How your last name could cost you an election

If your last name is Andreychuk, you might want to consider running for municipal office. But if it's Zalapski, you might want to keep your hat out of the ring.

A new study in the March issue of the Canadian Journal of Political Science suggests that the order in which municipal candidates' names appear on ballots can have an impact on how many votes they receive — particularly when they don't have the party affiliations that help signal to voters where they stand on the issues.

The paper, co-authored by Charles Tessier of Laval University and Alexandre Blanchet of McGill University, examined the results of municipal and provincial elections held in Quebec between 2008 and 2014.

It found that a municipal candidate without a party affiliation whose name appeared lower on an alphabetically-arranged ballot did significantly worse than candidates nearer the top.

In dozens of cases, the effect may have been enough to spell the difference between victory and defeat.

In provincial and federal elections, political parties send strong signals to the electorate. Voters may not be familiar with their local candidates, but party affiliations make the choice simpler.

Parties receive widespread media coverage and have longstanding policy positions. So voters don't necessarily need to closely follow a federal or provincial election campaign to be informed enough to make up their minds.

Municipal elections are different. Candidates have no party affiliations in most Canadian municipal elections. There are a few exceptions to the rule — such as Montreal, Quebec City and Vancouver — but even in some of those jurisdictions, the 'parties' are little more than vehicles for mayoral candidates.

"Municipal elections are characterized by almost no party cues and much less visibility," Tessier and Blanchet wrote. "These elections ask for a little more effort from voters compared to provincial elections.

"In this more demanding setting, our results show signs of ballot order effect, indicating that this small increase in complexity has an important impact on voting behaviour."

The 'As' have it

Studies of this effect have been conducted in the United States, but the authors believe their work is the first of its kind in Canada.

Analyzing the fates of 16,000 municipal candidates in the 2009 and 2013 elections in Quebec — and taking into account the different political landscapes in Montreal, Quebec City and the rest of the province, where party affiliations at the municipal level were far weaker or non-existent — Tessier and Blanchet found significant differences between candidates at the top of the ballot and those further down.

It found no systematic differences between the first- and second-ranked candidates, but found that the third name on the ballot tended to be penalized. Where there were no party affiliations, the study found that the third-ranked candidate finished an average of five points below the first-ranked candidate, and four points below the second-ranked candidate.

Candidates in the fourth ballot slot and those lower down did not experience as much of a disadvantage as those in the third position, though these lower-placed candidates finished an average of 3.5 points behind the first candidate on the ballot.

This suggests that the middle spot in a list of names is the worst place for a candidate to be. The bottom is better than the middle, but still worse than the top.

In municipal races with party affiliations, Tessier and Blanchet found a less significant effect — but the third-ranked candidate was still at the greatest disadvantage.

Voters need cues

The authors also looked at the results of the three provincial elections held in Quebec between 2008 and 2014. They found ballot order had no impact at all on how the candidates did — which suggests that party affiliation offers voters a helpful shorthand guide to policy positions.

"The fact that the ballot order [effect] seems to be only present in elections without party cues indicates that [these cues] are indeed a crucial piece of information to voters," the authors wrote.

The results suggest that the ballot order effect could have real-world implications. Tessier and Blanchet identified 99 candidates ranked third on the ballot in the 2009 and 2013 municipal elections who lost by five percentage points or less.

In other words, the letter those candidates' surnames started with might have been the reason they failed to win — even though it's hardly a relevant factor for judging their suitability for office.

A healthy democratic process "assumes that people are voting for the candidates they prefer," the authors wrote, "which entails that they actually have meaningful preferences to begin with."

Random ballots wouldn't solve the problem

The findings might argue in favour of randomly placing candidates' names on individual ballots to avoid this effect, as pollsters do when conducting surveys. Researchers in other fields have long been aware of the "primacy effect" and conduct their studies accordingly.

But Blanchet says randomization wouldn't solve the core problem posed by voters lacking the information provided by party cues. The ballot order effect is merely a symptom of that bigger problem. Though it might prevent candidates at the top of the ballot from having an unfair advantage, randomizing names would simply hide the problem by spreading it equally (and randomly) across the ballots of low-information voters.

"Parties are often criticized for limiting debate," said Blanchet, "but voters need parties to help them understand the political universe."