Gaming the system -- model at 270towin.com

[pbrower2a]

270towin.com has an analysis of several possible means of allocating electoral votes. As we all know, President Obama won 332 electoral votes and the election by winning enough of 48 states and the District of Columbia on a winner-take-all basis and the two states that allocate their electoral votes based on Congressional districts. (The results for Maine and Nebraska were the same as for winner-take-all).

Alternatives include:

1. The current hybrid in which Maine and Nebraska allocate electoral votes by district with the winner of the plurality getting the other two electoral votes, and other states and DC go winner-take-all.

Obama 332, Romney 206. 

2. Winner-take-all in all  50 states and DC. No change in 2012, but it would have prevented Barack Obama picking up a stray electoral vote from Nebraska in 2008 because Greater Omaha voted differently from the rest of Nebraska.

3. An interstate compact to allocate electoral votes to the winner of the popular vote irrespective of how the State votes on a winner-take-all system within such states. So long as states with at least 270 electoral votes so allocate their votes, President Obama would have won every state participating in the compact and won re-election no matter what other states did.

4. Near-proportional allocation of electoral votes. States allocate two electoral votes based upon winning the plurality of the vote within the State and the others proportionately. With this system the Presidential election would have gone

Obama 283 -- Romney 254 -- G. Johnson 1

(the libertarian nominee winning one electoral vote in California) 

This is probably the allocation closest to the proportional distribution of the popular vote, which would have been

Obama 274 -- Romney 254 -- G. Johnson 5 -- Stein 2 -- others 3

5. Congressional District - Popular  awards two electoral votes to the popular vote winner of the state, with one each allocated to the popular vote winner in each individual Congressional District.   Due to gerrymandering, some states have ultra-strong Democratic districts concentrated in urban areas  so that the remainder of the state can be split nearly at-large in the rest. Example: Michigan has five safe D districts in the Detroit-Ann Arbor-Bay City triangle but the other nine districts that effectively dilute Democratic votes in Grand Rapids, Lansing, Kalamazoo, and Battle Creek all went Republican in Congressional and the Presidential elections in 2012. Outside the Ann Arbor-Detroit-Bay City corridor Michigan as a whole votes much like Texas. See also Pennsylvania, Ohio, Virginia, Wisconsin, and Florida, where state legislatures obeying ALEC gerrymandered districts to maximal benefit to Republicans in Congress.

Here's the result:

264 Obama, 274 Romney

6. Congressional District - Majority (CDM) awards two electoral votes to the party winning the popular vote in a majority of the Congressional Districts, with one each allocated to the popular vote winner in each individual Congressional District.

This system would allot the two at-large seats associated with Senators on a majority of wins of Congressional districts. Based on gerrymandering, a state whose Congressional districts go largely to one Party would go to the winner of the most Congressional seats even if the state as a whole went nearly 10% for the other Presidential nominee. This would have been even more of a distortion of the popular vote:

252 Obama, 286 Romney

7. "Arranged hybrids". By selecting the method to give the greatest advantage to one of the Parties as is imaginable in any state, one could get results as severe as

389 Obama, 149 Romney -- or --

194 Obama, 343 Romney, 1 G. Johnson

Barack Obama was reasonably close to the optimum, suggesting that he ran as competent a campaign as possible. Tellingly the mean between those extremes of scenario indicate such a split as

291.5 Obama, 246 Romney, .5 G. Johnson

which is about 20 away from a proportional allocation of electoral votes to the popular vote.
 
 

     
 
http://www.270towin.com/alternative-electoral-college-allocation-methods/

[pbrower2a]

This is the

194 Obama, 343 Romney, 1 G. Johnson scenario:

The state's winner gets the first number in any split. There are no ties.

Romney won the state or district on a WTA or district basis
Romney would have won the majority under an allocation of a majority of districts
Obama would have won a majority under the most R-friendly distribution
Obama would have won all electoral votes on any basis


Florida goes 18 Romney, 11 Obama -- in case such is hard to see.
New Jersey goes 10 Obama, 4 Romney, likewise.
Massachusetts goes 8 Obama, 3 Romney.
California goes 34 Obama, 20 Romney, 1 G. Johnson.
Connecticut goes 5 Obama, 2 Romney.
*Gary Johnson, Libertarian, gets one electoral vote in California but nowhere else -- and is not shown in any other state.

Stock tip under such a scenario -- Rosetta Stone (software that teaches languages).

