by Michael Heeney and Andrew Yan
With the looming federal elections on October 19, 2015 , BTAworks wanted to look at the population densities of ridings and their voting patterns. With data from Elections Canada and Statistics Canada, we examined the population densities of every riding in Canada and the party that they voted for and see if there might be any patterns between the riding density and party preference. While we focused on the 2011 elections, we also studied the patterns of the 2006 and 2008 elections. Given the vastness and geographic diversity of Canada, these riding population densities ranged from the riding of Nunavut at 0.01 people per hectacre at its lowest to 112 people per hectacre in the riding of Papineau in Montreal at its densest. To visualize what a hectacre is, it is about the size of the field within a 400 metre running track. It’s important to note that as a result of the 2012 federal electoral redistribution, the number of electoral districts was increased to 338 from 308 for these previous elections, with additional seats based on population assigned to Alberta (6), British Columbia (6), Ontario (15), and Quebec (3). Nevertheless, the question would be how riding population density might be an indicator or even motivator of party choice.
The above chart is what happens when you take every seat won by party organized by riding population density from the 2006, 2008, and 2011 federal elections. The Conservative seats were all from ridings of less than 50 people per hectacre, the Liberals were between 0.0001 to 112 people per hectacre, the New Democrat Party were between 0.0001 to 101 people per hectacre, the Bloc seats were between 0.003 to 113 people per hectacre, the Greens at 2.3 people per hectacre, and Independents held ridings that were between 0.09 to 0.13 people per hectacre. The slide below illustrates the average and the median population density per hectacre by party. Not surprisingly, there are very different patterns depending on what party one is examining over three elections.
As perhaps, a possible predictor of who might or might not win and where, here is the party riding by riding population densities for 2011. It is perhaps the subject of further, much deeper study to look at the specific reasons why these patterns occur.
Some final notes:
A special thanks to Mark Heeney, the BTAworks Summer Intern extraordinaire for helping processing the elections and demographic data. And finally regardless of what riding and riding population density, please remember to vote on October 19, 2015 and click here for more information.