I understand how the US Electoral College works, and how the aggregate popular vote is meaningless.
I also understand that with the 9 swing states there were a number of combinations and scenarios that could have made Mitt Romney the next president, but has anyone figured out the least amount of votes, in certain key states, it would have taken for Romney to win?
For example if he got X more votes in Florida, and Y more votes in Ohio, Romney would have reached 270 and become president.
I realize it doesn’t change the outcome, and that President Obama took almost all of the swing states, but it would show how close this election really was, and how easily it could have gone the other way.
FL is 29 votes, VA is 13 votes, OH is 18 votes, and NH is 4 votes for a total of 64. Romney “currently” has 206, so those four states would have given him 270.
The map dolphinboy linked to didn’t award NC to Romney; NC is worth 15 electoral votes.
Or half that if you assume you’re changing voters’ minds rather than motivating the stayed-at-homes. But are those four states the ‘cheapest’ way to get the extra EVs? Maybe winning two other states would take less than getting FL or VA.
I don’t think it really means that much as an ex-post-facto measure. In any presidential election there are going to be close states, and switching some small number of those will change the outcome. As fumster points out, it 2000 the “small number” of states was one.
Conversely you could ask, “what is the largest number of votes that could be switched from Obama to Romney *without *changing the result?” The answer would be the sum of Obama’s margin (less one vote) in each of the blue state. Without bothering to do all the math, that’s at least 20% of the electorate of California, so many millions of votes that Obama could have done without.
Does that mean Obama’s win was crazy-secure? Of course not. It’s just post-facto trivia.
The largest number of votes you could switch and maintain an Obama victory is much more than that. First you switch all states he lost to 100% Romney. Then you switch several selected states he won (probably OH NH VA FL) to 100% Romney. Finally you switch the remaining states he carried down to a one vote margin. And that may not even be the biggest switch you could make.
I’d predict that you could construct scenarios in which you could win the electoral college with < 30% of the popular vote even assuming equal percentage turnout in all states. This involves winning all the small states which get a disproportionate number of electoral votes. If you let me adjust turnout, it’s theoretically possible to win the election with 11 votes. Simply carry California (55), Texas (34), New York (31), Florida (27), Pennsylvania (21), Illinois (21), Ohio (20), Michigan (17), New Jersey (15), North Carolina (15), and Georgia (15) 1-0 and lose the other states with no votes.
57787 gives it to Kerry in 2004 (the actual switch is in NM IA and CO. He needs 59301 to just win OH which is a bit more).
269 in FL gives it to Gore in 2000
575515 in 1996 gives it to Dole
Even in the blowout year 1988, it only takes 527766 votes to switch the win to Dukakis. This is less than 2% of the votes cast.
Interestingly it shows you have to switch 5 states to Nixon from Kennedy though I seem to recall a “well-known” meme that the election was stolen for Kennedy by switching the IL vote to him. If you switch IL to Nixon, he still only has 247 electoral votes to 290 for Kennedy (DC had no vote then)
This question (or its various hijacks) caught my interest, so I did some tinkering. Without changing the voter turnout, but assuming we can adjust individual votes as needed, the key to winning the election with the lowest possible percentage of the popular vote lies in winning the states with the most powerful voters, i.e. the fewest voters per elector.
So, I summed the popular vote counts for each state, and divided by that state’s number of electoral votes, producing the following “voter power” rankings (from most to least powerful):
State Total votes Electoral votes Voters per elector
Alaska 220596 3 73532
District of Columbia 243348 3 81116
Wyoming 245785 3 81928.33333
Vermont 297274 3 99091.33333
North Dakota 319825 3 106608.3333
Hawaii 429494 4 107373.5
Rhode Island 436853 4 109213.25
South Dakota 363695 3 121231.6667
West Virginia 661643 5 132328.6
Delaware 413844 3 137948
Nebraska 765083 5 153016.6
Utah 920384 6 153397.3333
New Mexico 771356 5 154271.2
Montana 479386 3 159795.3333
Idaho 649653 4 162413.25
Nevada 1011105 6 168517.5
Arkansas 1055112 6 175852
New Hampshire 705874 4 176468.5
Maine 710738 4 177684.5
Arizona 2041519 11 185592.6364
Kansas 1131244 6 188540.6667
Oklahoma 1332019 7 190288.4286
California 10508184 55 191057.8909
Mississippi 1213035 6 202172.5
Texas 7962799 38 209547.3421
New York 6186864 29 213340.1379
South Carolina 1921363 9 213484.7778
Tennessee 2438431 11 221675.5455
Connecticut 1562187 7 223169.5714
Kentucky 1796736 8 224592
Alabama 2064699 9 229411
Washington 2802117 12 233509.75
Indiana 2602883 11 236625.7273
New Jersey 3374407 14 241029.0714
Georgia 3877038 16 242314.875
Oregon 1697276 7 242468
Maryland 2475956 10 247595.6
Louisiana 1988576 8 248572
Illinois 5091061 20 254553.05
Iowa 1563943 6 260657.1667
Colorado 2408844 9 267649.3333
Missouri 2744935 10 274493.5
Pennsylvania 5596499 20 279824.95
Massachusetts 3128134 11 284375.8182
Virginia 3747858 13 288296.7692
Florida 8459112 29 291693.5172
Minnesota 2921173 10 292117.3
Michigan 4717728 16 294858
Ohio 5368334 18 298240.7778
North Carolina 4499039 15 299935.9333
Wisconsin 3055484 10 305548.4
If Obama were to win the top 32 states on that list (Alaska through Alabama) by one vote, and every other state went entirely to Romney, Obama would win the election with just 25,430,095 votes — or 20.68% of the total.
Interestingly, that’d actually give him 273 electoral votes, so he might be able to do even better if you switch out one of the more heavily-populated states. I’d wager you could get it under 20%, but I’ve procrastinated long enough with this, so I’ll leave that task as an exercise for the reader.
The simplest improvement you could make would be to drop one of the 3-vote states (presumably Montana, the largest of them), leaving him at exactly 270. Though I won’t guarantee that there isn’t some better option out there.
…wow. Yeah, that’d work, wouldn’t it? Even considering we’re in GQ, I owe you one for the tact in that correction; I was due at least a 4 on the snark scale.
If you look at the list of states in order of percentage points for Obama as seen here, you’ll see that Colorado was the tipping point, which works in both directions. So, let’s say there were just enough people in FL, VA, OH, and CO who switched votes to tip them for Mitt, the number needed would be 203,178, which is half the difference in votes in each state plus one. (But keep in mind that’s using numbers where final numbers are not yet verified due to stuff like absentee and mail in ballots).
So, if Mitt had convinced the right 203,178 people to switch votes, he’d be President-Elect right now. Alternatively, if he couldn’t convince them to switch votes, he’d have to convince the right distribution of 406,352 eligible voters in those states who didn’t vote, to have gotten out to vote for him.
Of course, with such close results, we’d be counting and recounting for a month.
Colorado may have been closer percentage-wise, but NH has far fewer voters, meaning far fewer votes the Mittster would have to flip. I still contend that for this exercise, NH is the state to use.
