The Twitter ElectionsContinuing an analytical exploration of the UK government and electorate as they are on TwitterChris BrownlieBlockedUnblockFollowFollowingJul 6The first piece of analysis in this series analysed politicians of the UK and how they compared to each other on Twitter.
This article will consider the other side of the coin.
Instead of analysing the ‘House of Tweeters’, I will instead investigate the the Tw-electorate (the UK public on Twitter that follow MPs – excuse the terrible attempts at wordplay).
A note on TrumpIn the first post, I made a point that caught the attention of several readers: the fact that Donald Trump has more than 60 million followers and the entire UK government has a combined following of around 20m.
Even if you take into account the difference in population, he still has vastly more followers than any individual UK politician.
However, I mentioned that this didn’t take into account people who followed more than one UK MP, so it wasn’t in fact a fair comparison.
The actual difference in followers is far more significant.
There are 5.
8 million unique Twitter accounts that follow at least one member of the UK parliament.
There are 61 million unique Twitter accounts that follow Donald Trump.
The population of the US is roughly 5 times larger than that of the UK so even if you account for population size, Trump has more than twice as many followers as the entire UK government.
The rest of this post will look at who these 5.
8 million people are, who they follow and how they interact with each other.
The (Unique) House of TweetersMy original analysis includes a table that shows a combined following for each political party.
While this is a valid visualisation to show the total level of engagement each party receives on Twitter, it doesn’t properly represent the real life following of each party.
So the table below shows how many unique followers each party has.
League table showing the political parties of the UK arranged by their unique following on Twitter.
As is to be expected, this change in metric affects parties with many MPs more significantly.
The only change in the ‘leaderboard’ is the SNP moving below the Green Party (Caroline Lucas).
Despite the SNP having 35 elected politicians compared the Green’s solitary Caroline Lucas, their collective reach is not as wide.
The Twitter ElectionsThis still doesn’t accurately describe the Twelectorate though as it violates the conventional voting logic of one vote per person.
So for each of the 5.
8 million individuals in the Twelectorate, I’ve calculated their political allegiance by the MPs they follow – e.
if they follow 4 Labour and 2 Conservative MPs I will mark them as a Labour follower.
For ties I will divide their vote equally, as I have no other way of breaking the tie.
Below you can see the results of this Twitter Election:League table showing the number of Twitter-votes each party got and the number of seats they won in the Twitter parliament.
This was calculated by dividing the 650 seats equally across the 5.
8 million ‘voters’.
The table above indicates the number of seats a party wins when determining allegiance in the way I have set out above.
Note that seats in the Twitter Parliament are a) proportional to the number of votes the party received, b) geographically unrestricted (constituencies don’t exist) and c) only considers currently elected MPs as able to compete for votes.
Below you can see the comparison between the two parliaments visually (with the majority government on the lower set of benches)Visualisation showing the difference between the current UK parliament and the theoretical Twitter parliament.
Each circle represents an MP, with the two sections showing each side of the House (the lower section representing the majority government).
Visualisation created using Rob Hickman’s ggparliament package.
The first and most obvious difference is that Labour have an outright majority in the Twitter parliament, claiming 343 of the 650 available seats.
On the other side the Conservatives struggle considerably, only managing to win 233 seats – 79 less than in real life.
The other big winners in the Twitter Elections are the Liberal Democrats (+20) and the Green Party (+11).
All other parties perform worse on Twitter than in real life, showing that Labour, Lib Dems and Greens are the top performers on social media.
A well balanced feedOne of the main areas I was keen to investigate was how much cross-over there is between the followers of different parties.
One of the biggest problems in modern politics is the electorate’s unwillingness to hear views that don’t match their own.
This manifests online as a ‘filter bubble’, whereby their own ideologies and biases are continually reinforced and never challenged.
So to investigate this I looked at all the insulated people on Twitter and the results were worse than I was expecting.
Of the roughly 5.
8 million people who follow a UK MP on Twitter, 4.
25 million only follow members of a single party.
Of these, 3.
4 million only follow a single MP.
Now perhaps this isn’t necessarily as bad as it seems because it only takes into account the MPs that people follow (not news corporations or other influential figures etc.
) and you could also argue that it is reasonable for people to only follow their local MP.
However, I think if you consider that of all the people on Twitter who are interested in the UK government, the fact that 73.
3% get their online picture of the government from a single party is both significant and worrying.
Below you can see which parties have the highest number of these ‘isolated’ followers.
Table showing how ‘isolated’ each party’s followers are.
A follower is isolated if they only follow MPs from a single political party.
The two main parties have the highest proportions of isolated followers.
This is likely because they are higher profile: if someone is only going to follow a single MP on Twitter it is far more likely to be Theresa May or Jeremy Corbyn than an MP for Plaid Cymru or Sinn Fein.
The Liberal Democrats and The Independent Group (Change UK) are the two least isolated parties.
This makes sense for The Independent Group as they were all a member of a different political party until recently, so their followers are likely to also follow MPs from their old party.
I think it is again concerning that for all parties at least half of their followers have no exposure to different political parties.
Now to consider at the 3.
4 million people who only follow a single MP.
Below is a list of MPs arranged by how isolated their followers are, showing the 10 most and 10 least isolated.
Here the ‘Isolated Followers’ column is the number of people who follow only that specific MP and none others.
Table showing the top and bottom 10 MPs arranged by how isolated their followers are.
As mentioned in a previous point, if a non-political person is going to only follow a single MP it is disproportionately more likely to be the leader of one of the two main parties – Corbyn or May.
Having said that, the fact that almost a million of Corbyn’s Twitter followers don’t have any other interaction with the Government on Twitter is very surprising – and perhaps a side effect of his unique (for a politician) involvement with popular culture.
Aside from the expected inclusion of these two, the rest of the table doesn’t appear to have a particularly obvious pattern.
The other MPs with isolated followers tend to be less well known and are more likely to be Conservative – these isolated followers are likely members of their constituency.
Also as expected, two of the lesser known members of The Independent Group are some of the least isolated MPs.
As Gapes and Leslie were both Labour MPs until early 2019, their followers likely still follow other Labour party politicians.
I think overall this analysis shows thata) Labour have the edge over the Conservatives in the Twitter Elections,b) Trump would have the edge over the UK in the Twitter Elections andc) the ‘filter bubble’ is a real and dangerous part of the Political Twittersphere of the UK.
What next?Thanks for reading, I hope you enjoyed following along in these Twitter elections!.If you found this analysis interesting then follow me here or on Twitter for updates on the next instalment.
To follow on from this I’m going to see if I can find any interesting patterns among the characteristics of our 5.
8 million-strong electorate, for example is it possible to guess which political party a Twitter user follows based on their profile?.I’ll also soon be switching back to looking at the politicians again- analysing the words they use in their tweets to see if there are any common themes among parties!.