When he hasn’t been dressing his party workers up as nurses, the Lib Dem candidate in Cardiff North, John Dixon, has been making a rather unusual case to the local voters based on the supposed weakness of the local Labour vote. Check out these quotations from a recent election leaflet of his:
“With Labour and Plaid out of the race locally, only John Dixon and the Lib Dems can be trusted to stand up for people in our area!”
“With Labour now a spent force both locally and nationally, I believe I am the clear alternative here to Cameron’s Conservatives.”
“Remember, with Labour and Plaid out of the race here, only the Lib Dems can keep the Tories out!”
“Don’t forget. In our area, only the Liberal Democrats can stop Cameron’s Conservatives. Labour are a spent force — they don’t even have any councillors in Cardiff North!”
“The race to be Cardiff North’s next MP is set to be a close-run contest. Local Lib Dem campaigner, John Dixon, is providing a strong challenge to the Tories.”
And to top all that we get one of the Lib Dems’ customary bar charts:
Get the message? That’s just from one leaflet.
If you didn’t know better, you might be forgiven for thinking that Cardiff North is a Tory marginal seat where the Lib Dems are the main challengers and Labour are “out of the race” with no realistic prospect of winning.
Nothing could be further from the truth. Cardiff North is a Labour marginal under serious threat from the Tories. Labour have held the seat since 1997 when they took it from the Tories. John Dixon and the Lib Dems are lying to suggest otherwise.
But don’t take my word for it. Let’s look at the facts.
We all know that the Lib Dems are very fond of bar charts. I’m keen on them too. Unlike the Lib Dems, I like ones that are drawn to scale, accurately labelled and that show data relevant to the point I’m attempting to make.
First, let’s see the most recent general election result in Cardiff North:
We can see that Labour hold Cardiff North with a majority of 1,146 votes. It would take a 1.27% swing to the Tories to unseat Labour here. This seat is very marginal based on the 2005 general election results. The Lib Dems are in a distant third place with less than half the votes of Labour. Based on these results, the other parties could quite genuinely be described as “out of the race”. Catherine Taylor-Dawson standing for the Rainbow Dream Ticket polled just a single vote in 2005. Never was a deposit lost with such panache.
This bar chart looks very different from the Lib Dems’ one which, using a statistical airbrushing technique that would shame Stalin, eliminates the bar for Labour entirely and fails to show the very slender gap between Labour and the Tories. How did that happen? They cherry-picked their data not just from a convenient election but displayed it in a way that’s entirely irrelevant to how voting works in the general election.
In other words, John Dixon and the Lib Dems in Cardiff North have lied with statistics.
Here’s how they did it.
Based on recent elections, all the results show that the Tories and Labour are the two most popular parties in Cardiff North. We’ve seen the 2005 general election results above. Labour won with the Tories very close behind.
In 2007 there was an election for the Welsh Assembly. Usefully, the Welsh Assembly constituencies are the same areas as the constituencies for the UK Parliament at Westminster. Here’s how that election turned out in Cardiff North:
Once again we see the Tories and Labour as the two biggest parties by far. The Tories won this seat in the Welsh Assembly with a comfortable margin over Labour. The Lib Dems were in a distant third place. Obviously, these election results don’t help the Lib Dems’ case that they, and not Labour, are the main challengers to the Tories in Cardiff North. Which, of course, is why the Lib Dems don’t mention them.
In desperation, the Lib Dems turned to the local council elections in 2008. While Cardiff Council is a different body to the Westminster Parliament, the Lib Dems have chosen to use the council election voting patterns as a guide to the relative strength of the parties in the area. You can draw your own conclusions on the extent to which this is a valid exercise.
There are eight Cardiff Council wards in the same area as the Cardiff North Westminster constituency. They are:
- Gabalfa
- Heath
- Lisvane
- Llandaff North
- Llanishen
- Portprennau/Old St. Mellons
- Rhiwbina
- Whitchurch and Tongwynlais
Let’s have another look at their bar chart:
I’m going to label this Lib Dem effort as Chart A.
Given that I’m the kind of pedantic, dull chap that prefers their bar charts drawn to scale and with a bar for every party, I’ll redraw it properly. I’ll call this Chart B:
While my Chart B is based on exactly the same data as the Lib Dems’ Chart A, drawing it properly shows something very unusual that isn’t apparent in the Lib Dems’ own chart: Labour and Plaid Cymru don’t have any councillors in Cardiff North.
We can also see more clearly that when the correct scale is applied, the Lib Dems have fewer councillors relative to the Tories than their own bar chart would suggest. The Lib Dems’ bar chart shows them as having at least half as many councillors as the Tories. In fact, the Lib Dems have five councillors and the Tories have 13. Scale matters.
It’s also notable that there are three independent councillors. Should we infer from this that independents are more popular in Cardiff North than Labour? Hardly.
