U.S. Public Sentiment on Nuclear Energy: Positive Trends Amid Concerns Over Waste, Cost, and Safety
A recent analysis of 300,000 posts on X (formerly Twitter) conducted by University of Michigan researchers reveals that the U.S. public generally holds a positive view of nuclear energy. However, concerns about waste management, cost, and safety persist.
The study, published in Renewable and Sustainable Energy Reviews, highlights the need to address public misconceptions and worries about nuclear energy as it is expected to play a critical role in decarbonizing the energy sector by 2050. Replacing oil and gas with nuclear power could provide a stable baseload electricity source.
"Understanding and addressing how the public feels about nuclear energy is essential for a just transition to clean energy," said Majdi Radaideh, U-M assistant professor of nuclear engineering and radiological sciences and corresponding author of the study.
While traditional surveys offer detailed insights from specific locations, such as communities near proposed nuclear facilities, social media analysis can significantly expand the sample size while reducing costs and time.
Researchers compiled 1.26 million posts from X, spanning 2008 to 2023, using an extensive list of keywords related to nuclear energy. They employed large language models (LLMs) to annotate the posts as positive, negative, or neutral and summarize the content.
Of the 300,000 posts geotagged to the U.S., neutral sentiments—those presenting facts without advocating for or against nuclear technology—were the most common, accounting for about 50%. Positive sentiments comprised 30% of the posts, while negative sentiments made up roughly 23%. In a state-by-state breakdown, 48 out of 50 states exhibited more positive sentiment, with the national average at 54% positive.
Positive sentiment was primarily driven by technological advancements, with users highlighting innovations that make nuclear energy safer, more reliable, and economical. Posts emphasized nuclear power's high energy density and its ability to operate continuously, as well as job creation and the necessity of transitioning to clean energy.
Conversely, concerns about radioactive waste, its long-term dangers, and disposal challenges fueled negative sentiments.
In addition to gaining a deeper understanding of U.S. public opinion on nuclear energy, the study developed a method for using AI to label data with reduced bias. Instead of relying on a single data labeling tool, the researchers employed seven different programs and determined the final label based on a majority vote.
"Labeling with multiple tools reduces bias as each tool struggles with certain types of texts and tones," explained Katie Vu, an undergraduate majoring in electrical engineering and computer science and co-author of the study.
The labels were scored as high confidence if five or more programs agreed and low confidence if only three or four programs agreed. Notably, LLMs trained exclusively with high-confidence posts showed a 15% increase in accuracy, achieving 96% accuracy.
For this initial analysis, the researchers chose X over other social media platforms like Instagram, Facebook, or LinkedIn due to its concise, text-based format. As a next step, the team plans to develop a near real-time dashboard that aggregates nuclear sentiment from various social media platforms and news headlines.
