Review: Merchants of Doubt (Naomi Oreskes, Erik M. Conway)

I hadnt planned on reading this book, just saw a random comment about it on a discussion board. I figured i might as well read it since i found a free pdf:

The book turned out to be moderately interesting. It is mostly about the history of denying pollution problems, usually by marked-friendly people. It underlines the problem of lobbying and the need for transparency.

Call it the “Tobacco Strategy.” Its target was science, and so it relied

heavily on scientists—with guidance from industry lawyers and public re­

lations experts—willing to hold the rifle and pull the trigger. Among the

multitude of documents we found in writing this book were Bad Science: A

Resource Book—a how-to handbook for fact fighters, providing example af­

ter example of successful strategies for undermining science, and a list of

experts with scientific credentials available to comment on any issue about

which a think tank or corporation needed a negative sound bite.14

Sounds too good to be true, so i checked, and its true:

These findings shouldn’t have been a surprise. German scientists had

shown in the 1930s that cigarette smoking caused lung cancer, and the Nazi

government had run major antismoking campaigns; Adolf Hitler forbade

smoking in his presence. However, the German scientific work was tainted by

its Nazi associations, and to some extent ignored, i f not actually suppressed,

after the war; it had taken some time to be rediscovered and independently

confirmed.21 Now, however, American researchers—not Nazis—were calling

the matter “urgent,” and the news media were reporting it.22 “Cancer by the

carton” was not a slogan the tobacco industry would embrace.

Typical case of guilt by association. Im still waiting for eugenics to be free from that association, just like anti-smoking is today, and vegetarianism, etc.

C. C. Little was a renowned geneticist, a member of the U.S. National

Academy of Sciences and former president of the University of Michigan.34

But he was also well outside the mainstream of scientific thinking. In the

1930s, Little had been a strong supporter of eugenics—the idea that soci­

ety should actively improve its gene pool by encouraging breeding by the

“fit” and discouraging or preventing breeding by the “unfit.” His views

were not particularly unusual in the 1920s—they were shared by many

scientists and politicians including President Theodore Roosevelt—but

nearly everyone abandoned eugenics in the ’40s when the Nazis made

manifest where that sort of thinking could lead. Little, however, remained

convinced that essentially all human traits were genetically based, includ­

ing vulnerability to cancer. For him, the cause of cancer was genetic weak­

ness, not smoking.

Well, there is no ‘the’ cause of cancer, unless one thinks of things like the breakdown of apoptosis, i.e. programmed cell death (cells kill themselves when they detect genetic anomalies in themselves, a process which necessarily has to be disabled for cancer to spread, otherwise it just kills itself).

The treatment of eugenics above is also unfair. See e.g.:

And surely genetics has a lot to do with cancer. I havent looked into the issue before, but i did a quick check for a twin study on the heritability of lung cancer. I found this:

Lichtenstein, Paul, et al. “Environmental and heritable factors in the causation of cancer—analyses of

cohorts of twins from Sweden, Denmark, and Finland.” New England Journal of Medicine 343.2 (2000): 78-85.


Bolin and his Swedish colleagues had made “mass balance arguments” :

they considered how much sulfur could be supplied by the three largest

known sources—pollution, volcanoes, and sea spray—and compared this

with how much sulfur was falling as acid rain. Since there are no active vol­

canoes in northern Europe, and sea spray doesn’t travel very far, they de­

duced that most of the acid rain in northern Europe had to come from air

pollution. Still, this was an indirect argument. To really prove the point,

you’d want to show that the actual sulfur in actual acid rain came from a

known pollution source. Fortunately there was a way to do this—using


Scientists love isotopes—atoms of the same element with different

atomic weights, like carbon-12 and carbon-14—because they are excep­

tionally useful. I f they are radioactive and decay over time—like carbon-

14—they can be used to determine the age of objects, like fossils and

archeological relics. I f they are stable, like carbon-13—or sulfur-34—they

can be used to figure out where the carbon or sulfur has come from.19 Dif­

ferent sources of sulfur have different amounts of sulfur-34, so you can

use the sulfur isotope content as a “fingerprint” or “signature” of a partic­

ular source, either natural or man-made. In 1978, Canadian scientists

showed that the isotopic signature of sulfur in acid rain in Sudbury was

identical to the sulfur in the nickel minerals being mined there. In later

years, some skeptics would argue that the acid in acid rain came from vol­

canoes (they would say the same about fluorine and ozone depletion, and

about C0 2 and global warming), but the isotope analysis showed that couldn’t

be true.20 In any case there are no active volcanoes in Ontario.

pretty cool.

Political Action and the U.S.-Canadian Rift

In 1979, the United Nations Economic Commission for Europe passed the

Convention on Long-range Transboundary Pollution. Based on the Decla­

ration of the U.N. Conference on the Human Environment—the one for

which Bert Bolin’s report had been prepared—the convention insisted that

all nations have responsibility to “ensure that activities within their juris­

diction or control do not cause damage to the environment of other states

or of areas beyond the limits of national jurisdiction.”24 Henceforth, it would

be illegal to dump your pollution on someone else, whether you did it with

trucks or with smokestacks.

