But back to our main subject.

What I have been suggesting to you is that nuclear winter was ameaningless formula, tricked out with bad science, for policy ends. Itwas political from the beginning, promoted in a well-orchestrated mediacampaign that had to be planned weeks or months in advance.

Further evidence of the political nature of the whole project can befound in the response to criticism. Although Richard Feynman wascharacteristically blunt, saying, "I really don't think these guys knowwhat they're talking about," other prominent scientists were noticeablyreticent. Freeman Dyson was quoted as saying "It's an absolutelyatrocious piece of science but…who wants to be accused of being infavor of nuclear war?" And Victor Weisskopf said, "The science isterrible but---perhaps the psychology is good." The nuclear winter teamfollowed up the publication of such comments with letters to theeditors denying that these statements were ever made, though thescientists since then have subsequently confirmed their views.

At the time, there was a concerted desire on the part of lots ofpeople to avoid nuclear war. If nuclear winter looked awful, whyinvestigate too closely? Who wanted to disagree? Only people likeEdward Teller, the "father of the H bomb."

Teller said, "While it is generally recognized that details arestill uncertain and deserve much more study, Dr. Sagan nevertheless hastaken the position that the whole scenario is so robust that there canbe little doubt about its main conclusions." Yet for most people, thefact that nuclear winter was a scenario riddled with uncertainties didnot seem to be relevant.

I say it is hugely relevant. Once you abandon strict adherence towhat science tells us, once you start arranging the truth in a pressconference, then anything is possible. In one context, maybe you willget some mobilization against nuclear war. But in another context, youget Lysenkoism. In another, you get Nazi euthanasia. The danger isalways there, if you subvert science to political ends.

That is why it is so important for the future of science that theline between what science can say with certainty, and what it cannot,be drawn clearly-and defended.

What happened to Nuclear Winter? As the media glare faded, itsrobust scenario appeared less persuasive; John Maddox, editor ofNature, repeatedly criticized its claims; within a year, StephenSchneider, one of the leading figures in the climate model, began tospeak of "nuclear autumn." It just didn't have the same ring.

A final media embarrassment came in 1991, when Carl Sagan predictedon Nightline that Kuwaiti oil fires would produce a nuclear wintereffect, causing a "year without a summer," and endangering crops aroundthe world. Sagan stressed this outcome was so likely that "it shouldaffect the war plans." None of it happened.

What, then, can we say were the lessons of Nuclear Winter? I believethe lesson was that with a catchy name, a strong policy position and anaggressive media campaign, nobody will dare to criticize the science,and in short order, a terminally weak thesis will be established asfact. After that, any criticism becomes beside the point. The war isalready over without a shot being fired. That was the lesson, and wehad a textbook application soon afterward, with second hand smoke.

In 1993, the EPA announced that second-hand smoke was "responsiblefor approximately 3,000 lung cancer deaths each year in nonsmokingadults," and that it " impairs the respiratory health of hundreds ofthousands of people." In a 1994 pamphlet the EPA said that the elevenstudies it based its decision on were not by themselves conclusive, andthat they collectively assigned second-hand smoke a risk factor of1.19. (For reference, a risk factor below 3.0 is too small for actionby the EPA. or for publication in the New England Journal of Medicine,for example.) Furthermore, since there was no statistical associationat the 95% confidence limits, the EPA lowered the limit to 90%. Theythen classified second hand smoke as a Group A Carcinogen.

This was openly fraudulent science, but it formed the basis for banson smoking in restaurants, offices, and airports. California bannedpublic smoking in 1995. Soon, no claim was too extreme. By 1998, theChristian Science Monitor was saying that "Second-hand smoke is thenation's third-leading preventable cause of death." The American CancerSociety announced that 53,000 people died each year of second-handsmoke. The evidence for this claim is nonexistent.

In 1998, a Federal judge held that the EPA had acted improperly, had"committed to a conclusion before research had begun", and had"disregarded information and made findings on selective information."The reaction of Carol Browner, head of the EPA was: "We stand by ourscience….there's wide agreement. The American people certainlyrecognize that exposure to second hand smoke brings…a whole host ofhealth problems." Again, note how the claim of consensus trumpsscience. In this case, it isn't even a consensus of scientists thatBrowner evokes! It's the consensus of the American people.

Meanwhile, ever-larger studies failed to confirm any association. Alarge, seven-country WHO study in 1998 found no association. Nor havewell-controlled subsequent studies, to my knowledge. Yet we now read,for example, that second hand smoke is a cause of breast cancer. Atthis point you can say pretty much anything you want about second-handsmoke.

As with nuclear winter, bad science is used to promote what mostpeople would consider good policy. I certainly think it is. I don'twant people smoking around me. So who will speak out against banningsecond-hand smoke? Nobody, and if you do, you'll be branded a shill ofRJ Reynolds. A big tobacco flunky. But the truth is that we now have asocial policy supported by the grossest of superstitions. And we'vegiven the EPA a bad lesson in how to behave in the future. We've toldthem that cheating is the way to succeed.