Why blockchain voting won’t solve the trust problem in election results

Office work and public services are moving to digital formats, but governments are not rushing to implement online voting. Among the reasons: technical problems, a never-ending string of failed experiments, and public distrust. The obvious way out of this impasse is blockchain technology — after all, it was designed to solve the problem of trust. ForkLog investigates whether this is feasible in practice.

The United States presidential election reverberated worldwide; not everyone trusts the official results, largely due to one of the candidates, incumbent President Donald Trump. He refused to concede and pledged to challenge the outcome in court. On Twitter, Trump speaks of numerous falsifications.

Reasons for such distrust include the complex voting and vote-counting procedures at polling stations, as well as mass mail-in voting. The latter measure is due to the coronavirus pandemic and aims to minimize contact between voters.

Indeed, mail-in voting with ballots in envelopes may seem strange and, frankly, outdated. If the pandemic has achieved anything, it has accelerated the already ubiquitous digitization of various services. It would be much more convenient to vote online in such times.

Why did the U.S. authorities decide to use such an archaic and unreliable method? And in general, why in the 21st century should people go to a polling place to mark a ballot, rather than solve this task, as with many others, online?

The answer is that traditional voting using paper ballots is trusted even more than online voting.

Why online voting isn’t trusted

Online voting differs from electronic voting—the latter also includes on-site voting systems, but using the devices installed at polling stations where voters insert their filled ballots. The device scans the ballot, determines the voter’s mark, and transmits the result to a central server. In online voting, the voter can vote without leaving home.

There are examples of countries where electronic voting using devices is widely used — Brazil, India and the United States.

At the same time, there is only one country that has implemented internet voting without the need to visit a polling station — Estonia. It has been in use there for a long time — since 2005. In the 2019 parliamentary elections, about 44% of eligible Estonians voted online.

Online voting in Estonia takes place in advance. On the designated day, the election commission’s website opens the voting app: to cast a vote through it, voters must use a phone with a SIM card and Mobile ID — this is an electronic identity.

In no other country has internet voting fully worked at scale. Public trust in such services is even lower than in paper voting, due to the opacity of the online voting system itself.

Much of this is due to the fact that the development and testing of such systems are government orders, for which a specific contractor is chosen and which fulfills the order according to formal criteria. As a result, the code almost always contains errors and vulnerabilities (if the code is even published).

For example, in 2019 researchers found a cryptographic backdoor in a voting system designed for Switzerland. And in 2020, MIT scientists criticized the Voatz voting app, which was used locally in several U.S. states. Notably, Voatz’s developers claimed that their app used blockchain to protect votes during transmission, but scientists disputed this claim.

How open-source code could help online voting

The problem with all online voting experiments is their secrecy from the expert community and the concealment of the codebase, says Evgeny Barkov, the development manager of the Russian blockchain project Polys for voting.

According to him, experts should be brought in from the moment the concept of the future online voting platform is discussed, and there should be a public dialogue with them with a detailed breakdown of all critical points.

He believes that using open-source systems can help:

\”Publishing the code is an excellent method for increasing trust and reducing the risks of hacking. In case the development is led by a private company or third-party code is used, disclosure of the source code raises questions of intellectual property\”, — noted Barkov.

Arguments against blockchain

In theory, blockchain technology is well-suited for organizing voting, as it allows data to be collected and processed more securely and reliably than traditional databases connected to the internet. It is not surprising that the community has again started discussing blockchain-based online voting in the wake of the U.S. presidential elections and related scandals.

One of those who spoke in favor of blockchain-based online voting was the head of the Binance exchange, Changpeng Zhao:

\”If there were a mobile voting app on blockchain (with a proper KYC system, of course), we wouldn’t have to wait for results or raise questions about its authenticity. Privacy can be protected using various encryption mechanisms\”

There were enough critics in the comments, ranging from unwillingness to pay for the app to a lack of smartphones among voters.

CZ also received support from Ethereum founder Vitalik Buterin:

\”The technical challenges in developing a secure cryptographic voting system are indeed substantial (and often underestimated), but in my view this direction is 100% correct.\”

Some specialists believe that blockchain will not make online voting reliable — the MIT researchers’ paper is cited in a November 16 publication.

Researchers caution that voting should not be equated with buying goods or performing financial transactions online. If attackers gain access to a user’s bank card, the bank can block the card or even return stolen funds. In blockchain voting, if a vote is somehow altered or not counted, the ability to change the result is almost nonexistent.

Another problem is the guarantee of anonymity. After voting, no links between the vote and the voter’s identity can remain, but it becomes impossible to identify and prove errors, fraud, falsifications or other acts of vote manipulation.

One of MIT’s main concerns is the scale effect: a single error could affect many voters simultaneously and influence their expression of will. For example, the system could show that a voter voted for one candidate, while the vote would be counted for another.

