Calgary Centre Federal By-election - Nov 26 (UPDATE - Crockatt wins for CPC)

[Rerun]

Originally Posted by longsuffering

So far you're the only person who believes that polling results could have an "undue influence" on voters. Did I miss the post where you had data to support that assertion?

There are plenty of studies that support this theory.

http://en.wikipedia.org/wiki/Opinion...fect_on_voters

http://www.kaaj.com/psych/abstract/pollsabstract.html

http://www.oireachtas.ie/parliament/...olling_web.pdf

[Rathji]

Did you seriously just link Wikipedia as a study?

[SebC]

Originally Posted by Rerun

should your voting preference be determined by how you believe your neighbor plans to vote?

No, and that's why we need electoral reform. Because under the current system, you can be punished (in that you get an outcome that is less desirable to you) for voting without taking other voters into account.

Polling helps our first-past-the-post system function more like a runoff system, where you actually need a majority to get elected. We should have a system where voting with your head and voting with your heart are the same... but since we don't, strategic voting is the way to produce results that more closely reflect the will of the electorate, and the information provided by polling is a key part of making that possible.

[Rerun]

Originally Posted by Rathji

Did you seriously just link Wikipedia as a study?

It references the studies.

[longsuffering]

Originally Posted by Rerun

There are plenty of studies that support this theory.

http://en.wikipedia.org/wiki/Opinion...fect_on_voters

http://www.kaaj.com/psych/abstract/pollsabstract.html

http://www.oireachtas.ie/parliament/...olling_web.pdf

Thanks for the links. I appreciate you taking the time to find them.

My next question would be - now that we've established that polling may influence voters, have be established that the influence is a negative one?

It seems that polls are just as likely to influence voters to cast their ballot for the front runner, as it is to influence them to vote strategically.

[Rerun]

Originally Posted by longsuffering

Thanks for the links. I appreciate you taking the time to find them.

My next question would be - now that we've established that polling may influence voters, have be established that the influence is a negative one?

It seems that polls are just as likely to influence voters to cast their ballot for the front runner, as it is to influence them to vote strategically.

In my opinion it is. The bandwaggon effect may result in a majority government where without polling the result would probably be a minority government. The strategic or tactical voting effect can result in a 3rd place party being virtually wiped off the face of the electoral map because a lot of people who would normally support the 3rd place party move their vote to the 2nd place party in the attempt to defeat or reduce the number of seats the the 1st place party wins.

Both effects are in play at the same time.... this may have contributed to Harper's majority and the large number of NDP wins.... all to the detriment of the Liberal Party.... (God, I can't believe I am saying this!!!)

[Rathji]

Plus, it really screwed the WRP over this past spring!

[Iowa_Flames_Fan]

I have no issue with polls; they should be interpreted with caution, particularly where it's a multi-way race and there are demographic issues, such as no cell phones. We don't really have better options yet. Online polling is starting to be a viable alternative, and a number of online pollsters got results that were at least no worse than those of traditional pollsters in the recent US election.

On balance, if we are going to have polling, my wish is that it would be of higher quality. If someone posted a poll in the recent US Politics thread which was land-line only and only spent one day in the field, and had a same size under 500, someone would have noted that it was methodologically much poorer than the standard polling outfit.

Here, that almost describes all of the polls. I understand there are logistical difficulties and cost barriers, but I still think it's a problem.

With that said, to me the polls are telling a pretty consistent story right now. Crockatt will benefit from the vote split, and will likely win with a plurality in the mid-to-high thirties. If her GOTV effort is superior to Locke's, she may hit 40%.

The wild card is undecided voters and leaners. I don't buy that undecided voters automatically break for the challenger--history shows that to be nonsense. Usually a voter who is undecided going into the last weekend is not going to vote. The difference here is that I suspect some voters are waiting and seeing, and will break to the strongest challenger.

At that point, the true gap between Locke and Crockatt will matter a lot. If its 5% or higher, he's in trouble. If it's more like 2 or 3 %, then it could be a coin toss for him at best.

Those are just my feelings. I'll be voting for Locke and hoping for the best, but I wouldn't bet on his chances. At this point his best hope is an influx of strategic voters.

[mariners_fever]

Just to give ReRun some credit, it has been documented that polling close to voting day can have effects.

A little different than actual polling, but early results from Connecticut (or Massachusetts, I can't fully remember) are often said to have affected western votes in the 1960 Presidential Election. Obscure, yes, but something fun to chew on.

[nik-]

Originally Posted by Iowa_Flames_Fan

I have no issue with polls

Statements that are a surprise to no one for $100 Alex

[edslunch]

Originally Posted by Rerun

Yes... but is the truth or a lie?... and should your voting preference be determined by how you believe your neighbor plans to vote?

I have to say that, sadly, almost all my voting since I moved here has been strategic. I voted PC for the first time to elect Clark over the Alliance. I joined the conservatives for a day to vote against Morten. I voted Nenshi in part for him and in part against McIvor. This time I'm thrilled to have two progressive candidates I like but I will choose the one who has the best chance of defeating the Conservatives.

Well I would prefer to vote for someone rather than against someone I think it's the current lot of progressives in Alberta that any outcome other than the Conservatives is a good step

[Iowa_Flames_Fan]

Originally Posted by mariners_fever

Just to give ReRun some credit, it has been documented that polling close to voting day can have effects.

