Department rejects count errors in e-voting tests

16 Apr 2004

Evidence of inconsistencies between the paper ballots and electronic counts in pilot runs of e-voting in 2002 has emerged from reports into those tests. However the Department of the Environment, which is overseeing the full nationwide rollout of electronic voting this June, has categorically refuted these latest claims.

The information about e-voting anomalies came to light following the release of documents under the Freedom of Information Act to Joe McCarthy, a prominent opponent of the Government’s proposed e-voting plan.

The data is contained a Dublin County returning officer’s reports on the use of e-voting for the 2002 Dáil election and Nice II referendum. It has revealed anomalies in the results of the polls where the Nedap/Powervote electronic voting system was used. This is the same hardware and software combination that the Government has chosen for nationwide rollout of e-voting, scheduled for the elections and referendum this June.

In the case of the Dáil Election in the Dublin West constituency, the reports indicate the presiding officer recorded 29,272 votes as having been cast. The Nedap/Powervote modules recorded 29,988 votes as having being cast, a difference of 716 votes. This additional number of votes represents 2.4pc of those who voted in Dublin West.

In the results of the Dublin North election, in which Nora Owen lost her seat, the documents indicate that the presiding officers recorded 45,236 votes as having been cast. However the Nedap/Powervote ballot modules recorded only 43,942 votes. This difference of 1,294 votes represents 2.9pc of total poll.

The documents appear to show that it is not possible to reconcile the findings of the paper ballots against the electronic votes recorded. Speaking to, McCarthy said that the discrepancies may be attributable to one of four possible factors. “It could be human error, a mistake in the ballot modules, a mistake in the voting machine or a mistake in the counting machine. Nobody knows and nobody can say where the error occurred.”

He claimed that the report proves that the Department of Environment’s stance that the e-voting test process “is completely unproven”. Documentation about other internal tests conducted by the department has also turned up errors, McCarthy alleged. Mock elections, held using votes from Buncrana Urban District Council and Athy, also returned numerically inaccurate results.

McCarthy also called into question the software development methods used by the Dutch programmers to write the software for the e-voting machines.

Colm MacCárthaigh of Irish Citizens for Trustworthy E-voting called again for a paper audit trail that would act as proof of votes cast as the voter intended. He acknowledged that it was too late to add a paper trail to the current system in time for the 11 June elections. “From a pragmatic point of view, the only option for this June is to run the election with the old system. We’re pro-e-voting, we just would like to see a voter-verified audit trail.”

However the Department of the Environment has refuted McCarthy’s assertions. It has indicated that the discrepancies are only due to mistakes by presiding officers. These centre on differences between the number of permit tickets issued and the number of voters marked off the electoral register. These were subsequently reconciled by returning officers and they did not affect the votes, a spokesperson said. For the referendum later in 2002, these procedures were modified and none of these discrepancies arose, it is claimed.

A department official has also confirmed to that the e-voting machines will be able to handle the proposed citizenship referendum – only recently introduced and now due to take place on the same day as the local and European elections, not to mention the full nationwide introduction of e-voting. “All we have to do is activate the referendum along with the local and European elections. No testing is required because it’s already part of the program”.

By Gordon Smith