The case for home smoking bans, Voxsplained

This one headline perfectly encapsulates why’s coverage of tobacco policy is terrible:


That’s a lie, actually. Vox’s tobacco coverage is bad for more reasons than can be encapsulated in one headline, and it’s not really much better or worse than any other publication’s, but I’m trying to keep with the current form of writing on the internet. “Not that the old way was perfect,” writes my old school blog pal Jason Kuznicki, “but nobody fisks anymore, and for that we ought to be ashamed.” I agree. So how about some fisking? Perhaps my rank on Technorati will go up if you link to this.

Here’s Vox’s German Lopez writing on the Department of Housing Urban Development’s announcement that it will be banning smoking in all 1.2 million of the nation’s public housing units. “One major problem with this policy is it seems to single out low-income people,” writes Lopez, providing a glimmer of hope that he might think twice about intruding into 1.2 million homes of the less well off. “But there’s an easy solution to that: Indoor smoking should be banned everywhere — inside bars, restaurants, your home. Full stop.”

Never mind the casual authoritarianism behind the sentiment, or the constitutional issues it raises. (Has the Commerce Clause been interpreted so broadly as to allow this sort of thing? Eh, probably.) And never mind whether enforcing this particular ban by evicting some of the poorest members of society from their homes is a humane idea — Megan McArdle covered that here. Let’s just look at some numbers, the stock in trade of explanatory journalism. Lopez supports his advocacy for enforcing a smoking ban even in privately owned homes by noting the scope of the problem:

According to the Centers for Disease Control and Prevention’s most recent data, smoking kills 480,000 people each year. Secondhand smoke alone kills nearly 42,000 people. To put that in perspective, that’s around 8,000 more people than die to either car crashes or gun violence.

The 480,000 deaths figure got picked up in just about every story about the smoking ban in federal housing, including additional Vox posts this weekend from Julia Belluz and Dylan Matthews, from whom I’d expect better. (Matthews also seems to wish that we could ban smoking in private homes, but regrets that people would “freak out” if we tried. Imagine!) Both the total number of deaths and the number attributable to secondhand smoke are dubious, however, and none of the writers appear to have put any effort into understanding where they come from.

Let’s take the total number first. As Belluz notes, the smoking rate in the United States has dropped from 21 percent of Americans in 2005 to 17 percent in 2014. The decline gets more significant the farther back you look. Yet the CDC’s estimates of mortality caused by smoking don’t seem to reflect this. The CDC currently tells the public that smoking kills 480,000 people per year. That’s nearly 40,000 more annual deaths than it attributed to smoking from 2000-2004, when it claimed that smoking killed 443,000 people. And that’s more than the 438,000 deaths per year the CDC assigned to smoking from 1997-2001.

It’s worth asking why smoking mortality seems be increasing even as smoking rates are doing down. This is partially due to population growth and demographic changes; mortality rates would be a far more useful metric for comparison over time. And since smoking is associated with chronic diseases, some lag in the figures is to be expected. But still, the 480,000 figure is controversial.

Unlike deaths due to car accidents and gun shots, which are discrete events that can be tallied, there’s no direct way to count deaths caused by smoking (or especially deaths caused by secondhand smoke). Death certificates don’t say things like, “Bob died from heart disease made marginally worse by his cigarette habit.” Instead, researchers compare the prevalence of causes of death that are associated with smoking to the rate of smoking among different age groups, attempt to account for other confounding factors, and give their best shot at an estimate of how many people who died would still be alive in a tobacco-free world. Exactly how the current guess of 480,000 deaths is arrived at isn’t really clear.

The CDC’s numbers have been criticized in academic journals. A 2007 paper by Brad Rodu and Philip Cole in Nicotine and Tobacco Research offers a different model, which Rodu summarized in an article for Cato’s Regulation that criticizes the lack of transparency in the CDC estimates. With the Rodu-Cole model, you see the good news you’d expect from fifty years of declining smoking rates: “The U.S. mortality rate attributable to smoking declined about 35% between 1987 and 2002.”

OK, that’s just one paper, and from a source you may consider less reliable than the CDC. But in 2012 the same journal published another paper by Brian Rostron, whose affiliation is with the Food and Drug Administration’s Center for Tobacco Products. Rostron is also critical of the CDC’s methods, noting that they “have not been substantially revised since their introduction in the 1980s.” Rostron’s revised estimate for annual smoking-related deaths in 2004 is 380,000. The paper concludes, “we have estimated smoking-attributable mortality over time and found that the number of estimated deaths has peaked and finally begun to decline for both men and women in the United States.” If that’s correct, then we should have experienced a decade of declining deaths since 2004, and the CDC’s 480,000 figure is wildly off target.

[Update: German Lopez brings up a new paper in the New England Journal of Medicine, which suggests some excess mortality missed by earlier figures due to diseases that have not yet been definitively linked to smoking. Some of the relative risks are low, and there may be confounding variables, but the sample size is large. If those links bear out, then the sets of estimates above would be revised upward.]

That’s the total number of deaths attributable to smoking. How about the fraction attributed to secondhand smoke? If it’s difficult to figure out how many smokers are dying because of tobacco use every year, guessing how many people die from secondhand smoke is even more problematic. The CDC currently pegs the number at 41,000 deaths per year, of which a little over 7,000 are due to lung cancer and about 34,000 are due to heart disease.

Let’s tackle lung cancer first. If you want a rigorous, scientific indictment of secondhand smoke as a cause of lung cancer, you probably can’t do better than the 2006 report from The Surgeon General. The report concludes unequivocally that environmental tobacco smoke increases the risk of lung cancer in non-smokers. But by how much? This is expressed as a relative risk. A relative risk of 1 would indicate that people exposed to secondhand smoke are no more likely to develop lung cancer than those who are not exposed. (People who smoke habitually have a relative risk in the neighborhood of 16, which is huge.) What would you guess is the relative risk for non-smokers who are exposed to secondhand smoke over the long-term at their home or workplace?

