In Fair Voting BC’s submission to the BC Special Committee on Democratic and Electoral Reform, we put forward a recommendation for a more inclusive and proportional voting system carefully tailored for use in BC. This “Regional Representation” model was developed in consultation with Fair Vote Canada and a number of voting system experts.

Key Democratic Principles Underlying Regional Representation Model:

Regional Representation aims to satisfy a number of key democratic principles which our current First-Past-the-Post system significantly violates:

• Make Voters Equal: To the greatest extent possible, each citizen’s vote should have equal effect in electing an MLA who shares their point of view and the number of ‘ineffective’ votes should be minimized (votes that do not affect the make-up of the legislature).
• Empower Voters: Each voter should be able to vote for a specific candidate, and should be able to choose between candidates of the same party; MLAs should principally be accountable to the voters who elect them.
• Ensure Majority Rule: The system should not give a government majority power without the support of a majority of voters.
• Ensure Geographic Representation in Line with Regional Voter Preferences: Each region of the province should be represented by MLAs who live in that region, and each region should elect MLAs reflecting the diversity of political views in the riding.
• Promote Sincere Voting: Voters should not feel torn between voting against who they don’t want to win and voting for who they do want to win.
• Respect Minority Views: Our voting system should better support and foster political pluralism and competition – voters with minority political views and emerging political parties should not be discriminated against, and independents should have a fair chance of being elected.

What Regional Representation Avoids:

At the same time, it seeks to avoid any elements that are not compatible with BC’s political culture or would pose significant obstacles to implementation. In particular, it avoids:

• “Closed” Party Lists – British Columbians have expressed opposition to voting systems where parties can place candidates in favoured positions on a list. 
• Many Small Parties – Having too many small political parties can make it harder to form stable governing arrangements. The approach we recommend below avoids this issue.
• Non-Local Representation – British Columbians are likewise concerned about losing representation in the more sparsely populated areas of the province, and would likely oppose any approach in which top-up MLAs come from outside their own region.
• Any Significant Increase in the Number of MLAs – Adding new MLAs to represent currently under-represented voters would be helpful, but the number of MLAs added should be modest enough in number to garner public support. Regional Representation could work with no added MLAs.

The Concept:

Regional Representation is a hybrid of the two main types of proportional voting: the Single Transferable Vote (STV) and Mixed Member Proportional (MMP). It has three key features:

Multi-Member Districts: Like STV, Regional Representation uses multi-member districts (MMDs), merging 2 or 3 current ridings and electing 2 or 3 MLAs from each merged district to ensure that the views of at least the two main groups of voters in each district are reflected in the MLAs elected from each district.

Regional Top-Up Seats: Like MMP, Regional Representation defines regions (typically covering 10-20 current ridings) and uses top-up seats (one for each 5-6 seats in a region) both to (1) balance out any discrepancies between vote share and seat share that might arise in the MMDs and (2) make it possible for smaller and emerging parties to win representation in a region if they secure on the order of 5-10% of the popular vote in a region.

Simple “Mark an X” Ballot: Voters will find it easy to vote with Regional Representation. Just as they do now, they will mark an “X” beside the name of the candidate on the ballot that they most prefer (the larger parties would likely list 2 or 3 candidates, as shown below). These votes will determine how many seats each party or independent candidate wins in the multimember districts, as well as which parties or independents are most under-represented at the regional level.

Example: How Regional Representation Might Have Worked in 2024

Organization by Regions: In the example below, we illustrate how Regional Representation might have played out in the 2024 BC election. The figure below illustrates a possible organization of the province into seven regions, each electing between 10 and 17 MLAs (we recommend that the actual organization be determined by the Electoral Boundaries Commission):

• Northern Interior (10 seats)
• Souther Interior (15 seats)
• Vancouver Island (15 seats)
• Fraser Valley (10 seats)
• South of the Fraser Suburbs (12 seats)
• North of the Fraser Suburbs (14 seats)
• Vancouver-North Shore (17 seats)

Multi-Member Districts (MMDs) Within Regions: Within each of these regions, we would merge 2 or 3 adjacent ridings to form MMDs (with an occasional riding left as a single-member riding, where desired or appropriate):

Outcome of 2024 BC Election: The figure below illustrates the outcome of the actual 2024 BC election. Orange dots indicate ridings won by the NDP, blue by the Conservative Party, and green by the Green Party. Note the significant regional concentrations: the NDP won almost all the seats on Vancouver Island, as well as in Vancouver and the near suburbs, while the Conservative Party won almost all the seats in the Interior and Fraser Valley.

Predicted Outcome in Multi-Member Districts: The next figure shows the likely results in the MMDs if we used Regional Representation. Note that the regional monopolies are largely broken up – voters would have been able to elect both NDP and Conservative MLAs in almost every MMD across the province.

Predicted Top-Up Winners: The final figure shows who would likely have won the top-up seats if we used Regional Representation. These seats are determined by looking at the results in the MMDs and then, based on the regional vote, identifying which parties are most under-represented. There are two top-up seats in the smaller regions and three in the larger regions. Note that the teal colour refers to independent candidates who were previously BC United MLAs.

Overall Results

Overall, Regional Representation delivers results that very closely align with how voters voted:

• At a provincial level, the NDP received 45% of the vote, the Conservatives 43%, the Greens 8%, and the former BC United MLAs 2%. Under FPTP, the NDP won 47 seats, the CP 44, and the GP 2 (51%, 47%, and 2% of the seats, respectively), so the two largest parties are over-represented (with the NDP winning a majority of seats with less than a majority of the popular vote), and supporters of the GP and UP are significantly under-represented.
  
