Controller’s Race, Update 6: Fears that Contra Costa’s Count Will Be Conclusive

We’ll be updating you on the updates as they occur.  Or soon after.  Or not so soon after.  Cities in the roster down below will be changed to green type as they report.

UPDATE 6: And it looks like we do have another update coming this weekend after all!  Thanks to Cal Newsroom for informing us that Santa Cruz did come in Friday evening after all — and the news was a little disappointing for Betty Yee.

Here’s what things looked like before the update:

Vote Count Percent
DEM – BETTY T. YEE 15,319 36.2%
REP – DAVID EVANS 8,325 19.7%
DEM – JOHN A. PÉREZ 8,160 19.3%
GRN – LAURA WELLS 4,330 10.2%
REP – ASHLEY SWEARENGIN 3,895 9.5%
DEM – TAMMY D. BLAIR 2,248 5.3%
WRITE-IN 64 0.15%
Total 43,562 100%

And here’s what they looked like afterwards

Vote Count Percent
DEM – BETTY T. YEE 15,724 36.10%
REP – DAVID EVANS 8,551 19.63%
DEM – JOHN A. PÉREZ 8,381 19.24%
GRN – LAURA WELLS 4,518 10.37%
REP – ASHLEY SWEARENGIN 3,994 9.17%
DEM – TAMMY D. BLAIR 2,330 5.35%
WRITE-IN 64 0.15%
Total 43,562 100.00%

So, Yee’s added votes were 405 and Perez’s 221 — a net increase of 184 compared to the 270 that had been projected.  (Actually, 1,230 more from Santa Cruz had initially been projected, but I fixed that.)

Amador County also brought in 4 more votes for Yee, 13-9.  So we’re at 188.

Add that 188 to the previous 659 and Yee’s total with five counties left is 847.  What’s left?  We’ll leave Lake County for last.

Trinity is projected to bring in 26 net votes for Yee.   (A handwritten note on its report says that it’s done, but the numbers haven’t changed since election night and the Secretary of State says it still has 521 votes outstanding, so I’m going with the SOS.)  Sutter should bring in 19 net votes for Perez and Los Angeles 5 more, meaning 24.  San Mateo’s 4 votes are expected to split 1 for Perez and 1 for Yee — if anything.  Let’s call those four counties a push and go to Lake County.

Lake County has 6,053 votes out: 5,263 VBM, 743 provisional, and 47 “other.”  It has had a 7.4% advantage for Perez so far — and since these are mostly VBMs out, that’s probably a decent estimate of what’s out there.  Multiply it out and we see an expectation of 448 more net votes for Perez — which would mean a Yee victory margin of 395.  If those five other counties with outstanding votes equal out to 0, what percentage of the Lake County vote does Perez need to tie?

Lake County has thus far cast 8,950 votes in the Controller’s race, out of  9,703 cast and counted overall — meaning that 92.24% of its ballots include a vote for Controller.  So let’s expect 5,583 of those votes to count.  Let’s toss out three provisionals as rejected, making it 5,580.  55.44% of those votes, or 3094, would be expected to go to the other four candidates, leaving 2,486 to be divided between Yee and Perez — 1,243 apiece if they split them 50-50.

So, if all but three of the ballots count and if the same proportion include votes for Controller and if the other four candidates take the same proportion of the vote as they have before — then for Perez to undo a margin of 843 (we’ll call it 844 votes to eliminate the fraction) votes he would need to win the two candidate vote by a margin of 1243+422 to 1243-422.  That’s a raw vote total of 1665 to 821, and that’s 67.0 % of the two-person vote — a 34% margin, or beating her more than two to one!

It’s possible, but not likely.  Yee took 64.7% of the final 2-person vote in Santa Cruz — but that county had had her up by 16.9% before it’s update, and Perez was only up in Lake by 7.4%.  So I still think that Yee wins by triple digits — 395 votes — but the low triple digits wouldn’t surprise me, and double digits wouldn’t shock me.  And so the likelihood of a full or partial recount rises.

