Updated (3 hours before polls close, June 23)

Interesting piece in The Telegraph that shows slight lead for Remain.

There will not be an election-style exit poll, but YouGov plans to release an election day poll at the moment the polls close (2pm PT / 5pm ET).  If there is going to be a shocking outcome, the first glimpse may be right then.

Gamblers are 84% certain of a Remain victory.  Do they know something we don’t know?  They are probably reading the polls as their main source of information.  If the polls are wrong, they’re wrong. UBC’s Sauder School Election Prediction market is made up of bettors betting real money to predict election outcomes.  In 2013, 85% predicted an NDP majority government; in 2015, only 20% predicted a Trudeau majority.  So, a sucker is born every minute.  We’ll see if the Brits are better bettors.

Professor John Curtice reiterates today that the polls can’t be trusted and it’s basically a crapshoot.  The public pollsters notoriously got it wrong in May 2015 and it was Curtice who conducted the election day exit poll that predicted the majority no one expected.  We might just have to wait until the votes are counted!

———-

John Curtice (@whatUKthinks), the polling expert who shocked the UK when he predicted a Conservative majority one minute after the polls closed in the 2015 General Election, says Brexit is too close to call.

The markets and the bettors are predicting a victory for Remain.  Political betting analyst Mike Smithson (@MSmithsonPB) has provided data that shows gamblers moved toward Leave about a week ago, but there was a sharp upturn for Remain after the Jo Cox murder.

Figure 1: Remain is a better bet for 75% of UK bettors.

What are the markets doing?  British Sterling surged over the past few days, pricing in a Remain outcome.

Figure 2:  GBP climbs in final week

And finally, what are the polls saying?

Figure 3: Compilation of Brexit polls (@MSmithsonPB)

There has been a slight advantage for Remain via Telephone surveys and a slight advantage for Leave via Online surveys. While the markets and the bettors have turned toward the Remain camp as the likely outcome, those responding to polls are still very divided.

YouGov’s detailed tables reveal some of the underlying divisions in British society concerning Brexit.

The numbers shown in Figure 3 above are Decided support, however, as YouGov reminds us, there are still undecided voters (9%) and those who say they won’t vote (4%).  On decided vote, YouGov has it at 44 Remain, 43 Leave.

According to YouGov, here are some of the dividing lines:

Remain voters

• Labour (64%) and LibDem (59%) voters
• 18-24s (64%) and 25-49s (45%)
• London (50%) and Scotland (56%)
• Upper/Middle class – ABC1 voters (53%)

Leave voters

• Conservative (55%) and UKIP (95%) voters
• 50-64s (49%) and 65+ (58%)
• Rest of South (45%); Midland-Wales (51%); North (47%)
• Skilled working class/ working class/non-working -C2DE voters (52%)

When asked who is most certain to vote, Remain was at 79% and Leave at 84%.

On the dividing lines, there is a fundamental generational difference.  The range between young and old is stark.  The Euro debate, and underlying views on immigration, shape partisan leanings as evidenced in the Party ID splits. Class is also significant.  Then in Scotland, attitudes are tied somewhat to Scottish identity – leave Britain, but stay in EU.

But who will vote?

One would expect a high turnout.  YouGov indicates a tilt toward Leave voters.  Young people are expressing a strong preference for Remain but election turnout studies consistently show they vote at a lower rate.  Will they close the gap, like they did in Canada’s federal election, in the Brexit vote?  Will lower income voters vote at the same rate as higher income voters?  How will the UK’s sizeable immigrant communities vote and will they turnout to vote?  We don’t know this from the YouGov poll, possibly because it’s online which is typically less representative of people who don’t speak or read English well.  Telephone surveys are better in including those populations which may explain a slight leaning to Remain.

Going back to John Curtice, he is the most credible voice in the UK on polling and he believes it’s too close to call.   Curtice says the result could split the difference between the aggregate of phone polls which have Remain at 51% and the aggregate of online polls that have Leave at 51%.  A cliffhanger like the Quebec referendum of 1995.

This process ends in a vote and an outcome, but this discussion of the cold, hard numbers comes just days after a shocking murder of Labour MP Jo Cox.  This is no ordinary vote.  The referendum campaign has exposed the fault lines of UK society.  The stakes are extraordinarily high, especially in the context of a campaign that appears to be a photo-finish.

I’ll bet on Remain, based on voters pulling back due to perceived risk, like they did in Quebec in 1995 and in the recent Scottish referendum.  We’ll see if the gamblers and traders got it right.