Masters Theses

Date of Award


Degree Type


Degree Name

Master of Science



Major Professor

Thomas L. Bell


The purpose of this thesis was to ascertain the factors governing a metropolitan area's (MSA's) retail provision and to determine the degree of influence these factors imposed upon the particular metropolitan area. This thesis was an extension of earlier research performed on randomly selected cities throughout the United States. For this analysis, I selected 78 metropolitan areas, comprising all census-designated Metropolitan Statistical Areas (MS As) in the Southeastern states of Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Tennessee. I utilized the National Research Bureau's 2003 Shopping Center Directory as the predominant source of data to determine each MSA' s aggregate retail shopping center square footage. The total area of an MSA devoted to retailing is not available in a comprehensive and consistent format. Using central place theory and marketing principles as guidelines, I first hypothesized that the total amount of retail square footage in a city is an approximate function of that city's population. I performed a simple ordinary least squares (OLS) regression analysis with 2000 MSA Population as the independent variable and MSA Total Retail Square Footage as the dependent variable to test this initial hypothesis. This regression yielded an extremely high coefficient of determination of 0.95 between the two variables. Larger cities obviously have more retail selection than smaller ones. The amount of retail space per person, however, will not necessarily be higher in the larger metropolises. I then focused the remainder of my study on retail square footage per capita in order to understand what dynamic forces other than population affected the number of square feet of retail space allotted to the average resident within an MSA. Using descriptive statistics, I sorted the MSAs on the basis of each area's Gross Leasable Area (GLA) per capita and selected those areas deviating by more or less than two standard deviations from the aggregate Southeastern MSA mean of 20.26 square feet per capita for further in-depth analysis. Before investigating the individual MSAs that were categorized as outliers, I attempted to understand the underlying trends within the aggregate data. Performing more single regression analyses with MSA Population, 1990-2000 Population Change, and Median Household Income as the independent variables and MSA GLA per capita as the dependent variable yielded few conclusive correlations. I next entered the three independent variables simultaneously in a multiple regression analysis to explain statistically variation in the dependent variable GLA per capita. This analysis yielded a multiple coefficient of determination of 0.45. More qualitative factors, such as tourism influences, proximity to larger markets, presence of major employers or universities, transportation infrastructure, and governmental policies and restrictions, were then used to understand the remaining 55% of variation left unaccounted for by the three independent variables. These catalysts or inhibitors, along with other non-quantifiable ones, occurred, to some extent, in every MSA and were not well-predicted by the multiple regression analysis. Certain overall trends were evident. Median household income, growth rate, and the effects of outside tourism contributed the most in explicating an individual area's retail dearth or excess. These trends translated into major causes for retail growth or contraction on the individual level. I detected the presence of one or more of these previously unmeasured factors in all eleven MSAs that were deemed to be vastly under or over-performing on the initial GLA per capita analysis. The five "true" overperforming MSAs (Myrtle Beach, SC; Fort Myers-Cape Coral, FL; Naples, FL; Raleigh-Durham-Chapel Hill, NC; and Atlanta, GA) have all enjoyed high growth rates, continuous inflow of outside tourism-based revenues, and overall higher-than-average income levels, thus allowing the retail stock within their MSAs to thrive and expand. The other slightly overperforming MSA (Knoxville, TN) conquered the hindrance of a linear spatial configuration, which would normally bring a city's GLA per capita score down, by having high levels of tourism (e.g., Great Smoky Mountains National Park and Dollywood) and the lack of viable retail alternatives within reasonable driving distances. Those MSAs displaying below average GLA per capita (Anniston, AL; Sumter, SC; Johnson City, Kingsport, Bristol, TN-VA; Fayetteville, Springdale, Rogers, AR; New Orleans, LA) were also influenced by many of the same factors as the high-scoring ones, but these factors acted as inhibitors rather than catalysts. These areas lacked most of the GLA per capita-raising factors such as elevated household income or a large tourist base and, thus, were kept from sufficiently expanding their respective trade areas and retail bases. The final chapter of the thesis focuses on the shopping center and its place in contemporary and future United States society. The construction of smaller open-air centers, referred to as lifestyle centers, and larger, but also unenclosed, power centers continues as Americans seem to have tired of cookie-cutter antiquated indoor malls. The fates of many smaller regional and inner city malls appear uncertain, and possibly sealed as they become renovated as office buildings or erode into vacant, uninhabited structures, while most suburban super-regional centers will continue to thrive. The overall number of these sometimes-overwhelming giants, however, is expected to increase only marginally. There is no territory left for them to conquer.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."