Author Topic: Improving democracy through math  (Read 2180 times)

0 Members and 1 Guest are viewing this topic.

Offline daqqTopic starter

  • Super Contributor
  • ***
  • Posts: 2321
  • Country: sk
    • My site
Improving democracy through math
« on: November 06, 2020, 09:16:55 am »
While the US elections are the hot topic now, this topic annoyed me well before that in my own country, where I regularly get to strategically decide to vote the lesser evil. There are alternative voting systems that are more mathematically and ethically sound. Enjoy the videos. It's a great channel by the way:







Complete list here: https://www.youtube.com/playlist?list=PLNCHVwtpeBY4mybPkHEnRxSOb7FQ2vF9c
Believe it or not, pointy haired people do exist!
+++Divide By Cucumber Error. Please Reinstall Universe And Reboot +++
 
The following users thanked this post: pardo-bsso

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: Improving democracy through math
« Reply #1 on: November 06, 2020, 05:02:43 pm »
I offer a simple mathematical proof that you cannot improve democracy:

If you vote for lesser evil and the lesser evil gets in, it will know damn well not to allow you to vote for anything else ;)
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Improving democracy through math
« Reply #2 on: November 06, 2020, 05:35:57 pm »
Election design and voting systems are examples of what should be technical problems, but become political problems when the proposed technical solutions affect different political factions differently.  The first technical article I read about election design (Scientific American magazine, sometime before 1980) defined a bad system thusly:
In a three-person race between A, B, and C:  if A beats B and A beats C in two-person races, but B wins the three-person race.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15795
  • Country: fr
Re: Improving democracy through math
« Reply #3 on: November 06, 2020, 06:01:33 pm »
There is indeed a fundamental known problem with the voting systems we use in almost all democracies these days. To begin understanding what is wrong, you can start with the Condorcet paradox (and derive corollaries from there):
https://en.wikipedia.org/wiki/Condorcet_paradox

We sure knew something was fishy, a long time ago.

Even basic survey techniques these days (unless they are willingly biased) are designed to avoid this problem, and often repeat related questions in different ways and contexts to make sure the answers are consistent. Voting the way we do to select candidates is one of the worst possible ways of eventually selecting the preferred candidate for a majority of people.

Now "less evil" doesn't mean much in reality. So that's yet another debate, but I do agree it worsens the voting problem as many people vote this way (each with their own - or through propaganda of some form - notion of "less evil"), which by nature can't get you anything representing the people in general. So that's just adding to the voting paradox.

Which gets us to another issue: is representative democracy really democracy? I for one am not so sure.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Improving democracy through math
« Reply #4 on: November 06, 2020, 06:34:19 pm »
Which gets us to another issue: is representative democracy really democracy? I for one am not so sure.

If plurality (first past the post) voting is used, then no democracy represents the voters.

How much better are the people represented when there are two choices instead of one?
 

Offline asmi

  • Super Contributor
  • ***
  • Posts: 2860
  • Country: ca
Re: Improving democracy through math
« Reply #5 on: November 06, 2020, 09:31:17 pm »
I think the fundamental problem of democracy is that you get to vote for a person or people, and not idea or plan or ideology or whatever. So you have to accept "the whole package" that is a person, as well as a good chance that he won't actually do what he said he would. And this leads to combinatorial explosion of possible permutations of ideas/plans/ideologies, and also practically guarantees that the vast majority of people won't find a single candidate they can fully support.
Now, if promising something and not delivering would become a criminal offense with long prison time - that would change quite a bit. But no politician would ever implement something like that, because he knows one day he might be charged under this article...

