Poll by Poll analysis

Poll by poll analysis of the Ottawa West-Nepean Riding.

OttawaWestNepean-pollbypoll_bureauparbureau35066.xls

Replies to This Discussion

Permalink Reply by David Balson on April 5, 2011 at 12:24pm

We are purchasing 40 bright yellow T-shirts with “Vote for Democracy Not Baird” on 20 of them and “...Not Harper” on the other 20, with the Catch 22 url and authorization statement on the back of them. This has been endorsed by the national campaign with costs shared between the campaign and members. We have analyzed the polling data and it is fairly clear where there are concentrations of low turn out and in some cases also close polls within the riding. To get through to people one can go door to door to give pamphlets and to discuss, attend events to give speeches and/or ask questions, and/or go to areas where there are concentrations of people to extend a message, give out material and to discuss. Some of the Ottawa group are focusing on the third approach...spending considerable time over the last three weeks of the campaign, especially Thursdays, Fridays and Saturdays, in groups, in the t-shirts, engaging potential voters at Bayshore, College Square, Lincoln Fields, the new Walmart and several Community Centres.

The intention is to attract attention, engage people and encourage strategic voting.

We do not intend to spend time with conservative voters; we do not have the time or resources to convince them of anything; rather we want to engage people who support the other parties or those who do not vote at all to think about what kind of Canada they want and to point out how they actually CAN make a difference by voting strategically.

In addition to convincing some people to vote strategically, we will be advertising the Campaign through the url and distribution of printed material if available, we hope to create some buzz, get some media attention and hopefully encourage other potential agents of change to join the campaign both as active demonstrators/communicators and as passive supporters.

Details of this part of the campaign will be discussed at the next face-to-face meeting of the group (hopefully on Monday).

Permalink Reply by Francis Fuller on April 5, 2011 at 4:55pm

David:

Are all the shirts spoken for? What is the price for a large "Not Harper?"

I am presently sifting through the 2008 polling data to identify swing polls, creating google maps with poll information and intend to pass out URL to Catch-22 page that hosts all similar URLS for polls in the riding via means of print flyer distributed in poll precinct.

I am somewhat worried about time constraints. I want to try and get a flyer together for a trial run fairly quickly and see what response that generates, get a valid door-to-door time estimate and also costing. Then evaluate, fine tune if required, and go for volume.

Question 1

How, who, did you go through to obtain Catch-22 approval?

Question 2

Do you have an XLS or ODF with your target polls that I can cross reference with my results?

Question 3

Do you have any information on how out of pocket costs are to be handled? I had a communication from Gary S on need to submit expenses, get approvals from Catch-22 org as it is a registered 3rd party. I clearly can donate my labour. I am not clear on other "donations in kind." For example if I print out 100 flyers on my machine using materials I paid for, does this constitute an "in kind" contribution to Catch-22? If I purchase 100 flyers do I need to send a bill for costs plus a matching donation to Catch-22? These are possibly stupid questions but I expect that if we are successful (In OWN or elsewhere) there will be recriminations, attacks, we may be ejected for hostile writings on a public blog, condemned for bumper stickers, even de-citizenized for T-Shirts unfavourable to the great man himself. Maybe the amounts are too small to bother with but I always like to know the legal jeopardy I may be facing. This is true even though I have never inhaled or had relations with Bruce Carson.

Cheers!

Permalink Reply by Larry Rochette on April 5, 2011 at 8:41pm

Gary has to clarify the donations and expenses issues. We could run into problems with Elections Canada if we go over their limits for ridings.

Personally I'm at at loss how to get donations to our group. It appears that Gary is working on this issue.

Permalink Reply by John Elliott on April 6, 2011 at 11:49pm

Francis,

Did you look at the spreadsheet that is at the top of this post?

Suggested polls are in the spreadsheet.

jre

Permalink Reply by Francis Fuller on April 7, 2011 at 3:39am

Thanks John

I started with that spreadsheet and then went to the official data as downloaded from Elections Canada?? as the Green vote was omitted.

My hunch is that we will see a lot of the Green vote migrate to the Libs as the Greens are rational decision makers and they will be likely to vote strategically.

NDP may be more party bound and less likely to shore up the Libs.

Now all I need to do is write an XLS macro that will go door to door and distribute flyers and I'm all set.

Cheers!

Permalink Reply by John Elliott on April 7, 2011 at 9:55pm

Francis,

Had you asked if the data for the Greens could be made available we would have done that rather than you having to dig it all out.

Permalink Reply by David Balson on April 11, 2011 at 11:26am

Francis, sorry for the delay in responding..to answer your questions:

1. Yes, there are t-shirts available and we are seeking $10 each for 20 of them

2.As you have seen from other postings, approval of expenses must come from the national campaign (Gary)

3. Our analysis was a rough one...identifying those polls where there were several or more contiguous with low turn out and some cases a close race, we then identified the streets where the polls were located and determined "public" locations that would likely be servicing them

4. Re donations (even in kind if there are transfers of money [not time]), I understand the rules to say that they must be approved by the national campaign as well and also be supported by names and addresses of any contributors of money towards these donations...and these all count towards a ceiling set for each riding..thus care and planning must be taken.

Cheers David

Permalink Reply by David Balson on April 11, 2011 at 3:30pm

An update on the T-shirts...

Just picked them up..look great..stood in the parking lot of Vanier Loblaws by my car for 1 minute and two separate people passed me 50 metres away..both of them stopped, looked at the shirt, went out of their way (opposite direction of where they were going), approached me and asked what it was all about...small initial sample but a pleasure to see the exact reaction that was desired. Both were also interested and listened carefully to the narrative.

Cheers David

Permalink Reply by M Kumar on April 30, 2011 at 6:49am

Larry,

Is there a volunteer group I can join? I know there is just two days, but I would be happy to do what I can. Thanks.

Permalink Reply by Fran Balson-Nugent on April 30, 2011 at 12:31pm

Kumar,

We are going to be at Bayshore at 2;30 today. Just look for the yellow shirts and ask for fran,David or Joy.

We are also going to the Market at 5:00 in front of the CTV building. Again look for the yellow shirts and ask for any of us.

Thanks,

fran Balson-Nugent

Permalink Reply by Glenn Brauen on April 13, 2011 at 10:24pm

Hi John (and Francis)

I like what you are doing with the poll by poll analysis. Francis's ideas to look for low turnout or close ridings in the 2008 data to target focus areas seems pretty sound.

I spent some time today and pulled the poll by poll analysis data into a GIS along with the polling division boundaries from Elections Canada. We could colour this based on a number of variables.

The attached kml file, which you could download and open in Google Earth, colours the polygons according to a "cushion" variable I defined based on what was in John's spreadsheet. It's a two-ended ratio.

If Baird won the poll, the cushion is the difference between Baird and the closest challenger divided by the total non-conservative vote (and coloured in the blue range). If another candidate (mostly Pratt) won the poll, the cushion is the difference between that candidate and next highest divided by the conservative vote (and coloured in the red range).

So the cushion variable tries to represent how safe the winner in each poll is, using the ratio of the difference against the votes not taken by the winner, breaking all of the votes into two piles: conservative and non-conservative. To separate them, the cushion values are negative if Baird won the poll and positive otherwise.

Voter turnout does not factor into this colour scheme but another map could look at that.

So, as Francis has been experimenting with, if a map (or two) like this would be useful, we could make it available for use with Google Maps. Or we could simply use it for our own analysis.

I can make some time available to play with ideas for looking at this data if you have suggestions.

Attachments: