A quick first look at the 2017 Calgary municipal election results

This is a quick look at the unofficial results from the 2017 Calgary municipal election. With record-breaking advance vote number (almost 75,000) and a level of turnout unseen for decades (58%), I thought it might be interesting to see how the vote broke down both geographically and between early and election day voting.

Since elections are decided by who gets the most votes, I look at the results using raw counts. This is a bit different than how the media and most observers look at elections, especially in the age of horse-race journalism. However, it is important to remember that elections are mobilization contests. From the perspective of a campaign team, the campaign period is 27 days of identifying your supporters and then getting out the vote (GOTV) on election day. There’s some GOTV during advance voting, but by and large, most voting happens on the election day.

Figure 1 shows the votes that Naheed Nenshi and Bill Smith received in each ward plus across all “city-wide” polling stations, the vast majority of which are the advance polls. There are three themes in this graph I discuss: the consistent strength of the Nenshi vote, the inconsistency of the Smith vote, and the importance of early turnout.

Figure 1 – Nenshi versus Smith Vote by Ward

First: Nenshi was consistently strong across most of the city. He got more votes than Smith in almost every ward, except for Wards 13 and 14 in the deep south. Those areas of the city are heavily conservative in both provincial and federal politics, so it’s not a surprise Smith “won” them, but he only won them by gaps of 595 votes in Ward 13 and 1,992 votes in Ward 4–both below the mean gap of 2,247 votes (SD=1,848). Looking at raw votes instead of percentages also shows that, while Nenshi “lost” Wards 13 and 14, he still got more votes there than wards he “won,” such as Wards 3, 5, 9, and 10–all of which are in the northeast or east-central and typically have lower voter turnout.

Nenshi had big leads over Smith in Ward 7 and 8, which is unsurprising, given these wards are in the inner city and are made up largely of young, liberal urbanites. But, the gap only tells part of the story. In terms of raw votes, Nenshi’s vote totals in those wards are only average, and he actually picked up more votes in Ward 1, Ward 4, and Ward 11–all of which are suburban wards.

Second: Smith failed to stockpile votes in the outer suburbs to compensate for his predictable weakness in the inner city. It’s no surprise Smith did best in the outer suburbs in Wards 1, 4, 6, 11, and 14. Even then, there are plenty of suburban wards where he should have been able to stockpile huge amounts of votes, such as Ward 12 (which shares much of the same territory as Jason Kenney’s old federal riding) and Ward 2 (which shares much of the same territory as Jim Prentice’s old provincial riding). He beat Nenshi in Ward 13, but only barely. If the suburban upper-middle-class was as angry about property taxes that we were led to believe, they would have showed up to the polls to “throw the rascal out.” A true tax revolt would have seen huge numbers of votes for Smith in Ward 12, 13, 14, which are some of the more affluent and most populous wards in the city.

The inner city and the central-to-north east are Smith’s areas of weakness, which was to be expected. While even Nenshi also had low vote numbers on the east, Smith did far worse, and Smith wasn’t competitive in the inner city. Still, it’s more than possible Smith could have “stockpiled” enough votes in the suburbs to overcome his weakness in the inner city, but that didn’t happen.

Third: Nenshi handily out-performed Smith in the advance vote (which comprises the vast majority of the city-wide voting stations). Some commentators speculated that high advance turnout was another indicator for a tax revolt and Nenshi’s defeat. Others countered that high turnout is more indicative of increased participation of groups who typically participate in lower-than-average rates (i.e. youths, low socio-economic status, ethnic minorities, etc.). Nenshi “banked” 3,477 more advance votes than Smith, which comprised 11.7 per cent of his total gap of 29,755 votes over Smith. Another way to look at those 3,477 advance votes is they more than make up for Nenshi losing Wards 13 and 14.


A quick look at the breakdown of the Nenshi and Smith vote by ward and in the advance stations show the strength and consistency of Nenshi’s support. It also shows that, while Smith had pockets of strong support, he under-performed in suburban areas that are traditional conservative strongholds. If I had to speculate, I’d say that a strong showing in the outer suburbs was Smith’s only realistic path to victory, given Nenshi’s decent-to-strong showing in the city core and the inner/older suburbs.

Nenshi also handily out-performed Smith in the advance vote, which repudiates the theory that the high advance vote was a harbinger of Nenshi’s doom. It’s possible that the brouhaha over horse-race polls served more to motivate the Purple Army more than it did to motivate opposition to Nenshi, but that is just conjecture.

We cannot infer individual motives based on aggregate results, so we don’t know why the suburbanites did not show up for Smith. Elections Calgary has not yet released ward-by-ward turnout rates, but the low vote totals in Wards 12, 13, and 14 might indicate many of those true-blue conservatives in the deep south simply stayed home rather than fight long lines only to have to chose between three candidates they didn’t like. This is not to take away from the Nenshi campaign’s achievement of growing their total vote in an election where some expected them to lose, but it is possible Nenshi’s competitiveness against Smith in the populous deep south has as much (if not more) to do with Smith’s inability to mobilize suburban voters than it does with Nenshi’s actual strength in those areas.