In the recent election, some of the parties put charts into their literature. In this post I analyse their accuracy.

Last month in the UK we had a general election, this is where we elect an Member of Parliament to represent our little area (called a 'constituency'). During the election period, I received a large amount of A4 paper from each of the various parties. I threw out hundreds of them, but I still managed to find a representative sample of them lying in my hall.

A couple of them (1 2) are leaflets for the neighbouring constituency! These are a complete waste of time.

Many of the leaflets share an interesting feature, little bar graphs or pie charts.

This is a commonly used tactic of the Liberal Democrat Party, to depict themselves as the second candidate in the local area, as opposed to their national profile as the third party. Here are some leaflets from the Lib Dem candidate in my area, showing these pictures.

The bar charts are a not very subtle appeal to people who would otherwise vote for the Conservative party. The argument is churlish and not in the British sporting tradition of fair play, candidates should campaign on their own merits.

However, it is not inaccurate from a demographic viewpoint. My area mainly consists of white working class workers, public sector workers and a growing Asian population. None of these groups are disposed to vote Conservative, and in the vote, the Conservatives came a pathetic forth.

Not that is stopped the Conservative candidate making her own graph in her literature. I found it hilarious so I will zoom in and show it in its full glory. It is dodgy (even dishonest) on several levels.

Firstly, it is not even a general election result, it is showing a local council by-election. Councillors have an important role collecting rubbish and other local issues, but it is not the same thing at all.

Secondly, as a council by-election, it is a much smaller area than the general election consistency. So there is no way to tell how the larger area will vote based on the east end (Sparkbrook) of the consistency. This is especially the case because Sparkbrook, as the Balti capital of the world, is a majority Asian area, whereas the other parts of the constituency have a more dispersed ethnic population. Anyway, here are the results in that 2009 council by-election:

So we can see here that the Conservatives came third, not even getting a third of the votes of the winning candidate. Therefore the most dishonest feature of the graph is showing the change in vote without supplying the total vote. When armed with the figures above, we can see that the Conservative vote went up by a few dozen votes and that they started from a very low base indeed.

The dishonesty is confounded with the annotations on the chart. It says "Can't win here!" pointing to Green, Labour and Respect. Even if we ignore my caveats about this data being irrelevant for the general election; as you can see from the absolute numbers from the by-election result, Respect and Labour were the top two results, so they could in-fact win here.

This is further proved by looking at the actual vote last month:

So the Convervative chart looks extremely selective and misleading as Labour did in fact win the seat and Respect came second. Many media commentators were predicting a possible Respect victory.

The Respect party also had a chart. It is a well presented chart, showing the different results in 2005. It is the previous general election in 2005, so it is not a complete fiction like the Conservative chart.

The annotation is also much more positive, instead of "someone else can't win here", it is the rousing "She can do it!". Unlike the Conservative annotation, it was not a lie; Yaqoob had a real chance.

The Respect party was actually founded by George Galloway, a Scottish Catholic, however locally in the campaign, Respect was perceived as the Muslim party, which limited its appeal among the indigenous population, and probably cost Respect the seat. However, demographic changes are on Respect's side, so they may well get over the finishing line next time when many new Asian voters have come of age. Assuming that the party does not run out of steam causing Asian voters return to the Labour party.

The chart is not without a problem. It is not the correct constituency. It is actually a neighbouring constituency. This is perhaps unavoidable because the constituency that Yaqoob ran for in 2005 was abolished due to boundary changes, and this current constituency did not have a respect representative in 2005. However, I think the chart should have put a little note pointing out that the boundaries have changed and this is the nearest relevant result.

Lastly we are onto Labour's literature:

As you can see, no chart. (The independent candidate's leaflet does not have a chart either.) Instead, on the Labour leaflet, there are pictures of the incumbent meeting people and a record of how he voted. For the incumbent candidate, especially when his/her party is doing worse nationally than the candidate is locally, then is less point to a chart like the type above. However, an accurate and honest chart would help potential voters to understand the context behind any dodgy charts, such as the one in the Conservative literature.