UPDATE 5: This is the last update of the night — except for perhaps the next one and the one after that.  The current score is: Yee up by 659 votes.  At this point, I have to number things:

(1) What’s up with Lake and Trinity Counties?  These are the last two on FENU status.  With Contra Costa in, Lake now is the biggest prize still up for grabs, with 6,053 outstanding votes.  Trinity has 521.  Lake has a 7.4% advantage for Perez, and most of those votes are just regular VBMs.  Trinity has a 5.0% advantage for Yee.  Both are subject to invasion, if need be, by the numerically superior forces of neighboring Mendocino County.  Our plain vanilla model has Lake giving Perez 450 votes — and Trinity taking 25 of them back.

(2) Only five other counties have votes outstanding — one big, one medium, two small, and one minuscule.  The reason that Yee probably doesn’t have to worry about Lake County stealing her lead is that Santa Cruz County remains outstanding, with its 1,600 provisionals and hefty 16.9% tilt so far towards Yee.  Projection: 1230 270 more votes for the frontrunner.  Next highest CCU County is Sutter with 468 votes out, about half VBM and half provisional, and a 4.0% tilt towards Perez.  After that, Amador has 122 provisionals (and a 1.2% tilt towards Yee) and Los Angeles has 110 provisionals (and currently a 4.8% tilt towards Perez.)  Finally, San Mateo has 4 provisionals out — apparently a desperate plea for attention.  The expected additions to the margin from the last four are 23 for Perez, 1 for Yee, 5 more for Perez, and 0.

(3) So the final projection for Yee’s margin, not counting CCC corrections, IF the information on who ha and hasn’t updated is correct, is:

659 [current lead] – 450 + 25+ 1230 270 – 23 19 + 1 – 5 = 2096 1433 528
[Correcting mistakes in Santa Cruz and Sutter totals and in including previous vote twice]

But, if Santa Cruz County disappoints Yee like Marin County did, and if provisionals skew hard to Perez, it could be much closer.

(4) Who came in this afternoon?  We left off with Contra Costa coming in and Monterey attaining CCC-hood.  Since then:

  • At 3:38, Tulare went CCC
  • At 4:04, Monterey and Sierra went CCC
  • At 4:42, Madera went CCC
  • At 4:56, Ventura went CCC
  • At 5:02, Merced went CCC

(5) Who’s still at CCU status?  These are the counties, not mentioned above, who also might adjust their numbers a few in what should be random directions.  Each of these is still a threat to report more new numbers prior to its 1% check, if the COMPLETED COUNT page is wrong — as it seemingly has been at times.  Each bullet point lists counties in order of the date in June they last reported.  Latecomers Lake and Trinity are not yet included.

  • 5:   Colusa
  • 9:   Alpine, InyoMariposa
  • 10: Plumas, Tehama
  • 11: Del Norte, San Benito, San Mateo, Yolo
  • 12: Calaveras
  • 13: Amador, Imperial, San Diego, Santa Clara, Santa Cruz, Sutter
  • 16: Kings
  • 18: Kern, Sonoma
  • 19: Fresno, Humboldt
  • 20: Contra Costa, Los Angeles, Placer
 The underlined counties are the CCU-status ones reported to have outstanding ballots.  So, going into the weekend, it’s that close to being over.  And then, we’ll see about whether — and where — we’ll have a recount.

UPDATE 4:  They’re coming in fast and furious now.  The bottom line is: Yee by 667 votes — so far.  She’s running behind projections, but still barely managed to avoid a victory margin being the Number of the Beast.  Congratulations to Anhley Swearengin for finally breaking 1,000,000 votes and to Tammy Blair for finally breaking 200,000.

San Luis Obispo came in at 3:11.  (I could swear it wasn’t there at the time!)  Yee by 12.2%, 6,959 votes.

Nevada County turned CCC at 3:16.  Yee by 12.2% here as well, but with exactly twice as many votes as Perez, 6,054 to 3,027.

Monterey, which was supposed to be “completed,”  came in with another update at 3:36.  It’s still listed as CCU.  Yee by 6,170, or 12.9%.

Oh — and Contra Costa came in at 3:23.  Yee’s margin here is 346, with her percentage rising from .2% to .3%.

Except for Nevada, I don’t know whether any of those additional returns are final — but why would a county turn in intermediate returns (excluding the CCC 1%) at this point?  That just becomes cruel.

