Today sees the release of new data from TNS that shows whose policies are preferred for Education and Health by a representative sample across Great Britain.
It's especially interesting to me because the poll used the data and format from Vote for Policies, a service I set up before the 2010 election, where participants compare policies taken directly from the manifestos of the main political parties - but without knowing which parties they belong to.
Vote for Policies users are a self-selecting survey group, so the results from a national poll were always going to be different. And while the outcome is certainly unexpected, this is exactly the kind of research that has been needed for a long time - not just because of the policy preferences it reveals, but also the impact the poll had on the participants' voting intention. That's where the real gold can be found.
But as I said, the results were pretty surprising so let's get those out of the way first...
The most preferred Education policies were those of the Liberal Democrats (23%) just ahead of UKIP (21%), followed by Green (15%), Labour and Conservatives (10% each). Other parties consisted of the SNP, Scottish Greens and Plaid Cymru (2%, 2% and 1% respectively).
The most preferred Health policies were UKIP's (28%), followed by Lib Dems (16%), Green (15%), Labour (14%) and Conservatives (11%). Other parties consisted of the SNP, Scottish Greens and Plaid Cymru (2%, 1% and 1% respectively).
Vote for Policies is a completely non-partisan service so I'm reluctant to interpret the data and its meaning for each party (not to mention the fact that I'm not qualified to do so). But whatever your thoughts about the results I think it's fair to say they are indeed surprising.
The area that I can comment on, and what to me is even more revealing, is how the participants responded to the survey. Specifically:
A third of the public (35%) agree they are now more likely to vote in the general election as a result of taking part and seeing their results.
This supports the 'policies matter' drum that I have been banging on about for the last 5 years. But it's not a concept unique to me. It's what the creators of the many brilliant voter advice services also know: that if you want to engage more voters, you just need to give them the information to make an informed choice. What is that information? Policies.
With the release of this research, the message couldn't be simpler, or more conclusive. If we want higher turnouts, and to know what we're actually voting for, we have to know what the policies are.
Half (50%) agree that they are surprised by the results.
If you have ever thought there was a disconnect between the party we think we align with and the party whose policies we actually prefer, this says you were right. What the data also reveals is that it doesn't matter which end of the political spectrum we (think we) are, there's a fairly equal chance we'll get a surprise.
The main take is, wherever we get our information about who to vote for - whether the media, our friends, family or guided by our own preconceptions - that information probably needs updating. That might not be easy for those who consider themselves life-long supporters of a particular party, but given how little we appear to know about what we're voting for, the incentive is there.
A fifth (20%) said they are now reconsidering who they might vote for.
So clearly not everyone who gets a surprise is going to change their vote off the back of it, but 20% is still enough to dramatically change the balance of power as it stands today. This proportion varies, with one quarter (25%) of those planning to vote for Labour at the beginning of the survey reconsidering who to vote for. This compares with 20% of UKIP supporters, 18% of Conservatives and 16% of Liberal Democrats.
And the big question is where will the changed votes go to? Seeing the results from the representative poll you may think that's a good thing, or you might be praying nobody else takes the Vote for Policies survey between now and polling day.
What can we learn?
For me this is further evidence that policies need to be 'mainstream' in the political discussion. And I hope the rest of the media will take heed and start supporting what the burgeoning digital democracy sector of non-profits and charities have been doing so effectively on their own - getting more people to the ballot box by helping them make an informed decision.
Where do we start? For me, the lowlight of the 2015 election campaign has been been how late the manifesto policies were announced. It's bewildering to see that political parties don't know what their policies are at any given time, but it also prevents us from getting informed and having useful discussions around them. Rather than allowing political parties to delay the release of their manifestos until they see fit - and clearly as part of their campaign tactics - manifestos should be released with the convenience of voters in mind. For me that would mean at least 2 months before polling day, with an agreed deadline.
I'd reluctantly describe myself as a campaigner, but this is one campaign I am prepared to fight for, and something I hope the rest of the media will come together and fight for too. It's the route to better advice services, more informed voters, and higher turnouts. Policies matter.
TNS Omnibus interviewed a representative sample of 1,199 adults in Great Britain between the 21st and 23rd April 2015. Interviews were conducted as online self-completion and results have been weighted to make them representative of the general population.
Full data tables and details on the methodology can be found at: www.tns-bmrb.co.uk.
Matt Chocqueel-Mangan is founder of Votes for Policies, an independent, not-for-profit service where users compare policies on topics such as education, health or the economy without knowing which party they belong to. The policies come direct from the political parties' manifestos, written in their own words. Vote for Policies is crowd funded by over 900 individuals and received match funding from the Joseph Rowntree Reform Trust.