Léim ar aghaidh chuig an bpríomhábhar
Gnáthamharc

COMMITTEE OF PUBLIC ACCOUNTS díospóireacht -
Thursday, 3 Feb 2005

Chapter 8.1 — Electronic Voting and Vote Counting.

Mr. N. Callan (Secretary General, Department of the Environment, Heritage and Local Government) called and examined.

We are here to discuss the 2003 annual report of the Comptroller and Auditor General and appropriation accounts, Vote 25 — Department of the Environment, Heritage and Local Government — chapter 8.1 on electronic voting and vote counting. The relevant correspondence dates 18 January 2005 from Mr. John Kelleher, finance officer in the Department of Environment, Heritage and Local Government, replying to the committee's letter of 19 November 2004 concerning an Exchequer contribution of €63,487 or £50,000 towards the cost of providing a car park at Trim Castle. That issue is up for discussion if members wish to question it. It is under correspondence with the main item.

Witnesses do not enjoy absolute privilege before this committee although members do. The attention of members and witnesses is drawn to the fact that as and from 2 August 1998, section 10 of the Committees of the Houses of the Oireachtas (Compellability, Privileges and Immunities of Witnesses) Act 1997 grants certain rights to persons identified in the course of the committee's proceedings. These rights include the right to give evidence; the right to produce or send documents to the committee; the right to appear before the committee, either in person or through a representative; the right to make a written and oral submission; the right to request the committee to direct the attendance of witnesses and the production of documents, and the right to cross-examine witnesses. For the most part, these rights may only be exercised with the consent of the committee. Persons invited before the committee are made aware of these rights and any persons identified in the course of the proceedings who are not present may have to be made aware of them and provided with a transcript of the relevant part of the committee's proceedings if the committee considers it appropriate in the interests of justice.

Notwithstanding this provision in the legislation, I remind members of the long-standing parliamentary practice to the effect that they should not comment on, criticise or make charges against a person outside the House, or an official, either by name or in such a way as to make him or her identifiable. Members are also reminded of the provisions of Standing Order 156 that the committee shall refrain from inquiring into the merits of a policy or policies of the Government or a Minister of the Government or the merits of the objectives of such policy or policies. I will open the chapter first and see how much progress we can make. I will subsequently open the Vote if time permits.

I ask Mr. Callan to introduce his officials.

Mr. Niall Callan

Accompanying me are Mr. John Kelleher, Department of the Environment, Heritage and Local Government finance officer, Mr. Maurice Coughlan, principal officer and head of the franchise section, Mr. David Walsh, also working in the franchise section, and Ms Maria Graham, head of the Department's housing policy section.

Will the officials from the Department of Finance introduce themselves?

Mr. Jimmy Doyle

I deal with Vote 25 in the public expenditure division of the Department of Finance.

Mr. Jim O’Farrell

I organise the management and training division.

Will Mr. Purcell introduce chapter 8.1? Chapter 8.1 of the report of the Comptroller and Auditor General reads:

Chapter 8 — Department of the Environment, Heritage and Local Government

8.1 Electronic Voting and Vote Counting

Background

The Department of the Environment, Heritage and Local Government is responsible for the various legislative codes dealing with the registration of electors and the conduct of elections and referendums. This involves an ongoing review of electoral law, the provision of information and advice to registration authorities, returning officers and the general public together with the publication of election results.

In the exercise of its responsibilities in this regard, the Department took the lead role in the planning, procurement, testing and deployment phases of implementing Government policy on electronic voting.

The scope of the electronic voting and vote counting project as postulated by the Department was to replace the existing manual voting and counting processes with a solution which met the following objectives at an acceptable cost

·Simple for public to use

·Easy for electoral staff to use

·Maintains integrity in the electoral process

·Count software which applies the count rules more accurately

·Improves efficiency of electoral administration

·Reliable and robust technology suitable for Irish electoral conditions

·Minimum maintenance pre-election, at election and post-election

·Provides early results after polling has concluded

·Supports a positive image of the country in the use of information technology.

The electoral system in Ireland, Proportional Representation by means of the single transferable vote (PR/STV), is in use in three other jurisdictions worldwide — Malta for national elections and limited use in Northern Ireland and Tasmania. None of these used electronic voting when it was first considered here in 1998.

Research undertaken by the Department indicates that electronic voting is in use in a number of countries — some to a limited extent — viz. Belgium, Brazil, Canada, Germany, India, the Netherlands, the USA and Venezuela.

Audit Objective

The audit objective was to examine

·The administrative planning for the introduction of electronic voting and vote counting, including adherence to capital appraisal guidelines and the robustness of the cost benefit analysis

·The scope and nature of the testing/piloting undertaken

·Management of the procurement of the hardware and software

·Costs associated with the introduction of electronic voting.

Schedule of Key Events

November 1998 — the Department of the Environment, Heritage and Local Government invited companies with products which would facilitate electronic voting and vote counting in PR/STV electoral systems to furnish details of these products.

May 1999 — the Local Elections (Disclosure of Donations and Expenditure) Act 1999, included a provision that ballot papers from the 1999 local election could be used for research into the use of electronic methods of vote recording and counting.

February 2000 — the Government approved, in principle the introduction of direct vote recording and the drafting of enabling legislation, with a view to introducing electronic voting and vote counting at the 2004 European and Local Elections.

June 2000 — tender notice issued for procurement of suitable hardware and software system.

December 2000 — the Government noted the proposal to begin testing the chosen system proposed by Nedap/Powervote.

August 2001 — Six voting machines were purchased for testing purposes.

November 2001 — The Electoral (Amendment) Act 2001 provided the statutory authority for the introduction of electronic voting and vote counting at a Dáil Election.

November 2001 — 600 voting machines were ordered for use in three constituencies at the 2002 General Election.

April 2002 — Statutory instruments approving the use of electronic voting at a Dáil Election in three constituencies — Dublin North, Dublin West and Meath were signed.

June 2002 — a further 400 voting machines were ordered for use in four more constituencies at the Second Nice Treaty Referendum in October 2002.

March 2003 — 6,000 more voting machines were ordered for use countrywide at Local and European Elections in 2004.

1 March 2004 — Commission on Electronic Voting established to report on the secrecy and accuracy of the chosen system for use in the June 2004 Elections.

29 April 2004 — Interim report of Commission could not recommend proceeding with the introduction of electronic voting for the June 2004 elections

Administrative Planning

Research

The Department of the Environment, Heritage and Local Government in November 1998 invited companies who had suitable products which would facilitate electronic voting and vote counting at elections, using the PR/STV electoral system, to furnish details of their products to the Department on or before 31 December 1998.

Five companies responded with products which were capable of facilitating direct vote recording and vote counting. One such product was viewed in operation at elections in Cologne (Germany) and in the Netherlands.

Departmental officials visited the UK, the Netherlands, Germany and the USA during their consideration of the project.

The Department indicated that it had taken steps to test historical data on a number of machines supplied for preliminary testing, demonstrate them in operation to the key players in the process including the main political parties and pilot use by 165 electors at a bye-election in 1999.

The Local Government Computer Services Board (LGCSB) and returning officers were involved with the Department at the testing and the tender preparation and evaluation stages of the project.

Business Planning

Department of Finance Guidelines

The Department of Finance produced in 1994 Guidelines for the Appraisal and Management of Capital Expenditure Proposals in the Public Sector to assist public sector managers dealing with capital projects. While it is not a detailed planning and cost control handbook, it sets out the main steps which should be followed in evaluating and managing capital expenditure projects, considers the major issues of principle involved, and describes the principal methods of appraisal. It reiterates the principle that the systematic appraisal and professional management of all capital projects helps to ensure that the best choices are made and that the best value for money is obtained.

It contends that it is not enough to be satisfied that investment is justified but that it is also necessary to ensure that it produces its planned benefits at minimum costs. Cost includes the ongoing current costs generated by the use of a capital asset, as well as the initial capital cost.

The Business Case

The business case for the project, which was set out in a Departmental report in December 1999, pointed to the desirability of providing a higher level of service for voters and the public in the conduct of polls and counts. Other positive considerations were the possibility of long-term savings and flexibility if alternatives to voting at polling stations were introduced in the future. In relation to costs and savings, the report gave broad indications only at that stage, as no actual tenders were available at that time.

The report indicated that the cost of holding an election was increasing — about €2.9 million (excluding free postage and returning officer postage) based on costs incurred at the 1997 general election. It was estimated that the cost of counting was approximately €1.4 million. More specifically the cost of the count in the Dublin area was €500,000 approximately and the cost of purchasing equipment for this area would be about €2.5 million. Payback on the expenditure would require 8 polls and perhaps 10 polls in areas outside Dublin, as there would still be expenditure on the count process.

It suggested that the cost of election counts would increase faster than inflation due to the tendency to hold elections on Fridays (and maybe on Saturdays and Sundays) and a demand for higher fees — some returning officers had indicated a resistance of some staff to do election work due to level of fees.

The report pointed to other savings expected to arise on equipment used at polling day e.g. polling booths, ballot boxes, stamping instruments, ballot papers and other stationery as well as the hidden cost of storage of equipment between elections, preparation, dispatch and return of equipment to the Government Supplies Agency before and after elections. In the context of costs, it concluded that, essentially, the main reason for introducing direct vote recording and electronic vote counting was not only long-term savings but the accuracy and speed of the vote counting process and perhaps flexibility if alternatives to voting at polling stations were introduced in the future.

In January 2000, the Department of Finance stated that the proposals did not adequately address a number of issues. It requested that more details should be supplied regarding the security and reliability of the system and the contingency arrangements necessary in the case of system or equipment failure. It also asked that more details should be provided on the costing of the system. The total cost of the system as a whole must be stated and maintenance costs included. The cost of contingency arrangements must be built into the overall costings.

Regarding costs, the Department undertook to provide detailed costings after tenders had been received and evaluated. In December 2001, following the tender competition, it advised the sanctioning authority that the system recommended for testing would cost approximately €31.7 million, excluding VAT.

Later costings were provided at stages when large scale purchasing of voting machines were being undertaken.

Savings

In February 2000, Departmental papers put the estimated cost in six Dublin Constituencies at approximately €1.3 million, while it estimated savings of about €160,000 per poll. Unquantified publicity and training costs would also be needed. Overall, it was anticipated that there would be staff and equipment savings of about €1.4 million per poll nationally at traditional count centres.

A Departmental submission prepared in November 2000 estimated that savings of €13.8 million would accrue over a twenty-year period (i.e. the projected life span of the voting machines) based on an assumption of the minimum of possible elections occurring in that period. More elections would result in greater savings. The figures assume that staff savings would be approximately 40% of costs arising on vote preparation and counting which in turn were estimated at one third of total election costs and were informed by data provided by returning officers. The submission proceeded to estimate a more realistic scenario for the incidence of elections and referendums over the period, and concluded that cost recovery would be achieved in twenty years on the basis that there would be about 25 polls over that period. The Department is of the view that this cost recovery projection was conservative on the basis of staffing costs increasing at 4% a year and points out that actual costs have in fact been rising at a significantly faster rate.

In its submissions of November 2001, June 2002 and October 2002, the Department simply indicated that savings on the conduct of each poll would arise due to reduced vote counting costs and other administrative efficiencies. No financial data on savings was offered.

The Department has, however, pointed out that the overall objectives of the project did not as a pre-requisite require it to be self-financing.

While it is acknowledged that the decision to move to electronic voting and counting was primarily influenced by factors other than cost, the project should have been subject to more rigorous cost/benefit analysis in view of the scale of the financial commitments involved.

Scope and Nature of testing/piloting

Equipment testing

The Department informed me that the system had been tested by returning officers, the Department and the LGCSB during the two years prior to its use in the May 2002 General Election. In addition, the voting machine had been tested by three international testing institutes, while the election management and count software had been tested by two independent bodies.

Details provided by the Department indicate that testing was carried out on its behalf between 2001 and 2004 in the following areas

·consultancy for general examination and assessment

·functionality testing (including the embedded software)

·testing and certification of Nedap voting machine for compliance with international safety standards

·architectural evaluation of election management software

·report on security of the voting machines in polling stations

·testing of Integrated Election Software (IES) count-rule software

·validation and source code review, including reviews of code amendments

All the testing and reports concerning the voting machine and software were positive. The Department considered that the system, albeit with different vote counting rules, had a proven track record of use in the Netherlands for over ten years and for a number of years in some German cities.

Prior to the pilots, the voting machine had been used — with sample ballot papers only — by voters at two Dáil bye-elections in order to gauge the ease of use of the machine by the public. The voting machine had also been demonstrated to political parties, returning officers and featured in radio and television programmes.

Pilot Implementation

The Department has informed me that given the Government's decision of 8 February 2000 to commence the use of electronic voting at the 2004 European and Local Elections, it was considered appropriate to pilot the system in three constituencies with different electorate profiles and sizes (a three-, four- and five-seat constituency). It was not an option to select part of a constituency as a pilot and retain the manual voting system in the rest, because of the statutory requirement to mix the ballot papers on a constituency-wide basis.

Against this background it was decided that there should be "live" use of the system at the May 2002 general election.

Evaluation of General Election Pilot May 2002

The Department provided me with a copy of its "Review of the Pilot Project — General Election 2002" prepared on 2 July 2002.

The review's summary indicates that the pilot project to test electronic voting and counting in 3 constituencies at the general election was successful. The voting and counting were completed correctly, results were available, at the latest, within 4 hours of the close of poll and the public reaction was very favourable. Approximately 470 voting machines, including spare machines, were used in the 3 constituencies. Few technical difficulties were reported on polling day.

It also made the following suggestions for improvement

·Modifications to the screen of the voting machine to facilitate voters

·Some modifications to the IES software to facilitate use by election staff

·A more co-ordinated approach to procedures at the count centres, particularly in relation to presentation of the count information and results to candidates and media

·Improved technical specifications for printing ballot papers including candidate photographs and political party logos.

Market researchers interviewed 1,207 people as they exited polling stations and 96% of those surveyed found the voting machine very easy to use or quite easy to use. Only 1% of the 1,207 voters surveyed mentioned a delay or queue. This percentage rose slightly among those voting during the evening time, but never higher than 5%.

Questions Posed in the Review

The Department's review of the General Election Pilot noted that the following matters needed consideration

(a)Is the Powervote system the most suitable having regard to cost, life span of the equipment, and changes in technology?

(b)If the answers at (a) are positive, is such a system preferable to paper ballots and manual counting of votes?

(c)Should the system be extended countrywide in 2 or more phases having regard to the time and staff resources available in Franchise Section?

(d)Study of changes in STV count rules to omit the arbitrary and random features of the present count rules.

The Department concluded that it had satisfactorily answered these questions.

The review went on to state that it would appear from reaction to the recounts at the election, that candidates, subject to changes in presentation of results, would also favour the use of electronic counting. Election staff in the 3 constituencies at the general election all expressed a preference for the use of the system in place of manual systems.

Extended Pilot — Nice ReferendumOctober 2002

The decision to extend the pilot to four more constituencies (with a further purchase of 400 voting machines) was taken on 5 June 2002 some three weeks after the 3 constituency pilot. I asked the Accounting Officer the extent to which the Department's assessment of the pilot was used to inform the business case for full implementation.

The Accounting Officer informed me that while the formal review document of the pilot became available in July, its substance was known within the Department when the submission to Government to extend the pilot was made. In light of this, it was decided that, given the objective to promote the use of the electronic system at the European and Local Elections in June 2004, it was appropriate to extend the use of the system to a further four constituencies at the October 2002 referendum.

He stated that the use of the system at a referendum is a less complicated process (voter preference choice is limited and counting is much simpler) and the extended use of the system in this context was intended to provide more information on the public use of the system for a different type of poll before confirming its use for the country as a whole. The use of the system in the extra four constituencies also provided for more efficient and cost-effective conduct of the referendum for the Returning Officer for County Dublin as he did not have to cope with both the manual system in four constituencies and the new electronic system in two constituencies for which he is the Returning Officer, as occurred at the 2002 general election.

The dominant consideration was that a further pilot for a different type of poll over a larger population base would better inform any decision on the countrywide use of the system.

Following the October pilot, the Government's objective to promote use of the electronic voting system on a nationwide basis for the June 2004 European and local elections was confirmed.

Commission on Electronic Voting

The Commission on Electronic Voting (CEV) was established in March 2004 and mandated to report by 1 May 2004 in relation to the application or non-application of the Nedap Powervote system to the June 2004 elections. The Commission's Interim Report was not able to endorse the use of the Nedap-Powervote system at the June polls. However, the Report made it clear that this conclusion was based not on any findings that the chosen electronic and counting system would not work but rather on the desirability of allowing more time for further testing and quality assurance by, or under the supervision of, the Commission.