A little different than actual polling, but early results from Connecticut (or Massachusetts, I can't fully remember) are often said to have affected western votes in the 1960 Presidential Election. Obscure, yes, but something fun to chew on.

It's definitely the case that polling affects vote intention, and even as I am a poll junkie I do recognize this issue. This is why people get so exercised about polls that they feel are inaccurate: they worry that through the so-called bandwagon effect, these polls become a self-fulfilling prophecy. See the "unskewed" crowd in the US for a good example.

To be honest, I am sympathetic to Rerun's point. But I don't see how you can lawfully curtail something like that. People WANT polls, and they especially want them (as SebC pointed out) when they are hoping to maximize the value of their own vote, which is a reasonable thing to want. As long as people want them, someone will be willing to produce them.

But there absolutely is a bandwagon effect. Subconsciously, people do not want to vote for a losing candidate.

[c.t.ner]

Alright, for those curious here is some easy to digest info on 1CC.

* Highlights/Comments - http://www.1calgarycentre.com/1calga...ll-highlights/
* Visual Dashboard - http://www.1calgarycentre.com/the-results-dashboard/

Now back to the regularly scheduled conversation.

[SebC]

Originally Posted by mariners_fever

Just to give ReRun some credit, it has been documented that polling close to voting day can have effects.

I think we all agree on that. What we disagree is on whether these effects are nefarious or beneficial.

[Rerun]

Originally Posted by SebC

I think we all agree on that. What we disagree is on whether these effects are nefarious or beneficial.

And what if polling results are completely out of wack from reality? Polls can be skewed to get the result you want.

At least if a person utters a lie you can rebut it... how do you do that to a poll. My faith in polling is currently at an all time low... for many of the reasons brought up in this thread.

Many voters however, take poll results as the gospel truth.... and vote accordingly.... to the detriment of some parties that don't deserve it and vice versa.

[Iowa_Flames_Fan]

Originally Posted by Rerun

And what if polling results are completely out of wack from reality? Polls can be skewed to get the result you want.

At least if a person utters a lie you can rebut it... how do you do that to a poll. My faith in polling is currently at an all time low... for many of the reasons brought up in this thread.

Many voters however, take poll results as the gospel truth.... and vote accordingly.... to the detriment of some parties that don't deserve it and vice versa.

Here's something that should restore your faith. American polls were given a rough ride this last election cycle. Response rates were at historic lows, in some cases lower than 10% (though they crept higher as election day neared). The bulk of pollsters were using robo-calls to land-lines only, for cost reasons. Pollsters were also modeling the electorate according to demographics, which some people thought was apt to skew the result toward democrats.

But in the end... the polls mostly got it right, especially at the state level. Non-cell-phone polls were on balance slightly biased toward republicans, but the effect was in most cases modest. A few pollsters had really bad results (Gallup, Rasmussen, Gravis and Mason-Dixon come to mind) but for the most part the polls did a very good job of handicapping the race--and the evidence (that Obama was winning) was there for anyone who cared to see it.

I don't buy that polls are very often "skewed," at least not deliberately. Nor do I buy that there are very many people who take them as gospel and "vote accordingly," whatever that would mean.

As for detriment to parties... sometimes the truth hurts. Right now I'm guessing Chris Turner would like to see a little less polling. That's just how the cookie crumbles.

[SebC]

Originally Posted by Rerun

And what if polling results are completely out of wack from reality?

Flawed polling is better than no polling at all. You also typically have multiple polls from different companies, so if one of them is cheating, chances are they'll get caught.

[Rerun]

Originally Posted by Iowa_Flames_Fan

I don't buy that polls are very often "skewed," at least not deliberately. Nor do I buy that there are very many people who take them as gospel and "vote accordingly," whatever that would mean.

I'm a little more pessimistic than you and believe that when it comes to politics people will answer a poll the way they feel will benefit the party they most support. And polls can be skewed by just who they ask the questions to and where they live.

As for "voting accordingly", I mean people who, because of the polling info, will either jump on the bandwagon to vote for the #1 party or others who utilize the polling info to determine the party they should support in order to vote strategically against the #1 party.

[Rerun]

Originally Posted by SebC

Flawed polling is better than no polling at all. You also typically have multiple polls from different companies, so if one of them is cheating, chances are they'll get caught.

What is your explanation of the polling during the last provincial election?

The polls were definitely flawed there. Would we have had a different result if the WRA wasn't deemed to be the front runner in the polls? Perhaps.

Would we have had a different result if the polling numbers actually reflected the actual election results? Perhaps.

[Rathji]

Originally Posted by Rerun

What is your explanation of the polling during the last provincial election?

The polls were definitely flawed there. Would we have had a different result if the WRA wasn't deemed to be the front runner in the polls? Perhaps.

Would we have had a different result if the polling numbers actually reflected the actual election results? Perhaps.

The polls were not flawed, they clearly showed that the WRP was in the lead. A lead which they lost because a whole bunch of moderate Albertans were scared (rightfully or not) about electing a party which seemed to be running a disproportionate number of 'displeasing' candidates.

You are 100% right that those poll results likely gave the PC party a larger majority than they deserved, because a rather large number of people in the province decided to vote strategically, to avoid the potential issues of electing the type of party that had these candidates.

This wasn't a polling problem, it was a problem with the WRP managing its image at the most critical part of the campaign.