Sunday, October 10, 2010

Why it is so difficult for Harper to win a majority

A column in yesterday's Star by Jim Travers renews the endless Ottawa speculation about a Harper majority and concludes that Harper "is closer than it appears to the majority he covets."  He quotes pollster Nick Nanos as saying:
Conservatives... are both efficient in converting votes into seats and unusually skilled at "repelling people from voting for others".
TC has discussed this issue before but it is worth reviewing again.  The largest obstacle to any party winning a majority under our current party system is that the Bloc Québecois controls between 40 and 50 seats in Quebec.  A majority must come from the seats that remain. If we assume the Conservatives can win at least 10 seats in Quebec (although most recent polls suggest it will be a few less) then the party must win 145 of the 233 seats outside Quebec to get a bare majority of 155 seats.  That 145 number is 62.2% of all non-Quebec seats. (It would be a little less if the Conservatives could win additional seats in Quebec.)

One can assess the likelihood of this by comparing it to the era prior to the advent of the Bloc when parties did not face the same obstacle to a majority.  I am ignoring the post 1993 period because as TC wrote before:
Chrétien's majorities were flukes in the sense that they depended on an even split between Reform/Alliance and PCs in Ontario and very low NDP numbers, which themselves were a product of a temporary decade-long depressed support level caused both by unpopular provincial governments (Harcourt/Clark in BC and Bob Rae in Ontario) and unusually weak federal leadership, principally Audrey McLaughlin. The NDP did begin to revive a bit under Alexa McDonough and gained new ground in Atlantic Canada but remained very weak in Ontario.
If we ask which majority governments won 62% or more of the seats in the pre-1993 period of Canadian history but after 1921 when third and fourth parties began to emerge, we find the PC majorities of Mulroney in 1984 and Diefenbaker in 1958 as well as the Liberal victories of St. Laurent in 1949 and 1953, and Mackenzie King in 1935 and 1940 (King's wins, however, were strongly affected by the depression and the war).  There weren't that many, they tended to be early wins (Mulroney in '84 and Dief in '58) and there have been just two since 1958. Trudeau was close in 1968 but he won just 58.3% of the seats.

As for Nanos two points: first, the Conservatives are not particularly efficient at converting votes into seats.  For example, look at all the wasted votes they accumulate in their huge majorities in Alberta.  Second, with respect to vote suppression, one should note that in the last election turnout was 58.8%, down from the previous three elections. It appeared to TC that it was particularly concentrated among Liberal voters who lacked confidence in Stéphane Dion. How much greater vote suppression can one realistically expect when over 40% of Canadians already don't show up at the polls?

What Travers and Nanos are asking us to believe (without reviewing the math outlined above) is that a government that has been around for six years can win more than 62% of the seats realistically available to it.  It is possible in the sense that anything is, but it is also not very likely.