<%BANNER%>

Military-Base Impact on a Local Economy: A Case Study of Three Military Bases in Two Metropolitan Statistical Areas


PAGE 1

MILITARY-BASE IMPACT ON A LO CAL ECONOMY: A CASE STUDY OF THREE MILITARY BASES IN TWO METROPOLITAN STATISTICAL AREAS By KENNETH E. HAWKINS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Kenneth Eugene Hawkins

PAGE 3

This thesis is dedicated to my three child ren; Rachel Anne, Beatrice Rose, and Quinton Zachary with the hope of inspiring them to achieve higher goals than mine.

PAGE 4

iv ACKNOWLEDGMENTS I thank my wife Maria for her support, understanding, and help during the time of separation and hardships, especially since she also earned her Doctorate while working and keeping our son. I thank my mother a nd Aunt Faye for instilling in me the importance of an education and the pursuit of my dreams. I thank my mother-in law, Dr. Veronica Free, for the advice and support from a professional educator’s perspective. I thank my father-in-law and step-mother-in-la w (Drs. Mario and Carol) for their support and assistance during my years in the graduate program at the University of Florida. I thank the professors in the Department of Geography and the office assistants for their time and effort in helping me through the years in graduate school. I thank Dr. Paul Zwick (a member of my supervisory comm ittee) and Professor Stanley Latimer. Their classes and wisdom in uses of geostatisti cal analysis and spatial analysis were instrumental in my research. I thank Dr. Jane Southworth for her assistance and advice as a member of my supervisory committee, and fo r her classes in remote sensing. Finally, I thank my mentor, tutor, and inspiration in studying quantitative methods: Dr. Timothy Fik. I hope to do him justice with my work a nd will remember the efforts and assistance he gave me for many years.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW....................................................1 Base Impact on Urban and Economic Growth of the Localized Area.........................1 Problems Associated with Base Realignment and Closures.........................................1 Base Realignment and Closure, Economic Impact, Accessibility, Military Bases, and the Central Business District..............................................................................5 Determining Base Realignment and Closure (BRAC)..........................................5 Concern over Economic Impact..........................................................................19 Alleviating Negative Economic Impact with Accessibility................................20 2 METHODS.................................................................................................................23 Economic Growth and Development because of Military Bases and the Central Business District: Economic Base Theory and Accessibility.................................23 Military Bases, Central Place Theory, and the Central Business District (CBD)...............................................................................................................23 Economic Base Theory........................................................................................24 Methods for a New Approach to Predicting Economic Impact from BRAC.............25 Military Bases Applied to Economic Models.....................................................25 Previous Economic Impact Studie s from Military-base Closures.......................28 New Approach to Predicting Economic Impact from BRAC....................................34 Base Realignment and Closure ( BRAC) 2005 and Florida's Bases....................34 Jacksonville and Tampa, Florida.........................................................................35 Jacksonville NAS and Mayport NS Study Areas and Diversity of Industrial Employment: 1980 and 2000...........................................................................37 MacDill AFB and Tampa, Florida.......................................................................38 Recovering from Economic Impact A ssuming Tampa or Jacksonville Bases Are Selected for BRAC 2005..........................................................................43

PAGE 6

vi Software Used for Our Study..............................................................................43 3 MAPPING URBAN GROW TH AND CHANGE.....................................................44 Software Used to Create the Images for the Study Areas..........................................45 Creating Shapefiles for the Study Areas.....................................................................48 Establishing Study Area Bounda ries Using Shapefiles..............................................49 The Jacksonville Metropolitan Statistical Area...................................................49 Mayport NS Study Area......................................................................................52 Jacksonville Naval Air Station (Jax NAS)..........................................................54 Tampa-MacDill AFB Study Area.......................................................................59 Empirical Approach to Study Areas...........................................................................62 Dependent Variables and Each Study Area.........................................................62 Independent Variables and Each Study Area......................................................63 Results of Choropleth Maps for the Jax NAS Study Area..................................65 Results of Choropleth Maps for the Mayport NS Study Area.............................66 Results of Choropleth Maps for the MacDill AFB Study Area..........................68 Review of Choropleth Maps Results...................................................................69 4 REGRESSION ANALYSES OF THE STUDY AREAS..........................................84 Jacksonville NAS Study Area Re gression Analyses Results.....................................85 Mayport NS Study Area Regre ssion Analyses Results..............................................89 MacDill AFB Study Area Regression Analyses Results............................................92 5 SUMMARY AND CONCLUSION.........................................................................112 Summary...................................................................................................................112 Concluding Remarks................................................................................................116 APPENDIX KEY TO INDEPENDENT VARIABLES.................................................120 LIST OF REFERENCES.................................................................................................131 BIOGRAPHICAL SKETCH...........................................................................................137

PAGE 7

vii LIST OF TABLES Table page 2-1. Employment populations for Jacksonville Naval Air Station (Jax NAS) study area........................................................................................................................... 39 2-2. Employment populations for Mayport NS study area..............................................40 2-3. Employment populations for MacDill AFB study area............................................42 3-1. Key to dependent variables.......................................................................................64 3-2. Jax NAS dependent variables...................................................................................70 3-3. Mayport NS dependent variables.............................................................................72 3-4. MacDill AFB dependent variables...........................................................................73 4-1. Jacksonville NAS study area regre ssion results for percentage change in commercial area between 1973 and 2000................................................................96 4-2. Jacksonville NAS study area regression results for percentage change in median household income between 1980 and 2000..............................................................98 4-3. Jacksonville NAS study area regre ssion results for percentage change in residential area between 1973 and 2000..................................................................99 4-4. Mayport NS study area regression resu lts for percentage change in commercial area between 1973 and 2000..................................................................................102 4-5. Mayport NS study area regression re sults for percentage change in median household income between 1980 and 2000............................................................103 4-6. Mayport NS study area regression resu lts for percentage change in residential area between 1973 and 2000..................................................................................105 4-7. MacDill AFB study area regression resu lts for percentage change in median household income between 1980 and 2000............................................................107 4-8. MacDill AFB study area re gression results for percentage change in residential area between 1980 and 1999..................................................................................109 A-1. Key to Independent Variables...............................................................................120

PAGE 8

viii LIST OF FIGURES Figure page 2-1. Income flows............................................................................................................ 25 3-1. Jacksonville Metropolitan Statistic al Area (Image created using FGDL and Census Bureau Data)................................................................................................46 3-2. Mayport NS study area census-tract boundaries.......................................................50 3-3. MacDill AFB study area census-tract boundaries.....................................................50 3-4. Jacksonville NAS study area census-tract boundaries..............................................51 3-5. Modes of transportation in the Mayport NS study area..............................................53 3-6. Urban and economic growth in Mayport NS study area (1973-2000)......................54 3-7. Transportation routes to Jax NAS.............................................................................56 3-8. Residential growth in th e Jax NAS study area (1973-2000).....................................57 3-9. Commercial growth in th e Jax NAS study area (1973-2000)...................................58 3-10. Tampa MSA............................................................................................................61 3-11. Modes of transportati on accessible to MacDill AFB..............................................63 3-12. Commercial area pe rcentage change 1973-2000 for Jax NAS study area..............75 3-13. Residential area percentage ch ange 1973-2000 for Jax NAS study area................76 3-14. Median household income percentage change 1980-2000 for Jax NAS study area........................................................................................................................... 77 3-15. Commercial area per centage change 1980-2000 for Mayport NS study area.........78 3-16. Residential area percentage cha nge 1973-2000 for Mayport NS study area...........79 3-17. Median household income percentage change 1980-2000 for Mayport NS study area..................................................................................................................80 3-18. Commercial area per centage change 1980-1999 in the Tampa study area.............81

PAGE 9

ix 3-19. Residential area percentage ch ange 1980-1999 in the Tampa study area...............82 3-20. Median household income percentage change 1980-2000 of the Tampa study area........................................................................................................................... 83

PAGE 10

x Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science MILITARY-BASE IMPACT ON A LO CAL ECONOMY: A CASE STUDY OF THREE MILITARY BASES IN TWO METROPOLITAN STATISTICAL AREAS By Kenneth E Hawkins May 2005 Chair: Timothy Fik Major Department: Geography Military bases have had a profound impact on urban and regional economic growth. Our objective was to assess the impact of military bases on local economic-growth rates in three distinct metropolitan statistical areas (MSAs). We examined the extent to which proximity to a military-base or the Central Business District (CBD) affects local economic growth rates and the degree to whic h variability in growth is explained by the distance to a base. Our study population included three metropolitan statisti cal areas: Jacksonville Naval Air Station (NAS) and Mayport Naval Station (NS) located in Jacksonville, Florida, and MacDill Air Force Base (AFB) located in Tampa, Florida. We used spatial analysis and multiple regression analysis to determine a discernable impact on economicgrowth rates of the localized areas from each of the three military bases at the 95% confidence level. Our hypothesis was that military bases have a discernible impact on economic growth rates at a geographical scal e (census-tract level) because of proximal

PAGE 11

xi distances and accessibility along transportati on corridors between the base and major commercial and financial nodes at the 95% confidence level. Spatial analyses showed di scernible impact on the economic growth rates of the study areas; however, the cause of economic growth is not discernible among impact from the base, commercial nodes, economic node s, demographics, distance variables and accessibility variables in the study areas. Regres sion analyses revealed possible positive and negative causes for economic-growth ra tes of the study areas. However, the significance of base impact on economic-growth rates was negligible in all cases. The study areas showed no evidence of localized spillover effects.

PAGE 12

1 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW Base Impact on Urban and Economic Growth of the Localized Area Military bases have had a profound impact on urban and regional economic growth. Our objective was to investigate to what exte nt variability in urban and economic growth rates are explained by proximity to important nodes in 3 distinct st udy areas. Specifically, we investigated the degree to which variability growth ratios are ex plained by distance to a military-base vs. other prominent node s in the metropolitan areas (e.g. CBD, commercial, residential, and indus trial) by spatial relationships. Our hypothesis was that military bases have a discernible impact on economicgrowth rates at a geographical scale (censustract level) because of proximal distances and accessibility along transpor tation corridors between the base and major commercial and financial nodes at the 95% confidence level. Problems Associated with Base Realignment and Closures Base Realignment and Closures (BRAC) ha ve occurred since the formation of the Department of Defense (DoD). Before the 1980s, BRAC was a defense function handled primarily by the DoD. The DoD did not have to answer to any governmental offices when BRAC actions were considered. After the Vietnam War, Congress began to get involved because they felt that BRAC was target ed toward those states that did not fall into line with defense spending. Several congressional members began organizing to prevent BRAC; and by the late 1970s, Congres s passed regulations that prevented the DoD from approving any BRAC actions w ithout congressional authorization.

PAGE 13

2 Congressional intervention prevented furt her BRAC actions until 1988. The events that occurred between 1981 and 1988 concerning the United States and the Soviet Union led to the eventual collapse of the Soviet Union and a change in the DoD’s mission and planning. The DoD had to consider the downsiz ing of personnel and th e ability to rapidly respond to changing military and strategic needs. The development of new weapon systems and training led to a “modernization” of the DoD and a reevaluation of their infrastructure. The DoD had to deal with another probl em from Congress while modernizing. By 1988, defense spending had been an integral part of the Reagan administration’s plan on crippling the Soviet economy. The intense defense development programs during the 8 years of Reagan succeeded in bringing an end to the 40-year-old Cold War. Problems associated with the post-Cold War defense spending brought about a change in Congress and cuts in defense budgets. The DoD argued that cuts in their budget would affect development of training, and weapons, and the creation of rapid deployment forces. Congressional leaders argued that downsizing military strategies would include a review of weapons development programs and potenti al base closures. The DoD argued that some bases were no longer vital to the new mission of post-Cold War defense, leading to new rounds of BRAC. Eventually, Congress agre ed that there was a need to consider BRAC actions. The new rounds of BRAC were considerably different than BRAC actions before the 1980s. The new rounds of BRAC that C ongress agreed to would be more of a congressional function than BRAC actions of the past. Although the DoD would have a say as to what bases should be considere d, a commission would be es tablished to review the bases and make recommendations to Congr ess. BRAC would evolve over the next 7

