Step 4(b)

The UK government worries about unemployment

Unemployment is generally agreed to be a good thing for governments to worry about.

• There is an argument that employed people are happier
• There is an argument that employed people will vote for you
• There is an argument that employed people pay more taxes

Either which way, if you are in government, unemployment is worth worrying about.

So the UK government applied a little improvement methodology

They had a go at using the improvement cycle. It went like this:

Step 1. They measured unemployment rates

Step 2. They changed some things they thought would make a difference

Step 3. They checked to see if they had made a difference

Did they reduce unemployment?

I don’t really know.

But what I can tell you is that they made a methodological mistake. They didn’t use the recommended approach.

The recommend approach is:

Step 4. Act on what you have learnt and try again

What the UK government did instead was:

Step 4(b). Change the way they measured the problem

In the 1980’s the UK government liked this approach so much that they changed the basis for calculating unemployment rates over 23 times.

No doubt somebody would argue that getting a “more accurate assessment of unemployment” helped refine policies and the intervention strategy.

Personally (cynically?) I guess that the only real change to employment was that a lot of statisticians found jobs.

A rule of thumb about Step 4(b)

It doesn’t work.

If you are spending time changing the metric so you can manage the message, then all you are doing is managing the message, not the situation.

Party Political Note

I have little doubt that the current incumbents of Number 10 Downing Street have also been seduced by step 4(b) on occasion (and the White House, 24 Sussex Drive and the Steward’s Lodge as well for that matter). But that is, of course, only conjecture on my part.

If you enjoyed this post click here for updates delivered to your inbox

Read another opinion

Image by The Prime Minister’s Office