THERE is a story we are often told by the current Westminster Government, and it is this: that our former social bedrock, the once wonderful institution of marriage, has been broken and if we could only patch it back up then much that is wrong with society – crime, riots, poverty, abuse – would dissolve away.

This notion explains why earlier this year, Work and Pensions Secretary Iain Duncan Smith gave his backing to the newly formed Marriage Foundation. It also explains his speech last week, as he launched a new "Social Justice Outcomes Framework" and a scheme in which, among other things, charities, private companies and social enterprises will be paid by results for improving the statistics of parents who stay together. Quite what this might involve is unclear, but relationship counselling seems a likely candidate. The Government, he outlined, has set its yardsticks for social progress. One of them is to improve upon "the proportion of children who have a stable family free from breakdown, and the proportion of such families that report a good-quality relationship".

In part, at least, this is yet another story about people on benefits and how we might reduce their cost to the country. The Daily Mail reported the plan was designed to "help keep Britain's 120,000 problem families together" (this figure is, itself, controversial). The suggestion is that those bogey families need to be mended by, among other things, an institution like marriage and the wonders of a traditional family structure.

Of course, there is no disputing that reuniting couples who can make a decent attempt at a not-too dysfunctional relationship is a good thing. Nor that the aspiration for kids to grow up in happier, more stable homes is to be encouraged. But even if this policy appeals to our emotions and common sense, the principles are shaky. Duncan Smith is right to point out that the statistics regarding the life outcomes of children who have experienced divorce and separation are grim, but those figures don't tell the full story. We don't know what other factors – social, genetic or other – create this picture, so we can't assume that if the relationships are put back together, everything else will happily fall into place.

This policy seems like just another strand in a pro-marriage agenda, which in turn is driven by the myth that we have fallen from some golden age in which the principles of society were upheld by a vast majority of households headed by a resident mother and father. American writer Po Bronson has drawn together research on this belief as it applies to North America and shown that the traditional family has hardly ever been entirely the norm. A far higher proportion of stepfamilies existed in the US of the 17th century than do today, and even in 1960, around 40% of children were being raised in "non-traditional families". In the UK, similarly, we only need to look back half a century to find that this non-traditional structure, the single-parent household, served the post-war generation pretty well.

Yet still politicians cleave to the idea that the family, ideally in its two-up, two-down form, is the institution to save us from social ills. They forget to mention that, while two people are better than one when it comes to the hard work of child-rearing, a village is far better again. And in the meantime, they use accusations about "poor parenting" and even poor pair-bonding skills, as another stick to beat the public with.

But what is perhaps most depressing about this policy is the way it makes us view relationships. Payment by results creates a world that is black and white, in which the only good and financially rewarded end for the agencies and professionals involved is a relationship saved. A break-up, even if it really is for the best for one particular family, is a failure and non-payment. Who would want to be a professional or agency working in this system? And what lone parent or divorcee would not feel criticised by this strategy?

The proposal echoes structures in the work programme, and attempts through current welfare reform to get those on long-term benefit, particularly the long-term sick, back into work. There is the same payment-by-results approach, and the same questions are raised. How realistic are the aims of either project? Are there even jobs and an appropriate workplace for those long-term unemployed, some of whom have mental-health problems and significant disability issues? And, when it comes to the "outcomes framework", are there really relationships for these individuals to return to – ones which would not be abusive or highly dysfunctional? In many cases, there will have been good reasons why these couples split, or were not even together in the first place.

But perhaps the biggest flaw of Duncan Smith's plan is that his marriage-promotion agenda appears to be a money waster. As American social historian Stephanie Coontz has written: "We cannot afford to construct our social policies, our advice to our own children and even our own emotional expectations around the illusion that all commitments, sexual activities and care-giving will take place in a traditional marriage. That series has been cancelled."

Research shows, for instance, that welfare strategies don't alter the way people make relationship choices. And the Department for Work and Pensions knows this. A review of research on the subject, published by the department in 2009, concluded that "on balance, the reviewed literature shows that there is no consistent and robust evidence to support claims that the welfare system has a significant impact upon family structure".

The Government might as well give up, support our changing relationship choices and find a way to work with them. The family is carving out its own multi-laned path with plenty of options for all.