There are many reasons why pre-election polls do not reveal the facts.
And bear in mind that these are survey polls -- a pollster does not talk to every voter but only a small - a very small sampling - and then guesses how that relates to the whole. Although polls do not say this, and professional journalism would make a point of it - polls are only accurate 19 times out of 20 so one in each 20 polls on the average is sheer bunkum.
1. Pollsters only contact those who have landlines - this automatically excludes many young voters
2. People lie to pollsters - afraid of retribution for stating opposition or minority sentiments. Or they may tell the posters truthfully whom they support but then in the polling booth they misrepresent their sentiments by engaging in strategic voting.
Misrepresentation is how strategic voting was described in old writings ( such as The UFA Circular No. 12 "Proportional Representation", reprinted in the Grain Growers Guide, Oct. 26, 1913).
How can an election work properly if voters do not vote for who they want elected? Conversely, how can democracy work when politicians say one thing in election campaigns and then do something else once elected?
In 1913, activists saw that votes should not change how they vote (strategic voting) but instead should work to change the game, to make it truly representative.
Nowadays both polling and Strategic Voting are seen as cool, novel gadgets that provide a way to game the system. But better in my opinion to settle down and adopt STV to allow true representation, without cheating/misrepresentation/placing your vote as if it is a shot in the dark. STV provides true representation but without extreme parties possibly given undue power through FPTP or party-list proportional.
3. People change their minds between the poll and the election. Some of this is due to strategic voting described above or just basic change of mind - Schmeer's saying in the last days of the campaign that the NDP would raise GST may have shifted some away from the NDP or the insanity of his claim may have shifted some away from him.
4. Pollsters may predicate the question with set-up and thus slant the results. They may do this to suit the body paying for the survey - what group pushing a cause wants to hear that their cause is unpopular?
(How to do this is shown on one episode of Yes Prime Minister for example. And of course an episode of "This Hour has 22 minutes" where the choices that people have to respond to "Are you racist some of the time, most of the time or all of the time?" and then the pollster goes away to report that everyone he contacted admitted to being racist.
It is easy to construct a set-up == do you like to pay taxes... would you prefer to pay less taxes ... would you vote for a party that promises to lower taxes?
OR do you worry about your retirement ... would you vote for a party that promises to ensure that you live in dignity in your old age, even if that meant higher taxes?)
And anyways there is an obvious problem trying to place the sentiments of more than 12M voters over the 338 ridings in Canada. To know the result of the election, it comes down to district elections electing MPs one by one. At least until we get STV or other form of proportional representation.
Strangely the Wikipedia page for the 2019 Alberta election even lists the election as a poll with a sample size of 1.9M. This shows a misunderstanding of the process.
Pre-election polls are intended to show how people will vote. All the wide human interests and individual characteristics are put aside to just nail down how many are expected to vote each way.
The election is not a poll, a guess about what will happen or how people will vote - it is the thing the polls were guessing at - it is how people voted. It may be wrong about every other thing about people - it does not concern itself with age, income, the car they drive, whether they are happy or not -- it is only a measure of how people vote that day.
And that is enough. It is how we choose governments. It is all.
And that makes it so important to have a system where what happens that one day has the best chance of providing a representative government that will represent people's sentiments.
An old widely-distributed print article said it well:
"The aim of our electoral machinery should be to secure justice and freedom [by ensuring] to all citizens, whether in a majority or a minority, that they may be fairly represented in parliament." (Monitor News, Oct. 3, 1919)
Energy spent analyzing polls or devising strategic Voting would be better spent in my opinion working to improve election process to better reflect voters sentiments and decrease regionalism through STV.
How does STV work?
In short:
Candidates run in multiple-member districts, electing usually 3 to 7 members in a district.
Each voter casts a single vote but marks back-up preferences on the ballot.
The ballots in each district are counted and sorted. Some votes elect one or more candidates on account of specific support.
The surplus votes not needed by the winners are transferred according to the back-up preferences on the ballots. Some go to candidates of the same party to further create proportionality of representation. Some go to other candidates.
These votes plus others transferred from eliminated low-ranking candidates are used to elect the best of the rest through formed consensus among the remaining voters.
Simple as that....
The best thing about STV is it was used for 30 years to elect MLAs in Edmonton and Calgary. It does not have to be invented. It is sitting right there in the can. Just open and heat.
During that time in each election each city elected a mixture of representatives from different parties, based on the sentiment of voters. Unlike first past the post where a party dominates almost all the seats in a city each time, under STV Edmonton elected five to seven MLAs each time and they were spread over three or four parties - Liberals, Conservatives, Social Credit-ers and CCF-ers. No extremists were elected. All substantial minorities in each city were represented each election.
That was good and it could happen again. Just open and heat.
Comments