UNISON General Secretary election & voodoo density figures - a response: It's all in the numbers‏

UNISONActive has posted before about the worrying uses of statistics by Paul Holmes and his supporters. The main gripe was about the fact the claim of 86% density was not set in any context and it was misleading to quote it isolation. In response Holmes' supporters said that the analysis was “pointlessly obscuring the clear evidence that Kirklees UNISON has a very impressive membership density in its lead employer."

One of the first things said in the original post was that no-one should accept the claims as they are without verifying the data. Apparently this was spot on. The 86% figure has been exposed as being false with a figure of 58% being the true figure. I'm not sure this was “pointlessly obscuring” anything. Given the figures have been exposed as being wrong you might think that would be it. But, I'm a numbers geek and I think there are three issues relating to Holmes' use of statistics that still need addressing, and that demonstrate why Holmes is not the right person to lead our union.

ISSUE 1
Holmes and his supporters have variously claimed his branch has “the highest membership density of a UNISON Branch“ and “Paul Holmes is branch secretary of Kirklees Local Government branch of UNISON - the branch with the highest proportion of members in UNISON in the country”. Setting aside the fact the 86% claim is patently erroneous the question arises how can Holmes' supporters claim that this is the best density of any UNISON branch unless they have the density figures for every UNISON branch in the country?

I think this is unlikely as detailed density data is not in the public domain. These unsubstantiated claims show an alarming lack of objectivity from someone who wants to lead the country's biggest public services union.

ISSUE 2
So now the 86% figure has been shown to be wrong at 58% other question that arises is how did Holmes and his supporters get this so wrong? It's one of two things. Either they set out deliberately to mislead. Or, they just don't know how to get the right data and calculate the density properly. Deceit or incompetence? You choose. Neither of these two are desirable traits in a trade union leader.

ISSUE 3
On his blog Holmes has a message of support from Bernadette Gallagher, one of the other of the small number of NEC members supporting him. The posting says “of the 33 women elected to the Regional seats on the NEC in the last elections, Gallagher polled the highest number of votes.” Sounds impressive. But again, Holmes quotes a figure in isolation without any context.

This is a theme picked up on in his original article and rightly so. This is something Holmes does with alarming regularity. Gallagher is elected by members in the North West region which I believe is the biggest region in the union. Some regions are half the size of the North West. It is not particularly impressive that she had the highest number of votes given it was the biggest electorate. But the way it has been quoted, without context, is clearly misleading.

And finally....
Some might call this the setting up of a straw man argument but the use of statistics and numbers is of vital importance in our union and wider society. We've seen enough of politicians who can't put the decimal point in the right place or misuse crime statistics. We really don't want our union to be doing the same thing.

We have a real battle in the coming years to fight off attacks on our public services and we will need to make the arguments against cuts with a credible economic and financial analysis. We cannot let the misleading and incompetent approach to numbers shown by Holmes and his supporters to be adopted by our union otherwise we will lose all credibility.

Nicolas Bourbaki
Why we're backing Dave Prentis . Dave Prentis election website