Now the Lib Dems aren’t actually trying to hide the fact that Labour don’t have any councillors here. In fact, they mention that point specifically in their leaflet to bolster their case that “Labour are a spent force locally”.
But how do we explain the anomaly between Labour having no councillors in Cardiff North, yet they hold the seat at Westminster and put in a strong second place in the Welsh Assembly elections?
That’s down to the first-past-the-post system used for the council elections. In the 2008 council elections, Labour were the second biggest party in the eight “Cardiff North” wards in terms of votes but got no councillors at all. They came in second place in six of the eight wards but won none of them. In the “winner takes all” first-past-the-post system, Labour’s strong showing across all eight wards counts for nothing as it’s not sufficiently concentrated in any ward to win them councillors.
Here is Chart C, the number of votes each party got in the eight Cardiff North wards in the 2008 council elections:
And now we’re back to the familiar story of Cardiff North: The Tories and Labour are the two biggest parties with the Lib Dems a distant third.
Chart C is the bar chart that the Lib Dems should have used but that they don’t want you to see. Unlike Charts A and B, it shows the relative strength of the vote for the parties in the Cardiff North area in a recent election. And unlike the Lib Dems’ chart, it correctly reflects the fact that the council ward boundaries within the Cardiff North area mean nothing whatsoever in a Westminster election. The number of councillors the parties have locally is an entirely inaccurate reflection of local voters’ preferences across the whole area. It is utterly disproportionate representation.
In Cardiff North, all the votes for each party across the whole constituency are added together to elect the MP. It makes no difference at all what the relative strength of the parties in each ward is. But that’s what the Lib Dems chose to show, because it’s the only set of figures that supports their entirely bogus case that Labour are weaker than the Lib Dems in Cardiff North.
Out of all the data on recent elections and all the ways of presenting that data, John Dixon and the Lib Dems in Cardiff North have chosen the one anomalous case that least represents local opinion as expressed at the ballot box and best represents something entirely untrue that they want you to believe.
That’s how to lie with statistics, Liberal Democrat style.
The Lib Dems have got form when it comes to this kind of thing. In last year’s European Parliament elections, Lib Dems across London used similar tactics to confuse voters into placing a tactical vote, even though the European elections are run under a proportional representation system in which tactical voting is not advisable unless you support one of the very smallest parties.
The Lib Dems were advocating voting tactically against both Labour and the Green Party on the entirely false claim that they “couldn’t win”. As it turned out, Labour polled much higher than the Lib Dems in London and the Greens got a single MEP, albeit on a lower vote, just like the Lib Dems. For a party that supports proportional representation and campaigns strongly against what it sees as the shortcomings of first-past-the-post, this was the most unbelievable hypocrisy.
Remember these bar charts whenever you hear the Lib Dems talk about how we need a new kind of politics. Well we certainly do — and one where the voters can at least expect the parties to be truthful about straightforward things like previous election results and how the electoral system works.
Thanks to the following organisations and people who helped me get the data behind this article, though I should make very clear that they didn’t know what I was working on and can in no way be associated with the content of this article, its arguments or its conclusions:
- The Straight Choice where I downloaded John Dixon’s election leaflet. You can add any leaflets you receive to this website so that all parties’ claims can be better scrutinised.
- Matthew Somerville (in a personal capacity) advised on administrative areas and boundary changes
- Cardiff Council who advised on wards and from whose website I downloaded election results
If you like this kind of thing, I recommend Darrell Huff’s classic book How to Lie with Statistics and the complementary volume by Mark Monmonier, How to Lie with Maps.
Tags: Cardiff, Cardiff North, elections, General Election 2010, John Dixon, Liberal Democrats, statistics
Nice to see JD is now a liberal.
At Cardiff Uni, he had slightly more authoritarian tendencies. The Students’ Union Executive, of which he was a prominent and senior member, censured the student newspaper for reprinting the cover to a Pixies album, then banned the editor from the building for using the headline ‘Bombay Mix’ over a review of one of Bollywood’s most prolific composers because he wasn’t from India. John was then installed as the Exec’s minder on the newspaper, with his approval required before publication.
Scratch a Liberal and all that…
Wonderful how those bar charts just keep coming. The Cardiff North Lib Dems have been doing this for ages. Rather to my amusement, they issued a very similar leaflet prior to the 2007 Assembly Election, urging voters to vote Lib Dem to keep the Tories out. This time, someone seems to have realised just how dubious the stats are, as witness the business of labeling the Independents (dissident Conservatives) as ‘others’. In fact, this whole sorry exercise would have been impossible if Rhiwbina’s councillors had not fallen out, as there would have been only Tory and Liberal Councillors.
Old Pal, Cardiff Students Union seem not to have changed.
I’ve misremembered the last bit I realise – it wasn’t John who did the vetting, but the leader of the Executive.
The newspaper – which despite the Executive vetting and Editor banning, went on to win the Guardian Media Award for student newspaper of the year – was ironically called ‘Free Word’ (in Welsh).