Who wins Election 2019 below a ranked-ballot system

Shortly after the “Proportional Representation Thought Experiment” I revealed close to the tip of the federal marketing campaign, a number of readers requested me to carry out other forms of simulations to check numerous voting techniques. Although electoral reform will probably not be on the radar of this 43rd legislature, I’m all the time keen to play slightly politics-fiction—particularly if knowledge of high quality is offered.

How a few single transferable vote system (STV), or ranked-ballot? Beneath this technique, voters don’t put an X or a verify mark subsequent to their most well-liked candidate, however relatively ranks them with numbers “1”, “2”, “three”.


This electoral system was the one most well-liked by the governing Liberals when the electoral reform debate raged on two years in the past. Those that adopted these consultations intently will do not forget that the NDP and Greens wished proportional illustration, the Liberals insisted on a ranked-ballot and the Conservatives wished no reform in any respect. And not using a consensus, the Liberals deserted electoral reform altogether, forsaking a number of disenchanted (and indignant) voters.

Close to the tip of the final marketing campaign, Léger (among the most accurate pollster on this election) revealed a ballot that measured voters’ second selections—and people had been tabled each in accordance with area and celebration preferences. Yow will discover the Léger/Canadian Press poll here. Scroll to pages eight and 9 for the total breakdown of voters’ second selections.


We are going to use this valuable knowledge, together with the precise election outcomes to simulate the outcomes of a ranked-ballot election. This simulation went in accordance with the next guidelines:

  1. We take the outcomes from the federal election in all 338 electoral districts. Candidates from events that received over 50 per cent of the vote in a single driving are mechanically elected. A complete of 123 candidates reached that threshold on Oct. 21: 57 Conservatives, 54 Liberals, 11 Bloquistes, and one New Democrat (and no Greens);
  2. If no candidate reaches 50 per cent, the last-place candidate is eradicated and votes that had been forged for him/her are redistributed in accordance with these ballots’ second selections. We repeat this cycle till one candidate reaches 50 per cent.
  3. Clearly, now we have to work with the info now we have, not the info we want we had—that means that now we have no info on voters’ third and fourth selections. Subsequently, this simulation solely makes use of voters’ second selections in accordance with the aforementioned Léger ballot. It’s an imperfect system, however I wouldn’t wish to simply make numbers up. Furthermore, since now we have no particular info on second selections of voters who voted for smaller events or unbiased candidates, these had been merely ignored on this simulation.

As soon as once more, that is an train of politics-fiction: It’s only a strategy to image what issues may appear to be. Because it was talked about on this hypothetical Proportional Representation column from last month, they’re mere “thought experiments”, or Gedankenexperiment.

Think about the next graphs. On the left are the precise election seat outcomes, and on the suitable the seat averages in accordance with the ranked-ballot guidelines said above:


At first sight, it’s placing how related these outcomes really are. The Liberals, which received 157 first-past-the-post (FPTP) seats final month, win a mean of 166 seats when second selections are thought-about. When digging slightly deeper into the outcomes, we discover the LPC additionally completed second in 115 districts, probably the most of any celebration (the CPC got here in second in 114 districts). The Léger ballot exhibits the Liberals had been the second selections of 13 per cent of respondents, and it is sufficient to maintain them in comfortably in first place, simply in need of a majority.

The present model of the Conservatives would lose probably the most below a ranked poll system. In keeping with this simulation, they might go from 121 seats in FPTP to a mean of 107 seats.


This shouldn’t be shocking contemplating the CPC was the second selection of solely 9 per cent of respondents (together with practically half of PPC voters). A number of observers famous that ought to the CPC have received probably the most seats on election evening, it will have been onerous pressed to search out companions to help their insurance policies. (For example, below a CPC minority situation, which different celebration would have supported scrapping the carbon tax?) These numbers inform us that though the Conservatives did win the favored vote, the anti-Conservative vote was considerably stronger than the anti-Liberal vote.

It shouldn’t be shocking that the NDP positive factors probably the most from this ranked poll simulation. In keeping with the Léger ballot, 25 per cent of respondents had Jagmeet Singh’s celebration as their second selections, by far probably the most of any celebration—together with 21 per cent in Quebec and 29 per cent in Ontario.

Nonetheless, we must always state once more that these outcomes are extremely related as these of the election. Would a ranked poll actually produce a Parliament extra consultant to Canadian voters’ needs?


Let’s have a look at a number of the regional outcomes.

Listed here are the leads to Atlantic Canada. On the left, precise election outcomes; on the suitable, the simulation outcomes:


Even by contemplating second place votes, the simulation leads to Atlantic Canada are virtually similar to the precise election outcomes.

In Quebec, the NDP would have fared slightly higher than it did, because it has probably the most second place votes of any celebration. However nonetheless, the outcomes are extremely just like the Oct. 21 seat depend:


In Ontario, the Conservatives would drop a mean of eight seats to the good thing about the NDP and the Liberals:

Within the Prairies and in Alberta, the simulation outcomes are similar to the election outcomes:


The case of Alberta is fairly straight ahead: of the 33 seats the CPC received within the province, 32 had been received with greater than 50 per cent of the favored vote, that means that second selections weren’t thought-about in any respect in these ridings.


Lastly, in British Columbia, each the Liberals and NDP would take a modest chunk out of the CPC seat totals, however general the outcomes are extremely related:

Naturally, we might now need to query voters on the next: what number of Liberal voters on this election would have put the NDP or the Greens forward of the Liberals in a ranked poll system? In keeping with the ballot common, the NDP and Greens barely underperformed on election day—may or not it’s fraction of these voters determined to vote strategically for the Liberals hoping to dam the Conservatives? That is very believable, but additionally onerous to quantify with at the moment accessible knowledge. We may think about each the Greens and NDP profitable a handful extra seats apiece on this situation.


Nonetheless, it seems to be as if the Liberals would nonetheless have received probably the most seats below a ranked-ballot system, and the Conservatives would little question have completed worse.

It’s considerably curious that the celebration that received the favored vote additionally has the fewest second place help among the many most important nationwide events, however that is one more strategy to emphasize that the Conservatives, though they’ve probably the most stable and trustworthy base of voters within the nation, completely should broaden their tent with a purpose to win. Pleasing your base at the price of antagonizing others dramatically narrows the accessible paths to victory.