Debunking Electoral Reform's "Urban" Myth

Proposed Reform Will Not Result In America's Biggest Cities Controlling The Presidential Election

October 13, 2017

The organization Equal Citizens is working to mount a court challenge to change how states allocate their Electoral Votes. Their proposed court case would not do away with the Electoral College (which would require a Constitutional Amendment), nor does it force states to award its Electors to the winner of the national popular vote (which would require action in many state legislatures and their governors). It would require states to distribute their electors in proportion to their own states' popular vote.

Detractors of this reform complain this would result in voters from a handful of cities and states deciding who becomes President. Let me show you why this is not so.

As of July 2015, there were 143 counties or county-equivalents (parishes, boroughs, independent cities, etc.) having populations of more than 480,000. At least 35 of the 50 states have at least one such county (36 if you count DC, and the Electoral College does, so I will). Collectively, a slight majority of Americans (160,720,317... 50.03%; ~11,000 more than half) lived in those counties. Let's call them "the top-half counties." The rest of America lives in not quite 3,000 other counties. Let's call those "the bottom-half counties."

• Takeaway #1: Turning the fact above upside down, a majority of Americans live in counties having 500,000 people or less. Following election reform, anyone who would aspire to be President intuitively would need support in these communities, regardless of that candidate's level of support in high-population centers.

• Takeaway #2: Politicians who decided to focus on this half of the population necessarily would be putting forth a material campaign effort in 70% of the states. That's far more than the roughly 20% (8-12) states that matter most for a winner-take-all Electoral College.

Breaking it down further, more than 1/8 of America (more than 40,000,000 people) lives in counties having a population of 50,000 or less.

• Takeaway #3: With the proposed electoral reform, a politician who writes off 40-million rural Americans is worse off than a politician who takes a pass on 39,5-million Californians.

The argument that reform would enable large population centers to hold the key to the Presidency only could be true if electoral reform paved a path to the Presidency that allowed a candidate to win the office with a strong majority of the highly-urban vote and just a smattering of less-urban/rural vote. In other words, for the urban-bias theory to hold, the vote of the top-half counties would need to matter more to the final result than the vote of the bottom-half counties, even though the two sets of counties have equal population. Electoral math says that cannot happen.

First, in the 2016 election, the top-half counties accounted for 65.74 million of the 137.06 million popular votes cast for President... just short of 48 percent (despite having 50%+ of the country's people). Demographics explain this; the top-half counties generally have proportionally more children and non-citizen adults than the bottom-half counties.

Secondly, you still have the Electoral College... not the by-and-large winner-take-all Electoral College of today... but you still have states sending Electors to determine the Presidency. Thus, the top-half-bottom-half split of votes cast by state would matter to any analysis trying to determine if the result of reform would be urban bias. The following shows the number of ballots cast for President in 2016, by state and top/bottom-half county.

# of Presidential Popular Votes by State and County-Population Split

State	Top ½	Bottom ½	EV
AK	0	318,608	3
AL	304,191	1,819,181	9
AR	0	1,130,635	6
AZ	1,989,478	615,183	11
CA	11,953,032	2,228,563	55
CO	1,476,412	1,303,808	9
CT	1,207,079	437,841	7
DC	311,268	0	3
DE	262,979	180,835	3
FL	6,021,848	3,480,899	29
GA	1,421,608	2,719,839	16
HI	285,790	143,147	4
IA	0	1,566,031	6
ID	0	690,433	4
IL	3,418,451	2,176,374	20
IN	569,680	2,188,285	11
KS	486,006	708,749	6
KY	353,100	1,571,050	8
LA	0	2,029,032	8
MA	2,776,351	548,695	11
MD	1,769,423	1,012,023	10
ME	0	743,941	4
MI	2,181,850	2,642,692	16
MN	953,120	1,992,113	10
MO	824,373	2,003,893	10
MS	0	1,211,088	6
MT	0	501,822	3
NC	1,257,886	3,483,678	15
ND	0	344,360	3
NE	240,433	603,794	5
NH	0	744,296	4
NJ	2,781,053	1,125,670	14
NM	274,662	523,657	5
NV	767,156	358,229	6
NY	5,080,906	2,640,889	29
OH	2,125,734	3,410,813	18
OK	520,996	931,996	7
OR	668,335	1,333,001	7
PA	2,967,667	3,199,062	20
RI	248,474	214,942	4
SC	215,165	1,887,862	9
SD	0	370,047	3
TN	586,451	1,921,576	11
TX	5,535,636	3,457,530	38
UT	627,385	516,216	6
VA	551,183	3,431,569	13
VT	0	315,067	3
WA	1,979,251	1,337,745	12
WI	750,407	2,225,743	10
WV	0	721,233	5
WY	0	255,849	3
	65,744,819	71,319,584	538

