Do you like Skittles? Do you like them so much you'd eat one from a packet even if you knew a couple were poisoned? Makes u think, ey?
This was the analogy Donald Trump Jr made this week about refugees and terrorism, a tweet which caused anger, not least from the company that makes Skittles, who responded: 'Skittles are candy. Refugees are people. We don’t feel it’s an appropriate analogy.'
This particular meme seems to have originated among feminists, who were making the point that it's no good saying that most men aren't violent rapists, because enough of them are to make problems: 'Imagine a bowl of M&Ms. 10 per cent of them are poisoned. Go ahead. Eat a handful.'
That seems like a reasonable point to me; denying the link between masculinity and violence, when 90 per cent of murders are committed by males, would be absurd. If you get in a car driven by a man rather than a woman your chances of being violently assaulted are far higher, and that is not an insult to men. It's a question of relative risk, and so it's quite reasonable for men to expect initial discrimination in some employment areas.
My only objection would be with the actual proportion. Nowhere like 10 per cent of men ever commit violent crime, and the percentage of Syrians who are terrorists is much smaller than the Skittles analogy suggests. A better one might be to suggest that one in every 100 packets contains a poisoned sweet - small overall, but I wouldn't buy any.
And the risk of terrorism posed by refugees is not non-existent. There are at least four studies showing that accepting refugees does increase the risk in the receiving state, although it's smaller in wealthy countries, largely due to better security infrastructure. The risk to non-OECD countries, who do take most refugees, is far higher.
The issue is essentially to what extent you think the risk is acceptable. Conservatives tend to perceive fear more readily than liberals. A liberal might say that x terror incident is worth the benefits of accepting y number of refugees, while conservatives are more likely to err towards zero risk.
It is also about probability; rates of crime vary hugely across the world, so if one in a thousand people in the host country behaves violently, and one in a hundred of an incoming population do, then that difference will make a significant impact on a lot of people's lives.
The best way of looking at immigrant behaviour risk is not by comparing it with weird, random occurrences, such as the Economist do, but with natives. In Denmark, for example, Somali men commit rape at an estimated 26 times the Danish rate, adjusted for age: most Somali men are not rapists, but this is still relevant information people should know before allowing large numbers of migrants into their country. The media lose people's trust when they fail to address such matters.
I'd have more sympathy for Trump's opponents if they made rational arguments for allowing more migrants - aside from economic ones - rather than relying on sentiment and guilt. In Britain, we have the 'I am an immigrant' poster campaign, the numerous historically dubious comparisons to Russian Jews or Huguenots, or the emphasis on totally atypical individuals like Mo Farah or Steve Jobs.
The reason pro-migration advocates like sentiment is because that way the debate will always be asymmetrical, for opponents can never make the same arguments without appearing unpleasant or even inflammatory. Someone could just as easily bring out an 'I am an Immigrant' campaign showing pictures of rapists, one of whom was recently caught trying to pick up and drug girls close to where my 7- and 6-year-old daughters go to school.
I wouldn't advocate that, because when people feel personally threatened they react emotionally and sometimes violently. It's why, when we're discussing the potential movements of hundreds of thousands of people across borders, changes that will have huge effects on millions of people, we need to keep the debate as calm, rational and detached as possible.