If they’re all done, then the significant remaining Perez counties are Lake and Tulare and the significant remaining Yee counties are Santa Cruz and Yolo.  (Plus Humboldt, if it isn’t done yet.  Just guessing there.)  If San Mateo, Sutter, Trinity, and Tuolumne are decisive, then things will have gotten really bad for Yee.  I wouldn’t be shocked to see these counties, which I have projected increase Yee’s lead by exactly 800 votes, instead be a draw — but I would be very surprised to see them eat up Yee’s remaining lead.

UPDATE 3:   At 1:44, Marin County reported in.  At 1:46, Los Angeles reported in as well.  The net effect was that Yee’s lead goes up to 656 votes.  I had projected her margin to receive a bump of 965 votes from Marin and a dip of 240 from Los Angeles, which would have been a net increase of 725 — putting her 624 votes behind my projection.  (It looks like her margin in Marin also dropped from 15.6% to 13.6%, which may mean that my figure had a typo.  If it was 13.6% all along, then my prediction would have been reduced by 124 votes, putting her only 500 votes behind.)  Marin is now CCC status, but is not yet clear whether Los Angeles County (still at CCU status) is done.  (One would think so, but — still.)  The results suggest that provisionals are a bit better for Perez than for Yee — something that becomes important if it also applies to Lake, Tulare, Humboldt, Santa Cruz, Yolo … and Contra Costa, whose mysterious 10,000 “other category” ballots are still the biggest fish in this pond.

Here are the current — and maybe final — figures from Los Angeles:

   JOHN A PEREZ                       Pref: DEM             198,417    27.84
   ASHLEY SWEARENGIN                  Pref: REP             175,052    24.56
   BETTY T YEE                        Pref: DEM             163,591    22.95
   DAVID EVANS                        Pref: REP             103,247    14.48
   LAURA WELLS                        Pref: GRN              38,770     5.44
   TAMMY D BLAIR                      Pref: DEM              33,717     4.73

Comparing them to the results we saw last time we checked LA, Perez gained 969 votes, Yee gained 1,062 — a net increase of 93.  If LA did have about 4900 votes, those totals fit well with that.  The nicest thing about that is that it might make Perez less likely to try a recount in LA, as it may put him even further in the hole.  BUT: this means that Yee only netted 8 votes (!) from the new ballots in Marin, which is a little terrifying for my model — but I’m not going to redo all of my provisional assumptions now!

UPDATE 2: San Joaquin goes to CCC status at 11:34.  Madera updates (still CCU) at 11:56.  Between them, they cut Yee’s lead to 555, a loss of 153 votes.  We had estimated that Yee would get 14 votes from San Joaquin and that Perez would get 70 from Madera — a net pickup for Perez of 56, so this is obviously well ahead of that expectation for him.

UPDATE 1:  Placer County reported in at 10:20 a.m.  Our prediction was that it would give Yee a net of +325 votes.  Instead, Placer gave her a net of +390 votes, putting Yee up by 708.  All eyes remain on LA (except for the ones on Contra Costa.)

Original post:

When we left the Controller’s race yesterday, which looks like it may be the closest statewide race in the history of California, Betty Yee had a lead of 318 votes.  I projected her to win by 2,400 votes — IF a bunch of untested assumptions held.  We’ve now had a few chances to test those assumptions — and they’re not as good as Yee might have hoped … but probably still good enough.  We’ll know a whole lot more once Los Angeles turns in what is rumored to be its last report — sometime today.

Contra Costa County

Contra Costa County, with Mt. Diablo being its most arresting feature, seems like a lovely place to live — but when it comes to predicting election results its role in this race scares the hell out of me.

(1) This way lies the weeds.  The faint of heart should skip this section.

Here’s where the counties stand:

FENU status: five counties — Lake, Madera, Placer, Sierra, and Trinity — still have not given any reports beyond their final election night status.  Sierra has got to be mislabeled, though, because it’s also listed as having completed its count — and its report of election results is listed as final.  So let’s presume that that’s only four counties plus one, Sierra, that we’re going to move into CCC..

