Churchill famously said that “democracy is the worst form of Government, except for all those other forms that have been tried”. But our democracy has some unique features, which make it worse than others. The new report from the Parliamentary Select Committee on Levelling Up, Housing and Communities notes that the the Electoral Integrity Project ranks the UK 58th out of 169 countries. The Electoral Commission describes the current system as “neither effective nor efficient” and has called for “urgent reform”. One often neglected issue is voluntary registration to vote, which means that some communities, predominantly left-leaning ones, are systematically underrepresented.
Voting: a right or a duty?
In many countries, voting is seen as a civic duty. Most famously, in Australia, voting is compulsory, and you can be fined for failing to vote. By contrast, in the UK it is a right, which you claim by registering. The Committee estimates that eight million people of voting age are not registered, through choice, inability or ignorance, and many more are inaccurately registered.
The official figures for the 2019 general election report a turnout of 67.3%. That is two thirds of registered voters, but only 58% of the voting age population. This means that although only a quarter of people of voting age voted for the Conservatives, they won a very large majority in Parliament. Hardly ‘the will of the people’.
The ‘missing voters’ are not evenly spread around the country
The Electoral Reform Society (ERS) has compared the population figures for each constituency with the number of registered voters. It is no great surprise that the proportion of unregistered adults is much higher in ethnically diverse areas and places with large transitory populations – young people, students and private renters. Disabled adults are also less likely to be registered. In five constituencies (in London, Leeds, Bristol, Sheffield and Liverpool) more than 20% of adults are not registered. Those are, in general, Labour voting areas.
The same pattern is evident in the East of England. In 2019, there were 58 constituencies, with an average of around 77,000 registered voters. Among these 58, nine had fewer than 11,000 ‘missing’ voters. All these have Conservative MPs, with the fewest missing people in the safe Conservative seats of Castle Point and Rayleigh and Wickford. At the other end of the scale, there are six constituencies with more than 15,000 missing voters, with the largest numbers in Cambridge and Peterborough: both with Labour MPs. In five of these six, the number of missing voters is larger than the sitting MP’s 2019 majority.
However, there is a second problem.
Some constituencies are much larger
Constituency boundaries are reviewed from time to time, to reflect changes in population. In the latest revision, which will apply in this year’s general election, the Electoral Commission was asked to ensure that constituencies were of roughly equal size. But this was calculated, not using the size of the population, but the number of registered voters.
As a result, the new constituencies all have similar numbers of registered voters, but some have much larger populations. These constituencies have more social problems, and since a major part of an MP’s role is dealing with constituents’ problems, more constituents means more work, and the MP’s effort is spread more thinly. This is not equal representation.
Systematic discrimination
So the system discriminates systematically against young people, renters, ethnic minorities and disabled people, and the broad left. It does this: firstly by simply leaving out anyone who doesn’t actively try to register, and then by giving those communities less representation. Peter Kellner has calculated that the average Labour MP represents 114,000 residents, while the average Conservative represents only 97,000. As he says
MPs representing young and diverse inner city seats have to serve much larger populations of constituents than MPs representing older, rural seats with high registration rates.
He calculates that, had population been used, rather than registered electors, London would have had 85 MPs rather than the 75 now allocated. It is highly likely that most of those extra ten MPs would have been Labour. He estimates that, if population had been used to set boundaries, the Conservatives would probably not have achieved a majority in 2010, 2015 or 2017.

Finally, even among those who are on the register, there are some who do not have the recognised ID documentation. In some. predominantly Labour-voting. areas of London, that is as high as 20%.
Solutions
To encourage registration, there are usually registration campaigns in the run-up to general elections. However, these are expensive, and put the electoral registration system under great pressure. In 2019, such campaigns resulted in three million people applying to register in the last month. The ERS estimates that, as a result, as many as one in ten current registrations may be inaccurate.
The Electoral Reform Society has welcomed the Parliamentary report, and especially its proposal to make voter registration automatic, as it is in many countries. People would be automatically added to the electoral roll using data which government already holds – like driving licences, passports or NHS records. This would substantially enlarge the roll, and might increase participation in elections. In Sweden, where everyone on the Tax Agency’s population register is automatically sent a polling card, 96% of voting age adults were registered, and election turnout was 84%.
Of course, the system would be far more democratic if constituency sizes were based on population, and if we switched to some form of proportional representation. But these are bigger issues, and implementation will take longer, and more debate. In the meantime, the next general election will be conducted under the current rules.
So, for the time-being, the system will continue to favour the old, the rich, and the right. Not very democratic at all.










