ABSTRACT
Food deserts have received considerable attention in recent years. Research has shown that individuals living in these areas have less nutritional diets and experience worse health outcomes compared with those not living in food deserts. Though the ramifications of living in food deserts are well known, less is known about efforts to alleviate these effects. One proposal for increasing food access while improving nutrition is the farmers market. While the farmers market has been championed as a potential solution and celebrated when it succeeds, little work has explored how often farmers markets are in food deserts nationally. This article explores this by using data from the United States Department of Agriculture and the American Community Survey. Results suggest that though much praise is given toward the promise of farmers markets in food deserts they are infrequently found there. This research contributes to illuminating the present state of US food deserts and proposes several questions about the efficacy of farmers markets as a tool to alleviate the impacts within food deserts.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. Despite the controversies and being mindful of these critiques, I believe that the term food desert still has analytical value toward examining access to healthy food.
2. Supplemental security income is also a form of public assistance; however, the ACS separates it into its own distinct variable.
3. Meaning that within a tract the sum of doing okay, struggling, and doing poorly totals 1.00.
4. The six types were farmers market neighborhood, non-farmers market neighborhood, food desert neighborhood, non-food desert neighborhood, food desert and farmers market neighborhood, and food desert without a farmers market neighborhood.
5. There is an exception to this where median income is lower in farmers market neighborhoods when compared with the US average.
6. An odds ratio lower than 1 suggests a decreasing likelihood of a neighborhood being a food desert, whereas an odds ratio greater than 1 predicts an increasing likelihood of a neighborhood being a food desert. The size of the odds ratio matters, where increasing odds ratios above 1 reflect a stronger likelihood of a neighborhood being a food desert (or vice versa for ratios less than 1).
7. Five tracts were dropped in the second model because their social support scores were significant outliers (more than five standard deviations from the mean of social support).
8. Quadratic terms are created by squaring a variable. They are included in regression to test linearity, specifically non-linearity, between an independent and dependent variable. The variable used to create quadratic terms predicts the general direction of the relationship between itself and the dependent variable, whereas the quadratic variable predicts any non-linearity to the relationship.
9. Odds in quadratic terms follow the same logic as singular variables, meaning that those with predictions of more than one have an increasing effect; whereas those with predictions of less than one have a decreasing effect. Since the singular proportion white effect predicts a decreasing effect, the quadratic term (somewhat counterintuitively) predicts the likelihood to continue to decrease, but at an increasing rate.
10. The model does not provide direct evidence of this. Equally as possible is that other forms of food distribution may be causing these predictions to occur.
Additional information
Notes on contributors
Justin Schupp
Justin Schupp is an Assistant Professor of Sociology at WheatonCollege MA. His research and teaching interests revolve around theintersections of food systems, social movements, and stratification.If he is not engage with these interests, you will probably find him inhis backyard building a treehouse with his children.