The idea of voting online has been floated around policymaking circles for as long as the internet has existed. Yet, decades later, there are still few government-endorsed standards for online voting procedures and systems. In this article, we set out to look at what standards exist now, what they are lacking, and what should be done to ensure that all online voting systems will be secure, trustworthy, and understandable.
What Do We Have Now?
The inconvenient truth of the matter at the moment is that many of the “standards” for online voting systems now are just suggestions and ideas rather than ironclad, widely-adopted rules. As noted in the US Voting Foundation’s 2015 report, “little national or international election standardization exists” when it comes to voting systems, with one key factor being the fact that “vendors…are uninterested in interoperability.” As a result, online voting systems may vary widely in their levels of not only security and transparency, but also usability and accessibility for users.
On top of recommendations which have no legal weight, there is also the problem of laws and guidelines for online voting which have not been sufficiently updated since the early days of the internet. Notably, in the United States, thirty states and territories currently allow some voters – including those in the military, overseas, and with disabilities – to cast their ballots online either via email or a government-run portal. Aside from obvious privacy and security issues stemming from submitting a ballot via one’s personal email or a portal which does not require multi-factor authentication to log in, the most concerning part is that many of these measures were adopted years ago (on average around 2008) and have only occasionally been updated.
Below, we’ll take a look at some examples of specific guidelines and recommendations which currently exist.
Council of Europe Revised Recommendation to Members (2017)
Issued originally in 2004 and last revised in 2017, the Council of Europe’s (CoE) “Recommendation of the Committee of Ministers to Member States on Standards for E-voting” are currently “the only international legal documents specifying more detailed requirements for using electronic technologies in elections.” Though not legally binding, the absence of other widely known public standards has led to these standards influencing online voting projects in several European countries and they remain a valuable source of guidelines for public authorities when pursuing voting online.
These recommendations state, amongst other things, that electronic voting systems must:
- Be accessible and understandable, meaning that all who can vote should be able to do so independently and comprehend the system(s) they are using;
- Be fair, meaning that official information about events must be equally shared with all parties, only eligible voters can be cleared to vote, and that any tampering with the system or ballots is immediately visible to voters and/or administrators;
- Allow “free suffrage,” meaning that voting systems must not influence how a voter casts their ballot, inform them if a ballot is cast improperly, and present “authentic information” to all eligible voters;
- Uphold “secret suffrage,” which involves keeping voters’ choices secret from casting to tallying of results as well as mandating that voters’ personal data is properly protected, appropriately used, and swiftly erased following the conclusion of an election (following GDPR rules);
- Be introduced in a “gradual and progressive manner,” indicating that time is needed not only for systems to be built to handle large-scale, high-stakes elections, but also that voters themselves must be eased into the system and not forced to use something they do not properly understand.
The guidelines also mention that CoE member states must be “transparent in all aspects of” voting electronically, properly inform voters about the planned use of electronic voting systems, develop “technical, evaluation, and certification requirements,” and establish “electoral management bodies” to oversee the implementation and use of any voting systems.
“The Future of Voting” Study (2015)
Carried out in 2013 and published in 2015 by the US Vote Foundation, “The Future of Voting” study is not a set of recommendations per se, but instead a detailed research effort by academics, election officials, and private sector developers into the current state and future feasibility of online voting.
Specifically, the study mentions that any online voting system used in public elections must be:
- Secure – an online voting system must keep both votes and voter data safe from tampering.
- Usable – an online voting system must not only be accessible to those with disabilities, but also understandable for voters and observers alike.
- Transparent – it is “not enough for election results to be correct” according to the study, as all parties should be able to confirm the accuracy of results via “open public review” of the online voting system.
- End-to-end verifiable – an online voting system must show voters that “their vote was included in the election outcome, the system is recording the content of their votes correctly and that the number of people that voted for a given candidate is accurately calculated…without revealing any of the individual votes.”
The study concludes that all of these guidelines for using online voting systems are contingent upon the systems being “widely deployed” in in-person, small scale contexts before moving onto high-stakes, large-scale elections.
Voluntary Voting System Guidelines (2021)
Drafted for the first time in 1990, these guidelines issued by the American Election Assistance Commission (EAC) have been updated regularly to advise US states on standards for any voting systems they wish to adopt in their elections. As the “voluntary” in their title indicates, these stipulations are not legal requirements for states but instead are suggestions and guidance for adopting any sort of voting system, online or not.
These guidelines encourage states to adopt systems which, amongst other things, are:
- Transparent – a voting system’s “processes and transactions…are readily available for inspection” while voters can “understand and verify the operations of the voting system.”
- Accessible – “all voters can access and use the system regardless of their abilities.”
- Able to guarantee votes are “marked, verified, and cast as intended” – voters can easily mark and submit ballots correctly to reflect their choices.
- Auditable – the voting system enables “evidence-based elections,” “produces readily-available records,” and “supports sufficient audits.”
