Cross-posted from The Wardman Wire:

Whilst claims of huge increases in US electoral turnout this year have turned out to be myths, with turnout only rising by around 1% on 2004, the continued gradual improvement in turnout in British elections is going largely unremarked.

The improvement is not yet sufficient to cause rejoicing, but there are solid grounds for being cautiously optimistic, particularly as the next general election is likely to be the first since 1992 in which the outcome is in real doubt before the votes are counted, which should give a further boost to turnout levels.

The evidence for improvement comes from both council elections and Parliamentary by-elections. The eight point increase in turnout in May’s London Mayoral election was promising, though that may have as much reflected the high profile personalities in this year’s contest as much as any underlying change in the level of public engagement with elections.

However, there have also been increases in turnout across the board in local elections. Given the different mix of councils and seats up for election each year, the best indication of broad turnout trends is to compare turnout in each time of English council with that same type from four years previously, the types being Metropolitan Boroughs, Unitaries, Shire Districts and London Boroughs.

(Scotland and Wales are excluded from the following figures as their changing patterns of combining or not combining local elections with Assembly/Parliament elections distorts turnout trends.)

In 2006, there was a noticeable across the board increase in turnout: up 6% in London Boroughs compared with four years previously, up 4% in the Unitaries and up 3% in both the Shires and Mets.

In 2007, the picture was more mixed, with Mets and Shires both up 2% on four years previously whilst in the Unitaries it fell by 1%. (2008 figures not yet available). However, across all three types of councils turnout was firmly higher than in 2007 than in 1999: +9% in the Mets, +4% in the Unitaries and +2% in the Shires. Overall, then, the trend is in the right direction.

But what about general elections? 2005 saw a modest improvement on 2001 (+2%) and it is worth noting that on a like-for-like calculation (i.e. adjusting the British figures to calculate them on the same basis as the American figures, which are worked out as a percentage of the theoretically eligible population rather than as a percentage of the people on the electoral register/voter lists), turnout in Britain in 2005 was only around five percentage points lower than the much-lauded turnout in this year’s American Presidential election.

Parliamentary by-elections since 2005 give grounds for believing that turnout will rise again. Although it is a well-established pattern that turnout in Parliamentary by-elections is lower than at the previous general election, this fall in turnout has been getting smaller.

The clearest way to demonstrate the figures is to compare the turnout at each by-election with the turnout in that seat at the previous general election, producing a percentage score. For example, turnout in Ipswich was 57.0% at the 2001 general election, falling to 40.2% in the Ipswich by-election during the 2001-5 Parliament, giving a score of 70.5% (40.2/57.0).

Calculating these scores for all the by-elections in the 2001-5 Parliament and then in the current Parliament (excluding Haltemprice and Howden, given the special circumstances) gives the following result:

2001-2005
Ipswich 70.5%
Ogmore 60.5%
Brent East 69.7%
Birmingham Hodge Hill 79.1%
Leicester South 70.0%
Hartlepool 82.1%

2005-
Cheadle 79.2%
Livingston 66.0%
Dunfermline and West Fife 81.3%
Blaenau Gwent 78.2%
Bromley & Chislehurst 62.4%
Ealing Southall 76.1%
Sedgefield 67.2%
Crewe & Nantwich 95.7%
Henley 73.1%
Glasgow East 87.2%
Glenrothes 93.9%

The average across the two Parliaments has risen from 72.0% to 83.4%, with four out of the five scores of over 80% coming in the current Parliament. Again, it’s a sign of steady and promising progress – even if nothing to widely celebrate yet.

All in all, the evidence gives some grounds for optimism for the next general election, particularly if it is seen to be a close or uncertain contest, thereby encouraging people to participate. Perhaps the biggest unknown is whether at some point the media will switch from a “people aren’t voting” to “people are starting to vote again” narrative. This could become a self-fulfilling prophecy given the frequent powerful bandwagon effect where people change their behaviour based on what they think other people are doing. But even without this, there is welcome steady progress.