Overview of Featured Simulations

This web site presents the results of simulating 18 electoral systems. Others have done individual simulations of their favourite systems, but this is believed to be the first which uses a consistent approach on so many. This page features the 18 systems I view as most interesting. For the full list of 18 simulations, please see the "All Systems" item in the "Overview" menu, above.

Each of these 18 systems is described at the bottom of this page.

There is a lot of data here. It's not for the faint of heart. Perhaps the best way to dive in is to start with the associated submission to the Parlimentary Electoral Reform Committee.

With respect to the Committee's first principle of Effectiveness and Legitimacy, the recommendations based on this modelling is that:

• the Committee issue a preliminary report stating that the Alternative Vote would be a step backward from FPTP and should not be considered further;
• the Committee strongly consider Rural-Urban PR, a highly proportional, made-in-Canada system that effectively deals with our diverse riding sizes;
• if choosing STV, the Committee think carefully about whether having smaller multi-member ridings is worth the decreased proportionality;
• if choosing MMP, the Committee should stipulate that FPTP (rather than AV) continues to be used in the local riding elections;
• the Committee avoid MMP-Lite’s substantial increase in complexity for very little gain in proportionality.

It should be noted, of course, that all simulations make assumptions and that any change to the voting system would change how people vote. So actual results will surely be somewhat different from any simulation. Nevertheless, these simulations show us important properties about the systems we are considering that can help guide our decision-making.

• FPTP in 2015 gave the Liberals an undeserved 15% over-representation of MPs, just like it did for the Conservatives in 2011. These graphs shows that as the spread in the vote becomes larger, the spread in MPs grows even faster. For every percent the Liberals can increase their vote, their share of MPs goes up by more than 3%.

• Many commentators have noted that AV favours centrist parties like the Liberals. These simulations illustrate just how true that is. Replaying the 2015 election with AV gives the Liberals 24% more MPs than deserved based on the popular vote.

• Some MPs might consider keeping the current 338 local ridings and adding 10% - 15% more MPs to give a measure of proportionality. These simulations and the table above show, however, that such a tepid response moves us towards proportionality but doesn't really get us there. With similar design parameters (338 MPs in local ridings; 15% more MPs) the Rural-Urban Proportional (RUP) model does far better.

• Simulating swings in voter preferences shows that most of the proportional systems remain proportional even as the gap between the first-place and second-place finishers increases. These can be seen in the Vote Swing Analysis graphs produced on the Summary page for each model as well as the "Average Gallagher Index" in the chart above.

  The Gallagher Index is a measure of disproportionality. Numbers below 5% are good. The Average Gallagher Index is the average taken across many simulations done for the voter swing analysis.

  Unfortunately, MMP with only a small top-up layer does not perform well on this measure.

• All of the parties are currently disadvantaged by FPTP. The Liberals are disadvantaged in the prairie provinces while the Conservatives and NDP are disadvantaged in the eastern provinces and particularly in the Maritimes. The Greens, of course, are disadvantaged nearly everywhere. By the same token, a proportional system would advantage each party in different areas of the country.

Election strategies are the specifics of how ballots are counted to determine which candidate fills a seat. Each strategy has three parts: how single-member ridings are handled, how multi-member ridings are handled, and finally how top-up or adjustment seats are handled.