The Register of Electors needs overhaul, but it’s not just a tech problem An article in the Irish Examiner yesterday reported that “The Government ‘Ignored Calls’ on Improving the Voter Register”. Seán McCárthaigh reports that the city and county management association have been warning the Government that “the practices used to maintain the register were outdated and in need of urgent improvements” for the past three years. McCárthaigh notes the following reforms have been proposed:

  • The introduction of a centrally hosted database of all electors.
  • Provision for applicants to be required to provide a unique identifier such as their PPS number to go on the register.
  • The provision of an online portal to allow electors to register themselves.
  • A move to a ‘rolling’ register and ending the existing ‘draft/live’ versions.
  • Automating the import of ‘life event’ data such as deaths.
  • The use of a range of modern media including text messages to issue polling cards.

The concern stretches back longer than that though. My boss and I had been discussing the Electoral register recently, and a few weeks ago Daragh unearthed a paper he wrote in 2006 analyzing the proposals for electoral reform that had been under discussion ten years ago. In that paper he notes that the proposed solutions tend not to address the root causes of the problems with the electoral register. Some of the proposals certainly look similar. Among them:

  • The establishment of a central authority in a Commissioner for Electoral Registration to oversee the creation and maintenance of the electoral register.
  • The use of the PPSN as a unique identifier for the register
  • Validation of the register against reality (a few different ways to do this were proposed)
  • Using data from the GRO to remove recent deaths from the register

Clearly the Government has been aware of the outdated register and its problems, and has been proposing reforms to fix it for at least a decade. You could say that the more things change, the more they stay the same. We know the Register of Electors is in disarray, but getting it in shape does not have an easy fix and does have a lot of risks for complete disaster. Anyone who remembers the electronic voting machine debut of 2002-2009 has a relevant example of what introducing updated technology change into the electoral process without due diligence in information governance and security might look like. We’re still using “stupid old pencils” because they were better fit for purpose than the machines as designed.

Plus ҫa change. But, one thing that has changed incredibly in the last ten or so years is the amount of information about people available, and the potential uses it may be put to when it comes to analytics and targeted campaigning. Campaign strategists have been looking across the pond to the pioneering use of analytics and social media in recent US elections, particularly the Obama campaign. Particularly since 2012, people interested in the data-enabled political campaigning have become increasingly enthusiastic about the ability to create a “single view of constituent” and use it to target political messaging and get out the vote.

The amount of data that can legally be and has been sliced and diced on American residents is frightening and far more invasive to privacy than is legal under European data protection laws. But the fact is, Americans don’t want that invasion any more than Irish citizens do, and knowledge of use of their personal data to target political messaging is likely to damage their opinion of candidates and affect their likelihood to vote for them. In fact, there have been several cases where US political parties have stepped a little bit back from going full “creepy” because they realized doing something perfectly possible and legal would have a very bad reputational effect.

European political parties and candidates are more restricted by comprehensive data protection legislation than in the US, but the technological capabilities exist and without clear guidance it may be easier to stray into the “because we can” without due consideration of “whether and how we should”. For instance in the UK, the Conservative Party has brought in the US Republican Party’s voter database system to aid their canvassing. Robust Information Governance practices, with proactive guidance from the DPC, is essential.

If one models the theoretical data flows already possible with information currently available about Irish citizens, including the possible availability of the marked electoral register, one really appreciates the importance of the Data Protection Acts and the constitutional protection of privacy of the ballot. The idea of a “Single View of Constituent” sounds great to people working political campaigns, but it is fraught and high risk. There are dots that should not be connected in order to preserve individuals’ privacy rights. Any implementation of a “single view” or “universal identifier” will have to be extremely carefully constructed to ensure it does not violate the fundamental privacy rights of the individual and the privacy of the ballot. Safeguards must be built in to prevent its abuse or subversion of the democratic process. Privacy of the ballot is constitutionally protected right in Ireland and fundamental to the democratic process. A successful design of a “single view of constituent” may be possible, but it must be extremely carefully constructed to preserve fundamental rights.

This is one of the reasons why the repeated proposal of using the PPSN as a unique identifier for the electoral register is concerning. The PPSN has seen a lot of scope creep, but this particular proposal may be particularly problematic. Risks raised by this proposal include:

  1. A too comprehensive Universal view of Citizen. Or “Single View of Constituent” linking voting record to social benefit eligibility
  2. Marked registers and access to marked registers with PPSN attached.
  3. Vulnerability to fraud
  4. PPSN was not designed for this purpose and thus may not actually be fit for this purpose.
  5. PPSN is still not actually a unique identifier in all cases and it could exclude persons living here who are entitled to vote in certain elections but don’t have Irish PPS numbers.

Electoral reform is needed. But effective electoral reform must be implemented in a way that addresses the root causes of the current problems with the electoral register, preserves the privacy rights of individuals or “doesn’t cross the creepy line”, and upholds the constitutional right to privacy of the ballot. With advances in data analytics and increased connectivity of the last few years, the risks of invasiveness are much higher than ten years ago. What do we need for effective electoral reform that does not cross the creepy line?

  • Governance: This is the foundation for reform, upon which everything else can be built properly.
    • A centralized governance structure: the current system is extremely siloed and fragmented, which makes it far more difficult for changes to be recorded across authorities. This is a clear problem when it comes to mobile populations.
    • Clear lines of stewardship, responsibility, and accountability from the local election workers, through the local authorities, through the central authority.
    • An audit trail.
  • Technology: Technology is an essential part of the solution when it comes to increasing efficiency, but without proper governance structures in place upgraded technology is likely to increase the efficiency of error making.
    • A system that accounts for name changes, location changes, and population changes.
    • A central system that can verify that an elector is registered only once.
    • A system that has adequate safeguards in place to ensure data is kept safe and secure and access is restricted only to those authorised.
    • A system that enables an audit trail, ensuring changes and access are logged and can be accounted for.
  • Communication: Clear communication is absolutely necessary not just between local and central authorities, but also with political parties and campaigners and perhaps most importantly, with the public. Communication is required on several levels:
  • Clear communication of accountabilities and stewardship
  • Clear communication of what can be done and what should be done
  • Clear communication to the public. . .
    • How to get on the register
    • How to record changes
    • Criteria for entry on the register and penalties for submitting false or inaccurate information
    • How their sensitive personal information is being protected and their fundamental right to privacy is being upheld.

It’s generally well recognized that electoral register reform is necessary. Effective reform requires addressing the root causes of current problems and designing and communicating a strategy that implements best practices in governance and Information Quality. Unfortunately, the Government’s recent track record here has not been good.