A lot has been written about the seismic shift in the 2020 Democratic Primary starting with South Carolina and continuing through Super Tuesday. In the aftermath, some have correctly noted that a proportional delegate race is a long and drawn out process. As it stands this morning per The Green Papers Joe Biden (681) has a 73 pledged-delegate lead over Bernie Sanders (608) with just under two-thirds of pledged delegates still to be allocated in the race to a 1,991 pledged delegate majority. Some of the remaining 63.11% of pledged delegates outstanding are to waiting to be allocated from Super Tuesday, but most will be decided in the upcoming remaining contests.
Over the past week I’ve seen variations of comments that Biden’s lead isn’t that significant i.e. “It is only 73 delegates” or “There are still 10 states left that have more delegates in their state than the gap between Sanders and Biden.” I personally think this line of thinking under-estimates how difficult it can be to make up ground in a proportional delegate race. Joe Biden’s current lead means as of this morning, he needs 52.17% of the remaining pledged delegates to reach the 1,991 majority. Sanders would need 55.08%. That may look close to some people too, but it is the difference between Sanders needing to win a state on on average by 10% from here on out versus Biden only needing to win any given state by 4%. That is fairly significant.
Others have noted that it is still early on in the process. With just over 1/3 of the delegates having been awarded I agree, but the window starts to close quickly. 365 delegates will be awarded in contests tomorrow bringing the cumulative total up to 46.8%. By next week on 3/17 that rises to 61.5% with only 38.5% left to try and obtain. Dealing with large delegate numbers, smaller delegate differences and constantly changing math can sometimes obscure the picture of a pledged delegate race. There is a different way of looking at the race, one that can at times provide a clearer picture.
Fundamentally for any given candidate at any given time we can ask are they on track to win the majority of pledged delegates? In 2016, Dave Wasserman of the Cook Political Report teamed up with Five Thirty Eight to create a demographic target model of the primary called Who’s On Track For the Nomination? The model started with the assumption the race was tied nationally and asked in a given state based on demographics, what number of delegates would you expect Sanders or Clinton to get if they were performing on track for the nomination? It is essentially a race versus yourself within the race. A candidate could win a given state but fall behind because they failed to win it by their target. Similarly a candidate could lose a given state by less than expected and actually come out ahead in the model.
Fellow Daily Kos Community member Subir continued to update 538’s original work throughout the 2016 race here not only to see where his candidate stood, but to update Sanders’ target numbers to account for his accumulated delegate deficits in the model. We were fortunate enough that Subir attempted to replicate 538’s work for Sanders’ 2020 path in a excellent diary two weeks ago "Nevada opens the door to Bernie winning a majority of pledged delegates."
Subir’s stated methodology was was follows:
My tabulation of the upcoming primary calendar is below and factors in various things including Bernie’s relative strength in the state in 2016 and current polling. To win on the first ballot, a candidate will need 1,990 pledged delegates at the convention. I’ve factored in a buffer, with a target of over 2,000 delegates.
Intuitively by nearly every metric, Super Tuesday seemed bad for Sanders. I was curious on how the Senator from Vermont performed last week by looking at the race in this manner.
The Early States
| Date |
State |
Pledged Delegates |
Sanders Expected based on Polling |
Subir’s Sanders Target |
Sanders Actual Result |
Pace Difference |
Cumulative +/- Pace |
| 2/3 |
IA |
41 |
19
|
16 |
12 |
-4 |
12 |
| 2/11 |
NH |
24 |
12 |
12 |
9 |
-3 |
9 |
| 2/22 |
NV |
36 |
15 |
15 |
24 |
9 |
18 |
| Total |
|
101 |
46 |
43 |
45 |
|
18 |
|
Columns 1-3 and 6 are from The Green Papers. Columns 4-5 are from Subir’s projections. The last two columns are the cumulative math. A positive number in the cumulative +/- pace column means at any given time Sanders is on pace to win the Democratic nomination. The only change I made was that Subir’s diary had Sanders needing 2,007 delegates as a cushion. I gave Bernie 16 extra delegates off the top to reflect that he only needs to get to 1,991. This obviously is something that benefits him.
As you can see, overall Sanders was on track in the first 3 states due to an excellent performance in Nevada. Sanders’ was 2 delegates ahead of his target in Subir’s diary and is 18 delegates ahead of his target at this point (again reflecting the 16 delegates I gave him off of the top).
