Why the equalization referendum isn't likely to sway Calgary's municipal election

EDITOR'S NOTE: In the lead up to the municipal election, Jack Lucas, who teaches political science at the University of Calgary and is part of the Canadian Municipal Election Study (CMES), is writing a series of columns to help us understand the mechanics of how a city election works. The CMES is a multi-year series of surveys analyzing how and why we vote as we do.

In June 2020, when the Alberta government announced it would hold a referendum on equalization alongside the 2021 municipal elections, many municipal representatives were angry.

"I think it is disrespectful to citizens who want to vote in their local elections," said Mayor Naheed Nenshi at the time, "to have their local elections — the coverage of them as well as the voter turnout — influenced by non-local issues."

In the best-case scenario, critics suggested, the referendum would be a distraction. In the worst-case scenario, it would affect the actual outcome of the municipal election.

Since last year, the questions have multiplied.

Alongside our standard votes for mayor, councillor and school board trustee, Calgarians who visit the polls on Oct. 18 will also cast a vote on equalization, daylight saving time and community water fluoridation, as well as choose up to three nominees for the Senate of Canada.

It's an abundance of questions on an abundance of topics — a challenge for voters, especially when the municipal vote follows so quickly on the heels of the federal election.

Could this affect the municipal election? Will we see an outcome different from what we might have seen in the absence of the equalization referendum?

Probably not. Let me explain why.

Pathways to influence

There are two main ways a referendum can affect a municipal election. Let's call them the issue pathway and the mobilization pathway.

In the issue pathway, a referendum issue becomes the only thing anyone talks about, and comes to define the choices that voters make about municipal candidates.

For example, imagine if the equalization debate became so contentious that Calgarians simply voted for the municipal candidates who shared their view on that one issue — to the exclusion of all others.

If that happened, we could reasonably claim that the presence of the equalization question changed the outcome of the municipal election itself.

The mobilization pathway is when a referendum influences the outcome of a municipal election by reshaping who actually turns out to vote.

Suppose that the people who ordinarily vote in municipal elections prefer Candidate X, and those who don't vote in municipal elections prefer Candidate Y. A referendum question comes along that gets the non-voters so fired up that they feel compelled to participate. While they're at the polls, these voters also cast a vote for Candidate Y.

As a result of this influx of new voters, Y wins. Here, too, the referendum has influenced the outcome of the municipal race.

Last year, when the provincial government first announced the equalization referendum, the issue pathway was the biggest worry for many.

Brooks Mayor Barry Morishita, then president of the Alberta Urban Municipalities Association, said the equalization referendum would "just drown out the people running on local issues for local consideration. There would be no other discussion except what's on that referendum."

Several Calgary councillors shared these concerns.

So far, there's no evidence that this is happening.

Not much influence

Municipal candidates in Calgary, knocking on thousands of doors across the city, have told me almost no one mentions equalization (or the other questions, for that matter).

Perhaps, in the election's closing days, we'll see a surge of attention to equalization. But it's unlikely.

So it looks like the issue pathway won't have much influence.

As for the mobilization pathway, research from the United States suggests the presence of a referendum can increase turnout by a few percentage points — enough to affect the outcome of only the very closest races — but the additional turnout tends to come from citizens who are most likely to vote regardless of the presence of a referendum

This would seem to leave little room for the Candidate Y scenario in Calgary; we have little reason to expect that the various questions will draw in big numbers of people who wouldn't otherwise vote.

Even so, data from the CBC-CMES survey, which we conducted in July and August, suggest that some ingredients are in place to make a mobilization pathway plausible in Calgary.

They are just unlikely.

What voters support

Here's the breakdown of decided voters (1,144 of the survey respondents) and their vote intentions on the equalization referendum. And remember, this is about voters, not an indication of candidate positions. To keep things simple, I focus here on the three candidates with the most support at the time of our survey in July.

Mayoral vote intention by equalization referendum intention

Among decided mayoral voters who plan to vote "yes" on the equalization referendum: 49 per cent support Jeromy Farkas, 14 per cent support Jeff Davison and 10 per cent support Jyoti Gondek.

Among "no" voters on equalization referendum: 43 per cent support Gondek, 16 per cent Farkas and 12 per cent Davison.

In other words, if the issue of equalization was of such motivation one way or the other that it drove a lot of people on one side to vote, and a lot of people on the other side chose to stay home, this would influence the mayoral vote.

So citizens' attitudes on the equalization question are related to their mayoral preferences.

But that's just Step 1 on the mobilization pathway.

The referendum must also mobilize voters who would otherwise stay home — and do so disproportionately, with those on just one side of the debate turning out in droves.

That's unlikely to happen here.

The data suggest that only about one per cent of Calgarians who intend to vote in the equalization referendum consider themselves unlikely to vote in the mayoral election. In other words, the pool of Calgarians who will show up to vote specifically for the equalization referendum and choose to cast a mayoral ballot while they're at the polling station is very small.

And while "yes" supporters are more motivated to turn out in the equalization referendum than the "no" side, the difference is small and has no apparent effect on motivation to vote in the mayoral election.

If you're curious, the same is true of the water fluoridation plebiscite question: those on the "yes" side are a little more motivated to vote than those on the "no" side but with no apparent effect on mayoral turnout intention. (We didn't collect data on the other Oct. 18 ballot questions: daylight saving time and the Senate.)

If we were to see a razor's-edge mayoral or council outcome — the sort of race that comes down to just 50 or 60 votes — we may want to consider the possibility that referendum-related mobilization was a factor in the outcome.

It's not impossible, but our data suggest it's very unlikely.

Questions and consequences

It's fair to worry about the influence of high-profile referendums on municipal elections, especially when they're imposed on municipalities by the province.

But an unlikely chain of events has to happen for these referendums to affect the outcome of our elections.

After the election is over, the Canadian Municipal Election Study team will have data that allows us to probe how important the equalization issue turned out to be. We'll report back with our findings.

In the meantime, we have a municipal election with a competitive mayoral race and an unprecedented number of open council seats — an election whose outcome is very much in question and whose consequences for the policy direction of our city will be profound.

Polling details and methodology:

The poll was conducted by Forum Research on behalf of the Canadian Municipal Election Study with the results based on a telephone recruit-to-web survey of 2,209 randomly selected eligible voters in the city of Calgary. The poll was conducted between July 6 and Aug. 4, 2021.

For comparison purposes, the margin of error for a probability sample of the same size would be plus or minus 2.1 percentage points, 19 times out of 20. Results at a ward-level and other subsamples have a larger margin. For more methodology information, see here. See here for more details on the data, methods and sources in this analysis.