Earlier this week the Electoral Commission published a new report, The completeness and accuracy of electoral registers in Great Britain, looking at how electoral registration is working in the UK.

Although it’s been widely covered, the coverage has been very similar – taking the top line figures from the report and covering press release without digging in to what the report really says. So if we venture in to the inner reaches of the report, what do we find?

The report is a very welcome piece of path-breaking research, based on in-depth local studies. Given the importance of registration, and the number of policy and organisational options available to politicians and council officials, gathering this sort of information is extremely useful.

An interim report was published in December (which we covered here) and this final report updates that with more evidence collected.

The use of in-depth local studies is a good move, but it immediately raises a caution about the quoting of figures as if they apply to the country as a whole. The report itself says, “the findings cannot be used to report on national rates of completeness and accuracy.”

However, the report went on to say, “Under-registration and inaccuracy are closely associated with the social groups most likely to move home. Across the seven case study areas in phase two (therefore excluding Knowsley), under-registration is notably higher than average among 17–24 year olds (56% not registered), private sector tenants (49%) and black and minority ethnic (BME) British residents (31%).”

As a result, the 56% has been widely quoted in the media as if it were a national figure, despite the report explicitly saying it isn’t. Take the BBC (“the Electoral Commission has released results that suggests 56% of 17 to 24-year-olds may not be registered to vote”) or the Evening Standard (“The Electoral Commission says that just 56 per cent of young people are registered to vote”). You wouldn’t guess from either of those that “the findings cannot be used to report on national rates”.

What’s more, despite the implicit negative tone of the media’s coverage, the report actually suggests there is good news on electoral registration overall with a long-term decline halted:

Evidence available from electoral statistics and surveys of levels of response to the annual canvass of electors suggests that there was a decline in registration levels from the late 1990s to 2006. The same evidence base suggests that the registers have stabilised since 2006 although it is likely that the completeness of the registers has declined since the last national estimate in 2000.

In addition, the return rate for electoral registration forms across the country, which dropped sharply in 1996-2003 and then declined a little further in 2004 has quickened its recovery: 2007 was up on 2004 and 2008 was up on 2007 by a larger margin. Though the figures are still below the 1996 ones, the trend is heading in the right direction and the figures are higher than in 2005.

Moreover, the figures in the report are based on data taken at one of the worst points in the year for electoral register accuracy.

There is a full update to the electoral register each year, with a new register published on 1 December. It then steadily deteriorates in accuracy through the next year. The register can get updated through the monthly rolling register updates, but people usually leave it until the full register is redone to update their records. If a general election is called, they can however then update their records and still get a vote at their new address.

Therefore, it is normal to see registration levels drop through the year and it isn’t necessarily a cause of worry. By doing their studies on very old registers (eight to ten months old in all the cases used to get the 56% figure and other similar ones), the Commission (and to be fair, they know this and the report makes it clear – if you get to page 16) produced figures which are much lower than if the evidence had been gathered on a new register. Depressing the figures further, the research was done when there was no election in the offing and so people did not have any particular incentive to use rolling registration to update their records.

In other words, the registration figures found are much lower than we’d expect either on a new register or for a general election.

What’s more, the reason for low levels of registration amongst young people in the local studies may have little to do with levels of interest in politics but more to do with mobility:

92% of people who have lived at their current address for five years or more are registered, compared to just 21% among those who have been at their present address for a year or less.

So is it registration or journalism we should be worried about?

One other thing this report tells us is something about the how journalism is works – or doesn’t work. It’s easy to sympathise with hard-pressed journalist taking story and data from reputable source and turning it into story without much questioning. But the data isn’t nearly as uncontroversial as the uniformity of media stories would suggest.

Are the figures for youth registration bad because they’re low, okay because of the time of year they were taken or good because a long-term decline has been halted? You can argue any of the three – and were these figures a matter of political controversy, we’d have had talking heads and quotes arguing the case on each side.

But because there isn’t a National Association for Electoral Registration and Turnout Optimists and there is no argument between the political parties on the statistics, the figures don’t get an external sceptical eye cast over them. Add to this the Electoral Commission’s need to emphasise the importance of people getting registered, which provides an incentive to stress the pessimistic in its figures, and we get just the bad news reported. The good news doesn’t get a look in.

The full report is below and if you need any help to register yourself, visit www.aboutmyvote.co.uk or call the Electoral Commission helpline on 0800 3280 280.

(UPDATE: The Evening Standard, one of the media outlets to get the figures wrong, has now corrected its report.)

Electoral Commission Report on Electoral Registration