2016 Presidential Election Prediction

Place your bets...

Now for the moment probably only a few people have been waiting for: the Official Nuking Politics 2016 Presidential Election Prediction™.

This is going to be a numbers-heavy post, so bear with me. First though, I am going to explain my methodology. Most elections can be pretty frustrating to predict and generally all the media seems to go on is polling numbers. Unfortunately, polls are notoriously unreliable. Some elections they seem to give a decent indication of how it is going to turn out, while in others they fail miserably (1980 Reagan-Carter is a good example).

So my conclusion is, there is no real correlation between polls and outcome other than, if the Democrat ends up winning, then the polls seem to have been correct, but if the Republican wins then the polls seem to have been mistaken. This is because the polls always show the Democrat winning (at least most of the time). I'll not speculate as to why that is in this post, but I'm guessing you can figure it out.


So I decided to look at another source of data to base a prediction on. One that is not so prone to errors or manipulation: Primary voter participation. It is well established that presidential elections generally swing one way or another based on whether or not the "sometimes" voters show up or not. The country is relatively 50/50 split between generally voting Republican or Democrat when it comes to sometimes or all-the-time voters. Those who rarely or never vote only show up once in a blue moon, and they tend to vote Republican when they do, historically speaking. This may be one of those times, but it's too hard to figure out the numbers involved there, so I am not considering it (although they may be hidden in the primary data, which I suspect they do).

It turns out there is a strong correlation between primary participation and election outcomes, as this site explains in detail. My methodology is a little too complex to fully explain, but basically I took the participation of the recent primaries and compared it to Republican participation in the 2012 primaries and Democrat participation in the 2008 primaries, both adjusted for estimated 2016 voter registration (+2.5% for Republicans based on 2012 registration numbers, and +7.3% for Democrats based on 2008 registration numbers).

The overall trend was a massive increase in Republican voter participation in the primaries (approximately +60%) and a noticeable drop in Democratic voter participation (-19%). I used slightly different calculations for each state, based on factors such as the timing of the primary (massive upswing in New York for Republicans, but the primary is usually over by the time it gets to NY), or special considerations specific to that state or primary compared to previous ones. I very rarely used an increase in Republican numbers (which were through the roof) and focused mainly on the drop in Democratic participation. In all cases I used what I termed "the most significant factor" to calculate a "swing" percentage to apply to voting numbers compared to the 2012 Presidential Election.

A few states had either not enough data available to determine anything of value, or had negligible changes in the identified "most significant factor". In these cases, I merely increased the vote count by 2.5% based on the estimated change in voter registration compared to 2012.

I want to stress that the following numbers aren't "real", as in, there is no way to predict what the actual numbers will be. These should only be regarded as qualitative in nature, not quantitative. They merely show a trend in a certain direction and aren't to be taken as anywhere near exact. In some cases, I guarantee they are way off, but in every one of those cases the outcome is still solid.

I also want to point out that no considerations such as 3rd party candidates, #DemExit/#NeverHillary, #NeverTrump, or current polls were taken into account. This is strictly R vs D based on primaries. Ultimately, my judgement on all of that stuff is that it would further hurt Clinton more so than it will effect Trump. Depending on how severe the effect is of former Bernie supporters refusing to vote for Hillary for instance, some states like Colorado might swing to Trump. I also think there is a slim possibility that many New York Democrats might actually vote for Trump.

After doing the math, I have given up any hope of California voting Republican. Based on past elections, it is so far in Democrat territory, that it would be impossible. Washington is close, however, so that is an outlier possibility. Hawai'i and D.C. are solidly Democrat.

The ultimate conclusions are that the Republican nominee, Donald Trump, will win in a massive landslide: 356 electoral votes to 182 over Clinton and with 57.8% of the vote. Key states that will swing include: Pennsylvania, Wisconsin, Michigan, Ohio, Florida, New Hampshire, New Jersey, Rhode Island, and Virginia.

The numbers have Minnesota winning, but it is close and I am not very confident in Minnesota voting Republican for the first time since 1972.

Here is the prediction map:

The following is a list of each state plus D.C. with the 2012 numbers (12 Election R and D) and my predicted numbers for 2016 (Pred 16 Elect R and D):

UPDATE!  I don't know if people aren't reading my warning above about the following numbers, or they just aren't understanding it, but I'm getting questions about the point spread on some of these. They don't matter, so don't take them as a literal prediction.

As an example to explain how to read this, I have Trump winning in Florida by 21 points and winning Georgia by 38 points. I am not predicting that will be the actual outcome. You should merely read this as a prediction that Trump will win Florida and Georgia with a high level of confidence. The larger the margin, the higher the level of confidence. The actual margins don't mean anything more than that. I don't see Trump winning in either state by that big of a spread.

