SMMP compared to Law Commission Recommendations

In 2004, the Canadian Law Commission issued a report entitled “Voting Counts: Electoral Reform for Canada”.  This document is very well researched and written, and I would highly recommend it to anyone who wants to understand the issues involved in electoral reform.  The full report can be found on the Government of Canada website at

The report recommends, among other things, a mixed member proportional system for Canada where 2/3 of the seats are elected as they are now, and 1/3 of the seats are top-up seats from party lists based on regions of 5 to 7 ridings. There are additional recommendations in the report which are not linked to the voting system, such as eligible age for voting.

In my opinion, this recommendation is one of the strongest and most positive suggestions that have been widely considered.  I humbly believe that Simple MMP is a better than the system proposed by the Law Commision in four key areas, outlined below.

Single vote versus two votes

Simple MMP uses a single vote to tally both the local winners and to tally the provincial party support used to select the top-up seats.  The Law Commission (LC) recommended system uses a two vote ballot where you vote for one person locally and a separate vote for the party you want to support.

To me, there are very few situations where this kind of voting system would make someone happier than just casting a single ballot.  One example that was raised was the case where you like a certain party, but you don’t like the local candidate.  The LC system would allow you to vote for an individual you like while supporting your party for the top-up seats.

My feeling is that if the local candidate is unacceptable, then you need to get involved with your local party and fix that. I find it hard to imagine that I would dislike somebody so much that I would abandon the party whose values I support in favour of a candidate from a party whose values are not my first choice.

I suppose the opposite might be true as well. I might have a fantastic candidate that I think would do a good job, but they are not with the party that usually support.  Again, my feeling is that if I admire the candidate so much, then why wouldn’t I respect the party choice they have made.  If I truly admire this person and want to vote for them, could their party’s values be so different from my own?

I further think that having two votes might be confusing to older voters who have spent a lifetime casting a single ballot. Two votes is an unnecessary complication in what could be a very simple solution.

For these reasons, I believe a single vote ballot is superior to a two vote ballot.

Candidates with the most support versus party lists

To select top-up representatives, Simple MMP uses the candidate pool that remains after the local representatives are given their seats. To select a seat for Party X, we select the candidate in Party X who received the most votes (by percentage in their riding) to occupy the top-up seats.  The LC recommendation suggests that all top-up seats would come from lists provided by the parties.  it is not clear if people on this list could also run as candidates, but it is clear that people can be on the list without being listed on a ballot.

To me, I think that legitimacy in democracy comes from having a mandate from the voters. I believe this applies to individuals as well as parties. Canada’s system has historically been one person voting for another person. In fact, listing party affiliation on the ballot is a fairly recent change in our electoral history.

Simple MMP rewards those candidates who came in second or lower in their riding, but still have the strongest support in their province (based on percentage of vote obtained). They represent those people who voted for them, as well as many other people who supported that particular party in other ridings.

Because these candidates come from particular ridings, they could participate in the constituency work now done by MPs. On average, ridings would have two elected MPs, although each riding is only guaranteed one seat for the local representative. Constituency office budgeting may need to be reviewed after each election to ensure that all Canadians receive reasonable access to their elected officials.

One of the challenges of mixed member proportional systems is that there is always a possibility that you will need more representatives than there are ridings in a province. There are some situations, even in Simple MMP, where using a list is necessary. If a party has over 50% of the popular vote in a particular province, then it is highly likely that ALL of their candidates will be elected, either as local members or as top-up members. If more top-up members are required, then a provincially based top-up list would be needed to select the extra seats. Using the 2015 election as an example, this would result in a total of 4 list seats being selected. A good example is Alberta, where the Conservative party received over 50% of the vote. Under Simple MMP, all of their candidates would have been selected, and two additional would come from a provincial list.

It would seem to me that having 4 list seats across Canada is a superior outcome to having a third or a half of the Parliament come from lists. There is an analysis here which shows that 50% is a near optimum mix for minimizing both democratic deficit and list participation.

In conclusion, Simple MMP offers voters a choice, and their vote for a local candidate increases their chances of being the local member and also increases their chances of being a top-up member.

Provincial versus regional top-ups

Simple MMP uses a provincial top-up pool of candidates taken from the ridings in the province, compared to regional top-up lists in the LC recommended system. The variable in question here is the size of the top-up region.

At one extreme, it does not make sense to have a top up from a single riding, as this would simply be a system where the two candidates with the highest vote would be elected.  While this would actually improve the representation of the popular vote, it is highly biased towards large parties.  Parties that have a lot of third place finishes would not have representation for those votes, and a large number of votes would still be discarded.

The opposite extreme would be to have a national pool of top-up candidates. Simulations of this model provided an extremely representative Parliament, but runs into constitutional issues. It should be noted that using such a model with the 2011 general election data would have resulted in a single seat for the Christian Heritage Party. While some people believe that small parties need to pass a threshold in order to be represented, it makes sense to me that if a seat in Parliament represents 1/338th of the power, then if you have 1/338th of the people behind you, you should get a seat.

In general, the larger the pool, the more representative the result. The Ontario Citizen’s Assembly in 2005 (Click here to see report) recommended a province wide top-up region.

The LC system uses regions of 5 to 7 ridings.  This improves the representation and reduces the democratic deficit, but only for large parties, and in fact, maintains a bias for those large parties.

The primary argument for regions smaller than provinces is that the top-ups would have more connection to local issues. While geographic accountability is important, Simple MMP provides a significant amount of local representation by candidates with the greatest support within their parties.

The other argument against forming regions, is that these regions may not be “natural” divisions which represent truly unique needs and desires. We have enough divisions in our country already, and I am not convinced that creating more will help us. It is important to remember that we are electing a Federal government which will act for all Canadians, rather than pit regions against regions.

50% local seats versus 67% local seats

Simple MMP starts from a base which has 172 elected members (50.9% of 338) and 166 top-up members (49.1% of 338). A sensitivity analysis on this number shows the democratic deficit is improved as the percentage of elected members is reduced from 100 and reaches its optimum at about 50%.  Reducing the number of local seats below 50% does not improve the democratic deficit, and at the same time increases the number of top-up seats which must come from list seats. You can see this analysis here.

Using 67% local seats reduces the democratic deficit from 50.6 seats to 18.6 seats which is a clear improvement. However, by going to 50% local seats, another 12 seats can be allocated to the proportionally appropriate parties, leaving the democratic deficit at only 4.1.

Comparison Summary

Both Simple MMP and the Law Commission recommendations propose a mixed member proportional system for Canada using a (compensatory) top-up model.  There are some differences in how these systems are formed. The primary differences are listed below:

Simple MMP Law Commission Recommended System
Ballot contains a single vote Ballot contains two votes: one for the local candidate, and one for the preferred party
Top-ups are selected from the candidates who had the most support from the voters for their party Top-ups are selected from lists provided by the party prior to the election
Party lists are only used when the popular vote is over 50% for a particular party (estimated maximum 6 seats using simulated data from the last 5 elections) Party lists are used for all top-up seats (approximately 112 seats, 1/3 of Parliament)
Top-ups are selected from a large provincial pool which provides maximum proportionality Top-ups are selected from regional pools which improves proportionality, but still favours large parties
In each province, about 50% of the seats are local seats and 50% are top-up seats, with a higher level of proportionality In each region, about 67% of the seats of local seats and 33% are top-up seats, with a lower level of proportionality