Mr No Vote wins 17/18 …

logo from scraperwiki.comA bunch of hacks (programmers) and hackers (journalists – real and citizen!) attended ScraperWiki‘s workshop in the University of Ulster Belfast campus on Saturday.

The idea was to unearth benefit and stories from the public data, if necessary using robotic data scraping techniques to lift facts and figures out of ill-formatted webpages into a more useful database format to explore the connections.

Three teams formed, looking at NI Court decisions, comparing pay levels across public, private and voluntary vacancies advertised on Jobfinder, and looking at NI electoral turnout and how this maps onto other social statistics.

Ivor Whitten’s written up a comprehensive post over on Hand of History. Our team’s hacker Rob Moore worked with Ivor, Matt Johnston and myself to scrape some sample data out of Ark and then reprocess it. If we’re honest, it’s been a while since any of us programmed in Python (or PHP) … but we’ll achieve more next time!

Map of NI showing 2010 Westminster election results - from Ark Election site

Given the levels of turnout at the May 2010 Westminster election, in 17 out of the 18 local constituencies the number of people not voting beat was greater than the winning candidate. Here’s what the map would look like if all those people had been forced to vote, but had chosen “None of the above” on the ballot paper.

Map of NI showing how turnout affected 2010 Westminster election results - adapted from Ark Election site

To repeat, not only does no candidate achieve a 50% mandate of the potential electorate (and few even achieve a 50% majority of those who did vote), but in nearly all constituencies the votes for the winning politician are outnumbered by the number of people who couldn’t be bothered to cast any vote.

I’d assumed that Gerry Adams’ large majority in West Belfast would be the constituency where he beat the stay at home voters. But no. It was in the split seat of Fermanagh and South Tyrone where both the main unionist and nationalist candidates just outpolled the no shows.

As Ivor asked:

… was the only motivator to get bigger numbers out to vote sectarianism? Looked like it as a riled divided community seemed determined not to let the other side in.

It’s not rocket science, and it’s not novel, but it’s still a stark statistic.

With a bit of work to extend the proof of concept – using some of the raw data from EONI too – we should be able to turn around interesting stats within hours of the May 2011 election results. And the data will be available to everyone else through ScraperWiki.

Map adapted (greyified!) from Conal Kelly’s maps over on the Ark election site.

, , ,

  • JAH

    Whilst the work is commendable, the point eludes me. Was there any seat contested anywhere in the UK in 2010 that the winner had 50% of the potential vote?

    Some of the Labour strongholds may return very high majorities but often also huge numbers of non voters.

    And so what? In a democracy every vote supposedly counts so that even if you win by one vote, that’s still sufficient.Only North Korea probably returns votes with huge majorities…

  • Hopefully I’ll get to the bottom of the pan-UK question at some point in the coming weeks. It would certainly be interesting to compare the margins of victory (or should that be the margin “below the non-voters”?) across Wales, Scotland and England.

    Business owners are often quick to point to strike ballots by unions, claiming that action has no mandate, that a majority of those who voted isn’t a majority of those who they were representing. I suspect that politicians have uttered those words too at times.

    Yet, what’s good for the goose should be good for the gander.

  • The 50% mandate is a bit of a red herring. The point being that only in one constituency did apathy get defeated. I would be interested in seeing some historical data for these elections as well as maybe an animation showing the changing vote percentages.

    If only a politician could appeal to the apathetic.

  • abucs

    But would they have chosen ‘none of the above’ if they were forced to vote?

  • JJ Malloy

    If only that was available! That would be a nice thing to petition and get on the ballot. Talk about a vote of ‘no confidence’

  • JJ Malloy


    Enjoyed the post over on hand of history. Thanks for the link.

    1.) Did you guys pull the numbers from all registered voters in a district, or just all those of the population eligible to register and vote? (Or do people even have to register ahead of time to be able to vote up there?…[ive been in the states for a good while now])
    2.) If people who were eligible HAD to vote, I wonder how the current percentages for each party would be effected. Which party’s supporters are really pumped up to go vote and which party has a lot of lukewarm supporters who stay home?
    I’m sure if there was a poll on a UI ‘Mr No Vote’ would not win.

    A few countries make voting compulsory (I believe Australia is one?). I don’t think it’s an idea without merit.

  • Mick Fealty

    It’s an interesting way of tipping the electoral system over in order to look underneath.

    You can say that it demonstrates sectarian head counts are (as yet) the only set piece plays that get the vote out. But it also demonstrates just how uncompetitive most constituency elections are under FPTP.

    It would be interesting to run the same experiment on the elections of the last 20 years to see if other patterns emerge.

  • dewi

    “It was in the split seat of Fermanagh and South Tyrone where both the main unionist and nationalist candidates just outpolled the no shows”

    Yep – but even there the turnout was the worst in world history

  • 1) From the total electorate registered to vote – ie, the number of people polling cards were issued to for that election.

    2) If people who stay at home blinked and discovered that they’d been magically transported to a polling station, many of them would definitely be happy to vote for a particular party. But until they do, Mr No Vote gets their name1

  • That’s the plan … an animated GIF of greyness spreading across NI perhaps?

  • This is very interesting, but please attribute the map you have used to Conal Kelly; and you should have asked him or me before using it – it is not a public domain image and the copyright is his. Just because it is on a website doesn’t mean you can copy it at will.

  • It’s already attributed – see the italicised comment at the bottom of the post. I’ll contract you offline too.

  • Brian Walker

    How would the same technqiue translate to Great Britain? A Commons with 2 or 3 three members? Or Scotland Wales? France? US?

  • OK, noted – sorry I didn’t spot it at first.

  • Excel and the Electoral Commission’s spreadsheets to the rescue.

    For Westminster election in May 2010, No Vote wins …
    17/18 Northern Ireland constituencies
    37/40 Wales
    48/59 Scotland
    329/533 England.

    So winning candidates were much more likely to have out-polled the no voters in England than anywhere else.

    English turnout was highest at 66.2%, Wales 65.5%, Scotland 64.3%, and NI 58.1%

  • Forgot to say, that means we’d have had a House of Commons with 219 seats filled (204 of them English MPs)

    There would have been 173 Conservative MPs, 27 Liberal Democrat, 18 Labour and 1 Sinn Fein. Parliament would have been electing a new Speaker.