Offline coppice

  • Super Contributor
  • ***
  • Posts: 10031
  • Country: gb
Re: Improving democracy through math
« Reply #6 on: November 06, 2020, 09:36:03 pm »
Look up Arrow's impossibility theorem. It doesn't really apply to a case like the US, where there are only 2 effective choices (i.e. ones which have any change of winning) in most elections, but it applies to most countries.
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Improving democracy through math
« Reply #7 on: November 06, 2020, 10:03:33 pm »
“First past the post” vs. proportional representation in the legislature is an example of a legitimate political question, where the technical analysis shows that the former inhibits small parties and promotes two-party elections.  Politically, one can argue the fairness of multiple smaller parties vs. the stability of two parties.  The question of “ranked voting”, now in limited use in the US, is more of a technical question.
« Last Edit: November 06, 2020, 10:26:03 pm by TimFox »
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Improving democracy through math
« Reply #8 on: November 06, 2020, 10:19:43 pm »
I do wish that I could vote for a first and second choice, either that or use the primaries to narrow down between two candidates for each position in a given race. Often I have voted for someone I am not real fond of, because I view them as better than a horrible one that I know will be popular, when the one I'd really like is a bit of a long shot.
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: Improving democracy through math
« Reply #9 on: November 06, 2020, 10:32:16 pm »
At present, the only US State with ranked-choice voting is Maine (first used in 2018), but several smaller jurisdictions  (counties and cities) use it for local elections.  New York City starts using it next year.
 

Offline VK3DRB

  • Super Contributor
  • ***
  • Posts: 2272
  • Country: au
Re: Improving democracy through math
« Reply #10 on: November 06, 2020, 11:58:42 pm »
The videos miss some things which distort democracy badly.

No matter the system, lying to get votes is not democracy. We had a Labor politician named Steve Bracks who told the electorate one week before voting day that there will be NO tolls on new major freeway - a major issue at the time. Almost immediately after winning the election, he reversed his decision and there were tolls.

No matter the system, deception to get votes is not democracy. Trump stood in front of a church recently holding a bible, whilst he got the cops beat up pedestrians around him to clear the way. That stunt helped him get 80% of evangelical "Christians" to vote for him.

No matter the system, those with money and power tend to get a bigger vote than a poor person. Rupert Murdoch has a much bigger say than a homeless person living in poverty.

No matter the system, any democracy is better than living in a communist country run by a ruling elite that brainwashes its population into submission and imprisons anyone whose beliefs are contrary to those of "the state".
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: Improving democracy through math
« Reply #11 on: November 07, 2020, 12:38:33 am »
California now enjoys "top two" voting.  Assume 3 candidates, 2 from brand D and 1 from brand R in a primary election.  Both brand D candidates win over brand R in the primary and are the only candidates running in the general election.  Brand R voters have nobody to vote for.

There are similar atrocities with gerrymandering and other practices but the "top two" must lead the list.

One thing we haven't quite signed up for is "electors go to the winner of the popular vote" such that it makes no difference who wins or loses the California vote, the 52 electors go to the winner of the popular vote.  Why vote at all?  Just wait to see what everybody else thinks and have the electors vote for that candidate!  Why even have electors, the result is determined elsewhere.

 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Improving democracy through math
« Reply #12 on: November 07, 2020, 02:30:50 am »
California now enjoys "top two" voting.  Assume 3 candidates, 2 from brand D and 1 from brand R in a primary election.  Both brand D candidates win over brand R in the primary and are the only candidates running in the general election.  Brand R voters have nobody to vote for.

There are similar atrocities with gerrymandering and other practices but the "top two" must lead the list.

One thing we haven't quite signed up for is "electors go to the winner of the popular vote" such that it makes no difference who wins or loses the California vote, the 52 electors go to the winner of the popular vote.  Why vote at all?  Just wait to see what everybody else thinks and have the electors vote for that candidate!  Why even have electors, the result is determined elsewhere.

They have that in WA too now, this last election both lieutenant governor candidates were the same party. I don't particularly like this, even if it ends up being the party I lean towards.
 

Offline Simon

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Improving democracy through math
« Reply #13 on: November 07, 2020, 10:49:29 am »
This topic is dommed to go sour. If we can stick to the theory it may be OK, if it goes tribal it will be locked.

Interestingly the BBC made a little video about the American electoral system pointing out how you can win with less than the majority of votes. What they carefully did was avoid naming the system, First past the post because here in the UK the debate around FPTP and some sort of proportional system never went away. The current system was sort of OK when there were only two parties but even them can introduce some distortions. Ultimately our "democracy" hinges on 650 ish votes, the rest of the millions of votes do not count.
 

Online EEVblog

  • Administrator
  • *****
  • Posts: 39025
  • Country: au
    • EEVblog
Re: Improving democracy through math
« Reply #14 on: November 07, 2020, 11:02:41 am »
This topic is dommed to go sour. If we can stick to the theory it may be OK, if it goes tribal it will be locked.

It's doomed!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf