Bernie Sanders And The Fifth Early State
The Democratic primary includes 57 states and territories with a total 3768 pledged delegates. To win the nomination in the first round you need 1885 of those delegates. Then the super-delegates don’t get a say. 416 of those delegates are up for grabs in California, almost double the next state, Texas, and accounting for 11% of the delegates overall, down from 14% in 2016. Because California goes so early in the race, with voting starting on the same day as the Iowa caucus, its effectively an early state. Its also by far the most expensive state and its a huge hurdle for campaigns who will not yet have had a break out media moment by doing well in one of the smaller traditional early states.
Bernie Sanders has signed up 400,000 volunteers, has 250,000 donors, held 1000 kickoff events, and plans to spend at least $25 million dollars in California during the primary. That means that, based on turnout for 2008 and 2016 of roughly 5 million voters both times, 8% of voters have pledged to volunteer for Bernie and 5% have donated money. No other candidate is even close. California accounted for 14% of the voters in 2008 and 17% in 2016. Its a massive and expensive state usually handled by a deluge of TV ads. Its going to be almost impossible for other candidates to spend as much money in California as Bernie Sanders, much less equal his volunteer operation.
Bernie Sanders And The 15% Solution
The Democratic primaries operate on proportional allocation. Every congressional district receives a share of pledged delegates that will be awarded to candidates who get greater than 15% of the vote in the state. California has 53 congressional districts and 272 pledged delegates by district. Some districts will receive more delegates than others, though all will be worth at least 4 delegates. There are an additional 90 at-large delegates allocated among candidates who receive 15% or more of the statewide total and 54 PLEOs, basically at large delegates but the slots are reserved for party officials, who are required to pledge to a candidate when they apply.
Delegates at the district level are allocated by a simple formula:
- Remove the votes of candidates under 15%
- Calculate the new vote share
- Multiply the % shares by the number of delegates
- Round the values to whole numbers
- Record whole number delegates
- Award delegate to candidate with highest fractional remainder
4 Delegate District:
|Candidate||Vote Share||Adjusted||Fraction x 4||Whole|
5 Delegate District:
|Candidate||Vote Share||Adjusted||Fraction x 5||Whole|
We can see from these models that depending on the district there are different cutoffs for delegates once you account for non-qualifying vote share. The math is below for two candidates:
The math starts to get a lot more complicated as you add more candidates because the % cutoff to get the highest fractional share starts to change.
Bernie Sanders And The 60% Super-majority
As we discussed in the previous section, due to the 15% threshold not only do small margins often add up to big benefits, but you will almost always win a larger % of delegates than your vote share, assuming you are above 15% state wide. California is effectively an early state which means that most of the candidates will still be actively campaigning or on the ballot when the majority of the votes are cast. While many candidates will be at 1% or less, with 24 candidates those votes add up. It’s very plausible to see statewide numbers where 30% or more or more of the votes will be set aside.
Here’s the polling in California so far:
The share of votes among qualifying candidates is 72, 61, 62, 67. An average of 34.5% of votes are excluded. Now obviously things can change as the election draws near but its hard to say how much of the vote will be to qualifying candidates for sure. While its possible that some of the *Other* votes move, whether from undecided or weak candidates, it could easily move in a way that kept the 4th and 5th or even 3rd candidate from qualifying. We’ll use these numbers for out analysis.
Essentially every % of the vote is multiplied by 1.5 in this case. If we assume that Bernie gains ground we can use a hypothetical result of:
- Bernie 40%
- Biden 25%
- Harris 12%
- Warren 10%
- Buttigieg 8%
- Other 5%
This would give us delegate split of:
- Bernie 62%
- Biden 38%
- Other 0%
Of course if 2 other candidates qualify Bernie can get a maximum of 54% of the delegates with 34.5% of the votes not qualifying. He could still win pluralities with 3 candidates qualifying. 30% of the delegates with 25% of the vote for instance. Of course he could also get 30% and get 40% of the delegates if 75% of the votes were for qualifying candidates. There’s a lot of ways it could shake out. We could see Bernie at 40% and 2 other candidates at 20%, awarding him 50% of the vote. He could break 65% by getting 50% of the vote with 20% of the vote to non-qualifying candidates. Indeed as we discussed earlier if he got efficient vote share margins at the district level he could actually do even better than the numbers using statewide totals would suggest.
You can see what the delegate allocation was for each district in 2016 and view the final results here.
Its Not Called The Golden State For Nothing
California is a massive state. I talked about how 5 million people voted in both of the last two presidential primaries in California. 8.5 million voters voted for the Democrats in those general elections. 12.5 million voted over all. California has roughly 25 million eligible and 19 million registered voters for 2020. Not only can Bernie reach the 2.4 million people who voted for him in 2016, but polling says that 14% of Clinton voters nationally now support Bernie. 46% of his voters from 2016 currently support him. He only needs 45% of the vote to get 60% of the delegates. There is a massive pool of voters out there who have never voter in the primary election.
If every person who has already volunteer to work for Bernie were to hit 100 doors and no one hit the same door that would be 40 million doors. There are 40 million people in California and they don’t all live in one person housing. Similarly only 25 million are eligible to vote. I hit more than 1000 doors in Iowa and Missouri in 2016. Its very plausible to average 800 doors per volunteer when you account for super volunteers. If you speak to someone 25% of the time that is 80 million voter contacts. That number is supported by reporting by 538 as being accurate: https://fivethirtyeight.com/features/voter-contacts/
Even if turnout is very high in 2020 you will likely only need 4 million votes to win 60% of the delegates. Bernie has 400,000 volunteers so far. He has 1.2 million people currently planning to vote for him according to polls. So he needs about 2.8 million more votes at the higher end. So 3.5% of voter contacts, or 1 in 28, need to result in a new supporter. Of course this is lowered by TV spending and online work and individuals talking to friends and family.
Assuming a 40 door sheet and 10 contacts per sheet each volunteer, using 400,000 as our total, would need to walk 20 sheets to hit the goal. We have roughly 8 months, or 32 weeks before the votes are in in California. I’ve walked as many as 3-4 routes a day but 2 might be more plausible for a weekend. That’s actually 64 sheets, or 3.2x as many as needed. Of course not every volunteer can walk every week. So 20 per volunteer is pretty reasonable. This is all envelope math and I’m sure a pro would absolutely reject this level of volunteer activity. The campaign would have to carry the weight for this to be successful.
The amount of money for ads and staff support and rallies and events is immense, hence the heading on this section. $25 million? Personally I’d prefer $40 million but it depends on how much people are willing and able to contribute. Even with a solid app for canvassing vs clipboard management the data entry implications are enormous. You also need to manage the different situations in urban vs suburban plus rural canvassing. You can knock on a lot more doors in 8 hours in the city or the inner ring suburbs than you can in the exurbs or the rural areas.
I had previously set my delegate target for California at 35% of the delegates. However because Bernie has so much early support and the field is so crowded with 7 candidates polling above 1%, there’s a lot more of a chance for vote total reduction. A lot of candidates are planning to rely on early state victories and momentum to boost their chances in California. Warren is behaving this way and Castro has outright stated that that is his plan.
That is incredibly unlikely to be effective for low name recognition candidates because of the massively front loaded primary schedule. The 4 delegate richest states this year account for 1087 delegates which is 29% of the total as of the current primary allocation. 863 of those are delegates are available on Super Tuesday when the polls close on California, Texas, and Florida. 34% of all delegates will be awarded by the end of Super Tuesday. Another 23% of delegates will be awarded by March 17th.
With 57% of all the delegates awarded 6 weeks after Iowa votes, the race will be all but over. The Democrats have done Bernie Sanders a huge favor and they don’t even know it. In the end Bernie has about a 30% chance to hit 60% of the delegates in California. That’s the same chance Donald Trump had to become president. It all comes down to volunteer organization and campaign leadership. California might cast its shadow over the 2020 primary, but that just means Bernie Sanders and his volunteers, 400,000 now and probably more by the time Iowa begins to vote, will walk the block in the shade.