If you guessed barely greater than 1, you’re correct. The report’s table of meta-analyses puts the relative risks of exposure at home for non-smoking spouses or at work for non-smoking employees in the range of 1.12 at the low end to 1.43 on the high end. That’s really low! Low enough to be sensitive to bias in how various studies are weighted, and low enough that it can’t be reliably detected in studies of secondhand smoke exposure. (One of the meta-analyses of childhood secondhand smoke exposure even suggests that children who are exposed to tobacco smoke are less likely to develop lung cancer. As they say on Twitter, “Whoa, if true!”)

How you interpret this ambiguity likely depends on your political priors. With that in mind it’s interesting to see the Journal of the National Cancer Institute publishing a headline like “No Clear Link Between Passive Smoking and Lung Cancer.” That’s a 2013 story about a cohort study of 76,000 women that “confirmed a strong association between cigarette smoking and lung cancer but found no link between the disease and secondhand smoke.” To go on:

The incidence of lung cancer was 13 times higher in current smokers and four times higher in former smokers than in never-smokers, and the relationship for both current and former smokers depended on level of exposure. However, among women who had never smoked, exposure to passive smoking overall, and to most categories of passive smoking, did not statistically significantly increase lung cancer risk. The only category of exposure that showed a trend toward increased risk was living in the same house with a smoker for 30 years or more. In that group, the hazard ratio for developing lung cancer was 1.61, but the confidence interval included 1.00, making the finding of only borderline statistical significance. […]

But many studies that showed the strongest links between secondhand smoke and lung cancer were case–control studies, which can suffer from recall bias: People who develop a disease that might be related to passive smoking are more likely to recall being exposed to passive smoking. […]

However, Silvestri finds some reassurance in the passive-smoking findings. “We can never predict who is going to develop lung cancer,” he said. “There are other modifiers. But you can say, with regard to passive smoke, it’s only the heaviest exposure that produces the risk. We kind of knew that before, but it’s a little stronger here.”

“We’ve gotten smoking out of bars and restaurants on the basis of the fact that you and I and other nonsmokers don’t want to die,” said Silvestri. “The reality is, we probably won’t.”

The study has not, to my knowledge, appeared in a journal, but if you’ve followed debates about secondhand smoke at all, then the candor in the commentary above is refreshing. The association between secondhand smoke and lung cancer has never been as scary as anti-smoking advocates make it out to be. Read Christopher Snowdon for a sense of the numbers; the ten studies with the largest sample sizes find risk ratios of 1.29, 1.11, .70, 1.03, 1.53, 1.10, .90, and .96. This is hardly the stuff of nightmares, and it’s difficult to imagine basing an accurate model of secondhand smoke mortality on such tiny risks.

How about deaths due to heart disease? Economist Kip Viscusi provides what I think is a fair summary in his 2002 book Smoke-Filled Rooms: “Despite the central role of lung cancer in the ETS [environmental tobacco smoke] debates, the heart disease linkage may be greater, and the risk levels much larger.” This accords with the CDC’s higher attribution of deaths due to heart disease than to lung cancer.

Studies on the short-term effect of smoking bans, however, are not encouraging. In the early 2000s, a wave of research conducted in small cities that had implemented smoking bans concluded that they could drastically reduce heart attacks. Skeptics criticized these studies for relying on data from small populations. Newer research confirms that they were right to be doubtful. The most notable example is a 2014 study of the state of Colorado in The American Journal of Medicine which was co-authored by researchers who had previously published papers showing that small towns in Colorado experienced declines in heart attacks after banning smoking. In their new research, they write:

Although local smoking ordinances in Colorado previously suggested a reduction in acute myocardial infarction hospitalizations, no significant impact of smoke-free legislation was demonstrated at the state level, even after accounting for preexisting ordinances.


These analyses support the hypothesis that small study populations may be more likely to find dramatic changes in acute myocardial infarction incidence, whereas increasing the study sample size attenuates the magnitude of the reduction. Also, review of the studies in aggregate reveals data asymmetry that suggests the potential for publication bias or heterogeneity not entirely explained by a random-effects meta-analysis. […] Overall, a review of published research shows that acute myocardial infarction RR reduction appears inversely related to sample size. […]

Available evidence suggests that acute myocardial infarction incidence has been decreasing dramatically, unrelated to smoke-free ordinances. […] This emerging evidence highlights the importance of accounting for secular trends in acute myocardial infarction incidence before definitive attribution to smoke-free ordinances can be made. […]

Overall, available evidence suggests that the decrease in acute myocardial infarction incidence associated with reductions in secondhand smoke exposure may be substantially lower than originally estimated.

The decline in hospitalizations due to heart attacks is an important factor to consider. From the same paper:

Data from the Centers for Disease Control National Environmental Public Health Tracking Network recently evaluated secular trends in 20 Network states from 2000 to 2008 using a longitudinal linear mixed effects model. The authors documented a statistically significant overall decrease in age-adjusted acute myocardial infarction hospitalization rates, with most states showing more than a 20% decline during the period. This temporal reduction in acute myocardial infarction incidence is of a magnitude that exceeds the reduction observed in many smoking ordinance studies. Despite this, some analyses have not accounted for secular trends.

This reduction in heart attacks doesn’t get much attention, but it comes up in Rostrom’s paper as well: “Deaths from ischemic heart disease for persons aged 65 and over decreased from 201,000 in 2000 to 158,000 in 2007 for U.S. men, and from 233,000 in 2000 to 170,000 in 2007 for U.S. women.”

And you see similar downward trends in the UK. A study of Scotland from 2002-2010 found that deaths from heart disease declined by 43%. Nearly half of this decline was credited to improved treatment. Changes in the rate of smoking were credited for only 4% of the decline. One has to wonder: Given all the advances of the past decade or so, why is the CDC’s estimate of secondhand hand smoke deaths caused by heart disease still so high?

The decline in heart attacks and deaths due to heart disease has occurred over a similar time period as the spread of smoking bans. The best source for information on the adoption of smoking bans in the United States is Americans for Nonsmokers Rights, who maintains a database of smoking restrictions. According to their tracking, the number of jurisdictions in the United States with 100% smokefree laws in all workplaces, restaurants, and bars has increased from two in 1993 to 790 in October 2015. Estimates of the number of deaths caused by secondhand smoke, however, have been strangely consistent. The current estimate from the CDC is 41,000. That’s a little less than the 2006 Surgeon General’s report estimate of nearly 50,000. And if you go back to 1990, you can learn from The New York Times that:

The newer understanding of the health hazards of passive smoking were underscored in a report at a world conference on lung health in Boston last week. Dr. Stanton A. Glantz of the University of California at San Francisco estimated that passive smoke killed 50,000 Americans a year, two-thirds of whom died of heart disease.

