"Today it's cigarettes," said R.J. Reynolds in a June 1994 newspaper ad that criticized proposals to sharply raise tobacco taxes. "Will high-fat foods be next?"

Anti-smoking activists have traditionally responded to this sort of slippery-slope argument by insisting that cigarettes are unique, "the only legal product that when used as intended causes death." To suggest that anti-smoking measures might pave the way for attacks on cheeseburgers and ice cream, they said, was just plain silly.

Yet six months after R.J. Reynolds tried to scare people with the outlandish prospect of a tax on fatty foods, a leading obesity expert endorsed the idea on the op-ed page of The New York Times, citing the precedent set by cigarette taxes. Now U.S. News & World Report, in its current issue, is presenting a "Twinkie tax" as one of "16 Silver Bullets: Smart Ideas to Fix the World."

It's remarkable how quickly the notion of better eating through taxation has been transformed from reductio ad absurdum to serious policy proposal. A 1993 article in The Harford Courant, headlined "Fat Tax? Fat Chance!," concluded with a quote from an economist who said such a levy "wouldn't pass the laugh test." A 1994 story in The San Diego Union-Tribune raised the possibility of "heavy taxes...affixed to all junk food" but quickly reassured readers, "No one is proposing anything so drastic yet."

Into the void stepped Kelly Brownell, a professor of psychology who directs the Yale Center for Eating and Weight Disorders. In his December 1994 New York Times op-ed piece, he likened people to laboratory rats who, when given "convenience store delights" such as cheese curls, chocolate bars, marshmallows, and cookies, "will ignore available nutritious food, even as their body weight doubles and triples."

Based on this phenomenon, Brownell concluded that "environment is the real cause of obesity." To "shift the focus to the environment," he recommended censoring McDonald's commercials and "taxing foods with little nutritional value." He has since repeated these proposals in interviews and articles.

The recent U.S. News piece elaborates on Brownell's scheme: "Such a measure would involve devising a calorie-to-nutrient index, with low-calorie, vitamin-stocked fruits and vegetables on one end of the spectrum and fat-drenched, low-nutrient fast food at the other. The scale would not be based on a simple measure of fat...but on a more complex analysis of nutrients per calories that took into account such things as the nutritional value of saturated versus unsaturated fats."

The author, Shaheena Ahmad, admits that trying to micromanage the diets of 260 million Americans "would be unabashed social engineering. But so is virtually anything the government does about public-health dangers, such as air pollution or drunk drivers, that pose smaller threats to most people's life expectancy."

Ahmad thus seems oblivious to the distinction between endangering oneself (by eating too much junk food) and endangering others (by emitting pollutants or driving while intoxicated)--a distinction that is fundamental to the classical liberal tradition of limited government. She is not alone in her confusion. Former Surgeon General C. Everett Koop compares efforts to discourage smoking and other risky habits to mandatory vaccination of schoolchildren and laws against assault. He says "the government has a perfect right to influence personal behavior to the best of its ability if it is for the welfare of the individual and the community as a whole."

Koop is merely articulating contemporary public health doctrine, which says the state has an obligation to discourage behavior that might lead to disease or injury, even when the risk is voluntarily assumed. That principle has been cited as a license for all sorts of meddling, from alcohol and tobacco taxes to mandatory seat belt and helmet laws. So why not a levy on junk food? There is no end to the interference that could be justified in the name of "public health," as that concept is currently understood.

A few years ago, I wrote a short item for National Review facetiously suggesting that a tax based on the number of pounds by which people exceed their recommended weight would be more fair and efficient than a junk food tax. Given how readily yesterday's joke can become today's "silver bullet," I am sorry I did not add a disclaimer. "Only kidding!" I should have said. "Please do not try this in Washington."