Bad graphs of the 2021 UK elections

Some rules can be bent. Others broken.

The United Kingdom held elections on Thursday 6th May. As volunteers count votes, we can look at some graphs in election literature and on social media.

One way to understand rules of graphical integrity is to see them bent and broken. This is a short article. It is not a complete list of all poor data visualisation during the election.

Show me. (GIF: Gfycat)

The elastic axis

One common fault with graphs in election leaflets are disproportionate bar charts.

Bar charts show their values through the length of the bars. When bar graphs do not start from zero, heights are disproportionate to the values.

If you do not stand, you do not stand a chance.

In the Chippenham Pewsham ward in 2017, there were only two candidates.

For the Wiltshire Council election, the Conservative candidate got 696 votes. The Liberal Democrat candidate had 740. The Conservative bar should be 96% as high as the winner. It is not. That gives the constituent an impression of a greater gap between candidates.

Canvassing returns

Another problem is the usage of canvassing returns. Candidates and campaigners knock on doors, asking people how they intend to vote. This is different to market research — the person asking has a plain stake in the answer.

In this Bristol Green Party leaflet, they extol how “over 2000 doors” give estimates of vote intentions. There are selection and responses biases answering the door here.

First, parties choose where they go canvassing. It is not a random selection of houses. Second, people may not answer that question if they dislike the questioner’s rosette. Third, there could be inaccuracies in the replies. Not everyone can be having a great response on the doorstep.

All those factors can lead to an overestimation of support through canvassing returns.

From the image, it is unclear what ward this is referring to. (Image: Facebook)

Models and uncertainty

There is a lot to think about here:

Polls alone do not give a seat estimate. That requires a model, transforming constituency and list vote intentions into seats.

There is uncertainty in both the surveys and the projecting model.

There is also the questionable claim of “momentum”. In the Savanta ComRes polling series, the Conservative constituency vote intention share was 25%. That estimate is the same as their poll from a fortnight ago. For the regional vote, the Conservative estimate of 23% is also the same as that earlier poll.

There have also been revisions to past independence referendum estimates. (Image: Savanta ComRes)

The momentum could be a mirage from sampling error.

Election literature offers interesting insight into how people make graphs and statistical claims.

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store