Incumbent Mayor Karen Bass conceded the Los Angeles mayoral race to City Councilmember Nithya Raman on Nov. 5. With more than 95% of the vote in, Raman has netted 54.7% of the vote — nearly ten points higher than Bass — according to the Associated Press.
A list of runoff election polls collected by the New York Times from different organizations — including some sponsored by Raman and Bass — consistently placed Raman ahead of Bass going back to May. However, on prediction markets Kalshi and Polymarket, it’s almost always been the exact opposite.
Data gathered by Annenberg Media showed Bass trending ahead of Raman from the period of April to September 2026. For most of this period, this gap hovered around 20 points.
But on Kalshi, for instance, odds for Bass once reached 80% on June 2 — the highest they would be for the then-incumbent mayor.
This spike coincides with Bass advancing in the runoff election for her second term on the same day, according to NBC. At the time, Raman was still competing with Spencer Pratt to join Bass in the November election.
Odds are determined on prediction markets by the participants deciding to buy in on the available options; the odds themselves being a consensus of everybody involved in the market, according to Investopedia.
On Polymarket, Bass experienced a dip on June 2, with Raman’s odds spiking at around the same time. The highest odds Bass experienced on Polymarket were 70.5% on May 27.
Earlier in the month Bass experienced an almost 20 point increase, coinciding with the debate that took place between her, Raman and Pratt on May 6.
Going into the election on Nov. 3, which Bass would lose, odds in both prediction markets still placed her ahead of Raman by a margin of over 30 points.
Mark DiCamillo, who leads the University of California, Berkeley Institute of Governmental Studies poll, said the difference in the results in surveying and prediction markets came down to a difference in expectations versus intentions.
“To me, prediction markets are some kind of measure of the expectations of certain people making bets on a website,” DiCamillo said. “They could be heavily weighted by one or two individuals making big bets, who knows? It’s all based on expectations and that contrasts with polls, which we’re asking people for their actual voter intentions.”
UC Berkeley IGS conducted three polls regarding the L.A. mayoral race. In September and May, their surveys showed voters preferred Raman over Bass. In October, both candidates were tied at 40%.
Prediction markets on Kalshi and Polymarket both showed fluctuations on the days where IGS released their polls, with Raman experiencing a brief spike on May 28 and a rise by over ten points on Sept. 23.
Bass overcame Raman on the prediction markets on Oct. 28, the same day IGS released a poll showing the candidates tied.
But while the October IGS poll showed Bass and Raman tied, respondents did not think Raman would win, according to DiCamillo.
“We had 40% saying they thought Bass would win, 32% saying Ramen would win,” DiCamillo said. “I’m trying to make things clear that prediction markets are about expectations, not voting intentions.”
Wei Xuan, a student at the University of Southern California pursuing his doctorate in economics, and who specializes in cryptocurrency and prediction markets, said prediction markets are an “essentially different mechanism” from polls.
“Prediction markets don’t have an edge on accuracy… So I’m not making that claim,” Xuan said. “I’m just saying those people who participate in the prediction market are more incentivized to make the bet as accurate as he or she can.”
Xuan said the discrepancy in polling compared to prediction markets could come down to who’s involved in the market. He said prediction markets could become more accurate if more people were involved in them to average out different viewpoints.
On Kalshi, the end trading volume for the L.A. mayoral race was nearly $150 million. On Polymarket, just over $18 million.
“The key question to me is who relies on the odds on the prediction markets,” Xuan said. “If voters, donors, reporters and campaigns outwash the market, then the price becomes powerful, and anything powerful is worth trying to move. The worry is simple: A group that supports a candidate could spend money to push that candidate’s price up. The goal isn’t to win the market; it’s to create a headline.”
Edited by Benjamin Gamson