Misleading by petition: just what is the consensus on global warming?

IS GLOBAL WARMING A REAL phenomenon, and if so, are humans causing or contributing to it by burning fossil fuels and will it lead to an increased frequency and/or severity of natural disasters? The American public looks to science or scientists to help answer these questions. A petition circulated by a small group of scientists is creating quite a stir, arousing considerable praise and disdain from groups on different sides of the global warming issue.

The petition drive was begun by Dr. Frederick Seitz, now deceased, and is presently led by Dr. Arthur Robinson and his son, Dr. Noah Robinson, both members of the Oregon Institute of Science and Medicine (OISM), who remain the chief organizers and expositors of the petition project. Seitz was a physicist and past president of both the National Academy of Sciences and Rockefeller University. The two Robinsons are both chemists and Arthur Robinson is the current director of OISM.

According to the website, presumably authored by Arthur Robinson, which reports the details of the petition and is, "the purpose of the Petition Project is to demonstrate that the claim of "settled science" and an overwhelming "consensus" in favor of the hypothesis of human-caused global warming and consequent climatological damage is wrong." (1) Robinson asserts not just that his collection of 31,072 signatures on a petition has refuted the claim of "settled science" and "overwhelming consensus" among scientists with regard to global warming, but that "The very large number of petition signers demonstrates that, if there is a consensus among American scientists, it is in opposition to the human-caused global warming hypothesis rather than in favor of it." (2) Not only has Robinson failed to substantiate either of his assertions, he is misleading the American public by implying that his petition fairly represents relevant expert opinion.

To understand the problems with Robinson's "Global Warming Petition Project", we must first examine how the petition itself was distributed and how signatures were collected. To a sample of persons on the mailing list of American Men and Women of Science, (3) Robinson sent a petition packet consisting of a petition card, a return envelope, a cover letter from Seitz, and a 12-page review of the literature on the human-caused global warming hypothesis authored by the two Robinsons and Willie Soon. (4) The two main assertions stated on the petition card were that there is no convincing scientific evidence that the human release of carbon dioxide and other gases is causing harmful atmospheric heating and climate change, and that the U.S. government should reject the Kyoto Agreement and any other similar proposals. Arthur Robinson not only requested that recipients return the signed petition card if they agreed with its assertions, but also arranged for the recipients to distribute petition packets to their colleagues. He also enabled other persons to obtain petition packets by simply requesting them through his website, and this procedure ultimately produced 5% of the returned petition cards. Thus, signed petitions were solicited in three different ways.

Although the website for the petition indicates that checks of credentials and identity were performed for signatories of the returned petitions, and invalid petitions were excluded, how the checks were performed is not described. Signed petition cards were accepted only if they came from persons who had "obtained formal educational degrees at the level of Bachelor of Science or higher in appropriate scientific fields." (5) In the end, "valid" and signed petition cards were obtained from 31,072 persons with degrees in file following fields: Earth science (3,697 persons or 12% of the total); computer science and mathematics (903 or 3%); physics and aerospace sciences (5,691 or 18%); chemistry (4,796 or 15%); biology and agriculture (2,924 or 9%); medicine (3,069 or 10%); and engineering and general science (9,992 or 32%). The breakdown according to educational level was: Ph.D. (9,021 or 29%); MS (6,961 or 22%); M.D. and D.V.M. (2,240 or 7%); and B.S. or equivalent (12,850 or 41%).5 On his website Robinson fails to report the cross-tabulations of fields of expertise and levels of education for his petition respondents. For example, it doesn't show what percentage of the persons with Earth Science expertise had Ph.D. degrees.

Although in one interview Robinson called his petition project a "survey," (6) it is definitely not a well-designed scientific survey of the views of a group of relevant experts, its results cannot be used to reach the conclusions about "consensus" that are asserted and hoped for by Robinson. In the first place, Robinson presents neither a dictionary nor an operational definition of "consensus." He wants to reach conclusions about a consensus, but he spends no time telling us what he thinks a consensus is. According to Merriam-Webster's Eleventh Collegiate Dictionary, a "consensus" may refer to general agreement, unanimity, judgment by most of those concerned, or group solidarity. Of course if we use the most extreme definition of consensus--"unanimity"--that every single person in the group must agree, then Robinson's Global Warming Petition Project shows that all people with science degrees, or even all people with relevant degrees do not support the hypothesis of human-caused global warming. But this is a trivial conclusion; we knew this before the petition drive! The "unanimity" definition of "consensus", however, is not the one in which the American public is interested.

If Robinson had been conducting a valid survey, he would have offered an operational definition of "consensus" before he started his inquiry. Robinson misleads the public to think that a consensus is defined by some large absolute number of persons. It is not. It is determined by a large percentage of persons in a relevant sample. Does Robinson, or the general public, think of a consensus as agreement within a given group at a level of 75%, 90%, or some other percentage? He does not tell us. He reports only the number of persons who sent back signed petition cards, but he reports neither the total number of persons to whom he sent petition cards in the first place nor the number of persons to whom he sent petition cards who subsequently returned only messages of disagreement. Since Robinson chose to conduct a petition project rather than a well-designed scientific survey, he ...