Lies, Damned lies and Statistics

As I write this, the latest unemployment figures have been released, indicating a reduction in overall unemployment, and specifically a reduction in youth unemployment.

I will return to these later, but it reminded me of something I came across in my research, which is that people tend to achieve according to whatever statistical measure you use. This means that if management see reports to indicate progress - however they understand that - those reports will tend to show good progress, because everyone want to show good progress. This is only a problem if the reports are not reflecting what management think they are.

So, if I am judged on the number of hours I work, I will probably work longer hours. That doesn't mean that I getting more work done, just that I am recording more hours. The fact that I could, possibly, achieve more working less hours is irrelevant because, according to those in charge, "achieving more" is the same as "working more hours".

This is not an issue as long as those in charge know that this is what they are measuring. Of course, as so many failed projects shows, this is not by any means always the case. So many failures are caused by this mismatch of managerial expectation and reality, because some people assume that information means one thing, when it doesn't.

So, back to the unemployment statistics. The thing is, Tory governments (especially, but not exclusively) have, for many years, taken these as the indications of the success of their policies. This means that they find a range of ways to reduce these figures - without necessarily meaning that people are in any better situations.

So moving people onto training schemes - some of which are good, others are not - keeps the figures down. Reducing the availability of benefits can also be used to reduce the figures. And how many of these people - especially given that the reduction is most significant in younger people - are in zero-hours contracts, or jobs paying less than a living wage?

This government has stated - wrongly - that people in work are better off than those not in work. That only applies if they are in reasonably paid work. Reducing the statistics while not actually helping people is deceptive. What I would like to see is how many people are existing on less than the living wage - that is far more of an indicator of how people are doing. I suspect the group of those working but on less than living wage (calculated on a weekly, received money rate, so zero-hours people can be based on how much work they actually get) is far higher now than it was a few years ago.

I do not want to dismiss the fact that some people who were unemployed are now in viable jobs. That is a good thing, and I want to see more of it. What I don't want to see is people being forced into unsuitable roles and low-paid jobs just to satisfy the politicians need to show certain statistics are down.

What is more, I know that the government will use reduced unemployment figures to "prove" that they are "doing a good job". That is, of course, the problem with statistics - they are presented as unarguable and unanswerable when they are anything but. They do not prove anything, except that the statistic being used has changed. It is academically poor to presume a changed statistic demonstrated something without looking at the whole picture.

So when the government claims that it is doing a good job because the unemployment statistics show this, remember - they don't. And they aren't.