Response to London Elects’ Manual Count vs Electronic count cost benefit analysis

Open Rights Group

The Open Rights Group (ORG) welcomes the Greater London Returning Officer (GLRO) having taken up the recommendation by ORG and the Electoral Commission to examine the cost benefit of manual versus electronic counting of ballots.

The ultimate finding of the analysis, that manual counting is cheaper than electronic counting by some £1.5m, supports ORG’s analysis of elections elsewhere. However ORG believes that the analysis strives at nearly every stage to close the cost gap between manual and electronic counting. For example while the electronic counting option considers combined count venues (as used in 2008) the manual counting option does not. There are potential cost savings in venue fees, staffing and security by combining venues which could have been explored for manual counts.

Page 2 of the response suggests manual counts would be slower because verification, adjudication and counting would have to be separate processes, however they are also separate for electronic counts.

Page 3 makes what we believe to be a spurious claim that re-counts are more likely with manual counts due to the potential for human error. There is plenty of potential for human error in e-counts whether when writing software, configuring systems, operating scanners or using the adjudication software (as observed in 2008). Agents have felt less able to challenge e-counted results because they do not feel confident querying the technical systems used, not because they are convinced the results have always been correct.

On page 5 the report warns that manually-counted GLA elections would be a step into the unknown. Yet electoral administrators across London have to manually count an election nearly every year whether it is for the local council, Westminster or the European Parliament. They are experienced and highly able in delivering manually counted elections.

The report suggests that procuring a multi-election contract for e-counting could reduce its overall cost but again does not consider what savings could be made by procuring printing services, for example, over many years for manually counted elections. Contracting out elections, which it must be noted are a democratic process and not a government service, is being phased out in some countries. For example Norway, concerned by the excess control contractors can hold, are looking to move all elections back ‘in-house’ into government.

The report argues on page 8 that the media will be impatient for the result with a manual count and that observers will find it difficult to monitor progress of counting. These are absurd arguments. Elections are not run to satisfy the media. Furthermore there is significant experience amongst party agents of how to closely and effectively monitor manual counts which are far more open to scrutiny than their electronic equivalents. At the 2008 London counts, as reported by ORG, many felt completely unable to scrutinise what was happening to ballots and information screens were difficult to interpret.

Detailed response to conclusions on page 10
A. Manual counts are far more accepted and widely understood in UK political elections than any other method, so stakeholder acceptance does not seem to much of a challenge compared to the serious doubts expressed by all parties at the 2008 count.

B. There is no national programme of electoral modernisation. There have been pilots to trial technologies. But it is misleading to imply there is any kind of clear strategy or programme of modernisation and no such political decision has been taken. Ministers currently indicate no plans to pursue e-voting or e-counting.

C. Programming and operating e-counting systems are also human activities and so also prone to error. In fact systematic errors introduced could well be much greater with automated systems than manual ones.

D. Progress information during the 2008 count was nearly useless to party agents so loss of these ‘big screens’ is not really a material factor.

E. It is a shame that additional levels of voting information would be lost from a manual count. We accept that is a disadvantage of the manual process as proposed.

F. CROs have to run counts for all European, General and Local elections so they are experienced and able to run London counts also.

G. The Electoral Commission is working hard to deliver standards to assist in providing consistent results, this is work which needs to be done for all elections regardless. Furthermore there is strength in this diversity — it would be very hard to trick or bribe a significant portion of counting assistants, one hack could alter the results from all e-counting scanners.

H. There is nothing inherently wrong with having to wait a bit longer for results to be declared. We were able to do it for the European results with no problems encountered for stakeholders.

The Open Rights Group believes that the very real financial costs of electronic counting and the risks of increased error or fraud are in no way balanced by the rather ephemeral benefits of potentially faster counts and more information on second preferences. In 2007 e-counting was not faster than comparable manually counted elections and we believe a manual count could well be completed in London more quickly than the analysis suggests.

As a result ORG urges the GLRO and Mayor of London to commit to manual counting for the 2012 London elections.