Résumé
Editors of scientific journals make central decisions in the publication process. Information peripheral to the content of a manuscript such as the editor's professional circle and authors' publishing record may influence these decisions. This constitutes reputation whose role in the publication process remains poorly investigated. Analyzing three decades of publications of 33 Nature journals in physical and life sciences, we show that reputation is tied with publications at the level of editors. The establishment of a non-financial conflict of interest policy and the journals' impact factor were associated with changes in the number of publications of editors' former co-authors and authors with a publishing record in Nature journals. We suggest changes at the author and journal levels to mitigate the role of reputation in the publication process.