Thoughts on the 2021 Canadian election

What is at stake for the climate

We’ve been lucky to have many different assessments of the party climate plans in the runup to this election. One feature that received a lot of discussion was the different climate targets, which makes sense seeing as how these are hard numbers that we can easily use to judge a plan. The Conservative Party of Canada is aiming to cut emissions by 30% in the year 2030, the Liberals by 40%, the NDP by 50% and the Green Party by 60%. While it is clear that 30% would not be as good for the climate as 40%, it’s hard to know just how meaningful that difference is. So I went and did a little analysis to shed some light on the gap.

In the figure above I chart projected Canadian emissions over the next decade under the Liberal and the Conservative plan. Because warming is driven by cumulative CO2 emissions, what matter for the climate is not just how much we’ve reduced emissions in the year 2030, but also the pathway getting there. In fact, assuming linear emission decreases from both plans, the Liberal plan would result in 365MtCO2e less entering the atmosphere than the Conservative plan (about the same as not running 130 coal plants for a full year). So for voters making the decision, this is pretty consequential (about 13 tCO2e per eligible voter).

How can we judge which parties are serious about their climate policy

Another question that I thought was missing from a lot of analysis was how we evaluate whether a party is serious about implementing their climate policy. Time and political capital are finite resources and if new issues arise (recessions, wars, pandemics) then governments may not implement policies which they do not view as priorities. So which parties view climate as a priority?

I can think of a few ways of answering this. We could look at where the parties receive campaign donations and how they voted on key climate legislation. I think both of these are promising strategies but, given the short timeframe of this election, didn’t have time to dig into them as much as I’d like. Instead I took a couple other metrics that were easy to access: how much party leaders talk about climate change (as judged by posts on Twitter) and how much their voters care about climate change.

I think some valid criticisms can be leveled at either approach. Party leaders are going to talk more about the issues that they “own”, or those that the public trusts them the most on. So talking a lot about climate change doesn’t necessarily mean that you are prioritizing it – you could be weaponizing it as a divisive issue, or just “playing to the crowd”. According to polls, the Liberal Party “owns” the issue of climate change. But I do think that if we see leaders who are not trusted speaking out frequently about climate change this might indicate a personal preference for the subject that could indicate a willingness to prioritize the issue at a later point.

Similarly, although a party cannot control the preferences of their voters, if the people who vote for them feel strongly about climate change that would give leaders motivation to enact policies even if doing so wasn’t politically convenient. Beyond that I’ll let the graphs speak for themselves.

For some additional, more in-depth analyses of the various climate platforms, try these sources:


To calculate the cumulative difference in emissions between the Liberal and Conservative climate plans I used projections from Canada’s Greenhouse Gas and Air Pollutant Emissions Projections 2020. The projections for 2020 include anticipated reductions from the COVID-19 Pandemic. To estimate reductions for 2021 I applied the downward linear trend needed to hit the Liberal target of 40% emissions of 2005 levels by 2030. For the Conservatives I kept the same 2021 value since the fall election of a Conservative government could not change emissions for the year. But for the CPC plan in years 2022-2030 I applied a linear downward trend such that emissions reached only a 30% reduction of 2005 levels in 2030. Note that For the visualization, 2018 and 2019 values were taken from the updated 2020 NIR but are only present to provide context (see Table below).

For the Twitter data: data was scraped using the Rtweet package on September 1, 2021. The most recent 3200 tweets from each party leader were included. Climate relevant tweets were identified by searching for keywords: “climat” ; “global warm” ; “carbon” ; “emission” ; “fossil” ; “greenhouse gas” and “effet de serre”. Word clouds were formed after removing punctuation and English and French stopwords (e.g. “we”). To calculate rate of climate-relevant tweeting I divided the number of climate relevant tweets for each party leader by the number of tweets in their sample. 

To determine the fraction of climate voters per party I used data from the 2019 Canadian Election Study. This is a high-quality, large-scale survey of voters before and after the election. I filtered out only high quality responses and those respondents who answered all three of the questions of interest. To identify climate voters I looked at those who selected “Yes” to two questions: the first on a belief that climate change is happening and the second that it is caused by human and those who selected either climate change or global warming as the most important election issue.


To make a comparison to coal power plants, I took the Canadian coal power plant with median levels of power generation in 2014 (Belledune: 2560 GWh) and estimated emissions for that plant if it had average emissions intensity for Canadian coal power plants (1070 kgCO2e/MWh). This worked out to 2.7MtCO2e in a year.

To make a comparison of emissions per eligible voter I used data from Elections Canada. Note that if this was a retrospective analysis we could perform more complicated emissions responsibility calculations as we described in a 2021 paper in One Earth (see below).


Leave a Reply