I'm Not Lovin' It...

We all know ‘that one street’: the street that seems to have every fast food restaurant compiled into a hot spot, creating an oasis of calories for the willing participant. A Taco Bell next to a McDonalds, across from a Wendy’s and adjacent to a KFC establishment. To some, this might seem like a paradise. To others, it only raises questions of concern. The abundance of fast food restaurants in urbanized areas begins to seem excessive, and even more so when one begins to observe nearby demographics of a region. This line of thinking leads to the concept of food deserts and the less commonly discussed food swamps. Food deserts are regions that have a small amount of access to any food options--especially healthy ones--and require people to travel far for nutritious meals or settle with the options they have. Food swamps, on the other hand, address areas that are overrun with food options, but none of them too good. Food swamps typically are flooded with fast food restaurants with the occasional grocery store, still leaving nutritious options out of reach or less accessible. Many researchers believe that food deserts and food swamps are mostly found in areas that have a higher poverty level or less of an average income.

In this case, a few years of reading essays and journals about the subject raised a specific question for me: are fast food restaurants deliberately placed in the more impoverished regions of Colorado? Where are these regions, and how can I come to that conclusion? The predicted result is that yes, areas with a higher percentage of impoverished individuals most likely have a higher density of fast food restaurants within their county. While people often assume the prevalence of fast food restaurants in impoverished regions, being able to visualize this data will be vital to take steps in the future to find healthier solutions for areas that may be troubled with food swamps and understand the first areas where action should be taken.

I intend to test my hypothesis by observing the amount of fast food restaurants per. square mile on a Colorado county level, putting this data against the estimated percentage of the population that is impoverished during the available data frame in RStudio. Understanding the amount of fast food restaurants by square mileage can identify the density that they occur, explaining if areas with a higher poverty percentage are targeted by food swamps. By pulling data from the US Census Bureau, I was able to curate a dataset that included the estimated percentage of impoverished compared to total population and the number of limited service restaurants (LSR). Let’s take a moment to detail the process of going from lLSRs to the density per. square mile.

LSRs are explained by the North American Industry Classification System (NAICS) to be restaurants that “[provide] food services ... where patrons generally order or select items and pay before eating.” (NAICS, n.d.). This data is compiled in census.gov as the number of firms and the number of establishments. I utilized the number of establishments, as this includes the firms and buildings the firm--or ‘home’ building--owns. This was not enough for me, though. I wanted to determine the density of fast food restaurants per. square mile to compare to the percentage of a whole population. In order to do this, I conducted the following equation:

After curating each map and developing a bivariate representation of the density of fast food restaurants compared to the percent of total population in poverty by county proved to create a significant amount of overlaps. In analyzing this map, it is rare to find a  county that has a low amount of both variables, the only examples being Morgan and Phillips county. A total of 7 counties were identified to have an overwhelming percentage of impoverished with a larger density of fast food restaurants, while many others contained a blend of the two variables with one overwhelmingly strong variable, such as Gunnison, Lake, Ouray, and Logan county. Interestingly, there were no areas with an overwhelming amount of impoverished populations and a small density of LSRs per. square mile, while conversely there were a fair amount of areas with a low rate of poverty and a large amount of restaurants. These data points lead me to believe that, more often than not, areas with a high rate of poverty often have a higher density of fast food restaurants in their county lines. Additionally, regions with a low poverty rate still often have abundant amounts of fast food restaurants.

An undeniable factor of this map and dataset are the multiple unaccounted counties, making up nearly half of the total counties of Colorado. Unfortunately, much of the data was not provided for the LSRs in each county, making the obtaining of data across the entire span of Colorado more difficult. On the source website, this is stated to be due to the disclosure of data from individual companies, where the data is only represented when it is in higher numbers. This can provide the idea that these counties may have little to no LSRs and may require data collection or need not be involved in this analysis.

Overall, my hypothesis was reflected properly in the data. There were a large amount of areas with a medium to high density of fast food restaurants in impoverished counties. While this did support my hypothesis, this is not the only work to be done on this subject. Something I want to correct with this data is understanding that many of these counties are not fully urbanized, and taking the square mileage of the county does not reflect the actual populated areas of the county. By doing this work on a census tract level, it would be easier to comprehend exactly what regions of a county need attention, and if any false information was reflected in this research.