Classic case of negative externalities.

Like most of his colleagues, Singer believed there was a need for more

science, but in 1970 he argued that one cannot always wait to act until mat­

ters are proven beyond a shadow of a doubt. Singer cited the famous essay

“The Tragedy of the Commons,” in which biologist Garrett Hardin argued

that individuals acting in their rational self-interest may undermine the

common good, and warned against assuming that technology would save

us from ourselves. ” If we ignore the present warning signs and wait for an

ecological disaster to strike, it will probably be too late,” Singer noted. He

imagined what it must have been like to be Noah, surrounded by “compla­

cent compatriots,” saying,” ‘Don’t worry about the rising waters, Noah; our

advanced technology will surely discover a substitute for breathing/ If it

was wisdom that enabled Noah to believe in the ‘never-yet-happened, ’ we

could use some of that wisdom now,” Singer concluded.63

But prevention is usually cheaper than clearing up afterwards. So it is better to prevent if one can.

Singer also presumed that the costs were mostly accrued in the present,

but the benefits in the future, and therefore the latter had to be discounted

in order to make them commensurate with the former. (That is to say, a

dollar in the future is not worth as much to you as a dollar now, so you

“discount” its value in your planning and decision making. How much

you discount it depends in part on inflation, but also in part on how much

you value the future.) Discounting would later become a huge issue in as­

sessing the costs and benefits of stopping global warming, as long-term

risks can be quickly written off with a sufficiently high discount rate.111

Im pretty sure there is a name for this kind of bias, but i cant recall its name or find it on Wikipedia’s list.

In later years, emissions trading would be used to reduce acid pollution—

and today many people are looking to such a system to control the green­

house gases that cause global warming. Yet economists (and ordinary

people) know that markets do not always work.114 Indeed, many economists

would say that pollution is a prime example of market failure: its collateral

damage is a hidden cost not reflected in the price of a given good or service.

Milton Friedman—the modem guru of free market capitalism—had a name

for such costs (albeit an innocuous one): he called them “neighborhood


Or ”negative externalities”.

David Hounshell is one of America’s leading historians of technology.

Recently he and his colleagues at Carnegie Mellon University have turned

their attention to the question of regulation and technological innovation.

In an article published in 2005, “Regulation as the Mother of Innovation,”

based on the Ph.D. research of Hounshell’s student, Margaret Taylor, they

examined the question of what drives innovation in environmental control

technology. It is well established that the lack of immediate financial bene­

fits leads companies to underinvest in R & D, and this general problem is

particularly severe when it comes to pollution control. Because pollution

prevention is a public good—not well reflected in the market price of

goods and services—the incentives for private investment are weak. Com­

petitive forces just don’t provide enough justification for the long-term in­

vestment required; there is a lack of driving demand. However, when

government establishes a regulation, it creates demand. I f companies know

they have to meet a firm regulation with a definite deadline, they respond—

and innovate. The net result may even be cost savings for the companies, as

obsolete technologies are replaced with state-of-the art ones, yet the com­

panies would not have bothered to make the change had they not been

forced to.

That is unexpected under a rational choice theory of companies. Are companies also irrational?

Satellites don’t just “collect” data in the way that nineteenth-century ge­

ologists collected rocks or biologists collected butterflies; they detect sig­

nals and process them. The electronics and computer software involved

are very complex and sometimes things go awry, so procedures are in­

cluded for screening and rejecting “bad” data. This was the case here. The

satellite processing software contained computer code designed to flag

ozone concentrations below a certain level—180 Dobson units—as unre-

alistically low, and therefore probably bad data.44 Concentrations that low

had never been detected in the stratosphere and could not be generated by

any existing theoretical model, so it seemed like a reasonable choice.

When some of the Antarctic ozone retrievals had come in well below 180,

they were catalogued as errors. The instrument’s science team had a map

that showed the errors concentrated over the Antarctic in October, but they

had ignored it, assuming the instrument was faulty. A healthy skepticism

about their machinery led them to dismiss crucial data.

When Stolarski double-checked, he found that the depleted region

covered all of Antarctica—and the “ozone hole” was bom. It wasrft an instru­

ment error. It was a real phenomenon. It had been detected by the satellites.

And it defied expectation.

That is why one does not program such things into the primary data collectors. In general, discarding data is bad bad bad, and any such process shud be explicit.