PAGE 14

3 years because of differences between Congr ess and the DoD; however, intervention in the 1995 round by President Clinton preven ted any further BRAC action until 2005. The Government Accounting Office (GAO) and th e Congressional Budget Office (CBO) are the two primary organizations that review and report th e progress of previous BRAC actions and serve as part of a check-and-ba lance system between the DoD and Congress. The problem is that both the DoD and C ongress have alternative agendas concerning BRAC. The DoD needs to close unnecessary bases; however, Congress is concerned about economic impact from base closures. A lthough, there is a need to close bases, political considerations come into play. Congress paints a gloo m-and-doom picture of base closures (which previous research has shown is not necessarily true); and the DoD has a valid argument that base reduction has not matched the reduction in force, and that reduction in costs for unnecessary bases can be effectively applie d to other budgetary concerns. Preparation of BRAC in cities and metropo litan statistical areas (MSAs) with bases can prevent or reduce the e ffects of economic impact. When a new round of BRAC is announced, federal, state, and local government officials actually try to prevent base closures in their respective areas. Unfort unately, the local population believes in the gloom-and-doom scenario painted by Congress and will do anything they can to prevent the closure of bases to protect their local economics. Much preparation and planning must be done to prepare for the economic impact that occurs with base closure and/or realignment. City planning and commerce of ficials, and other business and government organizations can reduce the negative econom ic impacts with proper organization and planning. The land use associated with the loss of a base can be replaced with new businesses and industries by rezoning. If subs titute industries are situated to move in,

PAGE 15

4 when the military moves out, the impact shou ld be substantially reduced. There is one other useful alternative of proper planning and organizati on: the prevention of BRAC. One of the criteria that the DoD and BRAC commissions consider when choosing bases for BRAC actions is the urban growth of th e local area approaching the outer boundaries of the base (also known as encroachme nt). The DoD’s major concern about encroachment on a base’s boundaries mainly ar ises with bases that have airfields. Complaints of noise pollution a nd other civilian issues generally cause problems for the military in the local area. The DoD has argue d that local planning and commerce officials should keep in mind the problems associated wi th encroachment and other civilian issues when determining future land use in the local areas. BRAC assessments need to (and usually do ) consider the economic impact of the local area. A key to preventing real dangers to local communities and economic impact is to consider the size of the urban area a nd the diversity of its economy; most BRAC actions in the past have involved large military bases in or near MSAs. Recent studies (Dardia et al. 1996, Cockrell 1998, Hooker and Knetter 2001) that the economic impact in past areas have not been as severe as predicted. Recovery of the areas has been swift and their economies have typically recovere d fairly rapidly. The areas affected were typically in major MSAs with diverse industries that aided in economic recovery. Future BRAC assessments need to consider the impact of bases in smaller urban areas (such as Jacksonville, NC; and Manhattan, KS) whos e industries are predominantly servicerelated and thrive on the military presence. BRAC actions in these types of communities can be devastating. Accessibility of bases to the central business dist rict (CBD) may also play a vital role in how urban planning a nd economic growth rates in areas near bases

PAGE 16

5 may be potentially affected by base closures. Land use before, during, and after BRAC needs to be considered if negativ e economic impacts must be reduced. Base Realignment and Closure, Economic Im pact, Accessibility, Military Bases, and the Central Business District Determining Base Realignment and Closure (BRAC) The first BRAC Commission was formed in December 1988 and Congress authorized a new round of closures. Reminded of the difficulties of base closures in previous years, Congress acted by passing legi slation that would al low the DoD to close bases, yet allow Congress to keep control of the proceedings and actions involving these base closures. The Defense Base Closure and Realignment Act of 1990 was passed to correct deficiencies from the first rou nd of BRAC in1988. Many congressional members were not overly enthused with the select ions made in 1988, particularly those from California. Since 1988, economic impact of the urban areas affected by BRAC recommendations have caused resistance from Congress and individual states’ legislative bodies in an era where it was necessary to reduce the military infrastructure which meet the post Cold War goals of the Department of Defense (DoD). Th ere are a great number of government documents from the G overnment Accounting Office (GAO) and Congressional Budget Office (CBO) that expr ess in detail the concerns of the DOD and Congress and economic impact from BRA C recommendations. The GAO publications primarily discuss the importance of the re duction in military infrastructure. Yet the concern of economic impact because of BRAC is discussed in the same GAO publications. According to the GAO (Holman 1995, 1996, 1998, 1999, 2001a, 2001b, 2001c, 2004, Wiggins 1996) the increases in the DOD budget and spending during the 1980s led to significant changes in the future of military planning and policy. The DoD

PAGE 17

6 reevaluated the mission objectives and determ ined that excessive spending could be reduced in some areas and redirected to m eet the requirements for spending on research and development (R&D) of better weapons sy stems, equipment and improved readiness of units. Once the DoD determine what bases or functions on a ba se were unnecessary, a list of those facilitie s and functions were drawn a nd proposed to BRAC. BRAC later released those reports to Congress and the Congressional Budget Office (CBO) as well as the Government Accounting Office (GAO). Recommendations were made to determine the effect of cost reduction for the DoD and the negative impact on the economies of urban areas in close proximity to the base s targeted for closure or realignment. Of importance was the impact on the empl oyment of the local community, real estate values, and future us e of the land returned to the urban region. The DoD is interested in the cost to maintain, train and operate the facility and its personnel. Another issue that DoD takes into considerati on is the urban region's growth and the encroachment of growth toward the base a nd the subsequent; particularly bases with aviation units that require f light training and operations. Vi rtually no attention, however, is paid to the spatial relationships betw een military bases and the Central Business Districts and other important nodes in the urban area. Moreov er, the economic impact of military personnel leaving the urban region and the loss of linked income is of secondary concern. The studies need to look much further than just the loss of jobs in the civilian sector and changes in real estate values as future business opportunities from the availability of land may play a vital role in the recovery process once a base is closed. Urban planners must control the potential grow th in areas progressing into encroachment on the fringe of military bases, and plan in accordance to remove the possibility of encroaching land uses. The first step in planni ng is to determine the location of a base’s

