Speaking in the Seanad on 21 February 2001, on the 2nd Stage debate on the Electoral (Amendment) Bill, 2000, I said:

I wish to refer to the substance of Part 3 which makes provision for electronic voting. Senators who are aware of my enthusiasm for the information age will no doubt expect me to welcome this development with open arms. While I welcome it, I do so with caution as I am not convinced that the implications and practicalities have been fully thought through. 

I wish to look at the issue from two perspectives, first, a conceptual viewpoint and, second, the practical viewpoint of making it happen in real life with real people. Both perspectives offer food for thought.

At a conceptual level, what we are talking about is not a shift from paper-based voting to electronic voting. What is really at stake is a move from voting with an audit trail to voting without an audit trail. That is the basic difference.

Let us consider the present system, old fashioned as it is. Its key characteristic is the physical existence of a vote on a piece of paper which is tracked at every single stage. Before a vote becomes a vote, it is counted in the sense that the number of ballot papers is carefully controlled. The issue of each piece of paper is recorded and takes place in full public view.

A vote becomes a vote when a voter records his or her choice on that physical piece of paper. The voter can look at the paper and handle it before giving it up to confirm that the paper reflects his intentions.

The fact that the vote is secret does not mean that it disappears from view. It is counted going into the ballot box and again when it is being taken out and it is guarded while it is in the box. The counting takes place in public so that everyone can see it and if there has to be a recount the physical evidence is there to do it.

The entire process is so open and visible that the voter has total trust in the integrity of the system. To the best of my knowledge nobody has ever questioned the integrity of the balloting. Any accusations of fraud generally centre around the issue of personation, using a dead person's name etc. No one has sought to question the integrity of the ballot process.

The fundamental reason for this is the existence of a physical audit trail from start to finish. Our unquestioning trust in the system rests on that foundation.

I physically create my vote by marking the ballot paper and from that moment on the ballot is handled in a way which virtually eliminates the possibility of fraud or error. There is no fraud because the ballot is watched over, while there is no error because the raw material is there to recount if there is any question of doubt.

I want to compare that system with the one being put in place.

Under the new system a voter will indicate choices by pressing the relevant parts of a touchable screen - I assume that is how it will work - and these presses will be electronically recorded on a medium such as a cassette or cartridge.

The voter will not see the record and he must trust the technology to ensure that his voting intentions are fully and correctly recorded.

There is no feedback, confirmation or guarantee that the voter's intentions have been properly translated into digital impulses.

There is not even one of the notorious "hanging chads" to chew over - I only heard the term for the first time in December in regard to what happened in Florida - because there is no punchcard for a chad to hang on to.

Under the new system the moment of truth, the instant of voting, will disappear into a black hole and, to the best of my knowledge, it can never be retrieved from it.

The Bill provides for detailed regulations for the care of the recording medium before and after the voting has taken place. The regulations will ensure that the counting process can be done repeatedly.

However, it does not provide for any means - I stand to be corrected - to allow us to inspect whether the translation of the voter's intentions into digital impulses was properly carried out.

The new system will require an act of faith that the current system does not require.

The vast majority of people will make this act of faith but whether they are right to do so is another matter. Nobody who has ever been involved in the commissioning of a major computer system would put much faith in any system ever getting it perfectly right on the first occasion.

Developing software is notoriously a matter of trial and error. However, in most projects it is possible to build in space to undo mistakes. The Minister spoke with confidence about what will happen, but I wonder whether that space will be built into this project.

In a recent article in The New York Times, Edward Tenner, a researcher at Princeton University, stated:

"Many independent voting security specialists doubt that any electronic system can be both secure and anonymous".

He was making the point that the audit trail an ATM machine creates as a back-up against fraud is not possible in a secret ballot. Maybe we can prove him wrong, but this is also my concern. Electronic voting systems are, therefore, inherently open to fraud since a fraudster could easily bury among several thousand lines two lines of code which would affect the way votes are recorded.

I am not so worried about fraud as I am a trusting person and am sure nobody here would attempt this. I am, however, scared stiff about the possibility of mistakes being buried in the software.

Let me shift perspective and raise a totally different set of concerns from the practical, human and usability dimension.

For more than 25 years, I have sat down once a week with a group of customers who very generously tell me what my company is doing wrong. They tell me about small practical issues relating to trolleys, queueing and the way in which goods are displayed and offered.

Whatever some people might think, I am not totally stupid. One would think that after 25 years I would not have to go back to customers week after week and ask what are we doing wrong. However, the truth is we keep making mistakes all the time.

We keep making mistakes because it is virtually impossible to anticipate exactly how things will work out in practice. We design something using the best knowledge available to us yet, from a user's perspective, it always looks different. Matters we think might cause problems turn out to be no problem, other matters where we see no problem turn out to be real sticky points.

I say this to emphasise just how steep the learning curve will be for voters during the transition from the present voting system to electronic voting.

We need to have the system thoroughly tested from a usability point of view and we then need to invest massively in educating people to come to terms with the new approach. The approach will neither be quick nor easy.

We must introduce electronic voting in a way that will make it equally acceptable to everyone, whether young, old, a technical whiz kid or mechanically averse. I am old enough to recall when ATMs were introduced. Older people would not touch them and would prefer to go inside the bank to cash their cheques, as they had always done. A number of them still do that. I believe there will be no choice in this instance. We expect people to adopt the new technology regardless of whether they like it. All the more reason, therefore, to invest in getting it right the first time.