State		12 Election R		12 Election D		Pred 16 Elect R		Pred 16 Elect D
Alabama		1,255,925		795,696		1,331,280		747,954
Alaska		164,676		122,640		144,915		137,357
Arizona		1,233,654		1,025,232		1,196,644		1,055,989
Arkansas		647,744		394,409		855,022		268,198
California		4,839,958		7,854,285		5,420,753		6,911,771
Colorado		1,185,243		1,323,101		1,220,800		1,283,408
Connecticut		634,892		905,083		704,730		805,524
Delaware		165,484		242,584		173,758		230,455
Florida		4,163,447		4,273,756		5,327,932		3,504,480
Georgia		2,078,688		1,773,827		2,723,081		1,223,941
Hawai'i		121,015		306,658		133,117		275,992
Idaho		420,911		212,787		387,238		229,810
Illinois		2,135,216		3,019,512		2,199,273		2,928,927
Indiana		1,420,543		1,152,887		2,173,431		541,857
Iowa		730,617		822,544		971,721		551,105
Kansas		692,634		440,726		699,560		436,319
Kentucky		1,087,190		679,370		1,630,785		339,685
Louisiana		1,152,262		809,141		1,371,192		655,404
Maine		292,276		401,306		312,735		373,215
Maryland		971,869		1,677,844		962,150		1,694,622
Massachusetts		1,188,314		1,921,290		1,271,496		1,786,800
Michigan		2,115,256		2,564,569		2,749,833		1,795,198
Minnesota		1,320,225		1,546,167		1,439,045		1,407,012
Mississippi		710,746		562,949		796,036		495,395
Missouri		1,223,796		1,482,440		1,554,221		1,082,181
Montana		267,928		201,839		334,910		151,379
Nebraska		475,064		302,081		570,077		241,665
Nevada		463,567		531,373		621,180		350,706
New Hampshire		329,918		369,561		376,107		317,823
New Jersey		1,477,568		2,125,101		1,876,511		1,551,324
New Mexico		335,788		415,335		248,483		523,322
New York		2,940,431		4,485,741		3,013,942		4,597,885
North Carolina		2,270,395		2,178,391		2,633,658		1,829,848
North Dakota		188,163		124,827		192,867		127,948
Ohio		2,661,433		2,827,710		2,927,576		2,544,939
Oklahoma		891,325		443,547		1,007,197		385,886
Oregon		754,175		970,488		791,884		921,964
Pennsylvania		2,680,434		2,990,274		3,538,173		2,405,958
Rhode Island		157,204		279,677		216,942		173,400
South Carolina		1,071,645		865,941		1,296,691		684,093
South Dakota		210,610		145,039		259,050		107,329
Tennessee		1,462,330		960,709		1,608,563		864,638
Texas		4,569,843		3,308,124		6,306,383		2,051,037
Utah		740,600		251,813		859,096		211,523
Vermont		92,698		199,239		106,603		169,353
Virginia		1,822,522		1,971,820		2,278,153		1,478,865
Washington		1,290,670		1,755,396		1,458,457		1,527,195
Washington D.C.		21,381		267,070		16,891		323,155
West Virginia		417,655		238,269		551,305		162,023
Wisconsin		1,407,966		1,620,985		1,619,161		1,377,837
Wyoming		170,962		69,286		213,703		51,965
TOTALS		60,933,500		65,915,796		78,674,311		57,456,259

And here are the percentages of the votes and electoral vote outcomes for each state:

State		% R	% D	EV R	EV D
Alabama		64	36	9
Alaska		51.3	48.7	3
Arizona		53.1	46.9	11
Arkansas		76.1	33.9	6
California		44	56		55
Colorado		48.8	51.2		9
Connecticut		46.7	53.3		7
Delaware		43	57		3
Florida		60.3	39.7	29
Georgia		69	31	16
Hawai'i		32.5	67.5		4
Idaho		62.8	37.2	4
Illinois		42.9	57.1		20
Indiana		80	20	11
Iowa		63.8	36.2	6
Kansas		61.6	38.4	6
Kentucky		82.8	17.2	8
Louisiana		67.7	32.3	8
Maine		45.6	54.4		4
Maryland		36.2	63.8		10
Massachusetts		41.8	58.2		11
Michigan		60.5	39.5	16
Minnesota		50.6	49.4	10
Mississippi		61.6	38.4	6
Missouri		59	41	10
Montana		68.9	31.1	3
Nebraska		70.2	29.8	5
Nevada		63.9	36.1	6
New Hampshire		54.2	45.8	4
New Jersey		54.7	45.3	14
New Mexico		32.2	67.8		5
New York		39.6	60.4		29
North Carolina		59	41	15
North Dakota		60.1	39.9	3
Ohio		53.5	46.5	18
Oklahoma		72.3	27.7	7
Oregon		46.2	53.8		7
Pennsylvania		59.5	40.5	20
Rhode Island		55.6	44.4	4
South Carolina		65.5	34.5	9
South Dakota		70.7	29.3	3
Tennessee		65	35	11
Texas		75.5	24.5	38
Utah		80.2	19.8	6
Vermont		38.6	61.4		3
Virginia		60.6	39.4	13
Washington		48.8	51.2		12
Washington D.C.		5	95		3
West Virginia		77.3	22.7	5
Wisconsin		54	46	10
Wyoming		80.4	19.6	3
TOTALS		57.8	42.2	356	182