By
Rob
On
18 April 2017, the Prime Minister surprised the nation by calling an
early general election. As polling day (8 June) approached, the
British people braced themselves for the third major nationwide vote
in just over two years. Meanwhile, political pundits, pollsters,
politicians, parties, press officers and Pubquest all made plans in
preparation for the big day.
"Wait
a minute," you're probably thinking. "Did he just say Pubquest?"
That's
right, I did.
You
see, while Pubquest is ostensibly a politically neutral enterprise,
we had big plans in store for election day, which we'd concocted
about a year earlier. These plans fitted into our usual pattern of
trying desperately to find new and inventive ways to visit pubs.
Naturally, we hadn't expected to get an opportunity to put them into
practice before 2020.
That
was until Theresa May's unexpected announcement changed everything.
Despite
the rather impromptu nature of the whole ordeal, we were ready. As
the words 'general election' were still hanging in the air
outside Downing Street, Andy was busy requesting annual leave
for the big day. Meanwhile, my carefree and languorous lifestyle
required nothing more of me than to make a mental note to wake up, if
at all possible, before 1pm on the day in question.
As
for the plans, they would go as follows:
- There
are a handful of pubs in Sheffield that, when election day
comes around, also serve as polling stations
- We
would visit each of these pubs
- We
would drink a pint in them
Once
this intricate and multifaceted outline of the day's events had been
carefully drawn up, we turned our attention to logistics. After
extensive research, we'd discovered that there were five pubs that
would be serving as polling stations, none of which were particularly
close to the others. We also knew that we wanted to get all of these
pubs ticked off before the polls closed at 10pm, which meant that we
couldn't afford to embark on the sort of inordinately long
perambulations that we might otherwise have enjoyed. Therefore, we
realised that the only way forward was to get a taxi between each
venue.
Secure
in the knowledge that we'd be spending about as much on this election
as the average Tory Party donor, we pressed on.
As
the big day rolled around, Andy and I met at the first pub on the
list: the Cobden View.
A
pleasant, stone-built little building in the heart of Crookes, the
Cobden View looked like the perfect place to begin. Outside, numerous
signs confirmed that our research was correct and that this was,
indeed, a polling station.
As
it was 1:30pm on a Thursday, the barman justifiably assumed that we
had come in to vote and, as such, immediately pointed us in the
direction of the room that had been temporarily transformed into a
polling station. We assured him that we'd already performed our civic
duty and that we were, in fact, looking for refreshments.
Minutes
later we found ourselves sat in a small, rather quirkily decorated
room with two pints of Daily Bread and a pool table for company. The
pint proved to be a lovely hoppy ale from Abbeydale Brewery, who can
always be relied upon for quality. Looking around the pub, we were
fairly impressed: the seventeen-thousand different hats hanging on
the wall were a novelty. Overall, the pub had a cosy, friendly
atmosphere and a reasonable selection of beers.
It
was at this time that we began to work on our very own Pubquest exit
poll for the 2017 election.
The UK General
Election 2017 Pubquest Exit Poll:
Like
its famous, televised cousin – which is commissioned by the BBC, ITV
and Sky – the Pubquest exit poll worked on the basis of analysing the
voting intentions of those people who had just cast their ballots at
select polling stations. Also like the televised exit poll, ours
would then use the information collected to construct a prediction of
what the actual result would be.
Unlike
the televised exit poll, however, we were conducting a more focused,
micro-level study. For instance, where the wider exit poll took
results from hundreds of different constituencies across the UK,
we took ours from one (Sheffield Hallam). Additionally, while the
broadcasters' exit poll was assembled by a team of data analysts and
election experts, ours was constructed by myself and Andy, while
drinking beer. Finally, it's worth noting one last crucial
methodological difference, which is that the broadcasters' exit poll
involved actually asking people who they had voted for, whereas ours
rested heavily on me and Andy watching people walk into the polling
station and then guessing, based entirely on their appearance, who
they might have voted for.
Now,
having read the above, you're probably thinking that our exit poll
was not scientifically sound. You're also probably thinking that any
predictions made on the basis of such an exit poll must be wholly
unreliable and, frankly, wrong.
Not
so.
As
we sipped our pints and played some pool (winning one game each), we
watched as an unending line of youthful faces spilled through the
door, polling cards in hands. I don't think we saw a single voter
over the age of thirty in the entire time we were there. All of them
were young, and most were clearly students. It didn't take a genius
to surmise that very few of the Sheffield student cohort would be
turning out to vote for the local Tory candidate or, indeed, for the
local Liberal Democrat – who just so happened to be former party
leader and deputy prime minster Nick Clegg.
Our
prediction thus went as follows: despite what the bookies were
saying, the constituency of Sheffield Hallam would swing from the Lib
Dems to Labour, thus removing one of the most high profile MPs of the
modern era from parliament and turning Sheffield into an
all-Labour stronghold.
The
result?
Well,
as I'm writing this several days after the event took place, I can
confirm that the Pubquest exit poll was 100% spot on, as
Nick Clegg lost his seat to Labour. This level of accuracy is
notably higher than that of the broadcasters' exit poll, which got
the overall picture broadly right, but failed to judge the numbers
exactly. Furthermore, the broadcasters' exit poll usually comes with
a margin of error that stands at roughly 20 seats, while the
prediction put forward by the Pubquest poll was operating within a
margin of error that was 19 seats fewer!
So
maybe we should ask ourselves: do the BBC, ITV and Sky need to
rethink their methods? Does polling expert and social media
heartthrob Professor John Curtice need to spend less time in his
office and more in his local boozer? Is sipping real ale and making
guesses about people's voting intentions based purely on their age
and appearance an efficient way to predict election results?
No,
don't be stupid.
Cobden
View pool score: Andy 1-1 Rob
Pubquest pool score: Andy 60-36 Rob
Rating:
7/10