TORIES ENJOY CLEAR BUT MODEST LEAD AS ELECTION LOOMS - March 10, 2011
UNDERSTANDING WHERE VOTERS ARE COMING AND GOING
[Ottawa – March 10, 2011] – Following a brief breakout a month ago, the voter landscape has settled into a pretty stable pattern with the Conservatives enjoying a clear but modest 7.4-point lead over the Liberals. This survey provides some interesting analysis on the question of how stable that lead might be and where voter movements have occurred, and where they are likely to occur in the future. Coupled with an analysis of the underlying demographic and regional patterns, this provides a revealing portrait of the pre-campaign electorate.
The survey data were collected over the past two weeks (excluding weekends) and are based on random sampling of the entire population (land line and cell phone only households, internet and non-internet households). In order to enhance comparability and to isolate real effects we always present the voter intention questions in exactly the same order and position in the survey (to avoid unknown contamination effects). We also employ the same sampling procedures from poll to poll and gather large samples (in this instance, roughly 2,900 cases) in order to strengthen the ability to separate the real and the spurious. Our conclusion is that the electorate have now moved into a pretty stable wait and see mode.
Apart from the overall national lead, the most notable drift that we have seen over the past several polls is that the Conservatives are now significantly ahead in vote rich Ontario. This is a very significant and fairly newfound advantage for the Conservatives. The Liberals are significantly ahead in the Atlantic and have a small but significant lead over the Conservatives in Quebec, the only real sour note in the poll for Conservatives. The NDP are now within the margin of error of the Conservatives in Quebec. Coupled with other findings in the survey, there is evidence that the NDP could be poised for something of a breakthrough in Quebec. The Bloc remains solidly in the driver’s seat in Quebec; an advantage that may be understated given other findings we will discuss momentarily. There is nothing else particularly novel in the demographic or regional patterns seen in this iteration of our polling.
What is new is an analysis of where voters are coming and possibly going. In addition to our usual questions we revisited the question of how people say they voted in 2008 and compared that to where they are today. We also asked respondents what second choice they might move to in the future. Between these two indicators we see some indicators of loyalty, and potential flux which otherwise wouldn’t be evident.
First, the issue of voter mobility which is examined by cross classifying current preference with stated voting behaviour. Depending how we array the table, we can see both where parties have experienced losses and where they have secured gains. The degree to which they end up on the diagonal (“stayers” as opposed to “movers”) can be taken as a measure of party loyalty. On that issue, it is interesting to compare the level of voter loyalty in Canada with the United States. Canadians are much more politically promiscuous then our Southern neighbours. In the States, 90% (sometimes more) voted the same way they did in the last election. In Canada, roughly 2 in 3 stick with their last choice reflecting a more fluid, less ideologically entrenched electorate.
Looking at this indicator by party, we see that supporters of the Conservative are the most loyal, with nearly 80 per cent of 2008 supporters staying on board (a finding evident in other indicators as well). The Bloc also fares very well with voter loyalty, whereas the Liberals and NDP experience more voter churning and lower attachment levels. Both the Liberal and NDP supporters are less firmly attached and show more mobility. Unsurprisingly, the Green Party voters are the least attached, reflecting their younger age profile and the fact that the Green Party does better with eligible voters than it does with actual voters.
Another key finding is that a clear majority of all voter movement in Canada over the past three years has been an ongoing transfer back and forth across the Liberal and Conservative support groups. This pattern has a number of interesting implications.
Apart from the tantalizing but practically ludicrous notion that the most natural coalition would be a Conservative-Liberal alliance, these trends in voter movement also underline the depth of the challenges that Michael Ignatieff and the Liberals face. Many potential and actual Liberal voters are relatively ambivalent about the two parties, but the Conservative Party has the larger and more attached base and there is little evidence that the Liberals have had much luck prying these votes loose. In fact, the net transfer across the two constituencies mildly favours the Conservatives, and they are now ahead with a larger and more committed base. The more fluid opportunities exist to the left but to date the evidence is that the Liberals haven’t been able to provide much attraction in these quarters. The NDP have a similar problem with their voters casting both right and to the Green Party in a fairly indistinct manner. Finally, Green supporters move in an almost random manner across alternatives. All of this points to a modest Conservative advantage.
In the case of second choice, the news is less favourable for the Conservative Party. In fact, they have the lowest second choice appeal out of the four federalist parties and it may be that the very strategies which engender loyalty that lower their ceiling with other voters. The NDP fare best as second choice, and this is quite pronounced in Quebec. The Liberals fare better than the Conservatives on second choice, which may offer a mild sense of optimism as we approach an election. Finally, there is an interesting discrepancy between words and deeds where voters tell us they are more likely to move left but the actual mobility patterns suggest that this has decisively not been the case.
Click here for the full report: full_report_march_10_2011