New blog

All new content on my restarted blog is here

Thursday, July 3

Spanner in works

In their review of the use of vote counting machines in the London elections, Open Rights Group (ORG) have thrown a huge spanner in the works:

"There is insufficient evidence available to allow independent observers to state reliably whether the results declared for the May 2008 elections for the mayor of London and the London Assembly are an accurate representation of voters' intentions."
The company involved are Indra and The Register reported the very visible issues a while back from their observations at the poll count - not fraud but a lack of transparency.

As well, some of the issues sounded like those experienced during the 'electronic counting fiasco' at the Bedford poll count last year - counting blank ballots as valid votes, jams in the scanning machines and system freezes.

The margin of error ORG thinks there may be would affect some of the declarations for positions in the London Assembly - not the Mayoral vote.

One of the main issues raised by ORG was value for money: is it actually cheaper to do it all by hand? Their three recommendations are:
  • London Elects should perform a full cost-benefit analysis of e-counting against a properly costed manual count of similar scope.
  • There should be time for formal consultation at national and local levels before e-counting is used in an election.
  • There should be long lead-in times, at least 18 months, for the procurement and implementation of election technology.
More about this on Kable

Postscript: Ella Taylor-Smith posted to the UKIE-EDem list, reminding me of the other botched example of electronic voting and answering the point made by James Gilmour that "because the Scottish Parliamentary elections in 2007 were counted electronically, the Scotland Office has made voting data available at polling district level (subject to a minimum of 200 voters). This has never been done before in the UK. That information is of very great value to the political parties.".
"While the data may well be of value to political parties they don't pay for the elections, which, in the case of Scotland 2007, cost nearly £40 million compared to £17.5 million in 2003. Both were parliament and local government elections."
More on this from The Scotsman, though Ella notes that they "are not unbiased in this matter".


Post a Comment