Now suppose the Equal Citizens reform is enacted. Further suppose there is a 2-candidate race (Candidate A and B) where the same number of voters showed up to the polls in each county as 2016. Further suppose Candidate A were to win 75% of the votes in the top-half counties (extremely hard! see below) but just 25% of the votes in the bottom-half counties, with Candidate B winning the rest.

If top-half counties would be capable of exerting over-sized influence on the result of the Presidential election, this split of the vote would need to result in a White House win for Candidate A. However, the opposite actually is true. With more votes to be had in the smaller counties than the bigger ones, Candidate B wins by 2.6-million popular votes and 10 EV:

Candidate A: 75% Top ½ & 25% Bottom ½

Votes A	Votes B	EV A	EV B
67,138,515	69,925,888	264	274

In order for Candidate A to win the popular vote, their support in bottom-half counties would need to be about 27%. That still doesn't get them over the hump in Electoral Votes.

Candidate A: 75% Top ½ & 27% Bottom ½

Votes A	Votes B	EV A	EV B
68,564,904	68,499,499	266	272

What if A's bottom-half support was 29.9%? Not good enough (despite 4.2-million more popular votes)!

Candidate A: 75% Top ½ & 29.9% Bottom ½

Votes A	Votes B	EV A	EV B
70,633,167	66,431,236	268	270

  

Candidate A would need a full 30% of bottom-half county support to secure the required Electoral Votes.

Candidate A: 75% Top ½ & 30% Bottom ½

Votes A	Votes B	EV A	EV B
70,704,493	66,359,910	270	268

Now, do you think it's realistic for anyone to win 3/4 of the top-half county vote in a 2-person race? Hillary Clinton didn't do so in 2016... not nearly. In the top-half counties, Clinton garnered almost exactly five votes for every three cast for Trump. That's the equivalent of a 62.5%/37.5% split in a pure 2-person race. So, let's look at some scenarios where Candidate A's top-half county support is 62.5%.  

Again, with both halves being unequal for popular and Electoral-College purposes, someone with 5/8 of the top-half county vote to earn almost 4 in 10 votes from the bottom-half counties to win the White House. 40% is pretty significant support. Yes, 2016 wasn't a 2-person race, but Clinton failed to get 40% of the vote in 18 states, and Trump failed to get 40% of the vote in 10 states and DC.

Candidate A: 62.5% Top ½ & 37.5% Bottom ½

Votes A	Votes B	EV A	EV B
67,835,356	69,229,047	262	276

Candidate A: 62.5% Top ½ & 38.5% Bottom ½

Votes A	Votes B	EV A	EV B
68,548,555	68,515,848	266	272

Candidate A: 62.5% Top ½ & 39.6% Bottom ½

Votes A	Votes B	EV A	EV B
69,333,066	67,731,337	270	268

• Takeaway #4: Should Equal Citizens succeed in their aims, there is no scenario that would follow reform that would allow any candidate who succeeds wildly in high-population centers to become President without significant support from less-populated areas.

Sources:
County Population Data:
https://data.world/garyhoov/2016-pres-election-by-county

Voting Data By County:
same as sources used by https://uselectionatlas.org/

The author is not affiliated with Equal Citizens or the sources cited above.

Electoral Vote Allocation Alternative

A Lawrence Lessig article (https://medium.com/equal-citizens/the-equal-protection-argument-against-winner-take-all-in-the-electoral-college-b09e8a49d777) posed 4 options, in the form of questions, re: which candidates should be permitted an allocation of a state's Electoral Votes while best respecting 1-man-1-vote (including avoidance of throwing the choosing of the Presidency to the House of Representatives when no candidate receives 270 EV).
a) One candidate with most votes?
b) Two candidates with most votes?
c) All candidates whose share of the vote met or exceed (somewhere between 5%-15%) of votes cast?
d) All candidates whose share of the vote would round to a whole number of EV?

