I was perusing the voting results for Santa Cruz County and I stumbled across something weird. It has to do with a strange feature of this year’s ballot: the U.S. Senate election for California. After Diane Feinstein died in September 2023, Governor Gavin Newsome appointed Laphonza Butler to fill her seat temporarily. By law, after such an appointment a special election must be held at the next general election to allow voters to choose the replacement for the rest of the term. Feinstein’s senate term ends in January of next year, so there also needed to be a regular contest to determine the senator for the next six year term.
The upshot is that there were two identical-looking ballot entries on November 5th: first, the regular vote for the six-year term starting in January, and second, a special election vote to fill the spot for the month of December (the period after the election is certified, and before the new Congress is sworn in in January). Goofy, but there we are.1This law makes more sense when the replacement happens earlier in the term – in those cases the unexpired term might be a year or two, so a special election is appropriate.
Anyway, that’s not the weird thing. Here are the results so far from Santa Cruz County:
| US Senator (Full Term) | Party | Votes |
| Steve Garvey | R | 24,742 |
| Adam Schiff | D | 77,339 |
| Undervotes | 4,653 | |
| Overvotes | 29 |
Overvotes are when someone voted for both candidates – probably just a silly mistake of filling in both bubbles. If the voter noticed it maybe they didn’t feel like getting a new ballot. Anyway, it’s very rare.
Undervotes are when neither candidate was chosen. It’s perfectly okay to skip a specific ballot entry; maybe you don’t like either candidate, or you don’t care.
But things get strange with the vote for the unexpired term:
| US Senator (Unexpired Term) | Party | Votes |
| Steve Garvey | R | 24,557 |
| Adam Schiff | D | 75,842 |
| Undervotes | 6,352 | |
| Overvotes | 12 |
The Undervotes increased from 4,653 to 6,352. That means that 1,699 people who voted in the full term election couldn’t be bothered to vote for the unexpired term. The ballot items were consecutive, with the same choices. Having just filled in one, why wouldn’t you fill in the next one? Is it just laziness? Or maybe a protest against the ridiculous necessity of a special election to fill a seat for one measly month? Whatever reason, 1.7% of the people who voted for the first section abstained in the second.
But that’s not even the weird thing. This is: almost all of the undervotes came from Adam Schiff voters. Schiff’s vote total dropped by 1,497, while Garvey’s dropped by just 185. You would expect more undervoters from Schiff, because there were so many Schiff voters in the first place (he won the county 3 to 1). But even allowing for that, the count is still out of whack: Schiff voters undervoted at 2.5 times the rate of Garvey voters (1.9% vs 0.7%).
I spot checked a few other California counties. I chose from across the political spectrum: some had a strong Democratic lean (like Santa Cruz), others leaned more Republican, and some were in the middle. In all cases, there were more undervotes in the unexpired term election than the full term election. But in each of these counties the distribution of these excess undervoters (D vs R) closely matched the way the county voted. For example, Fresno County voted for Garvey over Schiff by 55% to 46%, and they had pretty much the same ratio of undervoters (52% to 48%). Only in Santa Cruz was there a wide discrepancy in the undervoter profile. Why? Who can say. Santa Cruz has always been a bit weird.
There’s one other possibility. My analysis assumes that no one changed their vote, e.g., no one voted for Schiff in the full term choice but Garvey in the unexpired term choice. That assumption leads to the conclusion that Schiff voters undervoted at a much higher rate in Santa Cruz. But there is another way for the math to work out. Let’s start with the assumption that undervotes occurred at the same rate for both Schiff and Garvey voters. The only way that could be true is if about 225 people who voted for Adam Schiff on the first ballot entry changed their vote to Steve Garvey in the second. Now that would be really weird.
I remember hearing an argument that Adam Schiff supporters were encouraged to vote for him for the full-term but not the partial-term, in case term limits were later imposed. (The partial-term might count as a term, hence limiting how long he could serve.) I have no idea if this questionable argument was presented by supporters or detractors.