Media post by-election analysis

There has been a fair amount of analysis of the by-elections in the media. Most of it, in TC's view, not worth much.

The media likes the Greens (partly because of their newness), but simply doesn't understand how weak the Greens really are. TC wrote about this before, demonstrating that the Greens have a very inefficient vote. The Greens are more harmed by the first-past-the-post system than other parties.

On the other hand the NDP has a highly efficient vote, and is good at focusing its resources to get results. While the NDP did not do well in this week's by-elections, the last time the NDP made a major effort, in Outremont last September, they achieved a breakthrough. TC suggests ignoring much of the analysis that has been written this week about the party's woes.

The best the Greens could do in the by-elections was to increase their vote share by a factor of 2.6 (in Vancouver Quadra). As a hypothetical exercise TC took the results of the 2006 election and multiplied the Green vote share by 2.6 in each province (adjusting the other parties appropriately). The result gives the Greens a national vote share of about 12%. However, when applied to TC's forecast model, even this large increase would still not deliver any seats to the party.

Such an increase is not likely to happen as it seems likely to TC that Elizabeth May will in some way endorse Dion in the next election. However, the risks of their collaboration were made evident in the Quadra by-election where the unexpectedly weak Liberal showing meant the Green vote almost handed the seat to the Conservatives. This may have some as yet undefined longer term impact.

TC's view is that more and more Green and NDP inclined voters are thinking about and actually voting strategically in order to prevent Conservatives from winning. One place to look for it in the next election would be in Vancouver Quadra where expectations will be changed by the results of the by-election this week. The increase in the Green vote owed much no doubt to the widespread perception that Quadra was a 'safe' Liberal constituency. Quadra indeed turned out to be a safe Liberal seat. A good definition of 'safe seat' for a party would be a constituency where, even if the party's candidate is weak and loses significant vote share to third parties, and faces a vigorous challenge from the party's major opponent, the party still wins.