Even one death is too many? No it isn’t. It’s important to treat people like numbers

Some people, yesterday

The laziest political argument that exists – with the possible exception of “don’t you trust the British people?” – is “you’re treating people like numbers”.

I wrote a piece last week about Atos, the French IT firm that has overseen the shambolic new system for disability support. A stat was doing the rounds that 10,600 people had died within six weeks of being declared fit to work in 2011. That stat was false: none of them had “been declared fit”, and the 10,600 figure included everyone who had died within six weeks either side of their support payments coming to an end. At least some of those 10,600 would have died while receiving payments and then had the payments stopped, for obvious reasons. My own suspicion, for simple reasons of likelihood, is that the died-then-stopped-getting-paid figure hugely outweighed the stopped-getting-paid-then-happened-to-die-within-six-weeks figure.

And last month, I wrote another piece about deaths and public policy. Except that time it was the NHS. Jeremy Hunt had declared that there were 12,000 “unnecessary deaths” on the NHS every year: “the equivalent of a jumbo jet crashing out of the sky every fortnight”. Twelve thousand sounds like a lot, but, I pointed out, the researchers who carried out the study from which the number comes were surprised not that it was so high, but that it was so low: previous estimates had put it between 60,000 and 255,000 “deaths or serious disabilities” caused by mistakes on the NHS. The NHS sees an awful lot of people, millions and millions a year, many of whom are gravely ill, and no system will completely avoid all mistakes: there will be some number of “unnecessary” deaths. The question is how many we can reasonably expect.

After both pieces, I received a lot of the same sort of emails, comments and tweets: “Even one death is too many”; “you’re treating people like numbers”; “why are you quibbling about statistics when people are dying?” The novelist John Niven, who had written an otherwise reasonable piece which quoted the Atos stat, tweeted when it was pointed out that it was wrong: “Tell me how many deaths are acceptable.” Regarding the Atos piece, I got an awful lot of those comments from Left-wingers keen to criticise a Tory government; after the Hunt piece, they were mainly from Right-wingers complaining about the NHS. I’d be interested to know if each group felt equally strongly about the other stat.

I know it sounds callous to talk about getting the numbers right and asking is that really a large number when we’re talking about real human beings who died. But if you don’t ask these questions, if you don’t get the numbers right, then it makes things worse.

For example, let’s imagine that there really was an available statistic of how many people died in the six weeks after Atos declared them fit to work. Judging by the FOI report on Atos, which as mentioned in my previous blog post is completely impenetrable, it looks like about 700,000 people came off disability support for whatever reason in 2011. Let’s say that all of them were declared fit to work (they weren’t, but for the sake of argument). How many deaths would be “acceptable”? Would one be unacceptable?

Well, obviously not, because if you have a group of several hundred thousand people, you’d expect some of them to die in any given six-week period just from unforeseeable bad luck: accidents, heart attacks, stroke. According to the UK national mortality statistics, people between the ages of 25 and 34 have about a one in 1,800 chance of dying in a given year. So even if the 700,000 people coming off disability support were young adults in average health (which they wouldn’t all be), you would expect about 45 of them to die in a randomly selected six-week period.

In that context, if there really were only one death, it wouldn’t only be “acceptable” – it would be a statistical miracle. It would still be terrible for the relatives of the person who died, of course. But none the less, from a population health point of view, it would be absolutely remarkable.

And that’s why this matters. If 10,600 people really had died in the six weeks after being declared fit, we could reasonably say even from my very crude look above that the number is hugely above what we’d expect, and that something was terribly wrong with the Atos assessment. If 100 people died, then we could say that the number was probably inside “acceptable” boundaries. What you then do about our disability support system entirely depends on what the real number is. And at the moment we don’t know.

A similar thing applies to the Jeremy Hunt/NHS issue: if you don’t give a context for your “12,000 unnecessary deaths”, if you don’t say what level of success you can reasonably expect in a modern health service, you don’t know whether that means the NHS is doing well or doing badly.

If you want to build policy – if you want to make the NHS work well, or disability support fairer – you need to know the difference. If you say that “10,000 people died after being declared fit”, you’ve got it wrong. That’s fine, people do that. But if, after it is pointed out that you’ve got it wrong, you say “well you’re just quibbling about statistics”, then you are actually making things worse; you’re denying people the chance to work with accurate statistics, and making it harder to make good decisions about how to use the limited resources available for public services, whether the NHS or benefits or anything else. If you throw around big-sounding, scary-sounding numbers without context, you do the same thing.

None of this – I really want to stress this point – none of this is to say that Atos and the DWP are doing a good job. By all accounts they are not (not least because they seem to be incredibly wary about releasing the REAL numbers that would let us make informed decisions). Nor is it to say, regarding the Jeremy Hunt stuff, that the NHS is perfect. Plainly it is not. But complaining about false numbers or out-of-context numbers is not “quibbling”, and it is not saying that deaths are good things. It is saying that if you want to actually do any good in a complex world full of millions of people, you need numbers. And if those numbers are no good, you can’t do any good with them.

The problem here isn’t that we treat people like numbers. It’s treating the numbers you get from people badly. Mouthing pieties about “even one death is too many” might make you feel better, but it doesn’t help anyone.

Get the latest comment and analysis from the Telegraph

Read more from our news and politics bloggers


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s