The article posits c) as the answer.

What about
e) a variant of c) which involves ranked-choice voting. Voters would be permitted a "first choice" and a "second choice." The state would tabulate the total number of occurrences of each choice combination of their voters. Where any candidates' first-choice votes fail to meet or exceed the chosen minimum percentage, the state would substitute the second-choices of their voters. Any votes where the second-choice candidate also does not meet the minimum percentage does not count.

Why? Because it maximizes the chance of each individual's vote becoming meaningful to their state's EV calculation while still minimizing the risk of throwing the election to the House of Representatives.

2016 example -- California.. % votes // unrounded EV // whole EV
EV under method c)
Clinton - 61.72% // 35.49 // 35
Trump - 31.62% // 19.51 // 20
Total - 93.34% // 55 // 55 (93.34% means 6.66% of CA voters, 1 in 15, would have no say)

Under method d)
Clinton - 61.72% // 34.40 // 34
Trump - 31.62% // 17.62 // 20
Johnson - 3.37% // 1.88 // 2
Stein - 1.97% // 1.09 // 1
Total - 98.68% // 55 // 55

Supposing the following:
Second choices of Johnson voters -- 50% Trump (1.69%), 40% Clinton (1.35%), 10% other.
Second choices of Stein voters -- 80% Clinton (1.58%), 10% Trump (0.20%), 10% other.

Under method c) modified for ranked choice voting...
Clinton - 64.65% // 36.22 // 36
Trump - 33.51% // 18.78 // 19
Total - 98.16% // 55 // 55 {leaving 1.84% w/o a say, much closer to d) than c)}

Alternative To The One-Game MLB Wild Card Round -- Weekend Of Truth

Background

Major League Baseball has a long tradition of playoff rounds set up as best-of-5 or best-of-7 games. Theory being... you play 162 regular season games, one single game shouldn't seal your fate.

Since the advent of a system where 5 teams per league make the playoffs, they've abandoned that. The two best teams in each league that don't win their division titles play a one-game playoff to determine two teams that will advance to the first 5-game-series round of the playoffs. Some don't like the do-or-die nature of that.

Proposal

I propose an alternative that respects the 162-game season, permits more than 4 teams per league to play in a meaningful post-season tournament where each round has 3 or more games... in such a way that you probably would not have to change the MLB Collective Bargaining Agreement.

Details

The 162-game regular season schedule shall be developed such that all teams would be scheduled to
* play Game 159 on the last Wednesday of September.
* play Games 160-162 on the Friday thru Sunday to follow.
* open the Thursday in between to be designated as a universal open date.

Determine the top 6 teams in each league at the end of the Game-159 Wednesday.  Any two-way tie for 6th place would be broken in a one-game tie-breaker on Thursday.  Ties for 2nd thru 5th place would be broken by regular-season statistics of MLB's choosing (i.e. better head to head record, net runs, etc).

Every team plays their scheduled Games 160-162 as and where the schedule-makers originally decided, except for the #3-#6 teams in each league and their scheduled opponents.

The #3-#6 ranked teams would begin their Weekend Of Truth (WOT) Series (instead of a one-game Moment Of Truth)... four series played to determine who advances to play in the Divisional Series round of the playoffs.

Team #6 would play at team #3, and #5 at #4 in each league.  All of these games count toward regular season statistical totals.

The teams that had been scheduled to be the opponents of these WOT teams would, instead, be rescheduled to have games against other teams whose original schedule was usurped by the WOT games.  These also count toward regular-season statistical totals.

Advancing from the WOT to the Division Series

Each WOT series will last at least three games (to keep all teams to a 162-game schedule). Other than that, it should be treated as a best-of-5 series where the home team (out of respect for their better regular season record) gets spotted a 1-0 series lead before the initial actual game of the WOT series. 

Thus, if the home team then wins at least two of the three initially-scheduled games against the visiting team, the home team advances.

If the visiting team sweeps all three initially-scheduled games, they advance.

If the visitor wins two of the three games held Friday-Sunday, the two teams play a do-or-die 4th game Monday.

Example:

Based on the standings at the end of September 28, 2017 (yes, that's a Thursday... but many teams hadn't played their 159th game on the real-life schedule until Thursday), the following would happen:

The Angels and Royals would be tied for 6th in the AL. The winner of a Thursday game between the two would visit the Red Sox on the weekend.
The Twins would visit the New York Yankees.
The Brewers would travel to Wrigley to face the Cubs.
The Rockies would play at the Diamondbacks.

