Harry Benson is Research Director for the Marriage Foundation and co-author of What Mums Want (and Dads Need to Know).

A friend of ours is an amazing woman who has brought up her now teenage children on her own, having split from the father soon after their youngest was born.

Her children are polite, motivated, sociable, intelligent and a credit to her. With limited resources, and not much encouragement from her own family, she has struggled through. She has depended almost entirely on housing and other benefits, surviving from hand to mouth just above the poverty line. Escaping the dual poverty trap – where every pound earned meant the loss of most of a pound of tax credits and benefits – was a deliberate choice. But for her own sanity and well-being, she persevered despite earning little more. Now in a secure full-time job, she is largely free of the welfare system.

Hers is a great success story because she leaned on the state for the time she needed it before becoming independent and self-reliant on the other side.

But could a more stability-focused family policy have improved her odds of avoiding the split in the first place?

If ministers are ever asked to explain their family policy, they might talk about general services – such as health, education, welfare, childcare, mental health and Surestart Children’s Centres. Or they might refer to more specific services that deal with the consequences of family breakdown and dysfunction – such as the criminal justice and care systems, and Troubled Families programme. These are all important aspects of family policy.

But the question that is rarely, if ever, asked is how and whether family policy can limit or reduce the scale of family breakdown. This is a massive social justice issue simply because of the consequences for children.

For much of the past two decades, Britain has been at or near the top of the Western European league table for family breakdown. (Having been top in 2012, we are now fifth due to falling divorce rates, though not government policy.) According to the Relationships Foundation, the taxpayer spends £51 billion per year in picking up the pieces.

Nobody disputes that some relationships are best ended. But when the majority of break-ups occur ‘out of the blue’, with no obvious evidence of serious conflict or unhappiness, it makes sense to ask why we do so badly and how policy could reduce its prevalence.

First, a genuine family policy needs to be rooted in robust evidence.

There is already a huge body of research that has identified factors associated with a higher risk of breakdown and explored the consequences of breakdown. Some of these factors are general – mother’s education, age, ethnicity. But others offer the potential for specific policy initiatives – marital status, parental happiness with the relationship, parental well-being.

However, the UK research base into family stability is almost non-existent. The vast majority of what we know about relationships and their outcomes comes from US research and journals.

Cambridge University has a world class Centre for Family Research, though their focus is ‘new family forms’ who probably represent around two per cent of families. With rare exceptions (e.g here, here, and here), what UK research there is for the other 98 per cent tends to come from a handful of think tanks such as the Marriage Foundation, Centre for Social Justice and CARE.

A Number 10 Family Policy Unit should encourage the development of UK research into mainstream family stability, instability, and its consequences.

Second, government policy itself can and does influence the decisions couples make.

Most couples start off wanting and seeking reliable love. The odds of achieving it improve massively if they make a clear formal commitment together. And staying together improves the likely outcomes for their children. The absence of a father from the home, for example, is one of the biggest predictors of teenage mental health problems.

This finding in no way undermines the heroic efforts of lone parents like my friend. But it would be odd if the cut in resources from two pairs to one pair of hands had no effect whatsoever.

Anything the state can do to improve the way couples commit will help maintain their resources.

The government was right, therefore, to extend civil partnerships to heterosexual couples because, like marriage, it requires clarity of decision and commitment. But it would be wrong to extend marriage-like rights to unmarried cohabiting couples because it negates the need for a clear mutual decision.

And it is bordering on the obscene that couples are paid thousands more in benefits and tax credits to live apart – or pretend to – rather than to formalise their relationship and live together. A marriage allowance of a couple of hundred quid can’t compete with the scale of this ‘couple penalty’ bribe, although it might if it were better targeted. No wonder half of all family breakdown takes place among unmarried parents either during pregnancy or in the first two years of their child’s life.

A Number 10 Family Policy Unit should be looking at how government policy encourages or discourages couples to make clear decisions about their future and to formalise those decisions.

Third, policymakers need the confidence to base their public policy on the same principles most of them apply in private.

Politicians clearly understand the importance of personal commitment in their own lives. The vast majority are married. This is a good decision because, across all backgrounds, couples who marry are generally much more likely to stay together.

Yet few ministers are willing to stick their head above the parapet and say this. The policy signals they send out are that marriage doesn’t matter.

The tragedy is that those with fewest financial resources are listening. Nine of ten parents in the top income quintile with young children are married. In the lowest quintile, just a quarter are married. Pile relational ambiguity on top of low income and a system that bribes you to live apart, and you have the perfect recipe for family instability.

It used to be thought that unmarried cohabitation would look increasingly like marriage as it became more widespread. In fact the gap is widening. Today’s marriages are more stable than any since the 1960s. Divorce rates have plummeted because those who choose to marry really embrace their commitment to one another. And yet the UK continues to languish near the top of the family breakdown table because unmarried breakdown rates are typically three times higher.

There’s an important lesson here about commitment and stability. Few businesses succeed without making a clearly agreed plan that everyone understands. This is why today’s marriages are doing so well and today’s cohabitees so badly.

My friend never married. Would she and her partner have stuck at it had they made a more explicit commitment in the first place? Who knows. But had the policy signals been more encouraging, a clearer plan might have shifted the odds in their favour.

A Number 10 Family Policy Unit should have the key role of giving senior ministers the confidence to promote clarity of commitment – and therefore marriage and civil partnerships – as the centrepiece of a bold new family policy that boosts the odds of stability.