Is the way we elect fair?

Athlone_Bhoy

Just wondering what other people think of this.

Say myself and other people go for elections results are

Paul 8000
John 7500
Aidan 7000
James 6800
Luke 6000
Peter 5800
Athlone_Bhoy 15000
The quota is 10000 so I have 5000 votes to give away.

I give each candidate a few hundred however me and Peter are good friends so I give him 4300 which makes him elected.

Does this system seem like it's fair or a democracy? After all the people wanted Paul in but actually got Peter?

Degag

That's not how it works.

Athlone_Bhoy

I shall wait on your reply to explain how it works!

Degag

You can't just "give" someone votes!

Athlone_Bhoy

Then why aren't all people given an equal amount of votes?

I seen some people getting one hundred and some getting one thousand.

SF for example weren't getting the transfers from FG.

Degag

Also, by my calculations, the quota would actually have been 7013, so John, Paul and Athlone Boy would be elected and Aidan would only be 13 off the qouta... so the people got Paul afterall!

Degag

Athlone_Bhoy wrote: »

Then why aren't all people given an equal amount of votes?

I seen some people getting one hundred and some getting one thousand.

SF for example weren't getting the transfers from FG.

Because it's proportionate on the ammount of people who voted for each canditate. If 3000 out of the 15000 who voted for Athlone Boy gave Peter their number 2, then he would get 1000 from your surplus papers. (ie. 15000/3000 = 5. 5000 surplus devided by 5 = 1000)

.BrianJM

Athlone_Bhoy wrote: »

I shall wait on your reply to explain how it works!

You could try http://www.rte.ie/news/election2011/mechanicsprstv.html

...and there are several threads on the topic.

Athlone_Bhoy

Only two get seats.

Anyways not the point, the point is people can be elected because of transfers but really their not actually the person they wanted elected.

Am I right?

I've had a few drinks so sorry

Degag

Athlone_Bhoy wrote: »

Only two get seats.

Anyways not the point, the point is people can be elected because of transfers but really their not actually the person they wanted elected.

Am I right?

I've had a few drinks so sorry

Only 2 seats now? You can't make up the rules as you go along!:)

In any case, in the case of surplus votes being awarded - it may not be the absolute fairest thing in the world - but it is done proportionately... so....

Edit to say: As Auld Yin said there are several threads on the topic, and the way in which the votes are distributed is certainly alot more technical than i can explain to you, but i think pretty much everyone agrees that it is the best way.

eagle eye

It is a terribly unfair system. If you went by first preference % votes in this election the seats would be split as follows

Fine Gael - 60 seats
Labour - 32 seats
Fianna Fail - 29 seats
Sinn Fein - 17 seats
Green Party 3 seats
Socialist Party - 2 seats
People before Profit - 2 seats
Independents - 21 seats

I do realise thats a bit off with Independents but not that much.

Degag

I think that you can argue the fact that it may not be the fairest system if you want, but within the system, the way the votes are counted and distributed is probably as fair as you will get.

Zubeneschamali

eagle eye wrote: »

It is a terribly unfair system. If you went by first preference % votes in this election the seats would be split as follows

If it went by first preferences only, I wouldn't have voted for the Green party first.

You can't look at the % first preference and pretend that's what people "really" wanted, as that's not the system people used to decide how to vote.

cavedave

By Arrow's Theorem no method of voting can be fair

takun

cavedave wrote: »

By Arrow's Theorem no method of voting can be fair

Woah! I'm sorta glad I didn't read that before voting, I'd have been frozen to the spot for fear that I was the pivotal voter.

That's not easy reading exactly, but fascinating. Thanks.

ixtlan

It does seem likely that we will have a list system in place for the next general election, assuming a referendum passes. I think this will happen since all parties are committed to the removal of the Senate, and this will be it's replacement.

This will be quite interesting... and there will be much discussion about how it would operate. I suspect in the end it will be a minor change, maybe 20 seats, which won't benefit the small parties the way it might have... meaning that if the Green Party gets 2% it still probably won't be enough to get a seat.

ix.

zig

Zubeneschamali wrote: »

If it went by first preferences only, I wouldn't have voted for the Green party first.

You can't look at the % first preference and pretend that's what people "really" wanted, as that's not the system people used to decide how to vote.

ok fair enough we chose the system we have, no problem, but just for arguments sake, why cant you look at the % of first preference and say thats what people really wanted? Alot of people vote with only the first preference in mind, not number 2/3/4 whatever.

Also , presuming by your post you gave Greens number 1, why wouldnt you have if it was first preference only?

ixtlan

zig wrote: »

ok fair enough we chose the system we have, no problem, but just for arguments sake, why cant you look at the % of first preference and say thats what people really wanted? Alot of people vote with only the first preference in mind, not number 2/3/4 whatever.

Also , presuming by your post you gave Greens number 1, why wouldnt you have if it was first preference only?

Am I missing something here? The system we have is...

Vote 1st preference only... that's the candidate I really want. If I can't have them elected I don't care about any choice of the others.
or
Vote 1st and more preferences. I really want the first person. However if that person does not have enough votes to be elected, then I'll take number 2, and then 3 and then 4.

There seems to be a suggestion that transfers are happening against the wishes of the voter. They are not. Any transfers that happen do so at the preference of that voter. So that is what they really want.

Ix.

Zubeneschamali

zig wrote: »

Also , presuming by your post you gave Greens number 1, why wouldnt you have if it was first preference only?

Because the Green candidate was a no-hoper, and it would have been a waste.

If we used a 1st preference only, we'd only have 2 parties.

ixtlan

Zubeneschamali wrote: »

Because the Green candidate was a no-hoper, and it would have been a waste.

If we used a 1st preference only, we'd only have 2 parties.

Indeed. Proportional Representation allows people to vote for anyone knowing that even if that candidate is eliminated the vote transfers to the next preference.

If you could only give one vote, you would really have to vote strategically, working out which candidates were likely to get elected.

Ix.

Respublica

If we used a 1st preference only, we'd only have 2 parties.

Unless we used list-PR, in which case there would be many parties.

Because the Green candidate was a no-hoper, and it would have been a waste.

It wouldn't have been a waste if we used a more proportional voting system than is currently in place.

If you could only give one vote, you would really have to vote strategically, working out which candidates were likely to get elected.

Again, "first past the post" in single seat constituencies is not the only voting system in which there are only first preferences. It is not even the most common one.

antomack

Not sure if it makes any real sense mathematically/statistically but would it not make sense to weight the preferences and determine an overall vote 'count' from all ballot papers for each candidate and then the top n candidates depending on number of seats are deemed elected.

So for each ballot paper each candidate is 'scored' according to their preference as below
Number of Candidates - Preference specified + 1

Example
Candidate 1 gets a 1st preference vote in a constituency with 10 candidates which equates to a score of 10 (10 - 1 + 1)
Candidate 5 gets a 2nd preference vote which equates to a score of 9 (10 - 2 + 1)
And so on for each ballot paper and candidate with a preference specified.


I'm sure there's a load of holes in the theory but it seems to me that the constituency's overall preferences would then be taken into account with the top n overall preferred candidates elected rather than which candidates meet the quota first.

It should lead to a simplified count system where there will only be one detailed count required and no need to worry about surpluses and transfers. It should be simple to develop a database application for the count process to allow count staff carry out a data entry process from each ballot paper where they enter the preference specified for each candidate and the application does all the calculations in the background. The system is so straight forward that each candidate can have their own version of the application for their representatives to register the votes for their own verification of the results.

Respublica

antomack,

What you've done here is you've reinvented a system called the "Borda count". (Look it up on Wikipedia)

The system has it's advocates but there are a few major flaws. For example:

(1) If the Borda count is used in multiseat constituencies (as you seem to be suggesting) the result won't be proportional representation. (There is a much more complicated version designed for multi-seat constituencies called the "quota Borda system")

(2) The Borda count is highly vulnerable to tactical voting. Basically voters who understand the system wouldn't vote honestly but would, for example, give "1" to their first preference and "16" to their real second preference in order to protect their favourite.

(3) The winner is affected dramatically by the number of candidates who happen to take part in the race. This makes the result semi-random.

(4) The biggest flaw with the Borda count is that the "score" it gives to each candidate is totally arbitrary. In your example the assumption is that the voter likes his second preference 90% as much as he likes his first preference. It's not reasonable to assume that. In reality there might be a big gap between the first and second preferences in the mind of the individual voter--or a small gap, you just can't know.

There is a genuine problem with STV that it does not take proper account of all of the lower preferences. But the Borda count is much worse. A much better solution is to use the Condorcet method for presidential elections, and CPO-STV for parliamentary elections.

[End of rant ]

calm_bull

cavedave wrote: »

By Arrow's Theorem no method of voting can be fair

i agree but it can be democratic and i dont think the current system meets that criteria>

cavedave

calm_bull

i agree but it can be democratic and i dont think the current system meets that criteria

How do you think it should be democratic? AFAIK the Irish republic which includes Corporatist elements is not democratic.

nesf

Respublica wrote: »

There is a genuine problem with STV that it does not take proper account of all of the lower preferences. But the Borda count is much worse. A much better solution is to use the Condorcet method for presidential elections, and CPO-STV for parliamentary elections.

[End of rant ]

CPO-STV needs computers to be involved in the count (though not necessarily electronic voting). Not a simple or cheap switch to make.

Voipjunkie

I think the op doesn't understand the prstv system that is his main problem.

The quota is not an abritary figure going on his first and second post

Total vote is 63,300 seats 2 so quota would be 2+1 divided into 63,300 = 21,100 + 1

So on his figures nobody would have reached the quota and bottom candidate would be eliminated and his votes distributed based on his second preferences. If someone is over the quota then the excess amount of votes are randomly selected and those votes next available preference are distributed to the available candidates.

Candidates have no say in how their votes are distributed that is purely how voters voted themselves.