Sound familiar? It’s almost as if the claim that secondhand smoke kills around 50,000 Americans per year is based on its political utility rather than any firm grounding in epidemiology.

To review: In the past two decades, the rate of smoking has substantially declined. Treatment and prevention of heart disease, which allegedly causes the vast majority of deaths associated with secondhand smoke, has improved dramatically. Doubts about the magnitude of the association between secondhand smoke and both lung cancer and heart attack incidence have spread to mainstream academic journals. And exposure to secondhand smoke has been greatly reduced thanks to declining smoking rates, the proliferation of smoking bans, and changing social norms. Yet the number of Americans dying from secondhand smoke exposure has, supposedly, barely nudged downward over twenty-five years. These facts don’t hang together.

So how many deaths does secondhand smoke cause per year? I’ve been writing about tobacco policy intermittently for nearly a decade and in-depth for the past year, and I’m reluctant to commit to a number. I can tell you that I’m extremely skeptical of the CDC’s figure of 41,000, and I think that any responsible journalist ought to be skeptical too.

This brings up problems with contemporary reporting on tobacco policy and with explanatory journalism in general. In the old days of tobacco reporting, a policy announcement such as the ban on smoking in public housing would have gotten just as much coverage as it did this week. But reporters covering it might have also sought comment from pro-smoking sources. A lot of what those sources said would have been total spin and bullshit of the sort satirized in Christopher Buckley’s Thank You for Smoking, but some of what they said might have been valid criticism that pointed writers to legitimate weaknesses in their stories. That dynamic has been much reduced since the Master Settlement Agreement of 1998 that dismantled pro-tobacco organizations.

A couple years ago I interviewed Michael Siegel, a professor at the Boston University School of Health who formerly worked for the CDC’s Office on Smoking and Health, and he was blunt about the deterioration of scientific integrity in the anti-smoking movement. Siegel is generally in favor of indoor smoking bans, but has become a vocal critic of how anti-tobacco groups’ exaggerate the dangers of secondhand smoke. He blames this in part on the loss of an adversarial process:

The current state of tobacco control I would describe, quite sadly, as misguided. It is now guided more by ideology and politics than by science. Ironically, I think one of the reasons it has lost its way is that some time around 2000 or so, the tobacco industry relinquished its watchdog role. Organizations in tobacco control used to be very careful because they knew the tobacco industry was watching and would call them on it if they exaggerated or distorted the truth. But after around 2000, the tobacco companies stopped playing this role and basically allow the tobacco control groups to say anything they want.

Good journalism is more than regurgitating a scary-sounding number from an authoritative source. It’s understanding the motivation and reliability of your sources, and seeking out potential opposition to see if there’s a counterargument that they failed to mention. Reporters know to be skeptical of tobacco companies. They haven’t learned to be skeptical of anti-smoking sources, and given the changes in the regulatory landscape of tobacco it’s long past time that they do. It’s the only way to avoid uncritically reporting that a smoking ban can decrease heart attacks by 60% in just six months, or that smokers are “contaminated” and “actually emit toxins,” or that people who use e-cigarettes are “inhaling Chinese-made antifreeze,” or that 41,000 Americans are dying from secondhand smoke exposure in 2015.

The past few days of coverage at Vox, with three different writers repeating the same dubious statistic and none of them investigating it, doesn’t raise my confidence in their model of explanatory journalism. This is explanation divorced from skepticism. The current media environment makes it easy to find numbers to support one’s political view with just a few minutes of Googling, appearing to write from a perspective of data-driven empiricism, and harvesting those all-important clicks. (Is Vox’s “The case for banning smoking indoors — even in your home” any less dumb than Slate’s recent piece arguing that spooning is sexist?) But when those numbers are taken at face value and without context, the writers are just one bad statistic away from calling in the cops to search for ash trays on your kitchen counter.

I like Vox. I really do, even if I find myself turning more often to their entertainment writing than their policy pieces. I view the site as a generally useful source of information about complex topics that it’s difficult for any one person to know in detail. Yet occasionally the site covers a topic that I do know in detail. And when it does, I have to wonder how much of their other coverage is equally superficial and credulous.

Gasp! New Orleans passes a smoking ban

In my first contribution to Eater, I put the New Orleans smoking ban into a national context:

As in most cities, New Orleans’ smoking ban debate centered on the trade-offs between the interests of business owners who fear lost patronage and the interests of customers and employees in avoiding secondhand smoke. It’s an apt time to inquire into what’s really at stake with smoking bans. Will the sky fall for bars and casinos in New Orleans? And will banning smoking save as many lives as advocates promise?

The answers may surprise you!

Vaping in the Oregonian

In today’s Oregonian, I look back at a column I wrote in 2008 and say, “I told you so.”

When the clock struck midnight on New Year’s Day 2009, Oregon ushered in its statewide smoking ban in bars and restaurants. I was at the Horse Brass Pub, one of Portland’s most notoriously smoky drinking dens, enjoying one last cigar with a bunch of other patrons who were none too happy about the new rules.

The ban, we were told, was necessary to protect employees and customers alike from secondhand smoke. Health researchers had conducted dozens of studies attempting to show that exposure endangered nonsmokers. Some of the results were medically implausible, but ban advocates at least made the effort of demonstrating actual harm to actual humans.

Many of us doubted that the evidence really mattered. As I wrote in The Oregonian/OregonLive at the time, “Protecting workers is simply the polite fiction by which nonsmokers have imposed their will on an increasingly unpopular minority.”

We suspected this, but how could we prove it? What if there were a device that looked like a cigarette and mimicked the effects of smoking, yet emitted a mostly harmless vapor instead of tobacco smoke? If authorities tried to ban that too, without bothering to establish that it endangered anyone, then our suspicions would be vindicated.

