First, the Spectator’s leader cites ONS research showing that married men are more likely to find work that single ones, and infers that “perhaps it’s time to chivvy the unemployed to church.” This inference suffers from two problems.
One is: why does marriage enhance employability? It could be because marriage causes men to want to work more, perhaps to escape the wife’s nagging. Or it could be that marriage is merely correlated with factors that make men attractive to employers: good social skills, reliability, a conventional mindset etc. There’s lots of research (pdf) on this question - none of which the Spectator cites - which is gloriously ambiguous.
This question matters. If the marriage-employability link is correlative rather than causal, then chivvying the unemployed to get married won’t improve their chances of getting a job, simply because it won’t give them the features that make men employable.
But even if the relationship is causal, there’s another problem - the fallacy of composition. It might be true that being married increases an individual man’s chances of finding work. But for all men, the chances of getting work depend upon aggregate demand. Yes, you could tell a story in which if men become more employable, demand for labour will eventually increase. But it won‘t do so immediately, and the Speccie doesn’t even try to tell this story.
Speaking of bad articles brings me to Jan Moir‘s now notorious piece. On first reading, I thought this was cunningly written, intended to convey an innuendo without exactly saying so. Reading her “apology”, however, makes me suspect it’s just slovenly rubbish.
Now, many people have objected to Moir’s frothing, hateful, sickening homophobia. There is, though, another flaw - she’s committed an elementary howler of reasoning. She’s taken two cases - the deaths of Kevin McGee and Stephen Gately - and drawn some kind of inference: “the happy-ever-after myth of civil partnerships“, a “more dangerous lifestyle”. But she hasn’t asked the basic question: what is the sample from which these observations are drawn? She seems to be drawing inferences from the extreme of a distribution without asking: what are the properties of that distribution? This is just irrational.
Had she began from those cases, then asked: are gay civil partnerships generally less happy than straight marriages? and produced good evidence on this question, she might have had a perfectly acceptable article.
As it is, she’s just failed statistics 101, and shown an inability to think.
I say this not to question the value of marriage; for some legitimate evidence, try this pdf. I do so merely to point out that some of its advocates sometimes produce mere irrational babble.