31 January 2007

Who's afraid of il-liberal hegemony?

Uh-oh. I may be about to step on some blogosphere toes.

Stumbling & Mumbling (Chris Dillow) and Tim Worstall are two of the most intellectual British blogs. They lend a touch of class to the UK blogosphere — which some claim it sorely needs. I ought to be happy about their existence, and I've been trying hard to like them. They are both obviously written by pretty clever people. They are also very popular, with loads of people linking to them, including The Times's Finkelstein.

I'm still making my mind up about TW (I enjoyed his recent attack on the latest trendy piece of pseudo-psychology) but I've seen two posts in succession now on S&M I didn't much like, as they struck me as implicitly reinforcing the ideological bias of the cultural establishment, by quoting its goings-on uncritically.

1. Nobel Prizes and longevity

Dillow: If you want a long life, win a Nobel prize. This paper by Andrew Oswald and Matthew Rablen estimates that Americans who won Nobel prizes for physics or chemistry between 1901 and 1950 lived 2.08 years longer than scientists of equal age who were nominated for Nobels but never won them. This is not because Nobel prize-winners were richer than nominees and richer people live longer; the longevity premium, they estimate, is uncorrelated with the real money value of the prize. Instead, this confirms that high status prolongs life; here's other evidence, from Oscar winners.

This matters enormously. The difference in status between a Nobel prize winner and a nominee is small; both are enormously successful and respected people. So, if small differences in status at high levels lead to big differences in longevity, isn't it possible that the bigger status differences across society might explain some of the large differences in life expectancy? If so, isn't it important that we at least ensure that inequalities in status are merited?
So giving someone a Nobel is supposed to extend their life span by an average of two years. But hang on a moment. What about the possible explanations that (a) there is a correlation between intellectual and physical superiority (gosh, doesn't that sound politically incorrect), or (b) that higher ability leads to better dietary behaviour? These are given scant space by either of the papers Dillow refers to, and Dillow himself doesn't mention them.

(A reader of Dillow's blog facetiously commented on this by quipping: "I love the idea that Nobel runners-up are sat around each tea time eating Iceland chicken dippers with oven chips and tomato sauce." Very funny, but we are talking about small statistical effects. I don't see anything inconceivable in the possibility that (i) Nobel Prize winners might be, statistically, a lot cleverer than nominees, and (ii) that the kind of cleverness involved makes the people concerned good at working out the best way to extend life span. Think Linus Pauling for instance. The whole point about research is that you can't prefer some explanations to others just because they strike you as more plausible.)

The first paper refers to a study which allegedly controlled for these possible factors in monkeys by "manipulating rank". The study on monkeys claims to have demonstrated a causal flow from status to physical health by allegedly showing that whichever monkey was artificially set up as leader, that monkey had lower levels of stress hormone. Reference to this study is as close as we get, in the paper on Nobel winners, to a serious consideration of the alternatives. I would be surprised, incidentally, if monkey troops are really as amenable to exogenous leader selection as is implied by this.

It's not that I reject the possibility that social status has psychological, and hence physical, benefits. I expect it does. But there is clearly a hidden bias in this research. It prefers to focus on the idea that social factors determine individual differences, and to avoid the possibility that genes are more important than environment — the old nature versus nurture thing. Which, notwithstanding any apologias from people such as Steven Pinker, still appears to be interpreted in favour of nurture, in most academic departments.

Since the genetic factor has not been controlled for, the finding in question (i.e. Nobel Prize correlates with two years extra life span) is utterly useless as a predictor of the effect of status on longevity. Which effectively makes this research, as presented, highly tendentious and misleading.

2. "Benign" inflation

This carp on my part is perhaps a little unfair. Dillow quotes Anatole Kaletsky in The Times referring to "the relatively benign social character of inflation — i.e. hitting the rich much harder than the poor". The primary villain here is Kaletsky, from whom one expects better. His statement tacitly approves of the notion that an adverse economic effect is "socially benign" if it affects rich more than poor, in effect expressing approval for redistribution. The fact that someone as normally anti-left as Kaletsky uses this phrase demonstrates how pervasive the leftist assumption that redistribution is desirable has become among economists of all persuasions.

My beef with Dillow is that he quotes this statement but confines himself to criticising the assumption that inflation does have the redistributive effect Kaletsky claims, without questioning the hidden ideological assumption. By doing so he is himself giving tacit support to a claim which is political rather than economic.

On the plus side, I rather like Dillow's diatribes against managerialism.

Part 2 of this post tomorrow, with further discussion of the blogosphere's take on ideological bias in contemporary culture.