first generation and the second would be the same as that between the first generation and the twelfth. Measuring only the former might give the illusion of a mobile society.
RESTORING THE ETHICAL FAMILY?
Some aspects of the ethical family were a stultifying veneer for relationships of power and abuse. We are well shot of them. But other aspects of the ‘liberation’ from it were little more than selfishness masquerading as self-discovery. Similarly, the juxtaposition of Utilitarian concern about ‘the poor of the world’ and denial of responsibility for family was less of a moral awakening than the easy pleasure of moral posturing: Dickens skewered such attitudes through the character of Mrs Jellyby, in Bleak House.
More fundamentally, the triumph of individual fulfilment through personal achievement over meeting obligations to family is beginning to look psychologically flawed. In a profoundly subversive book, The Road to Character, David Brooks starts from the familiar celebration of fulfilment through achievement only to turn the tables on it, suggesting that the future trend will be towards a restoration of fulfilment through meeting obligations to others.13 The seductive proposition that we find ourselves through focusing on ourselves is opposed by a powerful counter-narrative, which is perhaps best expressed by Dietrich Bonhoeffer in Letters and Papers from Prison, his testimony while awaiting death at the hands of the Nazis: we find ourselves through ‘losing ourselves’ in the struggles of the other people in our daily lives. Freedom is not found in servitude to the self, but in escape from the self. Bonhoeffer and Brooks have the new evidence of social psychology on their side. Our regrets about insufficient personal achievement are dwarfed by our regrets of obligations that we failed to meet. The distinguished psychologist Martin Seligman has conducted a sustained programme of research on the attainment of well-being. His conclusion is unambiguous: ‘If you want well-being, you will not get it if you only care about accomplishment . . . Close personal relationships are not everything in life, but they are central.’14 The replacement of the ethical family by the entitled individual is revealed to be more tragedy than triumph.
Seemingly a world apart from this, a major breakthrough in economics showed that ‘weaker’ could be ‘stronger’. In order to benefit from making credible commitments, it may be necessary for a person to shed some power. Being able to make commitments was an instance of enlightened self-interest. Fancily expressed, a ‘commitment technology’ solved the ‘time-inconsistency problem’: the discoverers received the Nobel Prize. The commitment technology to solve inflation was to give central banks independence; that which solved child-rearing was marriage. Paradoxically, over the same period that Western societies were establishing the commitment technology that tamed inflation, they were systematically tearing up the commitment technology that had defended the right of children to be brought up by the people who conceived them. Just as politicized central banks create an initial sugar-rush of printing money, so tearing up the bonds of marriage created the sugar-rush of liberation.
In many Western societies, marriage is tainted by its religious associations, and so we need a purely secular equivalent. This is not revolutionary: in all Western societies marriage preceded Christianity and religious and secular forms of public commitment can readily coexist. In each case the commitment technology draws its strength from the public and explicit acceptance of mutual obligations: esteem and shame are the force on which it draws. If you recall, a commitment technology is in the self-interest of those who use it. It is ‘enlightened’ self-interest in the same sense as previous instances – it infuses compliance with purpose. Once you understand the true causal chain that leads to desired consequences, mutual compliance becomes rational. Just as enlightened self-interest complements and reinforces other reciprocal obligations, so the economic insight of the value of public commitment complements the psychological insight on the value of meeting those obligations.
Between them these insights can powerfully counter the somewhat jaded aspirations of fulfilment through achievement. But this does not address the new reality of the shrunken domain of the family, the transformation from the extended ethical family to the nuclear dynastic family. How might this be countered? Fortunately, there is one magnificent consequence of technological progress that can offset this process: increased longevity.* While families have shrunk horizontally, they have grown vertically, and many families now span four generations instead of three. The most senior generation in such a family commands an extended span. If each generation has two children, any surviving senior will encompass four nuclear families and twenty people spread across the three younger generations. Such patriarchs and matriarchs need not retreat into fossilized purposelessness: give them a role, that of regenerating the force of esteem that polices the obligations of the extended ethical family.
