Erin Brockovich Numbers
Those of you who've seen Erin Brockovich may recall the scene near the beginning of the movie where George (played by Aaron Eckhart) asks Erin Brockovich (Julia Roberts) for her number and she replies saying “Which number do you want?”
George: How many numbers you got?
Erin Brockovich: Oh, I got numbers comin’ outta my ears. For instance: ten.
George: Ten?
Erin Brockovich: Yeah. That’s how many months old my baby girl is.
George: You got a little girl?
Erin Brockovich: Yeah. Yeah, sexy, huh? How ‘bout this for a number? Six. That’s how old my other daughter is, eight is the age of my son, two is how many times I’ve been married — and divorced; sixteen is the number of dollars I have in my bank account. 850–3943. That’s my phone number, and with all the numbers I gave you, I’m guessing zero is the number of times you’re gonna call it.
This is how the NHS does its unscheduled care numbers. Headliney bite-sized bullet point numbers. Random factoid-type numbers that describe how bad things are. Erin Brockovich numbers.
“You want my unscheduled care numbers? Which numbers do you want? I got unscheduled care numbers comin’ out my ears. How about ten? Ten is the number of people on trolleys in A&E who've already been waiting over six hours for a bed. Thirty-five is how many medical patients are currently boarding in surgical beds…”
Erin Brockovich numbers are fine if all we want to do is describe how bad things are. But they don't get us anywhere close to helping us solve the unscheduled care problem. Erin Brockovich numbers don't help us with cause-and-effect. This isn't to say that numbers don't help us. They do. They absolutely do. But not Erin Brockovich numbers. Not on their own anyway.
We need other numbers, too.
We need System 2 numbers.
System 2 is part of a nomenclature stolen from the Nobel prize winning psychologist Daniel Kahneman. He put forward the idea that human beings have two ways of thinking about things: System 1 thinking and System 2 thinking. 1 is fast; 2 is slow. System 1 is thinking that we don't have to think about. Reflexes. Fear, anxiety, joy. But System 1 also includes things we've learned but which are so deeply learnt that we no longer have to think about them. If you're asked what the answer to two plus two is, you answer four without having to work it out. System 1 thinking is easy. And System 1 is always switched on.
System 2 thinking, by contrast, is harder. And we have to actively switch it on when System 1 fails to provide us with a satisfactory answer to a problem. For example, most of us would need to switch System 2 on if we were asked to multiply thirty-seven by sixty-one. But it’s not just about long multiplication without a calculator. We also need System 2 for the wicked, messy problems.
System 2 thinking is hard work. We have to concentrate.
One of my most overused anecdotes in the training room over the last twelve months has been about how I only got as far as page 100 (there are about 400 pages altogether) when I tried reading Daniel Kahneman’s Thinking, Fast and Slow. Not because I was finding it hard work (it’s a beautifully written book that’s easy to read) but because I was finding it too depressing.
Kahneman tells us not only that System 2 thinking is hard work, but that human beings will do pretty much anything to avoid having to do it. We hate it that much.
And at this point in the book (and it’s only about page 50), I’ve already worked out that numbers, tables and graphs usually require System 2 thinking. In other words, I’ve already worked out that most — if not all — of a data analyst’s output is something that requires its recipients to concentrate. And concentration means hard work. And human beings will avoid that at pretty much whatever cost.
It gets worse.
By the time you get to about page 75 of Thinking, Fast and Slow you’ve discovered that not only will human beings avoid doing System 2 thinking whenever possible, but we will also fool ourselves into believing that there’s a System 1 solution to a problem even when there isn’t.
And it was that discovery that finished me. I had figured out in my own mind how this plays out in the NHS. There are these difficult, challenging, wicked problems that we’re all grappling with. The Unscheduled Care Problem is one of them. Deep down, in their heart of hearts, health service managers know that data has an important role to play in the understanding and solving of these problems. But because looking at data will require them to think hard (“Here comes the science bit — Concentrate!”), they blank the numbers out. System 2 thinking. Too difficult. And they rely on hunch, gut feel and anecdote instead. And when they do use numbers, they use System 1 numbers. Or, as I have now renamed them: Erin Brockovich numbers.
So that is the long-winded story of how, with a heavy heart, I abandoned Daniel Kahneman’s Thinking Fast and Slow at page 100. And it is the story of how I have learned to be very sceptical whenever I ask an NHS manager to describe their Unscheduled Care system using numbers and all I get back are Erin Brockovich numbers.
About eighteen months ago I read a blogpost by a researcher called Sarahjane Jones. She made the point that when you are trying to improve outcomes, you need to look at the data that describes the processes at least as carefully as you look at the outcome measures. It’s processes that generate the outcomes. So if you want to improve the outcomes in a system, you need to be able to measure and describe and understand the dysfunctional processes that are causing those poor outcomes.
But our bias in favour of System 1 thinking prevents us from analysing the all-important process indicators effectively. It’s our human instincts that are getting in the way.
Fixing this depressing state of affairs is going to be a long haul. But I want Julia Roberts and Aaron Eckhart to help us make a start. The next time you hear a health service manager describing their Unscheduled Care system by telling you how many breaches they've had, how many outliers they've got, or how many delayed discharges there are, I want you all to fix a look of insouciant disdain on your face, curl your lower lip and say:
“Pah! Erin Brockovich numbers! Is that all you've got?”