The Angels/Royals, Twins, Brewers, and Rockies would need to win 3 games against their opponents to advance.
The Red Sox, Yankees, Cubs, and Diamondbacks only would need to win 2 games to advance.

To make these 8 teams opponents whole for a 162-game slate, the following teams (with their originally-scheduled opponents in parenthesis) would meet that Friday-Sunday.
The Tigers (MIN) would play the Blue Jays (NYY).
The Reds (CHC) would play the Cardinals (MIL).
With a Royals win...
...The Dodgers (COL) would play the Astros (BOS).
With an Angels win...
...The Astros (BOS) would play the Royals (ARI).
...The Dodgers (COL) would play the Mariners (LAA).

Thoughts and Questions On Equal Citizens - Part 1

For those of you who read my blog but haven't visited the Equal Citizens website, let me give you its "elevator pitch."

1) A flaw of the Electoral College (EC) system is that states choose to award Electoral Votes (EV) mostly on a winner-take-all (WTA) basis. While the EC is Constitutionally mandated, WTA is not.

2) WTA makes votes in a handful of swing states worth much more than those in other states, depriving voters in other states of equal protection under the law and the expectation of one person, one vote.

3) Legal precedent appears to strongly disfavor the status quo and it is thought that a persuasive argument would find a receptive audience in the Supreme Court as constituted.

4) Equal Citizens wishes to mount a legal challenge to abolish WTA in favor of having ALL states award their EV in proportion to the statewide popular vote.

I do not represent Equal Citizens; if you want to know more or want clarification, go to their website.

I do have a few questions... two of which I posed directly to its organizers.

Q1 - On the post https://medium.com/equal-citizens/the-equal-protection-argument-against-winner-take-all-in-the-electoral-college-b09e8a49d777 you show a map indicating proportional allocation of the 2016 popular vote would have resulted in a 270-263 Clinton win. My math/methodology would lead to a result of the House of Representatives picking the President: Clinton 266, Trump 262, Johnson 8, Stein 1, McMullin 1. What methodology underlies the proportional allocation associated with that map?

Q2 - Suppose SCOTUS hears the case and rules in favor... but then allows the states do decide their own allocation methodologies. If states opt for the Maine/Nebraska method, one man one vote suffers just as much because of partisan gerrymandering of Congressional districts in various states. Describe how you plan to convince SCOTUS not only to strike down Winner-Take-All but compel the states to adopt a method of proportional EV allocation that's NOT subject to manipulation through the redistricting process?

Equal Citizen's answers:

A1 - We had a voting data expert (who's a political science professor at U of Michigan) crunched our numbers and he said that the different results are most likely due to how third party candidates are treated. Unfortunately he won't be able to answer individual questions like this (the fall semester just started, after all), but I'll talk to my team about writing a blog post on how a proportional allocation scheme would work and how the votes could be divided. If they see there's a demand for an explainer on this, I think they'd be on board.

A2 - You're right that allocation by congressional district would lead to even more unequal representation. We believe we could file the lawsuit in a way that pushes the court to recognize that congressional district allocation violates the Equal Protection clause as well.  We'll roll out more contents that explain our plan in the coming weeks, so please follow us on Facebook as we post our informational contents there.


In the absence of receiving Equal Citizens' methodology of proportions, I'll show you mine.

First, let's look at the 2016 popular votes from each state, expressed as a percentage of all Presidential votes cast in that state.

