Bad Food Data and Science Still Make for Bad Food Policy
USDA's diet guidelines are a mess because the information it uses is suspect.
A new study in the journal Current Problems in Cardiology by researcher Edward Archer, Ph.D. and several colleagues has pointed to serious flaws in the data the U.S. Department of Agriculture (USDA) relies on to calculate the average number of calories that are present in the American diet.
The researchers looked at the number of calories the USDA says Americans consume and compared them with the number of calories people generally need to stay alive. Using data from 1971-2010, the researchers found that if the USDA data were correct, then a reference person (a hypothetical American established using algorithmic analysis of the data) would have lost nearly eighty pounds between 1971-1980 and also gained more than 215 lbs. between 1988-2010.
While such weight fluctuations aren't impossible, they're a rare occurrence at most.
If it were simply the case that the USDA compiled bad data, then there'd be little reason to express alarm. But it turns out the USDA uses these flawed data to inform and set federal dietary policy.
It's not just the data that's rotten. The laws and policies that are based on that data are inherently rotten, too.
The new research builds upon previous research by Archer, including a study published in the Mayo Clinic Proceedings last year, which found—as I wrote at the time—that the federal dietary guidelines "and the research used to support that work… is so off base as to be scientifically useless."
The new study doubles down on Archer's earlier work.
"These results demonstrate that the USDA's caloric data are meaningless and should not be used to inform public policy," Archer told me this week by email. Besides poking fatal holes in the federal dietary guidelines, Archer's research shows more broadly the perils of relying on bad data to inform law and policy.
Those who develop federal dietary guidelines are hardly alone in relying on incomplete, wayward, contradictory, or inconclusive data as the foundation for various food laws and policies. In fact, the push to adopt laws that seem to contravene what data tells us about those laws—namely, that they are uniformly bad ideas—continues headlong.
One recent example— menu calorie labeling—illustrates this point. A new study by NYU researchers reports that menu calorie labeling is a totally ineffective tool for helping consumers make lower-calorie food choices. That's just piling on. The fact that menu labeling doesn't achieve its goals is nothing new.
The data don't support it, yet mandatory calorie labeling is coming to chain restaurants, vending machines, and movie theaters (and, likely, grocery stores and pizza parlors) in every state in the land in mere months.
Data supporting soda taxes as a tool to combat obesity is virtually nonexistent. Yet cities proceed to adopt them.
The FDA's own data on the likely impact of the Food Safety Modernization Act (FSMA), as I detail in my new book, Biting the Hands that Feed Us: How Fewer, Smarter Laws Would Make Our Food System More Sustainable, shows these bad rules clearly aren't worth their enormous cost.
[T]he FDA's own estimates predict these rules could—if implemented to absolute perfection—reduce foodborne illnesses by a maximum of 1.23 million cases. That would represent just a 2.6 percent reduction in total foodborne illness cases. Again, this is the FDA's own best-case scenario for the impact of these two key rules.
Relying on bad data to justify food and dietary laws is as absurd, indefensible, and unscientific as it sounds. If we can't trust the government to base those food laws and policies that call for science on actual, you know, science, then maybe that's evidence the government should have far less power to craft those laws and policies in the first place.