Karen Bass never needed the late ballots, and the late ballots never gave her anything. From the first returns on the night of June 2 to the final projections a week later, the mayor of Los Angeles sat parked at roughly 35 percent of the vote while the count churned on around her.
That fact alone should reorder how we think about what happened in this election, because the conventional assumption, that California’s weeks-long ballot counting machinery exists to protect powerful incumbents, simply does not fit the data. The machinery did go to work in Los Angeles this month. It just was not working for Bass.
It was working on two other projects entirely. On election night, Republican Spencer Pratt held second place in the mayoral primary and a ticket to the November runoff, while a half-cent county sales tax called Measure ER was losing. Six days later, both results had reversed.
Pratt was out, replaced by Democratic Socialists of America-aligned Councilwoman Nithya Raman, and Measure ER had clawed its way into the lead. Every contested outcome in this election was decided not by the voters who showed up on or before Election Day, but by the ballots that showed up afterward.
This is the part of the story the legacy press has no interest in dwelling on. They will tell you, accurately, that late-arriving mail ballots in California skew Democratic, and they will treat that observation as the end of the inquiry. It should be the beginning.
A system in which the election-night loser of one particular party routinely gets a second bite at the apple, week after week, batch after batch, is not a neutral system that happens to produce lopsided corrections. It is a system whose corrections always run one direction, and in Los Angeles this month, those corrections landed with surgical precision on the only two outcomes that were actually in doubt.
- Karen Bass’s vote share held steady near 35 percent through the entire Los Angeles count; the late ballots changed nothing for her.
- Spencer Pratt held second place and a runoff spot for six straight days of counting before late ballots pushed Nithya Raman past him.
- The Associated Press now projects a November runoff between Bass and Raman, a Democrat and a DSA-backed Democrat, with no Republican on the ballot.
- Measure ER, a half-cent county sales tax projected to raise about $1 billion a year, trailed from election night until June 8, when it flipped to a lead of roughly 13,000 votes.
- Around 148,000 ballots remained to be counted at the flip, with updates running through June 26 and certification due by July 10.
- Election analysts attribute the late surge to Democrats holding their mail ballots while deciding the governor’s race, an explanation that accounts for the mechanism but not the system that enables it.
- Whether by design or by structure, the late count’s only measurable effects were removing the Republican from the runoff and reviving a billion-dollar tax.
The Trendline Nobody Wants to Examine
Start with the numbers, because they are not in dispute. When the first returns posted on election night, Bass led with 36 percent, Pratt held second at 30, and Raman trailed at 20. By Friday morning, with the count grinding forward, Pratt still held second at 29.4 percent to Raman’s 23.4, and Bass sat at 35.1.
Then came the weekend ballot drops. By Monday, Raman had surged to 27.1 percent, edging Pratt’s 26.7, and Decision Desk HQ projected her into the runoff. The Associated Press followed, and the November ballot was set. Bass versus Raman. A Democrat versus a Democrat the DSA once endorsed, in a city where a Republican whose own home burned in the Palisades Fire had just spent six days in second place.
Notice what did not move. Bass began the count in the mid-thirties and ended the count in the mid-thirties. If you believe Los Angeles runs an industrial-scale ballot harvesting operation to shield its incumbents, this election is a problem for your theory. The harvest did not touch the incumbent’s number. It operated entirely in the space below her, deciding which opponent she would face and ensuring that opponent would come from her left rather than from the right.
A Tax Rises From the Dead
The same six days performed the same resurrection on Measure ER. The half-cent sales tax hike, sold as a rescue package for county healthcare services, trailed from the moment polls closed on June 2 until June 8, when yes votes pulled ahead for the first time. As of that flip, the measure led by roughly 13,000 votes out of nearly 1.9 million cast, 50.35 percent to 49.65, with about 148,000 ballots still outstanding. Counting continues through June 26, and certification is due by July 10, so the result is not final. But the trajectory is the same trajectory that carried Raman past Pratt, and there is little reason to expect the remaining ballots to behave differently than the batches that preceded them.
That measure is worth about $1 billion a year to county government. It was losing among the people who voted by Election Day. It is winning among the ballots counted afterward. Angelenos already staggering under some of the highest combined tax burdens in America rejected this thing in real time, and then watched it pass in slow motion.
The Official Explanation, Taken Seriously
The professional election watchers have an answer for all of this, and intellectual honesty requires engaging it. Geoffrey Skelley of Decision Desk HQ told The Hill that Democratic voters sat on their mail ballots until the last minute because, in his words, “they’re trying to figure out who they’re going to vote for for governor” in a chaotic primary. Late ballots therefore skewed Democratic, lifting Raman and Measure ER alike.
The problem with this theory is that neither Raman nor Measure ER should have seen such a giant boost compared to Bass. Deciding late for governor should have helped Bass as much if not more than Raman. As for Raman voters, one would think Tom Steyer would have gotten the nod over Xavier Becerra, but he didn’t.
Now ask the question the explanation conveniently forecloses. Why does California operate a system in which ballots can arrive for days after the election, get harvested by paid operatives in the weeks before it, and get counted for nearly a month afterward, with no realistic capacity to verify the chain of custody on any individual ballot at that scale? A mechanism that reliably reverses election-night results in one party’s favor does not need to be a conspiracy to be a scandal. The scandal is that the system was built this way on purpose, defended this way on purpose, and produces these reversals so predictably that analysts now describe them as routine.
Woe unto them that seek deep to hide their counsel from the LORD, and their works are in the dark, and they say, Who seeth us? and who knoweth us? (Isaiah 29:15)
Who Actually Benefited
What follows is speculation, and it should be read as exactly that. But speculation disciplined by evidence is how citizens reason about opaque systems, so follow the beneficiaries. If an organized effort moved late ballots in this election, it was not aimed at saving Karen Bass, whose share never budged. The two outcomes it actually produced were a November runoff scrubbed of any Republican option and a billion-dollar annual tax stream pulled back from defeat.
A Bass-versus-Raman general election keeps every lever of city power inside the progressive coalition no matter who wins, and it hands the machine a contest it cannot lose. The tax pays for the government that machine runs. If you were allocating a finite harvesting operation, that is precisely where you would spend it, on the margins that were close, not on the incumbent who was safe.
Perhaps it was all organic. Perhaps hundreds of thousands of procrastinating Democrats independently delivered both reversals without coordination. The honest answer is that under California’s rules, no one can prove the difference, and that is the indictment. A trustworthy election system makes innocent explanations verifiable. This one makes them unfalsifiable.
Meanwhile, the conservative voters of Los Angeles, the ones who watched their neighborhoods burn while City Hall managed press conferences, now have no one to vote for in November. They did not lose that choice on Election Day. They lost it on the sixth day after, somewhere in a ballot drop, and they are expected to call it democracy.