That device exists. It’s called an e-cigarette. And sure enough, the Multnomah Board of County Commissioners is voting on whether to ban its use indoors. The Legislature, too, may expand the state’s smoking ban to cover vaping.

Read the whole thing.

Multnomah County misleads bar owners

As a follow-up to my article in yesterday’s Oregonian about the failed attempt to include e-cigarettes in the state’s smoking ban, today I’m posting a memo Multnomah County officials have sent to local bars and restaurants. In it, they mislead business owners about the dangers of e-cigarettes, telling them:

State law does not currently prevent the use of e-cigs; however business owners are encouraged to include e-cigs in no-smoking policies. E-cigs pose serious health risks and challenges to enforcement of the Smokefree Workplace Law as it appears people are smoking indoors.

The letter then recommends that businesses include e-cigarettes in their no-smoking policies, adopt completely smokefree outdoor dining areas, and adopt a completely tobacco-free policy for their entire properties. (Here’s a PDF of the memo.)

There are valid reasons why a bar or restaurants might ban the use of e-cigs, such as the fact that some guests find them annoying. But county officials’ claim that the devices pose “serious health risks” is completely unsubstantiated. There’s not even much evidence that e-cigarettes are dangerous for users, much less for bystanders exposed to vapor secondhand.

There have been two recent studies on exposure to e-cigarettes in realistic indoor conditions. They are summarized here and here. Conclusion of the first:

… the quality and quantity of chemicals released in the environment [by vaping] are by far less harmful for the human health compared to regular tobacco cigarettes. Evaporation instead of burning, absence of several harmful chemicals from the liquids and absence of sidestream smoking from the use of the e-CIG are probable reasons for the difference in results.

And the second:

The study showed that e-cigarettes are a source of secondhand exposure to nicotine but not to combustion toxicants… Using an e-cigarette in indoor environments may involuntarily expose nonusers to nicotine but not to toxic tobacco-specific combustion products.

Even in the case of nicotine, exposure from real cigarettes was ten times higher than that from e-cigarettes.

Those are studies of indoor use. Multnomah County’s advice is to ban them outdoors too. The idea that indoor e-cigarette use could be harmful to bystanders is at least worthy of investigation, although the evidence so far is that it’s nothing to worry about. The idea that outdoor use presents serious health risks is wildly implausible.

This is yet another example of how the crusade against e-cigarettes is driven by unscientific alarmism rather than any empirical evidence of danger. County officials have shown that they have no credibility on the issue by misleading local business owners about the alleged risks.

Oregon’s war on vaping

The Oregon legislature recently failed to pass completely sensible restrictions on selling e-cigarettes to minor, an effort undermined by more extreme anti-smokers who were more intent on banning vaping in workplaces, bars, and restaurants. In today’s Oregonian, I write about lawmakers’ misguided attempt to include e-cigarettes in the smoking ban and their next proposal to impose new taxes on them. An excerpt:

I never thought I’d say this, but I’m actually getting nostalgic for the original smoking ban debate. Advocates exaggerated the dangers of secondhand smoke, but at least they made an effort to ground their views in science and demonstrate that non-smokers were being harmed.

The same cannot be said for those seeking to extend current bans to cover vaping. They’ll be the first to tell you that more study of e-cigarettes is needed. But why wait for results? They’re ready to ban first and ask questions later.

Read the whole thing here.

How grilling meat really is like a smoking a cigar

In his “Explainer” column at Slate yesterday, Brian Palmer raised alarm about grilled and smoked meats, suggesting that by eating them we may be, as the headline puts it, “Cooking Up Cancer”:

A growing body of research suggests that cooking meats over a flame is linked to cancer. Combusting wood, gas, or charcoal emits chemicals known as polycyclic aromatic hydrocarbons. Exposure to these so-called PAHs is known to cause skin, liver, stomach, and several other types of cancer in lab animals. Epidemiological studies link occupational exposure to PAHs to cancer in humans. When PAHs from a flame mingle with nitrogen, say from a slab of meat, they can form nitrated PAHs, or NPAHs. NPAHs are even more carcinogenic than PAHs in laboratory experiments. The reasonable conclusion is that grilling meat may be hazardous to your health.

The evidence linking cancer to cooking meat over a combustion source has been accumulating for decades. Epidemiologists first noticed a connection between the consumption of smoked foods and stomach cancer in the 1960s. Japan, Russia, and Eastern Europe, where smoking is a popular way to preserve meat and fish, became laboratories for gastric cancer research. Newer studies suggest that eating smoked meats may lead to cancer even outside the gastrointestinal tract. A 2012 study, for example, linked smoked meat consumption with breast cancer.

Palmer then compares current attitudes toward grilled meats to outdated acceptance of tobacco use:

In the mid- to late-19th century, doctors determined that lip and tongue cancer rates were higher among smokers of pipes and cigars. Despite this link, major medical journals mocked those who opposed smoking. The Lancet, the leading journal of the time and still one of the most important medical journals in the world, wrote in 1879, “We have no sympathy with prejudices against … tobacco, used under proper restriction as to the time and amount of the consumption. … A cigar when the mood and the circumstances are propitious [is] not only to be tolerated, but approved.” Moderation, not abstinence, was the order of the day.


It wasn’t until 1964 that the Report of the Surgeon General finally and firmly declared that smoking was indisputably linked to the surge in lung cancer. By that time, epidemiologists had a complete picture of the long-term effects of the increase in cigarette smoking that began around the time of World War I. The conclusions practically drew themselves. Still, it took the public health community decades to agree that smoking in moderation is a terrible idea.

I learned of the article from my friend Jeff Woodhead on Twitter, who took Palmer to task for sensationalizing the dangers of grilling and noted that habitual cigarette use carries far greater risks than exposure to charred meat. I don’t disagree. However I want to defend Palmer on one point. He is actually correct to compare grilling meats to smoking tobacco, though not in the way he realizes.

That pack-a-day consumption of cigarettes greatly elevates one’s risk of lung cancer is no longer disputed by any sensible person. Moderate use of other forms of tobacco, in contrast, carries much less danger. The Lancet article that Palmer mocks for suggesting that moderation in tobacco use is nothing to worry about overstates the case but was not too far from the truth.

