Statistikhengst
Diamond Member
- Banned
- #1
Trump Would Lose Badly In A Third-Party Bid But He Could Take The Republican Down Too FiveThirtyEight
A lot of what Nate says is just plain old common sense, but he does give this some historical context.
It's a good read, overall.
Nate does one thing, however, that is bothersome. In comparing a third party candidates initial polling value to his actual performance, he only relies on Gallup's polling data. Well, after Gallup's horrible track record from both 2010 and 2012, I can't recommend that. I recommend much more the aggregate approach in measuring the strength of candidates in polling.
However, his point stands, namely, that of the FIVE third-party candidacies since and including 1948, excepting Wallace in 1968, each third party candidate did worse in the actual results than in his first polling values, at least from Gallup. He also puts out a chart about that. So, if Trump is at 20%, he could land at maybe 11% come election time, were a three-way race to develop.
In all of this is one more data-point that Nate didn't speak about, one that pointed to the historical electoral volatility of much of the deep South. So, I'm gonna bring it up:
In 1948, States-Rights nominee Strom Thurmond only got 2.41% of the NPV (Silver lists it at 2%), but he racked up 39 electoral votes, from: South Carolina, Alabama, Mississippi and Lousiana. That's pretty good bang for his buck, I would say. But Thurmond wasn't on the ballot in most states. He was truly a regional phenomonen. And BTW, there was a FOURTH party candidate that year, Henry Wallace (Progressive Party), who got 2.37% of the NPV and absolutely no electors. He was on the ballot mostly in the big sky states. Without going to the 100th of a percentage point, officially, both candidates were at 2.4% that year.
Fast forward 20 years, to 1968. American Independent Party nominee George Wallace was the third party candidate on the ballot, he got 13.53% of the NPV (which Nate Silver erroneously rounds down to 13%), but he too only won states in the Deep South, then worth 46 EV: Georgia, Alabama, Mississippi, Louisiana and Arkansas. He got less bang for his buck in spite of 7 electors more than Thurmond, because was on the ballot in considerably more states: every state except Maine and Hawaii, also DC. Wallace was also very close to Nixon in Tennessee, he almost took the state. Outside of the deep South, his best showing was in Maryland and in Nevada and generally, the farther north on the electoral map, the worse he did.
Fast forward 24 years, to 1992. Ross Perot (Reform Party) took 18.91% of the NPV and absolutely zero electoral votes. His two best showings were in Maine and Alaska and in both cases, he almost unseated the 2nd place candidate (Bush in Maine, Clinton in Alaska). But in spite of a boatload of money, he received no electoral votes.
All three of these men (Thurmond, Wallace and Perot) were third-party candidates from the South.
So, these kinds of figures tells us what a long, hard, climb a third party candidate has these days just to win some states in the Electoral college, much less piece together a national victory. The only time since 1856 that a party other than the Democrats or Republicans was in 2nd place was in 1912, with former Pres. Theodore Roosevelt, then as a "Bull-Moose" candidate. And he did so well that Republican incumbent Taft only won 2 states, worth all of 8 EV: Vermont and Utah.
So, Nate's postulation that Trump, as a third-party candidate, would lose badly, is correct. And considering the type of voters he is attracting, it is logical to assume that in a three-way race, he is going to take more of the GOP nominee's voters than Hillary's voters. However, considering the anger and volatility on the Right, I could see a possibility that Trump could actually move into 2nd place against Hillary, leaving the Republican nominee a distant third, ala 1912.
More info at the link. Enjoy.
FYI, the poll is good for 365 days, until 23 July 2016, AFTER the GOP convention...
A lot of what Nate says is just plain old common sense, but he does give this some historical context.
It's a good read, overall.
Nate does one thing, however, that is bothersome. In comparing a third party candidates initial polling value to his actual performance, he only relies on Gallup's polling data. Well, after Gallup's horrible track record from both 2010 and 2012, I can't recommend that. I recommend much more the aggregate approach in measuring the strength of candidates in polling.
However, his point stands, namely, that of the FIVE third-party candidacies since and including 1948, excepting Wallace in 1968, each third party candidate did worse in the actual results than in his first polling values, at least from Gallup. He also puts out a chart about that. So, if Trump is at 20%, he could land at maybe 11% come election time, were a three-way race to develop.
In all of this is one more data-point that Nate didn't speak about, one that pointed to the historical electoral volatility of much of the deep South. So, I'm gonna bring it up:
In 1948, States-Rights nominee Strom Thurmond only got 2.41% of the NPV (Silver lists it at 2%), but he racked up 39 electoral votes, from: South Carolina, Alabama, Mississippi and Lousiana. That's pretty good bang for his buck, I would say. But Thurmond wasn't on the ballot in most states. He was truly a regional phenomonen. And BTW, there was a FOURTH party candidate that year, Henry Wallace (Progressive Party), who got 2.37% of the NPV and absolutely no electors. He was on the ballot mostly in the big sky states. Without going to the 100th of a percentage point, officially, both candidates were at 2.4% that year.
Fast forward 20 years, to 1968. American Independent Party nominee George Wallace was the third party candidate on the ballot, he got 13.53% of the NPV (which Nate Silver erroneously rounds down to 13%), but he too only won states in the Deep South, then worth 46 EV: Georgia, Alabama, Mississippi, Louisiana and Arkansas. He got less bang for his buck in spite of 7 electors more than Thurmond, because was on the ballot in considerably more states: every state except Maine and Hawaii, also DC. Wallace was also very close to Nixon in Tennessee, he almost took the state. Outside of the deep South, his best showing was in Maryland and in Nevada and generally, the farther north on the electoral map, the worse he did.
Fast forward 24 years, to 1992. Ross Perot (Reform Party) took 18.91% of the NPV and absolutely zero electoral votes. His two best showings were in Maine and Alaska and in both cases, he almost unseated the 2nd place candidate (Bush in Maine, Clinton in Alaska). But in spite of a boatload of money, he received no electoral votes.
All three of these men (Thurmond, Wallace and Perot) were third-party candidates from the South.
So, these kinds of figures tells us what a long, hard, climb a third party candidate has these days just to win some states in the Electoral college, much less piece together a national victory. The only time since 1856 that a party other than the Democrats or Republicans was in 2nd place was in 1912, with former Pres. Theodore Roosevelt, then as a "Bull-Moose" candidate. And he did so well that Republican incumbent Taft only won 2 states, worth all of 8 EV: Vermont and Utah.
So, Nate's postulation that Trump, as a third-party candidate, would lose badly, is correct. And considering the type of voters he is attracting, it is logical to assume that in a three-way race, he is going to take more of the GOP nominee's voters than Hillary's voters. However, considering the anger and volatility on the Right, I could see a possibility that Trump could actually move into 2nd place against Hillary, leaving the Republican nominee a distant third, ala 1912.
More info at the link. Enjoy.
FYI, the poll is good for 365 days, until 23 July 2016, AFTER the GOP convention...
Last edited:
