Cognitive biases everywhere

Yesterday I asked: do anti-racists irrationally prefer statist mangerialist interventions to market forces? This set me thinking: what sort of cognitive bias could explain why they do so?
One possibility, I thought, is the representativeness heuristic, the process whereby we assume that effects must somehow look like their causes.
Maybe this leads people to over-rate the ability of state intervention to combat racism because they assume that an anti-racist policy must somehow be representative of anti-racist outcomes.
However, they under-rate the importance of markets because market forces aren't representative of racial equality.
Having thought that, I then started my day job. I was discussing with a colleague the possible consquences of a slowdown in UK manufacturing output. What impact, if any, would this have on engineering stocks returns and equities generally?
One's instinct is that that it's bad for engineering  stocks but not for the wider market, as the vast majority of UK equities have little to do with UK-based manufacturing.
This is wholly wrong.
Here are the data for annual changes in manufacturing output and share prices since 1995:
All-share index: correlation with output = 0.55, R-squared = 30 per cent.
FTSE enginerring and machinery sector: correlation with output = 0.32, R-squared = 9.5 per cent.
So, manufacturing output matters far more for the All-share index than it does for the FTSE engineering sector.  Indeed, if we control for changes in the All-share, the impact of fluctuations in manufacturing output  on engineering stocks is negative and statistically insignificant.
Why did instinct tell us otherwise? It's the representativeness heuristic again. We assume bad times in manufacturing must be representative of bad times in engineering shares, but not representative of bad times in shares generally.
This leads us to ignore more a more subtle causal path. Cyclical weakness in UK manufacturing output is usually due to weakness in the global economy*. And this causes global investors to become more risk-averse and hence to sell shares around the world, and the All-share falls as this happens.
So, in two different areas, the representativeness heuristic crops up.
Does this show the strength of the Kahneman and Tversky-inspired research into cognitive biases? Or is it instead a weakness of this research - that once we start looking for cognitive biases, they appear everywhere and can explain pretty much anything?
* I stress cyclical here because UK manufacturing output has for years declined relative to that elsewhere.

Comments

[rjw]

Nice post. Very thought provoking. I'm guessing you would like Jon Elster's work on the scope and limits of rationality and on the methodology of the social sciences, if you don't know of it already. Some of the most stimulating stuff in social science I have read.

[dearieme]

Let us set aside that set of anti-racists who're in it for themselves, who're searching for jobs or political power, and those who uncritically approve of what is approved of. Could one cognitive bias arise from the belief that the most murderous cases of racism are mediated by large institutions: the Nazi State,the Church, the Ottoman Empire, whatever, and that therefore opposition should also be mediated by large institutions? If so, it's a rum bias, because it overlooks the obvious rejoinder that the list of horrors proves the imprudence of trusting power to large institutions and the wisdom of dispersing power e.g. through the markets.

[Paddy Carter]

I fear I'm in danger of missing the point, but hasn't the fact that most anti-racists align themselves with the left got more to do with the left and the right being bundles of policies and attitues that bear little logical relation, and that people just tend to apply labels and support their side rather like football fans, regardless of merit. What I mean is that there's not much point in searching for a explanation - cognitive bias or otherwise - to explain the cohabitation of anti-racists and lefties - because since when did things ever make sense? One might as well ask why racists are associated with free market economics. I reckon just about the only thing that the left has in common with anti-racism is concern for fighting injustice and sticking up for the oppressed (or at least that's what the left thinks it's doing).

[dearieme]

Nah, the Left was let down by the Working Classes who proved to be unrevolutionary, so they are looking for some other bunch of mugs, sorry, deprived groups, who can be forced to march towards the cannons to the beat of the revolutionary drum. It also explains their love of immigration: if the people won't back you, change the people. My comment above did start with the disclaimer that I was going to set these sods aside. And may I say that "One might as well ask why racists are associated with free market economics" is preposterous: they're not. Racists are often part of a market-opposing Trade Union structure. Look at the history of Afrikaaners and Apartheid or some of the Orange Lodges in the old Scottish coalfields: these guys were Trade Unionists opposed to what had been Irish strike-breaking labour. And, at least by the 1960s when I first understood what I was seeing, they voted Labour.

[Paddy Carter]

dearieme we are talking at cross purposes I think - that's the real world you're talking about there, or at least part of it - in another part (in the minds many) there is an association between the two: free markets => right wing, racists => right wing therefore racists free markets. As I said, since when did things ever make sense?