New York Times columnist Daniel Akst reported Sunday on a National Bureau of Economic Research (NBER) “working paper” blaming fast food restaurants for the perceived “epidemic” of obesity in America. A trio of NBER researchers took statistics on wages, hours worked, self-reported weight, and restaurant density, put them in a blender, and concluded that inexpensive, convenient dinners are responsible for 68% of our national weight gain.

But last fall, the New York Times told a very different story — also relying on an NBER paper. In a September 26 news story, the Times concluded that “most of the [weight] increase comes from too little exercise, not too many calories.” Indeed, NBER then claimed that 60% of our added pounds could be traced to sedentary work habits and general lack of exercise. Of the remaining 40%, only a portion could be pinned on fast food.

So which flavor of NBER research is correct? The Times apparently doesn’t know, and neither do we. But when the writer touting the 68-percent figure (Mr. Akst) also falls prey to the well-debunked myth stating that 300,000 Americans die each year due to obesity, we have our hunches about whom to believe.

In order to conclude that obesity kills 300,000 Americans every year, you have to assume that every dearly-departed obese American died as a direct result of being overweight. This logic comes apart, of course, when you consider car crash victims, suicides, and accidental deaths. For this reason, among others, the respected New England Journal of Medicine has written that “the data linking overweight and death… are limited, fragmented, and often ambiguous,” and that the evidence for an obesity-related death toll in the 300,000 range “is by no means well established” and “derived from weak or incomplete data.”