NYC Gazette

Much of the news coverage immediately following this week’s presidential debate focused on who won the contest between Vice President Kamala Harris and former President Donald Trump, as well as on the Republican candidate’s false claim that Haitian immigrants were eating pets in Springfield, Ohio. However, viewers’ top interest during and immediately after the debate, measured by online searches, centered on the issue of abortion, shows an analysis by New York University researchers.

Prior to the debate, the most-searched election topic pertained to crime. However, during and after the 90-minute event, online users’ top preference shifted to abortion—a topic both candidates spoke about multiple times—with searches about abortion rights, existing laws by state, and “post-birth abortion” among the most common queries related to this issue.

The study conducted by NYU’s Courant Institute of Mathematical Sciences reveals how candidates’ words can alter the focus of viewers, whose interests may diverge from what’s highlighted in media accounts.

“How debates are reported on and what viewers care about may indeed coincide, but they also differ in crucial ways,” says Anasse Bari, a clinical professor in computer science at the Courant Institute and director of NYU’s Predictive Analytics and AI Research Group, which conducted the study.

The authors add that while abortion was the most-searched topic, Google searches for “Springfield” grew 16 times during the debate while those for both “Haiti” and “Haitians” increased threefold during the same period. However, they note these terms were likely not on many voters’ minds prior to the event so significant growth is not surprising.

The research also showed that immigration was the second-most searched topic—after abortion—during and after the debate. Searches focused on immigration bills, candidates’ standing on the issue, and German immigration laws—the latter cited by neither candidate during the debate.

During the debate, search queries about Project 2025 surged by approximately 900% compared to before it. Trending searches included “Project 2025 website,” “Project 2025 key points,” “Project 2025 summary,” and “What is Project 2025.” These queries trended in several states with high activity from Pennsylvania as well as Connecticut, Massachusetts, Minnesota, New Hampshire, and Ohio.

During the debate, top queries about Vice President Harris included: “Kamala Harris’s father,” “Kamala Harris’s childhood,” “Kamala Harris and fracking,” and “Kamala Harris’s economic plan.” Related queries about former President Trump focused on his inheritance and his time at Wharton School of Finance.

“How debates are reported on and what viewers care about may indeed coincide but they also differ in crucial ways,” reiterated Anasse Bari.

Researchers also examined Reddit posts to understand overall sentiments towards different issues discussed online. Positive sentiment indicated text conveying positive emotions or approval (e.g., using words such as "improving" when referring to inflation). Negative sentiment suggested text conveying negative emotions or criticism. Neutral sentiment marked an absence of either positive or negative sentiments. The researchers considered where sentiment stood before analyzing shifts resulting from the event.

Sentiment findings included:

- The positive sentiment score related to economy increased by 14.34% during the presidential debate reflecting a slightly more positive reaction than existed before; it rose an additional 34.43% after indicating a much more positive view due to exchange.

- Regarding inflation: sentiment score dropped by 31.25% during indicating shift toward neutral; however increased by 23.39% post-debate suggesting partial return towards positivity.

- Sentiment score for immigration dropped by 81.78% post-debate shifting from very positive pre-debate stance towards neutrality overall.

The study's other authors included researchers at Courant Institute's Predictive Analytics & AI Research Group: Charles Wang; Suryavardan Suresh; Harrison Gao; Atmaj Koppikar; Yifei Xu; Dev Pant; Anway Agte.

To gauge online users' interests/sentiments three hours pre-debate through two hours post-debate researchers analyzed Reddit posts Google search trends YouTube comments tracked trending tags X related each candidate created indices analyzed near real-time data applied natural language processing sentiment analysis algorithms.