Gun crime and single parent families, yet again

(an answer to Shuggy's comment on my last post on this topic)

1. Heroic assumptions

You accuse me of not understanding Chris Dillow's methodology. Believe me, I do. The real problem is that you haven't noticed that he's talking out of his bottom.

Let's examine his 'heroic assumptions':-

1. All gun crime is committed by males. Nothing very heroic about that, is there?

2. All gun crime is committed by males who grew up in single parent households. Still not particularly heroic. In my last post I made my own assumption that it is 80%, and as I said that may well be conservative, given the overwhelming concentration of gun crime in communities where single parent families are the norm.

3. 11,084 crimes involving firearms in 2005-6 represent 11,084 criminals. Well clearly this is heroic in the sense that there will have been a lot of individuals who were responsible for more than one of the crimes. On the other hand...

(a) many of the incidents will have involved more than one person

and

(b) there is massive underreporting of gun crime because victims are too scared to go to the police.

4. Now for the really fatal flaw in Chris's methodology (Danny Finkelstein was indeed on the right track, as Chris acknowledged by ignoring the point). He is taking the number of males in the 16-24 age group - a range of nine years - and comparing the number of crimes committed in one year. So we are not taking account of individuals in this category who had committed gun crimes before 2005-6 but didn't during that year - e.g. because they spent the year in custody. Nor those who didn't commit gun crimes in 2005-6 but have done or will do subsequently.

So let's make some new assumptions. Let's assume that the number of individuals who committed a gun crime for the first time in 2005-6 is just a quarter of the number of reported crimes: 2,771 (since there were 2,365 convictions for indictable firearms offences in 2005 - see table 2.12 here, which unfortunately doesn't say how many were first offences - and since the proportion of offences resulting in a conviction is plainly abysmally low, I don't think this is unreasonable). Let's assume that 80% of them were males who grew up in single parent households: 2,217. Chris estimates 560,000 16-24 year-old men who grew up in lone-mother households. Divide that by 9 so that we're comparing a year's worth with a year's worth: 62,222.
2,217 is 3.56% of 62,222. You may still feel that that is insignificant and you are free to question my assumptions, but Chris's argument for setting 2% as an absolute ceiling lies in ruins, I think.

2. Hoist by his own petard

Chris estimates that 16% of the 16-24 year-olds grew up in lone-mother households. Coincidentally, the Unicef report (table 1.1) estimates that approximately 16% of children in the UK live in relative poverty. So if he has proved that single parent families are an insignificant factor in gun crime, he has also proved that poverty is an insignificant factor in gun crime. I rest my case, m'lud.

3. The evidence

If you choose to believe that, when researchers sit down to design a study of the impact on children of single parent families, it never occurs to them to adjust for income level, the best I can do is to refer you to the Unicef report. And remind you that its author is the founder of the New Internationalist - not exactly a Times-reading blimp, one would suppose. From page 23:

'The use of data on the proportion of children living in single-parent families and stepfamilies as an indicator of wellbeing may seem unfair and insensitive. Plenty of children in two-parent families are damaged by their parents’ relationships; plenty of children in single-parent and stepfamilies are growing up secure and happy. Nor can the terms ‘single-parent families’ and ‘stepfamilies’ do justice to the many different kinds of family unit that have become common in recent decades. But at the statistical level there is evidence to associate growing up in single-parent families and stepfamilies with greater risk to well-being – including a greater risk of dropping out of school, of leaving home early, of poorer health, of low skills, and of low pay. Furthermore such risks appear to persist even when the substantial effect of increased poverty levels in single-parent and stepfamilies have been taken into account [...]'

One senses that if Peter Adamson could find some way of rubbishing this research, he would. But evidently he can't.

Shuggy junior destined for a life of crime? No, of course not. Some reasons why not: 1. Dad's a teacher. 2. He doesn't live in Peckham and won't go to school there. 3. Dad is responsible and caring and spends plenty of time with him. 4. He will have lots of friends from stable families, maybe grandparents, aunts, uncles in stable relationships (the absence of this is what I was trying to get at by talking about a 'multiplier effect'.

So I don't believe there's any need for defensiveness on your part. Who's a perfect parent? I'm not expecting to be one, I assure you. In any case, though, whether there is a problem and whether anybody feels got at when the problem is discussed are obviously two entirely separate questions. And since the first is a matter of life and death in Peckham, the second really does have to take second place. As Mr Adamson acknowledges, sometimes facts are unfair and insensitive beasts.