According to a new working paper from the National Bureau of Economic Research, those groups include consumers who make less money, have less education, possess lower levels of nutrition knowledge and live in certain geographic areas — such as south-central Montana.
“Economists have a problem with the word ‘culture,’ ” said Hunt Allcott, an economist at New York University and the paper's lead author. But an area’s food culture, he acknowledged, “could make a big difference in the formation of food preferences.”
Allcott and his colleagues didn’t necessarily set out to study food culture. Or to call out Musselshell, which their data indicates is the country's least healthy county, according to grocery purchases. (Washington Post readers in the county note that residents often supplement their diets with items from hunting trips and home gardens, meaning that they may be eating more healthfully than measurable data indicates.)
Their paper, which has not yet been peer-reviewed, analyzed 12 years of Nielsen grocery purchase data from 100,000 households across the United States, looking for patterns that explain why low-income and high-income families buy such different foods.
Economists and public health advocates have long known that high-income households tend to buy more fruits, vegetables and proteins, and that a large dietary discrepancy exists between the rich and poor. The exact mechanisms of this gap are less clear, however. They can’t be entirely explained by the cost or availability of food.
For this paper, Allcott and his colleagues analyzed what happened to a household’s food purchases when a new, full-service grocery store opened in the neighborhood, or when the household moved from “food deserts” to areas with more abundant grocery options. But even major changes in a household’s immediate food environment, the paper found, had a limited impact on the foods that people purchased.
Instead, the economists write, there appears to be a great deal of variation in demand for healthful foods. And those big demand gaps tend to correlate with other, fairly predictable characteristics, such as income, education, nutrition knowledge — and geographic region.
The map below shows how unhealthy counties tend to cluster in the south central states. Each county is colored according to its “average health index,” a standardized measure of grocery nutrition based on the Dietary Guidelines for Americans, with darker shades representing the least healthful food purchases.
Among the healthiest counties: New York County, N.Y.; Pitkin County, Colo.; and Iroquois County, Ill. Among the least healthy: Menominee County, Wis.; Turner County, Ga.; McNairy County, Tenn.; and, of course, Musselshell.
“This suggests a separate role for culture and tastes that vary across regions,” the paper says.
Allcott acknowledged that the role of region was something of a surprise to him. He hypothesizes that a region’s dominant cuisine, be that barbecue or avocado toast, informs the meals that people eat as children. That, in turn, has a large effect on their lifelong food preferences. An upcoming paper from his team, looking at the same set of Nielsen data, will evaluate this thesis by tracking how eating habits differ according to where people lived at different life stages.
“We know that your early life experiences play a large role in how you eat,” he said. “That suggests there’s a possible intergenerational transmission of preferences.”
Geography can’t explain it all, of course. Allcott is emphatic that the link between region and healthy eating is not causative and more research will be needed to tease out its nuances. There are also other factors, which his model did not consider, that could explain lower demand for healthful groceries among certain groups, including time available to plan and cook meals, exposure to food marketing and stress levels.
Still, the idea that place could shape demand makes sense to many advocates. Candace Young, the associate director of research and evaluation at the national food-access nonprofit the Food Trust, agrees that nutrition interventions will need to address the healthful-food-demand gap. (Unlike Allcott, however, Young says that food deserts are a major part of the problem.)
The Food Trust offers free nutrition classes to both children and adults and has been a vocal supporter of programs like SNAP-Ed. When it comes to cultivating demand for healthful food, “there are a lot of barriers,” Young said.
“If you grew up in a community where you never saw an apple, but saw lots of fast food,” she added, “it could take a lifetime to shift those preferences.”
This story has been updated with comments from readers in Musselshell County, Mont.