[pbrower2a]

Now, rigging things the opposite way:

Romney wins all electoral votes anyway
Obama picks up only one electoral vote  
Obama picks up 2 or 3 electoral votes
Obama picks up 4 electoral votes
Obama picks up 6 electoral votes
Obama picks up 15 electoral votes (Texas only)
Obama won the state in reality and would not have lost anything

[Skill and Chance]

Does the 2 EV for winning the state, otherwise proportional system elect Gore or Bush in 2000?  Purely proportional allocation would send any election where no candidate gets >50% to the House, right?

[pbrower2a]

Does the 2 EV for winning the state, otherwise proportional system elect Gore or Bush in 2000?  Purely proportional allocation would send any election where no candidate gets >50% to the House, right?

My suggestion would be to deny any electoral vote to any nominee who gets less than one full fraction of the vote. Thus to get one electoral vote in California based on House seats one would have to get at least 1/53 of the popular vote statewide. To win a state such as West Virginia (which could still vote based by district because of a small number of electoral votes) one would have to get 1/3 of the popular vote statewide.

Electoral votes would have to be allocated on the basis of candidates who win adequate fractions of the popular vote, with shares less than a fraction going to the winner of the plurality.

Consider Missouri in 2008, in which nobody got a majority:

       

•    John S. McCain, III   Republican    1,445,814    49.36%   2+
     Barack H. Obama   Democratic    1,441,911    49.23%   0
     Ralph Nader           Independent          17,813       0.61%   0
     Bob Barr                Libertarian          11,386       0.39%   0
     Charles O. Baldwin   Constitution            8,201       0.28%   0
     Other (+)   -    -                                    3,986       0.14%   0

McCain wins the plurality (which would not have been established for days)

To qualify for an electoral vote one would have to win at least 1/8 of the total valid popular vote.  With 2,928,841 total valid votes for President one would need at least  366,082 popular votes to win one electoral vote.

Discard the votes of those no longer considered, and one gets

• John S. McCain, III   Republican    1,445,814    49.36%   
  Barack H. Obama   Democratic    1,441,911    49.23%

The 2,887,725 votes remaining after votes for minor candidates failing to qualify for an electoral vote can now be split proportionally.  John McCain won 50.06% of the relevant vote and Barack Obama won 49.93% of the popular vote. Such would translate, respectively, between 4.005 and 3.995 electoral votes. But -- only the winner of the plurality can win any fractional electoral vote, no matter how close. John McCain would have gotten 2 electoral votes by winning the plurality of the vote and 5 electoral votes by winning barely more than half the total relevant vote.  Barack Obama would have gotten 3 electoral votes. 

The system does not work well with a strong third-party nominee -- as in 1992 or perhaps 1968. Winner-take all is then a good back-up.

[muon2]

Congress has the power to require neutral redistricting for congressional districts. It would be interesting to see how the congressional district model would fare if political gerrymandering were taken out of the equation.

[traininthedistance]

Congress has the power to require neutral redistricting for congressional districts. It would be interesting to see how the congressional district model would fare if political gerrymandering were taken out of the equation.

I believe there have been several analyses of just that, and all point to the conclusion that the Republicans' structural advantage would be lessened, but would still persist.

"Fair" districts can very easily still lead to unfair results.

[barfbag]

It's fun to play around with different rules for the Electoral College but it's really one of the greatest things this country has to offer. It protects us from the majority.

[muon2]

Congress has the power to require neutral redistricting for congressional districts. It would be interesting to see how the congressional district model would fare if political gerrymandering were taken out of the equation.

I believe there have been several analyses of just that, and all point to the conclusion that the Republicans' structural advantage would be lessened, but would still persist.

"Fair" districts can very easily still lead to unfair results.

The analyses I've seem suggest that partisan gerrymandering is a much larger effect than structural differences in political geography. Even where there are structural biases, neutral rules tend to increase competitiveness which benefits the party without the structural advantage.

[pbrower2a]

Congress has the power to require neutral redistricting for congressional districts. It would be interesting to see how the congressional district model would fare if political gerrymandering were taken out of the equation.

I believe there have been several analyses of just that, and all point to the conclusion that the Republicans' structural advantage would be lessened, but would still persist.

"Fair" districts can very easily still lead to unfair results.

The analyses I've seem suggest that partisan gerrymandering is a much larger effect than structural differences in political geography. Even where there are structural biases, neutral rules tend to increase competitiveness which benefits the party without the structural advantage.

