On average a Canadian Member of Parliament [MP] elected in 2015 needed 25,063 votes to win, (all data used here is from this Elections Canada site) but the averages differ dramatically by party and region.
For example, the average maritime riding has about 75,000 people while the average Alberta riding has over 110,000 - just less than a 50% difference. Thus the 29 Liberals elected in the maritimes averaged only 23,581 votes for the win, while the 29 Conservatives elected in Alberta averaged 36,139 - somewhat more than a 50% difference, but a difference that means, in electoral terms, exactly nothing.
In the worst case, Ken Sorenson, the conservative who won 47,552 votes in Alberta's Battle River riding is a member of the opposition and therefore unimportant and invisible to those in power where Hunter Tootoo, a liberal from Nunavut elected with all of 5,619 votes, has a vote in the government caucus, some access to the Prime Minister's office, and correspondingly some minor role in government.
This situation is obviously unfair and undemocratic, but has been exploited by the Liberals since at least the the first Trudeau because the present system means that angering a western voter to please one in Quebec or the maritimes pays off in House votes at somewhat better than two to one - and a conservative running in Quebec or the maritimes can neither win by running against the unfairness of the system nor risk getting too close to party leaders from the west who might be opposed to it.
There are many reasons for these disparities ranging from historical accident through the economic and demographic effects of past political policy in the regions, to the deliberate shift of electoral power toward Quebec and the Maritimes orchestrated by the Liberals under the first Trudeau circa 1980; but the most proximate cause for the continuation of these disparities today is vote splitting in ridings where the Conservatives don't have overwhelming majorities.
From the NDP or Bloc perspective vote splitting in contested ridings meant, in 2015, that winning candidates needed, on average, only 20,400 votes - barely more than half the Alberta average for Conservatives.
From a liberal perspective the goal in ridings they can't win is to have the conservative lose because minority votes in the House don't matter, and seats held by the minority parties are largely irrelevant as long as those members are ideologically prevented from forming coalitions with the Conservatives if the Liberals manage a plurality, but not a majority.
The obvious remedy, seat redistribution, is impractical in Canada because the winners from the present system have the power to block change - and nothing short of a serious western Canadian threat to leave the Canadian federation in favor of eventual union with the United States seems likely to change that.
On the other hand Trudeau's outspoken commitment to electoral reform during the 2015 election and his subsequent repudiation of the reform commission's findings combine with the enthusiastically expressed support for both electoral reform and the one man, one vote, principle among the minor parties who expect to benefit from proportional representation to open the door to a fair and easy way to implement a simple solution: proportional voting by the elected rather than the electors.
Here's how that would work: winners in traditional first past the post elections would be thought of as representing the people who voted for them, with each of those votes counted equally. So if the average member across the entire House represents Y voters, and the specific member received X votes to win the seat, then that member's vote in the House is counted X/Y times instead of exactly once.
In effect members vote their shares of the popular vote.
In 2015, for example, the average winner across all 338 seats received 25,087 votes and member's votes in the House are counted as if each represents roughly the same number of voters. In reality, however, the average liberal represents far fewer votes than the average conservative and that disparity under-represents millions of Canadians. Counting votes in the House in proportion to the votes the member received rebalances the House according to the popular vote - thus Ken Sorenson's vote would be counted 1.9 times because he received nearly twice the average number of votes while Hunter Tootoo's vote in the House, representing less than a quarter of the average number of electoral votes, would only be counted 0.22 times.
Making this happen is simple: it requires nothing more than a rule change in the House - and that's both well within its purview and completely consistent with the expressed views of the parties: Green, New Democratic, Bloc, and Liberal, collectively making up about 72% of the House.
Making this change has very little impact on standings in the House after the 2015 elections: the Liberals retain their 184 votes but ten votes move from the NDP and Bloc in Quebec and the maritimes to the west - exactly the result redistribution would produce if done fairly.
The longer term effects, however, would be dramatic because this form of proportional representation removes the left's incentives both for vote splitting and for keeping left leaning ridings much smaller than those which lean Conservative. As a result it seems reasonable to expect that:
However, the membership of both have largely bought in to the claim that these parties are different rather than merely the careerist and ideological strata of the market for the left's claims, and so many will fight the merger. Thus a merger forced from the top down will fracture the leftist alliance that now all but guarantees power to the Liberals while a failure to merge will further shift power to the west.
Note: the principle behind the equalization program is that the provinces should provide roughly comparable services to their citizens, that taxpayers should share this burden regardless of their province of residence, and therefore -by the same logic as that animating the progressive income tax- that people in richer provinces should pay more per person than people in poorer provinces. In practice, however, equalization has allowed the Liberals to use western money to buy eastern votes. During fiscal 2010/11, for example, the federal government collected roughly $14,000 per person from Alberta and gave just over half this money to the governments of Quebec, PEI, New Brunswick, and Nova Scotia.
The equalization idea started in the 1920s but became critical to Quebec, and the liberals, only after the first Trudeau enshrined it in the 1982 Constitution Act. Since then Alberta has contributed an estimated $130 billion to equalization but has no real liklihood of ever receiving any significant benefit.
This would not be an issue for most Albertans if it resulted, like decades of insurance payments made without ever filing a claim, from Alberta's good luck or good management; but it does not: the disparity is purely political in origin with the Liberals and NDP leveraging population and riding disparities between the west and the rest of Canada to retain federal power in large part by giving politicians in Quebec and the maritime provinces the ability to deliver social programs that are effectively cost free to their voters.
Basically, everyone knows elections have consequences - and the rules governing those elections have less direct, but equally compelling, consequences. Thus proportional representation by the electorate gains enthusiastic support from the NDP and splinter groups like the Greens, Christian Heritage, and the Communists, but leads to governmental paralysis and a defacto transfer of power from the elected government to the bureaucracy. Proportional representation in the House, however, meets Trudeau's oft-expressed wish to treat each vote cast in federal elections equally, drives toward a clean two party system across the country, doesn't affect Trudeau's current majority in the House, and seems likely to eventually extinguish the present Liberal practice of using western money to buy eastern votes.