Why a smoking ban is bad policy

The Hennepin County Commissioners will hold yet another public forum on the smoking ban on November 15th. Participants -- pro and con -- will be limited to 2-3 minutes. If you're paying attention that says the commissioners are perhaps more interested in a timely meeting and giving the appearance of input rather than looking for any real information.

Nonetheless, one can say a lot in two to three minutes. For example, looking at "big government" data, one might make a pretty strong case that even accepting the health claims of pro-ban supporters, a smoking ban is still bad public policy. Using this chart, one might make the case outlined below.

California EPA Study: "Proposed Identification of Environmental Tobacco Smoke as a Toxic Air Contaminant"

From Table 7.2A "Risk of Lung Cancer with ETS Exposure"

Exposure..........Duration...............Cases/Controls...............OR (95% CI)

Work.................Ever......................729/1560.....................1.13 (0.97-1.31)

......................<8years................198/472.......................0.94>21 years.............262/543.......................1.25 (1.03-1.51)

............................................................................................p = 0.01


The debate you’re hearing today is between private economic loss and public health. As policy makers, your job is not to come down on one side or the other. Your job is to resolve the conflict with no more economic and social disruption than is absolutely necessary. That is your challenge. That is your obligation.

To resolve the conflict, you need data. You need context. This table provides both.

This table is from the California EPA study entitled “"Proposed Identification of Environmental Tobacco Smoke as a Toxic Air Contaminant." The data is from “Big Government,” not “Big Tobacco.” It suggests a correlation between exposure over time to secondhand smoke and lung cancer. Without examining the study methodology, assume this data is valid. It is data smoking ban proponents ought to accept without question.

Note the bottom cell in the far right column of the table (p = 0.01). That cell tells a statistician that there is little probability that the trend in the right-hand column could have occurred by chance. In other words, it is statistically probable that exposure to secondhand smoke over time increases the chances that one will contract lung cancer.

However, as policy makers, your concern is with the relationship between the “Duration” column and the “Odds Ratio” column. That relationship guides how you make sound policy that balances immediate harm against uncertain future risk.

Just a little math -- The number in the Odds Ratio column is the median of the confidence interval -- or range -- found in the parenthesis. An odds ratio of 1.0 or less tells a statistician that a casual connection cannot be made between two factors -- in this case between secondhand smoke and lung cancer.

As you can see, it is not until the fourth cell down in the Odds Ratio column [1.25 (1.03-1.51)] that the lower value of the confidence interval rises above 1.0. In other words, it is only in that cell that the risk of lung cancer is statistically significant. When you look at that cell in the context of the “Duration” column, the data tells you one must work in a smoking environment 21 years or more before secondhand smoke exposure has a statistically significant negative health impact. Working in a smoking environment less than eight years, one has no greater risk of contracting lung cancer than a person working in a smoke-free environment (OR = 0.94).

From your policy perspective, this data does not support a behavioral ban in bars and restaurants. A smoking ban unnecessarily creates the economic hardship and social disruption you see here today. It unnecessarily infringes on private property rights and individual choice. And why? To protect career hospitality employees from a risk that requires decades of exposure to become statistically significant and then at a median level risk only 25 percent greater than for the general population. Can it not be assumed that a person working in the hospitality industry for 21 years has made a conscious choice to accept that risk? A smoking ban imposed on bars and restaurants is simply unnecessary and bad public policy.

You are challenged to do better. You are obligated to do so. That data is there.

A Possible Question --

Question: The California EPA study is based on a “normal working environment.” The intensity of exposure to secondhand smoke in a smoky bar is more of a health risk than long-term exposure in a “normal working environment.”

Response: The California EPA study specifically states “It is not known how the intensity of exposure may have affected risk estimates.” On the other hand, in a pre-ban study completed in St. Louis Park, bars with ventilation equipment installed demonstrated air quality 100 - 500 times below the OSHA standard for nicotine particles -- the only airborne particulate exclusive to secondhand smoke. An air quality standard commensurate with the OSHA standard, rather than a behavioral ban on smoking in bars and restaurants, essentially nullifies the “intensity” argument.

Incidently, asking the question nulifies the ban supporter claims that the is no safe level of secondhand smoke exposure and there is not a threshold level below which exposure to secondhand smoke is not a helath risk. If those two propositions are correct, then the intensidty level of secondhand smoke ought to be irrelevant. If intensity is relevent, then there is an exposure level to secondhand smoke that is not toxic.

Category: Public Health, Smoking Ban