The Liberal Democrats are profiling every voter in the country by rating their political preferences.
Sky News can revealed how the LibDems seem be copying practices used Cambridge Analytica.
The percentage ratings – there are at least 42 in total, although the identity of only 37 are known – estimate whether someone voted Leave or Remain in the 2016 EU referendum and predict how they would vote if there was a second poll in 2019.
Other scored characteristics include “Likelihood of being a Labour voter in 2019”, “Likelihood of being a core Lib Dem” and “Net difference in likelihood of voting for the Conservative or Brexit Party in 2019”.
The system, which uses a sophisticated computer model to generate the scores, also assesses personal outlooks, giving a percentage to “Likelihood of being a pragmatic liberal”.
The Liberal Democrats also use software which estimates the age and first language of voters by analysing their names.
The name Rowland Manthorpe, for instance, is categorised as “older: probably older”.
This kind of scoring system is commonly used by political parties in the UK and the US, but new data transparency laws which allow anyone to ask for their data give an unprecedented glimpse into how a major party operates.
Matthew Rice, Scotland director of digital campaign organisation Open Rights Group, was assigned nine scores by the Liberal Democrats. In a possible 2019 election, it gave him a 15% chance of voting Liberal Democrat, a 22% chance of voting Brexit Party and a 20% chance of voting Conservative.
In a second EU referendum, Mr Rice was deemed 77% likely to vote Remain.
“The scores don’t necessarily represent my values, but for all intents and purposes, for the Liberal Democrats, this is me,” he told Sky News.
“They’ve decided where I sit on a political spectrum and in a general election they will define me based on these scores. That does feel quite intrusive. It doesn’t feel reflective of what you would want from a democratic culture.”
The data used to create the scores comes from a range of sources, including the UK electoral register, phone and doorstep canvassing, anonymous online surveys, and publicly available data such as census area classifications, which categorise different regions according to their populations.
The Liberal Democrats also employed “consumer/market research data”, which it bought from a third party. This information on lifestyle and financial behaviour is frequently used by political parties to target voters in specific areas.
Asked about the purpose of the scores, the Liberal Democrats told Sky News they were used to decide which people would receive campaign leaflets.
The party denied that the scoring system was used to target ads on social media, although Rob Blackie, a digital strategist who is standing as a Lib Dem candidate for the London Assembly, said that was technically possible.
“The voter scores will tell you that certain types of people are more likely to vote for a party. People who are younger who’ve been to university are much more likely to care about Brexit and to be anti-Brexit, and people who are older are much more likely to be pro-Brexit,” said Mr Blackie.
“You could use the Facebook targeting tools to target people you’ve identified through your voter scoring, although it is much harder to target individuals since the changes in the law, and the Lib Dems tend to be very cautious about that.”
Data protection law prohibits processing of “special categories of personal data”, including data on political affiliation, but grants political parties an exemption as long as there is “substantial public interest”.
However, data protection experts warned that the Liberal Democrats’ automated profiling fell into a legal grey area and could be challenged by the data watchdog, which is currently investigating political parties’ use of data.
“I’m at a loss to understand the lawful basis the Lib Dems have relied on for this,” said Pat Walshe, managing director of privacy matters.
“The information commissioner’s office should look at this and investigate again the use of personal data by political parties, because that profiling is deeper and more intense than the average person would expect and understand.”
Mr Blackie also defended the scoring system, saying it helped save weary voters from unwelcome canvassing.
“The key thing for all political parties is, we don’t want to knock on the doors of people who hate us, so one of the things it helps us to do is only talk to the people who want to hear from us,” he said.
“Generally speaking it does help people by improving the targeting.”
Labour Heartlands supplement.
We urge everyone make a Subject Access Request (SAR) to the Liberal Democrats after it has emerged that they have gathered data on EVERY registered UK voter.
Including political data which requires a valid excuse to be exempt.
It’s been labelled it a legally dubious use of the political data exemption under GDPR:
“Subject Access Request” give your,
• any aliases
“SUBJECT ACCESS REQUEST” SAR.
To whom it may concern,
I the above named, am requesting all data your party holds on me.
I am entitled to this information under the Data Protection Act 2018 & GDPR and have followed the guidelines as set out by the ICO here:
It must also include any percentage ratings you have given me based on your machine learning algorithms OR by any other method, and they must be broken down by each of the 42 different ratings your party uses, which your spokesman confirmed to Sky News.
I am entitled to know the ICO have advised.
The ICO have further advised me that I am entitled to know where you obtained each type of data. It must be broken down by data type and source
Please provide it to me in paper form as is permitted to the address listed above.
The easiest way to contact the Liberal Democrats is using this form.
If your query is urgent, you can also call them on 020 7022 0988 between 0930 and 1730, Monday to Friday. You can also contact the LibDems by emailing firstname.lastname@example.org
If your query relates to your data rights, please email email@example.com
Article explaining all the data being gathered and their scoring system.