An advantage in the current system goes to someone who, like Obama in 2012, who wins by smaller margins in the states that he wins and loses by huge margins in those that he loses. Thus if he wins by 6% in Iowa but loses by 11% in Mississippi he splits the electoral votes of those two states evenly (both have six electoral votes) despite winning fewer votes in Iowa and Mississippi together than losing in those two states.

That said, Barack Obama still won fair-and-square, and would have won by any means other than an artificial method that would give electoral votes based upon gerrymandered districts. The two states that now allocate electoral votes based on districts, Maine and Nebraska, don't seem to cause problems. Maine is homogeneous enough that there is no way to gerrymander Congressional districts; Nebraska has one urban area that is roughly one third of the state in population and very different from the rest of the state, and the other two-thirds in population is firmly R except for Lincoln.

Gerrymandered districts are intended to distort electoral results into something grossly unrepresentative and create disadvantages for the Party then out of luck. Congressional districts are artificial but often transitory constructions. Far more people know what county they live in than know what Congressional district they live in.       

[pbrower2a]

Does the 2 EV for winning the state, otherwise proportional system elect Gore or Bush in 2000?  Purely proportional allocation would send any election where no candidate gets >50% to the House, right?

Let's see what we come up with:

Blank map, 2000:

It will be easy enough to distribute the votes of states with three electoral votes and those two (ME, NE) that vote according to a plurality for two electoral at-large votes and the districts by plurality within them. The darkest color indicates no change from reality:
 

So far it is Dubya 18, Gore 11.

Any questions?

[pbrower2a]

A nominee or independent nominee would have to get a full share of the 'other votes' to get an electoral vote. There would be no fractional electoral votes. For a state with four electoral votes not divided by district, such would require the loser to get at the least a plurality.

Ignoring the wasted votes of those who voted for Third Parties that got less than one full share of the vote in any state, one finds that one would need (to three digits if not a whole number)

50% of the vote to win even one electoral vote in a state with four electoral votes
33.333% of the vote to win even one electoral vote in a state with five electoral votes
25% of the vote to win an electoral vote in a state with six electoral votes
20% of the vote to win an electoral vote in a state with seven electoral votes (or 40% to win two electoral votes in such a state)
16.666% of the vote to win an electoral vote in a state with eight electoral votes (or 33.333% to win two electoral votes in such a state)
14.286% of the vote to win an electoral vote in a state with nine electoral votes (or 28.517% to win two electoral votes in such a state, or 42.857% of the vote to win three electoral votes in such a state)
12.5% of the vote to win an electoral vote in a state with ten electoral votes (or 25% to win two electoral votes in such a state, or 37.5% of the vote to win three electoral votes in such a state)
11.111% of the vote to win an electoral vote in a state with eleven electoral votes (or 22.222% to win two electoral votes in such a state,  33.333% of the vote to win three electoral votes in such a state, or 44.444% of the vote to win four electoral votes in such a state)
10% of the vote to win an electoral vote in a state with twelve electoral votes (or 20% to win two electoral votes in such a state,  30% of the vote to win three electoral votes in such a state, or 40% of the vote to win four electoral votes in such a state)
11.111% of the vote to win an electoral vote in a state with eleven electoral votes (or 22.222% to win two electoral votes in such a state,  33.333% of the vote to win three electoral votes in such a state, or 44.444% of the vote to win four electoral votes in such a state)

Continuing:

Further 'no-change' results appear in all states with four electoral votes. Nebraska and Maine would have gone as they did either way. Gore got less than 33.33% of the popular vote in Utah, so he gets nothing there.  

So far it is Dubya 37, Gore 18. Dubya wins by a landslide? Not yet. 483 electoral votes have yet to be accounted for.




 

[pbrower2a]

Now it gets trickier.  I wish to show change and not absolute results.

A nominee or independent nominee would have to get a full share of the 'other votes' to get an electoral vote. There would be no fractional electoral votes. For a state with four electoral votes not divided by district, such would require the loser to get at the least a plurality.

