When the fictional character Tom Dobbs (played by Robin Williams) was elected as President of the United States in the 2006 movie, Man of the Year, it played on the fears of people across the world when the subject of electronic voting is raised. In the movie, the Delacroy voting company won the contract to provide voting machines for the Presidential election but a software glitch delivered the winner based on spelling rather than votes.
Across the world, only fourteen countries have dabbled in electronic voting and, of those, it is only widespread in four.
With the greatest respect to the various commentators and election analysts, one of the most attractive components of electronic voting is the instantaneous and incredibly accurate method of delivering a result. In recent Federal and State elections in this country, after watching hours of analysis by a variety of experts and witnessing numbers dribbling through into tally rooms, we all went to bed on the Saturday night not entirely sure who was going to be in charge. A further attraction is the dream that we won’t waste tonnes of paper for candidates to print beautiful glossy brochures to hand out to voters as they make their way through the scrum of volunteers.
The positives are obvious for all to see so why is it that we are reluctant to go that way. Australia in particular, with residents known for being early adopters of technology, logically would embrace electronic voting.
The fear of security breaches seems to overcome the attraction of the advantages. There may be other reasons our governments don’t head down the path with electronic voting but security is the main issue put forward.
The US has long had a variety of solutions to this problem and, via DARPA (Defense Advanced Research Projects Agency) they have recently awarded a $10 million project to an organisation to create an open-source voting machine that could prevent hackers from tampering with votes. The first real test occurred at DEF CON 2019. DEF CON is the world’s largest hacker convention with over thirty thousand people in attendance. At the Voting Machine Hacking Village setup in DEF CON since 2017, hackers typically take a few minutes to expose vulnerabilities in voting machines. At the 2019 event, the latest hardware from DARPA was on display at DEF CON and, ignoring a few minor setup issues, no hacker was successful in breaching the voting machine.
With Trump’s presidency still plagued by rumours of Russian interference, DARPA is keen to produce a system that will allow lawmakers to pass an election security bill in time for the 2020 presidential election.
I would also suggest there is a large potential return on investment if they manage to produce a secure system. Once security was removed as an issue, governments around the world would be interested in licensing the technology to use in their elections.
And maybe there is something there for a progressive Australian government (oxymoron of course) to look at. It is common knowledge that the CSIRO has generated over half a billion dollars for Australians from its Wi-Fi patent. I would propose that we could task the CSIRO with creating a secure voting methodology, patent it and use it in elections in Australia. Licensing that technology to other governments around the world to use in their elections would surely generate significant revenue for our government?
Perhaps I am over-thinking the first step though. Much of the focus for electronic voting is on securing the process for remote voting. I would propose a staged process to introduce electronic voting.
Stage 1. We continue to use voting booths and have people physically visit an election booth on election day with one significant twist. Instead of printing reams of paper and asking voters to write on those pieces of paper, I propose that we introduce electronic voting terminals into all polling booths. When you enter a polling booth in the normal way, you need to produce ID and you are electronically recorded as having voted. This immediately is more secure than the current method. Currently you don’t need to show ID and you could enter every different polling booth during election day and your name would be ticked off in a book. After the event you may be contacted and told you voted more than once but with no ID check, where is the proof that you actually did vote more than once? Once your name was marked off, you go to an electronic terminal to vote for the candidates. Being electronic, you would receive a warning if your vote was not completed correctly. If you are going to make the disappointing deliberate choice of voting informally, you can still choose to do so but accidental informal votes would be a thing of the past. You finalise your vote on screen and ‘submit’ the vote. Once it is entered, two pieces of paper print out from your terminal. These two identical printouts have two pieces of information on them. Firstly, it shows how you voted. Secondly, it creates a unique random number on the voting slip. As you leave the polling booth, you drop one slip into the ballot box the same as you would do now and you put the other piece safely in your handbag or wallet.
At the end of the election day, all of the votes are already recorded and it would take mere seconds for all votes and preferences to be counted. The results would be known immediately after the election. If there was any suspicion of a breach of the computer system, the physical pieces of paper would be on hand to allow a physical count to take place as well. Furthermore, a Web site would publish ALL votes on a Web site with the unique random number identifying how each vote was cast. There is nothing to identify an individual by name with those details but any voter who was concerned about the legitimacy of the process could look up the information on the Web site and check their unique random number against the piece of paper they took with them from the voting booth. This would satisfy each individual that their vote was accounted for correctly.
At this stage the main advantage gained is that the votes are entered quickly with less chance of fraud and the count is instant. Keep in mind that with the current physical voting scenario, all of the information is entered (by humans) into computers and then they are instructed to perform the count.
Stage 2. Once the voting public was familiar and comfortable with this method, the process could be extended to remote electronic voting. Using the same processes – ID check to access a portal then voting and printing a unique identifier etc. it would allow people to vote from the comfort of their homes – or from anywhere in the world, in fact.
We are now at the stage where we are comfortable with our entire life savings being accessible from our smartphone and our lives seem to be contained within that little device in our hands – yet we don’t seem to be able to vote electronically.
I am sure we can overcome this problem. It just needs the political will to make it happen.
Tell me if you would like to vote electronically at ask@techtalk.digital.
Mathew Dickerson