Tuesday, October 18, 2016

Could a 3rd Party Deadlock the Election?

Election Deadlocks Explained


According to the 12th Amendment to the U.S. Constitution, if no presidential candidate receives a majority of the Electoral College votes, the election is considered "deadlocked" and the decision goes to the House of Representatives to decide the president (among the top 3 candidates) and the Senate to decide the vice president (among the top 2 candidates).

This has only happened twice: once in 1800 before the amendment and again in 1824 during the crazy 4-way race between John Quincy Adams, Andrew Jackson, William Crawford, and Henry Clay, all of whom took at least 2 of the then 24 states.

One goal of running modern third-party campaigns could be to intentionally cause a deadlock so that the 3rd party can attempt to play kingmaker in exchange for policy concessions. This was George Wallace's plan in 1968 and, although he failed, my election simulator shows he actually had a 18% chance of success!

Enter Evan McMullin


Over the last few days (mid-October) polls have taken a surprising twist: dark horse independent candidate Evan McMullin has emerged as a serious contender in Utah, polling at ~22% of the state’s popular vote. In case you haven’t been following the story, McMullin is an ex-CIA, Mormon, moderate Republican who is offering himself as an alternative to both Hillary Clinton and Donald Trump.

McMullin as a Republican Spoiler


I’ve gone ahead and simulated McMullin from the perspective of a third party spoiler effect. As a caveat, I want to point out that at the moment there is very little polling data on McMullin outside of Utah and Virginia. The Federalist provided me with a list of states McMullin has a decent chance in, which I then corrected against states where he is either on the ballot or a legal write-in. For states that fell into both categories, I’ve provisionally given McMullin all the polling percentage not allocated to Clinton, Trump or Johnson. This crude estimate will have to do for now.

Unlike Johnson and Stein, McMullin is a potential spoiler for the Republicans. But as usual with third-party candidates, his partisan bias is not as extreme as I’ve seen quoted in online sources. I’ve seen lean factors suggested between 50-100%, but because we have very solid data polling data in Utah from both before and after McMullin entered the race, I can state with some accuracy that his base shows only a 21% conservative lean. That said, it may be more extreme outside of Utah.

As a result, I’m showing McMullin with a 1.34% spoiler chance. Interestingly, Utah is not likely to be at the heart of a spoiler. Pennsylvania, Virginia, Minnesota, Ohio, Michigan, and Colorado would be more likely.

Could McMullin Deadlock the Election?


FiveThirtyEight discusses the obscure possibility that McMullin could become president. The three stage plan would involve (1) McMullin winning Utah, (2) causing a deadlock where neither candidate has the necessary 270 electoral votes, forcing the decision to the House where, (3) McMullin can negotiate himself as a compromise selection.

But how likely is that?

I can’t answer for the backdoor negotiations that would occur, but I can check the odds of McMullin winning Utah and causing a deadlock. I’ve added that question into the simulator, but since data is currently scarce, I’ll look at the results in three different scenarios.

The odds that McMullin could deadlock the election:

  1. Based on McMullin’s recent average of Utah polls: 0.005%
  2. Based on the not-especially-reliable but most McMullin-generous Utah poll: 0.460%
  3. Making the assumption that McMullin wins Utah: 1.811%

Ranking States by Deadlock Power


The Utah deadlock value above leads me to another question:

If you were a third party presidential candidate hoping to cause an electoral deadlock in 2016, and you could only choose one state to win (presumably your home state), which would it be?

Well, after returning to the simulator, the answer is... not Utah. It’s ranked 31st for deadlock potential. Obviously, states with more electoral votes have a better chance, but it isn’t quite so simple. This election, Democrat-leaning states make better targets, since removing them does more to “even the odds” towards a red-blue split.

For each state, I've calculated the chance of a 2016 election deadlock, were it to be captured by a third party:
  1. California (55):   27.886%
  2. New York (29):   12.801%
  3. Florida (29):   10.961%
  4. Texas (38):   8.458%
  5. Illinois (20):   8.252%
  6. Pennsylvania (20):   8.095%
  7. Michigan (16):   6.288%
  8. Ohio (18):   6.134%
  9. New Jersey (14):   5.384%
  10. North Carolina (15):   5.263%
  11. Georgia (16):   4.945%
  12. Virginia (13):   4.790%
  13. Washington (12):   4.577%
  14. Massachusetts (11):   4.144%
  15. Maryland (10):   3.689%
  16. Minnesota (10):   3.652%
  17. Wisconsin (10):   3.562%
  18. Arizona (11):   3.553%
  19. Indiana (11):   3.374%
  20. Tennessee (11):   3.227%
  21. Colorado (9):   3.041%
  22. Missouri (10):   2.930%
  23. South Carolina (9):   2.699%
  24. Alabama (9):   2.669%
  25. Louisiana (8):   2.461%
  26. Oregon (7):   2.387%
  27. Kentucky (8):   2.373%
  28. Connecticut (7):   2.367%
  29. Oklahoma (7):   2.076%
  30. Nevada (6):   1.864%
  31. Utah (6):   1.811%
  32. Iowa (6):   1.799%
  33. Kansas (6):   1.789%
  34. Arkansas (6):   1.737%
  35. Mississippi (6):   1.688%
  36. New Mexico (5):   1.584%
  37. West Virginia (5):   1.402%
  38. New Hampshire (4):   1.151%
  39. Hawaii (4):   1.111%
  40. Rhode Island (4):   1.108%
  41. Idaho (4):   1.048%
  42. Delaware (3):   0.765%
  43. Vermont (3):   0.752%
  44. North Dakota (3):   0.743%
  45. South Dakota (3):   0.743%
  46. Wyoming (3):   0.726%
  47. Alaska (3):   0.710%
  48. Montana (3):   0.710%
  49. D.C. (3):   0.709%
  50. Nebraska - State (2):   0.376%
  51. Maine - State (2):   0.351%
  52. Maine - 1st D (1):   0.006%
  53. Nebraska - 1st D (1):   0.005%
  54. Maine - 2nd D (1):   0.004%
  55. Nebraska - 3rd D (1):   0.004%
  56. Nebraska - 2nd D (1):   0.001%

No comments:

Post a Comment