Mismatch

Yet more anamolies on the electoral math front.

This pretty chart shows the total number of electors on the voters list in Newfoundland and Labrador at each federal and provincial election (and other electoral events*) since 1970:

[Data sources: Elections Canada reports passim; Elections Newfoundland and Labrador reports passim; Statistics Canada.]

The blue dots are provincial electoral events; the red dots are federal ones; and the green line is the estimated number of persons 18 years of age or older as of July 1st in each year, according to Statistics Canada estimates.

The federal electoral qualification is as follows:

3. Every person who is a Canadian citizen and is 18 years of age or older on polling day is qualified as an elector.

This is functionally virtually identical to the provincial one:

23. Every Canadian citizen 18 years of age or more on polling day is qualified to vote at an election if he or she is ordinarily resident in the province immediately preceding polling day.

So it isn't entirely obvious why there should be such a gap between the number of electors that Elections Canada counts, and the number that Elections Newfoundland and Labrador does.

This is especially so, given that the federal voters list, in the "modern" era, since about the mid-1990s with the move to the permanent register of electors, has steadily converged on the estimated total population 18 years of age or older. (In the 2001 and 2006 censuses, resident non-citizens, who wouldn't show up in the electoral lists, amounted to 0.7% and 0.8% of the population respectively.)

Some of the gap can be explained historically: the 1993 federal election was conducted on the 1992 Charlottetown referendum enumeration with revisions; this was the last full federal enumeration. The provincial elections and plebiscite of 1996, 1997, and 1999, were conducted on the same list of electors.

But why, between the 2007 provincial election and the 2008 federal one is there a gap of roughly 40,000, and with such wildly divergent trend lines? And how do you square that late provicial downward trendline with the jump in individual district voter counts observed in late by-elections?

Just how permanent is the permanent list of electors? For that matter, to what extent is it a list of electors? Heck, to what extent is it a meaningful list of anything?


* These are the 1992 federal referendum on the Charlottetown Accord and the 1997 provincial plebiscite on denominational education. If anyone can provide the total number of eligible voters for the 1995 plebiscite, 'twould be greatly appreciated and 'twill be incorporated into the graph.