By John A. Tures, Professor of Political Science, LaGrange College
Polls show a close race between Keisha Lance Bottoms and Rick Jackson running for Governor, and a slight lead for US Senator Jon Ossoff against challenger Congressman Michael Collins. But prediction markets make it look like a blowout. What do these numbers mean for another pair of competitive races at the top of the ticket in Georgia?
Yesterday, at the LaGrange Kiwanis Club, I was dragged into virtual “stoppage time” by the members when I presented about polls and prediction markets for the Georgia Governor and Senate races. Even when I begged the audience that I had to leave to teach my questions, they kept peppering me with more. Clearly, Peach State residents are desperate to know more about the contests on the ballot. Here’s a summary of what I presented.
As of this week, the realclearpolitics aggregate of the polls show that the former Atlanta Mayor Lance Bottoms is beating businessman Jackson by 1.5 percentage points. But as I told the Kiwanis, you can’t read these like it’s a sporting match. For many of these contests where Lance Bottoms leads by 1-2 percentage points, the gap is within the margin of error, which varies between 3 and 3.6 percentage points. That number is determined by the sample size, which can even include polls with samples sizes closer to 800 respondents, and not those exceeding 1,000. Those properly polling would judge this race “too close to call.”
But prediction markets tell a very different story. In mid-July, there was a roughly 55%-45% Polymarket’s numbers on who was betting on the race in favor of Lance Bottoms. But now that gap has widened to a 60%-40% split in favor of the ex-Atlanta Mayor in this open seat.
It’s a similar story for the U.S. Senate race, albeit with an incumbent in the race, in terms of the gap between polls and prediction markets. U.S. Senator. The realclearpolitics polling average has Senator Jon Ossoff leading by 7.5 percentage points, across an aggregate of polls. That’s behind the margin of error which can vary from 2.9%-3.5% depending on the pollsters’ sample size. But when it comes to prediction markets, Ossoff appears to have a commanding lead. Those betting markets (Polymarket) give Ossoff a 94%-6% lead.
What is happening?
First of all, it’s important to note what the numbers mean. Again, to use the sports metaphor, if Georgia leads Georgia Tech by three points, it means they are ahead by three points. In a poll, that might be within the margin of error, where we cannot accurately say what’s up. Lance Bottoms could be up 4.5%, or Jackson could be ahead 2.5%. Ossoff could be up four percentage points, or 10 points. With small sample sizes, these are what the margins of error show.
Prediction markets can be even more easily misunderstood. Lance Bottoms is highly unlikely to win by 20 percentage points, barring an unprecedented event in Georgia history. Ossoff is even more unlikely to win his race 94%-6%. If he does, I’ll take a swim in West Point Lake on Christmas. It simply means X percentage of bettors think the Democrat will win, and it doesn’t say by what percentage. They are probably reacting to the sour national mood. In fact, when I covered Jackson’s events, barely a whisper was made of Trump’s name. Lance Bottoms and Ossoff seem eager to tie their opponents to the President.
Some think that betting markets are more accurate, because people are putting down real money, or have “real skin” in the game. But non-Georgians can bet. Those not living in America can bet. Some French guy is alleged to have dumped millions on Trump in the 2024 election. It’s analogous to those claims that if the Pittsburgh Steelers or Washington Commanders win will guarantee a Republican or Democratic victory. As Matt Towery with InsiderAdvantage and James Magazine reports, it’s a close race in both elections. The numbers generally support that.
John A. Tures is a professor of political science at LaGrange College in LaGrange, Georgia. His views are his own. He can be reached at jtures@lagrange.edu or on “X” at @johntures2. His first book “Branded” a thriller novel where corporate greed, media manipulation and academic intrigue collide in a deadly game of product placement, has been published by the Huntsville Independent Press (https://www.huntsvilleindependent.com/product-page/branded). His second novel, “Independent Thought,” about a third-party candidacy that brings America to its knees, is published by HIP (https://www.huntsvilleindependent.com/product-page/independent-thought).