The Establishment Math Problem
Please forgive the incoming rant. This is not an endorsement of Donald Trump (not by a long shot). However, what it is is an indictment of (deliberately) bad math skills on the part of punditry attempting to show that Trump’s path to the nomination remains very shaky.
Caveat 1: A moderate approach to looking at the numbers shows us that Trump will secure the nomination (1237) on the final day of contests, June 7. This makes the assumptions that current numbers remain equal.
Caveat 2: These numbers are based on polling and inference from previous results. Nothing here is set in stone.
Here’s how the current GOP delegate math works out (after the contests of 3/15 but without the addition of all of the delegates from Missouri where allocations are slow in coming because of the relative competitiveness of the race):
Other (Dropout or Uncommitted): 209
This leaves 1,013 delegates left to be decided in the race among the following states:
Arizona (58), Utah (40), Wisconsin (42), New York (95), Connecticut (28), Delaware (16), Maryland (38), Pennsylvania (71), Rhode Island (19), Indiana (46), Nebraska (36), West Virginia (34), Oregon (28), Washington (44), California (172), Montana (27), New Jersey (51), New Mexico (24), and South Dakota (29).
As we’ve discussed numerous times on this blog before, each state has its own rules for delegate allocation. However, they generally fall into three categories: Proportional, Winner Take All, or Winner Take All Statewide and by District. This final type means that the total state delegates are divided among those awarded to the statewide winner and those awarded to the winner of each of the state’s congressional districts.
Of the states that have currently allocated delegates, only Florida, Ohio, South Carolina have adopted a winner-take-all or winner take all statewide and by congressional district model. Other states have set proportional thresholds and allocated a supermajority of delegates to a candidate with the most votes in a given state or district, but not on a “winner take all” basis.
Now, here’s where the faulty logic comes in. There have, up to this point, been 1,459 delegates awarded of the 2,472 total. Pundits are quick to point out that Trump has only won 47.6% of the delegates awarded so far. They are correct. However, let’s take a look at the calendar.
Of the remaining states, above, the following are either winner-take-all or winner-take-all statewide and by congressional district: Arizona (58), Wisconsin (42), Delaware (16), Maryland (38), Pennsylvania (71), Indiana (46), Nebraska (36), California (172), Montana (27), New Jersey (51), and South Dakota (29). That’s 586 delegates, or, put another way, 58% of the remaining delegates.
The other 42% of the remaining delegates are awarded on a winner-take-most system (if there are 3 delegates to be awarded in a congressional district, the winner gets 2 and second place gets 1). The exceptions here are New Mexico, Washington, and Oregon that award delegates proportionally.
So, what’s the problem? The current punditry notes that Trump would have to win ~55% of the remaining delegates in order to get over 1237 by the end of the primaries. They cite that he has only gotten 47.6% as an indication that previous trends will hold. While this could certainly happen, the fact that the delegate award model has fundamentally shifted after March 15th means that this logic is, at best, faulty, and, at worst, deliberately misleading
Donald Trump needs 546 delegates to secure the GOP nomination. By winning only one more vote than Ted Cruz and John Kasich in any of the winner-take-all states, Trump will get delegate windfalls. (Again, as I mentioned in the first caveat, this is assuming current polling doesn’t skew in one direction or the other).
Let’s apply the current winner-take-all model retrospectively. Trump would have approximately 889 delegates (assuming 0 delegates in states he did not win and full slates in states where he was victorious). As you can see, he would have approximately 30% more delegates following this model. Again, if we assume that current polls hold, we would expect the winner-take-all models to drastically increase Trump (or any candidate’s) delegate allocations in relatively short order.
Using the calendar, we should still project, conservatively, that Trump would only get to 1237 on June 7th, but that his final delegate allocation would be north of 1,300. This analysis is subject to change, but you should not believe the pundits indicating that past is prologue.