Automatic re-registration in Britain?

We should have a system of automatic electoral registration in Britain to combat the crisis of millions of people missing off the electoral roll.  That was one of our core messages last week when we sent a report on elections in Britain to the government.

Most polling stations turned away would be voters at the 2015 general election, research from Alistair Clark and I showed.  On Radio 4’s Today Programme and Radio Norfolk I explained that many citizens think that they are on the register because they pay their council tax, renew their car tax and register for other government services.  But unlike many other countries, they still need to do more to register to vote.

Earlier this week John Penrose said that he is considering a form of automatic re-regstration. This is excellent news for democracy.

I flagged the problem of low levels of voter registration and how individual voter registration would worsen this in a report to Parliament in 2011. I have long argued that using methods used in other countries such as registering citizens when they use other government services would improve registration levels and turnout. This was based on my book and an article on electoral administration

John Penrose’s announcement is therefore very welcome news but there is a more to be done. Automatic re-registration stops names dropping off and might save local authorities time and resources. But what of those not currently ln the register? Upto 10 million could be missing from the register in December. And that is the register on which the boundaries for the 2020 general election will be fought.


Why we should keep postal voting in the UK

Richard Mawrey QC claimed yesterday that postal voting was ‘not viable’ in the UK yesterday.

I written a blog post for Eastminster and the Democratic Audit which has been published today, offering a reply, and explaining why we should not be so hasty to scrap it.

Similar arguments were raised today in separate posts from Stephen Twigg MP and Jessica Garland of the Electoral Reform Society.



Labour announces good news for British elections?

Will Sadiq Khan and Ed Miliband take electoral registration in the right direction?

Voter registration is a problem in the UK.  Approximately one in five citizens are not on the electoral register.  When it is reported that only 65% of registered people voted at the 2010 general election, we forget those people who are not on the register in the first place.

The Guardian is today reporting that Labour is looking to respond to this problem with three promises:

Labour is looking at bringing in a US-style system of allowing voters to register on election day….


….Labour will also ensure people are encouraged to sign up to vote every time they come into contact with a government services, including the DVLA, Passport Office, universities, schools, colleges, blood donation, council tax payment and parking permit applications….


….Labour will pull the plug on individual voter registration if it leads to a large decline in democratic participation


Good news?

I will write more in a fresh blog soon.  However, to give away the ending of the story, you might be interested in what I wrote in May 2012 on the Electoral Reform Society and LSE blog (and many later posts since), after my research had shown that individual electoral registration was likely to lead to a decline in registration.

Five things that you may (not?) know about voter registration on National Voter Registration Day

Bite the BallotBite The Ballot have made Wednesday 5th February 2014 National Voter Registration Day.

This is a great initiative.  There have been citizen-led projects to increase voter registration in the USA for decades.  The UK is only now catching onto the importance of this topic.

Here are five things you might may not know about voter registration in the UK.

  1. Registration levels are getting lower.  A recent piece of research from the Electoral Commission estimated that only 82% of people were registered in 2011.  This compares with rates of 91-93% in the 1990s.
  2. It is lower amongst some groups.  Research has identified it as being lower amongst the younger, Black and Minority Ethnic communities and those who privately rent.
  3. It is lower than many other countries.  Comparisons with other countries are difficult because of poor data quality.  A high registration rate might also result from duplicates or inaccuracies.  For example, one estimate put the registration in the Czech Republic at over 120%, which suggests other problems with the quality of the register.  But the UK still lags behind many other countries.
  4. Things might get worse.  We will soon be required to register individually and provide a National Insurance number, under changes planned by the government.  My research has suggested that this will reduce registration rates further.  This decline is likely to hit students, the young and geographically mobile the most (see: here for the full piece).
  5. There are things that we can do.  The government has invested resources into a scheme called ‘data-matching’ to register people whose identity can be confirmed using other government records.  This is a really important step in stopping registration rates falling further.  However, more can be done such as thinking about election-day registration and holding days and events to focus the media and public on registration, just like today.

