What Happened In Nova Scotia? Can It Happen Everywhere?

August 19th, 2021 | RR

With what was supposed to be a Liberal win, the province of Nova Scotia has stunned the country. Going into the province's 41st election, polls predicted a likely Liberal government. Instead, what unfolded was a strong Progressive Conservative majority of 31 seats. Like it was a walk in the park, the provincial PCs were able to easily unseat the long-serving Liberals and their new leader, Iain Rankin.

Expecting some kind of Liberal government, many Nova Scotians went to bed on Tuesday night only to wake up to a surprising PC majority on Wednesday. It became the first time since 2006 that Nova Scotia embraced anything but a left-wing government. It's even possible that Justin Trudeau called a federal election on August 15 to strategically coincide with what he expected would be a Liberal win in Nova Scotia. He probably intended to use a strong Liberal win in the East to rally federal Liberals, but instead ended up having to watch Erin O'Toole congratulate a new PC premier-elect.

What happened in Nova Scotia was surprising because of what the pollsters and media were saying for weeks and days leading up to the election.

Before we turn to the old argument about opinion polls being “rigged”, let's keep in mind that pollsters accurately predicted the past two provincial elections in Nova Scotia. In 2017, both Forum Research and Mainstreet were off by only 2%, respectively. In 2013, Forum and Abacus almost nailed the results down to the decimal. This time, something dramatic happened.

In 2021, the same pollsters got it completely wrong.


(article continues after ads)

The first indication that something was brewing would have been seen in the turnout. As is the case 90% of the time, a higher turnout from the previous election is generally a bad sign for an incumbent government. It happened to Donald Trump, it happened to the Alberta NDP, it happened to Stephen Harper and now it has happened to the Nova Scotia Liberals. Usually, the bigger the increase in turnout, the more likely it is for an incumbent to be defeated. In some cases, a small increase in turnout is not always fatal.

Turnout in Nova Scotia increased to an estimated 55% in 2021, up from 53% in 2017. It's not a massive shift, but it was enough to rattle the status quo.

A higher turnout is an indication of an unsettled and angry electorate. In rare cases it's a sign of enthusiasm, but usually it's just anger. In 2019, Albertans showed up in historic numbers to unseat Rachel Notley's NDP. It can be argued that discontent, fear, and panic about the pandemic led to Trump's removal—which also happened during a record-breaking voter turnout.

When people are angry or scared, they vote.

A mix of anger and enthusiasm ousted Harper in 2015, when turnout jumped to 68%. That was a 7% increase from the 2011 election. In BC's latest provincial election, turnout was brutally low and resulted in a big win for the incumbent NDP. Almost always, turnouts can be a more valuable indicator than opinion polls. The only problem is, we still can't definitively know the turnout until after election night.

Before the pandemic, early voting and voting by mail were less common—which made it difficult to see or predict turnouts. Things have definitely changed in the game of prognostication. In the 2020 presidential election, we could watch the turnout happening in real-time. On the eve of the election, more than 100 million Americans had already mailed in their ballots. In Nova Scotia, more than 10,000 ballots had already been received from early voters by July 30. At the same time in the 2017 election, just under 4,000 ballots had been received. In total, more than 105,000 people voted by mail in Nova Scotia.

More Canadians will vote in the federal election early and by mail than ever before. That makes it difficult to predict overall turnout before election night, but Nova Scotia may be an early indicator of what might happen in other provinces during the election and how voters may approach the government's pandemic response. As of now, it looks like Nova Scotians weren't willing to reward the provincial Liberals for their efforts during the pandemic, despite most polls showing they had a favourable opinion. It is possible that jobs and social issues have become a top priority for more Nova Scotians and that Liberals were given a failing grade on everything other than the pandemic. Therefore, the anger and discontent brought out the votes.

Since voting by mail is fairly new, numbers will be naturally higher than normal and may not indicate an overall higher turnout. However, there could be some indicators from elections that have happened since the pandemic began.

"A higher turnout is an indication of an unsettled and angry electorate."

In BC, where turnout was low, more than 700,000 people voted by mail. At first glance that seems like a big number, but not if you look at the average turnout from past elections. On average, in a normal BC election, two million people will vote. In the United States, up to 130 million people will vote. When mail-in and early ballots seem to exceed a certain percentage of the average turnout, we may be able to assume that overall turnout will be higher than normal.

There will always be a percentage of people that will vote in person, no matter what happens. If mail-in ballots reach a certain capacity within the normal turnout, we can assume that a significant percentage will still vote in person. In BC, 700,000 votes were cast by mail, which equates to 35% of the usual voter turnout. In the United States, 100 million votes were cast by mail, which equates to about 75% of the usual turnout. In BC, overall turnout was low. In the United States, overall turnout was high.

In New Brunswick's pandemic election, 133,000 people voted by mail, which is about 35% of the average turnout. New Brunswick's overall turnout dropped by 1% from the previous election and the incumbent PC government won a majority.

Newfoundland had the country's first mail-in only election in March, so there isn't much data to use there.

Nova Scotia is an outlier. There, approximately 105,000 people voted by mail, which makes up about 25% of the usual turnout. Turnout increased, but not dramatically. Unlike New Brunswick and BC, Nova Scotia's election came after most of the hype and hysteria about the pandemic had subsided. Some might argue the pandemic is officially over and that the rules have changed again, leaving us with no new data to rely on.

Pandemic numbers against post-pandemic numbers create different results. We probably won't be able to pinpoint federal turnout in September before election day, but if mail-in and early ballots exceed 35% of the usual turnout by September 19, Justin Trudeau might face the same fate as Iain Rankin.

On average, up to 16 million people have voted in the past three federal elections. 35% of that would amount to approximately 5.5 million. If we were to use the indicators from BC and New Brunswick, Justin Trudeau might be safe with that many mail-in ballots by September 19. If more than six million people have voted by mail on September 19, it would be safe to assume that overall turnout will be higher than 2019 and that Justin Trudeau could be in trouble.

If media and pollsters start indicating a higher turnout on election day, coupled with six million mail-in ballots, we can be sure that the incumbent-turnout  rule will give us a new prime minister, or a very slim Liberal minority—which could also lead to a new prime minister.

© 2021 Poletical