Monday 1 June 2020

Lies, damn lies and statistics

Benjamin Disraeli, apparently, didn’t use the phrase ‘Lies, damn lies and statistics’. Mark Twain said he did but we know that fiction writers aren’t welded to the truth. However, it was probably an observation that politicians know how to lie by quoting statistics at us.

Part of my routine, since schools stopped being for everyone, is that I watch the daily briefing and write down the numbers they give us in a little notebook. The journalist in the house rolls her eyes at me.
“You could just look them up on the government website.”
“But it helps me to write them down,” I say, “It gives me a feeling of control that I know I don’t really have and it’s something to do everyday at 5pm. A routine is important. It’s even better that weekends are 4pm because that marks them out as different.”

I have argued for a while that if the government  want to give us more confidence to get the economy going again and get back to normal, they might have to start reporting different figures. Getting excited that only 100 people died today isn’t going to cut it. They could report that 56.000 people have recovered, or that the chance of catching the virus has gone from 1 in 30 to 1 in a thousand. (I’m making these figures up because they don’t publish them. ) I know they have to change the statistic they tell us but this is not the way to do it. If anything, it has made me far more anxious than before.

Every day, they tell us how many people have now died from Coronavirus in the UK. Each day that figure goes up and the headline figure that they give us of daily deaths is the difference between that day’s number and the figure for the previous day. Even when they changed the reporting to include deaths in all settings on 28th April they only changed that for that day, telling us those that had died that day (765) and explaining that the extra 3811 were deaths from the 2nd of March until that date. After that the figures added up again.

Yesterday was different. The minister was so excited. Robert Jenrick stood up and said that the daily death toll was the lowest since records had begun.

I wrote all the numbers down and tried not to worry that there were still about 500 new hospital admissions every day. The Long Suffering Husband seemed pleased. Apparently I was wrong  and 111 deaths in one day is the figure that people are happy with. However, something bigger was nagging at me.
It took me a while to put my finger on it.

The problem is that I can do very simple maths. Yesterday’s cumulative death figure was 39 045 and the day before it was 38 489. The difference between those two numbers isn’t 111, it’s 556.
I went back and did the sum several times. I checked the government website to make sure I hadn’t written any of the numbers down wrong.

I mentioned it to my family. My son checked the numbers, my daughter explained that it will be to do with ‘reporting differences’. She told me that it had been very difficult for journalists not to confuse people with the figures because sometimes hospital trusts include the old deaths that weren’t reported before and sometimes they don’t. “It will be something to do with that,” she said.

So, the extra 545 deaths occurred before today?

They probably did but won’t there be deaths that happened that won’t be reported for days, weeks months ahead?

I had to check to see if this had happened before and I just hadn’t noticed because the numbers were similar.
These are the figures since the announcement of easing of lockdown measures.



You can see that they all work out, except the last one.
The trouble with changing the statistics to make people more confident is that you have to do it in a way that isn’t going to confirm that you are lying.

No comments:

Post a Comment