Popular perceptions exposed by numbers
What the public believes can be very far from reality, writes Tim Harford
‘A new survey by Ipsos Mori for the Royal Statistical Society and King’s College London highlights how wrong the British public can be . . . ’ Press release, July 9
That seems a bit harsh.
In fairness to the British public, the claim isn’t that we are always wrong. It is that when we are wrong we are very wrong indeed.
By a factor of 10, 15 – or more. Ipsos Mori was asked by the RSS to find statistical howlers and the market research company duly produced a list of its 10 favourite examples. These emerged from surveying more than 1,000 people, and focusing on areas Ipsos Mori had reason to suspect give us trouble.
So the pollsters set the Great British Public up for a fall?
No, they highlighted known weaknesses. For instance, everyone who has been paying attention knows two things about crime: it’s falling, and most people don’t know that it’s falling. Crime is less than half what it was in 1995, and a fair bit lower than in 2007. Violent crime is falling, too. But when surveyed, 58 per cent of people didn’t believe crime was down. This is an extraordinary mass hallucination.
Perhaps the statistics are wrong. You know what they say – damned lies and all that.
All statistics have their limits but I would be truly flabbergasted if some vast statistical conspiracy (or cock-up) had hidden the fact that crime has been rising for years after all. The statistics on immigration are probably more questionable – after all, some immigrants are illegal and so likely to resist being counted. The official view is that 13 per cent of the British population was born overseas; attempts to allow for uncounted illegal immigrants put that figure as high as 15 per cent. But take a survey and people reckon the true figure is 31 per cent.
Do they not know about the official figures, or do they not believe them?
Both. People don’t know about the official figures but when Ipsos Mori enlightened them, many were sceptical. And this matters. Ipsos Mori also found that three-quarters of people felt that too many immigrants were coming to live in the UK. Joe Public has spoken, and he says: “In my fevered imagination, there are too many immigrants around.” Sort that out, politicians.
Ah, well. Do politicians encourage these errors or pander to them?
A chicken and egg question, I fear. Some misperceptions are very convenient for the government. For instance, presented with a list of major spending categories – the National Health Service, defence, pensions and so on – people will tend to identify “interest payments on the national debt” as the biggest item. In fact it’s well down the list, less than half the cost of the NHS. And a substantial minority of people think Jobseeker’s Allowance costs the exchequer more than state pensions do. In reality, pensions cost 15 times more than JSA. People think that capping benefits at £26,000 a household will save a lot of money; it won’t. These misperceptions do play into George Osborne’s political strategy.
Come on, then – what’s the biggest error of all?
It’s probably not the most important error, but in sheer orders of magnitude, our view of underage pregnancy is pretty wrong. We think 15 per cent of girls under the age of 16 get pregnant each year, a figure that strains the limits of biology. The official statistic is that 0.7 per cent of girls aged 13-15 have either a live birth or a legal abortion. We have a grim view of some corners of British life, and fortunately that grim view is utterly wrong. One bright spot: people generally believe that their own area is closer to the way they like it with lower crime, lower unemployment, better policing, fewer immigrants. It’s the rest of the country they worry about.
That’s all very nice – but it doesn’t alter the fact that the British public are about as wrong as anybody can be about some basic facts.
You think so? You obviously haven’t heard of quantitative finance. The chief financial officer of Goldman Sachs commented, at the beginning of the financial crisis, that the investment bank was seeing “25 standard deviation moves, several days in a row”. It’s not totally clear what he had in mind but given some reasonable assumptions, it simply meant that Goldman’s mathematical models were wrong by, oh, 215 or 216 orders of magnitude – that is wrong by a factor of a trillion trillion trillion etc . . . with a total of 18 “trillions”. “Perceptions are not reality,” says the RSS. Well, quite.
Also published at ft.com.