When elections go wrong… four things we can learn from a big week of election mishaps

This week, several disaster stories from the world of electoral administration have made their way into the UK headlines. Ahead of his upcoming report on the impact of Voter ID, Dr Greg Stride, our resident local elections expert from LGIU’s Local Democracy Research Centre, shares some insights into the key takeaways from these events.

Part of my research on elections includes meticulously searching news websites for any tiny stories about potential election-related mishaps so LGIU can support our members to improve election delivery. But this week, following several big stories about electoral administration, other people started sending them to me! Although this saves me some work, which I appreciate, it is a pretty clear-cut sign that something has gone wrong.

First, there was the shocking news that the Electoral Commission had been hacked by ‘hostile actors’ and that the details of millions of electors had been accessible. Then, the news that over 1,000 late postal votes had not been counted at the most recent council elections in Brighton and Hove. And finally, that ballot papers would be counted again after a successful challenge to a result on Derry and Strabane District Council.

It’s worth noting the research that we’re doing at the LDRC on electoral administration, funded by JRSST-CT, where we have found that electoral administrators in England, as elsewhere in the world, are working incredibly hard under very difficult conditions to make elections happen. Although administrative mistakes do occur, and electoral administrators would be the first to admit that, they happen in a context where electoral administrators are under intense pressure to deliver an essential service with limited resources and in a very tight timeframe. More to the point, only one of the stories coming out here this week was due to electoral administration.

These three mishaps lead us to four very different conclusions on how elections are run in the UK, and where there are threats to election integrity.

1. Cyber-attacks are, and will continue to be, a threat to elections

Luckily, the role of the Electoral Commission in the administration of elections in the UK is relatively minor, responsible for overseeing elections rather than organising them, and regulating political finance. The information they held on electors was held only for research purposes, and, as the Electoral Commission Chief Executive said at the time the attack was revealed:

‘The UK’s democratic process is significantly dispersed and key aspects of it remain based on paper documentation and counting.”

This is definitely true, and this dispersal is one of the reasons why the data that could have been accessed is limited only to name and address, because the Electoral Commission does not hold other electoral data which is processed only by local elections teams. The second part of the Chief Executive’s statement is more debatable:

‘This [the dispersed and paper-based processes] means it would be very hard to use a cyber-attack to influence the process.”

This is true to an extent, in that vote tallies couldn’t be changed by cyber-attack because the paper record could always be checked. But large sections of the process are now digital, and therefore vulnerable to cyber-attack – especially if the aim of the attack is to disrupt electoral processes rather than change the result of an election. In 2016, the registration portal crashed at the time of the registration deadline, demonstrating, although there is no evidence that it was an attack, that there are processes essential to our elections that are fully digital and can fail at important times. Equally, electoral services departments around the country are reliant on outsourced election management software, and any successful cyber-attack on those could be damaging to an election in that area.

There is also the case made by experts in computer science and digital technology that data leaks of this kind can be used in conjunction with new technologies to target disinformation and further disrupt the electoral process, demonstrating how important it is for electoral infrastructure to keep up with advances in technological risks.

Electoral modernisation is good for participation, but where there are vulnerabilities in our electoral processes these should be recognised and guarded against, especially if intentionally disruptive hostile actors turn their attention to reducing confidence in electoral processes rather than the relatively more difficult task of attempting to change the results of an election.

2. Electoral integrity is dependent on external actors

The news on the uncounted ballot papers in Brighton and Hove has been floating around for some time, and the deputy returning officer there had asked Royal Mail to investigate the issue. We don’t know exactly what caused these 1,400 ballot papers to be delivered after the deadline, but the usual explanation – that people posted them too late, has been rejected because this is far beyond the normal number that appears after the deadline. The Royal Mail has since said that the ballot papers were delivered appropriately, so maybe we will never know why this happened.

But, that does not mean there are no useful conclusions we can draw from this. First, it demonstrates again that elections, as a uniquely complicated and time-limited process, are dependent on external actors. In the case of postal votes, they depend on Royal Mail for delivery, which means that any disruption Royal Mail faces for any reason that results in postal votes coming in late, means that those votes will not be counted. There are other external private actors that previous research has shown electoral administrators are dependent on: print companies for ballot papers, election management software companies for the software they use to manage the electoral register, and polling station providers to make sure people can cast their ballots.

This is somewhat inevitable, electoral administrators cannot take on every role in an election and must outsource some tasks, but it does introduce another threat to elections – that external providers can and do fail to meet the necessary requirements for an election to run smoothly. However, wherever the issue may arise, the Returning officer remains legally responsible for the election.

3. Normal accidents do happen – it’s correcting them that’s important

No administrative process can be completely free from all errors, the important thing is that when they do happen, there is a process to make sure we can act on them. Unfortunately, in the UK the only way to challenge an election result based on the sorts of mistakes spotted in Derry and Strabane is through a legal challenge called an election petition, a costly and difficult process.

In a more ideal system, it would be possible for administrative mistakes that are fully recognised by all involved to be solved faster and with less reliance on legal systems. In the case of Derry and Strabane, the Electoral Office of Northern Ireland and the council have already confirmed that there was an error. There is no dispute that something has gone wrong, but there has still been a necessary legal challenge and an 80-day delay to fix this error.

Any well-designed process should take into account the possibility that normal errors will occur, and put in place a simple and easy system for correcting them – a standard the current system fails to meet. Both the Association of Electoral Administrators and the Law Commission have made recommendations for changes to simplify this system.

4. The biggest threat (for now) is reduced confidence

The final conclusion we can take from all of this is that even in the case of UK elections, which are generally managed without significant problems, there is still a risk that mishaps can shake confidence in the process. None of these mistakes alone should result in reduced confidence, but we have seen how – especially if confidence in the process becomes a political issue as in the United States – that trust in the electoral process is not guaranteed.

As we will raise in our research, it is essential for confidence in elections that the administrative process should be reliable and that risks should be managed. Our research has suggested that pressure on administrators represents another risk, and one that should be mitigated before mishaps like those we have seen in the last few days become more common, and represent a greater threat to public confidence in elections.

To find out more about our elections research, please sign up for our free online report launch on the 13th September: The impact of voter ID in the UK. Where Dr Greg Stride will talk through our findings and what they mean for elections going forward alongside colleagues from the Association of Electoral Administrators (AEA).

Want more insights on this topic? LGIU’s got you covered:

For the May local elections, we produced an expert panel together with Ipsos, where we touched on topics such as trust in elections and attitudes towards local government. See the write-up here, and the expert panel here.
From exclusive LGIU Member content to articles which are free for anyone to read, this collection brings you a full set of resources on democracy to support better policy and practice for your communities.