Although mainstream food desert maps are useful in visualizing food access and the spatial inequities that stem from redlining, gentrification, and deindustrialization, they are also critiqued for oversimplifying community food sources, mobility, and priorities. This study seeks to address these limitations in North Nashville, an identified food desert, by creating a more refined and nuanced food access landscape map. Using direct observations in the form of market basket surveys, local demographic information, and suitability analysis, a combined weighted overlay of socioeconomic factors and food access was derived, which was joined with urban farm, community garden, emergency food, and supermarket locations. This study found that the resultant food access landscape provides finer resolution on previously identified food deserts and identifies overlooked areas that lack adequate food access. Despite the exclusion of additional food assistance programs and restaurants, the results of this study can assist communities prioritize where to mobilize resources to address food inaccessibility.
16 thoughts on “Nourishing North Nashville: Refining Food Access Landscapes”
Comments are closed.
Inspired by De Master et al. (2019), I wanted to map North Nashville at a much finer scale to better understand food access (or lack thereof) within the area. I combined direct observations on the availability of food to meet the USDA Thrifty Food Plan with local demographic information at the census block level to determine a range of factors shown in Table 1. I then used this information and community assets to create weighted overlays, which were then combined to result in the food access map shown in Figure 1. The refined food access landscape produced in this study was intended as a means of potentially assisting communities to pinpoint where additional resources could be mobilized to address issues of food access. I also want to emphasize that outside intervention or assistance must be made with community participation in the decision-making process. In a talk with David Padgett, Associate Professor of Geography at Tennessee State University, at Farm Burger (along Charlotte Ave.) last semester, he emphasized that food access is often an issue of transportation and despite the availability of foods at smaller grocers or corner stores, people may simply prefer to go to a large supermarket chain, which is less accessible due to bus route removal. As such, this study presents an opportunity for future community-based research. Thanks for tuning in!
This is very interesting and well researched. Your “figure 1” map really helps to clarify and demonstrate which areas of North Nashville are struggling compared to the USDA map in “figure 2”. This map could be also useful for city/state officials in providing better access (bus/transit) for those areas lacking in access to food providers/grocery stores, etc.
Thanks for your feedback! I completely agree with the use of this type of map in making public transit decisions. Within the last couple of years and especially now with COVID-19, many bus routes and bus stops are being removed due to budget-cuts and low-ridership, making public transportation a particularly relevant factor in food access.
Excellent project. Do you think that a higher resolution national food desert map would be both possible to create and worth creating? What do you think are the policy implications of such a map?
Thanks, August! Although a higher resolution food access landscape map would be a great resource at the national level, the time and financial aspect of data collection make this type of map impractical. I would say that a higher resolution map is particularly useful at the neighborhood level in helping communities identify sites for food distribution centers (or food banks), community gardens, etc. It is also useful at the city level to inform decisions about public transit, public assistance programs, and development.
The greater resolution of this project compared to the USDA map, as well as the nuance afforded by emphasizing informal community assets, is a testament to the insights able to be contributed by local community members. Very good — and very much needed — work!
Thank you, Isaac! I appreciate your feedback!
Hi Hannah,
Great project! I really found your point about how previous studies on food deserts have been conducted at the census track level instead of smaller blocks that you used in this scenario. Really good insight in this regard. You mention low-income factors throughout your piece, but do you think race could have played a role in the food desert scenario due to how limited stores can be for certain racial groups?
Thank you so much! I really wanted to create a map that looks a bit beyond the food desert concept and incorporate a greater variety of community food sources and assets. The USDA map uses low-income by a few different measures and I specifically use % below the poverty line. Redlining, gentrification, and even the legacy of highway expansion all disproportionately affect communities of color’s access to food. I definitely think race plays a role in this scenario; however, it would be difficult to weigh using suitability analysis. Instead, I would make a visual comparison of both the combined overlay and another map (by census block) of the % of each race within the AOI.
This is a very interesting and timely piece of research. The point that census level data can lack sufficient level of granularity to properly identify local levels of food poverty is an interesting one. To address this challenge, is the model that you have developed something that can be packaged for use across other locations to help local governments, NGOs/foodbanks better plan their activities?
Thanks for your feedback, JP! This is a great question. The methods used in this study can be replicated for other locations for which store locations (supermarkets, corner stores, specialty grocers, etc., if applicable) are known. The use of a ModelBuilder in ESRI software may also help optimize the methods for more efficient analysis across other locations.
Hello Hannah,
This is a very well thought out, interesting, and articulated project. As others have mentioned, the idea of evaluating food deserts at the block level vs. the tract level is substantially more nuanced and useful. Similarly, the poster looks nice and is easily digestible. I think your project has real-world application and policy weight. It really is remarkable to see how much of Nashville has difficulty accessing healthy food sources.
I am very interested to see how the results would change in relation to the Nashville MTA 2020 budget cuts and subsequent rerouting. In a hypothetical scenario, how would Let’s Move Nashville change your and the USDA’s map, and would it decrease the prevalence of food deserts? It is probably more difficult for light rail to eliminate food deserts due to its inflexibility. Another question I have is if you were to continue this project if you would weight distance in conjunction with travel time? I’m thinking anecdotally, but there is a Publix off of I-40 and state 70 in west Nashville (right at the left edge of your map by the river). It has a bus stop out front, maybe like a quarter-mile away, but to get into the Publix, one has to walk up the street, around the drive-in, down a steep hill, and across the entire mall parking lot. In your current model, how would this site be evaluated?
Hi Spencer! I appreciate your feedback and questions. In response to your first comment about the Nashville MTA 2020 budget cuts, I met with the CEO of WeGo, Stephen Bland, last semester for another course and that seems to be the general trend. He attributed much of the route changes (which were necessitated by budget cuts) to low-ridership, which often impact the most transit insecure populations. Looking at my map specifically, the % influence for distance to transit was 20%, using route and stop information updated last in October 2019 from Nashville’s Open Data Portal. Because of its relatively low overall influence, some changes in routes or stops may not have a drastic impact on the combined overlay. The Nashville MTA Budget Submission for FY 2020 shows a decrease in funding that was “used to support some services along the #10 Charlotte Pike,” which is located in the AOI. Additionally, the FY 2020 submission shows that in response to decreased funding, there may be “some reduction of bus service yet to be identified” and ” a fare increase or fare restructuring.”
In response to your second question, I would imagine that network analysis could be integrated to better account for walking up a steep hill, for example. The Publix you mentioned is just outside of the AOI, but in this current model, I use the distance from the centroid of a census block to the nearest feature (of each store type), which does not account for physical obstacles that may be present. I should also note that there was a supermarket (Aldi, Publix, Kroger, etc.) located within 0.25 mi for every bus route within the AOI.
Thanks again for your questions!
Great project, Hannah. Food access seems so basic and yet, I know, from courses and reading, that so many people in our country live in areas where their options are quite limited. Thank you for putting so much work into teasing out the inequities in Nashville! Have you put any thought into how the pandemic might have affected this data? I know where I live many store hours are reduced, food costs are significantly higher, and more people may now fall into that federal poverty characterization; however, on the other side, many places are now offering delivery services that were not seen before and, while probably not easily quantifiable, informal neighborhood outreach seems to be at a high currently. Can you imagine any way to account for these newer (though perhaps short-term trends) in a GIS analysis?
Hi Chris! Thank you for your input! I would venture to say that COVID-19 would impact a number of quantifiable measures, particularly household income and % of households below the poverty line (as in my map), which can be used to create an updated food access map. Additionally, budget cuts (see my response to Spencer’s question as well) to public transit may result in some route removal and fare increases. Social-distancing capacities may also present a concern for some high-ridership or commuter routes. My research for the development of this map did involve a calculation of a market basket price for each of the surveyed stores, which may be compared to the Consumer Price Index or future market basket surveys of these stores.
Very clearly presented model and exposition of the results; good going. Block group level census data bring the analysis much closer to lived experience, because there is less heterogeneity within block groups than tracts. You might even be able to specify further by looking at the distribution of houses/buildings within block groups, to weight population densities within them, likely through an interpolation of buildings to a surface. That would get you “inside” the block group. But I think your analysis is a big step forward, and also more holistic because you are taking into consideration a greater range of social factors and potential food sources. Most importantly, it produces a model that can be the basis for targeting market-based, governmental, and ngo-based interventions. Where specifically would you recommend interventions, and what kinds of interventions would you recommend to policy makers and stakeholders? Great job.