In a blog where I’m posing every post as a question, this was by far the biggest question of the week in the UK. In fact, probably the biggest political question of my lifetime. But this is a data visualisation blog, so there’s no way I’m answering that particular question.
The main point though, is that elections are a fantastic opportunity to visualise data. First, opinion poll after opinion poll allows us to track patterns in voting intentions, and then after elections large geographical datasets representing choices of political parties or yes/no decisions allow endless possibilities of maps and colour.
I’m a late starter to the practice of data visualisation, but have always been fascinated by what it’s possible to show. In 1983 in school, we used to check the latest opinion poll results each morning in the newspaper, then head to a chart on the wall to put a blue, red or yellow sticker in the right position for the day. I was just 11, I couldn’t vote, but I still wanted to know who was winning, who was catching up and who was catching up, and I used a three-coloured line chart on the wall to find out!
Fast forward to just one year ago the UK held a general election, at a point where I had only just started learning Tableau to practice data visualisation myself, learning what might be possible as I started the visualisations out there were fantastic, really piquing my interest in data visualisation once more. As results came in, you got an idea of the significance of every constancy’s vote from the numbers. But you really knew how important it was when it was visualised. What colour did it turn the map? Was the bar chart representing the majority larger or smaller than last year? In what direction did the vote “swing”? The 2015 election hooked me in – I’m not the most political person, but data and its visualisation gets me interested every time.
2015 is where the Guardian’s cartograms really came to prominence (click image below for full interactive version). The idea of representing each constituency with an area of the same size rather than its geographical area on a map might have given the country an unfamiliar shape, but it perfectly represented the share of seats among political parties.
The referendum this week had more great visualisations by teams of data journalists working hard all through the night. As the result unfolded and it became obvious that the overall result would be a Leave vote, visualisations were used to explain the differences – which regions and demographics made the difference?
The graphic below was produced by John Burn-Murdoch at the Financial times, and brilliantly shows that the presence of those with degrees was a key demographic in determining the Remain vote
The BBC showed a number of summary graphs, including the following, from a Lord Ashcroft Poll, showing the startling differences between age bands
And finally, geographical differences were striking. Simple choloropleth charts using divergent palettes show that London was strongly Remain, Scotland was Remain in every one of its councils and Northern Ireland was largely Remain, whereas the majority of England and Wales voted to Leave. One example of a great map was from the New York Times (below). The colours show at first glance the Remain and Leave areas, whereas the saturations tell the bigger story about greater levels of support for either side in certain areas.
Chances are if you’re reading or following this blog then I’m preaching to the converted – you’re a fan of data visualisation. But if not, then my point is that an important election is a real opportunity to delve into some of the visualisation output out there, or even to have a go yourself.
As for me, here’s my effort. Having focussed on so many of the great maps available, I decided there was nothing new I could add there. I wanted to show some of the differences within region by looking at every local council, while considering turnout. Click the image for interactive version.