CCC status: this is for counties that are supposed to be entirely done, including its review of 1% of its ballots.  (Orange County’s 1% test turned up 6 more ballots — which affected nothing.)  The 23 counties now showing CCC status (including probably mislabeled Sierra) are:

Alameda, Butte, El Dorado, Glenn, Lassen, Mendocino, Merced, Modoc, Mono, Napa, Orange, Riverside, Sacramento, San Bernardino, San Francisco, Santa Barbara, Shasta, Sierra, Siskiyou, Solano, Stanislaus, Tuolumne, and Yuba.

So that leaves 27 of the state’s 58 counties in a category.  (The hilarious saga of Santa Barbara eliminating all Democratic ballots for 2½ hours is told in yesterday’s report.)

CCU/Completed status: this includes the 15 counties that haven’t done their final review and tied the bow on the package yet, but which are believed to have already counted and turned in the estimate of all of their ballots.  This includes:

Alpine, Calaveras, Colusa, Del Norte, Imperial, Inyo, Kings, Mariposa, Monterey, [Placer], Plumas, San Benito, San Diego, Santa Clara, Tehama, and Ventura.

(Note that Placer is also listed as “FEMU” — so I’m taking it out of this category, because I think that it’s there by mistake.  The figures on its website haven’t moved since Election night.)  So now 42 of the state’s 58 counties are in a category — four of which are FEMUs.  So, twenty counties apparently remain to report additional vote totals that shouldn’t move much after the 1% recount.

I’ve listed them before, but I’ll do it again here:

Amador, Contra Costa, Fresno, Humboldt, Kern, Lake, Los Angeles, Madera, Marin, Nevada, Placer, San Joaquin, San Luis Obispo, San Mateo, Santa Cruz, [Solano], Sonoma, Sutter, Trinity, Tulare, [Tuolumne], Yolo.

(Note that Solano and Tuolumne are also listed in the CCC category, so I’m taking them out of this one — leaving 20 counties.)

(2) OK, you should probably start reading again

HOWEVER: my statistically oriented friend from UCLA, who has been keeping a spreadsheet of results, notes that Fresno, Kern, and Sonoma say that they’re done.  So let’s toss them out, leaving 17 to check:

Amador, Contra Costa, Humboldt, Lake, Los Angeles, Madera, Marin, Nevada, Placer, San Joaquin, San Luis Obispo, San Mateo, Santa Cruz, Sutter, Trinity, Tulare,  and Yolo.

I’ve previously identified a number of these as relatively insignificant to the results — though I’m taking one out:

Amador (122) — Y+1.2% = +2
Nevada (WAS 939) — Y+12.2% = ?
San Joaquin (2,687) — Y+.5% = +14
San Mateo (2,129) — P+2.5% = -55
Sutter (468) — P+4% = -20
Trinity (521) — Y+5% = +25
Tuolumne (183) — Y+2.7% = +5

Nevada doesn’t have an unprocessed ballots report, but they dumped a lot yesterday and I think they’re done.  Without them, Perez is forecast to gain about a net 30 votes out of these six counties.

That leaves 11 counties.  Four of them look good for Perez, projected (including the 3o from above) to bring in 960 more votes:

Lake (6,053) — P+7.4% = -450
Los Angeles (4.898) — P+4.9% = -240
Madera (3,458) — P+2% = -70
Tulare (2,300) — P+7.3% = -170

However: if you want to take heart for Perez, consider two things:

(1) I had forecast Fresno to come in with a net 105 votes for Perez — and they came in with a net of 280.  These are largely provisional ballots — less likely to qualify, but if they do qualify then they’re probably more likely to go for Perez.

(2) Los Angeles may have a lot more than 4,900 votes.   This is just an estimate.  If they have, say, 10,000 votes, and they have a 10% margin for Perez, then now we’d be looking at 1700 votes.

Yee had a lot of good counties out as of a day ago, but two important ones — Sacramento and Sonoma — have already been counted.  Humboldt reported 2,109 ballots (in this race), giving 419 votes to Yee — and it’s not clear how many of that 6,000 it previously had out still remain.  I’m going to guess that it has about 3,000 ballots left (meaning that this race appeared on 70% of the ballots), but the true number could be as low as 0.

