Jon Snow's claim that Tariq Aziz was a "nice guy" has been widely attacked: Nick Cohen called it "everything that's wrong with the British Left handily summarised in one tweet."
But what exactly is the error here? I don't think Snow is defending the Saddam regime: it's no defence of a murderous tyranny to claim that one of its members seemed like a decent bloke.
Instead, I suspect there might be one of two different mistakes - each of which is quite common.
One arises from the fact that, in judging Aziz's character, the personal impression he made on meeting him is only a fraction of the total information available. Any reasonable assessment would ask: would someone spend decades as a key member of a brutal dictatorship if he really were a nice guy?
The answer is: perhaps not. Even if Aziz started as a decent bloke, he might well have been corrupted by proximity to the worst form of power.
Snow seems therefore to be overweighting private information and underweighting public information.
But as I say, this is a common mistake: it's closely related to that most ubiquitous cognitive bias, overconfidence. We see often see it in financial markets. For example:
- The neglect of base rate information - for example the tendency for many takeovers to fail - and overweighting of private information ("this acquisition seems a good idea") can lead to catastrophically expensive takeovers, such as RBS's of ABN Amro.
- Investors often pay too much (pdf) for newly-floated companies because they overweight their private feelings and underweight an obvious public signal: if this is such a good business, why are the people who know it best selling it?
- One reason for momentum in asset prices is that people cling to their prior private beliefs. For example, if bad news hits a company, its owners are loath to sell because they cleave to the belief that it's a good stock. This causes prices to under-react to bad news, and so drift down in subsequent weeks.
Let's though, ignore all this and assume that Aziz really was a nice guy. The fact remains that his life did not improve the condition of humanity. And this poses a question which Snow ignores: to what extent does character matter?
Adam Smith thought not."It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest" he wrote. Marx agreed. The well-being of workers, he said, "does not, indeed, depend on the good or ill will of the individual capitalist." Both instead thought incentives and social structures were more important that individual dispositions - though they differed upon the effect that capitalist incentives had. And the famous experiments of Philip Zimbardo surely corroborate them.
In calling Aziz a nice guy, Snow was missing this point. Again, this is a variant of a common cognitive bias: the salience heuristic leads us to overweight obvious information - the impression a guy makes on us - and underweight less salient facts such as how social structures and incentives shape behaviour.
And again, Snow's error is common. Lefties commonly attack bankers and top footballers as greedy, when in fact their wealth arises not from greed but from the structure of 21st century capitalism.
In these ways, I'm sympathetic to Nick's point: Snow's tweet does express a lot that's wrong with how many people think about society.