.
/v3-uk/news/2008522/gla-presses-ahead-counting-despite-concerns
25 Sep 2009, Rosalie Marshall , V3
The Greater London Authority (GLA) is to press ahead with the tendering process for a firm to run its electronic voting system in the 2012 elections, despite ongoing concerns among regulators and civil liberties campaigners over the costs and security of e-counting.
In 2008, the Electoral Commission and Greater London Returning Officer voiced concerns about the e-counting arrangements for that year’s elections for the Mayor of London and the London Assembly, and asked the government to provide an analysis on whether the method made sense economically and democratically.
The GLA conducted the analysis itself and sent copies of the document to the Electoral Commission in June as well as civil liberties organisations such as the Open Rights Group (ORG) to review.
The Electoral Commission has said it will either endorse or reject the analysis by next Wednesday. The ORG, meanwhile, has criticised GLA plans to go ahead with the tendering process for e-counting without approval as undemocratic.
Since 2000, ballot papers for all three London elections and the 2004 European election have been counted electronically. In the 2008 election, discrepancies arose with the e-counting process used to vote the GLA in.
The Electoral Commission, which was set up by Parliament to ensure fair standards are adhered to in the electoral process, produced a report pointing out errors and concerns that need to be resolved if e-counting is to be used again.
“We understand that some scanner operators were recruited at short notice and had little to no experience of the electronic counting system before the day of the count itself,” the report said.
“We have significant concerns about the number and size of discrepancies between the numbers of ballot papers expected and the numbers verified as having being scanned,” it added.
The commission was also unhappy that the e-counting system “did not support the more detailed notes that might provide an audit trail of individual decisions or corrections to ballot paper account figures which is usual in a manual count”.
The GLA produced its cost-benefit analysis of e-counting three months ago and the commission said it will make its response available online on Wednesday.
The GLA’s report is heavily skewed in favour of e-counting. It said manual counting takes too long and the media would become impatient with such a lengthy process. The analysis also said because manual counting in the London elections has not been used before, there is more potential for errors.
However, the GLA did acknowledge that manual counting is significantly cheaper than e-counting. According to the GLA’s contract with technology vendor Indra, e-counting cost £5,159,018 in 2008. It is estimated a manual counting process would have cost £3,633,567.
London City Hall chief executive Leo Boland said on Wednesday that despite the additional costs to the tax payer of e-counting, and the fact that he had not had a final green light from the Electoral Commission, the GLA would begin the tendering process for firms to run the electronic process.
A spokeswoman for the GLA said, “Just because Boland said we would go ahead, it does not mean we are committed to e-counting. We will only go ahead if it makes economic sense. We might not get any successful tenders.”
However, the ORG said the GLA is being undemocratic in its rush to tender.
“The experience of 2008 showed that a lot of work must be done to ensure transparency, systems testing and analysis, and better verification of votes cast,” said ORG chief executive Jim Killock.
“Given the desire to charge ahead without proper analysis of its own cost benefit report, and analysis of its suppliers, ORG is not confident that the GLA properly understands the risks it is taking,” he said.
“Putting aside the risks from technical errors, failures or hacks, the GLA has completely failed to make the case that spending £1.5m more on the 2012 election is the best way to spend London taxpayers’ money,” Killock concluded.