It was much commented that turnout in yesterday’s Scottish referendum was very high (at 85%), and some also reported that there had been a surge in voter registrations, with as many as 97% of eligible voters registered to vote. Of course if 97% of eligible voters registered then turnout wasn’t actually 85%, but 82.5% (85*.97 – still pretty impressive). In most countries and many cross-country studies we take the turnout as the number of voters/ number of registered voters.

Of course the surge in registration in Scotland implies that it was much lower. It was. In the UK the percentage of registrations is reported to have fallen consistently over time to 86% in 2010 – so any official turnout figure was progressively over-estimating actual turnout. As such the fall in turnout there is greater than is reported. As we frequently use official turnout figures (based on no. of voters/ no. of registered voters) comparisons across countries aren’t always valid – registration rates in one country may be much higher than in another because of the rules and efficiency of administration. For instance research I did some time ago showed that the number of people on the Irish register was larger than the voting eligible population – 107% of it in one year. The register was cleaned up, and it appeared that turnout went up. In 2011 we had an increased turnout which many commentators put down to the crisis, and citizens re-engagement.

But turnout is probably bobbing along as this IDEA data show. IDEA collect data and estimate turnout on the basis of voting age population – which is itself inaccurate as it excludes Irish-resident British citizens who are eligible to vote (there are about 100,000 of these). We can see the increase in turnout in 2011 was in fact a fall in turnout. As a result the many academic studies of turnout that use official figures are probably wrong.

This matters because we are intrinsically interested in citizen engagement in politics and because politicians tend to come up with lots of hair-brained schemes to increase turnout (one of the implied rationales for the introduction of electronic voting was to increase turnout among the young). We will soon discuss whether to allow 16-17 year-olds vote, and the decision will in part be based on past turnout. Politicians may feel the ‘democratic revolution’ in Ireland is at work and there’s noting to worry about. Turnout is on the rise after all. It’d be useful to have and use an accurate estimate of this.

Because registration in Ireland is done at the local authority level it means there is possibly great variation in the accuracy of the register in different authorities. It’s likely big authorities such as Dublin have resources to ensure there is someone who’s in charge of this full-time, whereas in Leitrim the person might just get to do it coming up to elections, and then go back to running minor roads for the rest of their time. This could be solved by a well-resourced and permanent electoral commission which could discover if these are in fact problems, and give time and energy to thinking seriously about the integrity of elections, turnout, citizen engagement in politics and effective ways to address these issues.