Policy Documents

Scientific Consensus on Global Warming

Joseph Bast and James M. Taylor –
April 3, 2007

This booklet summarizes the results of international surveys of climate scientists conducted in 1996 and 2003 by two German environmental scientists, Dennis Bray and Hans von Storch. Bray is a research scientist at the GKSS Institute of Coastal Research in Geesthacht, Germany. Von Storch is a climatology professor at the University of Hamburg and director of the Institute of Coastal Research.

More than 530 climate scientists from 27 different countries provided numerical answers each time the survey was conducted. All responses were anonymous. The same questions were asked each time the survey was conducted, plus an additional 32 questions were asked in 2003. The 2003 survey was conducted online. Notice of the survey was posted in the Bulletin of the American Meteorological Society and on the Climlist server (Climlist is a moderated international electronic mail distribution list for climatologists and those working in closely related fields). Notices also were sent to institutional lists in Germany, Denmark, and the U.K. The survey was password protected to ensure that scientists in climate-related fields were the only ones with access to it.

The surveys presented dozens of assertions regarding climate change and asked respondents to give a numerical score, on a scale of 1 to 7, indicating the extent to which the respondents agreed or disagreed with each assertion. The entire results of both surveys can be found online at a site created and maintained by Bray and von Storch.1 The average responses to every question in both the 1996 and 2003 surveys are reported in the appendix of this booklet. This is all valuable and accurate data, of course, but it can be difficult for a layperson to interpret. What does it mean, for example, to say the average response to a question is 3.39?

To make the survey results more transparent, we singled out 18 questions from the 2003 survey and present the answers here in a simplified and less academic style. For each question, we combined the percentages of those respondents who gave numeric scores of 1, 2, or 3 and called this “agree.”