The main reservations of the CEV Interim Report related not to the system hardware (i.e. the voting machines), which accounts for the great bulk of the public expenditure involved in the project, but to the late stage (relative to June 2004) at which software modifications were being carried out. In a set of recommendations for action, the Commission set out a programme of action which it required to be in position to satisfy itself as to the secrecy and accuracy of the system. The Department will cooperate fully with the Commission in relation to this work programme.

While the CEV's report concluded that it could not recommend the use of Electronic Voting and Counting for the June 2004 elections it was not unreasonable for the Department to proceed as it did on the basis of the results of its testing and pilot projects.

Management of the Hardware and Software Procurement

The Tender Process

The Department published a Pre-Information Notice (PIN) (a requirement for public contracts in excess of €6.34million) for the purchase of an electronic voting and counting system for use at statutory elections and referendums in the EU Journal on 1 April 2000. Request for tender documentation was prepared and the full notice for tenders was dispatched to the EU Journal on 23 June 2000.

The Tender document proposed the phased introduction of an electronic voting and counting system:

Phase 1 — 6 voting machines, including all necessary hardware and software for counting of votes on the machines and ongoing support for hardware and software during testing which was expected to take 6 to 9 months (October 2000 to June 2001).

Phase 2 — subject to satisfactory testing of the voting machines and count software and Government approval to the use of the equipment at an actual election, assume the purchase of 300 machines including software for vote counting, for pilot use at a Dáil Election in 2002 or at a referendum (if any in 2002 onwards). In the event that a general election is held prior to 2002, the pilot of the scheme will be held with the European and Local Elections in 2004 — in which case assume the purchase of 1,200 voting units in addition to the 300 units.

Phase 3 — if there is a poll at the presidential election in October 2004, assume the purchase of a further 300 voting units.

Phase 4 — the position was to be reviewed at the end of 2004 for the post 2004 period. If it was decided to extend the system to the whole country, the total requirements for the State would be approximately 8,000 voting units (including phases 1 to 3), including necessary count software. This also provided that the counting rules might be simplified as part of the extension of electronic voting and counting to the whole country.

In phase 1, the six test units would be purchased by the Department. The contract to be entered into with the successful tenderer would stipulate, provided it was decided to proceed with the project, that the units required in phases 2 to 4 would be purchased by returning officers.

Tender documentation was sought by 30 companies and 7 tenders were received by the 14 August 2000 deadline.

A project board consisting of Departmental staff, staff from the Local Government Computer Services Board and representatives from the Returning Officers carried out evaluations on the various tenders received.

Two tenders were excluded from the process at this stage, as they did not meet all the requirements. Product presentations were made by the 5 remaining firms in October 2000. Tendered prices ranged from €25.4 million to €60.9 million.

The Department recommended acceptance of the second lowest tender from Nedap-Powervote at a cost of €33.4 million (after inclusion of €1.9 million for battery backup). The results from the tender process were submitted to the Government Contracts Committee who gave approval in December 2000. Tender prices were to remain unaltered until the end of 2004.

Description of the recommended system

The Nedap/Powervote voting system consists of the following:

Voting Machine — this replaces the ballot box, ballot paper and polling booth.

Ballot Module — this is a bespoke cartridge which is programmed with the election and candidate details. The set-up of the election, including candidate details, is programmed on to the ballot module and it is loaded into the voting machine prior to the election. It stores the votes and when the polling station closes the ballot module is sent to the count centre and its votes are collated with the other votes in that constituency for counting purposes using the election PC.

Programming Reading Unit (PRU) links the ballot module with the election PC and enables the ballot module to be programmed before the election, and then reads back in the data from the Ballot module once the polls have been closed.

Election PCs specifically provided for use in the elections and have been security hardened and programmed with the election software and supporting applications to ensure secure use and access.

Integrated Election Software (IES) is the election management software which enables the returning officers to set up the poll, take in candidate nominations and prepare the ballots for the election. The election data is then formatted onto the ballot modules using the PRU and the software counts the votes for each electoral area once all the votes have been read in and reconciled.

Purchase for testing system

Delivery of the six test machines was not made until August 2001, some twelve months later than was originally intended in the initial project timetable.

Purchase for Pilot for General ElectionMay 2002

In November 2001 the Department, proposed the purchase of 600 voting machines for use in the General Election pilot in 2002 — no date had been set but the supplier had indicated a need for a 23-week delivery period. The proposal was made notwithstanding the fact that testing on the original six test machines had yet to be completed. The Department had expressed itself confident that the testing would be satisfactory.

The order was placed with Nedap/Powervote on 16 November 2001. The Department indicated that protracted negotiations were required before the contract for the purchase of the equipment could be signed on 19 December 2003.

Departmental documents indicated that, in February 2002, functional and reliability tests on the machines had been concluded satisfactorily, but that functional testing of the software was ongoing and was expected to be satisfactorily completed by mid-February and fully completed by March 2002.

The Department of Finance again raised the issues of the security and reliability of the system and the details of the contingency arrangements in case of system / equipment failure and voter confidentiality.

The pilot at the General Election indicated that a number of physical modifications were desirable to make the voting machines more user-friendly.

Purchase for Second Pilot — Nice Referendum

On 5 June 2002, approval was given to purchase an additional 400 voting machines and ancillary equipment for use in 4 further constituencies for the then anticipated second Nice Referendum. The recommendation to order this equipment was made notwithstanding the fact that the review of the General Election Pilot, then known to the Department suggested the need to make modifications to the voting machine. Departmental papers indicate that the cost of these modifications, if carried out to machines before their delivery would be in the order of €615 each, while returning already delivered machines to the factory for modification would cost about €2,300 each.

I asked the Accounting Officer why the modifications were not sought before the order for these machines was confirmed to avoid the extra charge of returning them for modification at a total cost of approximately €680,000 (400 machines at about €1,685 each).

He stated that the proposed modifications to the voting machine panel involved the reduction in the number of rows per column to allow for better legibility of candidate details, and the use of brighter displays to make preference numbers clearer. The Referendum was scheduled for October 2002 and it was necessary to order the machines so that they would be available in sufficient time for the poll. While there was not sufficient time available to retro-fit the voting machines already purchased (and to re-test the equipment following the referendum), the nature of a referendum (where a simple "Yes/No" choice is required) did not necessitate that these improvements be made at that time.

Purchase for the 2004 European and Local Elections

After the Nice Referendum in October 2002, the Department recommended the extension of electronic voting and vote counting countrywide for the forthcoming 2004 Local and European Elections. This would involve the purchase of about 6,000 more voting machines and ancillary equipment. The recommendation also outlined the need for three modifications identified after the pilot use at the General Election in May 2002

·Reduce the maximum number of candidates per column on the voting machine from 28 to 20 or 18

·The size and lighting of the preference number were to be reviewed

·An investigation would be made on the provision of a facility for visually impaired voters to vote independently.

These modifications would also have to be made to the existing 1,006 machines. The replacement of the voting screen with the improved preference number was described as a highly technical procedure which must be carried out under production conditions and the modified machines have to be thoroughly tested by the manufacturer and retested by the independent test institutes. This would require that the machines be returned to Holland. The modifications to the existing 1,006 machines would cost approximately €2.3 million.

Contract for purchase of 6,315 machines

The Department issued a letter of intent to purchase 6,000 machines to Powervote on 28 January 2003 subject to 18 conditions. On 4 March 2003, the Department wrote to the Department of Finance informing them that it was placing this order and pointed out that a condition of the order was a deposit of 20% of the cost on placing of the order. Two returning officers had been requested to apply for an advance of the funding for their constituencies for voting machines (rather than 28 small requests) to meet the advance payment requirements. The Department endorsed the requests for funding from the Returning Officers for Dublin (€5 million) and Mayo (€3 million). 6,315 machines were actually purchased.

While the procurement of hardware and software was carried out in line with proper procedures the decision to purchase the additional 400 voting machines for the Nice Referendum without the identified modifications ultimately resulted in avoidable costs of €680,000 being incurred.

Costs Associated with the Introduction of Electronic Voting

In response to my enquiries, the Accounting Officer provided the following details of the costs incurred. In a number of instances detailed costs could not be provided because the information relating to the items concerned were held by Returning Officers and would be accounted for by them in their statutory returns to the Department of Finance.

Table 1 Hardware and Software Costs

Comment

2001 Test Purchases

6 voting machines and ancillary equipment

39,708

Election management software and equipment

488,490

2002 General Election Pilot

600 voting machines

Precise costing not available (payments by returning officers)

Estimated €3.4m.

Election management software and equipment

2002 Referendum

Purchase of 400 voting machines for 2002 Referendum on Treaty of Nice

Precise costing not available (payments by returning officers)

Estimated €2.3m.

Election management software and equipment

Modifications carried out to 1,006 voting machines

Cost for each machine €2,393 excl. VAT+ carriage. Precise costing not available (payments by returning officers)

Estimated €2.9m.

2004 Local/European Elections

Purchase of 6,315 voting machines

6,315 voting machines purchased @ €4,508 plus VAT. Precise costing not available (payments by returning officers)

Estimated €34.5m.

Electoral management software and equipment

Dependent on size of electorate, subject to negotiation with developer

General

PCs for Departmental use

28,623

Returning Officers’ PCs

Costs not available

Total

€43,657,000 (est.)

Consultancy and Testing Costs

The Department engaged consultants and experts to assist in implementing the project as well as test the chosen systems and software. Costs incurred included tests for compliance with international safety standards, architectural evaluation of election management software, security of voting machines, IES count rule software, source code modification, compliance with EU emission and immunity standards, ergonomics and international certification. These costs are itemised under each of the purchase phases as follows:

Table 2

2001 System Testing Phase

129,877

2002 General Election Pilot

46,827

2002 Treaty of Nice Referendum

37,746

2004 Local/European Elections (2003 Expenditure)

135,941

2004 Local/European Elections (2004 Expenditure)

187,788

Total

538,179

Advertising and Promotional Costs

The Department carried out extensive promotional campaigns to inform and educate the electorate in advance of the use of electronic voting and vote counting in the 2002 General Election and 2002 Referendum in selected constituencies and in anticipation of its use nationwide for the 2004 Local and European Elections. The costs incurred were:

Table 3

€(incl. VAT)

€(incl. VAT)

2002 General Election Pilot

Advertising/promotional

263,047

2002 Referendum

Advertising/promotional

270,485

2004 Local and European Elections*

Advertising (Media)

1,307,874

Advertising (Production)

636,628

Print

245,046

Video

10,523

Website

40,257

Roadshow/Events

270,342

Call Centre

15,093

Research

55,055

PR costs

733,669

3,314,487

Total

3,848,019

* With the decision not to proceed with the use of the systems in June 2004 savings under this heading of approximately €1,130,000arose on the original tender costs.

Training Costs

The Accounting Officer has informed me that the cost of training personnel for polling and count duties for the 2004 Elections is not available as these costs are the responsibility of Returning Officers and will be included in their returns to the Department of Finance.

Other Costs and Liabilities

I asked the Accounting Officer for details of any contingent liabilities arising out of the decision not to proceed with the deployment of electronic voting and vote counting for the elections in 2004. He provided the following information

·Batteries for voting machines — probable liability €75,000

·Voting machines ordered but not delivered (185) — maximum liability of €1 million if it is not possible to cancel the order

·6,000 extra ballot modules were ordered for use at the Presidential Election as the existing modules had they been used in the Local/European Elections would not be available for legal reasons — Maximum liability of the order of €690,000

The Accounting Officer also informed me that there are no annual maintenance costs required for the equipment involved. Annual storage and insurance costs are not yet available. He has sought details from the individual Returning Officers.

He also pointed out that although the Department is not directly responsible for the acquisition, maintenance or storage of ballot boxes, he understood that Returning Officers had purchased between 1,500 and 2,000 at an estimated unit cost of €50.

The cost to date of Electronic Voting and Counting is in the range €47.5m — €50m. The degree of value obtained for the vast bulk of this expenditure depends on future decisions on the use of Electronic Voting and Counting.

Mr. John Purcell

I will provide some background first. Proposals for the introduction of electronic voting and vote counting in Ireland first emerged in the late 1990s, culminating in an invitation to possible suppliers in November 1998 to furnish details of their products. Five companies responded with suitable products and over the next year or so, they were subjected to preliminary testing and use at pilot sites and for demonstration purposes.

In February 2000, the Government approved in principle the introduction of direct vote recording with a view to its use in the 2004 European and local elections. Following a competition held under EU procurement rules, the successful tenderer supplied six voting machines with all necessary hardware and software to the Department for testing at a cost of just over €500,000. The contract provided that the purchase of further machines was subject to satisfactory testing of the machines and count software supplied and Government approval for the use of the equipment at an election.

After the hardware and software were found fit for the purpose, it was decided to go ahead with the project and the covering statutory authority was conferred by the passing of the Electoral (Amendment) Act 2001 in November of that year. Six hundred voting machines and ancillary hardware and software were purchased for use in three constituencies in the 2002 general election. There was general satisfaction with the operation of that pilot scheme and it was decided to use the electronic system for the second Nice treaty referendum in four more constituencies in October 2002. A further 400 machines were purchased for use at the referendum. Again things went smoothly with the result that it was full steam ahead for the nationwide implementation of electronic voting and vote counting. To this end, 6,000 more voting machines were ordered in March 2003 for use in the European and local elections in 2004.

Towards the end of 2003 concerns began to surface about the security of the electronic system, ultimately giving rise to the establishment of the Commission on Electronic Voting in March 2004 to report on the secrecy and accuracy of the chosen system for use in the June 2004 elections. The commission's interim report advised against its use and consequently the Government decided not to go ahead with the system pending further testing.

The cost to date of procuring and testing the hardware and software, together with promotional and other costs, is in the region of €52 million. That is a potted history of the project that should provide some background to today's examination.

Going on to the substance of the chapter, it records the results of my examination of the electronic voting project from four perspectives — administrative planning, the testing and pilot schemes undertaken, management of the procurement and the cost. The report does not call into question the policy underlying the introduction of electronic voting, that is taken as a given. The implementation of that policy, particularly the financial dimension of the various elements, was examined.

My findings on each of the four strands of the examination are highlighted in the text but I will run through them briefly. My one serious gripe about the planning of the project was that there appeared to be a lack of rigour in the arrangements for costing the system and, by extension, it was difficult to come up with a clear picture of the likely cost benefit. I appreciate that achievement of cost savings was not the main objective of the project but good financial planning should always be a key element of effective project management.

The testing and piloting turned out to be the Achilles heel of the project. When the committee first raised the question with me in December 2003, I responded in a letter in which I stated that the integrity and security of the voting system is fundamental to the exercise of democracy. I went on to say that the importance of ensuring the integrity of the voting system dictates that the machines and their operating software should be subject to the most rigorous and comprehensive testing regime and that political and other considerations would suggest a strong case for independent validation of such testing before the machines are brought into use at elections.

Events as unfolded have tended to support that view. At the time, it was entirely a matter for the Department to decide on the scope and nature of the testing of this system. Within that context, I felt it was not unreasonable for the Department to proceed to full implementation on the basis of the results of the testing that had been carried out and the positive results from the two pilot runs.

In the third area, procurement, I found that the required procedures were complied with and that management generally of the procurement was sound, with the possible exception of the ordering and delivering of the extra 400 voting machines for the Nice referendum in the knowledge that they would later have to be returned for modification. If the machines could have been modified before delivery, costs of approximately €680,000 would have been avoided. I know the time scale was tight but the matter did not appear to be pursued with any vigour with the supplier.

Regarding the costs, these have mounted to €52 million to date. The late decision not to go ahead with electronic voting and vote counting in the European and local elections in June 2004 inevitably led to nugatory promotional and related costs totalling €3 million. Costs of €1.5 million were also incurred on hardware and software orders that were in train when the decision not to proceed was made. Both of these amounts are included in the overall computed costs of €52 million.

Clearly the degree of value for money obtained for the vast amount of the money expended to date depends on future decisions on the use of the system purchased.

I invite Mr. Callan to make an opening statement.

Mr. Callan

I thank the committee for allowing me to respond to the Comptroller and Auditor General's assessment of the electronic voting and counting project. My Department has lead responsibility, under the Minister, for policy, legislation and administrative arrangements around the electoral system. We operate in a complex organisational environment involving a range of other actors, including local authorities, returning officers, the Department of Finance, the Oireachtas, and a number of statutory, independent commissions.

The Comptroller and Auditor General's report acknowledges the various considerations behind the Government's decision to promote the major innovation of electronic voting. Central to these was the wish to improve the efficiency and accuracy of a system which had remained largely unchanged over the last century and to modernise the image of voting in a way that might encourage greater voter participation.

