FEATURE1 February 2009

Think of a number

Michael Blastland is on a mission to open the public’s eyes to the truth behind statistics. Robert Bain asks him why he believes data is so often mishandled.

?Michael Blastland confesses to being a data geek, but he insists it’s not numbers he’s really interested in, it’s “what they can do for us”.

“Most of our public conversation is driven by numbers,” he says. “If it’s health it’s waiting times, or the mortality rate of this surgeon compared with that surgeon. If it’s crime it’s the number of stabbings and whether it’s gone up or down. If it’s education it’s league tables… There isn’t a subject that gets into the news that doesn’t depend to a very high degree on some kind of quantitative analysis, and generally reporters are not very good at understanding the way those numbers are generated and what they mean.”

Blastland is not a statistician – he studied English and eventually found himself working as a producer of factual programmes on BBC Radio 4. “I put together an understanding of economics in reverse, if you like – from trying to understand the big issues that dominate the news headlines. I asked all the stupid questions that an English graduate feels entitled to ask, and it’s striking that, dumb though the questions are, the answers are often dumber. I assumed that the numbers that appear in the newspapers are well founded and well understood, and it’s not the case.”

This was how he, together with economist Andrew Dilnot, came up with the idea for their show More Or Less. A weekly programme about numbers was tough to pitch to the BBC, but persistence paid off, and Blastland and Dilnot were eventually given the green light “so that we’d go away”.

The pair co-authored a book, The Tiger That Isn’t, which sets out to help readers ask the right questions about the numbers they are confronted with. A revised edition has just been released in the US under the title The Numbers Game.

“People have an ambivalence towards numbers,” says Blastland. “One attitude is that they are ‘killer facts’ – the single piece of information that will tell you everything you need to know. The other is that it’s all lies, damned lies and statistics. So the choice seems to be between gullibility and surrender. There has to be a better way.”

“People know that one month’s fall doesn’t indicate a trend in the retail habits of a nation. They know that you can have a big wave on a falling tide or a small wave on a rising tide.”

The only people who can hold both of these views at the same time, he says, are journalists – who fill the headlines with the ‘killer facts’, while thinking at the back of their minds that it’s all a load of rubbish. They are one of the key audiences of his book, along with general consumers of news, and the people – researchers included – who generate all these numbers in the first place.

As he speaks Blastland poses many questions, with a habit of pausing just long enough for you to think he expects an answer before providing one himself. These simple questions are the essence of his approach to numbers. Many of them, including his favourite, “Is that a big number?” are so basic as to seem slightly insulting – until you realise you’re not sure of the answer.

Blastland’s belief is that “we all know more than we think we do” – no complex mathematical analysis is needed to answer these questions, just a sense of human proportion. After all, “it’s people who are doing the counting and it’s people they are counting, more often than not”.

In the book he sets out to demonstrate how easy it is to be duped by numbers. The recent financial crisis has provided plenty of fresh examples – Blastland picks one from the reporting of the bank bail-out in the UK. “If you look at the sum of money involved, £37 billion, it’s got a lot of zeroes on the end, and that’s the way it’s mostly been reported. Or as the equivalent of the primary schools budget or the defence budget. But it’s not the same kind of spending – it’s just a reallocation of the government’s assets so that we now own a few bits of banks. Is it at risk? Well it might be, but the risks are marginal. Is it a large number? The proper number to compare it to is the value of the government’s right to tax, which is about ten thousand billion pounds. Thirty-seven billion? We could do it again! But it was reported in some places as if people were wheeling barrows of cash outside the Treasury to put on number 37 red at the roulette table.”

Harmless nonsense
It’s lucky that Blastland can – most of the time at least – see the funny side of all this. But sometimes he doesn’t seem to know whether to laugh or cry, and it often has to do with research findings. “Of course there is some market research nonsense out there. Most of what falls into the category of nonsense is harmless, designed to get a laugh or a bit of attention as much as anything, and with, possibly, some truth in it somewhere if you don’t take the figures seriously. But then there is serious nonsense, where the subject needs treating and researching with care and rigorous independence of mind, with the methodology of the questionnaire properly reported, the interpretation of the data careful, measured and balanced and so on.”

One piece of “serious nonsense” that made him particularly angry came from a privately commissioned 2008 study on self-harm among girls, which, by failing to take into account response bias, came up with a much larger estimate of the extent of the problem than past surveys had produced. “One thing we know about self-harm is that people are more likely to do it if they think there’s more of it going on. So was any worthwhile data reported? I doubt it. Might it have encouraged other people to self-harm? Possibly. I think the whole exercise was appalling and I think the media was culpable for reporting it.”

The obvious question
Blastland says he is “baffled” by the failure of researchers and journalists to ask the obvious questions that would allow them to avoid this sort of mess. I suggest that this may be because asking these questions often leaves people who are desperate for an answer without one. “Yes, but there’s no point in having a bad one,” he says.

Through the book, radio show and numerous articles, Blastland is doing his bit to chip away at our bad habits. The clincher, he says, would be public embarrassment for those who peddle nonsense numbers. “It has to become embarrassing to say this sort of thing. If everybody smirks at this sort of casual political posturing, empty of any statistical meaning, then I think people will feel more inhibited about playing those tricks. One day it will be unforgiveable.”

The UK Statistics Authority has offered some hope, limiting the access that ministers have to official data before it is made public, and forcing apologies from the government for its spinning of figures on knife crime and immigration.

The silver lining
So despite his gripes Blastland remains optimistic. He points to the example of Norway’s national statistics authority, which compiles a compendium of data in the run-up to elections, providing a consistent basis for discussions on public issues with no prior access for the government. “It does wonderful things,” says Blastland. “You don’t get this nonsense about ‘record’ amounts of spending that aren’t adjusted for inflation.

“Here, I see government programmes where we decide to spend large amounts of money on the basis of some terribly flawed piece of evidence, or you see arguments about the level of spending on the NHS, and they make it impossible to have a balanced discussion because we haven’t got a number that’s comparable over time. The whole political conversation could be just that little bit more sober if we were more responsible in the way we used the data.”

Blastland is also a great admirer of the Gapminder Foundation, which promotes global development through better use and understanding of statistics. Gapminder’s freely available Trendalyzer tool, which animates statistical data to provide striking visualisations of global trends, was acquired in 2007 by Google, and is now being developed and integrated with the tech giant’s other information tools.

He rejects the idea that dealing with numbers is somehow inherently more difficult than, say, spelling or reading a map. “People understand that if you’re watching Who Wants To Be A Millionaire and the amount doubles, it makes a difference whether you’re starting from a fiver or £500,000,” he says. “People know that one month’s fall doesn’t indicate a trend in the retail habits of a nation. They know that you can have a big wave on a falling tide or a small wave on a rising tide.”

“I have to be optimistic that we can get better at this, because these are not difficult things, and there’s huge mileage.”

You know more than you think


Everyone recognises the folly of mistaking one big wave for a rising tide and, since we can do that, perhaps to our surprise, we can unravel arguments about whether speed cameras really save lives or cut accidents. In life, we would see – of course we would see – the way falling rice scatters and, because we can see it, we can also make simple sense of the numbers behind cancer clusters. We know the vibrancy of the colours of the rainbow and we know what we would lack if we combined them to form a bland white band in the sky. Knowing this can, as we will see, show us what an average can conceal and what it can illuminate – average income, for example. Many know from ready experience what it costs to buy childcare, and so they can know whether government spending on childcare is big or small. We are, each one of us, the obvious and ideal measure of the policies aimed at us.

From The Tiger That Isn’t by Michael Blastland and Andrew Dilnot, Profile Books