Cold Hard Wonk

No sentiment but politics

Prologue — Bases Unloaded

Posted by JJ in Federal Elections, Strategic Planning, Golden Tacks (Thursday September 21, 2006 at 10:57 pm)

With the first real signs of the result of the Liberal Party of Canada’s Leadership campaign imminent, there’s good reason to revisit what’s at stake. For Liberals warming to the idea that they might retake Parliament sooner than expected, it’s important to think about how to do just that; and that task demands that they know where they stand.

Electoral support comes in two flavours: base and bonus. Base supporters are groups who feel so deep a connection with a party that their support tends to follow the party from election to election. Bonus supporters are attracted to a party by the position they assume by siding with it in a given election, whether because they identify with policies or want to distinguish themselves from other parties. Elections focus on both groups, but in different ways.

It’s virtually impossible to change your base during a campaign, but it is possible to motivate them. If you can get them excited enough about your chances, they become more likely to vote and encourage others on your behalf. If you alienate them, they’re likely to sit out.

Bonus voters are where there’s an opportunity to make real gains in support during a campaign. A masterful riposte during a debate, a well-chosen policy, or serious gaffes can turn these voters quickly; and without loyal starting positions, they may shift alleigances many times before the race is run.

Obviously, it’s easier to win with a large base than with lots of bonus voters. What isn’t so easy is building a base large enough to make it that much easier. The more people you add to your supporters, the harder it becomes to reconcile some differences and put off making choices between competing groups on others. Building a base means careful work over time, both in government and out. That’s why being in “campaign mode” while in government may or may not be a valid criticism. Campaigning to motivate the base isn’t very useful in office. But campaigning to fold other groups into your base is the most politically useful thing you can do. The problem is that it’s not the same kind of campaign you fight to win them over, come election time.

But enough of the obvious. Why all the fuss? It’s crucial to understanding where the Liberals now stand. So is looking back a bit.

In two elections against Brian Mulroney, the Liberal Party took around 30% of the popular vote. In 1993, the Liberals began the election polled at around 37%, not far from either the Tory total or their performance in the previous election (31%). Most interestingly, while Tory support dwindled in polls from 35% to 16%, Liberal suport rose only to 41%. While 4% of difference, concentrated in Eastern Canada, was enough to assure them of a commanding 177 seats in the 295 seat House of Commons, the Bloc Quebecois nearly doubled and the Reform Party more than doubled that theft of support, rising by 6% and 9% respectively.

Equally telling in 1993, though, was the decline in voter turnout from previous campaigns. From 1957 onward, only two elections drew less than three quarters of voters: Trudeau’s 1974 and 1980 Liberal majority governments, both of which came within two years of the previous election. Voter turnout dropped from 75.3% in 1988 to 70.9% in 1993. Had voter turnout remained roughly constant, the Tory base should have grown by 450,000 voters. That number, 3.2% of the total votes cast, couldn’t have defeated the Liberals (even if it remained loyal), but is so close to the growth in Liberal support during the campaign that it should illustrate the significance of demotivating your base, assuming that these were mostly alienated Tory voters.

1997 saw a further 3.9% decline in voter turnout, leading to nearly 700,000 fewer votes being cast than during the previous election — nearly the precise total lost by the Liberals in going from 5,647,952 to 4,994,277. The Liberals stood pat in British Columbia (six), Prince Edward Island (four), and the Northwest Territories (two), gained six seats in Quebec, and lost seats in every other province save one. While the Liberal total increased by three seats in Ontario, that province gained four seats in that election, meaning that proportionately, the party took fewer seats than in 1993.

The 1997 losses were largely ascribed to the public outcry over the Liberals’ failure to remove the hated Goods and Services Tax — widely regarded as the most important of their 1993 campaign promises. The correlation between the decline in turnout and in Liberal support would tend to support that conclusion. Moreso when you consider that the Liberal total would have been within about 5% of the same total vote had their supporters increased in line with the general population from 1988 to 1997.

Most of the Liberals’ gains retained in 1997 could therefore be attributed to natural growth of their base, and most of their loss to the loss of bonus voters gained in 1993, attracted to a position opposing the Tories and the GST alike. This is especially likely considering that the GST question was more an incidental policy than a central aspect of the party. But notice two things. The expected growth in the base accounted for almost all of the Liberals’ retained voters from 1993, suggesting that the Liberals either took little advantage of the opportunity to expand their base, or did, and alienated part of their existing base in the process. More importantly, though, there was no indication that the 5% of voters who failed to show in 1993 had returned.

The 2000 election provided the Liberals with an excellent opportunity to regain ground. They were faced with a clownish and unskilled opposition leader, who provided significant political fodder even for non-partisan observers. The Liberal vote grew, but only by 250,000 votes or so to 5,252,031. That growth is more significant when considering that turnout declined once more, falling by 2.9%.