The Department was pleased to take up this mandate and we progressed the project actively and responsibly against the challenging real time deadlines involved. I note that the Comptroller's report, with the exception of the cost benefit issue — and I will return to this — would generally support the view that proper risk assessment of the chosen system of electronic voting and counting was carried out. The Comptroller and Auditor General's conclusion is that "it was not unreasonable for the Department to proceed as it did on the basis of the results of its testing and pilot projects".

As regards the risk assessment we undertook, the most important strength of the Nedap Powervote system was its proven record of reliability and successful operation in a number of European countries. Extensive hardware and software testing of the various components of the system was commissioned by the Department and completed. The electronic system was deployed in stages, from mock elections through parallel voting and counting exercises to pilot use of the system at the 2002 general election and the second Nice referendum. A user survey, which showed a high level of satisfaction with the system, was carried out in conjunction with the pilot projects in the selected constituencies.

Our risk assessment highlighted the need for extensive training of returning officers and their staffs. Significant resources were committed to this purpose and to public education on, and awareness of the new arrangements. The fall back of using the paper ballot system was maintained at all times. In the event, this had to be deployed following the decision on 30 April 2004 not to proceed with the electronic system for the June 2004 elections.

The risk assessment did not factor in the establishment of a commission on 1 March 2004 to provide independent reporting and advice on the secrecy, accuracy and testing of the chosen system. This arrangement can help improve the transparency and independence of the validation process. In the short term, however, it entailed time demands beyond those catered for in our implementation schedule for last year's elections. This was critical to the conclusion in the Commission on Electronic Voting's interim report that it could not recommend the use of the proposed system at the June 2004 elections.

The Comptroller and Auditor General's report highlighted two key points: the cost benefit analysis and the additional cost of extending the system to four more constituencies. While an additional formal analysis on the lines suggested in the report, might have been advantageous, payments for the hardware items originally tendered for have stayed in line with the prices quoted at that time. In addition, other cost elements of the overall project — such as testing and ancillary equipment — were carefully controlled as the project proceeded.

The report acknowledges that the decision to move to electronic voting was primarily influenced by factors other than cost, including the efficiency and accuracy of the system and its potential adaptability to further forms of electronic voting, as technologies develop in this area. Ultimately, the precise extent of the cost savings will be determined only through extended use of the system. The cost of running manual elections has been rising sharply, a trend that is not likely to be reversed. For example, costs incurred by returning officers at the May 2002 general election increased by some 71% over the comparable expenditure incurred at the June 1997 general election — from around €3.6 million to just over €6 million. The costs inherent in the present system must be taken into account in any overall evaluation of the project.

The purpose of extending use of electronic voting to four more constituencies at the second Nice referendum was to build on and increase public support for the project following the successful pilots at the May general election when 138,000 voters used the system. In October 2002, a total of 113,000 voters used the system in the same three constituencies as in May and a further 156,000 used the system in the four extra constituencies. In the event, results from all seven pilot constituencies were available by 11.30 p.m. on the night of the referendum, just two and a half hours after the close of polling.

It was not unreasonable to approach the project in this way, keeping the momentum going from the initial pilot and incrementally building public familiarity and support. Otherwise, there might have been no opportunity for more extended live use of the system until the European and local elections in June 2004, the date by which the Government had indicated the system would be used nationwide.

The Electoral (Amendment) Act 2004 has renewed the mandate to the Department to deploy the Nedap Powervote system. Approval of the state of readiness of the system is ultimately entrusted to the Minister for the Environment, Heritage and Local Government, as the responsible member of the Government who is accountable to the Oireachtas and can take advice from the Commission on Electronic Voting. The commission has commented positively on many elements of the electronic voting system, and confirmed that it accurately counted the votes recorded at the pilot tests during real elections.

It emphasised that its conclusion to recommend against use of the system in June 2004 was not based on any finding that the system will not work. It has, however, made clear its wish to oversee a more transparent and comprehensive validation of the software packages for the different forms and combinations of elections; and for improvement of several security features and protocols.

The Department will co-operate fully with the commission in this work. The Minister advised the Dáil on 9 December last that, in parallel with the commission's efforts, the Department, in consultation with Government, is developing a programme of further assessment, testing and validation which is intended to provide the reassurance called for in the commission's reports.

In retrospect, we recognise that the timescale for national roll-out of electronic voting may have limited our ability to consult as widely as we might have wished on the project and to provide fuller reassurance about the system. The debate has also raised interesting issues around standards of reasonable certainty or satisfaction prior to introducing innovations in public administration or service. Final adjudication of these matters is entrusted not to experts or professionals, whether inside or outside Government but properly to the political process.

My Department believes the chosen system of electronic voting and counting is fit for its purpose and that, in co-operation with the Commission on Electronic Voting, this can be demonstrated more transparently and comprehensively to the public and the Oireachtas. We will continue to work towards this outcome.

Thank you, Mr. Callan. May we publish your statement?

Mr. Callan

Yes, certainly.

The lead questioner this morning is Deputy Michael Smith followed by Deputy Boyle. As we have a high attendance of members this morning I will keep to the time allocated. I remind members that the protocol we agreed is 20 minutes for the lead questioner, 15 minutes for the second person and ten minutes for each subsequent member. There is no difficulty about getting in a second time but because so many have already expressed interest I will keep strictly to the time limits.

I welcome the delegations from the Department of the Environment, Heritage and Local Government and the Department of Finance to this important meeting. Two questions arise here: whether we should have electronic voting and counting, and if we are to have it whether this is the right package.

Changing from a system which had been tried and tested over such a long time was bound to involve some turmoil. People trusted that system. There is nothing more transparent and accountable than throwing the ballot papers on the table — if one uses a coloured pen it is possible to see one's vote there. Some people equate people's passion for politics with sport. Getting the results of a game when it is over is not the same as watching the player and seeing where the ball came from.

Those who have very little interest in politics watch the count till 5 a.m. One who does not fare well in the count has some time to adjust to that. Changing such a system is historic and would probably prove difficult, no matter how much ingenuity and time and effort is given to change successfully.

My first question is for the Comptroller and Auditor General. In view of the rising costs for manual election procedures, does he believe that there would be a considerable saving in the new system, assuming that it met the other criteria which have been required of it?

Mr. Purcell

That kind of thing was not analysed to the extent that it should have been, bearing in mind the amount of money we were committing to the new system. Listening to what the Accounting Officer said this morning about the rising costs of holding general elections under the traditional system, ultimately, if the system worked and worked well, it would begin to pay back further down the road. There is clearly a big up-front capital cost. Assuming everything went according to plan, there should have been a saving on the running costs at each election. If we keep this system for 20 years, which is a big if with the development of technology, we would see cost savings.

I supported the Department in using this proven technology. The argument can be made that one should go for a more up-to-date technology. Having some experience in science and technology, it seems to me that the first question is how proven is this technology. We cannot prove something over a short time. We need a longer time-span to prove something with any degree of certainty.

The countries Mr. Callan visited do not use the proportional representation single transferable vote system. It is virtually unique to this country. Some of us could say that we have barely survived under it a few times. Proving that the technology will work is not as simple as for single transferable votes or the list system. In Mr. Callan's efforts to prove this technology, what independent authority supported it?

Mr. Callan

The system's hardware is currently serving over 11 million citizens in sophisticated democracies in western Europe such as the Netherlands, parts of Germany and some municipalities in France. In the Netherlands, the Nedap Powervote system has been in operation for over ten years, dealing with that country's PR list system. It has penetration in that country of about 80%. We felt that track record was extremely valuable in recommending it for adoption in the Irish situation. It is right that we should have many technical tests on the component parts of the hardware system, but the proven experience of its operation in mature democracies gives comfort in regard to its deployment. If the hardware of the system was seriously flawed, we would have to believe that 11 million citizens are being sold short on their franchise. We do not believe that that is the case.

There is a different job to be done with the software designed to apply the indigenous Irish form of PR. That is critically important to proving and ensuring the suitability of the system in Irish conditions. This committee is concerned about the financial implications involved. We are only committed to about €600,000 worth of investment in that work, compared to a €40 million investment in the hardware. If more investment needs to be made to perfect and give greater reassurance about the software, that will be worth doing. In terms of the proportionality of the investments made in this project so far, it would be foolish not to go the extra mile to validate the software elements in the system.

We commissioned consultants to test the software elements of the system very extensively. An Irish software firm called Natheon Technologies undertook a full architecture and code review of the system and judged it fit for its purpose. We had many tests of the system run through a UK company called Electoral Reform Services. That company is a leading world expert on testing the quirks of different voting systems and it worked with us at all stages. The Commission on Electronic Voting judged that the system had properly counted the votes at the pilots. It also carried out various mock exercises which showed that the system counted votes properly. For an independent body set up in the timescale involved, it was understandably difficult that the software package specific to the June 2004 election was still being finalised at the same time that the commission was being asked to sign off on it.

The decision to add a referendum to the set of polls conducted in June 2004 was taken in early April 2004. While there was a referendum profile already in existence, it had to be bolted together with the local and European election profiles. There were some interconnections to be made. Therefore, we did not have the definitive version of the software that would have been used in June 2004. The commission made its own judgment on the situation in which it found itself, and in accordance with its terms of reference. It determined that there was not enough time for it to sign off on the suitability of the software package. In no way should anyone believe that durable and adequately perfected software cannot be put in place for a relatively modest investment to allow this system to work to high standards in Irish conditions.

What modifications have already been added to the software package, arising from the various concerns expressed? What is the potential for other modifications, should they be required? We have to bear in mind that it will all have to be independently validated. The public must be able to fully trust the system. I read somewhere in a newspaper that one of the councillors said he would rather die the old way than in the electric chair. Can Mr. Callan tell us the modifications that have been successfully made to that software package, arising from the concerns expressed during the spring?

Mr. Callan

In my opening remarks I mentioned the tight timetable and I mention it again in terms of the commission's inability to recommend positively on the introduction of the system in Spring 2004. The other side of that point is that given the time we now have it is a manageable task to improve the software, to bring it more transparently through various processes of validation, working with the Commission on Electronic Voting, and in that way to provide greater reassurance to the public and all interested parties regarding the fitness of the software system for its tasks.

I am not a software expert and many issues arise from the commission's recommendations that need to be addressed — security as well as software issues. We will be addressing these. I understand we will be bringing in new and well qualified expertise to deal with these issues. I emphasise that the perceived difficulties and concerns which arose regarding this system had more to do with the time available to us, the adequacy of being able to demonstrate the fitness of the system rather than with any inherent flaws in it. Even back in the spring, the various tests then run showed that the system was counting votes adequately.

There are two aspects to the controversy over the introduction of electronic voting. One relates to the technological side — whether there was faith and public confidence in the system being proposed and whether the appropriate hardware and software was in place. However, this committee is concerned with the public expenditure aspects of how decisions were made, what costs resulted and whether there was value for money at every stage of the process.

Can I confirm with Mr. Callan the current costs, and the likelihood of what they will increase to? My information is that as of last week, the cost of introducing electronic voting up to now is €52,118,184.72. This does not cover the ongoing costs. Is Mr. Callan confident that this is a fair representation of what electronic voting has cost the State to date?

Mr. Callan

My figure of €51,650,000 is not markedly different.

I am informed that my figure relates to the situation at the end of last week. Recalculations may need to be done. However, that figure does not include the ongoing cost of the storage of the hardware, the voting machines, and the insurance costs involved. The figure currently stands at €658,2229.

Mr. Callan

We accept that figure.

The figure will increase by at least that amount every year. Storage and insurance costs will be subject to inflation.

Mr. Callan

We calculate the present annual cost of storage in total at €658,000. That was included in my figure of €51,665,000 but I agree it is a figure that will increase.

Taking inflation into account over the next five years it is fair to say this cost will be in the region of €750,000. Insurance and storage costs will increase.

Mr. Callan

In many cases, authorities have signed up for five-year leases but I will not quarrel over these figures, given the difference.

Is there a likely further cost in the sense that the contracts signed with the voting machines provider include a maintenance contract that concludes in 2007, and that from that date onwards there will be an additional cost to the State of maintaining these machines?

Mr. Callan

That is so but relatively little maintenance is envisaged for the voting machines. I am not saying there will be no cost but we do not envisage heavy maintenance costs.

Has a figure been factored in as to what the cost is likely to be, or what percentage it will amount to of the current contract it now costs?

Mr. Callan

It comes free with the contract until 2007. We have not factored in any figure.

Can I suggest that a maintenance cost might involve a modest six-figure sum, perhaps €100,000 annually?

Mr. Callan

We will accept that as a hypothesis.

Could I put all those additional costs together and suggest that the annual increase in cost of electronic voting for as long as the present situation continues will be €1 million?

Mr. Callan

In rough and ready terms, that might be so.

That figure does not take into account other changes that may occur. Software and hardware has been acquired to deal with the Irish single transferable vote, proportional representation voting system. That system is not unique since it is also in place in Malta, but it does not seem to have been in place in any of the countries where there is expertise in electronic voting. The company involved is in Holland and the providers are in the United States. The system is almost unique so that there would have been difficulties in acquiring software to properly reflect our voting system. Was that an ongoing difficulty throughout this controversy?

Mr. Callan

No, the system was an adaptation. That was never seen as a particular difficulty. Wherever they operate, electronic voting systems have to be adapted to the local circumstances. There were no off-the-shelf products. We had not the benefit of our supplier or any other having previously developed a software package for our proportional representation voting system. There would be differences in counting systems all over the world which I imagine would necessitate a bespoke job being done in most cases.

Did the Department ever factor in the possible future change of voting systems for different elections? In the United Kingdom they use different voting systems for the national parliament, for regional assemblies and for the European elections. It is possible that in the future there might be proposals for a standardised system of voting for European elections. Is the Department confident that the machines bought and the software provided can be adapted for changes in different elections?

Mr. Callan

We are quite confident regarding the deployment of the machines. Clearly if a markedly different system of counting is introduced in this country or anywhere else, new software development would be required. If we are moving away from proportional representation it would be more than a question of adapting.

Regarding the Deputy's tentative figure of €1million annually, there are currently ongoing costs of storage under the present manual system.

They are borne by the local authorities rather than by the Department.

Mr. Callan

They are also borne by the OPW. The Government supplies agency incurs expenditure for maintenance of election storage facilities containing stationery, polling information and ballot boxes, which will continue to be used, and for printing ballot papers and so on.

There is a qualitative difference between storing a cubic box measuring two feet by one foot by two feet and a voting machine measuring six feet by four.

Mr. Callan

Indeed, but there are some costs too. I have some figures in front of me here, but I will not try the committee's patience with them, since they vary from year to year, and I am not sure how accurately computed they are. I am only making the point that there are some costs capable of being offset that would not arise if we went over completely to the electronic voting system.

There is a second area of questioning to which I would like to move, namely, the outside expertise that the Department brought in regarding electronic voting, on the technical and promotional sides. When the idea was first mooted, the Department had to seek advice from outside, since it did not have the expertise within it. We had no practical experience of electronic voting. How was that sought? Were tenders issued for people with expertise in the field? If so, who was chosen, and how?

Mr. Callan

There were tenders for some of the consultants whom we brought in. Others had such specialised expertise that we regarded them as uniquely qualified, going for them on a single-tender basis.

Is there a cost for that first category of technical consultants brought in on the technology itself?

Mr. Callan

We brought in the Nathean Technologies company, the Irish firm centrally involved in examining the software, on tender. With some of the other consultancies, such as the Electoral Reform Society from the UK, which has a unique reputation, we decided to take the single-tender route. PTB is the German science institute that tested the voting machine software, and TNO, the Dutch electronic products and services specialist, tested the voting machines and supporting equipment for compliance with international standards for environmental conditions. We used those bodies partly out of consideration for the fact that they had been involved in previous testing of the machines. Though our machines had other modifications, we thought that there could be some economies in using firms that were familiar with the field.

Can you put a price tag on that outside expertise needed for the technical aspects of electronic voting?

Mr. Callan

The sum of €620,000 covers all the testing to date.

Mr. Callan

Yes.

I find that interesting, since, in the second category of outside expertise brought in for the promotion of electronic voting, the Department was involved in making a contract available for advertising, information and public credibility regarding electronic voting to the value of €4.5 million. Is that correct?

Mr. Callan

Yes. The amount spent in the event on electronic voting in that contract and an earlier one was about €3.1 million, as the Comptroller and Auditor General has mentioned. In other words, parts of that awareness and education campaign went beyond electronic voting and were stated to do so. There was a generalised campaign too.

I find it rather strange that the outside expertise needed technically, because it did not exist within the Department, was so small a factor in comparison with what was brought in for promotion, which is something of which the Department should have practical in-house experience.

Mr. Callan

That amount of money did not principally go into the pockets of the consultants whom we hired. It went into radio, television and other media. There was real dissemination of messages through various Irish media.

