Clear Full Forecast

Is There A Better Way To Elect Our Reps?

By 250 News

Thursday, November 25, 2010 04:00 AM

Presenter, Dr. Dan Ryan, at UNBC's 'Cafe Scientifique' at Cafe Voltaire in Books & Co.

Prince George, B.C.- While he believes our electoral system is a good one when it's working as it should, the new Dean of Science and Management at UNBC says declining voter turn-out should have us looking at alternatives.

 

Speaking to a small crowd at Cafe Voltaire last night, Dr. Dan Ryan offered up what he admitted was a radical idea, but one, he thinks, has merits. Ryan suggested that turning from our current census system to one where a random sample of voters in each riding was selected to cast ballots could be less expensive and, perhaps, an even more accurate way of electing our federal representatives.

Ryan said the census system -- where every person eligible to vote -- can cast a ballot is an excellent one when everyone votes. But he pointed out that only 59-percent of eligible voters cast ballots in the 2008 election. "So this becomes a 'self-selected sample' (of voters) and the challenge with a self-selected sample is it may not be representative," said the Dean. "And this is what the job of the candidates has turned into -- it's not to speak to everybody in your riding, but it's to bias the sample -- it's to get the demographic that's going to vote for you out (to vote)."

Ryan said one solution would be to go to a mandatory vote like Australia. He said it would never be a complete census -- closer to 90- to 95-percent -- but much more representative of what the people want than what's happening in Canada with sucha small, self-selected group electing the government.

The statistician said a better way -- less costly and with a more accurate result -- would be to select a random sample of voters from those eligible in each riding. In a riding of 90-thousand, you would need 10-thousand people to cast ballots to get a result that would accurately represent what 95-percent of the people in the riding wanted, with a margin of error of one-percent. Ryan offered that the numbers chosen for the random sample would have to be fine-tuned based on the number of candidates running in the riding and the degree of accuracy and margin of error, but said, "The key point is that for less than 20-percent (of the population voting), we can get a more accurate sample than what we have at the moment."

Ryan easily admitted to the crowd that his 'Random Sample' system would probably be a hard sell because not everyone would get to vote in every election and "that would be a challenge because it's not the way we were brought up." And he said the political parties would hate it because they'd hold a rally and not know how many of their supporters would be able to actually cast a ballot. But Ryan pointed out it would certainly force candidates to speak to all demographics in their riding, including the 'disillusioned' -- a few of which would now be required to cast ballots.

And while he conceded his idea would require more faith than our current system in the people running it, he said with a well set-up polling company, it could be done. "'Would you believe them?' I don't know, a lot of people wouldn't," said Ryan. "As a statistician, I'm quite comfortable with (the idea) because I've seen it work in large samples."

 


Previous Story - Next Story



Return to Home
NetBistro

Comments

Sometimes I do wonder. The current system, in which everyone is eligible to cast a vote, gives 100% accuracy on the wishes of the voters. Even those who choose not to cast a ballot are legitimately expressing an opinion by not voting, and his proposed statistical voting scheme would not capture that. The major defect in his proposal is that it is 99% accurate, but we have all seen elections turn on far less than 1% of the votes, sometimes less than 10. I think this is not the area in which there should be attempts to save money. There should be no price on democracy.
I think we should get rid of all the parties. Elect someone to represent us not because of party lines. When they all get to Victoria or where ever government may be, the elect a leader, chairperson, whatever name you wish to give them, and get on with things. They mess up and not represent those who they are suppose to be reperesenting, they get fired.
first,,, you have coruption from day one,you have power ,greed all stem from politics.this is an old machine that now is really showing its demise,you cant fix a political engine this bad sooo you get a new one !! and thats what has to happen ,as intelligent as we think we are ,we all fall into the old pattern,fix a little here and so on,real people with real honest and true ideas will have to take place to make a system that is truly transparent and willing to work for the people.untill then ,, get used to bieng run by criminals and immature fools.just look around ,not hard to see .is it ?
To back up his theory, he should use actual knowledge we have of the system. He may have done that in his presentation.

