Colorado: another nail in the elctral collg coffin

[MODU]

Reignman,

This thread is only 5 hours old.  People are still at work/school.  You can't rush to conclusions such as that.

As far as the electoral college goes, though it's original intent was the act as the "informed" buffer for each state (since the populous didn't have access to all the information we do these days), these days it acts as a "leveling field" of sorts for the nation.  If we went by a strict populous vote election, the Democratic candidates would normally win due to the liberal concentrations in California and New York, as well as in most of your large cities.  By using the census data to redistribute EVs across the nation, it allows for your smaller states like Alaska and North Dakota to fairly compete against California and New York.

[Fmr. Gov. NickG]

Question:

Let's say in 2012 afetr re-aportionment, Colorado has 10 EVs (or any even number will do)

Suppose the vote is say 54/46 - ie somebody has a fairly clear win.

A 5/5 EV split is closer than a 6/4 split.

Does this mean that even though you won a state by 8% you get exactly the same number of EVs as the guy who lost it...?

I think this is how it would work.   If Nader is able to get around 5% in CO this year, the state could split 4-4-1.

[Blue Rectangle]

I actually can vote on this and I'll definitely vote no.  So the results from a poll of likely voters are currently 0% in favor, 100% opposed.

The large majority of voters in CO are conservative in that they vote no on everything unless they have a very good reason to support something, so I think this proposal doesn’t stand a chance.

I also do not support any sort of CD-based system.  The last thing we need is to insert presidential politics into redistricting.  Reform the district drawing process and I would reconsider.

[AuH2O]

Few thoughts on this:

1) Horrible for the actual voters of Colorado, because it takes away any reason for EITHER candidate to really do much there. At best they get what, 1 more electoral vote?

2) Consequently, it will probably lose. Democrats figure it will use up some GOP money, which is true but not significant

3) I don't even like the CD thing- unless it was done nationally. States are mostly screwing themselves and also increasing the possibility of vote counting problems (extremely close counts are less likely over a state than a Congressional district)

4) Any Democrat that supports this should call for a similar system to be implemented in California, or admit to being a complete hypocrit

5) There could even be a little backfire potential here. If voters are educated, they aren't going to like it- and they're going to know who wanted it

[Fmr. Gov. NickG]

Out of curiosity, I ran the spreadsheet for using this system in all states in the 1996 election:

Clinton - 267
Dole - 224
Perot - 46
Nader - 1 (in CA)

Meaning this election would have been thrown into a Republican-dominated House and Senate!  Of course, I could definitely see the Nader elector and a couple Perot electors defecting to Clinton in this case...but this is still alarming.

Of course, this system in 1992 would also have thrown the election into the House, but Clinton would have almost certainly won the House vote that year.

[Nation]

So is the electoral college going to be used in Colorado this year or not? I'm a little confused.

[Fmr. Gov. NickG]

In 1980, under this system, Reagan avoids a House vote narrowly:

Reagan - 276
Carter - 223
Anderson - 37
Clark (Lib) - 2 (CA & NY)

[IowaLibert]

I also do not support any sort of CD-based system.  The last thing we need is to insert presidential politics into redistricting.  Reform the district drawing process and I would reconsider.

This is an excellent point. Fix gerrymandering first.

Posted from Iowa, where we have fixed gerrymandering.

[True Federalist (진정한 연방 주의자)]

The results depend a lot on how the left over fractions are divied up.  I just calculated assuming that if a candidate gets between k/n and (k+1)/n of a state's votes where n is the number of electoral votes the state has, then a candidate receives k electoral votes, with the left over electoral vote(s) a state has being received by whoever gets the most votes.  Under this method of "winner takes all fractions" the 2000 results would have been:
  Bush 274
  Gore: 261
  Nader: 3

[Andrew]

1) Horrible for the actual voters of Colorado, because it takes away any reason for EITHER candidate to really do much there. At best they get what, 1 more electoral vote?

