In October, Greenland was reported to be exploring the feasibility of an online voting platform for its national elections. Among the options being considered is a blockchain-based system.
That isn’t entirely surprising. Electronic voting, or e-voting, has long been viewed as a promising use case for blockchain technology. “It’s time for online voting,” wrote Alex Tapscott in a New York Times opinion piece in 2018. “Using blockchain technology, online voting could boost voter participation and help restore the public’s trust in the electoral process and democracy.”
It seems especially timely now as large swaths of the world’s population are raising questions about election integrity — most notably in the United States, but in other countries as well, such as Brazil.
Tim Goggin, CEO at Horizon State, for one, believes that blockchain-enabled elections represent a “significant improvement” over the way most elections are operated today. Voting machines break down, software fails and election irregularities often create uncertainty and doubt among the voting public.
With a public blockchain, by comparison, “it is much easier for voters to trace their vote,” Goggin told Cointelegraph, “and audit an election themselves.”
Moreover, if something untoward does occur in the voting process, it is easier to identify it on a decentralized ledger with thousands of nodes than on current tabulation systems “where counting is done behind closed doors,” says Goggin, whose company set up a public election for South Australia in 2019, the first time blockchain technology was used in the voting process for that Australian state.
Still, blockchain technology’s potential vis-a-vis public elections has been highlighted off and on for some time now. No country has yet to use blockchain technology in a national election.
Marta Piekarska, senior DAO strategist at ConsenSys, recalls working at Hyperledger in 2016, where blockchain voting was discussed as a promising use case. “Six years later, and we are still talking about this,” she told Cointelegraph. “We are still quite far from a situation where any kind of distributed ledger would be considered” — at least in a national election.
A few countries, notably Estonia, have been experimenting with systems that allow people to vote online, she further explained. On the other hand, “Netherlands abandoned the idea of doing electronic voting due to some of the concerns around security and authenticity of the votes.”
Then, there’s sparsely populated Greenland, where the vast distances make it difficult for people to vote in person. A group of researchers from Concordium Blockchain, Aarhus University, the Alexandra Institute and the IT University will soon be investigating “whether a blockchain-based system will be a more trustworthy e-election on the world’s largest island,” according to the Concordium press release.
Ensuring trust is critical
Any voting system requires trust, and trust requires a number of properties — any one of which can be a challenge depending on the circumstances, Kåre Kjelstrøm, chief technology officer at Concordium, told Cointelegraph. For in-person voting, these include: whitelisting: ensuring only eligible voters take part; identification: voters need to prove their identity when casting a vote; anonymity: votes are cast in private and can’t be traced back to the voter; security: locations are secured by the government; and immutability: cast votes can’t be altered.
“Any digital system that replaces a manual voting system needs to address at least those same issues to ensure trust and this has proven to be rather tricky to pull off,” Kjelstrøm explained. “But blockchain may prove to be part of a solution.”
A public decentralized blockchain ensures immutability by default, after all, “in that any transaction written can never be deleted.” The system is secured by cryptography and “transactions are anonymous, but are open for inspection by anyone in the world,” said Kjelstrøm, adding:
“The trick is to maintain privacy and anonymity while ensuring any eligible voter can only cast their vote once. […] This is a current research topic at top institutions.”
Permissioned or public chains?
“The main problems I see for public elections as opposed to say corporate governance is that there cannot be a permissionless [blockchain] system because voter information is private and we cannot trust all third parties,” Amrita Dhillon, professor of economics in the department of political economy at King’s College London, told Cointelegraph.
“The second problem is that of inputting the vote at a location of the voters choice: We cannot prevent anyone coercing voters at the point at which they submit the e-vote,” she added.
Others say permissioned chains aren’t the answer because they are run by a single entity or a group of entities that exert complete control of the system. “Worst case this means that a private blockchain can be tampered with by those self-same guardians and elections rigged,” said Kjelstrøm. This isn’t much of a problem in Western countries, “but in large parts of the world this is not true.”
On the other hand, if one can “weave self-sovereign identity (SSI) into the core protocol,” as Concordium, a layer-1 public blockchain, aspires to do, that “may be just the right technology to power public elections,” said Kjelstrøm.
That said, Goggin noted that many governments will probably opt to use private blockchains in line with their own privacy/data laws, and there are many ways to set up permissioned blockchains. But, if they don’t at least offer the public an auditable trace of voting records, then they aren’t likely to boost the public’s belief in election integrity. He calls himself “a big fan” of public and distributed blockchains.