We have a breakdown, but I am curious as to why the Department felt that it did not have the competence to do it alone. Surely that is the sort of ongoing work that it does anyway regarding elections and the franchise. Why was there a need to bring in outside expertise for this field? While I am on that theme, perhaps Mr. Callan might tell me the procedure followed to acquire those outside consultants, the selection procedure and who was involved in it.

Mr. Callan

I will start with the latter part of the question. The contract was advertised in the Official Journal of the European Communities in the middle of 2003, if I am not mistaken. We had received bids from 18 tenderers by the deadline of August 2003. We set up a tender evaluation board comprising four departmental officials chaired by the head of the franchise section, the predecessor of my colleague, Maurice Coughlan, together with the Department's communications consultant. That meant that there was a five-person board. In setting about its evaluation work, that board——

Were the communications consultants from outside the Department?

Mr. Callan

That is correct. It was Monica Leech Communications, together with four departmental officials. The exercise was headed and chaired by the——

Is it usual practice for someone from outside the Department to be involved in a selection procedure?

Mr. Callan

In this field, since the person was contracted to the Department specifically regarding strategic communications issues, it did not seem unreasonable that she should participate in this exercise.

Someone was hired by the Department regarding communications to help it choose other people to conduct the Department's communications?

Mr. Callan

That was obviously not the sole purpose of her employment, but as she was under contract to us working across the range of the Department's business and as this was a significant communications contract, there was seen to be merit in having her assistance in the process. However, she was only one of five people on the tender evaluation board.

Was the decision made by consensus? Was a recommendation made that had to go to the Minister or Mr. Callan for final approval?

Mr. Callan

I will go through the process, which had several stages. The tender evaluation board, at its own discretion, unanimously short-listed six companies from the 18 to be invited to make presentations of their proposals before the board. Following those presentations by six tenderers, there was a further unanimous decision by the board that two consortia should be further short-listed for interview before the board made its final recommendation. That was done, and the winning consortium was selected and recommended to the Minister at that point for award of tender. We believe that the selection process was correct and robust, with appropriate procedures followed. The relevant papers have been available to the Comptroller and Auditor General.

Is it a procedure that would have been followed in any other case regarding the award of contracts and tenders in the Department?

Mr. Callan

The structure of short-listing, evaluation and so on that I have described is commonly carried out. We obviously place frequent advertisements in the official journal, and all those things were done in this case. The outcome was as it transpired.

Has the use of an outside consultant working with departmental officials in a tender process been followed in any other situation?

Mr. Callan

The situation of having a communications consultant in the Department is by no means unique. On the other hand, there are periods during which we do not have that sort of service available. The Deputy is asking me to span a considerable period.

Has there been another example of it? Is it common practice within the Department? Will it be repeated?

Mr. Callan

The present consultant was involved in the assessment and awarding of three such communications contracts.

Was that the first instance of this?

Mr. Callan

No, there was a "race against waste" contract and there was one for assisting in publicity around the national inventory of architectural heritage. Both, I believe, were earlier than this.

What was the size of those contracts?

Mr. Callan

The "race against waste" contract was worth approximately €3.5 million. The national heritage contract was much smaller. I do not have the figure but it might have been €200,000 or €300,000.

Perhaps it was €8.5 million between the three contracts in which an outside consultant had an influence over how the contract was presented for final decision.

Mr. Callan

Influence is perhaps putting it strongly——

Either she had influence or she had not.

Mr. Callan

As is common in these cases consensus is reached between responsible people who are entrusted with these tasks. We feel that in all cases a proper recommendation from a process, also properly conducted, came before the Minister.

Was this the first time these three contracts were agreed in the Department and does the Secretary General believe that it is a process that can be replicated?

Mr. Callan

As I said, different circumstances arise under various Administrations, involving people either on a consultancy contract basis or on staffing contract, who come to supplement the Department's efforts as regards communications. Those types of situation are not unprecedented. There are numerous precedents. These matters, moreover, have been documented in the recent Quigley report, and commented on. We accept the commentary, recommendation and analysis of the Quigley report, as far as that goes.

The Deputy's time is up.

Not literally, I hope.

One question has crossed my mind, before I call on Deputy Fleming. As my constituency was not involved in any of the pilot schemes, I have not seen the system in operation. What staff are in the polling stations? Is a person assigned to each voting machine?

Mr. Callan

Yes.

Are there returning officers and clerks, as well authenticating, who can vote?

Mr. Callan

There is a presiding officer and a clerk. The clerk is the official who assists at the machine. Perhaps I am anticipating wrongly the thrust of the Chairman's question. There is no staff saving in the polling station situation. Staff savings arise at the count process.

I know that, but I am trying to establish whether there are additional staff in the polling station.

Mr. Callan

No.

Is there a cost increase, in numbers?

Mr. Callan

No, not in regard to how the system was deployed in the pilots.

I want to make a general observation, following up on the last point dealt with. I know of numerous examples in the public sector, the National Roads Authority, for example, where expenditure runs into hundreds of millions, where consultants commissioned to design a scheme, who are either employed by a Department of Government or directly, through a State agency, have a major influence in recommending the particular contractor who gets the job. In most of the large infrastructure projects funded by the Department of the Environment, Heritage and Local Government, such as water, sewage etc., it is quite normal for a local authority to appoint a client's representative as its consultant, if it did not have the expertise, itself. That individual would then have a major say and influence as regards the recommendation of the final contractor ultimately appointed to do a job. The point alluded to is standard practice and happens on an ongoing basis as regards large publicly funded contracts. Without getting into the details of the specific case, I want to emphasise that I see nothing unusual in this.

There is a vote in the House. I just want to indicate that Deputy Fleming is in possession and Deputy Rabbitte is next, followed by Deputy Michael D. Higgins.

Is the Chairman going to continue?

We are going through because Deputy Fleming will pair with me. If Deputy Rabbitte wants to remain, he is next on the list.

Might I be back?

I have only ten minutes.

Time will be allowed. Deputy Rabbitte will be next on the list when he returns.

I want to go back to the Secretary General's opening statement and concentrate in particular on the Commission on Electronic Voting. He says the commission has already commented positively on the many elements of the electronic voting system and confirmed that it accurately counted the votes recorded at the pilot test during real elections. In the seven constituencies where the system has been used so far, the Commission on Electronic Voting found that the votes were counted accurately. Is that correct?

Mr. Callan

Yes, that is correct.

It emphasised that its conclusion to recommend against use of the system in June 2004 was not based on any finding that it would not work. The essence of what the Secretary General has been telling the committee is that there was not sufficient time to go into the necessary details on all aspects in terms of transparency and comprehensive validation, in advance of the local and European elections and the referendum, in June last year. The commission's recommendation not to use the system was not based on any finding that it would not work. Is that correct?

Mr. Callan

That is correct.

Perhaps Mr. Callan will talk about the criticisms made by the independent Commission on Electronic Voting as regards the accuracy and security of the system. From what he has said, the commission appeared to be happy with the accuracy of the voting, as counted. Much has appeared in the public domain to date on this topic and does the Secretary General agree that the independent commission was critical as regards the security and accuracy of the system?

Mr. Callan

Yes, the commission developed criticisms of the system under both of those headings. We believe that these criticisms can be addressed and remedied, working with the commission. The fact that the final integrated software package for all of the four potential polls being run in June 2004 was not fully ready for the commission was a difficulty. The Commission on Electronic Voting was dealing with versions of the software which were not yet signed off by the Department, let alone by the commission. At a certain point the Department and the software developers discovered what we regarded as a fairly minor flaw in the counting software and we notified this. I will not go into the technical means by which it was fixed, but it was basically fixed by moving from decimals to fractions. We notified this technical glitch to the commission, which commented that there was an error in the accounting software that could have affected the distribution of votes. I do not profess to speak for the commission, but our view was that this was a very minor glitch. It was in the process of being fixed, but for reasons of completeness, we made it known to the commission.

The commission and its experts sounded cautions about the security protocols around the computers which extract the data from the modules and produce the final results. It felt that those security protocols could benefit from considerable strengthening. That is not to say that they were wide open to fraud in the arrangements which were in place at the pilot projects and which would have been applied in June 2004. We would not readily allow people with malicious intent to come in and interfere with computers during the running of an election. However, we fully accept the commission's concern. With the benefit of the time we now have, much greater reassurance can be given on these issues. Security can be improved and demonstrated more transparently.

I cannot see how we can proceed with the electronic voting system in the light of the commission's report. We will end up writing the whole thing off. Does Mr. Callan agree?

Mr. Callan

No, I do not agree with that.

The Chairman stated that this morning before Mr. Callan came in. It is published on the front page of today's The Irish Times. A statement like that by the Chairman, where he has reached and published his conclusions on this issue before we have heard from Mr. Callan, has compromised the ability of the Committee of Public Accounts to look at this issue here today in an open manner. It is unfortunate that a person in such a position would publish such conclusions prior to hearing Mr. Callan’s views and the views of the other 11 members of this committee. I am impressed that we have a full turn-out of the committee this morning. I wondered what the point was in coming to the meeting when the Chairman has unilaterally published his own conclusions. We have debated whether if at the end of each meeting, perhaps the committee would be wise to issue a press statement on our findings. It is a new level altogether when the statement is released in the morning newspapers before we even get here.

Following the Chairman's publication of his conclusions this morning before we had the courtesy of meeting Mr. Callan — which I read in a newspaper article at 10.40 a.m., just 20 minutes before this meeting started, I looked up the Fine Gael website. There is a leaflet produced on the website dealing with the Meath constituency, which states that it was produced by the Fine Gael parliamentary party in the run-up to the 2002 general election. It states "Your handy guide to electronic voting, it is as easy as 1-2-3." The Chairman of this committee, Deputy Noonan, who was then the leader of the Fine Gael Party, had the parliamentary party publish this document prior to the last election when he was leader. I am confused as to how that can be on the party website today, yet as Chairman of this committee he has reached diametrically opposed conclusions on the function of the committee's consideration of this report before the meeting even began today. That is not something on which I expect Mr. Callan to comment, but it is unfortunate that we come to this meeting with these conclusions already on the record.

The Deputy's time is up. If he wants to continue to chastise me, I will give him injury time.

I think my point was well made. It is an example of what will prevent rational debate by the committee on topics like this. I would have preferred to discuss the real issue, rather than what was already published by the Chairman before we even got here.

In reply to the Deputy, I was called by a journalist and I gave a statement. As a Member of this House, I am quite free to comment on any public issue. The Deputy will notice from the statement that I did not comment on the statement from the Secretary General of the Department of the Environment, Heritage and Local Government for today's meeting, which was forwarded in advance. The Deputy will also notice that I did not comment on the Comptroller and Auditor General's report, as set out in chapter 8.1. I commented on the report of the Commission on Electronic Voting, which was circulated to all Members of the Oireachtas in early January. That document, although relevant to the committee's work, is not before the committee. What is before the committee is the Department Vote on chapter 8.1 of the report by the Comptroller and Auditor General. I did not comment on either document, so I am not in breach of any protocol procedure or even courtesy in commenting in the manner I did.

The Deputy also spoke of the information leaflet circulated by Fine Gael in Meath and in the other pilot constituencies. What did he expect Fine Gael to do when the system was being tested in three constituencies? We were getting feed-back that some voters, particularly elderly people, were uneasy about going to polling booths and using machines. We were trying to encourage people to go out to vote. As a matter of fact, our policy position was to support electronic voting at that point. I am reviewing my position since then. I am reviewing it in light of what happened, particularly in light of the first report of the Commission on Electronic Voting on the secrecy, accuracy and testing of the chosen electronic voting system. That is my position and if the Deputy is upset, he is upset——

I am not a bit upset.

He is upset on the wrong grounds. He is accusing me in the wrong.

The comment in today's paper is about writing off the expenditure. That is a matter for consideration. The expenditure relates to the work of the committee.

In appendix 2M of the report on electronic voting, there is a feasibility audit which was commissioned by the commission. It was put together by Dr. Frank Bannister of the Policy Institute in Trinity College. Reading through the submissions made to the commission about electronic voting, the constant theme from voters and computer experts is that there must be a full audit trail. This has come from the professions, from the academics in universities and it is the stated position of the Computer Society of Ireland. It was also a principal issue among the public. In his conclusions, Dr. Bannister states that the Nedap Powervote system does not provide a full audit trail. We know that. His second conclusion is that the system is not feasibly capable of modification to provide a full audit trail, without unrealistic cost. The Comptroller and Auditor General states in his report that the proof of the pudding will be whether these machines are used. If the policy position changes to the one taken up by the experts and a majority of the public, namely, that there must be an audit trail, these machines will not be able to be used, according to Dr. Bannister, without unrealistic cost. As a consequence, €52 million of taxpayer's money will be written off.

Before the meeting, in reply to questions, I stated this would happen. It will not happen tomorrow, next week, next month or next year but before electronic voting is introduced, if it ever is, a Minister will make a decision to comply with the demand for an audit trail. As a consequence, these machines will be scrapped. That is a personal view and I accept I could be wrong. However, I gave it as a personal view which I am entitled to give.

Are the machines capable of being adapted to provide for an audit trail? It is obvious there will be a cost involved.

Mr. Callan

That is not something we contemplate at this stage. The commission did not at any stage posit the voter verifiable audit trail as a necessary characteristic of a good electronic voting system. The Chairman is quoting from an expert report. The commission made it clear it did not necessarily endorse or rely on all of the conclusions arising from the expert's work. Moreover, it made it clear that the absence of a verifiable audit trail significantly raised the standards and quality of other system testing required. We accept this fully. However, it has in no sense stated one cannot have a good electronic voting system without an audit trail.

While I do not wish to get caught in a political discussion of this issue, my mandate comes from the legislation which asks us to implement the Nedap Powervote system. It is our professional view that this system is capable of being well implemented working with the Commission on Electronic Voting.

Therefore, what we have just heard was not the view of the independent Commission on Electronic Voting but part of the submission included in an appendix in respect of a report presented to the commission. The commission did not necessarily accept that opinion.

Mr. Callan

It was a report commissioned by the Commission on Electronic Voting. However, the commission made it clear it did not necessarily endorse or rely on the conclusions arising from this work.

It is also true that the commission did not reject it.

Mr. Callan

Yes. However, in defence of the system mandated to the Department to implement, it is abundantly clear that the commission at this point of its work is not stating one must have a voter verifiable audit trail. That is entirely clear to us.

It is entirely clear because the commission was very careful not to go beyond the position of the report published before the June elections. It is simply providing the supporting documentation. However, it is ongoing work. At best, the commission is neutral on the issue. If one reads the full report, it is easy to see where the advice is tending.

I wanted to track that issue but the Chairman and Deputy Fleming have done so. However, for the record, what is the present position? Do Mr. Callan and the Minister accept that verification of the voter trail is essential?

Mr. Callan

The present position is as mandated twice under legislation, most recently the Electoral (Amendment) Act 2004 which was passed after the publication of the interim report of the Commission on Electronic Voting. It tasks us with implementation of the Nedap Powervote system in co-operation with the Commission on Electronic Voting.

Does that mean we will not have a verifiable audit trail?

Mr. Callan

That system does not comprise a voter verifiable audit trail in the sense intended by those who advocate such an arrangement.

For example, will the Department do this in Meath and Kildare North?

Mr. Callan

No. The only system of electronic voting that can be applied under legislation is the Nedap Powervote system.

Will we have it in Meath and Kildare North?

Mr. Callan

That is a matter for determination by the Minister in the context of making various orders which will have to be made for the by-elections.

Has the decision been made?

Mr. Callan

No.

If the experts who say we require a verifiable audit trail are proved correct and it does go ahead, would that be consistent with the hardware and software in operation at present?

Mr. Callan

No. I do not want to speculate about adaptability. One can adapt anything to anything else but at a cost. It might not be sensible to think of doing this.

As the Deputy asks, I stated the formal position first, that we are mandated by the Oireachtas to take forward the Nedap Powervote system which does not have an audit trail in the sense intended by those who advocate such an arrangement. To go back to my introductory remarks, we felt a proven system should be recommended in the context of introducing electronic voting into the Irish system. We believe the voter verifiable audit trail, although much has been spoken of it, is far from proven in the experience of electoral administrations worldwide. There have been a few attempts at such an arrangement which, in the main, have not been proceeded with. There is an aspiration within state policy in California to move towards such an arrangement but there is no proof of its feasibility in that jurisdiction.

In conclusion, Mr. Callan is saying the present system cannot do it and the Department is not persuaded it is necessary. Is that a fair summary?

Mr. Callan

Yes. We believe it would introduce dualism into the voting system.

I do not want to hear the rest of the note as our time is limited. Mr. Callan stated: "In retrospect, we recognise that the timescale for national roll-out of electronic voting may have limited our ability to consult as widely as we might have wished on the project and to provide fuller reassurance about the system". What does that mean?

