Ladies and gentlemen, welcome to Quebec, and thank you very much for your invitation.
This commission has been given the mandate of recommending a voting system that both incorporates proportional representation and maintains local representation in a non-complex manner, yet all the many systems that have been proposed to this commission involve unacceptable trade-offs between these two criteria.
Let's take two examples. In the party list methodology, one votes for the party, not the candidate. While achieving proportional representation, many ridings are assigned candidates from a different riding or a different party instead of the candidate the majority wanted. That's an unacceptably high cost for proportional representation.
The second popular contender is the MMP methodology, in which one has two classes of representatives, constituent and list. The list members provide the necessary seats to approximate proportional representation. But the cost here, according to Professor Massicotte, whom you know, is either to reduce the number of constituent members from 338 to somewhere between 160 and 200, or to increase the number of seats to somewhere between 500 and 675. The cost of achieving PR in this context is either a much lower level of representation, or a much larger, more expensive bloated House of Commons. Neither is going to be a political blockbuster; in fact, there would be quite the converse.
Canadians simply don't like these alternative systems. B.C. rejected STV by 61% in 2009. Ontario rejected party list by 63% in 2007. P.E.I. rejected MMP by 63% in 2005. They just don't like it.
So where do we go from here? Winter is coming. December 1 is getting close.
There is a huge number of these methodologies, but ultimately, they are all ad hoc and they all have unacceptable costs, and they all fall into the same trap. The concept of requiring that the number of seats held by a party matches the percentage of votes that is received by that party is not the appropriate metric. The relevant metric is that the voting power of a party in the House matches the popular vote that party received. If a party receives 40% of the vote, then its voting power in the House, its clout, has to be 40%.
This can be easily achieved by simply mandating a weighted vote. If the Conservatives have 50% of the seats and 40% of the vote, then each member of the Conservative Party is given a weight of 0.8 when voting. The vote in each party is simply the percentage of the party vote divided by the percentage of the party seats: 40 over 50 is 0.8.
If the Conservative Party votes as a bloc, it has 40% of the vote, 0.8 times 50%, which exactly matches the 40% it got in a popular vote. If the Green Party receives 5% of the seats and 10% of the votes, then it would have a weight of 2, for each member of the Green Party. This I call fractional representation. It's first past the post with weighted voting in the House: instant, painless, proportional representation.
As for precedents, in Quebec we have 11 agglomeration councils, regional councils, in which each municipality has a weight proportional to the population of the municipality.
The Council of the European Union similarly uses weights, where the weight is proportional to the size of each member state. The IMF and the World Bank—they're significant, aren't they?—have weights, and those weights are proportional to each member's contribution. Of course, in every public company, when you vote in a proxy vote, you have a weight that's proportional to the number of shares you hold.
Weighted voting is something that we are all familiar with, and there's nothing unusual about it. How you assign the weights is simply a matter of what your criteria is. Our criteria here is proportional representation.
The advantages are clear. It is well understood, since there is no change required from the existing electoral system. It's familiar, since proxy voting occurs for every company. It provides a one-to-one relationship between a riding and its elected representatives. It provides exact proportional representation at the party level. It does not require any new administrative structure. It can be implemented immediately and it is far and away the most cost-effective method of achieving proportional representation.
Let me go through a couple of details and then we can take it from there.
Who forms the government? Government is formed in any system by that party which has the most clout. Under our current one, it's the party with the most seats. Under this one, its the party with the highest proportion of the popular vote. Minority governments would form coalitions, as they do today.
Is change complicated? No. At a general election, the weights are ascertained in the way I've described for each party. Those weights are then assigned to each riding and are fixed until the next general election. If there's a by-election, the weights stay the same for that riding. If a member decides to cross the aisle or become an independent, again, the weight is fixed and doesn't provide any incentive for such a move. The weights change once every four years.
Regarding free votes, the government can declare that it will treat a particular item, excluding the budget, as a free vote, and defeat does not amount to a vote of non-confidence. If all parties permit a free vote, then it's no longer a party vote; it's a unitary vote, and that would be appropriate. You can't have some parties having unitary votes and some parties having weighted votes. That's illogical. A free vote with unitary voting would only be permitted if all the chief whips unanimously agreed.
Constitutionality is a good question. Democratic rights are covered by sections 3, 4, and 5 of the charter. Section 3 gives every citizen the right to vote and doesn't change under this system. Subsection 4(2), which permits the continuation of the House of Commons beyond five years, would require a unitary vote. There's no constitutional right to have a vote counted in a certain way, so there's no provision that will make fractional representation unconstitutional.
Every PR system has a threshold. Typically, a threshold would be a minimum percentage of votes, say 5%. If a party receives more than 5%, more than the threshold, and does not have a seat, then a compensatory seat is provided. However, irrespective of the threshold, a party gains representation if it gathers at least one seat.
Let me summarize.
The concept of requiring that the number of seats held by a party match the percentage vote received by that party is not the relevant metric. What is important in proportional representation is that the voting power of a party matches the vote received. A fractional representation system does exactly this. In a simple and direct manner, it maintains the current, well understood system with local representation, without any additional complexity, and at no extra cost. Since it maintains the current electoral system, it does not create dissent, neither from the public who generally don't like change, nor from MPs fearful of losing their seats under the new system. It's politically acceptable.
Ultimately, the recommendations of this committee, if they're to have any impact at all, must be such that consensus is assured.
Each party is going to look to protect its own partisan interests, so anything radical will be rejected out of hand, or at least require a referendum which, based on past history, will also be rejected. By making a small incremental change through a weighted vote, fractional representation presents the best chance of ensuring the consensus that is necessary to make any change to Canada's federal election process, certainly within a realistic time frame.
May I propose to you that fractional representation which is an incisive, intuitive, and innovative solution that satisfies all five principles of your mandate be seriously considered as a viable alternative federal voting system for Canada.
Thank you very much.