Benjamin Leinwand, assistant professor of mathematical sciences delivers a talk about his study findings. Credit: Stevens Institute of Technology

In order to get elected, politicians need to signal their philosophy, principles, and values to potential voters. Beyond their policy positions, many candidates also strive to cultivate a personal brand by carefully choosing their language and visual style.

With so much political messaging migrating online, Benjamin Leinwand, assistant professor of mathematical sciences at Stevens, wondered if elected officials might similarly try to subtly position themselves on social media by strategically associating with other members of their coalition.

Using tools from network science, he built a statistical model that observed how senators and congresspeople interact with one another on X, formerly known as Twitter, but hid each person’s true position and political affiliation. The model managed to extract some insights from X’s digital cacophony by sorting politicians into groups based solely on how they interact online. In other words, by analyzing the individual’s connections, the model can deduce their positions.

Leinwand and his collaborator Vince Lyzinski, a mathematics professor at the University of Maryland, analyzed 475 members of the 117th U.S. Congress who each issued at least 100 tweets between February 9, 2022, and June 9, 2022. Their model found that these politicians could be split into three different groups or “communities” based on their patterns of interaction with one another.

Although their model was not furnished with any information about existing political blocs, the inferred communities fell along familiar lines. Broadly speaking, the first group is made up of Senators, the second of Democratic congresspeople, and the third of Republican congresspeople.

The two researchers developed a new method to understand how elected officials connect to one another. “We call two people ‘connected,’ if either one in the pair tweeted at the other one or retweeted the other one during this period,” Leinwand explained.

The researchers found that most people within the three groups tended to tweet most often at the people from their own group. “Republican congresspeople talked among themselves a lot, and Democratic congresspeople talked among themselves a lot, though Democratic congresspeople were somewhat more likely to interact with Senators than their Republican counterparts,” said Leinwand.

As Democrats controlled the Senate during the observation period, “one could imagine that Democratic congresspeople might be incentivized to amplify senate leadership messaging in addition to their allies in the House,” he continued.

Benjamin Leinwand, assistant professor of mathematical sciences at Stevens Institute of Technology talks about his recent study. Credit: at Stevens Institute of Technology

However, there were a few nonconformists. Out of the 475 members, 463 appeared to behave like the rest of the people in their respective groups, but 12 were such outliers, which researchers described as “exceptions” because they tweeted more like people in other groups.

“It seems as though certain Democratic congresspeople talked to Senators more, and certain Republican congresspeople talked to Senators more,” Leinwand said. “And so based on their patterns of interactions, they behaved like someone in the other group. They were tweeting at Senators also, rather than only tweeting at their own Congresspeople.”

Of these 12 miscategorized congresspeople, two went on to be elected Senators, one in 2022 and the other in 2024. The first one was Democratic Senator Peter Welch from Vermont. The second one was Democratic Senator Andy Kim from New Jersey. Another, Chris Pappas, is currently running for Senate in New Hampshire. A fourth member, David Trone, had an unsuccessful run for Senate in Maryland in 2024.

“The thing that jumps out to me is that we have 475 members, and 463 of them are grouped correctly,” Leinwand said—meaning that they act in accordance with the rest of their group. “And of those 12 exceptions, two ended up in the group where our model thought they should be assigned,” he noted—and that may be telling us something.

Does it mean that tweeting at Senators can help a Congressperson get elected as one? Clearly, tweeting alone won’t win you a seat in the Senate, but there may be a strategy to it, Leinwand thinks.

“They have public facing jobs, so I suspect they’re often tweeting strategically, particularly when they’re including other members of Congress,” he says. “I think if you start positioning yourself in a certain way, if you start interacting with senators on X, you may be seen as more senatorial to voters.”

Leinwand and Lyzinski outlined their findings in the study titled “ACRONYM: Augmented Degree Corrected, Community Reticulated Organized Network Yielding Model,“ published in the Journal of Computational and Graphical Statistics on October 7, 2025.

Leinwand is quick to point out that their study used an exploratory approach to search only for patterns and in order to derive more specific conclusions, different analytical techniques would be needed. But, he adds, “I suppose some behavior betrays some intention.”

More information: Benjamin Leinwand et al, ACRONYM: Augmented Degree Corrected, Community Reticulated Organized Network Yielding Model, Journal of Computational and Graphical Statistics (2025). DOI: 10.1080/10618600.2025.2540356

Citation: Can politician tweeting strategy betray their ambitions? Scientists map congressional positioning through X engagement (2025, November 4) retrieved 4 November 2025 from https://phys.org/news/2025-11-politician-tweeting-strategy-betray-ambitions.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.