Mr. Callan

It means there were various real time-——

Who pushed the timescale?

Mr. Callan

The Government indicated from the year 2000, as noted in the report of the Comptroller and Auditor General, that, in principle, it wished to have a national roll-out of electronic voting by the June 2004 local and European elections.

Is the Secretary General saying this did not allow for the consultation and professional advice he would otherwise have taken?

Mr. Callan

No, not professional advice. My remarks referred to communication with interests, even interests which disagreed with what we were doing, and the time that would allow for greater reassurance and persuasion by the Department.

What interests?

Mr. Callan

I am talking about a range of stakeholders, including the public and people from various walks of life who advocated other forms of electronic voting. In the absence of indefinite time, the situation became more vexed than might have been the case if we had more time.

Did we not spend the budget for communicating with the public? What else might the Department have done?

Mr. Callan

We did not spend it all. When the decision was taken on 30 April, we sensibly retrenched.

Had the Department already been running the campaign?

Mr. Callan

Yes. There was also an earlier campaign in connection with the pilots.

As I understand it from the experts, the Department did not show any willingness to engage with them.

Mr. Callan

We were at their disposal. Departmental officials appeared twice before the Joint Committee on Environment and Local Government with the Minister accompanying us on one occasion and the manufacturers on the other. If this system was to be rolled out successfully in June, we had many other calls on our time and attention such as the training of local staff for the major transition to come.

The Secretary General made a point that costs were rising for the traditional system. He instanced that they increased by 71% from 1997 to 2002. What were the main causes of this rise? Inflation was perhaps 15% in that period.

Mr. Callan

I understand it was approximately 20%.

Why did costs increase by 71% in that period?

Mr. Callan

These costs are negotiated and are not pay costs in the conventional sense. They may have to be negotiated under a certain amount of duress having regard to the proximity of elections.

Mr. Callan

Weekend work is a factor and may influence comparisons because staff will demand more pay to work on Saturdays.

If the Secretary General has precise information on this issue, perhaps he will make it known to the committee.

I wish to return to a point which the Chairman raised briefly. The advice I received suggests that a control operator must be provided to operate the machines. Is that control operator additional to the existing staff in each case?

Mr. Callan

No, the operator is redeployed from the same complement that operated in a manual situation.

I am curious about this. The note I received came from the same people who worked with my colleague, Deputy Gilmore, on producing a report which we now know was borne out. It makes clear that the electronic voting system would have cost more to run than the old manual system, even if we accepted the system was safe and secure. The note states:

It appears that the promised cost savings resulting from reduced manpower during counting may be outweighed by the increased personnel required to operate the voting machines. Each voting machine requires a control operator to enable the machine for each voter.

Mr. Callan

In some situations in Dublin there are three people at tables because of longer polling hours. I understand that in typical situations, however, this does not add to numbers.

Is an assessment available? The public attitude on this question is focussed on concerns about safety and security and people are persuaded by the arguments for voter verification. In the event of a turnaround in public opinion and the adoption of this system, does the Department have a figure for the savings per election that are likely to accrue vis-à-vis the traditional system?

Mr. Callan

We estimated a saving of €13 million over 20 years. This estimate was made in 2000 and took a conservative view of the incidence of elections. If more elections are held, the saving will be greater.

How many elections did the Department estimate would take place in that 20-year period?

Mr. Callan

We estimated there would be 12 elections.

In that case, is the projected net saving per election a little more than €1 million?

Mr. Callan

Yes.

The Department put out the tenders for the voting systems in June 2000. What was the statutory authority for that process?

Mr. Callan

We sought to action the Government decision to move in this direction.

When was that Government decision made?

Mr. Callan

It was made in February 2000.

On what statutory basis was this decision made? Was it the 1999 Act?

Mr. Callan

No, the 1999 Act referred to electronic voting but only provided that for research purposes we could have a certain type of access to election records. The first full statutory authority for conducting electronic voting, either partially or on a nationwide basis, was provided in the Electoral (Amendment) Act 2001 which was enacted just before the summer of that year.

It became law on 24 October 2001 but the Department put out the tender to purchase the machines in June 2000.

Mr. Callan

Yes. There was no final, contractual commitment to a large number of machines. That took place later. We started the procurement process but did not cross the line of major financial commitment until considerably later.

When was this financial commitment made?

Mr. Callan

Some 6,000 machines were ordered for pilots in November 2001.

The order was made after the Act became law?

Mr. Callan

Yes.

The committee to which the Secretary General refers raised all the public fears in December 2003. Some Deputies had professional advice from information technology experts. How much was the Department committed to at that stage?

Mr. Callan

We were committed to the bulk purchase of the 6,500 extra machines which were required in addition to the pilot machines. This comprised a total commitment of more than 7,000 machines from late 2003.

Was the committee's work a wasted exercise in that the Department would go ahead with its plans regardless of the committee's conclusions?

Mr. Callan

The debate on this issue was not a new one and decisions had been signalled much earlier. It was evident from 1999 that there was an interest in pursuing electronic voting. In 2000, the Government made a major announcement that indicated its intent in this regard.

Who is making the arrangements for the storage and security of the machines?

Mr. Callan

That is the responsibility of returning officers.

Do they get the costs from the central fund?

Mr. Callan

Yes.

What is the situation now facing us if a decision is not made and if the public and public representatives are not satisfied with this system?

Mr. Callan

The situation in which we find ourselves is that we have a mandate from legislation enacted in July 2004 to proceed with this system subject to validation by the independent Commission on Electronic Voting. That is the mandate we are working towards. We believe validation reassurance in the capabilities of this system can be proven and we want to vindicate the system and the public investment made in it.

Is the commission currently seized of the issue?

Mr. Callan

Yes.

Has it given clearance for the by-elections, for example, in the event that the Minister should decide to use it?

Mr. Callan

It has advertised for a testing consultancy of its own at this point. That is in the public domain. It did that sometime before Christmas. It is not in communication with us on the by-election issue.

The Minister is confronted with the decision when he does not know whether he might make the decision and the commission might again have to voice public disquiet.

Mr. Callan

He will take account of all the factors the Deputy mentioned, no doubt, but it is his decision.

Will we ever use these machines?

Mr. Callan

We carry out the wishes of the Government and the Legislature under the mandates that I have mentioned. We have no doubt that the system is fit for its purpose and usable. We do not accept a voter verifiable audit trail is a necessary test of a good electronic voting system. That thesis is not accepted on a widespread basis in other electoral administrations. If a system which services 11 million people in advanced democracies of western Europe is doing a good job it should be well capable of adaptation to Irish circumstances. That is the path we pursue at the moment.

We should go very easy on the 11 million and the advanced democracies. It does not take in much when one goes outside the borders of the Netherlands.

Mr. Callan

There are three million Germans using the system.

Is it not the situation, whether Mr. Callan is correct in his conviction about the usability of the system, that the handling of it has been a shambles and whether it is a mitigated shambles depends on whether it can be made usable to the satisfaction of the public and our public representatives in the years ahead? As of now we are faced imminently with two by-elections and we are not in a position to say whether we will use this equipment in respect of those two by-elections, which cost €52 million of taxpayer's money.

Mr. Callan

For the reasons I set out in my opening statement I believe the risk assessment of this project has been actively and responsibly carried out. We have procured a system that is well capable of serving its purposes and of strengthening arrangements for Irish elections removing the many inefficiencies of the present system. It is noteworthy that the recent legislation, the Electoral (Amendment) Act 2004, has somewhat amplified the terms of reference of the Commission on Electronic Voting in order that it will now also have to compare the electronic system, the Nedap Powervote system, with the operation of the manual system. I have not gone into this much today but that system labours under many difficulties and produces inaccuracies up to a considerable point.

What is the Department of Finance view is on the management and operation so far of this project?

Mr. Doyle

The procurement of a system etc. is a matter for the Department of the Environment, Heritage and Local Government. The Comptroller in his report has stated that the procurement has been carried out in accordance with the procedures and we are not aware of any problems either with the procurement or any other aspects of the system.

Whether it works or will ever be used are not matters for the Department of Finance?

Mr. Doyle

The Department of Finance is working under the mandate of the Government decisions that have been taken in regard to this system and it will continue to operate under decisions taken by the Government on the operation of the system.

There are a number of issues on what has happened up to now but also on the future. Forgive me, Chairman, I appreciate Deputy Fleming may have asked about the future of e-voting and your own suggestion that it might never happen and might never work. Our historical record is not good on e-voting. In 1989 or 1990, as members of this committee, we were sent off to look at the electronic voting system in a number of parliaments and we finally got around to installing it here in 2002 or thereabouts. It took 12 years to get the Oireachtas members to use it. The history of it is not great. I could make a case for or against, Chairman. You changed your own mind on it. You were for e-voting but now you have more than an emotional attachment to the old system. I certainly would have an emotional attachment to it because I have gone through the wringer as it were, on a couple of occasions, particularly at our last election.

I would be concerned if, for whatever reason, Parliament decided to turn its face totally against the possibility of using new technology at any time in the future. I am glad that Mr. Callan referred to the reasons for bringing it in. We know from the figures, and all of us who have tallied are aware that 95% of the error rate, the spoiled votes, is human error, for example, three ones, two ones or whatever. Perhaps two or three people use the opportunity to write an unpleasant message. I suggest it was an effort to rectify that problem. I have had the experience of 54 votes being counted twice as 50 in a knife-edge election. If we can improve that position I would be happy to do so. There were difficulties.

I was disappointed with the Minister's proposals that visually impaired people were not being facilitated with e-voting. I felt that with a simple add-on they could have been facilitated. I am worried that, for whatever reason, we would say that e-voting will never happen and that we might decide to write it off. I presume Mr. Callan answered earlier on the future of e-voting but does he foresee it taking pIace?

Mr. Callan

We hope the system can be enhanced to make it much more user friendly for people with disabilities. There is a plug-in port on the machine which will accept a special adapter — I have not got the technical details — which would be a great facility for people with disabilities. Unfortunately that adapter has not yet been fully developed anywhere. The system we have procured has the capacity — probably more than its competitors — to accept add-ons that will be friendlier to people with disabilities. In the shorter term we have procured tables that are suitable to people with disabilities more than the present manual polling booths.

There is an audio aspect also but I will leave that. I participated in a recount in 1982 when two of my colleagues had four, two votes initially and then five, votes between them. The aspect that jumped out at me was the random selection for distribution of votes. It is a lottery. It is like "Winning Streak" on Saturday nights. With e-voting is it possible to remove that element, where somebody decides to take a bunch of votes at random and depending on the district it could decide the outcome of an election in many cases?

Mr. Callan

Electronic voting has the capability of readily producing a purer arithmetical resolution of preferences than the manual system. We are not deploying that capability at the moment because the law states that the preferences are to be distributed in a certain way. We have required the electronic voting system to follow that arrangement, which reflects the weaknesses of the manual system, rather than move directly to a more arithmetically rigorous solution. Any change in the counting system would be a matter for legislation.

Can I take it the reason the random system is used currently is that the alternative would be to manually count the votes for the 15 or 16 candidates? Is that why the law is phrased that way? Obviously, if an electronic method could be used it would be used but that is not happening because of the way the law is phrased.

Mr. Callan

If we bring forward legislative proposals to deal with preferences in a new way, I would imagine that legislators and the public will be quite interested in the form of those new arrangements. It may seem simple to say that the computer will do it but there are policy choices to be made in terms of how we do the counting of preferences, even with a computer. That is a matter for wide political and public debate before the necessary legislation would be finalised.

Mr. Callan said he had a mandate from the Oireachtas. Was that imprimatur widespread? I recall when the decision was made around 1989 to hold by-elections, a programme was set out for a number of years. Can Mr. Callan recall if there was much opposition to that or did the events in Florida highlight many of the difficulties? Can Mr. Callan recall, from listening carefully at the time, whether many objections were made to the principle? I read in the newspapers that people will question the role of the Minister but I presume the Minister had the imprimatur of the Dáil to do this?

Mr. Callan

The user surveys we carried out after the pilots showed a high level of satisfaction and, anecdotally, it appeared to us that was reflected in wider society in that there was a buy-in to the novelty, as it were, of electronic voting. The difficulties arose later. As Deputy Dennehy rightly suggests, much of the advocacy for a voter verifiable audit trail comes from the United States where the electronic voting machines used there are much differently configured than our system and where cultural and political factors also differ.

In terms of election changes, an area in which there was human error involved the failure to stamp, at least on one occasion, a full book of ballot papers. Would that human error difficulty be removed with this system or is there any difficulty, in terms of what we saw in Florida, with e-voting that Mr. Callan knows about? Would that be a plus in removing the difficulty of not stamping ballot papers?

Mr. Callan

It would be a net gain. It is interesting that Deputy Dennehy mentions that category of spoiled vote. We have done some analysis of spoiled votes over different kinds of elections and, overall, we reckon that nearly 7% of all spoiled votes are down to the cause the Deputy mentioned — ballot papers not being stamped. I can give the Deputy the other categories that make up spoiled votes. Failure to mark "1" on the ballot paper is the biggest cause of grief regarding spoiled votes. Approximately 45% of spoiled votes over a number of years are down to that factor. The figure for putting "1" against the name of more than one candidate is 45%, and the figure for writing on the ballot paper in a way that might identify the elector is 6.5%. I apologise; I have given the wrong figures. I was quoting from a particular referendum. The average figure is 6.75% for failure to stamp the vote; 32% for not indicating "1" sufficiently clearly or not marking it at all; 53% for doubling up on the "1s"; and 7% for writing something on the ballot paper that is unacceptable or that identifies the voter.

I can recall getting three "1s" on a few occasions. Following the pilot schemes in 2002 in Dublin North, Dublin West and Meath, which represent a good cross-section of the community, were there widespread complaints? Did difficulties arise such as people not getting their ballot paper and so on?

Mr. Callan

The general feedback was good. The user surveys were overwhelmingly positive. A total of 96% of those surveyed found the machine very easy or easy to use. Since the exercises were pilot ones, we were looking to find enhancements in service out of those pilot exercises and we decided it would be more user-friendly if the illumination on the machines was increased and if more space was given in order that names of candidates could be read more clearly. That, and the presentation of results, was something we intended to address in June 2004. We felt that could be streamlined and handled in a better way but that does not really affect the performance of the machines.

I am glad to hear of the results on that because there were suggestions that Irish people would be incapable of using this new technology. I had concerns that others knew the Irish better than ourselves. Were observers sent to India to examine the system there? Was there any interest in that?

Mr. Callan

We took some soundings and it is interesting that large countries like Brazil and India, with developing economies, are buying in quickly to electronic voting and using it extensively.

We deprive a fairly large percentage of the electorate of the possibility of voting because they are ill and so on. I appreciate some changes have been made to allow people working in offshore fishing and so on to vote but would there be any benefit in e-voting? Could we facilitate people who are hospitalised, maternity cases and so on because we disenfranchise a percentage of the population in terms of their vote? Is there any way of catering for them by some outreach method or any other way Mr. Callan might know of?

Mr. Callan

The issue of remote voting using technology is under discussion in many countries, including in the United Kingdom where there is a certain appetite for determining if forms of communication like texting and the Internet might be deployable at some stage in the future. There are wide and understandable security concerns around the prospect of those forms of remote voting, which we consider have not yet been resolved. Our machines are stand-alone and completely separated from any possibility of interference through any network.

If I can withdraw money from a bank at a distance, we should surely be able to do it. However, we disenfranchise people. The Comptroller and Auditor General had one serious gripe. Can Mr. Callan deal with the aspect of the independent validation of the testing? It was suggested in the Dáil that the people in Government might put in a disk and control the whole process. Can Mr. Callan deal with the issue with which the Comptroller and Auditor General was concerned?

Mr. Callan

Now that we have been given more time, we believe there are ways in which we can better demonstrate the accurate performance of the system. A parallel voting exercise can be mocked up and it is likely that in the effort to provide greater confidence in the system we and/or the commission will develop more of these exercises in different ways. There would also be possibilities of, for example, taking various machines randomly out of commission on the morning of an election as a test against the possibility of any fraud or interference, and testing those machines as part of the greater security awareness that we must now develop.

I am told that visitors to Cork these days are not sure whether they have arrived in India, with rubbish piling up on the streets as the bin tax bites.

It is no thanks to some of the parties involved. The influence is travelling from Dublin, I can tell you, Chairman.

The Deputy is saying that the people in Cork are mugs, after all — that a few individuals can lead them by the nose.