• Under Regional Representation, the NDP would likely have won about 52 seats (including top-up seats), the CP 50, the GP 7, and former BC United MLAs 2 (47%, 45%, 6%, and 2%, respectively). These results would have been much closer to each party’s vote share than with FPTP, and the NDP would likely not have won a majority of the seats unless they had increased their vote share by a few percentage points.

Comparing Systems

The chart below summarizes how Regional Representation and various other alternative voting systems compare with our current FPTP system on the key democratic principles outlined above (see justifications in next section).

Overall, our current voting system (FPTP) does poorly on all key democratic principles, including on geographic representation (usually considered one of FPTP’s strengths) because it does not represent multiple political views within each region. MMP Lite (just adding a few top-up seats to the existing ridings) would represent only a minor improvement over our current system.

STV and MMP both perform significantly better than MMP Lite, but Regional Representation builds on the strengths of both, more effectively addresses all seven core democratic principles, and has the virtue of being extremely simple for voters to use.

We therefore recommend that BC proceed to implement a Regional Representation model or an alternative system that aligns well with the democratic principles outlined above. 

Justification for Ratings

In developing the ratings presented above, we assumed that the STV model would be similar to what the Citizens’ Assembly recommended (MMDs averaging approximately 4-5 in size), the MMP model would be a regional model in which the regions average approximately 15 seats, with 9 single member districts and 6 top-up seats (60:40 mix), the MMP Lite model adds 20% MLAs, and the Regional Representation model uses 2-3 member districts in regions averaging ~15 seats, with approximately 15% top-up seats (~1 for every 5-6 current seats).

The details of our justifications for these ratings are presented below:

Equality

FPTP does worst on this measure, as typically more than half the voters do not end up with an MLA for whom they’ve voted. MMP Lite partially closes the gap. MMP more fully balances the outcome at a regional level, though half the voters would not be represented by an MLA at the district level, but only at the regional level. Regional Representation delivers comparable overall equality in the effect of one’s vote to MMP, but with a higher proportion of MLAs elected directly at the district level (~85% with Regional Representation vs 60% with MMP).

Empowerment

FPTP again does worst on this measure as voters cannot choose between different candidates from the same party, and therefore cannot switch their vote to another candidate with similar political views if they are displeased with their current MLA. MMP Lite performs similarly to FPTP, but does slightly better as voters can potentially have some influence on the overall results if they switch their vote to another party. MMP does better than MMP Lite because voters have more influence on the overall seat outcome. STV does best on this measure because voters can specifically choose between more candidates from the same party. Regional Representation would perform similarly to STV, with additional influence over the overall seat outcome, if the MMD district sizes were similar; however, with district sizes of 2-3 (compared with an average of 4-5 with STV), voters might only be able to choose between two candidates for the larger parties.

Majority Rule

FPTP does worst on this measure because of the significant over-representation of the leading party, frequently delivering “false majorities” on as little as 35-40% of the vote. MMP Lite would help slightly here, and STV would do even better as the overall correspondence between vote share and seat share for the largest parties would be quite close. Both MMP and Regional Representation would deliver the best performance because they compensate at the regional level for discrepancies between vote and seat share.

Geographic Representation

FPTP only performs moderately on this measure, despite it often being claimed that geographic representation is its strength, because it only elects one MLA per riding and so does not reflect the range of political views held in each region of the province. This is exacerbated by its tendency to “sweep” all or almost all seats in a region, locking out voters with opposing political views. MMP Lite would only slightly mitigate these effects. MMP would also deliver similar sweeps in the single-member districts, but would elect opposing voices at a regional level. STV would do very well in electing MLAs with a range of political views within each local region, while Regional Representation would perform similarly (at least for the two or three main local political perspectives) and would better ensure that the overall diversity of voices in the region is accurately represented in the legislature.

Honest Voting

FPTP is well-known for incentivizing “strategic” (insincere) voting, in which a voter feels significant pressure not to vote for their most-preferred candidate, but for someone they like less in hopes of defeating the candidate they like least and fear will be elected. MMP Lite only slightly reduces this dynamic. MMP does not change this dynamic in the single member seats, but can give voters a reason to vote sincerely for the top-up seats. Both STV and Regional Representation strongly encourage voters to vote sincerely. 

Diversity

FPTP performs worst because parties can only run one candidate per riding, so they tend to be disproportionately drawn from the most privileged demographic groups (e.g., male, white, and older). MMP Lite only modestly changes this dynamic. MMP perpetuates this dynamic in the SMDs, but can provide some openness at the top-up level, though this tends to most affect the smaller parties in each region rather than the dominant party. Both STV and Regional Representation incentivize parties to run more than one candidate, but typically not more than 2-3; there would be potential to increase diversity further if more seats were contested in the MMDs of the two systems, which is why we rank them both as ‘very good’ rather than ‘excellent’.

Minority Views

FPTP again performs worst on this criterion because it typically significantly suppresses the seats won by all but the top two parties. MMP Lite would do only marginally better, as many of the top-up seats would likely be won by the second-place party in each region rather than smaller parties. STV lowers the threshold for entry for new parties to about 15% if the average district magnitude is about 5 and an independent candidate can win a seat if they secure this same fraction of the vote in the district. MMP and Regional Representation do best on this criterion because the threshold for entry is reduced further to about 5% for a region with ~15 seats in it, and independents could compete for these seats if they are listed on a regional ballot under MMP or on ballots in different districts under Regional Representation.