The rules for winning electoral votes in a state that one loses in states that he loses that have five to twelve electoral votes are as follows:

 
33.333% of the vote to win even one electoral vote in a state with five electoral votes
25% of the vote to win an electoral vote in a state with six electoral votes
20% of the vote to win an electoral vote in a state with seven electoral votes (or 40% to win two electoral votes in such a state)
16.666% of the vote to win an electoral vote in a state with eight electoral votes (or 33.333% to win two electoral votes in such a state)
14.286% of the vote to win an electoral vote in a state with nine electoral votes (or 28.517% to win two electoral votes in such a state, or 42.857% of the vote to win three electoral votes in such a state)
12.5% of the vote to win an electoral vote in a state with ten electoral votes (or 25% to win two electoral votes in such a state, or 37.5% of the vote to win three electoral votes in such a state)
11.111% of the vote to win an electoral vote in a state with eleven electoral votes (or 22.222% to win two electoral votes in such a state,  33.333% of the vote to win three electoral votes in such a state, or 44.444% of the vote to win four electoral votes in such a state)
10% of the vote to win an electoral vote in a state with twelve electoral votes (or 20% to win two electoral votes in such a state,  30% of the vote to win three electoral votes in such a state, or 40% of the vote to win four electoral votes in such a state)

Green is for gains by a Republican, and orange for a gain by a Democrat. With 21.88% of the vote in Oklahoma, Gore actually gains one electoral vote of  eight. With 47.85% of the vote in New Mexico, Dubya picks up only one.  Do much the same with Arkansas, Kansas, and West Virginia.  So one shows a very pale shade (20% saturation) and the split for the winner first:

I am obviously not finished. At this point, Al Gore has picked up a net of three electoral votes, which was the margin of his loss in the Electoral College.

[pbrower2a]

Now for other states with seven to twelve electoral votes as of the 2000 election:

Washington -- Bush gains 4.
Oregon -- Bush gains 2.
Arizona -- Gore gains 2.
Colorado -- Gore gains 2.
Minnesota -- Bush gains 2.
Iowa -- Bush gains 2.
Missouri -- Gore gains 3.
Louisiana -- Gore gains 3.
Wisconsin -- Bush gains 4.
Mississippi -- Gore gains 2.
Indiana -- Gore gains 4.
Alabama -- Gore gains 2.
Kentucky -- Gore gains 2.
Tennessee -- Gore gains 4.
South Carolina -- Gore gains 2.
Maryland -- Bush gains 3.
Connecticut -- Bush gains 2.
Massachusetts -- Bush gains 2.

A difference of 2  or 3 -- have the state at 30% saturation. 
4? 40% saturation (I will also use this for a change of 5 or 6 EV).

[pbrower2a]

13 to 23 electoral votes. Eight states, but higher stakes.

Pennsylvania -- Bush gains 9.
Illinois -- Bush gains 8.
Ohio -- Gore gains 8.
Michigan -- Bush gains 7.
New Jersey -- Bush gains 5.
North Carolina -- Gore gains 4.
Georgia -- Gore gains 4.
Virginia -- Gore gains 4.

50% saturation applies to 7, 8, or 9 electoral votes.
 

[pbrower2a]

Now for the giants -- California, Texas, and New York -- and the state that decided the 2000 Presidential election (Florida):

California -- Bush gains 20
Texas -- Gore gains 11
New York -- Bush gains 10
Florida --  Gore gains 11

Nader gets electoral votes in California and New York.

What do you know? Proportional distribution of electoral votes changes 10 or 11 electoral votes in Texas, New York, and Florida. Give those states a saturation of 60%. California is an extreme outlier in electoral votes, so I will give it a 90% saturation.

 

[pbrower2a]

Could someone make a table out of this data? I can do maps, but I stink at tables. 

[pbrower2a]

Table created.

 
State   Bush  Gore Other

AL           7      2
AK           2      3
AZ           6      2
AR           5      1
CA         20     33        1
CO          6       2
CT           2       6
DE           0      3
DC          0       3
FL          14     11
GA          9       4
HI           0       4
ID           4       0
IL           8      14
IN           8        4
IA           2        5
KS           5        1
KY           6        2
LA           6        3
ME          0        4
MD          2        8
MA          2       10
MI           7      11
MN          2        8
MS          5        2
MO          8        3
MT           3         0
NE           5         0
NV           4         0
NH           4         0
NJ            5        10
NM           1         4
NY           10      22    1
NC            9        4
ND            3        0
OH          13        8
OK            7        1
OR            2        5
PA            9       14
RI             0         4
SC            6         2
SD            3         0
TN            7         4
TX           21       11
UT             5         0
VT             0         3
VA             9         4
WA            4         7
WV            4         1
WI             4         7
WY            3          0
               270     267   2      


It would have still all boiled down to the Florida count.