Officially the mission of the Alexis de Tocqueville Institution is to pro­

mote democracy; in 1993 the Institution decided to promote democracy

by defending secondhand smoke. “EPA and the Science of Environ­

mental Tobacco Smoke” was written by Fred Singer and Kent Jeffreys.75 The

Tocqueville Institution had anointed Jeffreys with the title of “adjunct scholar,”

but he was in fact a lawyer affiliated with the Cato Institute, the Competi­

tive Enterprise Institute, and the Republican Party. He was well-known for

his attacks on Superfund—the federal fund designed to pay for the cleanup

of toxic waste sites—and for his advocacy of “free-market environmental­

ism.” One of his slogans was “behind every tree should stand a private . . .

owner.” To prevent overfishing, Jeffreys wanted to privatize the oceans.76

The defense of secondhand smoke was part of a larger report criticizing

the EPA over radon, pesticides, and the Superfund, but the center of it—

and the focus of the accompanying press releases—was what Singer and

Jeffreys called “Case Study No. 1: Environmental Tobacco Smoke.” It be­

gan by accusing the federal government of seeking a ban on smoking—

although there was no pending legislation to do so—and asserting that the

vehicle of the alleged ban would be the EPA. But the EPA had not asked

for a ban, so how did Singer and Jeffreys build their case? By asserting that

“scientific standards were seriously violated in order to produce a report to

ban smoking in public settings.”77 What was the alleged violation? The

EPA panel had assumed a linear dose-response curve. They had assumed

the risk was directly proportional to the exposure.

Singer and Jeffreys argued that the EPA should have assumed a “thresh­

old effect”—that doses below a certain level would have no effect. Citing the

old adage “the dose makes the poison,” they insisted that there might be a

threshold value below which no harm occurred. Since the EPA had failed to

provide proof that this wasn’t so, the linear-dose response assumption was


A threshold model is indeed possible, and one cannot just assume a linear non-threshold model. Not for smoke and not for radiation either. One needs data points that establishes the LNT model before one can use it for predicting things.

The real culprit in smoking, Seitz argued, was smoke, and this was “no

more wanted by smokers than coffee grounds by cappuccino addicts or a

hangover by drinkers of red wine.” So Seitz suggested that the U.S. gov­

ernment should figure out how to remove the smoke from cigarettes.

“Only one-tenth of one percent of a cigarette is nicotine, and it should not

take a rocket scientist to devise a means to volatilizing that small drop of

active ingredient without generating a thousand times its weight in burn­

ing leaves.”122 Seitz was proposing that the government should spend tax­

payer money figuring out how to safely deliver nicotine—an addictive and

toxic substance—to the American people.

This sort of exigent approach might make sense for methadone since it

helps people get off heroin, whose dangers to individuals and society are

both grave and immediate. But what public good would be served by the

government deliberately enabling people to continue to smoke?

The authors apparently do not understand harm reduction. The goal is not to stop people from smoking. The goal is to reduce harmful effects. If one can invest less dangerous alternatives, this is a great idea. Some people will never give up their nicotine dependence, and the only way to reduce harm for them without overly draconic regulation (banning smoking, and with a huge surveillance society to back it up), is to have less harmful alternatives.

See also:

The Global Climate Coalition meanwhile had circulated a report enti­

tled “The IPCC: Institutionalized Scientific Cleansing” to reporters, mem­

bers of Congress, and some scientists. By chance, anthropologist Myanna

Lahsen interviewed Nierenberg about his “skepticism” about global warm­

ing two weeks before the Working Group I Report was published, and

found that he had a copy of the coalition report. He had evidently accepted

its veracity, even though there was no way to compare its claims against

the real chapter 8 (since the latter had not yet been released). He quoted its

claims to Lahsen, telling her that the revisions had “just altered the whole

meaning of the document. Without permission of the authors.” Moreover,

he claimed, “Anything that would imply the current status of knowledge is

so poor that you can’t do anything is struck out.” 144 That was hardly true;

Santer7 s panel had included six pages of discussion of uncertainty in the fi­

nal text. But Bill Nierenberg knew all about altering scientific reports for

political reasons, so perhaps he followed the adage that the best defense is

offense. Or perhaps he was guilty of “mirror imaging,” as Team B had

accused the CIA of in 1976: assuming that his opponents thought and op­

erated the way he did.

The term for this is projection bias.

Ignore the Freudian crap.

Fred Singer gave his game away when he denied the reality of the ozone

hole, suggesting that people involved in the issue “probably [have]. . .

hidden agendas of their own—not just to ‘save the environment” but to

change our economic system . . . Some of these ‘coercive Utopians’ are so­

cialists, some are technology-hating Luddites; most have a great desire to

regulate—on as large a scale as possible.”33 He revealed a similar anxiety

in his defense of secondhand smoke: ” I f we do not carefully delineate the

government*s role in regulating [danger]. . . there is essentially no limit to

how much government can ultimately control our lives.”34 Today tobacco,

tomorrow the Bill of Rights. Milton Friedman said much the same in Cap­

italism and Freedom: that economic freedom is as important as civic free­

dom, because i f you lose one, it is only a matter of time before you lose the

other.35 And so one must defend free markets with the same vigor and vig­

ilance as free speech, free religion, and free assembly.

He wasnt completely wrong about this. There is a crazy amount of overregulation in some areas, altho probably not in the areas he was complaining about here.