THE scandal of the hanging chads in Florida that decided the 2000 presidential election sparked an electronic revolution. For this year鈥檚 race between George Bush and his challenger, whoever he may be, the hope is that touch-screen voting machines will ensure a fair election and restore confidence in US democracy. But are these 鈥渆-voting鈥� machines the answer to free and fair elections? Not as they exist today.
At the heart of the Florida controversy was the issue of voter intent: election officials spent hours peering at ballot papers trying to decide if hanging chads were real votes. At least they had something to peer at. With e-voting machines, once a ballot is cast and stored in computer memory, any information about the voter鈥檚 intention is lost. If election officials were to order a recount, either because of a breakdown or suspected fraud, there would be no way of knowing if their machines had accurately recorded the votes.
It is time to ensure that e-voting machines are viable before they are adopted by democracies around the world. The machines have potential advantages (see 鈥淐an we ever trust e-voting?鈥�). But trusting technology blindly, especially in elections that will change the world, is foolhardy. E-voting systems should at least print paper ballots that a human can read and verify. This paper trail should then be kept separately as an auditable record. There are other sensible safeguards. The machines should be tamper-proof and 鈥� to avoid hacking 鈥� should communicate with central servers only via private, secure networks.
Advertisement
Then there is fraud. We know from banks that computer fraud is not uncommon and is mostly committed by insiders. It would be naive to think that elections will be immune to similar interference. People are also right to be concerned about the political affiliations of companies that develop e-voting systems.
One ingenious way to protect against fraud is, curiously, to make things as open as possible. Australia鈥檚 e-voting software is open source, examined by a diversity of experts of differing political persuasions. If the software stays in the public eye, any changes will be spotted easily. This is the nearest equivalent to counting ballots in an open, transparent way.
In the frenzy to go electronic, a couple of basic issues seem to have got lost. First, trust is everything in elections, and must be preserved by whichever system is chosen. The second is a question: what is the appropriate level of technology to deliver that trust? New Hampshire chose optical scanners that read (and count) voters鈥� forms before they leave the polling station. If a form is filled in wrongly, the machine tells the voter, who tries again. This enhances trust in a way that may not be possible with touch-screen machines. It is not too late to think again.