Just how dangerous is it to enjoy an occasional cigar “when the mood and the circumstances are propitious?” A study in Preventive Medicine compared rates of lung cancer among smokers of various kinds of tobacco to those of non-smokers. Cigarette smokers were 16 times more likely than never smokers to get lung cancer. Smokers of cigars only, pipes only, and cigars and pipes all fared much better. Further, lung cancer among pipe and cigar smokers was concentrated among those who were the heaviest consumers. “Among pipe and/or cigar smokers only, patients with lung cancer were more likely than controls to have been long-time smokers of 5 or more cigars or 5 or more pipefuls per day and to have inhaled. The odds ratio for those smoking 5 to 9 cigars or pipes per day was 3.2 and for those smoking 10 or more units 6.7. The odds ratio of those cigar or pipe smokers who inhaled was 12.3.”

A cohort study published in The New England Journal of Medicine followed about 17,000 men enrolled in Kaiser Permanente health plans who reported never smoking pipes or cigarettes. Over a course of more than 20 years, the study compared health outcomes for non-smokers and smokers of cigars. Cigar smokers carried a relative risk of cancer in the aerodigestive tract of 2.02 and in the lungs of 2.14.

A third study in BMJ examined the risks of dying from three smoking-related diseases among former cigarette smokers who had switched to smoking pipes or cigars. Their mortality risk relative to users of pipes and cigars who were never cigarette smokers was 1.51. To put that in perspective, their relative risk compared to people who had never smoked at all was 1.68. In other words, the study found that even former cigarette smokers who switch to pipes and cigars lowered their mortality risk to a level not much above that of never smokers.

Brad Rodu, an advocate of harm reduction approaches in tobacco control, summarizes many of these risks. What it basically comes do is this: Cigarettes are uniquely dangerous because they are inhaled directly into the lungs and are very addictive. Other forms of tobacco that are mainly enjoyed in the mouth and lend themselves less easily to habitual consumption are significantly less dangerous. The risks are real, but much, much lower than those associated with cigarettes.

How does all of this compare to eating meat? Palmer doesn’t cite many sources that specify the dangers, but one study he links to does associate consumption of smoked meat with breast cancer. It finds an adjusted odds ratio of 2.31-3.13. If that’s accurate (and it may be on the high side), then eating lots of grilled meat may actually be a lot like enjoying an occasional cigar — which is to say that it’s a reasonable choice many adults may decide to make.

(Note: Some of the studies cited in this post use relative risk while others use odds ratios. These are not identical measures but should be roughly comparable.)

It’s also interesting to ask how this compares to the dangers of exposure to secondhand tobacco smoke. Smoking bans have proliferated on the justification that workers should not be put at risk. Whatever one’s opinion of these policies, consider the findings from the landmark 2006 report from the Surgeon General. The report concludes without equivocation that environmental tobacco smoke increases the risk of lung cancer in non-smokers. By how much? The reports’ table of meta-analyses of studies estimates relative risks of exposure at home for non-smoking spouses or at work for non-smoking employees in a range of 1.12 at the low end to 1.43 on the high end. Those are low relative risks! If they are enough to justify comprehensive bans on indoor smoking — not to mention the outdooor bans that are now so popular — then one may well question the ethics of allowing restaurant cooks to expose themselves to grill smoke on our behalf.

One of the big problems with reporting on cancer is that it focuses on the wrong question. Journalists ask, “Does X increase the risk of cancer?” The answer is very often yes, but they don’t follow-up with, “By how much?” Lifestyle choices carry trade-offs and better reporting would help readers put them into perspective. Palmer, to his credit, does write that the “risk-reward equation for smoking differs from that of grilling or frying meat” and acknowledges that the epidemiology on the latter is not yet clearly established. His mistake is to carelessly lump all forms of tobacco use into one, ignoring the fact that different types and different use patterns carry substantially different levels of risk.

Of course, few yet advocate completely giving up grilled or smoked meat, much less passing legislation to restrict it. “Everything in moderation,” said one of the toxicologists quoted in the Slate article. One assumes that the toxicologist’s definition of “everything” doesn’t include tobacco, because who would say that these days? But Palmer’s comparison of grilling to smoking isn’t crazy. It just tells us a lot more about contemporary bias against tobacco use than it does about the dangers of meat.

DC wants to ban e-cigs in bars

People in DC bars are “vaping,” or using electronic cigarettes, indoors. A couple members of the DC city council — Yvette M. Alexander and David Grosso — have introduced a bill to include e-cigarettes in the city’s smoking ban:

In an interview, Alexander said e-cigarettes are being “used to usurp the smoking ban.”

“It is smoking, is an inhalant and it’s similar to smoking,” said Alexander, chairwoman of the Health Committee. “We don’t know what the ill effects of this are, and it’s still a bother to some people.”

“Similar to smoking” and “a bother.” Time was city officials at least made a show of finding evidence of harm before imposing bans. E-cigarettes may be annoying to other patrons, but there’s no evidence or reason to believe that secondhand vapor (is that a thing now?) is something to fear. And to point out the obvious, bar and restaurant owners are perfectly free to set their own policies if guests prefer to avoid it.

California brings smoking ban inside the home

It was only a matter of time:

Millions of Californians would not be able to smoke tobacco inside their own homes under new legislation that would raise the bar nationwide for fighting secondhand smoke.

No state ever has ventured into personal bedrooms and living rooms with its smoking restrictions, but California is going even further than that by targeting owner-occupied residences as well as rental units.

Specifically, the measure would prohibit lighting up a cigarette, cigar or pipe in condominiums, duplexes and apartment units. […]

Levine’s bill would permit outdoor smoking near apartments or condos, but only in a clearly marked area that is at least 20 feet from any housing unit and 100 feet from a playground, school or pool.

Who could have predicted that ignoring property rights and letting bad science go unchallenged would lead to this?

