Blog Post

It’s time to take election voting online

With the 2012 Presidential election only days away, we’re reaching the culmination of a long and hard-fought campaign—one that has been distinctly 21st century, thanks to the ever-increasing role of technology. Consider just a few of the ways that it is altering the political process: SMS is being used to reach voters and seek contributions; online data is being put to work to target ads; and Twitter has seen incredible traffic during the recent Presidential and Vice Presidential debates. How strange then that the actual process of voting is still rooted in analog technologies of the past.

Our current voting process has a recent history of nagging problems: from hanging chads, to provisional ballots, to voting machine malfunctions, to incredibly shrinking fonts, to accidentally purged voter lists and the still-yet-to-be-proven claims of voter fraud. These issues continue to haunt us, despite the existence of technological solutions that could eliminate many, if not all of them. Improving the process—while engaging more voters—should be a priority.

Voter participation in U.S. Presidential elections has changed dramatically over time. Interestingly, it peaked in 1876, when nearly 82 percent of eligible voters cast ballots. (Of course voter eligibility in the 19th century was very different than it is today.) As the chart below illustrates, over the past hundred years the percentage of eligible voters who have cast ballots has dropped precipitously since 1876, and typically hovers between 50 to 60 percent. How then might voting be made easier, more accessible, more efficient and more secure?

This answer may seem counterintuitive, but it’s time to take voting out of the booth and move it to the computer, smartphone or tablet—bringing digital devices and democracy together just makes sense. According to research from The Pew Internet and American Life Project, ownership of digital devices among adults is high and will almost certainly continue to grow.

Device ownership is especially high among young people, who also happen to be the segment that is least likely to vote. Supplementing the voting booth with the types of digital devices so many Americans already rely on and are comfortably using could truly democratize the electoral process by allowing voters to cast their ballots whenever and wherever is convenient for them. (This is already the case for early and absentee voting with paper ballots here in the United States and online voting is being explored in several countries.) Without suggesting a causal relationship, if the ballot were to become accessible via, say, an app or even SMS , it would only be logical to expect voter participation levels to rise.

Obviously there are questions about implementation—some technical and some procedural—that need to be considered. Here are four big picture challenges, as well thoughts on possible ways to address them:

Voter Eligibility Voters would need to state their intention to vote via a device and register that device with election officials prior to using it to cast their ballot. As with early or absentee ballots, digital ballots would need to be cast within a specified time frame. Voters would also need to receive email or SMS confirmation that their vote has been received, in the event questions were to arise later.

Voter Verification Strong authentication would be needed to ensure each voter cast only one ballot. This could be accomplished with biometric security, two-factor authentication or device recognition/registration.

Device Verification There would also need to be a way to ensure a single device wasn’t responsible for an inordinate number of votes. Device recognition coupled with user verification or a device registry could limit the potential for this occurring.

Voter Privacy While ensuring an individual is casting only one vote and that a single device isn’t casting multiple votes, a very real and important concern is maintaining the secrecy of every voter’s ballot. One can imagine solutions (such as creating dual anonymized hashes, one for the voter and the other for the device, or encrypting voter information once the ballot is entered, etc.) that would allow a vote to be cast without tying it to a specific individual.

None of these issues are insuperable. On the contrary, with current technology and ingenuity they could be addressed and make digital voting a practical alternative. Which is also to say that for the near future it wouldn’t become a substitute for all current analog voting systems, as there will always be people that need or want the experience of going to their local polling station and casting their vote in person. Providing a simple, secure and readily available digital alternative  however just makes sense in this day and age.

Ori Eisen is the founder, chairman and chief innovation officer of online security firm 41st Parameter.

8 Responses to “It’s time to take election voting online”

  1. Electronic voting opens the door to massive fraud, computer hacking, fixed elections, and worse. A quick visit to will detail some of the scams attempted to change the outcome of elections. The person above who said we need to go in the opposite direction is right. We need an irrefutable paper trail when elections are contested. You can’t count electrons when wrongdoing is suspected.

  2. Robert Luand

    This I very interesting concept, but I would worry about security. The idea of using two-factor authentication on your cell to telesign in your vote in would also ensuring an individual is casting only one vote and that a single device isn’t casting multiple votes.

  3. Sorry, I think we should go in the opposite direction. More paper, more polling places, hand counts visible to everyone. It’s a construct of mass media and their “treat elections as a sporting event” attitude that we need to know who won the election before polling is complete.

  4. I can’t help but feel that all of the security concerns are overblown. Online banking (and, indeed, online shopping) works just fine for untold billions of transactions, and I’d certainly be more upset over money being incorrectly withdrawn from my bank account than I would be over my vote being accidentally miscounted.

    In order to be effective, stealing votes (even more so than stealing money, despite the fact that stolen credit card numbers are sold by the thousands) has to be large scale and organised, which makes it much easier and more sensible to pursue it at that level.

  5. Brad Hill

    “As the chart below illustrates, over the past hundred years the percentage of eligible voters who have cast ballots has dropped precipitously since 1876…”

    The chart does not illustrate that. It illustrates a precipitous decline from 1876 until 1924. Since then the voting percentage seems to have plateau’d or possibly exhibited a rising moving average. Not disagreeing with the article, but I don’t see a reason for an assertion which is unsupported by the supporting graphic. :)

  6. I agree that it’s past time for us to move voting online. The only way to make it secure and trusted would be to make the code open-source. Otherwise people will always be suspicious that the election could be bought — much like the issue with the voting machines in Ohio right now. Device identiers can easily be spoofed and probably shouldn’t be a part of the security equation.