Well done Bite the Ballot.

A blueprint for fixing elections? The Presidential Commission on Election Administration publishes its report

The US Presidential Commission on Election Administration has today published its report on how long-running problems with American elections can be fixed.  It has claimed that:

problems that hinder the efficient administration of elections are both identifiable and solvable.

The whole report can be downloaded here

Here were my suggestions:

Voter ID in Britain? A Note of Caution from Academic Research

New Research on Individual Electoral Registration - Policy ImplicationsVoter ID in Britain is being put forward to fix the problem of voter fraud.  Toby James claims that there is little evidence of widespread voter fraud.  Rather, research shows that voter identification requirements might restrict turnout unevenly across British society and is the policy option that is likely to be considered strategically by the main political parties.  

The Sunday Times has reported that the voter ID proposals are being considered by the government.  The Electoral Commission has been reviewing electoral fraud in Britain and will publish its report tomorrow.  According to The Sunday Times, it will propose that citizens should show identification such as their passport or driving licence to prove their identity at polling stations.

My research shows some evidence of declining confidence in the process of casting a vote.  Speaking to electoral administrators recently, many explained that citizens expect tight security procedures around the voting process.  We have passwords and verification processes for banking and other services.  It therefore surprises citizens that they can arrive without any form of identification (not even their polling card).  But is there actually any voter fraud? And what other effects might voter identification have?

Elite Statecraft and Election Administration considers the politics of election admnistration in the UK, USA and Ireland

Elite Statecraft and Election Administration considers the politics of election admnistration in the UK, USA and Ireland

Is there really evidence of electoral fraud?

There have been some cases of fraud in Britain, for sure.  Stuart Wilks-Heeg recently analysed the number of people found guilty of electoral malpractice in the UK.  The most famous case of electoral fraud was in 2004 when judge, presiding over the case declared that the ‘evidence of electoral fraud that would disgrace a banana republic.’  This headline has been banded around again and again by journalists, academics and commentators.

However, it is very feasible that the problem of electoral fraud is of one smoke without fire.  Voter fraud is commonly claimed to be widespread in the US.  Systematic studies, however, show that voter fraud almost non-existent.  One study has suggested that as many people commit voter fraud, as think that they have been abducted by aliens.  No, really.

Is Voter ID voter suppression?

Requirements to produce passports and other forms of identification may sound reasonable enough, but they will have unintended (or perhaps intended?  See below) consequences.  They stop people voting.  Imagine this scenario.  You decide to vote on the way back from work.  But after a busy day you realise that you forgot your passport.  What do you do?  Head back home, collect your passport and then vote?  The answer for many citizens would be head home and stay there.  That extra time is the difference between voting and not voting.  Most studies do show that voter turnout is adversely affected by voter ID requirements.  It will be interesting to see whether such academic research will have been included in the Electoral Commission’s review.

A decline in turnout is unpalatable for democracy in Britain, especially given that so few of us vote already.  However, the problem is more than that.  Studies from the US show that some people are more likely to have the forms of ID that state governments require than others.  This means that introducing voter ID can introduce a systematic bias against particular groups.  In the US these groups are racial and ethnic minority groups, those on lower incomes or with lower education qualifications.  Is there a reason to think that a British system would have different results?

The Politics of Voter ID

There is an even bigger problem.  As my book shows,  those that support voter ID and do not support voter ID have a difference: a partisan difference.  There is a rich history of US politicians using burdensome administrative requirements to try to win elections. Close elections can be won, as much by demobilising your opposition supporters, as mobilising your own supporters.  Republican politicians have therefore admitted that voter ID has been a tactic to be deployed to win them power.  Analysts agree that attempts to introduce voter ID are  ‘highly partisan, strategic and racialized’.

