The last few days have been busy ones for the Electoral Commission, with most of the headlines caught by their report into when election counts should take place (overnight or the next day):
The Electoral Commission has recommended general election counts should continue to be held overnight.
Before the 2010 election, a number of councils made plans to count votes the day after polling day.
But a campaign by MPs and others resulted in a change of the law requiring counts to start within four hours of the close of polls…
Chair of the Electoral Commission Jenny Watson said: “We are rightly proud of our democratic tradition of overnight change in government and the theatre of election night.
“Candidates and political parties, having campaigned hard, want results as soon as possible, and returning officers and their staff work hard to deliver accurate and timely results.
“But as more and more elections are combined, it’s right that decisions about timing of the count are made in the interests of voters.” (BBC)
As I’ve pointed out before, when it comes to counts being done inaccurately, the problems lie in poor systems and not overnight counts. Despite the apparently obvious inference that overnight counts mean tired people and therefore that is the cause of errors, the evidence is otherwise: bad systems cause the problems, and they do so whether it is overnight or the next day (as with the London 2008 counting problems with votes reported from more wards than exist in London).
You can read the Electoral Commission report into the timing of election counts in full here (and be truly stunned by my appearance on p.49).
Meanwhile, the Electoral Commission’s verdict on the May local elections in England was:
The analysis shows that the polls went smoothly with 89% of voters saying they were confident that they were well run.
But the report found that there were some areas for concern, particularly with regard to turnout and concerns over electoral fraud.
The turnout for the local elections was 31.1% compared to 35.1% at the nearest equivalent elections held in 2008. Turnout for the Mayoral referendums ranged from 36.5% in Bradford to 23.9% in Nottingham, where there were no local elections taking place on the same day…
Just over a third of people (35%) thought that electoral fraud had taken place at least ‘a little’ at the polls…
The supply of information was seen by voters as the primary reason for lack of confidence that the polls were well-run, especially in mayoral referendum areas.
That final statement may encourage the Electoral Commission to push the point further in future years about providing candidates with full election freepost services. However, the battle seems to have been lost for this November’s Police and Crime Commissioner elections and in the current financial climate extending it to local council elections is considered unthinkable. I say ‘considered’ because almost no-one calls for it, despite it being the obvious next step in the train of thought in many Electoral Commission statements.
The turnout level was down on 2008 which in turn was down on 2004. 2004 however was sharply up on 2000, and as a result the 2012 turnout was still higher than in 2000.
On postal voting, there are two figures which continue to be kept away from the headlines – one buried in the details of reports like this and one not calculated at all. The former is the proportion of postal votes which people return but which are then ruled out for a problem with them (possibly fraud, much more often human error and just occasionally due to deliberately spoiling of the paperwork):
For the local government elections, a total of 82,000 postal votes were reported by ROs as having been rejected as invalid or otherwise not forwarded to the count; 4.6% of all those returned.
The figure no-one even collates* is the number of postal ballots returned but not counted because they come back through the post too late. Even if that number is low, added to the former it suggests there is a significant risk to postal voters of their vote not counting. The law is, however, changing to at least improve the situation with the first figure.
Finally, the Electoral Commission also took a look at the London elections:
Jenny Watson, Chair of the Electoral Commission, said: ‘‘Problems encountered at the count must be addressed to prevent them occurring at future elections. We’ll be asking the Greater London Returning Officer to thoroughly review the evidence in our report to make sure this happens.”
The report also covers a number of allegations of electoral malpractice at the elections and the action taken to prevent and detect fraud before polling day.
Almost all the allegations are still under investigation and The Electoral Commission will report on these as soon as the Metropolitan Police investigations are concluded.
Given the repeated problems with London Assembly and Mayor election counts, it is far from clear that asking for another review like this will be sufficient to make the 2016 election counts a success.
The safeguards against serious machine failure, such as wrong configuration causing the votes for one candidate to be reported as for another, also continued to be weak:
In response to concerns about transparency in 2008, the GLRO [Greater London Returning Officer] decided that, in 2012, CROs [Constituency Returning Officers] would manually adjudicate a sample of ballot papers to ensure consistency with the verification totals given by the e-counting system. The GLRO asked each CRO to manually adjudicate one polling station ballot box and one postal vote ballot box. Observers reported that they were unaware of this process taking place, as there were no announcements made. One observer reported that “we overheard an agent complaining that he had been unaware that the manual verification of a box had taken place as there had been no announcements or any indication of where this might be happening”.
As with the council elections, put into a longer context the turnout figures look a little better – they were down on 2008, but 2008 saw a turnout surge. 2012′s turnout was higher than 2004.
London saw particular problems with Conservative Party produced postal voting application forms and a suggestion that the Conservative Party did not keep to the Code of Conduct on postal voting:
Certain aspects of the layout and design of these bespoke forms meant that they could not be recognised by electoral management software when scanned, and administrators had to manually enter postal voters’ details into their systems…
One ERO reported receiving applications delivered in a batch from the Conservative Party that had been dated long before they were delivered, suggesting that the applications had not been forwarded to the ERO without delay (as agreed by the main political parties represented in the UK Parliament in the 2012 Code of conduct for political parties, candidates, canvassers and campaigners on the handling of postal vote applications and postal ballot papers in England and Wales). Party officers responded to the concerns raised by EROs by suggesting that a combination of factors may have been responsible for any delays, including electors not necessarily posting their forms immediately after completion, the Royal Mail not always delivering promptly to the party offices and possible delays with couriers delivering forms to EROs.
* Accurately, that is – as the Form K data gathered is of very variable accuracy and completeness.