ST	HC (D)	DT (R )	GJ (L)	JS (G)	EM (I)	OTHER
AL	34.358%	62.083%	2.094%	0.442%	0.000%	1.023%
AK	36.551%	51.282%	5.877%	1.800%	0.000%	4.490%
AZ	45.126%	48.672%	4.132%	1.335%	0.678%	0.057%
AR	33.653%	60.574%	2.638%	0.838%	1.172%	1.125%
CA	61.726%	31.617%	3.374%	1.965%	0.279%	1.039%
CO	48.157%	43.251%	5.184%	1.383%	1.040%	0.985%
CT	54.566%	40.927%	2.959%	1.389%	0.128%	0.031%
DE	53.086%	41.713%	3.325%	1.375%	0.159%	0.342%
DC	90.864%	4.087%	1.576%	1.368%	0.000%	2.105%
FL	47.823%	49.022%	2.198%	0.684%	0.000%	0.273%
GA	45.640%	50.771%	3.045%	0.187%	0.316%	0.041%
HI	62.221%	30.039%	3.719%	2.969%	0.000%	1.052%
ID	27.492%	59.261%	4.104%	1.231%	6.733%	1.179%
IL	55.825%	38.762%	3.786%	1.387%	0.211%	0.029%
IN	37.775%	56.940%	4.899%	0.287%	0.000%	0.099%
IA	41.740%	51.147%	3.779%	0.733%	0.790%	1.811%
KS	36.052%	56.655%	4.678%	1.985%	0.550%	0.080%
KY	32.682%	62.520%	2.794%	0.723%	1.184%	0.097%
LA	38.450%	58.089%	1.872%	0.692%	0.421%	0.476%
ME	47.830%	44.870%	5.095%	1.905%	0.252%	0.048%
MD	60.326%	33.909%	2.862%	1.292%	0.346%	1.265%
MA	60.005%	32.808%	4.151%	1.433%	0.082%	1.521%
MI	47.275%	47.498%	3.587%	1.072%	0.170%	0.398%
MN	46.445%	44.925%	3.836%	1.256%	1.802%	1.736%
MS	40.115%	57.941%	1.194%	0.309%	0.000%	0.441%
MO	38.135%	56.772%	3.466%	0.905%	0.252%	0.470%
MT	35.746%	56.168%	5.640%	1.603%	0.462%	0.381%
NE	33.699%	58.747%	4.613%	1.039%	0.000%	1.902%
NV	47.918%	45.501%	3.322%	0.000%	0.000%	3.259%
NH	46.826%	46.459%	4.135%	0.873%	0.143%	1.564%
NJ	55.453%	41.350%	1.871%	0.975%	0.000%	0.351%
NM	48.256%	40.043%	9.337%	1.237%	0.730%	0.397%
NY	59.006%	36.516%	2.287%	1.398%	0.134%	0.659%
NC	46.173%	49.828%	2.744%	0.255%	0.000%	1.000%
ND	27.227%	62.956%	6.224%	1.098%	0.000%	2.495%
OH	43.558%	51.688%	3.175%	0.842%	0.229%	0.508%
OK	28.932%	65.323%	5.745%	0.000%	0.000%	0.000%
OR	50.072%	39.094%	4.708%	2.498%	0.000%	3.628%
PA	47.465%	48.183%	2.380%	0.810%	0.105%	1.057%
RI	54.407%	38.898%	3.177%	1.340%	0.111%	2.067%
SC	40.673%	54.939%	2.340%	0.620%	0.999%	0.429%
SD	31.737%	61.531%	5.634%	0.000%	0.000%	1.098%
TN	34.716%	60.722%	2.807%	0.638%	0.478%	0.639%
TX	43.235%	52.235%	3.161%	0.798%	0.472%	0.099%
UT	27.459%	45.538%	3.501%	0.834%	21.538%	1.130%
VT	56.678%	30.269%	3.199%	2.145%	0.203%	7.506%
VA	49.728%	44.407%	2.968%	0.694%	1.357%	0.846%
WA	52.539%	36.833%	4.850%	1.761%	0.000%	4.017%
WV	26.426%	68.499%	3.220%	1.130%	0.155%	0.570%
WI	46.454%	47.218%	3.584%	1.044%	0.398%	1.302%
WY	21.877%	68.173%	5.193%	0.983%	0.000%	3.774%

Table 1 - 2016 Presidential Election, Percentage of Ballots Cast for each Candidate, by State.

Now... to the above information, let's add the following:

a) the number of Electoral Votes to be awarded by each state.

b) a decimal computed by calculating 1 divided by (# of EV x 2). This comes under the heading "0.5EVTh" for "0.5 EV Threshold"... the percentage of votes one would need in a state that's theoretically worth 1/2 EV.  In a 5 EV state, that threshold is 10% (1 divided by (5 x 2) = 1/10 or 10%); in CA, with 55 EV, it's 1/110 or 0.9091%.

c) each candidate whose vote total in that state does not exceed the 1/2 EV threshold gets their vote total zeroed out. This is to prevent a quirk of (eventual) rounding from resulting in a candidate with less than 1/2 EV worth of votes in a state being awarded 1 full EV from that state.

d) a cumulative of the vote percentages of all candidates whose totals were not reset to zero in step c).

