The trial of online voting for the October 2019 local government elections was called off last December, for cost reasons. The online voting working party, which had been organising the trial, is now calling on central government to fund online voting for the 2022 local government elections. The Local Electoral Matters Bill is before Parliament, making some small changes to the laws around online voting for local elections.

I have previously written about online voting (here, and here). In this short paper I address some common questions about online voting.

Q: Does online voting help raise turnout?

This is a difficult question to answer simply. In some cases it has, in other cases it hasn’t.

Online voting has the potential to make voting more convenient for many people, particularly the increasing number who do not routinely use the postal system. Overseas studies on voting reform show that making voting more convenient leads to small gains in turnout, typically in the +2-4% range.

There is no increase in youth (18-24 years) turnout when online voting is adopted. Just putting a ballot on a screen does nothing to encourage the not-engaged to vote. Addressing low engagement with local government is the elephant in the internet voting room.

Internet voting will help some disabled voters vote. People who are overseas at election time will find it easier to vote, although by how much depends on the overall system design: as Census 2018 demonstrated, carrying out a civic duty online can be a popular option for many, but if the system is poorly designed, overall participation rates can fall.

Q: 93% of New Zealanders have access to an internet connection so access isn’t an issue is it?

Access to the internet does not mean someone feels comfortable using technology for tasks like voting. People may lack skills, or up to date equipment and operating systems, or not trust the system with their information. Research in Switzerland revealed that people’s subjective ease with technology was an important variable in whether they would vote online (p.36).

Population groups who are most attracted to online voting are those with high levels of digital literacy, which tends to be men aged between 30-50, with higher levels of education and income, and who work in white collar jobs. This is a group who are already likely to vote, and whose interests are already relatively well-represented in society. People who are digitally disadvantaged in New Zealand are often those experiencing other social and economic disadvantage – population groups already less likely to vote.

Making voting more convenient for the digitally literate may widen the gap between population groups in a way that entrenches existing unequal turnouts.

Q: I can bank and shop online so why can’t I vote online?

Banking and other financial transactions online are not 100% secure; people do get ripped off. But when that happens, there are important features which mean the problem can be put right. It is possible for you to detect whether money has been stolen from your accounts, because you can see your transaction and account balances. Your bank can also see these and investigate your complaint, and – unless you have been seriously negligent – they will reimburse you for your losses.

When it comes to voting, voters aren’t given a receipt that says how they voted – because that would enable the coercion or bribing of voters – so you might not know that your vote had been changed or corrupted. The election officials also might not know that anything was remiss. It’s possible no-one would know there was a problem.

Online shopping and banking’s paper trail provides a form of security and redress that is impossible with casting a secret ballot. Not knowing if the results generated from an election accurately reflect the wishes of voters risks undermining trust in the system and the outcome.

Q: Is it less secure than postal voting?

No voting system is 100% secure. Postal voting is less secure than our in-person general elections: it is possible for voting papers to be stolen; because the casting of the vote is not supervised by election staff, it is possible voters are coerced to vote in a particular way (say, by a controlling relative) or that votes are bought; once completed papers are posted, a voter has no idea if they are received and counted by the election officials.

Postal voting’s weaknesses, though, do not mean that online voting’s security flaws are okay. The magnitude of potential problems are also different. A hack of online voting could be widespread, hard to detect, and come from anywhere – literally, anywhere worldwide. Generally, software is incapable of identifying fraud that hasn’t been anticipated by its developers; it is the scale and difficulty identifying hacks that makes security such a problem with online voting.

The New Zealand security agency GCSB has ‘ongoing concerns‘ about the security of online voting trials for local elections.

Q: Won’t blockchain solve the security problems with online voting?

Blockchain is a new technology and it is too soon to hail it as the saviour of online voting.

Blockchain solves only a small portion of the problems inherent in online voting. Blockchain distributes data across many servers, making it difficult for data to be altered – it must be altered in many different places. But blockchain only deals with the fidelity of votes that have been cast and their storage. Many other aspects of an online voting system continue to be problematic: issues such as untrustworthy hardware and system software on either the vote collecting servers or on voters’ own computers, hidden security flaws which could alter the vote before it reaches the blockchain, and issues with the reliability and capacity of the network over which the vote is transferred (which could, for example, be disrupted by a Distributed Denial of Service attack).

Blockchain can be designed to give voters verification how their vote was cast – but this verification opens up the possibilities of coercion and corruption (vote buying). It does nothing to address the digital divide.

Q: Why would anyone be interested in hacking New Zealand’s elections, especially our local body ones?

Local government is very important in New Zealand, contributing to the development of communities, granting or refusing resource consents, and investing in large infrastructure projects that generate valuable contracts for work.

There could be personal, financial, or political motivations for wanting to alter an election result, ranging from a curious hacker stretching their ‘lock-picking’ skills, to a business wanting a council make-up that views their activities more favourably, through to foreign governments wanting to undermine trust in our democracy.

Any system provided by commercial suppliers for use in New Zealand would almost certainly be sold elsewhere as well; our local elections could provide a useful testing ground for anyone interested in probing that particular voting system, and identifying vulnerabilities to exploit in other jurisdictions.

We would be complacent to assume no one is interested in our election outcomes or systems; whether there is evidence of a threat or not, it is important to have a system we can trust.

Q: Is online voting inevitable?

No. We have a choice how we construct our voting systems and political institutions. A number of countries (e.g. Finland, France, Germany, Ireland, Netherlands, Norway and the UK) have rejected online voting or machine voting for their national level elections, mostly for security reasons but also because of cost, and because the turnout gains were not perceived as sufficient to warrant the security risks.