I put it to Mr. Callan that his Department has driven the implementation of electronic voting with an incredible arrogance. There is now a very real risk of losing €50 million or more in public funding. I was interested to hear Mr. Callan saying, if I understood him correctly, that the same Leech agency was involved in the vetting of tenders for electronic voting, as was involved in the race against waste campaign that showed rats and debris falling all over the place. I am not sure whether we will have a remake showing hundreds of useless electronic voting machines hitting unfortunate taxpayers over the head. It seems there is a real chance that this will be the situation.

I want to put two or three brief points to Mr. Callan about that. In his report, the Comptroller and Auditor General points to the review of the pilot project, general election 2002, carried out by the Department. One point that emerged was the need for a more co-ordinated approach to procedures at the count centres, particularly as regards the presentation of the count information and results to candidates and the media. I have an interest in this because I was one of the guinea pigs in Dublin West in 2002.

A month before the election I wrote to the then Minister, Deputy Noel Dempsey. We were told at that stage that the final result of the three constituencies would be posted on the display screen. I asked the Minister if each count could be put up on the screen in succession, while the returning officer would simultaneously declare it verbally. I stated that this would not take more than ten minutes. First, however, for the candidates who should be considered, it is a much more reasonable way to see how they have fared, count by count. Second, it is better for television coverage and those watching at home. As we know, people follow election counts in large numbers. It would be much easier to digest the results if they were given count by count, even if that was done in fairly quick sequence. That letter was ignored.

The counts in Dublin West and Dublin North were an utter shambles. In the case of Dublin West, the candidates were hustled into a corner and given a sheet of paper with the results. We had to decipher there and then whether we had been elected. A thousand observers and supporters were desperately trying to decipher from our facial expressions whether we had gone up or out.

In the case of Dublin North it was even worse. The candidates were lined up in public and then the final result was announced. That illustrates the arrogance involved. Why was the simple procedure that I proposed not adopted and implementation directed by the returning officer?

Mr. Callan

The pilots were promoted as a learning experience. One of the lessons gained from the experiences described by the Deputy was that these things should be done much better if the system was to be extended nationwide. Arrangements, closely paralleling the Deputy's own suggestions, were actively worked on and would have been put in place for any subsequent deployment of the electronic voting system. The speed with which results were produced at the first pilot, and more especially at the second Nice treaty referendum, demonstrated something of the efficiency of the system.

That is not the point, Mr. Callan. The point is that a month before the election the Department had a letter to the Minister from a Dáil Deputy and candidate in which a simple procedure was proposed. It would have been much more readily understood at home, would have been more humane — if I can use that term — to candidates, and more efficient, but it was completely ignored. Mr. Callan says that steps were taken following the election but the shambles did not have to happen if somebody had read the letter and considered that it made complete sense.

Mr. Callan

The point is certainly well taken at this stage in terms of the protocol under which electronic voting will be deployed.

Why was the advice not taken at that stage? Does this not show that all professional advice concerning the system itself was equally ignored?

Mr. Callan

The arrangements to which the Deputy referred have more to do with the traditional skills involved in organising things on polling night, than with the operation of the new system per se and the accurate performance of the system in counting votes. Important as the Deputy’s concerns were, they concerned protocol issues rather than the fundamental performance of the system.

Elections are about people, not machines. In his review of the pilot project, Mr. Callan had to include it. Obviously, if it was included there it is an important consideration but one that was ignored completely ahead of the election.

Mr. Callan

The experience of the pilot scheme has clearly vindicated the thinking the Deputy had developed, even prior to the pilot. We are glad, even belatedly, to take the Deputy's concern on board.

From the second half of 2002, the Department had submissions coming in from experts, raising the most serious questions about the viability of electronic voting, security and various other considerations. Yet a full six months later, the Department ordered the machines for the whole country. Who was delegated in the Department to examine those submissions? What kind of consideration were they given? At what level were they considered and how seriously were they considered?

Mr. Callan

Obviously, these points were made to us. It was within a fairly dynamic environment where we had various targets to meet. We had confidence in the system. We were involved all the time in continually testing it. The main point being put up against the system at that stage was the principled one that electronic voting should not be implemented without a voter verifiable audit trail. That is an arguable point. It is not one we accept, but it was a legitimate argument. Looking at the experience of electronic voting, as it then stood and as it now stands worldwide, one finds 90% plus of electoral administrations using their electronic voting systems without deploying a voter verifiable audit trail. Where the attempt is made to deploy such a trail, there is proven difficulty. We felt that it was not appropriate to recommend to the Minister and the Government that a proven system should be jettisoned or subjected to major amendment against these kinds of argument.

The reason for the demand for the verifiable audit was concerns about the operation of the system itself. The commission held with those individuals who made submissions and did not hold with the Department on that subject.

Mr. Callan

The conclusions to which the commission came were more nuanced than Deputy Higgins has described. It determined that it did not have enough time, as an independent and newly established commission, to recommend with confidence, the introduction of the system in June 2004. It found many good things to say about the system but it also said that certain aspects of it could be improved and-or that the validation of some of the elements of the process could be made more transparent. The commission did not find in favour of the need for a voter verifiable audit trail.

It found that it had sufficient reservations that it should not proceed — reservations which had been expressed previously but which were overridden in the Department.

Mr. Callan has repeated on a number of occasions that the commission was not in favour of the VVAT system. The commission states that consideration of this did not fall strictly within its terms of reference. That is the reason its comment is soft rather than that it did not agree with it. The Secretary General can check it later but the introduction on page 77 states that it is making observations on it but that it does not fall within its terms of reference. Therefore it could not make a recommendation.

The argument some Government Deputies have attempted to put forward and Mr. Callan relies on as well in regard to how easy it is to use these machines, is not really central. It is important that they are easy to use but the fact 96% of people said they were easy to use does not cover up the other, very deep reservations about the system. If one asks people whether it is easy to pull the trigger of a gun, they will say it is. It is the simplest thing possible but the results or implications could be quite different. That is the point. It is not counterweighted by how easy people felt the machines were to use.

In regard to the rush to implement, the Comptroller and Auditor General said that while the procurement of hardware and software was in line with proper procedures, the decision to purchase the additional 400 voting machines for the Nice referendum without the identified modifications ultimately resulted in avoidable costs of €680,000 being incurred. I do not know if that point was put in my absence when I was attending the vote. If it was not, I would like it answered.

Mr. Callan

I will come back to the user survey because I only quoted one piece of information from it previously. Perhaps a more telling result from that user survey was that 87% of voters surveyed said they preferred the new system to the old paper one.

To come to the extension of pilots from three to seven constituencies, it was not physically possible to have the enhancements we were to have to all machines carried out on the additional number of machines in the time that would have allowed for the deployment in the four constituencies. Therefore, the Comptroller and Auditor General is right that an additional cost of more that €600,000 was entailed in pushing it forward into those four constituencies. It was slightly mitigated by reduced counting costs in the constituencies concerned.

I tried to explain in my opening statement that, at the time, a major transition from one form of voting to another was being carried out. There was a wish on the part of Government, to give as many voters as possible the benefit of experiencing the new system before it went nationwide and, to build confidence and demonstrate the system on a wider basis. I agree this entailed a cost. In the totality of what we were doing, I have suggested that the Minister might argue that cost was a reasonable one to take on board at that point.

Electronic voting was used twice in the general election and in the referendum that followed. Obviously, a fair amount of research has been done on how it fared. The commission has looked at it also. Nobody has suggested that the results were inaccurate, which is one of the key aspects separate from customer satisfaction. Nobody has suggested to date that the results were inaccurate. Is that correct?

Mr. Callan

Yes, I believe that is the case. The commission, in particular, has positively asserted that the votes were counted accurately.

If the Minister turned around in the morning and said that electronic voting would be used in the two by-elections, would the Department be physically and technically in a position to deliver that?

Mr. Callan

Yes, but obviously the policy situation has evolved. I am not speaking for the Minister but, with the sort of accuracy I described a moment ago, the same job of counting could be done.

After the two pilots and before the last general election, the issue arose and the Commission on Electronic Voting was established. Am I correct in stating that the stand-alone unit was not one of the primary concerns and that people's concerns primarily rested on the counting of votes and the fact that the process could be tampered with or manipulated?

Mr. Callan

Yes.

Mr. Callan stated earlier that the amount of money spent on software has been relatively small when compared to that spent on hardware. Is that correct?

Mr. Callan

Yes.

Are those responsible for the software development the same people who supplied the hardware? What is the relationship, if any, between them?

Mr. Callan

They are formally different companies. Powervote is supplying the software and it is aligned with a company called Groendaal which is based in the Netherlands. However, there are close relations between the manufacturers of the hardware and those who are developing the software. They are closely associated and there are some connections between them.

However, there are no legal or binding agreements between them. In other words, the software designers in place at present could be replaced by any other software designers.

Mr. Callan

That is correct. If necessary, we could do so.

We have spent most of our money to date on the hardware. Am I correct in stating that the company involved in that side of matters is virtually out of the equation? That company has no real concern except to ensure that the machines remain in operation.

Mr. Callan

The company is offering us maintenance support until 2007.

However, in terms of finding a solution to this matter, it is out of the equation. A software solution is what is required.

Mr. Callan

In the main. I have to be circumspect here. The commission was established under statute to make various reports. It has made extensive reports already and will make further reports on this entire situation. My interpretation — I stress that it is my interpretation — of what is coming out of this analysis to date is that the difficulties seen by the commission and identified by it are, to a greater extent, on the software side rather than the hardware side. The machines came out of the examination quite well.

I do not know where it originated but on the occasions when we discussed electronic voting most people stated that the system — hardware and software — would have a useful life of 20 years. Is that correct?

Mr. Callan

Yes, that is the premise of our analysis.

The initial batch of machines arrived in November 2001.

Mr. Callan

Yes.

Therefore, the design of the technology had been proven and it had been around for a considerable period and the Department was happy with it. Is that correct?

Mr. Callan

Yes.

We, therefore, have machines which are stand-alone and which work but we are concerned about the security of the system, particularly in terms of counting. The system has a 20-year lifespan and we are discussing adapting the software to enhance it. I am not an expert but the amount of development in this regard will be limited by the parameters of the machines. I am surprised that, apart from a maintenance contract until 2007, the original supplier will be out of the equation. I would have thought the supplier would have to be tied in for the entire 20-year lifespan of the system. Technology is going to evolve and security will always be an issue but the supplier of the hardware will not be involved.

Mr. Callan

I do not have all the details of the contract in my possession but there are options for engaging the hardware supplier beyond 2007, at some level, and for upgrading the system as required. The figures are not hard and fast but there is that understanding within the contract and I believe it is enforceable. In that sense, the supplier is not completely removed from the scene.

A number of people have looked at the software developments that would be required to ensure the security of the system. Did anyone indicate, at any stage, that there were limits due to the hardware currently available?

Mr. Callan

The voting machines we have bought are designed purely to record votes. They do not themselves carry out processing. In that sense, they are a fairly simple element. If there was a desire to pursue more sophisticated forms of remote voting, such as those to which Deputy Dennehy referred, in the medium term, we are of the opinion that these machines could be adapted and used for that purpose. Despite the fact that they represent a proven design which evolved some time ago, we are of the opinion that their function is simple, namely, that they are electronic replicas of ballot boxes, and that they will be compatible with various software enhancements. That is our belief at present.

The Department is considering reports from the Commission on Electronic Voting and it has its own experts. Does he see a time when the necessary adaptions will be made and when the machines will comply with the requirements of the commission?

Mr. Callan

That is our ambition and that is what we are working towards at present.

Does Mr. Callan believe that public confidence in electronic voting has suffered as a result of this exercise?

Mr. Callan

We are aware that the high level of satisfaction and confidence that was elicited in our survey in 2002 following the initial pilot——

I am not talking about 2002.

Mr. Callan

——was somewhat eroded by events. Various unofficial polls were carried out during the height of the controversy in the spring of last year which showed that the high level of support for the system did not seem to be holding up as well. However, we are satisfied that with good demonstration and more time to show the capabilities of these machines good public confidence can be built around the system.

That is the point, however. Even in his opening statement, Mr. Callan stated that the premise for introducing electronic voting was to encourage voter participation. However, that has been turned on its head because people do not trust these machines. The Department has managed to do the opposite of its stated intention.

Deputy Rabbitte asked Mr. Doyle a question earlier. I am not trying to be smart but I have no idea what he was trying to say in his response about the Department of Finance's management of this issue. Do our guests have opinions? It is important that they, as civil servants, have opinions as to the viability of this system into the future because €52 million may, potentially, have been lost. Our guests must consider the situation in so far as ordinary people are concerned. They do not have faith or trust in this system any longer. The initial idea was to engender trust in the system but that trust has evaporated. The game is up. Civil servants need to be able to tell their political bosses when something is over. In my view, the Department is living in denial.

There is no point in discussing this matter any longer. Members of the public do not trust these voting machines because of what has happened in the past two years and they will not trust them in the future. I will ask questions over the next five minutes and Mr. Callan will try to avoid this. However, that is the reality and an official in the Department of Finance or in the Department of the Environment, Heritage and Local Government needs to admit the game is up and the initial goal is not achievable. As Mr. Callan said in this opening statement, the intention to encourage greater voter participation is shattered. I do not expect a response but there needs to be a realisation within Government circles, including top civil servants, that that has happened and this issue must be finalised once for all.

Mr. Callan

We have our mandate from our Oireachtas to advance this system in conjunction and co-operation with the Commission on Electronic Voting. The commission's recently published first report followed an interim report. The commission states in the conclusion, "In fact, it is clear to the Commission that the chosen system is capable of gathering and counting votes at elections with a high degree of accuracy, secrecy and efficiency and this is evident from the results of the tests carried out by the commission, as described in this report."

That was not my point. I referred to the public's trust in the system. These are different issues. I can put the following question to everyone around the table, "Do you believe that public confidence in electronic voting has been eroded because of this exercise?" None of the witnesses can say it has not. It is not a question of a simple majority because a significant proportion of the electorate does not trust this system and will not do so for the foreseeable future. As civil servants who have potentially wasted €52 million through mismanagement, the witnesses should have an opinion and call it a day.

Mr. Callan

I cannot speak for my colleagues in the Department of Finance but I do not accept that thesis. We believe that the investment is valuable and it can and will be made to yield good benefits in terms of improved service for voters.

Mr. Doyle

I cannot accept what the Deputy said. As I pointed out in reply to Deputy Rabbitte, the Department of Finance operates under the mandate of the Government decisions taken in this regard. Six Government decisions have been taken and the Department of Finance through its Minister feeds its views into that decision-making process.

What is Mr. Doyle's opinion?

Mr. Doyle

I am not here to have my own opinion; I am here to represent the Department of Finance and I am telling the Deputy what is the position of the Department of Finance.

This is the kernel of the problem. Unless Mr. Doyle has an opinion, this could continue.

Mr. Doyle

I am here representing my Minister and my Department. I am telling the Deputy the mandate on which we operate. I am not here on a personal basis and I am giving the Deputy my answer.

Therefore, the Department of Finance in a case such as this operates without giving an opinion.

Mr. Doyle

I did not say that. The Department of Finance, through its civil servants, submits opinions through the Minister to the Government where decisions are taken in respect of the mandate under which we operate. I do not accept what the Deputy is saying.

Does Mr. Doyle believe public confidence in electronic voting has been eroded as a result of this?

Mr. Doyle

We are here to look after the costs in this. That is our role in the Department of Finance. The procurement method was the main responsibility of the Department of the Environment, Heritage and Local Government. It has been done in accordance with the capital appraisal guidelines that have been issued. The Comptroller and Auditor General stated in this report that the procurement process we followed was proper. That is where the Department of Finance stands on this issue.

If that is Mr. Doyle's position, that is fair enough. Does he understand the point I am making? This is an intolerable situation when the people who are dealing with the setting up of this system are not able to express an opinion when this amount potentially is wasted. That is the problem. We all need to sit up, behave like adults and say, "We got this wrong, the voters will not have trust in this system for the foreseeable future."

The witnesses are present in an official capacity and not to give personal views. We respect that position.

Mr. Callan made it clear there was an erosion of confidence. That says it all.

Who is responsible for the storage of the equipment?

Mr. Callan

The returning officers in each constituency.

Were the machines delivered to each constituency or have they been stored at a central location?

Mr. Callan

They are in each constituency.

Was a tendering procedure used to store them in the constituencies? How much is the Department paying for storage on an ongoing basis?

Mr. Callan

A variety of methods have been used by returning officers who have discretion in matters relating to the electoral machinery. We recently completed the compilation of details of storage costs this year in all constituencies, which we can circulate to the committee. Some returning officers use local authority accommodation and those costs are not above the line. The accommodation is cost free but in most cases there is a need to procure accommodation at a cost to the central fund.

There is a significant variation in the costs. Is Mr. Callan concerned about this? For example, the cost in Carlow-Kilkenny is €28,000 compared with €298 in Louth.