Humboldt (3,000) — Y+9.3% = +280
Marin (6,200) — Y+15.6 = +965
Placer (8,499) — Y+3.8% = +325
San Luis Obispo (785) — Y+12.2% = +95
Santa Cruz (7,271) — Y+16.9% = +1230
Yolo (7,182) — Y+3.3% = 235

Yee thus would be projected to bring in 3,130 more votes from these six counties.  Subtract 960 for Perez and that’s a net increase of 2170 votes — which, with a 318 vote lead, takes her very close to 2500.

Except for one thing: I’ve left out the county with the most outstanding votes: Contra Costa.

(3) Contra Costa County may decide it all — and that’s not good

Contra Costa (14,000) — Y+.2% = +28

Contra Costa reports 4,000 provisional ballots (a lot like LA) and 10,000 “Other” ballots (neither absentee nor provisional) — and I’m really not sure what that means.  It’s a large county and a diverse one — and it is split almost exactly down the middle between Yee and Perez.  Yee has 34,930 and Perez has 34,565 — a difference of 365, or less than a .2% margin.  About 134,000 votes have been counted in the race so far, an extremely high 91% of the about 147,300 ballots counted so far.  It’s heavily Democratic, so we can presume that either Yee or Perez will be listed on about 7,500 of those ballots, if they all count.

Contra Costa is the 9th largest county in the state, in terms of ballots cast, but its largest city in the County is just a little larger than Costa Mesa, while the second and third largest are a little smaller.  If you could imagine Orange County after removing any city larger than Costa Mesa — cutting the size of the county in half — then you’d be close to imagining Contra Costa (which is a third the population of OC.)  Of the largest 20 counties in the state, only San Mateo and Santa Barbara have a smaller largest city.

The idea of the primary election coming down to Contra Costa makes me nervous, because the county seems so unpredictable. We don’t know where those uncounted ballots are coming from.  The county’s largest cities are Antioch, Concord, Richmond, San Ramon, Walnut Creek; and just below those Brentwood, Danville, Martinez, Oakley, Pleasant Hill, and Pittsburg — not ones that have the statewide recognition of those in other large counties.   If the ballots are from wealthy San Ramon, Danville, or Lafayette — or even comfortable Brentwood, Walnut Creek, or Pleasant Hill — one might expect quite different results than if they come from much less affluent Concord, Antioch, Pittsburg, or Richmond.  But those latter four relatively larger cities make up about 40% of the county — so it’s  a good bet that they may come largerly from there.

While the county overall is a liberal place, it ranges from quite wealthy to quite not wealthy — and I have no sense of the county at all when it comes to which ballots are outstanding, nor of why they had been left uncounted.  Guessing the mean of the county’s vote total overall is easy: it is split down the middle between Yee and Perez.  But the variance in the range of possible outcomes based on the nature of those outstanding ballots is huge.  If the outstanding ballots go 60-40 either way, rather than 50-50, we’re talking about the county bestowing on the victor a margin of perhaps 1,500 votes!

Of all the wild cards in this election, Contra Costa is among the wildest.  But, right now, it looks like the biggest remaining contest in the state.

About Greg Diamond

Somewhat verbose attorney, semi-disabled and semi-retired, residing in northwest Brea. Occasionally ran for office against jerks who otherwise would have gonr unopposed. Got 45% of the vote against Bob Huff for State Senate in 2012; Josh Newman then won the seat in 2016. In 2014 became the first attorney to challenge OCDA Tony Rackauckas since 2002; Todd Spitzer then won that seat in 2018. Every time he's run against some rotten incumbent, the *next* person to challenge them wins! He's OK with that. Corrupt party hacks hate him. He's OK with that too. He does advise some local campaigns informally and (so far) without compensation. (If that last bit changes, he will declare the interest.) His daughter is a professional campaign treasurer. He doesn't usually know whom she and her firm represent. Whether they do so never influences his endorsements or coverage. (He does have his own strong opinions.) But when he does check campaign finance forms, he is often happily surprised to learn that good candidates he respects often DO hire her firm. (Maybe bad ones are scared off by his relationship with her, but they needn't be.)