Референдум в Онтарио о пропорциональной избирательной системе.

Canadian provincial referendum on establishing mixed member proportional representation

A referendum was held on October 10, 2007, on the question of whether to establish a mixed member proportional representation (MMP) system for elections to the Legislative Assembly of Ontario. The vote was strongly in favour of the existing plurality voting or first past the post (FPTP) system.

Currently, Ontario elects Members of Provincial Parliament (MPPs) using the single member plurality, or first past the post (FPTP), system. In this system, each voter gives one vote to a candidate in an electoral district; the candidate with the most votes wins. In most cases, the party with the most elected candidates is asked to form a government. The initiative to reform this system was first proposed in 2001 by the Liberal Party opposition leader of the time, Dalton McGuinty. The impetus for the proposal was at least in part the experience of the province with two successive majority governments elected in three consecutive elections with less than 50% of the popular vote, the NDP from 1990 95 under Bob Rae (elected in 1990 with only 38% of the popular vote), and the Conservatives, twice elected with 45% of the vote in 1995 and 1999 under Mike Harris. When the Harris government enacted sweeping changes in public services, a feeling developed that both of these governments lacked a sufficient democratic mandate to justify the extent of such changes. The Liberals won a landslide majority of their own in October 2003, winning 70% of the seats with 45% of the vote. On November 18, 2004, Premier McGuinty announced that a citizens’ assembly would be established to examine the FPTP electoral system and recommend possible changes to be voted upon by referendum in the next provincial election. Enabling legislation to implement these measures —the Election Amendment Act, 2005— received Royal Assent on June 13, 2005. It included provisions for Elections Ontario to select volunteers for a Citizens’ Assembly on Electoral Reform. An all party Committee on Electoral Reform was also created to consider and report on options for electoral reform. Its report was submitted in November 2005. It included a number of recommendations and defined the Citizen's Assembly's mandate. The Ontario Citizen's Assembly was modeled primarily on the British Columbia Citizens' Assembly on Electoral Reform established as the first of its kind in 2002, leading up to the British Columbia referendum in 2005. It was composed of 103 members, one from each of Ontario's existing ridings, including 51 women, 51 men, and one native member. Its deliberations began in September 2006 and ended with the Assembly's final meeting on April 28, 2007. Proceedings included a learning phase as members learned about the strengths and weaknesses of different electoral systems, a period of public consultations, and a "deliberative" phase during which the Assembly came to a consensus on its recommendations.

The Citizens' Assembly on Electoral Reform proposed a mixed member proportional representation (MMP) system. Although there exist slightly different variants of MMP, the model proposed for Ontario by the Citizen's Assembly was designed with simplicity and practicality in mind.

List members would be chosen according to a ranked list of candidates proposed by each party before the election. In the Assembly's proposal, there would be only one such list per party for the whole province, as opposed to several regional lists, and it could include candidates also running for office locally. If a candidate for a list member seat had already been elected locally by the time that list seats were being allocated, that candidate's name would simply be passed over in favour of the next candidate on that party list.

The number of seats won by a party would be calculated by multiplying a party's share of the party vote by the total number of seats in the province. As this multiplication would not automatically yield a round number, the fractional part would be allocated in favour of the party or parties with the largest fractions. For example, if there were four parties with party votes yielding 45.43, 31.54, 24.76 and 5.27 seats respectively for a total of 107, the initial allocation without the fractions would be 45, 31, 24 and 5, for a total of 105 seats so far. This would leave two seats to be allocated on the basis of fractions; in this example, these seats would go to the second and third parties. The final tally of seats would thus be 45, 32, 25 and 5. This approach to allocating the fractional seats is called the largest remainder method. The result is as close to proportionality as possible using a simple method given the need to allocate the fractional seats in some way. However it is not necessarily as close as that from the more complex Sainte Lague method used for New Zealand's MMP system, given that unlike the latter formula the largest remainder method may result in the so called Apportionment paradox in allocating seats.

Since the number of local seats is established by the election results at the local level, what remains to be done is to fill in the difference with the required number of list member seats for each party. In the above example, if the first party had won 39 local seats, they would now be allocated 6 more seats from the list, for a total of 45 as determined by their share of the party vote.

A particular situation that could arise, though rarely, is that a party might win more local seats than it would normally be allowed in total based on its share of the party vote. This excess of seats is called an "overhang" and in some systems, a form of compensation is made for this by adding "leveling seats" to the other parties. The Assembly did not allow for leveling seats, accepting instead that the system would be slightly less proportional in such cases (see the article "Overhang seat", section "Dealing with overhang seats" for the alternative method that was proposed).