Mr. Callan

Different property situations confront different returning officers. It would probably not be realistic to impose standard costings or limits on this service.

Was a tendering procedure followed?

Mr. Callan

I would suspect not in most cases. I would suspect the arrangement was made directly.

There were no tendering procedures.

Mr. Callan

In most cases. I cannot speak with complete assurance about everywhere. I would imagine it was done directly by the returning officer.

May I publish this document?

Mr. Callan

Yes.

I seek clarification about the numbers employed on election day. How many officials are required to operate the polling booth system? How many will be needed to implement the electronic voting system nationally?

Mr. Callan

In aggregate terms, we believed that the introduction of electronic voting and counting nationally would mean a saving of approximately 1,200 staff dealing with the count. It would have the major benefit of eliminating the need for 1,200 staff.

People are needed on polling day. How many people were needed under the old system and how many were needed under the new system?

Mr. Callan

Under the old system, 15,000 staff were needed nationally and the same number is required under the new system. There is no change.

There is no change.

Mr. Callan

The change relates to the counting staff.

Would the cost of employing people be the same under the new system? Would nothing be saved?

Mr. Callan

One would save, in that 1,200 fewer people would be needed to assist with the count. Otherwise it would be neutral.

That is interesting. From where did the pressure come to have the system ready for the election?

Mr. Callan

The Government signalled its intention to introduce the system in 2000, subject to feasibility. It was a natural ambition, if confidence was to be built in the system, to build up its use on each occasion. It appeared a reasonable target for the Minister and the Department to have things ready for national roll-out in June 2004.

With hindsight, does Mr. Callan think the advisers could have advised that more time be given in this regard? Mr. Callan admitted earlier that if he had more time, he would have looked at the issue in a different light.

Mr. Callan

With hindsight, I said that more time gives more opportunity to run more tests and to give a better demonstration. That will always be the case. That is in the abstract. In the concrete circumstances which faced us, it appeared reasonable to try to implement the system nationwide for June 2004.

Was there political pressure from the Government, particularly from the Minister for the Environment and Local Government, to have the system in place for the election?

Mr. Callan

As I explained, the Government made clear as far back as 2000 that it wished, if possible, to have nation-wide deployment of the system for the important suite of elections which was coming together in June 2004. That was the stated objective. Following consideration by the Government, successive Ministers had the job of trying to implement the system.

Did you advise at any stage that you may run into difficulties, given that public utterances were made by public representatives during and after the announcement? When did you realise the system would run into difficulties?

Mr. Callan

The Department regarded the timescale to be reasonably challenging but we committed more resources to it. At the end of 2002, we significantly beefed up the staff of the franchise section which was dealing with the project. We felt on that basis that we could assure the Minister the target was achievable.

I have a few questions for Mr. Doyle from the Department of Finance. In December 1999, the Department of Environment, Heritage and Local Government submitted a business plan for the project to the Department of Finance. In January 2000 your Department expressed the view that their business plan did not adequately address a number of issues. In particular, you said it did not address security or reliability of the system. You also expressed the view that no overall costs of the project, or contingency arrangements, were included in the business plan. Did you get that information subsequently?

Mr. Doyle

Yes. Most of that information was supplied subsequently in memorandums to Government. Our concern at the time was not about a particular system, we had a general view about any system that might be purchased. We felt that the Department should ensure that the security element, including the back-up facilities, would meet the requirements. We were not concerned about a particular system. We were concerned that any system should meet the general criteria. We wanted to be sure it would meet those criteria.

You got information which satisfied you that the system was secure and reliable.

Mr. Doyle

No. If the Chairman examines the different Government memorandums, we continued to request that the security element be in place and that there should be back-up facilities. We continued to raise this issue. One of the main recommendations of the commission was that more testing should be carried out on the security and accuracy aspects of the system.

Are you saying that the Department of Finance had concerns about the security aspects of the system right up to May last year when the commission reported?

Mr. Doyle

No, Chairman. In 2002, when we received the memo, we said that more details were needed on the security, viability and confidentiality aspects at that stage. We were asking for an assurance that whatever system was put in place would meet those criteria.

In your first reply, you appear to suggest that at some point subsequently your queries were satisfied by the Department, and in your second reply you appear to suggest that the commission reflected your concerns, and that your concerns were and are ongoing. Which is it?

Mr. Doyle

No. Let me clarify the position. In January 2000, we outlined our concerns about the security and reliability of an electronic voting system and the need for a contingency plan in the event of its failure. These observations were made prior to any tender being issued. They were general concerns about any particular system. We raised the matter again in a memorandum in February 2002. The Department of the Environment had received international reports and some of our concerns were dealt with in that context. Ultimately, we still wanted an assurance on the issue. Even at this moment, we want an assurance on the integrity of the system. We have no reason to believe there is any problem with the system, but we want an assurance that the system will have complete integrity and accuracy. We have no particular concerns.

Is it the case that you do not have a particular concern about the security or reliability of the system but is it correct that you have an underlying concern which has not been satisfied, and you share the same concern expressed by the commission?

Mr. Doyle

The commission's electronic voting board has stated that one of its main concerns is the need for a comprehensive security and risk assessment of the system. We agree with that and await the outcome of the commission's report.

Are you also saying that this has been your position since November 2000?

Mr. Doyle

Yes. We raised these issues in a memorandum in November 2000.

Correspondence from the Department of Environment, Heritage and Local Government to the Department of Finance indicated an estimated cost saving of €13.8 million over a 20 year period. Subsequently, a year later in November 2001, the Department of Environment, Heritage and Local Government made a submission stating that the savings on the conduct of each poll would arise due to reduced vote counting and administrative costs. However, no financial data on the savings were included. Was Mr. Doyle happy that his Department was asked to make a decision on savings that were not specified and on which no cost analysis was submitted?

Mr. Doyle

We asked for detailed costings. We were of the view that the main savings that would arise when the system was up and running would be on the counting and voting end of it.

Did the Department get any document from the Department of Environment, Heritage and Local Government which built up the saving in a manner with which we would be familiar?

Mr. Doyle

No, some material costings were provided in a memorandum for Government of 14 February 2002. We got some information on costings back from the Department of Environment, Heritage and Local Government.

Was the Department satisfied with what it got back?

Mr. Doyle

Yes, we were fairly satisfied with it. We would have liked far more detailed costings such as those we are talking about today and to have got costings on by-elections and their operation. We would have preferred if the costings were as detailed as that.

Was that not done?

Mr. Doyle

Not at that stage.

Moving from savings to costs, in December 2001 the Department of Finance got a cost from the Department of Environment, Heritage and Local Government. The estimated cost at that stage was €31.7 million. Was that the first time the Department got an estimate of cost?

Mr. Doyle

I must check my papers.

The reason I ask this is that this was after the tender. It appears from the papers I have seen that no estimate of cost was provided to the Department of Finance until the tender was completed. However, it did not make a whit of difference what estimate was received at that stage because the decision had been made.

Mr. Doyle

I will have to confirm that because I do not have that information in the papers I have with me.

All right. In June and October 2002 the Department of the Environment, Heritage and Government again made submissions pointing out that there would be savings, but no financial data on savings was included and this as late as October 2002. Is that correct?

Mr. Doyle

No, let me get the information. Material costings were provided in the memorandum for Government on 14 February 2002.

Mr. Doyle

We would have preferred, even then, if we had got a detailed schedule of savings such as we have spoken about today through a projection of future elections or some idea of the savings on a similar basis.

The Department of Finance is the central Department monitoring activities, in particular expenditure activities, of line Departments. This was a situation where savings were claimed but they were not specified or built up in any kind of analytic fashion. The cost only came in after the tender process and there were doubts about the reliability and security of the system, doubts that still exist. Why did someone not shout "Stop"? Is that not the function of the Department?

Mr. Doyle

It was clear from day one the savings would arise when the system was operational. It is only when a system is fully up and running that one has substantial savings.

Yes, it was agreed here with Deputy Rabbitte that savings would be slightly over €1 million a year.

Mr. Doyle

We went through the memoranda for Government. We got Government decisions on six occasions on the system and where it was going, etc. We fed our views into that scenario and worked on the mandate we got back from Government on how to proceed and where to go.

Many people on this committee and in the House share ministerial experience. Those of us who have that experience know that it is very difficult to get sanction from the Department of Finance to spend money, unless an excellent case is made. It seems to me that this issue was treated differently. The case was not made. I have outlined how this case differs from procedures in the sanction of other large capital sums, yet nobody in the Department of Finance shouted "Stop". I find that peculiar. I can see Mr. Doyle's position. However, can he throw any light on this?

Mr. Doyle

In October 2002 approval was given for the procurement of electronic voting machines for the whole country at a cost of €33 million. That was the cost in 2002. We were concerned still that we did not have the final details of the savings or a more detailed account of savings. However, the decision went ahead on the basis of the case put at the time and €33 million was the cost then.

What date was the Government decision?

Mr. Doyle

It was October 2002.

Was this not after the tendering competition? The Government normally sanctions in advance on the basis of an estimate. Then a Department proceeds to implement and goes to tender.

Mr. Doyle

After the tendering the Department did not decide to go ahead. It decided to go ahead on a private basis, but it did not go fully. My recollection is that——

Was the tender awarded? Will Mr. Callan tell us when the tender was awarded?

Mr. Callan

The overall tender was awarded at the end of 2000, in December 2000.

Then in December 2001 the Department of Environment, Heritage and Local Government gave an estimate to the Department of Finance and the decision was taken in 2002. That could not be right. I call Mr. Purcell.

Mr. Purcell

The contract was unusual in that apart from the initial purchase of the six voting machines for demonstration and research and testing purposes, there was no final commitment under that contract. It was dependent on the tests using those machines and the software being satisfactory and, importantly, the Government giving a decision to use this system at a real election. Therefore, it was not until the Department had satisfied the Government that the testing of the machines and software was satisfactory that the Government took the decision to use the system at the forthcoming elections. It was that decision a year later that galvanised the take up of the main purchasing of the machines from the contract. Initially it was 600 machines, for the first trial in the three constituencies in the general election of 2002 and then a further 400 for the second Nice referendum later that year, in October. Then it moved to the nationwide system which involved the purchase of over 6,000 machines. That was all under the same contract.

Thank you. That clarifies the situation and fills in the blanks.

Mr. Callan should have produced this document at the beginning of the meeting. Questions were asked at an earlier stage with regard to storage costs. Does he have any other documents he may or may not produce depending on the questions asked?

Mr. Callan

No, apart from my introductory statement, which I circulated yesterday and this one. I apologise for any discourtesy. We received this material only recently through the Department of Finance which deals at first hand with the recoupment of these costs.

It is fascinating reading. Who owns the voting machines?

Mr. Callan

The Department of Environment, Heritage and Local Government in a sense. It has the contract with the suppliers for the supply of these machines, the framework contract. It is an interesting question because they are placed in the care and custody of returning officers.

It should be quite clear. Does the Minister for Environment, Heritage and Local Government own the machines on behalf of the people?

Mr. Callan

Yes. We signed the contract for the supply of goods.

Does the Minister have an insurance policy covering the machines?

Mr. Callan

That is remitted to the returning officers.

The Minister has €50 million worth of hardware and Mr. Callan is telling the committee that there is not a uniform policy on whether the machines should be covered by insurance.

Mr. Callan

There is a well tried system under which people with well defined responsibilities are looking after the matters of storage and insurance.

How does Mr. Callan explain, for example, that in respect of the constituencies of Clare, Louth and Sligo-Leitrim there are no insurance costs mentioned while in other constituencies there are figures of up to €20,000 in insurance costs?

Mr. Callan

The information may not be wholly complete but in the vast majority of cases there is insurance which is looked after by returning officers.

Does it cover the replacement of machines or a percentage of the cost or how does it work?

Mr. Callan

It covers the total value of the equipment.

Let us compare Tipperary North and South with Waterford. The cost for Tipperary North and South is €8,500 and Waterford, €1,061. How can there be this discrepancy in insurance costs?

Mr. Callan

I cannot explain it immediately because these are the figures the Department has been given. Some arrangements may be more economical than others. Quotes may be obtained from brokers. All this will be audited in due course.

I suggest that if the Minister has €50 million worth of the people's property, he should have a policy with regard to its physical security. On insurance costs met from taxpayer's funds, they should be uniform and as low as possible.

Mr. Callan

Some of the places for which no costs are mentioned may have general insurance. They may be local authority or other insurance policies covering the goods involved.

In response to the generality of the Deputy's question, in introducing electronic voting and counting we did not wish administratively to turn the system upside down or dispense with the valuable partnership with returning officers in running elections. It is not on our agenda to have overnight the Custom House become the sole returning officer for the entire country. We respect the jurisdiction of and the good service provided by returning officers in constituencies and are working through that system.

Mr. Callan cannot tell me if there a uniform insurance policy and what is covered or explain why the cost of insurance in one constituency can be eightfold that in another.

Mr. Callan

The figures will be audited. The Department gave them to the committee in a spirit of co-operation. We have only just received them and they will be subject to audit in due course. This will allow us to scrutinise some of the issues to which the Deputy refers.

I suggest that some close auditing will be required. Approximately how many machines are there in the Waterford constituency?

Mr. Callan

There are over 200 but that is not a firm figure as I do not have the information in front of me.

Is it known how much floor space that would take up? Is it known why the cost is an incredible €50,000 a year? They must be housed in luxury penthouses.

Mr. Callan

All this will be audited. There is local discretion and different property arrangements. Different arrangements for the storage of equipment have been in place since before the introduction of electronic voting. It is not the case that storage costs are uniform in relation to other items of election equipment.

On an initial glance, it reflects the shambles of the electronic voting effort overall on the part of the Minister.

Mr. Callan

My response to the Deputy's statement is that it reflects a perfectly good working delegation of authority to returning officers in a way which has served the State well in the past 60 years.

Mr. Callan has not convinced me. A penthouse or shed in Waterford costs the taxpayer €50,000 a year while it costs nothing in Galway. It costs €6,000 in Roscommon. Miscellaneous costs range from nil to €472 to €11,000. What costs are these?

Mr. Callan

Local returning officers have discretion to operate on the basis of ad hoc arrangements. It is felt that in the round this is likely to be more advantageous to the Exchequer and more economical than trying to centralise arrangements. However, the Department will be guided by the results of the audit to be undertaken of these figures.

I have finished my questioning. The whole operation has been characterised by ad hoc arrangements on the part of two Ministers. That is why we are in this mess.

Next week the committee will deal with chapter 2.1, returning officers' expenses, as part of the Vote for the Department of Finance and the Deputy will have an opportunity to explore the matter again in that context.

I will lie awake for the next week in anticipation.

I wish to explore another area. Is it correct that the franchise unit in the Department of the Environment, Heritage and Local Government is responsible in this case?

Mr. Callan

I did not hear the Chairman's question.

Which unit has responsibility in this case?

Mr. Callan

The franchise unit or section is responsible for electoral administration.

It primarily has expertise about elections.

Mr. Callan

It also deals with policy on legislation.

Administration and management. Is there electronic expertise in the unit?

Mr. Callan

No. In principle, there might not be. In the main it is staffed by general Civil Service grades but there is expertise available elsewhere in the Department and the Local Government Computer Services Board, a partner body of the Department and whose services we brought into play when the project was being developed.

I note that a former chairman of the Local Government Computer Services Board was brought into the evaluation group in the Department. Is that correct?

Mr. Callan

He was a member of the steering group, not the tender evaluation group.

That is what I am getting at. Who steered it and what level of expertise was involved? While Mr. Ó Ceallaigh has had a distinguished public service career, I understand he had retired from the position as chairman of the Local Government Computer Services Board when he joined the steering committee. Is that correct?

Mr. Callan

He would have been chief executive of the Local Government Computer Services Board——

He was the former chief executive.

Mr. Callan

Yes, he was the chief executive before the present incumbent. We had a connection with people currently working with the board. We had people who were actively in service in the Local Government Computer Services Board involved with us also.

Did Mr. Callan feel there was electronic and computer expertise on the steering group or available to it at all times?

Mr. Callan

Yes.

Mr. Callan will be familiar with the report of the commission. The point has been made by experts before that this is not the first time a big organisation moved from paper to electronic systems. Banks, insurance companies and many big businesses have done so during the years. It has been argued that all big organisations would first run parallel tests before going electronic. Mr. Callan's favourite bank, in moving from a paper based to an electronic system, would have run the two systems initially and tested one against the other. Is Mr. Callan familiar with this? This is correct, is it not?

Mr. Callan

Parallel testing formed part of the preparations for this project. It was carried out at three locations, Buncrana, Laois and Carlow.

On how many votes was the test carried out?

Mr. Callan

We used the database or the number of votes cast at the relevant local elections in 1999.

Were these the votes for town councils?

