The logic of family policy

At the heart of this morning's Today programme encounter between Ed Balls and David Willetts over family policy was Willetts' assertion that "income tax should recognise marriage," which, he explained would be the "most effective way of helping families."

Effective? In doing what? Nobody quite said, but John Humphreys approached the question by asking whether this was a form of "social engineering" based on an idea about "the ideal family."

Willetts parried with a remark about "the symbolism of who you can help". 

All three seemed to accept that if the families of married couples enjoy advantages, then it would be a good idea for more people to get married: 

Ed Balls: "The evidence shows that if you're married, the children do better."

John Humphrys: "So why not support them financially through the tax system?"

Balls tried to say that if you adjust for other factors, such as class and age, the effects aren't so marked. 

What was never tackled directly was the possibility that getting married doesn't make a couple stay together; rather, it's the ones who think they'll stay together who decide to get married. 

But much of the debate on this issue has taken for granted that getting marriage brings advantages. Here is what a report by Iain Duncan Smith's Centre for Social Justice - had to say in December (my emphases):

Five ways in which marriage matters:

(1) Marriage brings stability: just one in 11 married couples split before their child's fifth birthday compared to 1 in 3 unmarried couples.

(2) Marriage is directly linked to better mental and physical health amongst adults. The same benefits are not found amongst co-habiting couples, it is specifically a Marriage Effect'.

(3) Marriage reduces the risk of violence and abuse. Children growing up in lone parent or broken families are between 3 and 6 times more likely to suffer serious abuse than those growing up with both biological parents, and the risk of domestic violence is significantly increased for co-habiting women than married women.9

(4) Marriage leads to better mental health for children. Children of lone parents are more than twice as likely to suffer mental health problems than children of married couples, and those of co-habiting couples are 75 per cent more likely to have mental health problems than their peers with married parents.

(5) Marriage leads to better life outcomes for children. Children of married parents are more likely to achieve at school, less likely to use drink and drugs and less likely to get involved in delinquent or offending behaviour.

It looks like hard-headed number-crunching. But policies to support married couples financially only make sense if being married causes these outcomes. 

In the above list of claims, only one ("is directly linked to") does not make this assumption. The rest assume a causal connection on the basis of a correlation. 

If this is the logic of the policy, it's like encouraging people to take private health insurance to make them rich. Or issuing everyone with sunglasses to make the sun come out. Or persuading blokes to behave badly in nightclubs to turn them into successful footballers.