THE BLOG

A Truly Open General Election? Not Quite

05/05/2015 15:17 BST | Updated 01/05/2016 10:12 BST

The UK is in the top spot on the Web Foundation's Open Data Barometer, but how transparent are our elections? It turns out, not very.

The 2015 general election is just days away but we're still struggling to find out how many people are registered, where they can vote and who is standing for election. Not to mention gathering details afterwards on how votes were cast. The UK has some great online democracy platforms from organisations like MySociety and Democracy Club, which allow us greater reach into the workings of parliament and its data. But shouldn't election information come from local authorities, who oversee elections?

Surely we should know:

Who votes where - During the local elections in 2013, the ODI and Deloitte conducted a piece of joint research into what makes UK voters tick. We wanted to see what we could learn from the available data about how to reverse the long-running decline in voter turnout.

The research wasn't straight forward. We were set back by a lack of relevant data on voter behaviour at electoral ward level, as well as difficulties in matching what little data is available to other open data sources, such as demographic data from the 2011 Census. Even though individual ballot papers are collected and verified for counting the number of votes per candidate - the primary aim of elections, after all - the only recent elections for which aggregate turnout statistics have been published at ward level are the 2012 local council elections in England and Wales. In these elections, citizens from approximately 3,000 wards voted, out of a total of over 8,000. Data published by the Electoral Commission for the 2013 local council elections in England and Wales purports to be at ward level but is, in fact, for 'county electoral divisions'. Given these fundamental limitations, drawing any robust conclusions as to the types of people who were voting, how, and where, was difficult.

How many people have registered to vote. It has been reported that voters have dropped off the electoral register due to the change from household to individual voter registration, and the fact that universities can no longer block-register their students. There's been a last minute push to get citizens, especially students, to register. But how do we know whether the push has been successful? It seems like we won't know until after the election. Accurate and timely data could make a huge difference. For example by comparing numbers on the electoral register with those in the census, or previous versions of the register, we could identify areas (and maybe demographics) who have dropped off, and do more targeted work to get people to sign up. It's too late for this election, but important for the future, especially if turnout continues to decline.

Where all the polling stations are in the UK. At the moment we rely on democracy apps which crowdsource data to tell us where polling stations are, such as Election Polling Places. Some councils are publishing polling station data, but its very piecemeal, and there are no consistent standards. The Local Open Data Incentive Scheme helped councils publish data in standard formats - including the planning data and the location of public toilets - a similar approach could work for polling stations. It's small amounts of data, but has practical uses for campaigners, the media, so they know how and where to post people, as well as for voters with disabilities who would value information on the accessibility of their local polling station.

Who is actually standing for election? Again this is something that currently has to be crowdsourced, rather than being provided by local authorities. In fact crowdsourced data from YourNextMP is currently powering Google election searches. It's a great project and encouraging to see the data being used like this, but it should really be available from official sources, especially considering all of the local elections that take place too.

A truly transparent election would have all of this information available as open data. We'd be able to understand better what affects voter participation, such as the impact travel time has on turnout, or the areas of the UK with the lowest voter registrations. We'd be able to create democracy apps, such as those giving more information about parliamentary candidates, far faster and with better coverage if we didn't have to rely on crowdsourcing. And we'd be able to get more insight into the kinds of campaigns and other interventions that lead to greater participation and make our democracy stronger.

Election-related data should be included in the Local Government Transparency Code which states which data local authorities should publish. We would like to see the Electoral Commission leading on both defining the format of data to be collected and on publishing aggregate data from across the country.

The 2015 elections are almost here. I hope that the 2020 elections allow UK citizens a far better insight to the democratic process.