- Secure – the voting system “authenticates” any users appropriately, is tamper-resistant, and uses “multiple layers of controls to provide resilience” against any attempts at malfeasance or failure.
What Is Missing?
By now, we can see that there are some recommendations which may provide a suitable foundation for more robust standards. But, as noted in the first section, there are some noticeable gaps in both current recommendations and our understanding of them. Here, we highlight some notable examples.
Awareness about the need for end-to-end verifiability (E2E-V)
The US Vote Foundation makes it clear: “any public elections conducted over the Internet must be end-to-end verifiable.” This is confirmed by nearly all academic works cited in this piece, with the consensus amongst even skeptical experts being that E2E-V “seems[s] to be necessary for secure voting via the Internet.” Despite this, only one of the three examples of guidelines above (“The Future of Voting”) actually mentions E2E-V as a necessity rather than a feature of electronic voting systems.
Actual rules beyond recommendations
It would be incorrect to conclude that standards for online voting today don’t exist – they just aren’t widely adopted and/or enforceable at this point. The most promising are perhaps the CoE recommendations we covered, which could form the basis of EU-wide directives, but even they do not specifically address online voting and member states are not bound to follow its guidance when deploying them.
The recognition of online voting as something distinct
In our digitalized age, where everything from shopping and networking to filing taxes and banking can be done online, a question which pops up often is “we do so much online already, why can’t we vote online?” Progress in the development of online voting technologies in recent years has been encouraging, but voting online in public elections is a high-risk activity in several aspects and should not be hurried to meet the demands of the “digital age.” As noted in a 2022 report from the University of California, Berkeley, “unlike most activities, failed – or just distrusted – elections can result in significant outcomes which affect everyone,” including “civil unrest” or distrust in democratic processes themselves.
Ultimately, should we treat exercising a fundamental democratic right over the internet as something akin to a transaction or participating in a survey, it may affect our ability to both develop standards and continue to respect the sanctity of democratic processes.
What Can (or Should) Be Done?
Up to now, we have touched upon exactly what is out there now for both vendors and policymakers to follow when seeking to develop and deploy online voting systems in elections. In this final section, we offer up several suggestions about how to proceed in developing trusted, thorough standards for online voting systems.
- Make standards possible
This point might seem a bit odd – are we saying that standards aren’t possible right now? Not exactly, but, as put by the UC Berkeley report, “it is currently infeasible…to develop the full set [of standards] needed to insure safe and secure internet voting” without more extensive testing, evaluation, and improvement of online (and E2E-V) voting systems in public election contexts.
This leaves us with a “chicken versus egg”-type dilemma: did the lack of standards lead to the conclusion that Internet voting isn’t secure, or does the insecurity associated with voting over the internet make it hard to create standards? The answer isn’t necessarily one or the other: developing fully secure and transparent online voting for use in sensitive contexts is in fact a gargantuan task while the lack of concrete rules developers must follow in bringing such systems to market has led to too many insecure systems being marketed as safe. We must thus follow experts’ advice and take action to gain more experience which can be used for drafting standards.
- More local, small-scale testing of technologies
Building off of the previous point, part of developing standards is more testing. As mentioned by computer scientist Josh Benaloh and his colleagues, any online voting system used in public elections must be E2E-V, but these systems “should not be used before end-to-end verifiable poll-site voting systems [without an Internet connection] have been widely-deployed and experience has been gained from their use.” The Berkeley working group adds that “advances” in technology “have moved us closer to secure and accessible internet ballot return, but there is still a long way to go.”
Democracy itself is very much a “learn by doing” concept; the more experience citizens gain in voting, building connections, and advocating for the betterment of society, the better the outcomes. Developing online voting systems and standards is much the same, as the key to their development is greater usage and understanding in a variety of democratic contexts.
- Insist on E2E-V being the starting point
As we hope is clear at this point, E2E-V in online voting systems is not just a unique feature or fancy add-on, it is the sine qua non of voting online in public elections – without it to guarantee integrity and transparency, it is just not possible for online elections to be secure, transparent, and trustworthy. It is thus our duty to make sure that policymakers and developers alike know this as well when thinking about carrying out voting online.
We take pride in advocating for the recognition and use of E2E-V in online voting systems and believe the advancement of secure, trustworthy online voting can only come with widespread recognition of E2E-V as an irrevocable requirement of any online voting systems. For curious readers, we cover the issue in greater detail here.
- Establish an international online voting industry association to negotiate, promote, and uphold standards
As mentioned by our founder Jacob Gyldenkærne in an interview last year, “we are dealing with something extremely sensitive, namely trust in democratic elections and in our democracy” when talking about voting online. It isn’t difficult to find an online voting system if you want one, but the lack of standards has encouraged the proliferation of so-called “black-box” systems which are not E2E-V and often fail to meet the criteria we have covered thus far.
It is therefore in the interest of vendors of online voting systems to come together to deliberate their own standards as part of an industry association in the hope that such initially private standards enter the public sphere by connecting with policymakers and setting good examples for others to follow in the pursuit of voting online.