Reform Oregon’s smoking ban

An Oregonian editorial last week was refreshingly libertarian, calling for same-sex marriage, tuition equity for some undocumented immigrants, restraint on gun control, and even opposition to the state’s smoking ban. I sent in a letter about the last item, which they published today:

The Jan. 12 libertarian-leaning editorial “Protect and expand personal freedom: Agenda 2013″ was a breath of fresh air, especially in regard to our state’s excessively stringent smoking ban.

Current law makes few exceptions for businesses that cater to smokers, making it essentially illegal for entrepreneurs to open new cigar bars or smoking lounges even in stand-alone tobacco shops. Regardless of whether one supports a broad smoking ban, it’s difficult to justify forbidding these businesses to open.

Sensible reform would replace the current exemptions, which apply only to venues that have been grandfathered in, with objective guidelines that would allow both existing and aspiring business owners to offer smokers an indoor refuge.

As I reported in the Oregonian in 2011, the promised decline in heart attacks that the smoking ban was supposed to usher in never developed.

Save the stogies

My forthcoming article that I’ve alluded to a couple times this week is now up at The Atlantic:

If a time traveler from the early 1990s were to arrive in the U.S. bars and restaurants of today, what would notice first? Perhaps that the food has become more interesting and varied, or that a perplexing number of diners are photographing it with their remarkable phones. The most obvious change, however, might register on the nose: the nearly complete absence of indoor smoking.

California implemented the United States’ first modern statewide smoking ban in 1998. Today twenty-nine states and 703 municipalities require bars and restaurants to be smoke-free, according to data maintained by the Americans for Nonsmokers’ Rights Foundation (North Dakota brought the tally to thirty states this month). Tobacco use has been banished from our culinary radar along with the question “smoking or non?” Most of us don’t miss it. Yet as a slew of new bans, taxes, and regulations drive smoking to the peripheries of society, it’s worth giving tobacco another look.

Read the whole thing. And for more context on some of the arguments, see my recent posts about the effects of new tobacco taxes and the failure of the FDA to establish an effective regulatory regime.

Quote of the day

Julien Guttman on the forthcoming outdoor smoking ban at George Washington University:

“We’re trying not to use the word ban,” says Julien Guttman, of the GW campus advocacy group Colonials for Clean Air. “We encourage people to talk about a smoke-free campus rather than a ban on smoking.”

This is the same Julien Guttman who lamented a couple weeks ago, “No matter how much science we have to back up what we are saying, there will always be individuals who see this as a restriction on their freedom.” Perhaps it’s a sign of progress that she feels the need to resort to such euphemism when defending the policy.

A bit more on the shaky science behind outdoor smoking bans here.

Where to smoke in Portland

horsebrass 021

Today at Drink Portland, I’ve posted a guide to the city’s few remaining cigar bars. There are only eight of them, so click through if you’re looking for a place to escape the rain with a cigar and a drink. Since Oregon’s smoking ban only exempts bars that can demonstrate cigar sales from 2006, it’s essentially illegal to open a new cigar bar. These eight are all we have and all there will be. An update to the law a couple years later also capped the number of tobacco shops that can allow smoking.

Ban advocates predicted that Oregon’s law would drastically reduce the rate of heart attacks in the state. As I noted in the Oregonian, that never came to pass.

Pre-ban, my favorite place to have a cigar in Oregon was the Horse Brass Pub. My ode the Horse Brass is here. It’s still a great bar, but you can’t light up there anymore.

“No safe level,” continued…

The idea that there is “no safe level of exposure to secondhand smoke” has to be one of the most misleading statements in public health. I mentioned it in a story about campus smoking bans yesterday. It pops up again today in a blog post about airport smoking areas at The New York Times by Nicholas Bakalar:

Five large-hub airports in the United States have designated indoor smoking areas. According to a study from the Centers for Disease Control and Prevention, they all have unhealthy air — even in places where no one smokes. […]

The researchers found that the pollution level in smoking areas of the five airports was 23 times as high as the level in nonsmoking airports, and the average in adjacent areas was five times as high.

The study, in the Nov. 20 issue of the Morbidity and Mortality Weekly Report, notes that no level of secondhand smoke exposure is safe, and even brief exposures can have adverse cardiovascular and respiratory effects.

Notice how much work the “no safe level” line is doing here. It allows the researchers to abdicate responsibility for showing that levels of particulate matter in the air surrounding smoking lounges are causing any actual harm. They have no need to relate this minuscule level of exposure to a level of risk — because, in fact, doing so is likely beyond epidemiology when the risks are this small, if they exist at all. As Jacob Sullum noted in 2006 when Richard Carmona’s Surgeon General’s report came out, the science in the report is much more modest than the “no safe level” hype that accompanied it:

Since it is difficult even to measure the health consequences of long-term, relatively intense exposure to secondhand smoke among people living with smokers for decades, how could one possibly demonstrate an effect from, say, a few molecules? It’s clear that the vast majority of people exposed to secondhand smoke suffer no noticeable injury, so in what sense is their exposure unsafe? “No safe level” is an article of faith, not a scientific statement.

And yet, unfortunately, the “no safe level” idea continues to be used as an easy shortcut by researchers — and then dutifully transcribed by health reporters.

Banned on campus

One of the quiet ways smoking bans have spread across the United States is via bans on college campuses:

Colleges and universities have become the latest target of anti-smoking groups. While schools have long banned smoking indoors, the new bans are addressing outdoor space.

So far, more than 800 schools have banned smoking on campus, according to the American Nonsmokers’ Rights Foundation. The list is dominated by medical schools and non-residential community colleges, although more and more residential colleges and universities are joining.

Encouraging schools to go smokefree is now official policy of the Department of Health and Human Services, which recently launched an initiative in collaboration with the University of Michigan to encourage campus bans on campus. The bans apply outdoors, which is a major inconvenience to smokers on sprawling campuses, potentially driving students to less safe perimeters. The measures are supposedly important for creating a healthy environment. Unfortunately, they seem to also create an environment of self-righteous attitudes and bad science:

“The CDC and surgeon general say there is no safe level of exposure to second-hand smoke,” said Julien Guttman, a GWU public health graduate student who is part of the advocacy group Colonials for Clean Air. “No matter how much science we have to back up what we are saying, there will always be individuals who see this as a restriction on their freedom.”