ST	E.V.	0.5EVTh	HC	DT	GJ	JS	EM	CUM.
AL	9	5.556%	34.358%	62.083%				96.441%
AK	3	16.667%	36.551%	51.282%				87.833%
AZ	11	4.545%	45.126%	48.672%				93.798%
AR	6	8.333%	33.653%	60.574%				94.227%
CA	55	0.909%	61.726%	31.617%	3.374%	1.965%		98.682%
CO	9	5.556%	48.157%	43.251%				91.408%
CT	7	7.143%	54.566%	40.927%				95.493%
DE	3	16.667%	53.086%	41.713%				94.799%
DC	3	16.667%	90.864%					90.864%
FL	29	1.724%	47.823%	49.022%	2.198%			99.043%
GA	16	3.125%	45.640%	50.771%				96.411%
HI	4	12.500%	62.221%	30.039%				92.260%
ID	4	12.500%	27.492%	59.261%				86.753%
IL	20	2.500%	55.825%	38.762%	3.786%			98.373%
IN	11	4.545%	37.775%	56.940%	4.899%			99.614%
IA	6	8.333%	41.740%	51.147%				92.887%
KS	6	8.333%	36.052%	56.655%				92.707%
KY	8	6.250%	32.682%	62.520%				95.202%
LA	8	6.250%	38.450%	58.089%				96.539%
ME	4	12.500%	47.830%	44.870%				92.700%
MD	10	5.000%	60.326%	33.909%				94.235%
MA	11	4.545%	60.005%	32.808%				92.813%
MI	16	3.125%	47.275%	47.498%	3.587%			98.360%
MN	10	5.000%	46.445%	44.925%				91.370%
MS	6	8.333%	40.115%	57.941%				98.056%
MO	10	5.000%	38.135%	56.772%				94.907%
MT	3	16.667%	35.746%	56.168%				91.914%
NE	5	10.000%	33.699%	58.747%				92.446%
NV	6	8.333%	47.918%	45.501%				93.419%
NH	4	12.500%	46.826%	46.459%				93.285%
NJ	14	3.571%	55.453%	41.350%				96.803%
NM	5	10.000%	48.256%	40.043%				88.299%
NY	29	1.724%	59.006%	36.516%	2.287%			97.809%
NC	15	3.333%	46.173%	49.828%				96.001%
ND	3	16.667%	27.227%	62.956%				90.183%
OH	18	2.778%	43.558%	51.688%	3.175%			98.421%
OK	7	7.143%	28.932%	65.323%				94.255%
OR	7	7.143%	50.072%	39.094%				89.166%
PA	20	2.500%	47.465%	48.183%				95.648%
RI	4	12.500%	54.407%	38.898%				93.305%
SC	9	5.556%	40.673%	54.939%				95.612%
SD	3	16.667%	31.737%	61.531%				93.268%
TN	11	4.545%	34.716%	60.722%				95.438%
TX	38	1.316%	43.235%	52.235%	3.161%			98.631%
UT	6	8.333%	27.459%	45.538%			21.538%	94.535%
VT	3	16.667%	56.678%	30.269%				86.947%
VA	13	3.846%	49.728%	44.407%				94.135%
WA	12	4.167%	52.539%	36.833%	4.850%			94.222%
WV	5	10.000%	26.426%	68.499%				94.925%
WI	10	5.000%	46.454%	47.218%				93.672%
WY	3	16.667%	21.877%	68.173%				90.050%

Table 2 - % of 2016 Ballots Cast by Candidate and State where Percentage met or exceeded the equivalent of 0.5 EV.

Next, let's compute the # of EV in the state multiplied by the candidates' vote totals divided by the cumulative percentage of above-the-threshold ballots... rounded to 3 decimals.