Mr. Callan

And county councils.

Which county council was used?

Mr. Callan

We may not have figures here. Obviously, they are available but I may not have the number of votes here. It was mainly urban district councils.

It related to urban district councils and was in very small proportions. My point is that the Department did not do what is normal in big business and parallel test the whole system, even in a constituency, much less across the country as a whole. If a big bank was going electronic from a paper based system, from headquarters to the smallest branch, it would have run parallel tests. I am informed this is normal practice in business. The Department did not conduct parallel tests, even in a constituency. It parallel tested in a few token town councils. I do not believe that was good enough at that point. This is why I am asking about the expertise available to the steering committee. If the information I have received is correct, parallel tests should have been carried out at least in the three pilot constituencies where the machines were initially run.

The Department never tested the software scheduled to be used in the local elections, the referendum and European Parliament elections in June 2004. Is that correct?

Mr. Callan

It was subject to continual testing. It was tested all the way by Nathean Technologies as it was being developed by Powervote.

No, that is not right. That is why I wonder about the level of electronic expertise in the Department. While I am no expert, I can read reports. The reports of the commission on electronic voting and their appendices state quite clearly that when a new version of software is established by modification of an old system of software, it is not sufficient to test the modification; the whole system needs to be tested. This is mentioned chapter and verse, time after time. Several experts have stated that one of the big flaws was that while the Department had required testing of modifications of the software, people who knew the business agreed it was necessary to test the lot because modifications could kick back into existing parts of the programme already tested and cause unforeseen errors. This is a fact, yet the Department did not test. We were going into local and European Parliament elections and the Department had not tested the software as I have described. I cannot understand how that happened.

Mr. Callan

The software was being continually tested. We note the concerns expressed in some of the expert reports and the conclusions of the CEV in this matter. With the benefit of greater time we will ensure testing is carried out in a more comprehensive and transparent way. That said, the parallel tests were carried out by the CEV as part of its work. It inputted 35,000 votes in 720 machines and the counting proved accurate.

That information was not available when the Department made the decision to hold four elections on the same day. It does not help the Department's case to say the commission on electronic voting took a large sample. The Department which had responsibility for the project did not.

Mr. Callan

Returning officers had equally run mock elections which I have not mentioned here. There had been use of the returning officer network to test the system in other circumstances also.

Computer experts find fault in another area which is well documented. They claim with accuracy that the whole system was never tested from start to finish. While elements of the system were tested, including the machine, the vote storage unit and the software, at an early stage in its development, no tests were ever carried out to test the whole system in sequence as an operating model. As a consequence, this testing was also inadequate which, again, is documented. I cannot understand how this would have happened. While I understand the constraints under which the Department operates, the manner in which this project was steered, brought forward and implemented appears to have been less than adequate from the Department's point of view.

I have dealt with the audit trail, as have other Deputies. When referring to other issues, the commission stated: "The Commission also makes the following observations in relation to the chosen system which, although not falling strictly within its terms of reference, have a bearing on the successful implementation of the system at elections in Ireland". It continues to refer to the VVAT. Later in its observations it states: "The absence of a VVAT significantly raises the standards and quality of other system testing that is required."

On which page does this appear?

I am on page 77 of the commission's report, part 6.4, indent 3. Mr. Callan has legitimately expressed serious doubt about having a VVAT system and there is a debate about the matter. Mr. Callan says it is not necessary or desirable and has other difficulties. Others believe it is absolutely essential to restore confidence in the system. However, that is an ongoing debate. The commission states clearly that without VVAT, the standards of system testing required must be significantly raised. I do not believe they were significantly raised; they were slipshod.

Mr. Callan

We take encouragement from the CEV statement the Chairman has just quoted because we believe it shows VVAT is not a sine qua non of a well functioning electronic voting system. We will work with the commission on electronic voting to show that the chosen system can meet high standards of testing and quality, just as it demands.

I would have agreed with Mr. Callan when the commission's interim report was published before the elections. We have to reflect on Deputy Deasy's comments. His experience is shared by many elected members. Time has passed and confidence has been sapped. I am not too sure whether it can be restored without a voter verifiable audit trail but we will see.

I would like to move on to the final area of discussion — the security of the hardware. The Department uses 1980s technology. It can claim there is merit in using technology that is 20 years old, even though matters have advanced, because it is tried and tested and worked in the Netherlands.

I am interested in a phrase used in the computer industry — "security by obscurity". In effect, it means that the user does not know where the secure systems are. I refer to personal identification numbers, for example. As a result, the user cannot bypass the machine's safety mechanisms. Experts claim that security by obscurity is the equivalent of hiding one's door key under the front door mat. While it is hidden, one does not have to be a genius to find it. Anybody with any modicum of knowledge of computer hardware can break through. One of the academics who reports in the document I have in my hand — I think it is the man from DCU — talks about getting through in a matter of seconds and being able to develop software that can do all sorts of things.

The Department of Finance was concerned about the security of the system. The commission on electronic voting stated the level of security should be significantly higher on the voter verifiable audit trail. According to the evaluations, however, there was scarely the necessary level of security in the hardware or the machines. Was the Department aware of this when it made the decision?

Mr. Callan

I do not think that is the rounded conclusion of the commission on electronic voting on the overall deployability and security of the system. I have said the commission does not necessarily endorse or rely on the conclusions of the detailed expert work. Having said that, the Department is determined to work with the commission to enhance and improve the security of the system. The hardware has been deployed in three leading western European democracies and not attracted any great degree of difficulty or criticism in such places. The Department does not accept that the system constitutes a fundamental security risk. It accepts that it can and should work to improve the system's security dimensions in line with the commission's recommendations.

The academics commissioned by the commission on electronic voting have a different view. The best one can say about the security system is that the fellow with the screwdriver may not exist. If he does and has access to the machine, to put it bluntly, the Department will quickly find itself in trouble. We hope such things do not happen. When one is designing a voting system for national elections, which is fundamental to democracy, it is not enough to say it works in the Netherlands. In the current climate — I do not wish to refer to those who robbed the bank in Belfast — people in many constituencies would have concerns about the security of any system around election time. One would like to have an absolute assurance that everything is under control, especially when one considers the number of electronics experts who have been redundant since they stopped planting bombs.

Mr. Callan

That balanced view will be a feature of the work of the commission which is required to work towards validating the security of the proposed electronic voting system. As the Chairman said, the present arrangements may be vulnerable to insider attack or other forms of malicious intent.

There are two further problems with the hardware, resulting from the fact that it is based on 1980s technology. I am not an expert on the matter but it seems that the great advantage of electronic technology is that it is totally predictable and, therefore, very hard to randomise. While the storage system for votes on the voting cartridges or modules purports to be random, it is not. The first vote is stored on a random basis but every other vote is directly related to the position of the first. If one can establish where the first vote is stored, the secrecy of the ballot for that voting module ceases. The current machines are unable to give completely random positions but modern technology on the market is able to do so. There is a disadvantage in using 1980s technology because one is stuck with a pretend random system that is not really random. The Department has claimed in the documents it has presented to the committee that random placement of the votes in the voter module is a feature of the system.

A second problem with the technology being used is that it does not provide for encryption which is available in modern machines that use up-to-date technology. That is another disadvantage in terms of the security of the system. It would not be fair to ask Mr. Callan to respond in detail to these comments which are based on this report, the academic studies and the appendices. My remarks are not a matter of opinion but of electronic fact. There are serious doubts as a consequence. I agree with Mr. Callan in some respects.

When we go to vote, we do not expect someone to take the panel off the machine and start interfering with the voting. We do not expect people to go in and establish a voting signature. The first voter in the morning might be faced with 15 names on the ballot paper and millions of options — more options than one has with the national lottery. When one marks one's ballot in a certain way, one establishes a voter signature. If one knows what the first vote was, one can work from this to know how others voted. Mr. Callan might be right to say such events are fantasies and would not happen in practice but all possibilities have to be taken into account as we examine the security of the system. The way in which the machine is constructed does not have to be retained because technology has moved on. The biggest advantage of 1980s technology is that it proved to be successful when it was tried and tested in elections in the Netherlands and Germany. Matters have moved on in the meantime and more advanced systems are available. Similarly, there are more advanced ways of getting inside a machine to upset its security.

Mr. Callan

Every system is capable of being made better. The Department and the wider parliament are not competent to design systems. We enter the marketplace and examine the products available which have passed various industrial certification tests. We decide whether they would do a good, middling or bad job. We do not manufacture or design computer services. One can engage in an infinite amount of speculation about how one could add various layers of perfection to any system being produced or marketed. What is a reasonable standard of satisfaction, proof or perfection for a system such as this which will be used in a highly sensitive set of circumstances? We have to be realistic. One will always be able to send a prototype to other experts who will say, "I would have added on this bell or that whistle." One will obtain that sort of comment if one goes looking for it. In the end someone must take responsibility for signing off and saying that in the real world the process in question is one which can work, has adequate security precautions built in and protocols surrounding its use in terms of who will have access to it such that we can be sure of it. That is the way in which we have approached this project. To some extent, the commission on electronic voting offers a second opinion which it was right to secure on what we are doing. We accept and respect it. While the commission will no doubt bring some extra scrutiny to bear on the matter, we will never have circumstances in which we will endorse a product which is incapable of being made more perfect.

I agree. There is great merit in the way Mr. Callan has put the case. All I say in reply is that in the last six years in Leinster House Deputies have had their computers updated three times. While technology is marching very fast, the predicted lifespan of the machines is 20 years. As Mr. Callan told Deputy Rabbitte, they will serve in 12 elections up to 2021. I had a very good car in 1987 but I do not drive it any longer.

While there is merit in what Mr. Callan says, there is merit in what I say too. I am simply drawing on the appendices from the academics and computer experts who say this was not a good decision. It may be the case that nothing will go wrong. It seems the votes in the three pilot elections were counted accurately which brings me to my last question. Somebody suggested to me that a freedom of information request had been made to returning officers which yielded information that in Dublin North and Dublin West one saw surplus votes while votes were missing in the other. Does Mr. Callan have any information on this?

Mr. Callan

I do not have a detailed note on the matter but can get one for the information of the committee. There was confusion around reconciliation during the course of the elections. I believe the matter was resolved to the satisfaction of the returning officer.

It would be useful if Mr. Callan could send us a note.

Mr. Callan

Yes.

A former chairman of Disabled Ireland contacted me this morning to say he went to see one of the voting machines in the civic offices. He is wheelchair bound in a high wheelchair. He said the face of the machine was tilted, meaning that while he could reach the lower buttons from a high wheelchair, he could not reach the names of the candidates on the top buttons.

That is fair enough.

I do not know whether it was in alphabetical order. I think the letter "R" was at the top. Among the responsibilities of Mr. Callan's Department is access by disabled persons to all public buildings. If it is primarily the responsibility of the Office of Public Works, the Department is certainly involved. How could machines be installed to run elections which could not be reached by wheelchair users? The man in question stressed that his wheelchair was the highest on the market and he is a normal-sized man. It is outrageous that he could not reach the top buttons on the voting machine.

Mr. Callan

We had discussions with Disabled Ireland and other groups representing the disabled in the preparation of the project, and built in features in the design of tables which took account of requests we had received through that process. If there is more that can be done, we should do it. We will look into the issue the Chairman has raised.

Because he cannot press the buttons, he must call for the assistance of the polling clerk who is now the machine technician. He must tell the polling clerk for whom he is voting and let her press the buttons for him. The secrecy of the ballot for some disabled people is gone.

Mr. Callan

We want to minimise those scenarios. As a result of the process I described, we are using a tilt table which may not have been in use in the case the Chairman describes. Certainly, improvements have been made to the design.

Fair enough.

Mr. Callan has had a long stint but will have to accept that it comes with the territory. Eminence in the Civil Service results in this scenario. In which case did recruitment of consultants by single tender occur? What was the amount involved?

Mr. Callan

Is the Deputy speaking of the Department generally?

No, just in respect of this project.

Mr. Callan

We have had a number of single tenders in some of the specialist services but I am not sure I can identify them on the hoof. The awareness contract, a significant one, was not among them. It was fully advertised and went through the system. The report on the hardware and software testing of the modified voting machine undertaken by the German testing institute involved a single tender at a cost of €21,000. The environmental testing of the voting machine involving stresses and physical interference was carried out by TNO of the Netherlands at a cost of €28,000 and also involved a single tender. The validation of revised STV software undertaken by the UK specialist company Electoral Reform Services also involved a single tender. They are the ones of which I am aware.

Is it generally the intention that the control operative to supervise the hardware on the day will be a person redeployed from existing staff? Are there exceptions or was that agreed with returning officers?

Mr. Callan

Yes. The machine controller is a poll clerk redeployed from the previous duty of tearing off ballot papers. Previously, a person was deployed to tear off ballot papers and hand them to others.

Did that happen in the limited experience with the system?

Mr. Callan

That was the experience in the sense that the process was neutral in terms of numbers and costs.

When I asked about storage, I did not have the figures. How is the owner of the storage facility recruited? Is it a decision for the local authority, the returning officer or someone else?

Mr. Callan

The returning officer decides in every case. The machines are in the custody of returning officers who will decide on suitable secure storage arrangements.

Why does it cost twice as much to store the equipment for a six seat constituency than it does for the four seat constituency of Waterford?

Mr. Callan

As we have yet to audit the figures fully, I cannot give the Deputy a precise answer. I surmise that local property opportunities and circumstances vary across constituencies.

Does Mr. Callan know who stores the machines in Waterford?

Mr. Callan

I do not have that precise information.

Perhaps Mr. Callan will get that information in respect of some of the locations in question.

Mr. Purcell

As a point of information for Deputy Rabbitte's benefit, fortuitously, the matter of returning officers' expenses will come before the committee next week because they are paid from the Central Fund. The question of the returns and auditing that expenditure is central to that chapter of my report.

With regard to the subject under consideration today, I will briefly address two main areas. On the testing which, as I stated, has proved to be the Achilles heel, I made comments in correspondence with the committee about the need to be fastidious and go for overkill in terms of security because of the very nature of the exercise being computerised. This has been borne out by circumstances. If there was a failure by the Department, it was in that area.

To a certain extent — no pun intended — the Department was lulled into a false sense of security because of the success of the two pilot studies. Retrospectively, even the commission on electronic voting has validated these two exercises in terms of vote counting and so forth. If the Department had to do it over again, it would probably say it needed a broader based steering committee, even though the one departed had expertise from a sister organisation. From the point of view of it clearly being independent, it would have been preferable to have had a broader based steering committee involving the expertise later made available to the commission on electronic voting.

There was also a problem with the timescale, as the Accounting Officer stated. This precluded matters such as a full dress rehearsal and so on which was also a factor. I have some sympathy for the Department because even the computer systems of the Pentagon are being hacked into despite the extensive expertise available to that establishment.

On the matter of cost benefit, this was not done as rigorously as it should have been, as borne out by some of the members' questions. We are talking about technology and so on and having a system in place for 20 years. I mentioned — I believe in response to a question from Deputy Michael Smith — that there was a very large "if" attached to cost savings, particularly when the period in question is 20 years. Security developments will inevitably be introduced in the area of recognition, be they voice or retinal recognition and on the like. We see, with regard to passports and other areas, that we have moved to a stage we did not envisage some years ago. There is a large question mark about the cost savings and benefits set out.

That is all I have to say. That is as objective as I can be as a dispassionate observer of this episode who has no political interest in it. I can say I voted in both pilot studies and found the system easy to use. At least, my confidence in it has not been dented.

How many times did the Comptroller and Auditor General vote?

Are there any final observations?

Mr. Callan

I thank members of the committee for their courtesy. To get a little above the battle, as a public service organisation, we were involved in innovation and pushing out the frontiers of e-government. While there is more work to do to validate and secure public buy-in to the system, we should not be penalised for trying to make changes. That would send a bad signal to the public service generally. Equally, we must take care. The project is proceeding in a modified way which the Oireachtas has mandated and with the full involvement of the commission on electronic voting.

I thank Mr. Callan, his officials and the Department of Finance. We did not get an opportunity to state the Department of Environment, Heritage and Local Government has made a major contribution to public life. Through the local authorities, it has significantly raised the living standards of generations. Even since I entered politics in the 1970s the work of the Department, through the local authorities, has been excellent. I would not like the officials of the Department to leave without acknowledging this and noting its excellent record. In the area under scrutiny we have a difference of opinion. We wish the officials well in their ongoing work and hope the issue can be satisfactorily resolved.

Is it agreed to dispose of the chapter? Agreed. The secretary will talk to the officials in due course about coming before the committee to discuss the Vote.

Mr. Callan

That is fine.

The witnesses withdrew.

The committee adjourned at 3 p.m. until11 a.m. on Thursday, 10 February 2005.

Barr
Roinn