The Irish Labour Party has produced an excellent report on the flawed introduction of electronic voting in Ireland. Shane Hogan and Robert Cochran, both labour members and IT experts, show with elegant precision just what is wrong with how e-voting is being introduced in Ireland.
It is depressing that in Ireland we often wait till a public policy has been introduced and discredited/heavily criticised elsewhere, and then implement it ourselves, taking no heed of others’ criticisms and the obvious problems. Tower-block public housing in the 1970s and the current push for healthcare centralisation come to mind. So too with e-voting.
All the problems that have arisen elsewhere in the last few years are present in Ireland, but with no clear plan to address them;
– absence of proper system testing, meaning the first ‘test’ will be next summer’s local and European elections.
– lack of paper trail for votes
– no formal, specific ‘on the day’ security measures for the software or voting computers.
It seems that in the rush to punch our scorecard on e-government, some basic questions have not been answered;
– is electronic voting better (i.e. more accurate, reliable, faster, cheaper) than hand-counting?
– if it’s not better on all fronts, then are the trade-offs worth it?
– if so, then what are we doing to make sure it really works?
The answers so far would seem to be ‘not sure’, ‘maybe’, and ‘diddly squat’. Hardly the ringing endorsement we should need for fiddling with tried and tested democratic processes.
The Labour report suggests that there won’t be any actual cost-savings on e-voting, because each machine will require a supervisor (what that does for anonymous voting, I couldn’t say.). It’s probably too early to say, but it looks like any cost savings may be marginal. So we’re not saving money. We might even lose some. What else are we going to lose?
I don’t know how other countries run things, but in Irish elections, the votes are counted in public. It’s actually quite an event. In a public hall, the votes are counted by the official counters, sitting at a large table which is cordoned off but visible to the public. Anyone can go and stand within three or so feet of the counters and count over their shoulders. And they do. Each political party or candidate makes damn sure to have their own ‘tallymen’ doing their own count. It means that at the end of the day (or more like 2 in the morning), everyone knows just how many votes and transfers each candidate won, and democracy lives to fight another day. Tallies are also an invaluable source of information for political parties and candidates, as they can let you know exactly where and how many your and your opposition’s votes were. They’re also extremely handy if your electoral system is the rather complex proportional representation with a single transferable vote in multi-seat constituencies. Multiple counts, eliminations and transfers are usually called for, and the wealth of experience and knowledge that passes from hand to hand goes a long way to ensuring the system works.
Foul play still happens of course, but it is easier to ferret out because you don’t have to be a statistician or a computer genius to spot it.
So again, the question, what do we gain with e-voting? Not to be too much of a luddite about it, but I don’t see that current e-voting systems add much to the democratic mix. They introduce what Mr. Ashcroft might call ‘unknown unknowns’, reduce transparency, have a marginal effect on costs, and create a whole new host of security concerns.
Shane Hogan and Robert Cochran propose a number of recommendations for the successful Irish adoption of e-voting. I support them all, and add that any such system should be independently run by an agency that is fully compliant with the security standard BS7799 / ISO 17799.
But above all, I would love for someone to tell me – ‘what is so wonderful about e-voting that we ditch a system that works, and works well, in favour of an unproven unknown?’
{ 3 comments }
Anonymous 12.08.03 at 7:39 pm
Ashcroft?
dete 12.08.03 at 8:55 pm
I had this exact discussion with a co-worker just the other day. I was talking about how it should be relatively easy to build a system with a paper trail, etc. and he asked me why we should go electronic at all. I’ll admit I was stumped. In Canada, all counting is done by volunteers (the theory being that any party with a non-trivial percentage of the vote can certainly muster up a volunteer or two for each electoral district). So electronic voting only adds cost (the only potential argument I can see).
Frankly, I don’t think that there is a good answer to your question. I think that — like myself before talking with my friend — folks just assume that it has to go electronic just like everything else. No one thinks to question why. Just imagine if Florida didn’t have any electronic voting equipment; we wouldn’t have had the butterfly ballot contoroversy or the chad controversy. And nothing was saved since they had to do hand counts in the end anyway.
You can pry my laptop from my cold, dead hands, but I think putting an X on a piece of paper next to my preferred candidate’s name is a fine way to choose government.
dondi 12.09.03 at 12:57 am
My own feeling – based purely on the anecdotal evidence of working as an election judge in an election that was partially electronically counted and partially hand counted – is that electronic systems are almost certainly more reliable. The possibilities of accidentally pulling two ballots instead of one (and thus missing a vote), or mistakenly viewing a stack of ballots as already counted, are great enough in hand counting that I have more faith in an electronic system. It is also clear that hand counting of a full U.S. ballot (with all of the different offices, ballot measures, etc.) would require a substantial time investment on the part of official counters. Assuming that official counters are paid employees (even though they are typically not well-paid), there would be substantially higher day-of-election expenses associated with hand-counting (although whether those costs increase enough to justify the use of moderately expensive voting systems could still require analysis)
Thus, I would slightly revise the answer to the first of Maria’s questions to “Unsure, but probably so”. Of course, that still leaves the problems tied to her second two questions, for which I would not quibble with her “maybe” and “diddly squat” answers.
Just to provide a point of comparison – in my experience in the State of Minnesota, the polls had to be staffed by a mix of people from each party. When we did the hand counting of ballots – one representative of each party sat next to each other, and counted the ballots together. This sounds like it may be slightly more costly than the Irish model (there are 2 official counters on the time clock – rather than just one overseen by volunteers from the parties), but should yield results of comparable reliability.
Finally, the question of how many hand counts were actually done, let alone should have been done, is fraught with some contention. However, dete is slightly incorrect – hand counts were not done in general across Florida (I believe that they were only conducted in three counties, and at least one of the three (Dade) did not complete the hand recount. I am not certain, but I seem to recall that in Dade county, none of the hand recount was included in the final official totals, since the Secretary of State would not allow a partial update.)
Comments on this entry are closed.