General Election 2015: The year pen and paper dies?

This upcoming UK general election may be the last where pen and paper are used for vote casting. In some ways, it is a little strange that electronic voting is not the default already. But then the stakes are high and there is a question of an audit trial when it comes to computer voting.

At least with paper voting, you can more easily keep a separation of vote cast versus individual. Security experts also acknowledge that there will always be some degree of fraud with electronic voting.

Nevertheless, the UK is seeking to trial online voting in 2020. It has already happened with local elections and other countries like Estonia have successfully trialled online voting; in that instance for their general election back in 2007.

The case for online voting

The 'no brainer' is reduced costs. A report from campaign group WebRoots Democracy states that online voting would cut the cost per vote by a third to £2.59 while also reducing the number of accidentally spoiled ballots. Turnout should also increase but not as dramatically as one might expect. Estimates are still less than a five per cent increase.

Electronic voting can be used to facilitate, improve and ultimately extend the exercise of democracy. The underlying core principle of democracy is an informed and engaged citizenry. Most governments get passing marks for “informing” citizens via digital communication however the vast majority has a long way to go to “actively engage” citizens or to effectively exert global influence using digital media.

With e-voting, citizens can vote from anywhere. Is this important? It is when you examine the most common reasons people cite for not voting which include "not being able to take time off from work/school," "family emergency," "out of town," "had no way to get to the polling station" and "lines too long at the polls."

The people who claim these reasons represent the majority of those who did not vote. Logic follows that they could have voted if an electronic internet-based means was available. Of course there are inherent problems that could arise in such an electoral system.

What if?

What if, for instance, an electronic vote is infected with a virus so that when a voter 'opens' it to vote, it then proceeds to vote for its candidate of choice and even disguises the fact to the voter who is none-the-wiser after the e-vote?

This is the nightmare scenario for e-voting as any election system must separate a voter's choice from the identity of the voter to protect ballot secrecy; it follows therefore that the voter would receive verification only that the ballot had been received and not what the actual choice was. The content voter thinks the correct choice has been recorded when in actual fact someone has ‘stolen’ their vote. There is no trivial way of detecting such a ‘theft.’ In this manner, elections could be manipulated wholesale if the virus author was successful in infecting sufficient numbers of computers.

Similarly, if the voting is cast through websites, those sites could be spoofed to reveal personal identification numbers and passwords of voters, then the vote could be automatically recast with those values to a different candidate.

The servers hosting the electoral sites could also fall victim to various denial of service attacks. It is a very daunting prospect to have an important nation-wide event using an inter-networked system that has to be secure from attack. I am not saying it cannot be done but perhaps great thought needs to be put into a national system.

What would it look like?

If it can be done, what would the process of casting a vote be like?

First there is the matter of registering people to vote. Since all individuals in most tax-paying countries already have unique ‘keys’ (e.g. National Insurance numbers in the UK), these could be used as an entry to login during an election. As soon as a person comes of voting age, their unique key could automatically be activated to allow them to vote. This would eliminate the process of registration altogether.

If every citizen of legal voting age is automatically registered, the first pro-active thing the citizen would need to do is obtain a password so that others could not simply run a brute-force programme to enter every possible unique key and vote on others’ behalf.

This password could be sent on the legal-voting-age birthday of the citizen. Once logged-in, the voter would simply vote and the voter’s effort would then be finished until the next election. There would be no need to keep tabs on voters’ addresses and have voters re-register every time they change their address.

I am not advocating this as a method to follow but rather more as a simplistic overview of the issues involved. Building a secure electronic voting system is difficult. The US Pentagon dropped their proposed online voting system which would have given overseas military personnel the opportunity to vote in the elections in 2005, citing the inability to ensure the legitimacy of votes as the reason.

Beyond the matter of voting to elect representatives, the digital age raises another important and interesting question that could be asked. Since we elect politicians to vote on our behalf because it’s inconvenient for all citizens to gather to vote on every issue, why could we not, in this digital age now vote on every issue ourselves?

Representatives could still be elected to propose laws (and a means should exist for non-representatives to propose laws if those proposals are given enough support), but the voting could be carried out electronically by referendum every week.

In this scenario we could expect citizens to be more aware of every bill, as they would have the opportunity to vote on every issue. Giving citizens this power would probably do more to beat voter apathy than any other proposal out there, but perhaps this is a bridge too far for the moment.

Dr Kevin Curran is a senior member of the IEEE, Reader in Computer Science at Ulster University and group leader for the Ambient Intelligence Research Group.