A glorious example of the pocket intellectual’s most basic argumentative error today from ‘Peter’ Hitchens:
“The Canadian figures show that a child is 50 to 100 times safer with natural parents than with a step-parent in the home. The British research found married homes were 33 times safer than those with serial boyfriends. Stable marriage safeguards children.”
Correlation is not causation, nor does it imply causation. I see ‘Peter”s research and I raise him, as an example, Sidebothama, Heron and ALSPAC Study Team (2006) Child maltreatment in the “children of the nineties”: A cohort study of risk factors Child Abuse and Neglect 30:497-522:
“This study supports previous research in the field demonstrating that a wide range of factors in the parental background, socio-economic and family environments affect the risk of child maltreatment. By combining factors within a comprehensive ecological framework, we have demonstrated that the strongest risks are from socio-economic deprivation and from factors in the parents’ own background and that parental background factors are largely, but not entirely, mediated through their impact on socio-economic factors.”
The Office of National Statistics don’t collect data on marriage by socio-economic status, but taking a couple of proxies, we can have reason to believe that it is deprivation which is underlying both marriage rates and incidence of child abuse in the UK. Taking the most recent data on marriage rates in the UK¹: starting on page 36, table 3.1 – married couples with dependent children tend to be better educated than either co-habiting or lone parents with dependent children². Turning to maps 5.2 on page 74, the geographical distribution of marriage/co-habiting/lone parent families broadly mirrors the geographical distribution of wealth. If you’re unmarried with dependent children, you are more likely to live in a poor area and have a poor quality education. Taking these as proxies for socio-economic status, unmarried people with dependents are more likely to be poor than their married counterparts. Given the research linking poverty to likelihood of abuse, it seems likely that poverty is co-determining abuse and marriage, providing a sensible alternative explanation to their correlation.
On this basis, ‘Peter”s conclusion is somewhat wide of the mark:
“But all those who have connived at the dismantling of marriage, and continue to connive at it, should recognise their own grave guilt in sacrificing the welfare and happiness of children to the selfishness of ‘liberated’ adults who ought, above all, to be shielding the young from harm.”
Marriage is not the good we should be pursuing, but a distraction. Consider how we might ‘promote’ marriage – the discussion tends to be around tax breaks and other financial incentives. Leaving aside for the moment considerations of the efficacy of such policies³, given that you’re more likely to be married with children if you’re middle class to start with, these policies would have the immediate effect of redistributing wealth to the middle-classes. This wouldn’t necessarily entail a decrease in funding for poverty reduction, but would be less effective in reducing abuse than a comparable increase in poverty reduction spending would be. If our object is reducing abuse, giving money to those who already don’t abuse for living lives emblematic of their low likelihood of abuse as an incentive merely to remain in that emblematic state seems counterproductive. To put it more simply, if marriage is merely a symptom of being well off, and abuse a symptom of being poor, we should be treating the root causes of poverty and the marriage will look after itself.
It’s easy to look at the world at some arbitrary point in the past (‘Peter’ chooses 1965 as the date when, for him, everything started going badly wrong), pick differences between then and now and see patterns. We could draw graphs linking global temperature and Tesco’s market share, women in parliament and divorce, number of countries in the EU and UK birth rates. Their correlation, however, would not be enough to imply a link, or enough to base policy on. The fact that you’re less likely to be abused if your parents are married is not a good reason to promote marriage if it’s only another way of saying you’re less likely to be abused if you’re well-off. It would be nice to believe, as ‘Peter’ seems to, that marriage is the panacea for all social ills, but it’s almost certainly more complicated than that.
¹ Incidentally, if you’re looking for an eloquent precis of the difference between correlation and causation, you could do worse than the box on page 39.
² Tables 3.9a and b, page 43, show the knock-on effect from this, as you’re more likely to continue in education if you’re in a family with married parents, educational status and marriage acting in a declining spiral.
³ Such policies seem to rest on the assumption that people will happily live together if only there was enough financial incentive – that a couple of hundred a year from the government will save troubled marriages, or that such money will encourage people whose relationship is not solid enough for them to choose to marry currently to form a stable and lasting marriage. Which n those terms, it seems more likely that such policies will merely reward people who were staying together anyway.