Anatomy of an Exit Poll

Working out why we got the rector election so wrong

So, Lady Rae is Glasgow’s next university rector. Who’d have thought it? Certainly not me, or this publication, both of us assumed that John Nicholson would win comfortably. In fact, I didn’t want to respond to the exit poll run by this paper because I assumed that I would only distort the responses and make the whole thing less accurate. I want to explore why the exit poll, and my own predictions were so wrong – both of us thought Nicholson would win comfortably, both assumed Lady Rae would come in a distant third place. And perhaps we find some  things we might see with tonight’s exit polls. 

Echo Chambers

I am the sort of person who likes to think that I don’t sit in an echo chamber. I have a politically diverse group of friends: those who have stood as MPs for the Lib Dems, some who worked as parliamentary staff for Labour MPs and others who have been profiled as “the only Tory on campus” for national outlets. The watch party I attended for the last election even had a non-Welsh Plaid Cymru supporter. So, when someone told me that they suspected Lady Rae might win I dismissed it as wishful thinking. The thing was even if I knew people with a surprisingly wide range of political views, we largely only met through a couple of ways – mostly through Debating which is directly affiliated with the GUU and the Dialectic Society (which helped found the GUU). The GUU, for the few of you reading this article who didn’t follow the rector elections, endorsed John Nicholson… strongly. Dialectic had initially backed Lady Rae, the president at the time being a law student, but a year later with a new president they also came out as endorsing John Nicholson. Nicholson was a former debater and so had been viewed favourably by many of my friends. Add to that a plethora of information about how the candidates interacted with people we knew and Nicholson seemed like the choice for the majority of people I knew, especially those who were politically active – therefore likely to vote. I think the fact that many of the editors of the Glasgow Guardian are in or adjacent to this circle means that it is likely that they also only saw the purple facebook profile frames and the endorsements in group chats. 

The Pandemic

The election has been long. Like the London Mayoral elections which were suspended around the same time we know the candidates.  Because of the pandemic, candidates didn’t have in person hustings or society meets in person. They also weren’t walking around, handing out flyers and having important discussions with those who might not have been otherwise reached by hustings and student politics. I think this turned out to be crucial. This year a lot of student life has seemed slightly more sluggish than normal. Turnout at a lot of society events has been much lower and first years have struggled to integrate beyond their tiny accommodation bubbles, as socialising in many ways has been made illegal. Because of this the importance of group chats and course chats has been increased dramatically. Society chats are fine if there is a reasonable amount of people in that society – for example the Cecilians are notoriously huge and so had they endorsed a candidate it is likely that would have led to a number of votes. Looking through the societies the more influential in terms of votes who gave an endorsement were probably:

1. The Catholic Society (Lady Rae)
2. GULGBTQ+ (Nicholson)
3. Medico-Chirurgical Society (Nicholson)
4. Adam Smith Economics Society (Nicholson)
5. The Law Society (Lady Rae)
6. The Accounting Society (Lady Rae)

There are a large number of societies who also endorsed each of the candidates which probably led to significant numbers of votes but my guess would be these were the top six. They have a few things in common. Firstly, these are big societies. I think these are quite possibly amongst the biggest societies on campus in terms of sheer numbers, outside of sports clubs. Secondly, aside from GULGBTQ+ and the Catholic Society these are all associated with a degree. Unlike some other societies associated with degrees many, if not most, students in these degrees are involved with these societies for at least part of their time at university. These are also degrees associated with very competitive jobs, so these societies offer significant networking opportunities while an undergraduate. Which is why of these four societies the most important with regards to the rector election is the Law Society. 

Because Lady Rae is a lawyer the Law Society was unique in engaging their members. Maybe various people in Med-Chir were excited for Nicholson to be rector, but it was unlikely they considered the campaign for him to be a potentially important future networking opportunity. This is clearly reflected in quite how active the Lady Rae campaign was with a number of very vocal law students. The number of vocal Nicholson supporters was also quite high but they seemed to mostly come from the GUU which this year has taken a hit to its visibility on campus as many of their most popular events did not translate well to remote hosting (Zoom Hive anyone?). Because of this I think that the number of GUU loyalists was probably lower this year than it usually is. 

Party Politics

Another point which is not as key in deciding the winner but still important to consider are the candidates themselves. Each candidate had their merits and I’m sure each candidate would have done a good job. However, two of the candidates were affiliated with the SNP and one was non-political. Therefore, those on campus who dislike the SNP felt they had just one option to choose, as well as those who just wanted a non-political candidate. Equally because Nicholson and Ashraf were both affiliated with the SNP, they split the vote. You might expect that Nicholson, as an MP would take precedence over Ashraf who is a councillor but the Scottish Nationalist Association (the SNP society on campus) backed Ashraf instead. The SNA are large and mostly dedicated so had there been one SNP candidate the result might have been different but because their vote was split no-one could capitalise from an “SNP” vote. 

Anatomy of an Exit Poll

Finally, back to the exit poll itself. In 2017, during the last election, there was a turnout of 8210, this year saw less than a quarter of that number. The Glasgow Guardian exit poll had 84 respondents. Even if they (rightly) assumed that the pandemic would depress turnout, this is still less than one percent of potential turnout. This is not a particular fault of The Glasgow Guardian. During normal times, it is likely that they would be out and about and gathering a higher number of exit poll votes but with the pandemic and no platform available to them to gather many responses they were left with only 84. The only possible platform freely available to them that I can think of would be posting the exit polls on Glasknow or sending a mass email, though this would still preference those who checked their email often or read Glasknow. The issue came instead when it was up to individual people who were associated strongly with a particular candidate to fill in and then share a poll. Much of the sharing of the poll didn’t permeate law chats or accounting chats which could have served to suggest a success for Lady Rae. The exit poll therefore had her coming a distant third, and I believed it because it was a poll of my friends and so was very accurate with regards to them. 

Next Time

This is easily remedied next time when there (hopefully) will not be a pandemic forcing people inside. Most students spend some of their week on campus and asking for exit poll votes outside the library or by other buildings with high footfall would likely provide a more representative sample. This might even play a role tonight, as many votes are likely posted and come from specific groups of people, like those who are elderly or feel like voting in person might still present significant risks of Covid infection.

That being said, for me, I don’t know if this is so easily remedied. I am part of a society on campus who didn’t officially endorse a rector candidate, but I would be surprised if any of them did not vote for Lady Rae first preference. To me they seemed like a small dissenting voice, almost alone. I was completely blindsided that they were the majority. In trying to figure out who was the largest society on campus I asked a number of my friends who gave me a similar list. The people with flatmates doing theatre preferred STAG, while those who did economics promised that ASES was king but when I moved outside of people I’d normally talk to I got a completely different list. To some extent I think these bubbles are inevitable and the only way to fix it is to speak to everyone freely and without discrimination. This is easier said than done, especially during a pandemic though. Hopefully as things open up chatting to people on the bus or overhearing snippets from people in cafes will once again be an option. Until then I’ve got a new set of rules around upcoming elections. Don’t believe what you see, read without prejudice, and finally; and no one is better at exit polling than Sir John Curtice.