There is also the risk of a zero-day attack, i.e., hackers exploiting a known vulnerability for which a fix has not yet been released. Such an attack would not be expensive. Unlike online voting, with paper ballots, ballots are not linked to each other.

Blockchain elections in Russia

There have been few cases of online voting via blockchain-enabled platforms, but Russia is surprisingly at the forefront. The first blockchain voting was tested during the Moscow City Duma elections on September 8, 2019. This experiment was marked by system outages and criticism of its operation. ForkLog has already written about this:

In a second instance, the blockchain platform was used in Moscow for the vote on constitutional amendments in the summer of this year. By that time the system had been refined: if in 2019 Ethereum was used, in 2020 the corporate blockchain Exonum, developed by Bitfury, was used. Still, it was not without glitches.

The official developer of Moscow’s online voting system is the Department of Information Technology of the Moscow mayor’s office, and Polys advises the officials. Evgeny Barkov noted that for the vote on constitutional amendments Exonum’s speed was 1500 transactions per second, enabling about 1.8 million voters per hour.

\”We continue to work on optimizing the structure of the solution and plan to increase performance next year,\” said a Polys representative.

By the way, blockchain voting in Moscow will be conducted over the next two years — that is what city deputies decided.

The Moscow experience is already moving to the federal level. Waves, together with Rostelecom, by order of the Central Election Commission, developed its online voting system using blockchain, which was used by more than 30,000 voters in the September by-elections for the State Duma in the Yaroslavl and Kursk regions. Recently the source code of this system was partially published. It is not excluded that this solution could be applied to the State Duma elections in 2021.

Blockchain in the Waves and Rostelecom system ensures immutability of information and transparency of smart-contract execution, protects and stores data in a decentralized manner, including the voter list and encryption keys, and allows tracking of transactions within the network. Nodes for data verification are servers where vote counting is performed.

Election commissions and observers on the blockchain

In addition to attracting external developers to refine the voting system’s code, consideration should be given to employing institutions that build trust in traditional elections — observers and commissions.

For observers, Polys proposes using so-called auditor nodes. These nodes are placed at candidate sites, public organizations, and international observers. Auditor nodes participate in the overall blockchain network and can monitor the voting process — incoming votes and issued ballots. In addition, organizations using auditor nodes can create external interfaces and give the public access to monitor the election process.

But the basic unit of the voting system in any country remains polling places — special venues where voters come to vote. In democratic countries, election commissions are the main source of trust in elections. The principles of their formation and the conditions of their work play a major role. The main goal is to make commissions independent of external interference and neutral in counting votes and their final tally.

The election commission, in particular, is responsible for the “reality” of votes cast using pre-compiled voter registers. According to Evgeny Barkov, Polys has managed to recreate the election commission for traditional elections and in its online voting system:

\”The polling station’s election commission is also a deliberative body, and members from political parties and public organizations can also be appointed. For every election, public organizations and parties can appoint additional commission members with a right of deliberative vote. Any commission member, according to 67-FZ, can study the voter registry and verify voters’ addresses. If violations are detected, a commission member can raise questions about errors in the registry, submit a request to correct the error, appeal the commission’s actions to a higher commission or to court\”.

Even the participation of a single independent member in the election commission can prevent falsifications by the commission. Yet during Moscow’s blockchain voting, Barkov said, the function of the election commission was minimized.

The online voting of the future — what will it look like?

Thus, a portrait of a future online voting system emerges, one that society could trust. It would have the following characteristics:

• Open-source software, with the ability for specialists to audit it, propose changes, and contribute.
• Use of blockchain for data verification and immutability, and cryptography for data encryption and guaranteeing voter anonymity.
• Monitoring and rapid remediation of issues during the voting process.
• Access to voting data and tally results for observers, likely via special “observer” nodes.
• Presence of virtual election commissions with the same functions and powers as in traditional voting.

Perhaps these characteristics will someday become the standard for online voting. Perhaps the online voting process will someday be fully automated — as in the DeFi sector.

However, it is impossible to avoid bugs, backdoors, and code exploits (as in DeFi) — therefore a large community of developers will be needed to continually update the code. This is a form of civil-society participation. Rules for recounting votes at individual online precincts will also be needed.

Yet online voting cannot fully replace traditional voting yet, at least because of the “digital inequality”: not all voters have internet access or sufficient skills to use digital services. It is also possible that voters will split by their preferences and trust in the traditional or “modern” voting system. At the very least in coming years, online voting will be a supplement to the overall voting system.

Even if blockchain-based online voting can address distortions in the voting process and in vote counting, it cannot withstand factors outside the electoral system: propaganda, disallowing candidates, the administrative resource and coercion to vote, and other manipulation tools used, in particular, by modern regimes of “electoral authoritarianism.” Online voting may be a more convenient way to vote, but it is not a panacea.

Subscribe to ForkLog news on Telegram: ForkLog Feed — all the news feed, ForkLog — the most important news and polls.