Researchers who collaborate with colleagues abroad may produce work that is less novel than those who have collaborators in the same country, a new study suggests. Caroline S. Wagner of the Ohio State University, Travis A. Whetsell of Florida International University, and Satyam Mukherjee of the Indian Institute of Management analyzed references in 850,000 papers that were indexed in Web of Science in 2005 (Res. Policy 2019, DOI: 10.1016/j.respol.2019.01.002). They found that coauthors from different countries tended to cite papers only from their own discipline, suggesting the research was less novel. In contrast, coauthors from the same country tended to cite papers from their own and other disciplines, suggesting the research was more novel. “We’re really, really surprised” by the findings, Wagner says, noting that she thought international collaboration would boost novelty. One reason for the trend could be that scholars working at the international level are more conservative about their ideas, Wagner says. “The people that are working internationally are highly reputed and their interest is to maintain a reputation, not to make a reputation,” she says. However, the study is based on old data, notes Cassidy Sugimoto, an information scientist at Indiana University Bloomington. In light of that, “I would not use these results to advocate for a disinvestment in international collaboration,” she says.