Thursday, December 04, 2008

The calculus of the caring society

I'm not really much of a numbers guy, but I was struck by a recent post by Ben Goldacre at his Bad Science blog which highlights the care that needs to be taken when using statistics to measure changing social attitudes.

More accurately, he--for, like, the thousandth time--points out how the media fail to take this care.

My interest in this issue has increased, as I'm currently making my first forays into using serious quantitative data for a historical study. Indeed, I spent much of the last two days trying to come to grips with pivot tables so that I can make some sense of the information I've been carefully collating and entering into a spreadsheet.

Some of you may have gotten drowsy simply reading those last two sentences, and not long ago I might have been with you; however, I'm starting to find something exciting about making numbers do various little dances. (And, after a period of wanting to throw my laptop against the wall out of frustration, I seem to have gotten my tables to do what they are supposed to.)

I have the good fortune of working with a colleague who knows his way around numbers, and I've become increasingly aware of how much discipline is required in teaching your figures the requisite dance steps. This is especially so in cases when you're trying to use that data to tell you about such nebulous topics as 'cultural attitudes' or 'social life'.

The specific topic of Goldacre post was a recent spate of stories suggesting that there had been a shift toward a more 'caring society' based on a figure showing a greater number of births of children with Down syndrome. The logic behind this, I think, is that the growth in Down syndrome births showed a greater willingness to bring such pregancies to term. A number of media sources took this story and ran with it, expanding on the notion that this figure might tell us something meaningful about social attitudes.

And, yes, maybe they have changed in some positive ways.

But Goldacre points out the key problem with drawing that conclusion from that figure: while more Down pregnancies are being brought to term, there is another factor that needs to be taken into account in order to place that in its correct context:

There has indeed been a 4% increase in Down’s syndrome live births in England and Wales from 1989 to 2006 (717 and 749 affected births in the two years respectively). However, since 1989 there has also been a far greater increase in the number of Down syndrome foetuses created in the first place, because people are getting pregnant much later in life.

What causes Down syndrome? We don’t really know, but maternal age is the only well-recognised association. Your risk of a Down syndrome pregnancy below the age of 25 is about 1 in 1600. This rises to about 1 in 340 at 35, and 1 in 40 at the age of 43. In 1989 6% of pregnant women were over 35 years of age. By 2006 it was 15%.

I was not aware of the age-at-birth link to Down syndrome. (My mother was about to turn 44 when I was born. 1 in 40 might sound like long odds to you, but they're not long enough if you ask me...)

But what does this shift in birth patterns mean for claims of a more caring society?

Well:

The National Down Syndrome Cytogenetic Register holds probably the largest single dataset on Down syndrome, with over 17,000 anonymous records collected since 1989, and one of the most reliable resources in the search for patterns and possible causal factors. They have calculated that if you account for the increase in the age at which people are becoming pregnant, the number of Down’s Syndrome live births in the UK would have increased from 1989 to 2006: not by 4%, but from 717 to an estimated 1454, if screening and subsequent terminations had not been available.

Except, of course, antenatal screening is widely available, it is widely taken up, and contrary to what every newspaper told you this week, it is widely acted upon. More than 9 out of ten women who have an antenatal diagnosis of Down’s syndrome decide to have a termination of the pregnancy. This proportion has not changed since 1989. This is the “decisions” that Felicity Finch, Radio 4, the Mail, the Times, the Mirror, and the rest are claiming more parents are taking: to carry on with a Down syndrome pregnancy. This is what they are taking as evidence of a more caring society. But the figure has not changed.

While I'm interested in this story largely as a cautionary tale for the misuse of statistical data leading to the misunderstanding of society (which is a problem that is magnified in a historical context, when the statistics are a bit more iffy and our other information about social context more partial), Goldacre points out that there is a more direct reason why such stories are problematic.

Crass and insensitive moral reasoning helps nobody. If I terminate a Down syndrome pregnancy, is that proof that society is not a warm caring place, and that I am not a warm caring person? For many parents the decision to terminate will be a difficult and upsetting one, especially later in life, and stories like this make a pretty challenging backdrop for making it. This would have been true even if their figures had been correct, but as is so often the case, for those with spare flesh to wave at strangers, their facts and figures are simply incorrect.
He also links to further background provided by the NHS, if you care to look.

3 comments:

The Honourable Husband said...

Important points, JCW.

I'll take a mildly opposing view, for a moment.

Inference from numbers (not statistical inference, in the technical sense) is a good thing. That's where scientific hypotheses come from. It would sadden me if every hunch from numbers which the huncher hadn't proven, were dismissed as worthless.

That said (and here I flip-flop to agree with you) the Down's Syndrome case is just plain wishful thinking.

Responsible speculation from statistics really needs to acknowledge alternative hypotheses.

A classic case in point is the recent exit-polling for Prop 8.

The gay community was aghast that 70% of African-Americans voted in favour.

What's the explanation? Black homophobia? The influence of churches in the black community?

All possible. But the single most plausible explanation (in my view) is that Prop 8 looked a little too much like rich folks who want weddings, rather than struggling families of every kind, who want protection.

The gay-friendly press has largely ignored social class as contributor to the Prop 8 result. Why? Because it begs further, broader questions about American inequality that would be painful to confront.

I have a blogpost vaguely swimming in my head about this.

JCWood said...

Thanks hb8, good points all.

I'd say we're on the same side as far as using good statistics to draw conclusions about society. This is important and useful.

The findings by many social historians that homicide rates have declined significantly in European countries since the Middle Ages is one good example close to my own heart. This has, then, raised a lot of new questions which can only partly be answered via quantitative methods.

I've seen both some very good and some highly questionable (and some just plain wrong) use of those methods in history.

And, given that I struggle with the kind of relative innumeracy that is not all that uncommon among humanities types, I feel that I'm doing reasonably well in being able to tell the difference between these different categories.

Your points on Prop 8 are well taken. Class is clearly an issue that doesn't get the attention that it should in the US, and the PR work by the people promoting an expansion of marriage rights might have been misguided on that score.

I'll have to think about that.

But I also think, given the strong religiosity in parts of many black communities, we shouldn't entirely overlook the possibility that many black voters were, genuinely, opposed to granting marriage to gay people.

And if they are, that is not only a shame but also points to some work that needs to be done in changing those minds.

I look forward to that blog post you mention...

Till then, Tschüß!

Anonymous said...

Another example of superficial statistics-based inferences is the accusation, common in the European media, that America must be a racist society because there's a dramatically higher percentage of blacks in prison than in the population as a whole.

The problem is that blacks also commit crimes at a significantly higher rate than their presence in the population, a fact usually ignored by Europeans. Europeans who indulge in this particular form of American-bashing can be brought up short by noting that minorities are also disproportionately represented in European prisons. They hasten to explain that this, of course, is simply a result of poor social adjustment or a "lack of desire to integrate", rather than intentional racism on the part of the majority ethnic group. When it's their own society they're called upon to evaluate, nuance and qualification and context suddenly become critical to informed discussion. As Czeslaw Milosz once put it, it's always the other country's minorities who are the noble, oppressed savages: your own homegrown minorities are lazy grifters.

Note that my point isn't that there's no racism in the American or European criminal-justice systems, but rather that superficial top-line statistics don't prove it.