The privacy question is especially knotty when it comes to public elections. “You should not be able to tell which candidate some individual voted for, or even if they voted at all,” wrote Vitalik Buterin in a blog titled “Blockchain voting is overrated among uninformed people but underrated among informed people.” On the other hand, you want to ensure — and if necessary prove — that only eligible voters have voted, so some information like addresses and citizen status may need to be collected. Buterin viewed encryption as a way to get around the privacy conundrum.
Goggin suggests something similar. Horizon State might ask a client to “hash,” i.e., encrypt or scramble, eligible voter identities “before we are provided them, and we then hash those identities again.” This means that neither the client nor Horizon State can readily determine who voted or how they voted. He added:
“Voters will be able to see their vote on the chain, but there is no way for voters to prove that it is their vote, given they can see other votes on the blockchain also.”
Dhillon, for her part, proposes a compromise where “some parts of the process are centralized,” i.e., voters come to a booth where their identity is checked and they submit their vote, “but subsequent parts of the chain can be decentralized to make them more secure and tamper proof.”
In 2014, the city of Moscow’s Active Citizen e-voting platform was created to let Muscovites have a say in non-political municipal decisions, and in 2017 it used the Ethereum blockchain for a series of polls. The largest of these tapped 220,000 citizens and the voting results were publicly auditable. It revealed some scaling limitations.
“The platform based on proof-of-work reached a peak of approximately 1,000 transactions per minute [16.7 transactions per second]. This meant that it would not be easy for the platform to handle the volume if a higher proportion of Moscow’s 12 million citizens participated in the voting,” according to Nir Kshetri, a professor at the Bryan School of Business and Economics at the University of North Carolina at Greensboro. From this, Kshetri and others concluded that this PoW version of the Ethereum blockchain “was not sufficient to handle national elections.”
Things might be different in 2023, however, when Ethereum 2.0 implements sharding. This could boost the chain’s speed to as high as 100,000 TPS, which in turn “increases Ethereum blockchain’s attractiveness for voting,” he told Cointelegraph.
But blockchains probably still need to be more secure before they are ready for public elections, though this is manageable in Kshetri’s view. “Blockchains are likely to become more secure with increasing maturity.”
Buterin, too, said in 2021 that security was still an issue vis-a-vis elections. For that reason, “in the short term, any form of blockchain voting should certainly remain confined to small experiments. […] Security is at present definitely not good enough to rely on computers for everything.”
Online transactions, unlike manual systems, “can occur in the blink of an eye,” added Kjelstrøm, and software-driven attacks on an e-voting system can “potentially foil or damage the system or the vote.” Therefore, “any new system would have to be introduced slowly to ensure the voting system remains intact and fully functional.” Governments might begin at a small scale and conduct proof-of-concepts for select non-critical elections first, he said.
Usability is critical
Technology isn’t the only obstacle that needs to be solved before blockchain voting attains wide adoption. There are political and social challenges, too.
“The technology is there,” said Piekarska. “We can do it right now. I mean, decentralized autonomous organizations are governed through online voting now, and they are managing trillions of dollars.” But national elections are a different beast, she suggested, because:
“On the government level, your problem is: how do you create a system that is usable by citizens?”
One’s constituency is not tech-savvy members of a DAO, “but people like my mom, who is still struggling with online banking,” Piekarska added.
How long will it be, then, before the first national election with blockchain voting? “Hopefully not decades, but surely we’re not there yet,” said Kjelstrøm.
“It could be tomorrow or it could be in 50 or 60 years,” opined Piekarska, “because there are so many things that need to align.” In Europe, most people trust their governments and the quality of voting is not really an issue, so the push for encrypted auditable ledgers may not be so urgent. In nations with weaker governance where elections are often manipulated, conversely, why would the powers-that-be ever consent to tamper-free blockchain voting?
Greenland, which struggles with participation in its general elections primarily because of the great distances that its citizens must travel to vote, might prove an exception.
“Yes, some solid governments want to do the right thing but they struggle with the accessibility of in-person voting,” Piekarska acknowledged. “That’s probably where we might see the first movers because there is a very high incentive for them to do it. But these are unique situations.”
All in all, it’s critical that people have trust in their voting system, whether manual, electronic or blockchain-based, and building trust can take time. But, as more people become used to accessing public services online, electronic voting should take greater hold in different parts of the world, and once that happens, blockchain voting could catch on, given its well-documented advantages, allowing individuals to audit their own votes.
Large-scale blockchain-enabled national elections are probably some years away still. Even so, Goggin has been engaging in discussions recently “about providing elections at that scale,” adding:
“While it isn’t the norm yet, governments are beginning to consider the value that online blockchain voting systems can offer in efficiency, accessibility, speed, security and transparency.”