PAGE 18

7 aviation component and second the base’s training areas. By creating a buffer or no growth zone between the urban land use and th e base would improve a region’s chance of surviving possible base closure. A good exampl e of land use encroachment can be seen in the MacDill Air Force Base study area because th e base is centrally located in the Tampa MSA. Jacksonville Naval Air Station has not been as fortunate as urban growth from Orange Park (south of the ba se) and Jacksonville (north of the base) have surrounded the boundaries of the base. Fortunately, the base’s airfield is located away from the urban growth and near the St. John’s River. While urban sprawl has occurred in the area surrounding the base, the effects of this urban growth pattern has been beneficial to the base because of the real estate values and th e industrial and commercial activity near the base. Thus, encroachment has not been a major problem even though the urban growth reached the boundaries of the base. One final factor that prevents recommendation for closure is the mission of the base: it is hom e to the air wings that support the carrier group based at Mayport NS. Closure of Ja x NAS could occur, however, the space required for the base’s mission is not sufficient at Mayport NS Therefore, if the Jax NAS closed, the mission of Mayport NS would be adversely affected. Without air support for the carrier group stationed at Mayport NS, th e mission of the base would be weakened enough for base closure consideration. By 1995, the DoD had established a list of cr iteria that appeared to satisfy both sides. The list of DoD criteria was pu blished by the GAO (GAO-95-133, 1995) and is the basis for the DoD’s recommendation for future rounds of BRAC. According to the GAO the list of DoD criteria is broken into three categories with individua l criteria for base selection in each category. DoD Criteria for Selecting Base s for Closure or Realignment:

PAGE 19

8 Military Value (priority consideration is to be given to the four military value criteria) Current and future mission require ments and the impact on operation readiness of DoD’s total force. The availability and condition of land, f acilities, and associated airspace at both the existing and pote ntial receiving locations. The ability to accommodate contingency, mobilization, and future total force requirements at both the existing a nd potential receiving locations. Cost and manpower implications. Return on investment The extent and timing of potential cost s and savings, including the number of years, beginning with the date of comple tion of the closure or realignment, for the savings to exceed the costs. Impact The economic impact on communities. The ability of both the existing a nd potential receiving communities’ infrastructures to support forces, missions, and personnel. The environmental impact. Once the bases are recommended on these crit eria, the recommendations are presented to the BRAC Commission for further investiga tion. The final BRAC recommendations are then presented to legislative and executive branches for approval. Although the process seems rather simple many factors can disrupt the process beginning with the DoD selections. A GAO re port (Holman 1995) stated that the DoD is sensitive to the economic impact on affect ed communities; however, later GAO reports (Holman 2001c, Holman 2004) discussed the importance economic impact has in DoD’s selection criteria, but military value has top priority in all selection processes. The GAO reported (Holman 2004) a list of requirements that the DoD adopted to establish a guideline for all services and DoD agencies to implement in the base selection process:

PAGE 20

9 All installations must be compared equally against selection cr iteria and a current force structure plan developed by the Secretary of Defense. Decisions to close military installations w ith authorization for at least 300 civilian personnel must be made under the BRAC pr ocess. Decisions to realign military installations authorized for at least 300 civilian personnel that involve a reduction of more than 1,000 or 50 percent or more of th e civilian personnel authorized, also must undergo the BRAC process. Selection criteria for identif ying candidates for closure and realignment must be made available for public comment before being finalized. All components must use specific models for assessing (1) the cost and savings associated with BRAC actions and (2) th e potential economic impact on communities affected by those actions. Information used in the BRAC decision-maki ng process must be certified—that is, certified as accurate and complete to th e best of the originator’s knowledge and belief. This requirement was designed to overcome concerns about the consistency and reliability of data used in the process. An independent commission is required to review DoD’s proposed closures and realignments and to finalize a list of pr oposed closures and realignments to be presented to the President and, subject to the Presiden t’s approval, to Congress. The BRAC Commission is requir ed to hold public hearings. The BRAC process imposes specific time fram es for completing specific portions of the process (see app. I for time fram es related to the 2005 BRAC round). The President and Congress are required to accept or reject the Commission’s recommendations in their entirety. In addition to GAO’s role in monitoring th e BRAC process, service audit agencies and DoD Inspector General (IG) personnel are extensively involved in auditing the process to better ensure the accuracy of da ta used in the decision-making and enhance the overall integrity of the process. One requirement caused some previous co mplications with the BRAC process: the acception or rejection of the BRAC Commission’ s recommendations in their entirety by both the President and Congress. Apparently Congress gives the final endorsement for all BRAC recommendations. Twight (1989) gave an insight to congr essional actions concerning BRAC recommendations. Accord ing to Twight (1989), congressional

PAGE 21

10 members appear to be more concerned with their own political ambitions than addressing the issue of needed BRAC actions, which according to the previously mentioned GAO and CBO reports, Congress agreed that there is a need for BRAC. Twight (1989) explained the question of congressional de cision-making in the BRAC process through several examples such as prio r authority for base closures and realignment before 1988, actions that occurred because of reco mmended base closures before 1988, and congressional actions that occurred to refuse the recommendation of DoD proposed BRAC actions before 1988. Schwalbe (2003) de scribed the actions and attitudes of Congress, the Presidency, and DoD duri ng the recent BRAC rounds and supported Twight’s research. Another problem that needs to be addresse d is the education of the general public concerning BRAC. As long as Congress de fines the economic impact as being devastating to local, regional, and state economies, the average citizen without a grounded understanding of econo mic principle will believe th at their own economic wellbeing will be jeopardized with any BRAC ac tions in their community. Dardia et al. (1996) conducted research for the RAND Institu te that studied the effects of BRAC in three communities in California between 1992 and 1995. When the bases in the study (Fort Ord, George AFB, and Castle AFB) were selected for closure many California legislators at both state and fede ral levels tried in vain to prevent the closures. The study eventually concluded that pred icting economic impact is extr emely difficult, especially in the before closure stages, and those predictions are not necessarily an end-all scenario as they may appear for the communities affecte d, however, they also concluded that waiting for long-term studies to be conducted is not reasonable. The answer to these problems is very complex because the problem is very complex. The past history between the DoD

PAGE 22

11 and Congress concerning BRAC is key to solv ing many of the problems that have arisen in the previous four rounds of BRAC. One goal was to meet the requirements of the Balanced Budget and Emergency Deficit Control Act of 1985. The Senate bi ll S.1702, sponsored by Senators Gramm, Rudman and Hollins, called for the elimina tion of the federal budge t deficit by cutting the domestic spending by one-half and the defe nse spending by one-half. Another mission the DoD was reevaluating was the situation th at was developing in the Soviet Union. The increased spending on defense by the Reagan administration had repercussions in the Soviet Union. In an attempt to keep up with American advances in military technology, the Soviet Union was quickly creating a hazar dous national situation because of increased spending in their own military efforts to k eep up with the United States. BRAC was created for the purpose of assisting the DoD in determining what course of action would be feasible in reducing cost s and making defense spending more efficient. Dunbar (2000) conducted research concerning BRAC while attending the National Defense University and National War College. In Dunbar’s report, the past history conc erning Congress, the DoD, and base closures reflects similar ac tions reported in the previously mentioned references: Congress’ inability to trust the DoD’s recommendations for base closures and realignments. An interesting aspect of Dunbar’s resear ch was the DoD’s method of base closure before 1988. According to Dunbar, most base closures and realignments were conducted to the nuclear response from the Soviet Un ion. The bases were made larger and placed farther from larger Metropolitan Statistical Areas (MSAs) to preven t the annihilation of population centers in the event of a nu clear war. Unfortunately the economic development between an urban area and the m ilitary bases increased the ability of urban

PAGE 23

12 growth between the two locations. Many cities that were near a military-base may have been a small city or mid-sized city when realignment occurred thirty or forty years ago. Some examples are the two areas chosen for our study: Tampa and Jacksonville, Florida. According to the U.S. Census Bureau in 1970, the population for Duval County Florida (Jacksonville) was 528,865 and Hillsborough C ounty Florida (Greater Tampa) was 490,265. Referring to Rugg’s (1972) research on urban growth patterns and the economic development of MSAs, the two MSAs were in early stages of economic growth patterns that resulted in a megalopolis in the Tamp a area and Jacksonville’s MSA spilled over into surrounding counties (Clay, Nassau, and St. J ohn’s). The difference between the two areas is the location of the military bases. The 1980 Census reports the population for Duval County Florida as 571,003 and Hillsborough County as 646,960. At this time the Tampa-St. Petersburg MSA had become a small megalopolis, whereas the growth in Jacks onville shows a slight increase in the MSA’ growth. During this phase of Tampa’s growth, MacDill AFB could have developed a significant impact on Tampa’s economy. The population of Hillsborough County in 1990 was 834,054 and almost matched the population of Pi nellas County reported at 851,659. By 2000 the population of Tampa was 998,948 and Pinellas was 921,482, thus the population Hillsborough County’s ability to grow toward the east permitted the population to pass that of Pinellas County, which had almost grown to the full exte nt of the county’s borders. Although there is no direct connect ion between MacDill AFB and either St. Petersburg or Orlando, the significance of thei r interaction with Tampa and its’ economic base should be consistent with previous re gional studies concerning base closure impact. The significance of military bases being s ituated near major accessibility routes increases their nonbasic activ ity value to the economic development of the local

PAGE 24

13 community. For example, Orange Park in Clay County was closer to Jacksonville NAS and Mayport NS was situated near the b eaches area of Duval County. The growth between both bases and Jacksonville and the growth between the ba ses and the smaller communities developed into the Jacksonville MSA. Another factor that enhanced the growth involved the earlier BRAC rounds clos ing several smaller Naval bases in Florida and realignment to the two bases in Jacksonville. The GAO detailed the problems involved in preparing the land and facilities for exchange to the local communities in many of their reports (Holman 1997, 1998, 2001a, 2001c). Normally the average time for transfer of land can be between 3 to 7 years. Most of the time consuming factors involved envir onmental cleanup; however, parcels of base property are turned over when they are ready for use. The means of overcoming a base closure and keeping the economic impact at a minimum would rely on several factors: prepar ation of the base for closure, alternative methods of turning property over at a quick er pace, privatization in place, finding new environmental cleanup technology, and preparin g the local economies for transition to newer nonbasic activities to name a few. The mo st important aspect of preparing the local community for a base closure is explaining th at the economy is not going to be destroyed as most politicians would have them belie ve. Often Congressional members agreed that there may be a need to redirect defense spending from unnecessary costs to R&D and improved readiness of forces, unt il the closure or realignment of a base is situated within their congressional district. On learning which facilities and functions are revealed many congressional members begin grass root operatio ns to “save” the base from closure. The most common method of organizing the gras s root campaign is through an official website of a federal or stat e legislator or in some in stances the governor’s office.

PAGE 25

14 Another method of starting a grass root campaign is the website of a local university’s political science, journalism, or public relations department. Every state that has a military-base has a website that is design ed to prevent the closures of their bases by presenting different studies conducted by local universities, institutions, or research facilities. State or federal legislators sponsor most of the websites. For example, U.S. Senator Dianne Feinstein has a webpage that discussed practically any congressional concern for citizens of California. Interest ingly, Feinstein has a history of voting for defense budget cuts; however, her webpage (Fei nstein 2004) has many sites that argue against any base closures in California. The fact remains that the DoD has a need to determine the means to reduce costs and improve budget efficiency. Congress needs to unde rstand that some bases have little or no military value and are wasteful spending of tax dollars or if the base and its functions are that important to the urban re gion then increase defense spending. However, the purpose of BRAC is to save tax dollars in the first pl ace and increasing the defense budget to save bases and functions of bases is detrimental in more ways to the economy of the whole instead of a small part of the state or nation. The GAO (Holman 2001a) discussed in great detail the DoD’s budget, costs, and savings because of BRAC. The importance of Congress’s role in the proc ess is they give the final approvals of all BRAC commission decisions. Congress has been difficult in approving past BRAC recommendations. Congress denied the la st two rounds of DoD requests for BRAC commissions because of previous discrepanc ies. However, one of Congress’ concerns with BRAC was possible interference after the recommendations ha d been approved by any party and outside the re quirements established by the 1990 act. Schwalbe (2003) explained in explicit detail th e interference that occurred in 1995. President Clinton

