"How to Lie with Statistics" Had a Tobacco-Funded Sequel
Woody Magazine
We write the things that aren't news
Sept. 29, 2026 (Tue.)
Tobacco Paid for a Sequel to "How to Lie with Statistics." It Stalled 58 Years Ago.
The industry needed more than a rebuttal. It needed the rebuttal to come from someone who seemed to have no stake in it.
In Three Lines
- A study reported last week found that 97 percent of the 175,000 public comments on the FDA's proposed menthol cigarette ban were form letters.
- Sixty years ago, the same industry paid Darrell Huff, author of "How to Lie with Statistics," to write a sequel called "How to Lie with Smoking Statistics."
- This is not a story about one man's lapse. The industry paid for an argument, and for a voice that made the argument look independent.
On Friday, STAT published an account of what happened when the Food and Drug Administration asked the public about banning menthol cigarettes. Over three months in 2022, the agency received 175,000 comments. According to Rutgers University researchers writing in JAMA Network Open last month, 97 percent of them were form letters.
The researchers traced nearly half of those letters to campaigns run by Reynolds American and Altria, the two companies behind the country's biggest menthol brands. Only 2 percent of the comments were written in the sender's own words. In April 2024 the administration put the ban off indefinitely, and the health secretary cited the sheer volume of feedback. The next year, the FDA withdrew it.
When a great many people say the same thing, we tend to listen. That is a sensible instinct. A pile of letters suggests a pile of people who care. The instinct fails when we cannot tell who assembled the pile.
Most of us have learned this lesson from the sponsored post. An influencer praises a face cream, and the praise works because the audience believes no one paid for it. Add the small tag that says "#ad," and the same sentence carries less weight. The words have not changed. What the reader knows about them has.
The tobacco industry understood this 60 years ago. Back then, it did not need 175,000 letters. It paid for one name.
1964: The name the industry chose
In January 1964, the surgeon general released "Smoking and Health." Drawing on animal experiments and population studies, the report concluded that smoking contributes substantially to mortality. In Congress, bills to put warnings on cigarette packs soon followed.
The industry answered that the report rested on weak statistics and that statistics alone could never prove cause. But the same words sound different coming from a tobacco company.
The industry turned to Darrell Huff. In 1954 he had published "How to Lie with Statistics," a slim, illustrated book about the ways numbers mislead. Colleges adopted it almost at once, and it has never gone out of print.
Huff was not a statistician. He was a former magazine editor who made his living as a freelance writer, with no formal training in statistics. What he had was a knack for explaining things, and that one book had made him a public figure in statistics. In 1964 the industry hired him to write a short pamphlet, tentatively titled "The Holes in the Case Against Smoking," for key congressional committees and the industry's own lawyers.
March 22, 1965: The storks
On March 22, 1965, Huff appeared at Senate Commerce Committee hearings on cigarette labeling and advertising. He picked at the surgeon general's report and peppered his case with anecdotes. One involved storks. In the Netherlands, he noted, the number of stork nests on a house rises with the number of children inside. Storks do not deliver babies. Big families live in big houses, and big houses have more room for nests.
Sen. Maurine Neuberger, an Oregon Democrat who had sponsored one of the first cigarette warning-label bills, pressed him on the comparison.
They seemed the same to him, Huff replied.
Two things moving together do not prove that one causes the other. That is true. Huff's own book made the point, and most statistics courses still make it in the first week. The principle was sound. The trouble was where he aimed it. The link between smoking and lung cancer rested on hundreds of studies, each adding a piece.
In 2014, Alex Reinhart, then a doctoral student in statistics at Carnegie Mellon, reconstructed the episode for Significance, a magazine of Britain's Royal Statistical Society. Huff, he observed, never had to knock down the whole case. He only needed to attach a small doubt to each study.
The sequel and the fee
The testimony apparently went over well, and Huff soon began turning the pamphlet into a book. He gave it a title that borrowed from his best seller: "How to Lie with Smoking Statistics." Payment came from a special industry account. Under the terms of a November 1965 industry memo, an industry committee would decide whether the work could be published at all.
An industry expense report from January 1968 shows Huff had been paid $9,000, with $1,000 more still owed. That year, the median American family earned $8,630. Spread over several years, the payments still added up to more than a typical household earned in one.
1968: One line of disclosure
In the spring of 1968, the industry was embarrassed. Articles defending smoking in True and The National Enquirer had looked like independent journalism. They turned out to be underwritten by the industry, and copies had been mailed anonymously to lawmakers and public figures. The backlash brought a Federal Trade Commission investigation.
To keep it from happening again, industry lawyers proposed a line for the preface of Huff's book. It was, in effect, the "#ad" tag.
The line mentioned a grant. It said nothing about the industry's power to decide whether the book would appear.
Macmillan, a major New York publisher, was weighing the book. It first sent the manuscript to an outside statistician. The reviewer, K.A. Brownlee of the University of Chicago, called it a credit to its publisher. The industry had already hired Brownlee as a paid consultant. In May, Macmillan decided to publish.
In July, industry representatives told Macmillan that Huff had given paid testimony. The publisher wrote back, surprised that he had any involvement beyond a grant, and asked for details. Tiderock, the industry's public relations firm, replied that his ties were "not really extensive." The deal went ahead. By September the terms were set: a hardcover first, followed by 100,000 paperback copies.
In November, as Macmillan was about to mail a final contract, the project collapsed. The record does not say why. The manuscript and letters surfaced 30 years later, when a 1998 settlement between the states and the major tobacco companies made tens of millions of pages of internal documents public. Huff died in 2001.
One name, 175,000 letters
Huff's sequel and last week's form letters differ in scale. One carried a famous author's name. The other was an unsigned flood. They did the same job. In both cases, words written by the industry carried weight only if they seemed to come from outside it.
An earlier study from the same Rutgers project shows how the letters were made. Tobacco companies emailed customers during the comment period and steered them to company-run websites. The sites assembled letters from menus of prewritten arguments and submitted them to the FDA. The senders may have been real consumers. The sentences came from the companies. Most letters lacked identifying information, the researchers wrote, so it was unclear whether they reflected genuine responses from citizens. When we count voices, we rarely see who gathered them.
In 1968, industry lawyers wanted one line of disclosure in Huff's preface. The book never appeared. Most of the letters in last week's report did not say who sent them. A sponsored post loses its value once readers see the tag. In 60 years, what changed was how to keep the tag out of sight.
Sources
- Source ↗ Alex Reinhart, "Huff and puff," Significance 11(4), 2014
- Source ↗ Alex Reinhart, "Huff and puff," preprint (2014)
- Source ↗ Alex Reinhart, "The history of How to Lie with Smoking Statistics," index of UCSF tobacco documents (2014)
- Robert N. Proctor, "Golden Holocaust," University of California Press, 2012
- U.S. Senate Committee on Commerce, "Cigarette Labeling and Advertising" hearings, March 22 to April 2, 1965
- Source ↗ U.S. Census Bureau, "Average Family Income Up 8 Percent in 1968" (P60-63), 1969
- Source ↗ Jason McLure, "How Big Tobacco flooded the FDA with public comments to kill a menthol cigarette ban," STAT, Sept. 25, 2026
- Source ↗ Rutgers University, "Tobacco Companies Flooded Federal Comment Process That Stalled Menthol Cigarette Ban," 2026
- Andrea C. Villanti et al., "Public Comments and Federal Rulemaking for Menthol Cigarettes and Flavored Cigars," JAMA Network Open, 2026
댓글
댓글 쓰기