More number-crunching

Decima releases its poll on which voters are committed to which party - and plenty of votes are still up for grabs:

To sharpen the focus on the uncommitted voters who will likely determine the outcome of the election, Decima broke out the truly undecided and leaning voters from a survey of 4,804 voters outside Quebec, conducted Dec. 29-31.

From that larger survey, which has a margin of error of 1.4 percentage points 19 times out of 20, Decima found 20 per cent were committed to voting Liberal, 20 per cent for the Conservatives and 10 per cent for the New Democratic Party.

It found 47 per cent were uncommitted, including 14.2 per cent who were truly undecided and another 32.4 per cent who said they may yet change their vote, although they were leaning to one of the parties.

In analyzing that 47 per cent, Decima, which is working with Carleton University's School of Journalism and Communications, found two distinct groups: 12 per cent who were torn between the Conservatives and Liberals and 14 per cent who were torn between the Liberals and the NDP.

The article goes on to discuss the apparent traits of the two largest swing groups:

Those torn between the Liberals and NDP were more likely to be female, aged 35-54 and members of a visible minority. They tended to think the Liberals were the best choice to govern and that Martin was the best choice for prime minister. However, they also tended to think the NDP had the best approach to issues they care about most...

Voters torn between the Liberals and Conservatives tended to be female, under 35 or over 55 years of age, and more likely to be on the right of the political spectrum. At the time of the survey, they were leaning equally towards the two parties.

Nevertheless, this group tended to think the Liberals were best to govern and Martin the best choice for prime minister. However, the Conservatives had a slight edge as the party thought to have the best approach on the issues these voters care about most (48 per cent for the Tories versus 42 per cent for the Liberals).

It's particularly interesting to see that swing voters on both sides tend to see the Libs as having a weaker approach to the issues they care most about. Clearly the very swing voters targeted by the Libs' attempts to be all things to all people aren't convinced by the policy of promising everything; the question is to what extent each of the opposition parties can cast doubt on the Libs' qualification as the most capable government.

But what happens if both opposition parties succeed in putting a dent in the Libs' share of the swing vote? Well, let's divide up the numbers if both the NDP and Cons manage to pull in 2/3 of their respective swing votes:

Con = 20 + (2/3)(12) = 28
Lib = 20 + (1/3)(12) + (1/3)(14) = 28 2/3
NDP = 10 + (2/3)(14) = 19 1/3

This breakdown would leave 24% of voters unaccounted for, including the NDP/Con swing (whose numbers aren't listed) along with voters planning to vote for other parties, and the truly uncommitted. Suffice it to say that it's hard to see how any party could take a particularly large advantage within that group - and it would take a landslide among these voters to bring about a majority government.

Note that this assumes that the Libs earn substantially less of the swing votes than they held as of the time of the survey - yet the Libs would still be just behind the Cons in the ROC (figuring that there are enough Con/NDP swing votes to add a bit to the above Con total), to go with their presumed seat advantage in Quebec.

Which means that to the extent that NDP/Lib swing voters may be swayed by concerns of a Harper majority, Decima has given yet one more example of the lack any reason to worry.