Technology Guardian reports that the London Assembly has agreed that the votes in the 2012 Mayoral and Assembly elections will as in the three previous GLA elections be counted electronically.
I have to admit that I am rather ambivalent about this. There is something that brings home the reality of the electoral process with traditional manual counts: the ballot boxes being opened up and emptied in front of observers appointed by each of the candidates; the ballot papers being unfolded and counted to verify that the number of papers issued matches the number in the box; the separation of the papers into piles for each candidate; the piles being counted into batches of fifty; the batches for each candidate being placed together on trestle tables; the adjudication of disputed papers; the candidates and their agents being summoned by the Returning Officer; and finally the declaration of the result itself. It looked and felt transparent. The votes themselves were tangible and real. The process was being scrutinised and checked: the result becoming clear as the bundles piled up and when it wasn’t clear the wait for the final tally (and possibly the request for a recount).
Electronic counts do not feel the same.
I remember my count in 2000 in a sports hall in Harrow. The promise had been that the electronic count would be quicker – with results declared three or four hours after the polls closed. I remember the count dragging on – and on – and on. There was nothing to do. The machines churned away, but you couldn’t see the ballot papers themselves, as they went through the machines face down. By 2am, the heating had gone off. By 3am the refreshments (only tea, coffee and biscuits anyway) had run out. By 4am most people had gone home – leaving a small core of supporters around each candidate. At around 6am the Returning Officer summoned the candidates and agents to say:
“When I press this button, we will find out who’s won.”
I had been confident throughout (although having by then been up for 24 hours I was virtually catatonic as well). It wasn’t until much later that I discovered that my agent had notified Party headquarters eight hours earlier that, on the basis of the turnout and canvass returns, her assessment was that I had probably lost. In the event, I was elected – but by a margin of less than 1% of the electorate.
It also turned out that the Brent and Harrow count had been the quickest of the fourteen GLA divisions. The final results (and therefore the Mayoral result and the outcome of the top-up list) took another six hours: ten hours later than predicted and fully fourteen hours after the polling stations had closed.
Four years later, I instinctively knew that I had lost. The count was held over until the Friday and I spent the morning packing up my office in City Hall. Again the Brent and Harrow count took about eight hours (although inexplicably the machines in some of the other divisions seemed to run rather more quickly this time, so my result was not the first). As before, there was nothing to do and nothing to see. (The count – along with those for two other divisions – was held in Alexandra Palace, so at least there were refreshments this time.) If anything, the process was even more anti-climatic than four years earlier. No button was ceremonially pressed. Instead, the Returning Officer simply appeared with a print-out in his hand – and this time the margin was less than 1% of the electorate the other way.
I wasn’t present at the 2008 count. I gather this time the machines were programmed to produce running tallies. While these were not publicly announced, the net effect was to remove whatever dramatic tension there might have been. The turnout was higher, so the counts took even longer than in 2000 and 2004. I think the Brent and Harrow count took about eleven hours and the final Mayoral result was not available until nearly midnight – 26 hours after the polling stations had closed and rather beyond the bedtime of the youngest children of the new Mayor who had been kept up to watch their father’s acceptance speech at City Hall.
Electronic counting is – as I have described – not any quicker than traditional manual counts, provided enough tellers are engaged.
In my view, electronic counting is less transparent. Everyone has to take it on trust that the machines are scanning the ballot papers accurately – if the machines were programmed (or hacked into) to count every fortieth vote for candidate A for candidate B, it would not be apparent to those present and would be hard to detect.
I want to emphasise that I am not for one moment suggesting that this might ever be the case. However, the point is that in an electronic count there would not be the army of people, like those who are involved in and take part in scrutinising a manual count, to ensure and be seen to ensure the electronic count’s transparency and accuracy.
Above all, some of the election drama is lost by electronic counting. And that drama is itself – at least as far as I am concerned – an essential part of the democratic process.