PAGE 26

15 interfered with the approved closing of two bases, whic h clearly was prohibited by Congress. Although the interferen ce did create a new concept (p rivatization in place) of possibly protecting military bases in futu re BRAC rounds, the Congressional response was overwhelmingly strong against any futu re BRAC considerations. The political weapon of Congress has been the economic im pact that is created by base closures. Although the short-term effect can be disruptive and possibl y devastating to the urban areas that are affected by BRAC, studies have shown otherwise. The upcoming BRAC round in 2005 will be quite different than the previous four BRAC rounds. BRAC 2005 is designed to improve DoD budget spending; however, it will include some significant changes that make it different that the previous four BRAC rounds. Schwalbe (2003) characterized this new round of BRAC as the “Mother of all BRACs” and explained Secretary of Defense Donald Rumsfeld’s plans to cut the same amount of surplus that was cut in the BRAC round of 1988 to 1995 combined. The 2005 round will also cut at least 25 percent of the DoD’s remaining real estate. Based on Public Law 107-107, Section 3000, there are some important details that are required before the next BRAC round can occur. Befo re any recommendations can be made, DoD has to prepare a force-structure plan. The Secretary of Defense concerning any possible national security threats be tween 2005 and 2025 must base the plan on an assessment. DoD must then submit that assessment to Congress detailing the inventory of DoD’s infrastructure based on the force-structure plan. The greatest imp act for DoD and BRAC requests occurred with the law amending the Defense Base Closure and Realignment Act of 1990 significantly changing the selection cr iteria. DoD has been directed by Congress in the upcoming BRAC round to make military value the primary consideration for recommendation for BRAC action. DoD is to as sign the bases values as before, however,

PAGE 27

16 when making the recommendation the values are to be based on the following: preservation of training a nd staging areas; preservation of military installations throughout a diversity of climate and terrain areas in the U.S. for training purposes; high consideration of joint war fi ghting, training, and readiness; and high consideration for contingency, mobilization, and future total fo rce requirements at locations that support operations and training. The selection crit eria based on military value was not only changed, but also mandated as the primary factor for BRAC consideration, however, selection criteria is not only relegated to military value. Other selection criteria include the extent and timing of potential costs and savings for DoD, the economic impact on the local communities affected, the local communities being able to support additional infrastructure and forces, and factors concerning environmental costs for cleanup, restoration, and disposal. Ten years have passed since the la st BRAC round. Since 1995, Senator John McCain and Senator Carl Levin have sponsor ed congressional action for two new rounds of BRAC. In two Congressional Research Service Reports for Congress, Lockwood (1999, 2000) reported the results of previous BRAC actions up to that time. Lockwood also reported concerns of the DoD’s reque st for two more rounds of BRAC and the reasons given by the DoD for the new rounds and congressional action concerning the need for new BRAC Commissions. Lockwood emphasized similar results that were published by the GAO and CBO concerning the costs and savings of the DoD after the previous four rounds. Lockwood reported the estimated savings at $5.7 billion; however, the GAO reported (Holman 2001b) the estimated annual savings had increased $5.6 billion in 1999 to $6.1 billion by 2001. The GAO reported (Holman 2004) the estimate is now nearly $7

PAGE 28

17 billion in annual savings from the previous four BRAC rounds. The BRAC savings have been substantial according to the reports. However, BRAC was not the only action by the DoD designed to reduce defense spending. Th e Balanced Budget and Emergency Deficit Control Act of 1985 called for the defense budge t to be cut in ha lf, BRAC alone would not reduce the defense budget by half. Lockw ood (1999) discussed an understanding of the DoD’s requests for BRAC. Lockwood expl ained that Secretary of Defense William Cohen released the Quadrennial Defense Re view (QDR) in 1997, which simply reported a major review of the military’s strategies a nd capabilities. Accordi ng to Cohen’s review, the reduction in force had dras tically surpassed the reduction of infrastructure. There was a significant difference in Cohen’s percen tages: force structure was reduced by 33 percent; infrastructure was reduced by 21 per cent. Cohen’ s conclu sion was a request for two new rounds of BRAC in 1999 and 2001. Although Cohen’s arguments for two new rounds of BRAC were creditable, Congress was s till stinging from th e interference of the 1995 BRAC round and was in no hurry to appease the DoD’s requests. Cohen continued to emphasize the need for BRAC by declaring the significant savings from BRAC would achieve the balance between force structur e and infrastructure, thus supplying the necessary funding for future readiness of fo rce and weapons development to bring the modern military up to speed with the military mission. The 105th Congress neglected to give Cohen se rious consideration because of their concerns of political and economic fallout w ithin their own distri cts. There are some instances that BRAC can be difficult for smaller communities to overcome. Fort Riley Kansas is a major Army base in the middle of the plains in Kansas. The city that would be impacted economically from Fort Riley’s closure if it were to occur is Manhattan Kansas, a small city by most standards w ith a population of 44,831 in 2000 (US Census,

PAGE 29

18 2000). Manhattan Kansas’ economic activity is predominantly agricultural; however, it is home to Kansas State University and Fort Riley. The university and military installation are the two largest nonbasic activ ities within the small city. Farming is a way of life for most of the people in the area surrounding Manhattan, but th e agricultural industry relies heavily on weather and climate. If the base were to close and seve ral years of drought occurred after the closing, the economic imp act could possibly destroy the small city. This is one reason Congress has a legitimate ar gument concerning BRAC. The fact is that most military installations are located near areas, cities or metropolitan areas that have a larger, diverse economy. Finally, recent global events have given Congress anothe r argument against future BRAC actions. Since September 11, 2001, the military mission and goals have been enhanced with the War on Terror and Opera tion Iraqi Freedom. Although there has been a great deal of bipartisanship concerning Am erica’s military actions in both operations, a majority of congressional members have used the military actions as an argument to prevent further BRAC until the actions have been resolved. A GAO report (Holman 1998) emphasized the importance of reduc tion of operations and maintenance of infrastructure (O&M) if the DoD is to meet the required expenses to modernize the force structure. If the infrastructu re costs are not met, then di verting the funds required for force modernization to O&M jeopardizes th e overall goals of the DoD. Another GAO report (Holman 1998) discussed the Quadrenni al Defense Review and the expected savings from personnel reductions might not be achieved. According to this report an expected $3.7 billion would be saved by 2003 if forces were reduced by 175,000 personnel.

PAGE 30

19 At the time of these reports it is po ssible that no one knew of the future consequences that occurred on September 11, 2001. Since that fateful day our forces have been stretched to a breaking point. Schwal be (2003) explained the Department of Defense’s position to counter the requiremen ts of meeting the demands of today’s military. The DoD will assign some bases to inactive status that are selected for BRAC. Once a base is relinquished to the private or civilian sector it cannot serve a military purpose or be applied a military function. Because of the possibility of future need the DoD will try to retain some of the properties in an inactive status similar to the status assigned naval vessels when they are deco mmissioned, but may be needed for future missions. If the DoD reduced the bases to m eet the same percentage of personnel and then a surge in personnel occurred to meet the military requirements for the War on Terror, there may not be enough bases to house the rise in force structure. Other methods could be considered fo r studying BRAC. Geographical Information Systems (GIS) have advanced to a high level in the past twenty years. Particularly, geospatial and imagery analysis have gr own more advanced and many avenues for research can be utilized to improve possible necessary scenarios that could create means of preventing or accepting BRAC recommendations in the future. The best use of GIS is the planning for land use of BRAC recommendations before the bases being recommended. If each urban area that has a base were to plan for the closure or realignment of the base, then planning for th e reduction of economic impact could negate the effect or at least lessen the economic impact. Concern over Economic Impact Gentrification plays a role in urban and economic growt h. A recent trend seen in many MSAs is a revitalization of Central Busi ness Districts. The ge ntrification consists

PAGE 31

20 of older buildings and structures being re novated or destroyed a nd new buildings and structures being built in place of the old a nd the environmental cleanup of parks or green spaces within the CBDs. Gentrification permitted large metropolitan areas to not only reclaim areas within the CBD, but also to renew economic activity within the CBD that was viable with the economic growth and activ ities that were continuing to spread from the city’s center. A Thesis presented to Virginia Polytec hnic Institute by Hogan (1997) discussed the negative economic impact on local communitie s that experienced BRAC actions. Hogan argued that the interferen ce from Clinton in 1995 set precedence for future BRAC considerations: privatization in place. The argument for priv atization in place has merit because it saves the DoD money in several wa ys without disrupting local economies. The method would allow portions of a base or functio ns of a base to be maintained by private industry and remove the operating cost from the DoD and place the burden of cost on private industry. A strong argument for base cl osures is made using the data reported by the Business Executives for National Secu rity (BENS) Hill Advisory (1997) on employment figures from base closures comp ared to the total U.S. employment and the job loss of Fortune 500 companies: appr oximately 120,000 jobs from four BRAC and over 250,000 from Fortune 500 companies in the first six months of 1996. Alleviating Negative Economic Impact with Accessibility Rugg (1972) explained the phenomenon as the “multiplier effect” using previous research conducted by Hoyt (1961). Utilizing the economic base theory and defining the military-base as a nonbasic activity of th e Tampa MSA, MacDill AFB had become a normal supplement to the new basic activity of Tampa MSA. The coalescence of Tampa and St. Petersburg led to newer transportation and accessibility routes over the bodies of

PAGE 32

21 water that separated the two MSAs. The new r outes were also located closer to MacDill AFB. The newer roads and the urban growth fed one another and in 1990 the growth of Tampa-St. Petersburg continued until 2000 when the megalopolis of Tampa-St. Petersburg-Orlando could be seen forming along Interstate 4 (I-4). Tampa MSA has accessibility to other areas other than the Ce ntral Florida Megalopolis. Two Interstate highway systems are significant to Tampa: I-75 and I-4. Interstate 75 gives Tampa accessibility to areas on a north-south axis. Inte rstate 4 is the highway that connects the Central Florida Megalopolis. Both interstates connect to other interstates (I-10 and I-95), which increases the basic economic activity fo r the Tampa MSA. Jacksonville is not as fortunate as Tampa because its economic ba se is relegated to greater distances. Jacksonville’s basic acti vities require accessibility to MSAs at greater distances. Atlanta, Georgia; Tallahassee, Florid a; Savannah, Georgia; the ci ties along Florida’s Atlantic seaboard; and eventually the Central Florid a MSA provide infrastructural exchanges to Jacksonville’s basic ac tivities. The nonbasic activities related to military bases in Jacksonville play an important role in the city’s economic growth and development. The presence of military personnel and bases provide an economic stimulus in terms of sales and services supported by military rela ted transfer income and consumption. Another aspect with the military bases is their location within the Jacksonville MSA. Both bases lie on important transportatio n routes into the hear t of the Jacksonville MSA. The urban growth pattern in Jacksonvi lle has shown a tendency for growth toward the bases rather than toward the major tran sportation routes leading to Savannah, Georgia (I-95 to the North); Atlanta, Georgia (I-10) ; and Tallahassee, Florida (I-10). Industrial growth patterns appear to follow the major transportation routes; however, residential and commercial growth appears to follow the rout es toward the bases. The relative locations

PAGE 33

22 of the CBD to the bases are ke y to certain urban growth pa tterns. The need to study the economic base of the region that immediat ely surrounds the CBD and the base and the major routes that connect the two is an area that should be considered when studying the economic impact of BRAC. This area will pote ntially absorb the greatest impact after BRAC actions. Presence of a military-base is a nonbasic function and transfer of payments are of great importance to the loca l economy, helping to boost the tax base and increase the commercial and residential activity in the area. The closure of a base doesn’t necessarily impact the MSA as a whole, but as the RAND study (Dardia, et.al. 1996) of three bases in California showed us, the impact on the local communities and areas juxtaposed to the base are profound.

