Blog See all Articles

That Favorite Son scenario, which wouldn’t work

Some NeverTrumpers have been considering a plan that involves throwing the presidential election into the House of Representatives through the use of a variation on Favorite Son candidacies.

A Favorite Son is a candidate, usually at a political convention, who carries his or her home state or district and uses that support as a bargaining chip. Wikipedia:

At the quadrennial American national political party conventions, a state delegation sometimes nominates and votes for a candidate from the state, or less often from the state’s region, who is not a viable candidate. The technique allows state leaders to negotiate with leading candidates in exchange for the delegation’s support.

In a new spin on the concept, Favorite Sons would be used to snatch states away from the major-party candidates, ideally (from the NeverTrump point of view) denying those states’ Electoral Votes to Clinton. Rick Santorum would run as an independent or third-party candidate in Pennsylvania, John Kasich would do likewise in Ohio, and others would run in other states. Remember: If neither Clinton nor Trump gets 270 Electoral Votes, the GOP-controlled House of Representatives, each state delegation having one vote, picks the president from the top three Electoral Vote finishers. (See below.)

A Capital Research Center supporter asked us to examine the viability of the Favorite Son (“multi-candidate”) scenario. This was my response.


From: Dr. Steven J. Allen
Sent: Tuesday, July 12, 2016 12:53 PM
To: [redacted]
Subject: Multiple candidate scenario (RE: question from [redacted])

> I’m simplifying things slightly, leaving out twists and turns that are extremely unlikely to happen this year – e.g., a high number of “unfaithful electors” who break their promises and vote for someone other than their parties’ nominees.

> For purposes of discussion, I’m assuming that Clinton and Trump are the Democratic and Republican nominees.

“Third-party” includes independent, fourth-party, fifth-party, etc.

The main points:

  1. Our presidential election system is “first past the post” by state. In other words, whoever comes in first in a given state wins that state’s electoral votes. (The exceptions: Maine, with four EVs, and Nebraska, with five, split the vote, two for the statewide winner and one for each Congressional District winner.)
  2. To win a state, a third-party candidate needs to get more votes than Trump AND more votes than Clinton. Otherwise, a third-party candidate serves mainly to take more votes from the candidate with whom he/she is more closely aligned.
  3. Based on my analysis of dozens of successful and countless unsuccessful third-party candidates, I have concluded: Except in cases in which a major-party candidate is revealed to have stolen the nomination or is otherwise involved in a _major_ new scandal (a “child molester” level scandal), or dies, the ONLY time a third-party candidate wins if when he/she is SEEN as a possible winner by the first week in October. The recent growth of early-voting probably moves that deadline back by a few weeks.  By that deadline, the third-party candidate MUST be perceived as beating or tying at least one of the major-party candidates. To reach that level in a presidential or statewide election, one must be taken seriously by the news media at a level closely comparable to the major-party candidates, AND one must participate in any debates featuring the two major-party candidates.
  4. In almost every case of a successful third-party candidate, he/she has first attained second-place position by knocking one of the major-party contenders out of the race. The race becomes “Republican vs. third-party” and the Democrats mostly vote for the third-party guy, or the race becomes “Democrat vs. third-party” and the Republicans mostly vote for the third-party guy.
  5. Ballot-access petitions typically require the signatures of 1-3% of voters. By COB today [Tuesday, July 12], the filing deadline will have passed in seven states. By COB three weeks from today, it will have passed in 26 states. Anyone not already on track to be on all or almost all state ballots at this point is not a plausible winner because he/she will not meet the criteria described above.
  6. Another problem: There are “sore loser” laws in some places that might prevent a former major-party presidential candidate from running as a third-party candidate.
  7. Given the closing of deadlines for third-party candidates to get on the ballot, and the need to be taken seriously by the news media to have any chance for consideration, only Gary Johnson would have a theoretical chance to meet the requirements I’ve outlined. With regard to being taken seriously by the news media, Johnson is helped by his libertarian views on social issues (in line with much of the media), by the fact that he and his running mate are both “serious” candidates and former governors, and _especially_ by the fact that he is perceived as taking more votes from Trump than Clinton (and the media despise Trump).
  8. In order to throw the election into the House of Representatives, the third-party candidate MUST win at least one state. Otherwise, Clinton or Trump automatically gets a majority. (That’s unless there’s a 269-269 tie AND electors don’t break the tie themselves. Republican electors tend to be establishment people, and I wouldn’t be surprised if, in a 269-269 tie, a Republican elector switched to Clinton. Certainly, the bribes would be flying.) A 269-269 tie would leave the House with only two candidates from whom to pick; a third candidate would need at least one EV to be among the top three.
  9. The more states the third-party candidate wins, the _greater_ the chance of deadlocking the Electoral College. HOWEVER, the better the third-party candidate does in the overall vote, the more lop-sided the vote tends to be in favor of the major-party candidate who’s FARTHER ideologically/politically from the third-party candidate, and the _less_ the chance of deadlocking the Electoral College.  Thus, it’s extremely unlikely for any candidate to deadlock the EC except with a mostly regional candidacy. (George Wallace, if he’d received a couple of votes more per precinct, might have pulled it off in 1968, because his votes were largely concentrated in the South.)
  10. It’s often said that the presidential election is “really” a state by state contest. It’s not. It’s a national contest, carried out at the state level. The states are basically the way we keep score. Multiple candidacies (such as in the Kasich in Ohio/Santorum in Pennsylvania scenario) would generate little more than confusion, even if the plotters could somehow get everyone to cooperate. Remember the various Stop Trump efforts involving splitting up the states during the nomination campaign? Did any of them last longer than an instant?
  11. If the election is thrown into the House, the House, with each state having one vote, must pick from the top three candidates in electoral votes—presumably, Clinton, Trump, and Mr. or Ms. X. In the multiple candidate scenario, we wouldn’t know who the Clinton/Trump alternative was until the election was over, and it could be someone who won a single state. Republicans (who would control the House delegations 33-14 with three ties, barring the effect of vacancies and special elections between now and November) would almost certainly pick Trump, who carried the vast majority of their individual districts, over Mr./Ms. X, who carried Pennsylvania or Ohio or some other single state.
  12. FYI, I’ve done thousands of Electoral College simulations. They resulted in an Electoral College _majority_ for the first-place finisher in every case in which one candidate (a) got at least 38 percent of the vote and (b) got at least two points more than the second-place finisher.

In 1860, Lincoln won 39.8% (180 out of 303 electoral votes) to Douglas’s 29.5% (12 EV), Breckinridge’s 18.1% (72 EV), and Bell’s 12.6% (39 EV). Note that, after the first-place finisher, the Electoral Vote count was based more on regionality than overall support. The same sort of thing happened in the 2015 U.K. election. After the top two parties, UKIP came in third with 12.7% but got one seat out of 650, the Liberal Democrats got 7.9% and eight seats, and the Scottish National Party got 4.7% and 56 seats.


Other key considerations:

  1. My rule-of-thumb is that a third-party candidate typically gets half his/her votes from people who would not have voted otherwise. People who stay at home or otherwise decline to vote in the presidential race obviously have no effect on the outcome between the two major candidates.
  2. My rule-of-thumb is that, with the exception of people who would otherwise not have voted, third-party votes come 2:1 or 3:1 from one major-party candidate over the other.
  3. Therefore, the math works out this way:
    1. A serious conservative/Republican-leaning candidate will take (rule of thumb) 50% of his/her vote from no one [people who would not have voted otherwise], 35% from Trump, and 15% from Clinton. Effectively, the third-party candidate takes a net of 20% of his/her vote from Trump.
    2. A serious Progressive/Democrat-leaning candidate will take (rule of thumb) 50% of his/her vote from no one [people who would not have voted otherwise], 35% from Clinton, and 15% from Trump. Effectively, the third-party candidate takes a net of 20% of his/her vote from Clinton.
    3. Voters who love Clinton or hate Trump will vote for Clinton. Voters who love Trump or hate Clinton will vote for Trump. Voters who love both will flip a coin.  Voters who hate both will mostly stay home. A few both-haters might vote for a third-party candidate, but, if that candidate is not considered a serious contender, they won’t show up at the polls or send in their absentee/early ballots, or they’ll vote only for candidates in lower-level races.

The bottom line: In my judgment, the multiple candidate (Kasich in Ohio/Santorum in Pennsylvania/etc.) plan is not viable.