I keep seeing this argument, but I don't really understand it.  Is there really some benefit to the voters from having Presidential candidates wandering around their state?

I guess the TV stations like being in battleground states, because they sell ads to the campaigns--but they can always sell that same time to grocery stores or something.  The money's the same.

[Fmr. Gov. NickG]

Having this system nationwide would actually be a HUGE benefit to the small states, because it would make it MUCH more likely that they election would be thrown into the House.  And when that happens, each state gets one vote, no matter the size.

[Posterity]

I also do not support any sort of CD-based system.  The last thing we need is to insert presidential politics into redistricting.  Reform the district drawing process and I would reconsider.

This is an excellent point. Fix gerrymandering first.

Posted from Iowa, where we have fixed gerrymandering.

How has Iowa fixed gerrymandering?

[sobo]

Looking at these results, if for some reason the states do switch to this system, we will need a better tiebraker than just sending it to the house, or it will go there everytime.  However canidates who clearly will not win could instruct their electors to vote for another canidate that they most agree with, for example, in the 1996 election, Perot decides he has more in common with Dole than Clinton, he could instruct his electors to vote for Dole.  Same thing with Gore and Nader in 2000.  This could partially solve the problem of third-party canidates splitting the vote.  I mean if in the current system were to come out 268-267-3 and whichever canidate the 3rd party guy supported was going to lose in the House, do you really think that he would not tell his electors to switch their votes.

[Andrew]

There is a lot of speculation here about how the electoral votes would be split up if this passes.  The text of the proposed amendment is here:  http://tinyurl.com/48u77

Here's the process:  

1.  Multiply the proportion of votes for each candidate by the number of electoral votes for Colorado (this year, 9).

2.  Round these numbers off to the nearest whole number.

3.  If the total is too big, take electoral votes away from the last candidate receiving electoral votes until the number is right.  If, after taking away all of the last candidate's votes, the number is still too big, take electoral votes away from the next-last candidate.

4.  If the total is too small, give any unallocated votes to the candidate receiving the most votes.

In 2000, it would have gone this way:

Bush:  .5075 x 8 = 4.06, rounded to 4.
Gore:  .4239 x 8 = 3.3912, rounded to 3.
Nader:  .0525 x 8 = 0.42, rounded to 0.
Browne, etc. all round to 0.

4+3+0=7, which is too few.  Bush won the state, so he gets the extra vote.

Bush 5, Gore 3, others 0.

[Fmr. Gov. NickG]

There is a lot of speculation here about how the electoral votes would be split up if this passes.  The text of the proposed amendment is here:  http://tinyurl.com/48u77

Here's the process:  

1.  Multiply the proportion of votes for each candidate by the number of electoral votes for Colorado (this year, 9).

2.  Round these numbers off to the nearest whole number.

3.  If the total is too big, take electoral votes away from the last candidate receiving electoral votes until the number is right.  If, after taking away all of the last candidate's votes, the number is still too big, take electoral votes away from the next-last candidate.

4.  If the total is too small, give any unallocated votes to the candidate receiving the most votes.

In 2000, it would have gone this way:

Bush:  .5075 x 8 = 4.06, rounded to 4.
Gore:  .4239 x 8 = 3.3912, rounded to 3.
Nader:  .0525 x 8 = 0.42, rounded to 0.
Browne, etc. all round to 0.

4+3+0=7, which is too few.  Bush won the state, so he gets the extra vote.

Bush 5, Gore 3, others 0.

This is a poor way to do this allocation.  It should really be done the same way the census allocates congressional seats (and electoral votes) to the states, which I believe is the "truncate-plus-largest-remainders" method described earlier.  

In the case of CO above, both Nader and Gore were much closer to getting the final electoral vote than Bush, yet it is still awarded to Bush.  

With a large field, the distribution could get even stranger.

Consider a 5-candidate field in a state with 10 EVs:
A - 36%
B - 24%
C - 14%
D - 13%
E - 3 %

Under the "Colorado" system, the EVs would be allocated like this:

A - 6
B - 2
C - 1
D - 1
E - 0

While under the "truncate-plus-remainder" system, they would work out this way:

A - 4
B - 3
C - 2
D - 1
E - 0

Which seems much more logical to me.

[tinman64]

Hey, I know...why not do away with the "new math" and eliminate the EC all together?

[sobo]

Well, it is still closer to the actual result than
10
0
0
0
0

[Nation]

So is Colorado going to use this system or not? Or is it still being decided from within the courts?

[IowaLibert]

This is an excellent point. Fix gerrymandering first.

Posted from Iowa, where we have fixed gerrymandering.

How has Iowa fixed gerrymandering?

By turning it over to a non-partisan legislative services bureau which has fairly strict standards to which it must adhere, one of which is ignoring such matters as voter registration and residence of incumbents. Furthermore, the legislature can only vote up or down the lines submitted to them.  Well, at least the first two times, if it goes to a third plan then they have that option.  But there would be hell to pay with the voters if they rejected the independently drawn boundaries - Iowans really like this process.

I can find a link to an official description of the process, if you're interested. (but it will have to be tomorrow....)

[True Federalist (진정한 연방 주의자)]

This is a poor way to do this allocation.  It should really be done the same way the census allocates congressional seats (and electoral votes) to the states, which I believe is the "truncate-plus-largest-remainders" method described earlier.  

In the case of CO above, both Nader and Gore were much closer to getting the final electoral vote than Bush, yet it is still awarded to Bush.  

With a large field, the distribution could get even stranger.

Consider a 5-candidate field in a state with 10 EVs:
A - 36%
B - 24%
C - 14%
D - 13%
E - 3 %

Under the "Colorado" system, the EVs would be allocated like this:

A - 6
B - 2
C - 1
D - 1
E - 0

While under the "truncate-plus-remainder" system, they would work out this way:

A - 4
B - 3
C - 2
D - 1
E - 0

Which seems much more logical to me.

Congressional approtionment uses the method of equal proportions which is based on the geometric mean and depends upon each state getting at least one representative.  However you could use it for proportioning the electoral votes of a state if the you set some other method for determining when a candidate gets his first electoral vote.  The easiest way would be to use 1 as the multiplier to determine the priority value of the first electoral  vote for that candidate.  (The multiplier for  the nth vote is 1/sqrt(n(n-1)) which is the same as used for the nth representative that each state gets.

For the five candidate race you gave, the method of equal proportions gives the same result as the truncate plus remainder method, but that would not be the case for all possible results.
 

[Fmr. Gov. NickG]

So is Colorado going to use this system or not? Or is it still being decided from within the courts?

It will be decided by the voters on elections day.  If it passes, it will probably also have to be decided by the courts...but only with respect to whether it applies to the current elections or not. If Maine and Nebraska's systems are constitutional, I don't see how there would be anything wrong with this one in general.

[muon2]

This is a poor way to do this allocation.  It should really be done the same way the census allocates congressional seats (and electoral votes) to the states, which I believe is the "truncate-plus-largest-remainders" method described earlier.  

In the case of CO above, both Nader and Gore were much closer to getting the final electoral vote than Bush, yet it is still awarded to Bush.  

With a large field, the distribution could get even stranger.

Consider a 5-candidate field in a state with 10 EVs:
A - 36%
B - 24%
C - 14%
D - 13%
E - 3 %

Under the "Colorado" system, the EVs would be allocated like this:

A - 6
B - 2
C - 1
D - 1
E - 0

While under the "truncate-plus-remainder" system, they would work out this way:

A - 4
B - 3
C - 2
D - 1
E - 0

Which seems much more logical to me.

