A couple of days ago I wrote a diary called “An Optimistic but Possible Path to the Nomination for Bernie Sanders.” In it, I outlined a possible path to the nomination for Sanders that relied on an optimistic, but realistic view of how he could win more pledged delegates than Clinton by the end of the primaries. Below I outline how Sanders’ wins in WA, AK, and HI affected this. He blew past my optimistic expectations in two of the three states, but fell a little short in Washington based on current estimates. However, he could end up meeting the delegate goals for the night when all the math is done.
A key difference between my estimates and those at 538 is that mine have been adjusted based on past performance and several factors that are important to consider. Thus, you will find that my estimates for yesterday were much closer to the final results than 538 was.
Before I list the results and future expectations, I would like to address the lone post at 538 regarding whether or not last night changed the dynamics of the race.
Sanders had a strong week, and this has been a crazy year in politics. But there’s nothing in the recent results to suggest that the overall trajectory of the Democratic race has changed. Clinton was and is a prohibitive favorite to win the nomination.
538 — Harry Enten
Here is my beef with this statement. The problem is not that this week changed the dynamics of the race, although it certainly gives Sanders some strong momentum. It is that we have misunderstood the dynamics of the race since the beginning. Clinton’s campaign and the media have put far too much emphasis on race as the driving factor of this election. It is a factor, but so is age, sex, and region. In fact, I believe these results (where Sanders won in some heavily minority states and areas) show that the overwhelming difference between some of the states is their region. Sure, caucus vs primary and open vs closed all play a strong role as well.
States where Sanders has outperformed 538’s model:
Washington |
Caucus |
Northwest |
Utah |
Caucus |
West |
Illinois |
Primary |
Midwest |
Kansas |
Caucus |
Midwest |
Idaho |
Caucus |
West |
Alaska |
Caucus |
Northwest |
Hawaii |
Caucus |
West |
Colorado |
Caucus |
West |
Vermont |
Primary |
Northeast |
Oklahoma |
Primary |
Midwest? |
Maine |
Caucus |
Northeast |
So what have we specifically misunderstood? Clinton was always going to outperform 538’s estimates in the South and Sanders was always going to do much better than 538’s estimates elsewhere. If you look at 538’s list of states where he has outperformed their estimates and where Clinton has outperformed, there is a clear pattern where she has outperformed in southern states while he has outperformed elsewhere. There is a state here or there that brakes that pattern (particularly Ohio), but each of those could be viewed as special circumstances. In Ohio, I believe the reason this state broke the pattern is because John Kasich, as the sitting governor and obvious choice over Trump, won many independent and perhaps some Democratic votes that would have otherwise gone to Bernie Sanders. This would also explain the terrible turnout in the Democratic primary in Ohio.
True caucus vs primary plays a big role, but if that was the only factor here, Sanders would never have tied Clinton in delegates in Illinois and likely would have lost big in Michigan as well. Instead, regional differences can explain major differences between the results in southern states and elsewhere. While it is true that race has played a huge role in the south, when the region changes, those racial differences seem to have a much smaller affect. For example, exit polls and wins in states like Michigan, Alaska, parts of Washington, and Hawaii (Illinois as well, although a delegate tie) show that Sanders performs much better with minorities in regions outside of the south. The media has neglected to conduct exit polling in many of these states, making these differences harder to see. However, the exit polls in Michigan confirmed this. The argument that he didn’t win minorities in states like Alaska and Hawaii (went 70-80% for Sanders) is just denial given their very high minority representations.
Anyway, here are the updated numbers based the (now less) optimistic potential path to the nomination for Bernie Sanders:
State |
Clin
Del.
|
Sand
Del.
|
Sand Net Gain
|
Actual |
Clin
lead
|
Actual |
DATE |
Factors (+/- for Sanders)
(Regional/caucus/primary type/Momentum)
|
Alaska |
5 |
11 |
6 |
10 |
290 |
284 |
3/26 |
++ Caucus/Northwest
(Closed but same day reg)
|
Sanders’ Possible Path (with clinton at 296 delegate lead via 538)
Hawaii |
10 |
15 |
5 |
9 |
285 |
275 |
|
++ Caucus/West?
(Closed, but same day reg)
|
Wash |
20 |
81 |
61 |
47? |
224 |
228 |
|
+++ Caucus/Northwest/Open |
Wisconsin |
31 |
55 |
24 |
|
200 |
|
4/5 |
+++ North/Open/MO
(Open and same day reg)
|
Wyoming |
3 |
11 |
9 |
|
191 |
|
4/9 |
++ Northwest/Caucus/MO/Closed
(Closed, no same day reg)
|
New York |
125 |
122 |
-3 |
|
194 |
|
4/19 |
Closed/MO/Northeast/Clinton Home
(Closed, no same day reg)
|
conn |
23 |
32 |
9 |
|
185 |
|
4/26 |
++ MO/Northeast
(Closed, but day before reg for indies in person)
|
Delaware |
11 |
10 |
-1 |
|
186 |
|
|
MO/Closed |
Maryland |
48 |
47 |
-1 |
|
187 |
|
|
MO/Closed |
penn |
85 |
104 |
19 |
|
168 |
|
|
+ Northeast/MO/Closed |
Rhode Island |
8 |
16 |
8 |
|
160 |
|
|
++ MO/Northeast
(Closed, but Dems and indies can vote)
|
Indiana |
37 |
46 |
9 |
|
151 |
|
5/3 |
+++ Midwest/MO/Open |
Guam |
4 |
3 |
-1 |
|
152 |
|
5/7 |
|
W VA |
6 |
23 |
17 |
|
135 |
|
5/10 |
++ Midwest/MO
(Closed, but Dems and indies can vote)
|
Kentucky |
17 |
38 |
21 |
|
114 |
|
5/17 |
+ Midwest/MO/Closed
|
Oregon |
14 |
47 |
33 |
|
81 |
|
|
++ Birdie/MO/Northwest/Closed |
Virgin Islands |
4 |
3 |
-1 |
|
82 |
|
6/4 |
|
Puerto Rico |
35 |
25 |
-10 |
|
92 |
|
6/5 |
- Caucus/Closed/Demographics |
California |
206 |
269 |
63 |
|
29 |
|
6/7 |
++ MO/West
(Closed, but Dems and indies can vote)
|
Montana |
3 |
18 |
15 |
|
14 |
|
|
++ Northwest/Open/MO |
New Jersey |
61 |
65 |
4 |
|
10 |
|
|
+++ Northeast/MO/Openish
(Closed, but voters can change party at time of voting)
|
New Mexico |
16 |
18 |
2 |
|
8 |
|
|
Southwest/MO |
South Dakota |
5 |
15 |
10 |
|
-2 |
|
|
+++ Northwest/MO/Openish
(Open, but Republicans can’t vote)
|
North Dakota |
3 |
15 |
12 |
|
-14 |
|
|
++++ Northwest/MO/Caucus/Open |
DC |
12 |
8 |
-4 |
|
-10 |
|
6/14 |
- South/MO/Closed |
For the actual column, I used 294 instead of 296 since this is what www.thegreenpapers.com have for their numbers, but I left my original estimates alone. The exact delegates from Washington are still being worked out, but a 47 delegate lead for Sanders is pretty conservative, it could be a bit higher. Some have also put Sanders’ lead in Hawaii at 11 rather than the 9 I used here, but my understanding of the math and the AP say 9 is likely more correct. Therefore, Sanders is close to being on track according to this path.
Edit: I realized that New York does not have early voting, so I updated to reflect that but did not modify my estimates. However, i think this is a big potential factor in Sanders’ favor. I think he actually could end up with more delegates if everything goes his way, but still expect a narrow loss.