South Carolina and Super Tuesday
| Date |
State |
Pledged delegate |
Sanders expected based on polling |
Subir’s sanders target |
sanders actual result |
pace difference |
cumulative +/- Pace |
| 2/29 |
SC |
54 |
19 |
22 |
15 |
-7 |
11 |
| 3/3 |
AL |
52 |
17 |
24 |
8 |
-16 |
-5 |
| 3/3 |
AR |
31 |
11 |
15 |
9 |
-6 |
-11 |
| 3/3 |
AS |
6 |
2 |
3 |
0 |
-3 |
-14 |
| 3/3 |
CA |
415 |
176 |
233 |
220 |
-13 |
-27 |
| 3/3 |
CO |
67 |
32 |
41 |
20 |
-21 |
-48 |
| 3/3 |
MA |
91 |
39 |
52 |
29 |
-23 |
-71 |
| 3/3 |
ME |
24 |
12 |
15 |
9 |
-6 |
-77 |
| 3/3 |
MN |
75 |
36 |
46 |
27 |
-19 |
-96 |
| 3/3 |
NC |
110 |
45 |
60 |
37 |
-23 |
-119 |
| 3/3 |
OK |
37 |
17 |
22 |
13 |
-9 |
-128 |
| 3/3 |
TN |
64 |
24 |
33 |
20 |
-13 |
-141 |
| 3/3 |
TX |
228 |
86 |
117 |
102 |
-15 |
-156 |
| 3/3 |
UT |
29 |
16 |
20 |
12 |
-8 |
-164 |
| 3/3 |
VA |
99 |
38 |
51 |
31 |
-20 |
-184 |
| 3/3 |
VT |
16 |
9 |
11 |
11 |
0 |
-184 |
| Total |
|
1499 |
625 |
808 |
608 |
|
-184 |
Disclaimer: these delegate allocations particularly in the case of Utah and California aren’t final as votes are still being counted there. +/- A few delegates here or there doesn’t materially change the outcome or analysis though.
This shows just how far Senator Sanders got off track on Super Tuesday in a way that saying he is only down 73 delegates doesn’t. Bernie only hit Subir’s target once in his home state of Vermont where he merely met it, not exceeded it. In every single other state he failed to meet that target, more often than not missing it by 10+ delegates. Victories in Colorado or Utah weren’t what they needed to be due to multiple candidates above delegate viability. Losses in the South were worse than they needed to be as Biden benefited from a more contracted field. And lastly, there were flat out states like Minnesota or Massachusetts that Sanders needed a win and didn’t get.
What’s Next?
The other implication (and this is where Subir did a great job of keeping this updated in 2016) is that in the remaining roughly 2/3 of contests is that Sanders would need to make up his 184 delegate deficit in this model. Merely hitting Subir’s targets (again something Sanders failed miserably at on Tuesday) means Sanders is still 184 delegates off path.
In the remaining contests Subir’s initial target was 903 pledged delegates for Sanders. The new target would be 1,087 (903+184). This means Sanders’ needs to outperform Subir’s initial targets by about 20.4% percent in the remaining contests. In this model things start to snowball right about now unless something major changes quickly. The next week and a half looks something like the following.
| Date |
State |
pledged delegates |
Subir Target |
New Target |
| 3/10 |
Abroad |
13 |
8 |
10 |
| 3/10 |
ID |
20 |
13 |
16 |
| 3/10 |
MI |
125 |
68 |
82 |
| 3/10 |
MO |
68 |
37 |
45 |
| 3/10 |
MS |
36 |
15 |
18 |
| 3/10 |
ND |
14 |
8 |
10 |
| 3/10 |
WA |
89 |
48 |
58 |
| 3/14 |
NoMA |
6 |
3 |
4 |
| 3/17 |
AZ |
67 |
33 |
40 |
| 3/17 |
FL |
219 |
103 |
124 |
| 3/17 |
IL |
155 |
81 |
98 |
| 3/17 |
OH |
136 |
69 |
83 |
| 3/24 |
GA |
105 |
46 |
55 |
Instead of needing to win Michigan with 54.4% of the delegates Sanders would now need to win it 65.6% of the delegates. Instead of a 6% loss in Florida (47% of the delgates) he needs to win Florida by more than 10 points, 56.6%. Any time one of those things doesn’t happen, the targets can be rebalanced and the remaining states get more difficult again. Looking at the race in this manner shows just how precarious of a situation Sanders finds himself in post Super Tuesday without even even factoring in what appears to be highly unfavorable territory ahead.