But capturing a larger share of a smaller pie when confronted with lackluster opposition doesn’t inspire much confidence. The Liberals gained only eleven seats from party standings just prior to the election, of which eight came from recovery of seats in the Maritimes. Only three provinces posted real seat gains — one apiece in Saskatchewan and the Yukon (not technically a province, true) and the remainder in Quebec.

Most importantly, more voters had been alienated, though not only by Liberals, and those voters frightened into voting Liberal by the prospect of a Creationist Prime Minister might just have come from the New Democrats’ stock of bonus voters. That party lost roughly 300,000 votes between elections, largely in the Maritimes, where the Liberals picked up most seats, while the combined totals of the Alliance and rump Progressive Conservatives lost 170,000.

In short, the Liberals relied on bonus voters again in 2000, showing no evidence of real growth in their base constituency over the previous twelve years. This lack of growth becomes all the more important when considering that the total number of eligible voters grew by 1.6 Million between 1997 and 2000.

And now, the interesting parts.

In 2004, voter turnout dropped to the lowest ebb in Canadian history, to 60.9%. Nevertheless, the total number of voters rose, due to the continuing growth of the electorate.

In those circumstances, and faced with the fallout of the sponsorship scandal in Quebec, the Liberals lost 33 seats and majority status, but lost only about as many votes (300,000) as they’d gained in 2000. What happened? Had the turnout been as low in 2000, 570,000 fewer votes would have been cast. Consider that the PC party was merged with the Alliance to create the new Tory party.

The PCs polled 1.6 Million votes in 2000, and the Alliance 3.2 Million. In 2004, the “combined” total came to 3.9 Million — a gain of 720,000. Combine that with the fact that the Green party gained 480,000 votes, and the possibility of alienated PC and Liberal voters sitting the campaign out helps to explain the drop in turnout. Once again, Liberal electoral support didn’t drift very far from earlier results.

Everything changed in 2006. Voter turnout rose by four percent — this was the first election to meet and exceed the number of votes cast in 1993 (by roughly 1 Million votes — exactly the growth in the Tory vote over their take in 2004), despite over a decade of population growth in the intervening three elections. And the Liberal vote total declined by nearly half a million.

It’s almost certain that the Liberals picked up few additional bonus voters in 2006 — their polls showed declining support; but more significantly, the numbers suggest little growth in the base. The voting population grew by 29.5% between 1988 and 2006. Had the Liberal numbers grown at a comparable rate, they would have taken 5.4 Million votes in 2006, a much more competitive result.

Stop for a moment and think about that. In 1988, the Liberals trailed badly and lost. Had they been only as popular among the general voting population in 2006 as they had been in 1988 (that is — losers), they would have done better by nearly a million votes. That’s not good. The increase in votes cast roughly matched the Tory gains. The rebound in voter turnout did nothing for the Liberals.

All of which is not to suggest that all 4.2 Million Liberal voters in 1988 were base support — they were, of course, a combination of base and bonus voters. The point is that the Liberals were the only party with a degree of continuity over that time, which could have been an opportunity to work on their base.

Impossible? Certainly not — Mackenzie King began an era of Liberal government in 1921 that was interrupted only twice — by a two-month Conservative government in 1926 and a narrow loss at the outset of the depression in 1930 — until it ended in 1957 — a span of 36 years over which the growth in Liberal votes regularly outperformed the growth in overall voting population.

Twice the length of time? True. But 1988-2006 was both a long enough stretch and a large enough change in the number of voters (+30%) to expect more growth in Liberal numbers than 10% or so from a fundamentally worse performance in 1988 (83 seats of 295 in 1988 versus 103 seats of 308 in 2006). The degree of stability over the 36-year span was comparable to that over the 18-year span. There is no evidence of growth in the Liberal base, and much evidence of relying on poor opposition and demonised opponents to swing bonus voters to their position. Short term, that can be a successful strategy. Long-term it’s a recipe for a drawn-out and gradual death.

And that’s what the Liberals have to worry about. The idea that drooping poll numbers for the Tories will translate into a Liberal chance at forming government will dangerously distract party members from the hard question: what should we do to grow our base?

Leadership contests aren’t good opportunities for that kind of discussion because they’re so highly focused on an appeal to the existing base. Moving from that to a snap election won’t help things any. As Liberals move towards choosing the delegates who will vote for their next leader, they need to consider how to grow the party, not simply how to attract a few bonus voters in 2007 or 2008.

Working on building the base could ultimately lead the party back to the easier dominance it had in its glory days, rather than into another sequence of lurching and desperate elections. The first question for Liberals, then, should not be changed to “how do we win the next election” in the light of favorable short-term polling. It must still be “how do we grow our base again,” if they’re serious about political power.

No comments for Prologue — Bases Unloaded »

No comments yet.

Leave a comment


(required but not published)

RSS feed for comments on this post. TrackBack URI