What impact does digitization of electoral process actually have?

The mayoral election is around the corner in my hometown of Montreal, so I wanted to confirm whether I was registered to vote at my current address. What I naively assumed would be a simple task turned out to require a trip to a local community centre. I gave my name, they flipped through a binder, and gave me the thumbs up within 3 seconds of arriving. It felt like the ultimate “this could have been an email” moment and I lost two hours of my day confirming something I should’ve been able to check online in those 3 seconds.

Technology isn’t a silver bullet, and I’m far from a techno-optimist, but it can be a powerful vector for improvement. When designed carefully, it can increase accessibility, surface critical information, maybe even make it easier for people to engage civically in this context. For a long time, I assumed the main reason we hadn’t digitized more of the democratic process was good old-fashioned bureaucracy, a sluggish, underfunded government with outdated tools. I held a fundamental belief that if we simply made it easier to vote by letting people cast ballots from their phones or laptops more people would! More access = more engagement.

So I decided to investigate some interesting examples of places where they’ve managed to digitize their electoral process and see what the results were!

Markham’s digital experiment

We’ll start with an interesting homegrown example. In 2003, the City of Markham, a suburb north of Toronto, launched one of Canada’s first pilot projects to enable online voting in a municipal election. The initiative was small at first but had the ambitious aim of testing whether residents could safely and securely cast a ballot from their home computers!

As is often the case with public sector innovation, the city didn’t have the money or in-house expertise to develop its own secure digital voting infrastructure so it turned to a third-party vendor called Scytl, a Spanish company that specialized in using cryptography to secure elections. The way the system works is each eligible voter receives a Voter Information Letter by mail, this includes a unique Personal Identification Number (PIN). Then during the advance voting period, Markhamians? Markhamites? People from Markham would go to Markham’s election website, hosted by Scytl, enter their PIN, and authenticate their identity with personal information (such as date of birth and postal code) matched against the city’s registry. Once verified, they could cast their vote through a digital ballot interface! And voila, the vote was then encrypted and transmitted to a secure election server, where it remained sealed until game day(election day).

Seems straight-forward enough right? And, cherry on top, according to city officials, online voting participation increased in each cycle, particularly among seniors and people with mobility limitations! But here’s the kicker: despite the relatively simple process, turnout didn’t budge. Most of the people who voted online were the same people who would’ve voted anyway.

Unfortunately, Markham’s experiment became a reference point across Canada informing other municipalities’ approaches to digital voting. Also more importantly, the third party system that Markham was relying on, Scytl, eventually filed for bankruptcy in 2020, citing over €75 million in debt. Causing us to ask the question: what happens when the company holding your digital ballots no longer exists? Without money to invest in an internal system and having to rely on third parties, can be potentially devastating. What if Scytl had gone bankrupt, during an election? Needless to say, Markham no longer allows e-voting, which makes me e-sad.

When e-voting went wrong

While Markham was a valiant attempt, it was small potatoes dealing with local politics and might not in and of itself be representative of the potential impact of online voting at a national level. So let’s look at an example of how things can go, very, very wrong and very right. Starting with the bad.

In 2015, New South Wales Australia rolled out the world largest ever e-vote system called iVote. This system allowed voters to cast their ballots online during the state election. Again the goal was to increase accessibility for people with disabilities and those living remotely, while also trying to modernize the voting experience by letting people vote from personal devices.

iVote purportedly was actually quite a nice app and the user experience was smooth and intuitive. Public communication about this new way of voting was also strong, causing nearly 280,000 votes to be cast using iVote in that election! Trouble is, in the days following the election, some pesky independent researchers discovered severe vulnerabilities in the live system. Among them: an insecure external analytics script that could have allowed attackers to intercept or alter votes, and a verification process that could be manipulated to provide false assurances to voters.

What’s worse, there is no clear evidence of tampering, but the possibility was there and because of this they will never try to pilot a program like this again. This may have very well affected who took office and the downstream impacts of that are no joke. The point of this example is to highlight what is at stake if the system goes south, or should I say New South (I’m sorry). In this case, there was no clear evidence of hacking but the vulnerability existed and may have been exploited to favour one candidate or another. With consequences this high, how important is the convenience of voting online in contrast?

iVote was quietly shelved in the years that followed. They had lost the trust of the public despite a lack of evidence of hacking and what began as a promise of a modern democracy ended up as a prime example of how important security is in this conversation. The lesson from New South Wales is clear: building digital voting systems isn’t just about making them easy to use, it’s about making them impossible to doubt. And that requires more than good design. It requires deep, structural transparency, independent oversight, and a willingness to stop and re-assess when trust begins to weaken.

The shining beacon of Estonia

Lastly, I’d like to pull up the most frequently cited example of a successful techo-democracy. Estonia! A small but mighty nation that is often on the cutting edge. They were the first country to launch internet voting at a national level in 2005. Their system called i-voting, allowed citizens to cast their ballots online from any internet connected device, for both national and municipal elections. When it initially rolled out, online adoption was around 2% in the first year but it has since climbed to over 40% of its population using online voting.

But, the reason this is possible is because unlike most countries, Estonia had spent the prior decade building a national digital identity card. This is a critical part of the puzzle unfortunately, and it’s what allows their digital voting program to be so successful. They already use their digital ID’s to file taxes, access their own medical records, and sign legal documents. So when the opportunity came to use this same, state-backed secure authentication method, it was a no brainer. They had the full infrastructure to roll it out.

Their i-Voting process is built for security and usability. Voters insert their digital ID card (or they can use a mobile app), log into a government portal, and are presented with a digital ballot. Once completed, their vote is encrypted and then sent to a secure server. To guard against coercion or vote buying, voters can re-vote multiple times online. This protects against a common concern people have in e-voting, which is buying votes! Only their final submission is counted. And if someone chooses to vote in person on election day, it overrides any previous digital votes. Pretty ingenious.

Crucially, Estonia’s system includes end-to-end cryptographic verification. Voters can use a smartphone app to verify that their encrypted ballot was received and counted as intended without revealing who they voted for. The entire system is publicly audited, with detailed post-election reviews conducted by independent experts.

The Estonian system really is as good as it gets when it comes to a thoughtful, comprehensive and secure approach to digitizing democracy, and because of its long history it also offers us some concrete data on what effect digitization has on voter turnout. So here it is: 0.2 to 0.8 percentage points. That’s it. From 2005 to 2025. The painful truth is that if you’re looking to get more empowered citizens, focusing on access to voting is actually the last step in the process.

There are tangibly good reasons to invest in e-voting and they genuinely matter. From an accessibility standpoint alone, giving people with reduced mobility the ability to vote independently and securely is a powerful step forward. That, in itself, is worth the effort.

But if the goal is to dramatically increase engagement, we’re nowhere near a solution and I’m confident that the solution won’t be a purely technological one. What I uncovered is, if you care about people voting, it’s clear that access, especially technological access, is not the main bottleneck.

We need to stop obsessing over the ballot interfaces and start asking harder questions. Why don’t people vote in the first place? What would make them want to?

That’s where the real design challenge lies and that’s where we should be putting our energy!

P.S Don’t forget to vote in your next election

Can technology fix voting? was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.