Those crazy individuals! The idea that there is “no safe level of exposure to second-hand smoke” is itself unscientific, a line of propaganda devoid of any measurement or relation of exposure to risk. If taken literally it would lead to absurdities like… well, like banning smoking in wide-open expanses of outdoor space.

Too hot for North Dakota

North Dakota, rugged North Dakota. Now home to one of the most restrictive smoking bans in the country:

Measure 4 bans smoking in places where North Dakota state law previously allowed for it, including bars, tobacco shops, taxis, motel rooms and private nursing home rooms.

The ban also requires smokers to be at least 20 feet away from buildings. Tobacco-less e-cigarettes are also included in the ban.

Violators could be fined $50, and bars that still allow smoking could be subject to having their liquor license revoked.

North Dakota will be the 30th state to implement a smoking ban in bars and restaurants, but few are this restrictive. The impact on cigar bars and tobacco shops could be devastating, as Stogie Guys reports.

This is speculative, but I imagine that the average North Dakota resident is not so intolerant as to object to smoking in these specialized businesses. However the anti-smoking groups who pushed this initiative didn’t give them the option to allow exceptions, forcing voters to choose between all or nothing.

Cigars for me but not for thee

The city of St. Louis implemented a smoking ban on January 1, 2011. There’s one place, however, where people still smoke with impunity:

The 109-year-old downtown Missouri Athletic Club may wriggle free from the city’s smoking ban.
City officials have prepared an agreement which exempts the private, invitation-only establishment — long frequented by judges, attorneys and politicians — from the municipal no-smoking ordinance.

The club, known as the MAC, has flouted the law since it was enacted Jan. 1, 2011, openly leaving ashtrays in the lounge, hosting hazy boxing matches and allowing men in suits to gather weekly at the bar with tumblers in one hand, cigars in the other.

The city cited and fined the club twice. The citations ended up in municipal court, where attorneys began working out a deal.

On Thursday, city Health Director Pam Walker presented a draft agreement to her advisory commission, the Joint Boards of Health and Hospitals, arguing that the nonprofit MAC is a unique entity, governed neither by rules for private clubs nor by those for businesses.

Hat tip to Michael Siegel, who inducts St. Louis health director Pam Walker into his Colonel Benjamin Church Hypocrisy Hall of Shame for carving out this exemption for local elites.

St. Louis isn’t the first city to engage in this kind of smoking ban favoritism. In Washington, DC, city councilman Jack Evans voted in favor of the District’s smoking ban, then took advantage of his position on the council to create special exemptions for organizations he likes:

The city’s smoke-free law provides an economic hardship waiver for struggling bars and restaurants, Evans said, but it leaves no wiggle room for a single event, like the St. Patrick’s Day gala or Fight Night at the Washington Hilton.

“Once a year, 1,000-plus people go there to drink Irish whiskey, smoke cigars and have dinner,” Evans said of the dinner. “Now they’re not allowed to do that. From my reading of the law there’s no other way to get an exemption but to legislate.”

Evans has continued to seek these one-time exemptions while leaving less connected charities who’d like to host cigar events out of luck.

As I wrote in 2009:

Evans has discovered the pain of having one’s treasured tradition banned by a bunch of meddling bureaucrats. I’d be sympathetic if not for the fact that Evans is one of those meddling bureaucrats. If he doesn’t like the law, he should introduce changes that open up smoking venues to everyone, not just to clubs that happen to have a city councilman in their membership.

Smoking: More than just a vice

horsebrass 021

Something Not Unlike Research” is a health care blog I recently came across written by two university professors. One of them, Bill Gardner, wrote this week in defense of employers choosing not to hire smokers. He concluding by noting that “Smoking is a vice that benefits no one.” I took him to task for this last line on Twitter:

“Smoking is a vice that benefits no one.” — @Bill_Gardner Oh please. I like it. It benefits me! Arrogant assumption.

To my surprise, he responded in a new post:

My point is that I do not see a compelling argument against employers choosing not to hire them. By extension, we should have no public policies protecting access to employment for smokers (or bone lugers). This is, I believe, consistent with Mill’s view that

The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not sufficient warrant.

I do want to see smoking go away, and I think social pressure is what makes that happen. (“Social pressure” is not, I think what Mill means by “exercising power”.) To that end, I think it is legitimate to ban smoking in public places, and for employers to refuse to hire smokers if they judge that to be in the interest of their firms.

Having conceded that smoking may benefit Jacob’s subjective well-being, can I still say that “smoking is a vice that benefits no one”? I certainly can, if it’s understood that benefit refers to the long-term well-being of smokers, and those who depend upon them, rather than immediate subjective well-being. In that sense of benefit, there is nothing to be said for smoking or binge drinking.

(NB: The binge drinking reference is about the bone luge, which happened to be the top post on my site when he visited. The photo definitely makes his post more awesome but I’ll stick to the smoking discussion!)

I’ll elaborate in more than the 140 characters allowed by Twitter. To start, I agree with Bill that smokers shouldn’t be a protected class of worker. I think refusing to hire smokers is generally a silly policy but that’s something the market can sort out. Where we disagree is in our assessment of smoking. He says that smoking is “a vice that benefits no one” in the long term, that “there is nothing to be said for smoking,” and that “I do want to see smoking go away.”

I know where he’s coming from because I used to make the same assumptions. I grew up believing everything bad about smoking and made it through high school and college without taking a single puff from a cigarette. My interest in tobacco began shortly after I started working as a barista. In conversation with a cigar-smoking friend, I realized that he talked about his stogies the same way I talked about coffee. The varietal of the plant and the origin of the leaf mattered, a Cameroon wrapper tasting differently than one from Nicaragua. Flavors ranged from light sun-grown tobacco to deep, dark maduro, much as coffee roasts came on a spectrum from light to dark. And he suggested that there was just as much difference between the hand-rolled cigars he was smoking and the mass produced cigarettes of Big Tobacco as there was between Folgers and the small batch beans I brewed.

Perhaps, I thought, there was more to tobacco than I realized. When I eventually tried a cigar for the first time I took to it at once. Doing that required getting over my own prejudices that had led to me viewing smoking as pure vice, an unhealthy and addictive habit without redeeming qualities.