Congressional approtionment uses the method of equal proportions which is based on the geometric mean and depends upon each state getting at least one representative.  However you could use it for proportioning the electoral votes of a state if the you set some other method for determining when a candidate gets his first electoral vote.  The easiest way would be to use 1 as the multiplier to determine the priority value of the first electoral  vote for that candidate.  (The multiplier for  the nth vote is 1/sqrt(n(n-1)) which is the same as used for the nth representative that each state gets.

For the five candidate race you gave, the method of equal proportions gives the same result as the truncate plus remainder method, but that would not be the case for all possible results.
 

Ernest, I agree that the method you describe would be the most sensible if vote are proportional. Not only does it match the way the congressional apportionment is done, but it also matches the way many multi-party democracies handle proportional representation. I haven't seen which method of allocation CO proposes in its referendum.

That said, if EV's are split by a state I would prefer the ME/NE system since it mirrors the weight of all states and districts in congress. I think there would be less chance of FL-type problems, since the most a recount could affect is one vote for a CD or 2 for the state.

[IowaLibert]

That said, if EV's are split by a state I would prefer the ME/NE system since it mirrors the weight of all states and districts in congress. I think there would be less chance of FL-type problems, since the most a recount could affect is one vote for a CD or 2 for the state.

In my view, you've identified one of the real positives that the ME/NE method produces - small impact of tightly contested EVs.

Unfortunately, the ME/NE method is fatally (in my view) flawed because of the horrorific problem of Congressional district gerrymandering.  I would definitely NOT add to the importance of congressional district lines so long as they're drawn in such a blatantly partisan fashion.

[muon2]

This is an excellent point. Fix gerrymandering first.

Posted from Iowa, where we have fixed gerrymandering.

How has Iowa fixed gerrymandering?

By turning it over to a non-partisan legislative services bureau which has fairly strict standards to which it must adhere, one of which is ignoring such matters as voter registration and residence of incumbents. Furthermore, the legislature can only vote up or down the lines submitted to them.  Well, at least the first two times, if it goes to a third plan then they have that option.  But there would be hell to pay with the voters if they rejected the independently drawn boundaries - Iowans really like this process.

I can find a link to an official description of the process, if you're interested. (but it will have to be tomorrow....)

The key to Iowa's districting is that each CD is made up of a whole number of counties, and a computer algorithm sorts through thousands of combinations to produce the combination that minimizes the population differences.

IA can justfy such a course unit as a county for two reasons. One is that state law uses county-based CDs in drawing its state legislative districts. This provides adequate justification to keep counties intact despite not making them exactly equal in population. The second reason is that no county in IA is larger in population than a CD. Polk Co. is the largest at 374,601, much less than the average CD size. This means IA does not need any exceptions to its procedure.

The IA method cannot directly translate to any state that has one or more counties of population in excess of a CD, or a significant cluster of Polk-sized counties next to each other. Then the population differences would be too great to survive a court challenge.

This leaves states using a smaller unit, such as towns or townships. As long as the size of the smaller units doesn't exceed a CD, this could work like IA. Most of the larger states would still fail. For instance, Chicago by itself is over 3 million people so one is forced to carve it into smaller units. Once the rule is broken, one can argue that it should be broken in other places in the same state so as to make population more equal.

[Posterity]

This is an excellent point. Fix gerrymandering first.

Posted from Iowa, where we have fixed gerrymandering.

How has Iowa fixed gerrymandering?

By turning it over to a non-partisan legislative services bureau which has fairly strict standards to which it must adhere, one of which is ignoring such matters as voter registration and residence of incumbents. Furthermore, the legislature can only vote up or down the lines submitted to them.  Well, at least the first two times, if it goes to a third plan then they have that option.  But there would be hell to pay with the voters if they rejected the independently drawn boundaries - Iowans really like this process.

I can find a link to an official description of the process, if you're interested. (but it will have to be tomorrow....)

That basic description is fine for me, thanks.  Sounds like a good way to handle it.

EDIT:  Based on muon2's post, Iowa's system won't work for Arizona (unfortunately).