There are tons of polling results at varying times during the campaigns in many countries. As I see it, this system is no more than another poll.

So, what do we know about the accuracy of polls when measured against the actual election results?

I have always taken the position that ammonra takes, a non-vote is a position that has been voiced. Just as with a vote, we do not know why that person did not vote, just as we do not know why the voter voted the way he/she did or even how they voted.

This might help to start the search for information of how reliable an official polling system might be.

http://www.huffingtonpost.com/2010/11/03/how-did-the-polls-do_n_778216.html
Why don't we just have "None of the Above" printed at the bottom of every election ballot below the Candidates' names, with a place to put your "x" beside it, and countable as a vote?

That way there is no question as to the extent of the electorate being too apathetic to show up to vote on election day.

Or that someone is spoiling their ballot for some purpose other than mischief.

One thing this would do is provide an onus on the Candidates, all of them, to show that they deserve your vote. Or "None of the Above" just might get 'elected'.

We could follow that up with a provision that in such an instance "None of the Above" Candidates could run again in a follow-up election. They'd have to wait until the next general election, or a by-election, to do that.

Right now, too many people likely do not vote because they are not impressed with ANY of the Candidates running, or their Party, or its Leader. They are not going to show up and vote at trying to pick the "best of a bad lot", but would quite likely be willing to vote if they could meaningfully express their disapproval if no one running is offering what they want.
Socredible: This can be done now, Just put an X in every Box. That will show them why you spoiled your Ballot .
Forget it, I pay $100 grand plus in taxes to get my vote every four years. and your going to let some socialistic visionary turn it into a lottery to voice my opinion.

You can pretty much take that idea and place it where the sun don't shine
Well, "He Spoke", I assume this means you're happy with the current system, which I'm not. As for Gus' take on things ala Huffington, just notice that all of the polls taken recently for the midterm elections consisted of 3,000 responses or less, usually a lot less. The UNBC fellow is suggesting 10,000 responses, with accuracy a function of the number of responses so that once you get past a certain number, the sample becomes an extremely accurate representation of the total population. In theory, there's no reason why this idea wouldn't work.
As for the notion that a non-vote says as much as a vote, that's nonsense, since non-votes don't get counted and mean squat in a first-past-the-post election system.
Do I really think the proposed system will ever see the light of day and come into use? Not a chance - anything that threatens the status quo and hints at truely representational democracy will be ridiculed and quashed by the powers that be and their sycophantic mouth-pieces in the media.
"Just put an X in every Box."

That's not much fun!

Put an X in all boxes but one.

Put a checkmark in that one box.

That may at least cause some discussion by the scruteneers, especially if the checkmark is for the party expected to come in second, but there is a chance they could come in first .... :-)
"The UNBC fellow is suggesting 10,000 responses, with accuracy a function of the number of responses"

Hmmm ...

Do you mean to tell me that straight numbers are a better indicator of representational accuracy than how well the cohorts of the population segments have been selected?

One has to consider regional issues, ethnicity, age, income, educational background, etc. etc.

Has anyone ever done any "ground truthing" where a larger scale predictive model has been established with the population characteristic input, issues input, vote results predicted, and then tested against actual votes?

You wrote: " hints at truely representational democracy"

Seems you might live in a virtual world of research.
It's really easy, folks. Just elect the opposition. They always have the answers to every problem all the time.
Wow, most of the comments on this article renew my faith in the ability for online comments to be more than drivel.

Sorry, Dr. Ryan, but "accurate" statistics cannot equal the higher principles of democracy. They can influence elections, though, as years of polling has shown.
Election is like a theatre, where all adults can play a role in the ritual (if they want to) and determine the ending of the story. In a statistical voting system, we don't have this broad mass ritual.

And whoever controls the selection of the sample could be in control of the ending of the story. The machinery of election "being unbiased" is as important as the precision of the results. Maybe that is why we haven't adopted voting by phones yet.