I am not addicted to tobacco. I go weeks and months without it. Of the three mood-enhancing drugs I enjoy — caffeine, alcohol, and nicotine — the last would be by far the easiest to never again partake of. It’s unlikely that my occasional indulgence in a cigar will have any significant effects on my health, though there is a small chance that it will. In exchange cigars have given me wonderful experiences, both in taste and in the friendships that have deepened over contemplative, smoke-fueled conversations. I expect that cigars are a net benefit in my life on both the short and the long view. Perhaps additional evidence could persuade me otherwise, but it can’t simply be assumed that I’m mistaken.

Admittedly I don’t fit the median profile of a smoker. However the value that smokers get out of smoking is almost never acknowledged by anti-smoking activists, who treat smoking as inherently wrong. Unlike them, and perhaps unlike Bill, I don’t want to see a completely smoke-free America. I want to see smoking substantially reduced. I’d like for people to smoke less and smoke better, using products other than cigarettes, which seem to be the most dangerous form of tobacco. (It’s also worth noting that the employment policies that inspired these posts are enforced via urine tests for nicotine, which wouldn’t discriminate between pack-a-day smokers and those who smoke rarely or those who are trying to quit with the aid of nicotine patches or gums.)

Semantics aside, activists’ unwillingness to consider the benefits of smoking leads to excessively restrictive policies. Let’s take smoking bans for example. Consider two businesses:

Business 1 is a tobacco shop with an attached lounge that offers beer and wine. Customers are allowed to smoke there. It’s a freestanding building with no immediate neighbors, so no one except customers and employees is affected by the smoking. Four people are employed serving drinks in the lounge. A smoking ban passes that forces the business to eliminate drink service. The day the ban takes effect those four employees lose their jobs.

Business 2 is a restaurant that serves Dungeness crab caught in the Pacific Northwest. Commercial fishing has one of the highest fatality rates of any occupation and crabbing in this region is often the highest of all. For comparison, the average annual fatality rate for all occupations is 4 per 100,000 workers. For fishing as a whole the rate is 115/100,000. For Dungeness crab fishermen in the Pacific Northwest the rate is 463/100,000. (Source here.) There are no proposals to forbid restaurants from serving Dungeness crab.

The comparison might seem silly, but why? Dungeness crab is delicious but it’s hardly a staple in the food supply. Fishermen are literally dying to put it on our plates. Though the level of risk associated with secondhand smoke exposure is in dispute, it would be astonishing if the danger of pouring beer in a smoky room was at all comparable to crabbing on a stormy ocean. So again, why the disparity in how we treat these workers?

The number of actual deaths resulting from Dungeness crabbing is low since it’s a small industry, but it provides a model for how we normally regulate occupational risk. We don’t ban dangerous jobs, we try to discover reasonable rules to make them safer. Safety regulations have apparently been successful in reducing the fatality rate among Alaskan Dungeness fishermen.

Let’s accept for the sake of argument that banning smoking in some places to protect patrons and employees is justified under Mill’s proviso. It’s much harder to argue that smoke-friendly businesses should be banned entirely. Our normal approach to worker safety would allow people to work in smoking venues, subject to reasonable rules about ventilation to minimize risk.

Yet exceptions to smoking bans are often unreasonably narrow, preventing consenting adults from making free exchanges with each other. This is because policy makers view smoking as inherently without value. The thought process goes something like this:

1) Smoking has no value.

2) Protecting workers has value.

3) Therefore it’s OK to ban smoking everywhere without worrying about smokers’ preferences.

This is why I vehemently object to statements such as Bill’s. They create an environment in which the rights of smokers, business owners, and workers are too easily violated. I’ve seen too many of my favorite places completely altered by smoking bans, to the dismay of owners, patrons, and employees. See my ode to one of them, the Horse Brass, in DoubleThink magazine. Business 1 above is another example; it’s an actual establishment outside of Portland.

In sketch form, here are two other arguments for how ignoring the benefits of smoking skews our thinking and leads to bad policy.

Electronic cigarettes — The view of many in tobacco control is that tobacco is inherently bad and that quitting should be smokers’ only goal. It would undeniably be a good thing if more smokers quit, but the obvious truth is that quitting is difficult and relapse is frequent. A new study concludes that even with increased use of nicotine replacement therapy and quitlines, the rate of successful cessation has remained unchanged.

E-cigarettes are a promising alternative for smokers who find it difficult to quit. They are undoubtedly safer than actual cigarettes and offer a substantial opportunity for harm reduction, whether used to quit entirely or just to reduce use of real tobacco. Their popularity is likely due in part to their similarity to cigarettes: their appearance, the ritual of lighting them up, the ability to manage nicotine levels through their use. They deliver many of the same benefits smokers’ get from cigarettes at a fraction of the health cost. Yet precisely for this reason they are treated with outright hostility by many in tobacco control, who blind themselves to e-cigarettes’ potential because of their puritanical view of smoking — including acts that mimic smoking — as inherently bad.

Flavored cigarettes — The legislation giving the FDA regulatory authority over tobacco included a ban on all flavorings except for menthol. The reprieve for menthol may only be temporary. Flavors are treated as dangerous additives but they’re essentially being banned because they make smoking more appealing and enjoyable. This is pure paternalism. If a consenting adult wants to have a clove cigarette, he should be allowed to buy a clove cigarette.

The preferences of smokers were never given a voice in this debate. Because smoking was simply assumed to have no value, smokers were dismissed as addicts who could only benefit from this legislation. Perversely, those of us who support consumer choice on menthol are accused of racism since menthol cigarettes are popular among African-Americans. It’s unclear at this point how FDA regulation may affect more artisanal forms of tobacco, but the industry is understandably worried: If the agency views tobacco as inherently bad, it’s unlikely to pass regulations in the interest of pipe and cigar consumers.

The costs of smoking are undeniably high, yet smoking has persisted even in the face of penalties as harsh as death. The habit is here to stay. Unfortunately, smoking has become synonymous with Big Tobacco and the lowly, lethal cigarette. There is much more to it than that, and more to say on its half behalf. But that will have to wait for a future post.