The association fallacy

���We didn���t mean that kind of Brexit��� say Daniel Hannan and Andrew Lilico in response to May���s anti-immigrant proposals. To which we Remainers reply: if you ride a tiger, you shouldn���t be surprised when it bites you. As I wrote before the referendum:

some of you have a vision of a Britain outside the EU that is a free, liberal socialistic country. These are ideals with which I have sympathy. But we are kidding ourselves if we think a vote for Leave will be a move towards such a society. Instead, it���ll be a mandate for Farage and the inward-looking, reactionary mean-spirited philistinism he embodies.

But was my reasoning sound? It might not be.

I say this not because of Daniel���s claim that Leavers were motivated more by a desire to reclaim sovereignty than to cut immigration: many wanted to take back control precisely because they wanted to cut immigration, whilst others just can���t articulate why they want sovereignty.

Instead, there might be an error here: the association fallacy. Wasn���t I trying to discredit decent Leavers like Andrew and Daniel by association with indecent ones?

Let���s assume I were. Was this necessarily a bad thing?

Sometimes, we can reach the right decision because errors cancel out. For example, the erroneous belief at the roulette table that red is on a roll can correct the error that ���black must be due to come up next���. Or mental accounting can protect us from spending too much through weakness of will by putting some money off limits. In similar fashion, I was trying to use the association fallacy to counter what I saw as wishful thinking by free market Leavers.

But there���s another defence of what I was doing. It���s that the association fallacy might not actually be a fallacy. It���s only one if it commits the sampling error.

Let���s take a clear example of the fallacy: ���You shouldn���t be a vegetarian. Hitler was a vegetarian!��� This fails because of the sampling error: the vast majority of vegetarians are not genocidal maniacs. But what if a disproportionate number were? Wouldn���t this be at least a clue that there might be something wrong with vegetarianism?

In the case of Brexit, the fact that a disproportionate* number of bigots were on the Leave side wasn���t just diagnostic of something wrong with the Leave case. It was also causal. The association of the Leave cause with anti-immigration sentiment invited the government to become hostile to immigrants. Yes, Daniel and Andrew claim that Ms May���s inference is mistaken. But this doesn���t acquit them of the charge of wishful thinking: it is na��ve to hope that governments will do the right, liberal thing.

If you think this is another post about Brexit, you���d wouldn���t be wholly right. When Nick Cohen and James Bloodworth attack Corbyn and the far left for being pro-terrorist and anti-west they are asking me the same question I asked Andrew and Daniel: aren���t my ideals tarnished by association with some bad characters? And the answer is along the same lines. It depends upon proportions: how many such characters are there, and how much power do they have to pervert my intentions?

I don���t know how to answer that. My point is instead merely the trivial one that in a second-best world of bounded rationality and bad people, there becomes much more to politics than simply asserting one���s ideals.

* In saying this, I���m not claiming that all Remain voters are pro-open borders: they are not.