Wednesday, 17 March 2010

Strange graph

As the General Election approaches, the parties are increasing the amount of not-quite-election-publicity sent out, to get as much as possible out before the campaign expenditure limits begin. So that explains the second leaflet this week from the local Lib Dems.

As is traditional for the Lib Dems in a close second place, it includes a bar graph encouraging voters not to vote for the third party (in this case, the Conservatives). They may be strongly in favour of PR but that doesn't (and shouldn't) stop them pointing out the obvious tactical votes under first-past-the-post.

As is also sadly traditional for UK politicians, and something that nationally the Lib Dems are especially infamous for, the bar graph has no respect for scale, and overestimates their share of the vote compared with a linear axis.

PartyVote (%)1Height of bar (mm)Ratio (mm/%)
Liberal Democrat401142.85

1 At the last general election

In an unusual break from tradition, however, they actually overstate the Conservative share more than their own. Using an average 2.74 mm/%, the Conservative bar should only be 24mm tall - so it's almost double the size it "should" be.

The net effect is that the gap between the Labour and Lib Dem votes is only 9mm, which would only make up 2 of the 9 percent of the Conservative vote. What I'm wondering is why? Surely for a "Conservatives can't win here" approach they'd want to make the size of that bar as small as possible.

I can think of four semi-plausible explanations:

  • Shrinking the bar to its "real" size would have made it difficult to actually fit the label into the bar. There's not a lot of spare space now, and it would look a bit unusual at the smaller font size required.
  • They want to overstate the level of Conservative support so that the remaining Con-LD swing voters who haven't already swung to the Lib Dems (and indeed Conservative core voters) feel that there's a bigger chance of them getting rid of Labour if they vote tactically.
  • They're aware of the usual "dodgy graph" criticism the Lib Dems get, and want to have at least reasonable doubt for a "it's not supposed to be to scale" argument.
  • They're all genuinely unaware of the concept of linear scale in the local Lib Dem office.

On the bright side, none of the last three Lib Dem leaflets I received repeats the ageism I complained about, so that's an improvement.