As I show in my book, politicians in Britain have been slower to the game.  Electoral administration laws remained unchanged for much of the twentieth century.  However, New Labour began to think that there was some partisan advantage in making it easier for people to vote and therefore changed electoral law accordingly.  Some Labour politicians thought that the higher turnout that all-postal voting and similar schemes would bring, would give them an advantage since it would get their voters out.  It was also a more politically convenient approach to solving the ‘turnout problem’  than electoral system reform.  The Conservatives have pushed the opposite approach: tightening up on  ‘voter fraud’ by introducing individual registration, despite the risk that registration levels will decline.   Further research might be able to say whether it has been partisan advantage that has driven the Conservative approach as well.  However, we can expect the largest two parties to take risk adverse positions on this issue and consider their electoral interests very carefully.

Caution is Required 

In short, there are concerns about voter fraud.  But unless the Electoral Commission reveals evidence of widespread systematic fraud, we still do not know how significant this problem really is.  We can expect that voter ID could reduce turnout unevenly, and party positions on the policy issue will be electorally motivated.  We frankly need more research.  And if Voter ID is to be considered, it should be trialled before becoming widespread.

This was originally posted on the University of East Anglia’s Politics Blog.

How could the PCC elections have been better run?

The UK is having its first elections for Police and Crime Commissioners (PCCs) tomorrow, (Thursday 15th November). Citizens will have the chance to elect an official who will have the powers to:

  • appoint and dismiss chief constables
  • set out a five-year police and crime plan 
  • determine local policing priorities
Teresa May has claimed that they will ‘be a voice for local people’.  Perhaps, perhaps not.  However, there is a real concern that their democratic role will be undermined by low turnout.  So what could have been done?  
Ways increase turnout – lessons from academic research
Low turnout is a problem because elected Commissioners may not be representative of broader public opinion.  It also reduces the legitimacy of Commissioners and perhaps their effectiveness too.
In some senses, low turnout would not be a surprise because voter turnout is generally lower in ‘second order’ and ‘third order’ elections.  Citizens vote in elections that they consider to be more important, such as general elections.  Turnout would probably have been low, whatever.
However, based on established academic research government could have done much more to avoid low turnout.  It could have:
  • Hold the election at the weekend, not a Thursday.  Why?
    • Some democracies hold elections at weekends and research suggests that this helps turnout.
  • Better still, allow voting over several days.  Why?
    • As the US election showed, many US states allow voting weeks in advance.  Why not allow voting to take place over several days?  Research shows how this can improve turnout.
  • Provide more information about candidates. Why? 
    • As I said before, research shows that providing free-post leaflets to candidates helps boost turnout.  The government decided against this and instead set up a website.  This is problematic because a significant amount of the UK still doesn’t have internet access.  A telephone line for the public has been set up, but it has received criticisms.
  • Hold the elections in May.  Why?
    • November is a bad time for electoral officials who are busy updating the electoral register.  This means that they their resources are drained and they cannot invest in public awareness activities, as they might at other points in the year.  I interviewed many electoral administrators and they were concerned about a November election.
    • More importantly, research shows that combining elections can help boost turnout.  If elections were combined with local or other elections in May, turnout might have been much higher.  Ideally, combine it with a general election.
    • There have been some claims that November’s ‘early and dark nights’ reduce turnout, I’m not aware of any research that demonstrates this.  But please correct me if I’m wrong.  
  • Plan better.
    • Research shows that errors made in the way that elections are run, voters’ interactions with poll workers etc., can undermine confidence.  The Electoral Commission issued some early warnings that planning had not progressed sufficiently at the earlier stages.  There are some reports of leaflets not being delivered on time, problems with the telephone helpful for voters and confusion about who can and cannot be a candidate.  Some teething problems are inevitable, but they might have been avoided, and won’t help turnout and public trust.
Electing Commissioners on the Cheap
The bottom line seems to be that the election could have been improved if more was spent on running it. Understandably, the government will be keen to keep costs down.   Primarily, they will be concerned about arguments from the Labour Party that the money spent on the election could have been spent on more police officers.
However, if we are to have PCCs then it is important that they are not elected on the cheap.  It undermines the eventual Commissioners, it undermines the police, and it undermines democracy