Demographics are the key

Posted October 24, 2008 by Sacha Peter - Link
Category: By-Election Comments Off

I was interviewed by Jackie Wong (who writes for the Westender) about the upcoming by-election of Vancouver-Burrard.

The article she wrote about is on The Tyee, but I will post a complete transcript of what I wrote to here here. Also note I have provided two more maps that were not already up on the by-election link.

1. You offer a comprehensive map of the voter patterns in Vancouver Burrard from 2005. The map suggests that northern and eastern regions of the riding (Coal Harbour, Yaletown) showed Liberal voting tendencies, while western and northeastern regions (the West End, Gastown/Downtown Eastside border) showed NDP tendencies. Why do you think that is?

The one word answer is demographics. More specifically, two key variables would be income level and residential ownership. For example, people that rent apartments downtown would more likely support the NDP, while the people with mortgages on their condominiums would be more likely to vote BC Liberal. Generally, the way that the parties have been messaging would lead middle to low income earners to support the NDP, while middle to upper income earners to support the BC Liberals. The key point here, however, is that the middle income earner (as well as the political centrists) is the ultimate “swing battleground” for both the major parties in terms of demographics.

One can also walk the streets of downtown in the support areas of the NDP and BC Liberal and it doesn’t take very long to discover the “flavour” of the neighbourhood that supports each political party.

A question you never asked was the impact of the Green party. You can see (by the attached map) that a typical Green voter is geographically in closer alignment with the NDP than to the BC Liberals. In the last election, the Green party received 13% of the vote, so they cannot be ruled out if they are able to do a masterful job of mobilizing their base. 10,354 people voted for Adriane Carr in Vancouver Centre in the last federal election.

2. Do you think regional voting patterns in this 2008 byelection will be similar to those of 2005?
3. What effect will swing regions of the riding have on the byelection?
4. Where are the swing regions in the riding? Have they changed since 2005?

Yes, the voting patterns should be similar, with the noted difference that in a by-election (with low voter turnout), the individual candidates’ public persona and local campaign becomes more relevant since the provincial parties will be taking less of the media spotlight.

I have attached a map with the polls that were within 5% of each other (between the BC Liberals and the NDP). You can compare this with the map showing the poll winner and see that this is all on the perimeter of the NDP “circle of strength”. Essentially the BC Liberals have to concentrate inwards, while the NDP have to concentrate outwards.

It becomes less of a matter of changing voters’ opinions, but rather getting your core demographic out to vote for you.

5. What effect will voter fatigue have on the byelection, when voter turnout for byelections is already generally lower than turnout for general elections? Do you know anyone who has conducted local research on voter fatigue?

There will be a negative impact to voter turnout in this by-election due to the proximity of elections. The last provincial by-election was October 28, 2004 in Surrey-Panorama Ridge, which received a whopping 52% voter turnout compared to the 69% in the previous 2001 election. I would be shocked beyond belief if more than 50% turned out to vote in Vancouver-Burrard. People in Vancouver-Burrard and Fairview will be facing three elections in a month – federal, provincial and municipal.

The Mayor’s race will have more of an impact this election – people ordinarily supporting the NDP would be assisting Vision Vancouver/COPE, while people ordinarily supporting the BC Liberals will be assisting the NPA. When you factor in a lack of volunteer support, plus the fact that the by-election will be giving only a 6 month mandate to the winner before the general election is called, it will not be surprising to see very low voter turnout.

Unfortunately I do not know anybody who has conducted research on voter fatigue, other than the real-life experiment that will be on October 29.