Example: Alabama... Clinton 9 x 34.358% / 96.441% = 3.206

ST	EV	HC	DT	GJ	JS	EM
AL	9	3.206	5.794
AK	3	1.248	1.752
AZ	11	5.292	5.708
AR	6	2.143	3.857
CA	55	34.403	17.622	1.881	1.095
CO	9	4.742	4.259
CT	7	4.000	3.000
DE	3	1.680	1.320
DC	3	3.000
FL	29	14.003	14.354	0.644
GA	16	7.574	8.426
HI	4	2.698	1.302
ID	4	1.268	2.732
IL	20	11.350	7.881	0.770
IN	11	4.171	6.288	0.541
IA	6	2.696	3.304
KS	6	2.333	3.667
KY	8	2.746	5.254
LA	8	3.186	4.814
ME	4	2.064	1.936
MD	10	6.402	3.598
MA	11	7.112	3.888
MI	16	7.690	7.726	0.584
MN	10	5.083	4.917
MS	6	2.455	3.545
MO	10	4.018	5.982
MT	3	1.167	1.833
NE	5	1.823	3.177
NV	6	3.078	2.922
NH	4	2.008	1.992
NJ	14	8.020	5.980
NM	5	2.733	2.268
NY	29	17.495	10.827	0.678
NC	15	7.215	7.786
ND	3	0.906	2.094
OH	18	7.966	9.453	0.581
OK	7	2.149	4.851
OR	7	3.931	3.069
PA	20	9.925	10.075
RI	4	2.332	1.668
SC	9	3.829	5.171
SD	3	1.021	1.979
TN	11	4.001	6.999
TX	38	16.657	20.125	1.218
UT	6	1.743	2.890			1.367
VT	3	1.956	1.044
VA	13	6.867	6.133
WA	12	6.6913	4.6910	0.618
WV	5	1.392	3.608
WI	10	4.959	5.041
WY	3	0.729	2.271

Table 3 - Preliminary Calculated Electoral Votes, rounded to 0.001

Now, we'll round these to whole numbers. Generally .500 rounds to up and .499 rounds down... but there are exceptions in MI and WA, where a strict rounding rule would result in an additional EV in each state. In each case, only two totals could be rounded up of three possible decimals that ordinarily would be rounded up. The total with the lowest amount to the right of the decimal was rounded down (in blue)... the others (in yellow) were rounded up.

ST	EV	HC	DT	GJ	JS	EM
AL	9	3	6	0	0	0
AK	3	1	2	0	0	0
AZ	11	5	6	0	0	0
AR	6	2	4	0	0	0
CA	55	34	18	2	1	0
CO	9	5	4	0	0	0
CT	7	4	3	0	0	0
DE	3	2	1	0	0	0
DC	3	3	0	0	0	0
FL	29	14	14	1	0	0
GA	16	8	8	0	0	0
HI	4	3	1	0	0	0
ID	4	1	3	0	0	0
IL	20	11	8	1	0	0
IN	11	4	6	1	0	0
IA	6	3	3	0	0	0
KS	6	2	4	0	0	0
KY	8	3	5	0	0	0
LA	8	3	5	0	0	0
ME	4	2	2	0	0	0
MD	10	6	4	0	0	0
MA	11	7	4	0	0	0
MI	16	8	8	0	0	0
MN	10	5	5	0	0	0
MS	6	2	4	0	0	0
MO	10	4	6	0	0	0
MT	3	1	2	0	0	0
NE	5	2	3	0	0	0
NV	6	3	3	0	0	0
NH	4	2	2	0	0	0
NJ	14	8	6	0	0	0
NM	5	3	2	0	0	0
NY	29	17	11	1	0	0
NC	15	7	8	0	0	0
ND	3	1	2	0	0	0
OH	18	8	9	1	0	0
OK	7	2	5	0	0	0
OR	7	4	3	0	0	0
PA	20	10	10	0	0	0
RI	4	2	2	0	0	0
SC	9	4	5	0	0	0
SD	3	1	2	0	0	0
TN	11	4	7	0	0	0
TX	38	17	20	1	0	0
UT	6	2	3	0	0	1
VT	3	2	1	0	0	0
VA	13	7	6	0	0	0
WA	12	7	5	0	0	0
WV	5	1	4	0	0	0
WI	10	5	5	0	0	0
WY	3	1	2	0	0	0
		266	262	8	1	1

Table 4 - Proportional EV totals, rounded to integers

Thus... by my methodology... with neither candidate having won 270 EV, the House of Representatives ends up choosing the President. Probably Trump, maybe Johnson, snowball's-chance-in-Phoenix for Clinton... unless...

That brings me to my next post... I'll posit that proportional allocation of EV doesn't do away with swing states... it just changes which ones they are.