The usually sensible James Fallows joins the debate about obesity in America with some arguments of, ahem, dubious logical value:

If you’ve been around the US as long as I have (ie, if you’re as old), you have seen very significant aspects of public-health behavior change in your own lifetime. When my dad went to medical conventions in the 1950s and 1960s, most of his fellow doctors smoked. By the time he retired in the 1990s, very few of them did. For better and worse, smoking has become a class-bound phenomenon in America: better for the people who don’t smoke any more, worse as one more disadvantage of being poorer and less educated. The difference is startling and obvious if you spend time in, let’s say, China, where many more people of all classes smoke. As individuals, Americans have the same human nature as they did 40 years ago, and the same nature as people in China. Will power, compulsions, addition-seeking instincts, etc. But their overall behavior about smoking has changed. Some individuals did not or could not change their behavior. (One of my grandmothers, who had started smoking as a flapper in the 1920s, died of a horrible case of emphysema, sneaking cigarettes on her last conscious days.) But average behavior changed dramatically. In my view, no sane person can deny that public anti-smoking campaigns have made a huge difference.

Well, certainly over the last 50 years, the proportion of Americans (as in other western nations) who smoke has declined significantly.   Some of this decline may even have been due to smokers quitting, and some of those quitters may have done so in response to government anti-smoking campaigns.  But the overwhelming reason for the decline in proportions is the death of the smokers and the failure of the newly-born to take-up smoking.   “Will power, compulsions, addition-seeking instincts, etc” (does he mean “attention-seeking instincts”?) has little to do with this. Average behaviour certainly changed, but so did the people who comprise the averaged population.  It is NOT the case that there are fewer smokers now than in 1959 in the USA simply because lots of smokers just exercised some will-power.
A second logical flaw Fallows makes is in the very analogy between smoking and obesity.  Unlike the use of nicotine, eating is necessary to live.   Unlike smoking, it is not possible to give up eating by complete withdrawal (ie, going cold-turkey).  It is even not possible to give up eating less quickly than complete withdrawal without usually suffering serious adverse health effects.   Moreover, it is not even certain that eating less food will lead a person to lose weight.   The human body is a complex adaptive system with many non-linear relationships between its components.   Decreasing food intake, for example, may lead the body to storing proportionately MORE of the nutrients in the consumed food than previously, so that total body weight does not necessarily decline in proportion to the fall in consumption.  Whatever it is, the relationship between food consumption and human body weight is certainly not linear, as the scientific evidence makes clear.
The main problem here for public health policy is that we do not yet know the ultimate causes of the obesity epidemic now seen in parts of the West.   It surely is not lack of exercise (since what little scientific evidence there is says that regular exercise seems to INCREASE weight, by stimulating appetite and adding to muscle mass); and the cause is surely not over-eating, since the epidemic has arisen faster than the major changes in people’s eating habits (20 years versus 60 years).  The ultimate cause could be a virus; it could be a consequence of particular food-additives (eg, the increased use of sucrose, or trans-fats, etc) or some particular adverse combination of these additives; or it could be a consequence of the particular combination of proteins, carbohydrates and nutrients in our diets; or indeed any number of other causes – medical, nutritional, lifestyle, and/or sociological.
The medical profession has such a shameful historical record of wrongly blaming the victim of an illness for the sickness before discovering the real cause (eg, cholera, stomach ulcers, RSI, CFS, ADD) that it ill-behooves anyone, in the current state of medical ignorance, to lecture people that the cure for their obesity is just to eat less.