The 1984 election is but one example: the Conservatives, with about 50% of the vote, nonetheless took 75% of the seats. Indeed, such misleading landslides have been much the norm in recent Canadian elections, several of which have seen power slosh violently to one end of the pool, whether in Alberta, Ontario, Quebec, or in the most extreme case, New Brunswick.

In part this can be put down to the craving for pork the nation has acquired over the years. As Charles Langlois, Tory candidate in Brian Mulroney's old riding of Manicouagan, so fetchingly put it last week, ''the important thing is to elect a member on the government side. Brian Mulroney will form the next government, so we want to be on his side, not in the opposition.''

But something else is at work here: a structural flaw in the electoral system that exaggerates majorities. The source of the trouble is our ancient first-past-the-post method of constituency representation. A mere plurality of the votes in any riding, say 40%, is enough to claim 100% of the members it sends to Parliament (which is to say, one).

It is for the same reason common for a party with the support of 20% of the voters in a region, for example, the Grits west of Ontario, to win barely a handful of seats, accentuating the tendency to regional ghettoes that so disfigures Canadian politics.

REFLECT SUPPORT

Proportional representation (PR) is designed to correct this failing. The idea is simple: voters should get what they vote for. The composition of parties in the Commons ought, more or less, to reflect their support in the electorate at large. There are limits, clearly, to how far this can be taken: if Parliament comprised, say, 100 members in total, then a party would require at a minimum 1% of the vote to gain entry.

Opponents of PR usually voice their objection in one word: Italy. Since small parties find it easier to gain admission, and since large parties find it harder to win majorities, we'd be condemned to the endless merry-go-round of governments Italy has endured since the war.

But for every Italy they cite, one could as well point to a Sweden, or West Germany, whose systems combine PR with a high degree of political stability. Indeed, most of the governments in Western Europe are elected under some form of PR; the difficulties Italy has encountered owe more to local factors.

For that matter, it's not wholly clear the situation in Italy is so untenable. A bit of paralysis in government hasn't stopped the Italian economy from enjoying its own postwar ''miracle,'' nor the recent surge in productivity. Who knows? It may even have helped.

A related objection is that PR tends to coalition governments. But governments in the Westminster model are as much coalitions. The Tory party houses several sub-parties under its roof; at least two parties coexist within the Liberal ranks, not to mention the phantom caucus of such exiles as Don Johnston, Donald Macdonald, and Doug Frith. PR's only contribution is to make the distinction between these ''parties'' explicit, and so bring the flows of competition, coalition and consensus out in the open, where we can see them.

The attendant benefits from releasing the pent-up pressure for realignment in Canadian politics are several. Reformers in Canada have paid much mind to freeing MPs from the constraints of party discipline. This is probably futile: smaller parties, on the other hand, could better reflect the views of individual members.

Similarly, one of the great failings of our democracy is the voters' inability to make choices in different policy areas separately, rather than in ''bundles.'' Most polls show a majority for free trade, for example, and against Meech Lake, but no party offers this tandem. Referendums are one solution; PR is another.

Constituency representation could be preserved, to meet another concern, whether by a system of multi-member ridings, or a hybrid of single-member ridings and members elected at large. The broader the constituency to which MPs would be responsible, the less would they be prey to the demands of narrow interests.

Whatever its merits, it might be supposed that the chances of PR ever being implemented are decidedly dodgy. But a third party, finding itself with the balance of power in a future election, might well make PR a precondition for support. Ed Broadbent's dreams of a two-party system make him an unlikely reformer in this vein. But the Liberals, if they wish to avoid the fate of their counterparts in Britain, might do well to take the proposal on board.