By comparing the topics of Editorials published in Nature and Science, we unearthed some fascinating features of the changing policy landscape over the past decade.

To analyse the text of the more than 1,500 Editorials published in Nature and Science between January 2000 and July 2009, we used 'VOSviewer' — a mapping technique that categorizes non-specialist documents according to the co-occurrence of words (N. J. van Eck and L. Waltman Scientometrics doi:10.1007/s11192-009-0146-3; 2009). We confirmed the results by manual classification of the contents of a large sample (around 20% of the Editorials).

Editorial topics over this period covered space and physics (5%), publication issues (10%), global political and environmental problems (18%), biomedical issues (almost 30%) and science policy issues (39%). These figures are roughly comparable in Nature and Science, which comes as no surprise as they presumably reflect what the global scientific community considered important at the time. Editorials on climate change, for example, have almost the same prominence in both journals (1 in 10 Editorials).

But there are differences. Nature devotes three times more Editorials to space and physics (and particularly to NASA) than does Science (6% and 2%, respectively) and twice as many to the National Institutes of Health (2% compared with 1%). Science pays more attention to developing countries, environmental protection and other global problems.

Science writes more often than Nature about the political influence of science and scientists, whereas Nature writes more about priority setting and the organization of science. It seems that Science, as a learned-society journal, may be more reticent about tackling internal science issues than the independent journal Nature.