Understanding Social Media’s Role in Climate Misinformation Spread

The rise in climate-skeptic terminology on Twitter following Elon Musk’s takeover suggests social media may be worsening the spread of climate misinformation, according to researchers Max Falkenberg and Andrea Baronchelli. However, they argue that focusing solely on classifying content as “misinformation” may be less productive than understanding the structures that enable information spread.

In their analysis originally developed for The Times, the researchers found a significant increase in climate-skeptic content on Twitter over the past six months. Their study published in Nature Climate Change reveals that while outright climate denial exists on social media, such direct statements represent only a small portion of climate-contrarian views circulating online.

The more substantial growth has occurred in indirect forms of climate skepticism, which the researchers refer to as “discourses of delay.” These include questioning the pace of the green transition or attempting to discredit those involved in climate action debates. Unlike outright denial, these indirect forms of skepticism can be more nuanced and sometimes blur the line between contrarian views and legitimate concerns.

“The best example is perhaps the accusation that ‘politicians are hypocrites’ in calling for climate action,” the researchers note, pointing to the surge in such statements over the past two years, particularly regarding politicians’ use of private jets. The complexity emerges as both climate contrarians and pro-action activists can make similar claims but with different intentions.

For instance, when discussing how climate action might affect economically vulnerable populations, an activist might highlight the burdens faced by low and middle-income countries while advocating for loss and damage funds. In contrast, a contrarian might argue that phasing out fossil fuels will disproportionately harm society’s poorest, including those in wealthy nations.

This nuance creates challenges for automated classification systems, which typically focus on identifying strong climate denial language while missing the larger community expressing more subtle forms of skepticism about climate action. The researchers argue that understanding this broader community is likely more important from a policy perspective than focusing solely on accounts heavily engaged in outright climate denial.

To address these limitations, Falkenberg and Baronchelli have taken a structure-based rather than content-based approach in their recent work. By analyzing the network of interactions between users—including likes, retweets, and comments—they can identify broader climate-contrarian echo chambers without being limited to specific terminology.

This method revealed a dramatic increase in climate polarization in recent years. Their study of discussions around UN Climate Change Conference (COP) summits found that engagement with members of the contrarian echo chamber increased 16-fold between COP21 (2015) and COP26 (2021)—four times greater than the increase among the pro-climate echo chamber. This growth appears to be driven by climate contrarians broadening their audience across the political right.

While these findings were possible because of data Twitter makes available to researchers, the authors express concern about the future accessibility of such data following Twitter’s change in ownership. Most other social media platforms maintain far more restrictive data access policies, severely limiting research capabilities.

A review by Warren Pearce and colleagues highlights this issue: of 35 papers studying climate change communication on social media published up to January 2018, 28 focused on Twitter, five on Facebook, and only a handful examined other platforms. Instagram, despite having twice as many active users as Twitter in 2018, was not represented in any studies due to data access limitations.

Although some platforms offer researchers access to public posts, this represents only a fraction of content on sites like Facebook. While privacy concerns are valid, the researchers worry these considerations may become excuses for social media companies to avoid transparency regarding their data and algorithmic choices.

Falkenberg and Baronchelli, who are part of the IRIS Academic coalition studying “infodemics,” emphasize that without intervention from politicians and regulators, social media companies will continue to hide behind data privacy concerns, limiting future studies to anecdotal evidence and simple term-based analyses rather than comprehensive examinations of how information spreads through these influential platforms.