Others claim society is now so different we need to start afresh. History might repeat itself, they say, but society does not. The truth is almost certainly somewhere in between. The common ground is that something is missing: the social “glue” that helps us all to relax.

But what is this glue? Some claim it to be morality. Others are now crawling out of the woodwork to herald military-style discipline. Even Tony Blair talks about duty and morality as part of his vision for new, young Britain. But visions are not enough. Many travesties have been perpetrated in the name of religion and morality. To get close to an effective and sustainable solution, we need to go deeper.

All human beings need two basic things. They need to feel safe and secure; and they need to have scope for self-determination, to feel valued for who they are and be in charge of their own destiny. These needs, however, conflict; they sit at opposite ends of the spectrum. A society, or institution, that offers great security and predictability, in which life is highly regulated and routine, can stifle opportunities for self-expression. At the other end, a society that allows self-determination and individualism to reign supreme can threaten security because the consequent loss of mutuality, responsibility and predictable behaviour means people never know where they stand. Where the sky’s the limit, people will rocket off, perhaps precisely to find the final frontier.

Looking at this security-autonomy dimension another way, as the American psychologist Maslow might have done, where people have predictability, they feel secure enough (or stifled enough) to seek “self-actualisation”. By the same token, when they feel out of control, the primacy of the security need re-establishes itself.

It is highly relevant that child-centred theories of parenting and teaching were spawned in the Sixties, after a period of unprecedented social stability when mothers were sent back from the factories to the kitchens. And this stability was craved because the trauma of war had ripped people’s lives apart. It is, perhaps, a social policy equivalent of the economic cycles of boom and bust. There is a trade-off between the two goals.

We need to redress the balance, not to re-establish morality or military-style discipline but to create greater predictability. Morality, as such, is not the issue. To establish a morality that either smothers autonomy or ignores children’s and adults’ need for continuity, trust and attachment would be counter-productive.

Our sense of security does not only come from institutions and routines. It also derives from our key relationships, our sense of where we have come from, where we’re going and where we belong. Some people now, particularly children, have little sense of their roots, and many others have no idea where they are going. Partners, jobs, child-care arrangements, incomes and life-styles: who knows what’s round the corner?

Teachers are acutely aware that the more change we experience, the worse we get at managing it; and it’s far worse for children because they don’t have the benefit of experience to help them to anticipate what might happen next, or to believe that things will ever get better.

Children are parties to change. They experience it, feel it, and their behaviour is increasingly showing they are extremely troubled by it. It is no surprise that child psychologists are revisiting the work of Bowlby, the father of attachment theory. In times of change and stress, everyone’s need for security and their sense of belonging gets stronger.

The policy implications that follow from this don’t fall cleanly into one political - or boot - camp or the other. Some re-thinking is in order. “Change” policies can be devised for each area of government. Take working parents, for example. It does not follow that they should all give up their jobs to embrace child care, but we do need to question how children acquire that all-important experience of secure attachment. High-flying parents rarely at home are unlikely to deliver. Streams of au pairs or care assistants won’t help, either. Grandparents, who “belong” to the child can be a good substitute as can long-term nannies. But there is no single model suitable for all children. Arrangements that enable particular children to thrive will depend, in each case, on their surrounding circumstances. The more uncertain and unpredictable a child’s wider environment, the greater will be their need for security, derived from both a predictable and reliable, mutually fulfilling relationship - and, incidentally, clear routines, expectations of, and boundaries for, behaviour.

Discipline is important at home and at school. Perhaps because of change, “permissive” child-rearing has not worked. But we don’t just want any discipline. If it is based on dictatorship or deference, punishment or humiliation, children already simmering with anger and resentment could explode. Instead, it must rest on mutual respect, balanced freedoms, managed choices, responsibility, consistency and, yes, authority.

Children need more discipline now not because they are “bad” and need to be punished, and not just because some are running riot, but because they need the structure and security that it provides, the positive feedback to be gained from compliance and the perceived care and attention that flows from monitoring and supervision.

What about divorce, single parents, self-help and responsibility? While there is a strong argument for refreshing people’s sense of responsibility, parents included, levering single mothers back to work may not serve the best interests of their pre-school children if their lives are already fragmented. And on divorce, it is futile to look for evidence that demonstrates either that children are always destroyed by divorce or that they are unharmed by it. The outcomes will be influenced, crucially, not only by the degree of surrounding conflict (which researchers acknowledge, but also by the extent of upheaval and change children experience. If they lose touch with a parent, separate from siblings, move to a new home and leave familiar schools, the difficulties will be multiplied.

And primary schooling? Dogma is less relevant than re-assessing the security-autonomy balance: paying more attention to attachment, smoothing change and putting trust back into relationships. Perhaps children could better manage a more open-ended style of group learning when they had more secure backgrounds. Now the personal lives of many children are fragmented and chaotic, they need a more structured, less risky school experience: to learn “in safety”, to paraphrase Carl Rogers. Many primary-aged children prefer set timetables to a seamless curriculum, because they know where they stand.

Tutors, or “mentors”, assigned to each child for the duration of their school life would offer greater continuity, engender trust and rebuild personal significance.Similarly, more attention could be paid to the end-of-year transfer, with children going into their old classes with their old teacher first before being handed over to the new. “Study buddies” and “care pairs” teach children commitment and responsibility.

Managing change is the great challenge facing the Government. For this, Britain needs a flexible workforce, not an army of automatons. If newly sculpted social policies ignore the need for continuity and trust, they’re likely to misfire. Alternatively, if new “moral certainties” extinguish opportunities for self-determination, we will almost certainly see history repeating itself as the suffocated children of Britain’s new millennium tear themselves free, and go for bust.

Elizabeth Hartley-Brewer is an educational and parenting consultant