Just Register is a national campaign to improve voter registration in the UK.
To give us a baseline of under-registration, we wanted to check to what extent voter registration rates varied — and by extension, our representation in Parliament — across 650 parliamentary constituencies in the UK.
While voter registration is often promoted as a civic duty that gives people the chance to participate in democracy, there is another important way that registration impacts people: representation in Parliament.
In the United Kingdom, parliamentary constituencies are drawn based on the numbers of registered voters on the electoral register and not the total population living in an area.
In the 2023 boundary review cycle, constituencies were required (with limited exceptions¹), to contain an electorate within +/- 5% of the UK electoral quota.² The electoral quota for this review was 73,393, making the range of electorates 69,724 to 77,062.³
For this review, this quota was based on the size of the parliamentary electoral register in March 2020.⁴
Despite this attempt at equalisation, differences in registration rates mean that MPs may represent very different numbers of people.
To find out how large this gap was, we compared the total population of each constituency with the number of electors on the electoral register, and ranked the constituencies based on their registration rate.
See the full dataset at this link.
To develop registration rates, and rank them from 1 until 650, we looked at three sets of data:
The registration rate can be calculated by dividing the total numbers of voters on the parliamentary electoral register by the total population living in the electorate.
The electorate data is straightforward: each of the Boundary Commissions have listed the electorates of the new constituencies for England⁵, Wales⁶, Scotland⁷ and Northern Ireland⁸, as of March 2020. These are the numbers that were used to draw the new boundaries.
We chose to measure the total population of each constituency, rather than the eligible voter population for several reasons.
While MPs are allocated based on registered voters, their job is to serve the entire constituency. They take on the responsibility of representing their constituency as a whole, both in a practical sense through casework, and in a more intangible but no less important sense of the obligation to look after the interests of all people in their constituency.
That includes those not eligible to vote in parliamentary elections, like children and young people under 18, and non-qualifying immigrants to the UK (some of whom are eligible to vote in local elections).
Many democracies allocate their representatives in at least one house of their legislature based on the full population. Countries including the United States, Canada, France, Germany, India and New Zealand all have census-based systems for allocating representatives.
The ONS (Office for National Statistics) has released population information for England and Wales from the 2021 census, for the new constituencies resulting from the 2023 boundary review process.⁹
NISRA (Northern Ireland Statistics and research agency) shared the same information for the new Northern Ireland constituencies.¹⁰
Comparing each constituency’s electoral register with the population allows us to compare the actual population with the number of people represented on the electoral register.
As of January 2024, Scotland’s census had not released population data for the new Westminster constituencies. Additional geographies are slated to be released in spring 2024, but in the meantime a different process is needed compared to England, Wales and Northern Ireland.
For 10 constituencies, in Scotland, the boundaries were unchanged¹¹. For these constituencies, we used the census 2021 data for each of the previous constituencies.
For the remaining constituencies, we used another method. While this makes it more difficult to compare across Scottish constituencies, we chose to prioritise matching the method used for all other nations where possible, to get the most accurate UK-wide basis for comparison.
It is possible to estimate the populations of the new Scottish constituencies, using the available ward-level data. Data for Scotland is from the National Records of Scotland Electoral Ward Population Estimates from 2021.¹²
The basic building block of the new constituencies is generally the electoral ward. This means that we can find the population of each constituency by adding up the estimated populations of the wards that make up the constituency. The Boundary Commission for Scotland has provided a list of wards in each constituency.¹³
In some cases, a single ward was split between different constituencies. Where this was the case, the Boundary Commission for Scotland has also provided the portion of the ward’s electorate within each constituency.
For these split wards, we divided the known population of the ward in the same proportion as the electorate was divided. This means that there is a slightly higher level of uncertainty for constituencies with split wards.
Our process was to identify which wards made up the new constituencies, match each ward with its reported population figure using the ward’s GSS codes, and then add those populations to estimate the population of the new constituencies.
While using a different method for these constituencies does mean that there is a greater level of uncertainty for these constituencies, we are confident that the results are still useful.
We tested this process with 8 Scottish constituencies where both the total constituency population and the ward-level data population were available. For these, estimates obtained by using the ward-level figures came very close to the census figures. Ward-based estimates ranged from 335 fewer people (out of a population of over 45,000 in Orkney and Shetland) to 1,138 more than the census figures (out of a population of over 97,000 in West Aberdeenshire and Kincardine). No ward-based estimate varied by more than 1.2% of the census figure in our sampling.
We plan to update this data when the Scotland data becomes available, but in the meantime are confident that these numbers provide a reasonable basis for comparison.