In mid-December 2013, the British Library did something wonderful.
“We have released over a million images onto Flickr Commons for anyone to use, remix and repurpose. These images were taken from the pages of 17th-, 18th- and 19th-century books digitised by Microsoft who then generously gifted the scanned images to us, allowing us to release them back into the Public Domain.
“The images themselves cover a startling mix of subjects: There are maps, geological diagrams, beautiful illustrations, comical satire, illuminated and decorative letters, colourful illustrations, landscapes, wall-paintings and so much more that even we are not aware of.”
In the weeks following the release, incredible images have surfaced on many of the blogs I follow. I was particularly taken with a set of odd hexagonal maps Ollie O’Brien posted on the Mapping London blog. One map in particular stood out to me.
This strange map comes from John Leighton’s 1885 book The Unification of London: The Need and the Remedy. O’Brien provides this concise overview:
“London is split up into neat hexagons, colour-coded according to their proximity to the centre of the metropolis (defined as St Paul’s Cathedral rather than the more normal Charing Cross.) […] It looks like John Leighton was proposing a wayfinding system for London based on each area’s “zone colour”. Lamp posts would be used, with one handle always pointing to the north to orientate people, and colours, numbers and letters to show the zone. Bus blinds would have multiple colours indicating the zones buses passed through, and taxis would use appropriately coloured lights to indicate where they were willing to go. In a way, the idea of a uniform signposting system across London, across multiple objects and devices, is kind like the London Legible project, only 110 years earlier.”
I chose to use Leighton’s London Indexed in 2-Mile Hexagonals for my Hindsight inspiration because I have a long-held fascination with the mathematical properties of hexagons. I started playing with possibilities, unsure quite which direction I was heading in, but I knew that I wanted to make a hexagonal map of New Zealand.
After a few false starts, I landed on the idea of exploring the question, “How might we use hexagons to make better election result maps?” One of my frustrations with media reporting of election results is that, even though all the electoral areas are roughly the same size in terms of population, the maps tend to draw the eye to sparsely populated rural regions and de-emphasise dense urban areas. This can be partially mitigated by including breakout inset maps, but these fracture the map into many parts and make visual comparisons between regions difficult. Consider, for example, this New Zealand Herald map of 2011 general electorate vote results. Large rural and wilderness areas, such as the West Coast and Kaikoura electorates, dominate. Small but densely populated places like Hamilton and Tauranga are almost invisible. This is problematic because it gives a misleading impression of how people voted.
In considering these maps, I recalled some cartograms — maps that warp space by substituting a thematic variable for conventional distance and area — that the cartographer Daniel Dorling produced in his 1996 book, A New Social Atlas of Britain. Specifically, I remembered that Dorling had built some of his cartograms by dividing England into a set of tessellated hexagons.
I sat down with scissors, paper and a hex grid from a 1977 edition of War of the Ring, and began building a hexagonal cartogram map of the New Zealand 2011 general electorate regions. After finding a satisfying layout that roughly preserves neighbour adjacency, I downloaded the electorate vote data from elections.org.nz and wrote a D3.js script to automate the creation of cartograms. I’ve made the final source code as a Github gist.
In the maps that follow, each hexagon represents one of the 63 general electorates. My first attempt simply coloured the electorates according to the colour of the winning candidate’s party:
- Blue for a National Party victory;
- Red for a winning Labour Party candidate; and
- Dark grey for electorates won by a third party. (John Banks won the the Epsom seat for Act New Zealand and Peter Dunne took Ohariu for United Future.)
Overlooking the absence of labels (we’ll get to those in a bit), I already find this map useful insofar as it enables me visually to compare the rural and urban seats on equal terms.
Next I shaded the hexagons according to the share of the vote the winning candidate received.
- > 60% of the vote results in a completely opaque hex.
- 50 – 60% of the vote and the hex is 75% opaque.
- 40 – 50% of the vote and the hex is 50% opaque.
- < 40% of the vote and the hex is 25% opaque.
I noticed that, although the electorates have roughly similar populations, they vary considerably in turnout rates and the number of votes cast. Of the general electorates, Wellington Central has the largest turnout count (39,816 from an electoral population of 58,799 voters) and Manurewa had the smallest (26,457 from an electoral population of 54,662). Consequently, I scaled the area of each hexagon according to the count of people who cast a valid vote in the 2011 general election. I think the area-scaling needs a bit more work, but I am happy with the general direction.
I’m unable to fit the entire labelled map into this blog’s format, however, you can play with the entire map here. I’ve broken New Zealand into three parts for the detailed view. First, the upper North Island. Note how strongly Auckland dominates the country’s population, holding 21 of the 63 general electorates. Also observe the low voter turnouts in the Labour strongholds of Manukau, Manurewa and Mangere.
The second map shows central and lower North Island electorates, from Waikato through to Wellington. As one might expect, the largest voter turnouts are in and around our capital city.
The South Island is a significantly larger landmass than the North Island, but it has a far lower population. This is apparent in the cartogram.
The full visualisation is available on bl.ocks.org. There are more things that I would like to do to improve the visualisation, including providing rollover interactions with detailed statistics, adding legends to explain the colour and size values, animating between years and applying a similar treatment to the party vote data. But I am happy with the experiment to date and surprised at where I got to from that initial strange hexagonal map inspiration image. I think this could turn into something genuinely useful.
… and I’m honoured to be the first Hindsight contributor!
CHRIS McDOWALL helps preserve and promote cultural heritage as Manager of DigitalNZ Systems at the National Library of New Zealand. Previously he worked as a researcher and cartographer at Landcare Research, and completed his PhD at the University of Auckland on the problem of representing vague, dynamic geographic phenomena. You can find him blogging, tweeting and otherwise making magic in various corners of the internet.