1 year, 1 month ago
One of the big themes of this election campaign has been, somewhat surprisingly, bar charts. As a group of people who have been banging on about bad bar charts for a long time this has been pleasing to us, albeit in a bittersweet way, like when an obscure band you’ve loved for years suddenly gets used in a beer commercial.
Anyway here are two more bad bar charts we’ve seen.
The Gateshead Liberal Democrats have taken an innovative, but highly misleading, approach to trying to show that “it’s a 2 horse race” in Gateshead between Labour and the Lib Dems.
They’ve decided to use the results from the area’s recent Police and Crime Commissioner election, showing Labour on 41%, the Lib Dems on 27% and the Conservatives on 14%.
But using an election where less than 15% of the electorate turned out to vote is... not a very helpful way to gauge how people might vote in a general election. Turnout in Gateshead at the last general election was 65%. The role of police and crime commissioner is also quite different to that of MP: being responsible for policing in an area.
And on top of this, the graph is incomplete. As you might be able to tell from the percentages on view, there’s a missing candidate - an independent received 17.9% of the vote.
More relevant to the decision voters are making in Gateshead is the result of the last general election, where it was very much not a two-horse race between Labour and the Lib Dems. The Lib Dems got only 4% of the vote in Gateshead, while Labour got 65%, the Conservatives got 24%, UKIP got 5%, and the Green party got 1%.
The YouGov MRP model released yesterday currently has the Lib Dems coming fourth in Gateshead, on 8%, behind Labour, the Conservatives and the Brexit Party.
The affliction of misleading bar charts crosses the political spectrum. Via BBC journalist Joey D’Urso, we’ve also seen this amazing leaflet from the Labour Party candidate for Suffolk Coastal, Cameron Matthews.
It contains a non-annotated bar chart, which presumably is meant to show the results of the last general election vote in the constituency in 2017. It shows Labour receiving roughly 80% of the Conservatives’ vote share, and the Lib Dems receiving about 10%, with the message being “only Labour can beat the Tories”.
In reality, Labour was much further behind the Conservatives than chart suggests. It received 30.5% of the vote compared to the Conservatives’ 58.1%, and the Lib Dems’ 7%. That means Labour received just over 50% of the Conservative vote share—not 80%.
The thing that makes this stand out, though, is that Labour’s bar chart has also added an exciting new feature: a candy striped arrow on top of their vote share, labelled “2019”, which puts Labour above the Conservatives.
This is presumably meant to indicate Labour’s ambition of winning the seat in 2019. It’s not strictly misleading to do this, but it’s something we (and, we imagine, most voters) haven’t seen before. As a result it makes a confused chart even more confusing.
For the record, YouGov currently thinks Labour will lose Suffolk Coastal by roughly the same margin as last time.