PAGE 34

23 CHAPTER 2 METHODS Economic Growth and Development beca use of Military Bases and the Central Business District: Economic Ba se Theory and Accessibility Military Bases, Central Place Theory, and the Central Business District (CBD) Walter Christaller conducted a great deal of research revealing the relationship between an urban center and economic growth. Central Place Theory (Christaller, 1933) suggested that a city tends to decen tralize as its economic base grows. Nonetheless, a disproportional amount of growth was associat ed with more central locations. The more recent literature (Rostow 1960; Rugg 1972; Palm 1981; Forkenbrock 1990; O’ Sullivan 1993; Boarnet 1996; Wu 1998; Vickerman et.a l. 1999; Banister and Berechman 2000; Nelson and Moody 2000; Berechman 2001; et.al.) hi ghlighted other factors that influence urban growth patterns and the ro le of commercial, financial, and residential districts in the centralization of growth along prominent nodes or corr idors. Military expenditures and income also played a vital role in the gr owth pattern of an urba n center. This is the argument that is normally presented by congressional members when arguments for selected base closures are recommended. Since Christaller’s research concerning Central Place Theory (1933) was published, most research conducted con cerning urban planning and grow th place a great deal of emphasis on the economic indicators within an urban area, its surrounding region, and the nearby trade centers. The significance of the economic impact of industry loss and gain may determine the scale in which an urban cen ter may grow, the rate of growth, and the pattern of growth. Moreover, employment mi x and a city’s function or economic profile

PAGE 35

24 will also play a role in determining a city’s ab ility to weather adjustments from the loss of various industries. Economic Base Theory The spatial relationship of the urban and regional economy can affect urban growth pattern. Economic base theory is one method of explaining the impact of industry loss or gain. According to Yeates and Garner (1976) economic activity may be expressed by the following equation: Total activity in the city = Total in basic activity + Total in nonbasic activity. (TA) = (BA) + (NBA) Total income of the city = Total inco me derived from basic activities + Total income derived from nonbasic activities. The basic-nonbasic ratio (BA/NBA) represents the ability of basi c (export-oriented) activity to support nonbasic ac tivities such as retail sale s, consumer and producer services, etc. Nonbasic activiti es are also supported by transfer payments and transfer income (government expenditures on military ba ses fall into this category). According to Fik (2000) the basic-nonbasic ratio creates a mu ltiplier, which is applied to determine the impact of transfer income on employment expressed: Total Economic Activity = Basic Economic Activity + Nonbasic Economic Activity Basic Economic Activity Basic Economic Activity Basic Economic Activity (TE / BE) = (BE / BE) + (NBE / BE) Whereas TE / BE is the multiplier and is represented with the symbol m the new equation is: m = 1 + (NBE / BE) In addition to the multiplier, Fik explains 4 different types of income flows or transactions, which have an effect on the econom ic growth trends on an area. The type of

PAGE 36

25 income flow that typically concerns the im pact a military-base has on the local economy is transfer of income. A visual representati on of income flows is given in Figure 2-1. The importance of this concept can be relative to military-base closures because transfer payments to the base support i ndustry at the metropolitan level. Figure 2-1. Income flows. Methods for a New Approach to Predic ting Economic Impact from BRAC Military Bases Applied to Economic Models Military bases are a nonbasi c function of urban centers and supply a transfer of income from the federal government as well as from the military labor force. Federal funding for military bases includes purchasi ng necessary goods and supplies from the local urban economy. Since most urban growth models are based on economic foundations, spatial characteristics and classifications, involve so cial and political factors,

PAGE 37

26 and must rely on transportation networks; a st rong argument can be given to reflect the possibility of the military factors in r ecent years having an influence on economic growth. Although the military impact has not b een commonly used in previous research of urban growth and planning or in economic models, military bases are usually lumped in with one or more land use categories (mos t often institutional) or certain labor force categories such as governmental employment. However with recent controversial issues concerning base closures, studi es are becoming more recogni zant of the economic impact an urban center may face on a base targeted fo r closure. Risa Palm (1981) explained in great detail interesting points of the defense spending in previous years and the impact on regional urban economic growth in the Sout h and West regions of the United States. Palm also referred to other research c onducted concerning federal defense spending (Rostow 1960; Sale 1975; Perry and Watkin s 1977; Weinstein and Firestine 1978) and the economic development of the regions in which the defense spending was concerned. The importance of the research may be the foundation of current de fense spending trends in the same regions, and the greater impact it now has with higher military salaries. Several factors in recent years also contribute the growing influence military bases have in the nonbasic function of an urban cen ter: increases in military pay, emphasis on research and development, support services, ci vilian labor force re quirements, and import of perishable goods are just a few. In r ecent years the military pay increases have normally been higher than the cost of liv ing increases in the South, Southwest and Midwest regions of the United States. Because of the restructuring of the military’s training operations and equipment (TO&E) has changed significantly since the Cold War, many of the urban centers that contained base s have industries that are reliant on federal contracts to supply goods and services requi red for the fulfillment of the military

PAGE 38

27 obligation to the United States A concern of many urban cent ers is the closure of bases can relegate those industries to closure as well. The economic impact on the urban center can be devastating to the urba n economy if there is a signifi cant reliability of the urban center’s industry on the nonbasi c function of the base. However, according to Economic Base Theory (Rugg 1972; Yeates and Garner 1976; Christian and Harper 1982; Mayer and Hayes 1983; Hartshorn 1992; Fik 2000; et.a l.) an urban center with a diverse economic base should overcome the loss of one type of industry given the depth and breadth of its other industrial linkages. Ther efore, the argument of many DoD officials is the restructuring of land use av ailable after a base closure sh ould be utilized efficiently to compliment and/or enhance existing basic-nonb asic functions of the urban center. The opposing view is the time it takes to replace th e lost source of revenue can be difficult or impossible to overcome in the short term and l ead to the eventual destruction of certain sectors of the urban center in the long term. Urban transportation systems play a significant role to military bases. Highway systems are vital to military bases. Mobility of military labor force is vital to location of the base. Most employees of the military mu st be on the job earlier than the civilian sector and thus having a reliab le transportation network to tr avel to work is necessary. Nelson and Moody (2000), Kim and Chung (2001) and Kim et al. ( 2003) discussed the importance of transportation corridors and their effects on urban growth models. The three study areas chosen for our study have major transportation corridors that connect the military bases to other nodes including th e CBD. Another important feature of the transportation corridors is the availability to the residential areas to and from the military bases. The access to transporta tion corridors outside the locali zed area may also allow for some military members to live outside the study areas. Availability of transportation

PAGE 39

28 corridors throughout the MSA may account for the spillover s een in previous research (Dardia, et.al. 1996; Hooker and Knetter, 2001) concerning economic impact from base closures. The land use classi fication given by St. John’s Wa ter Management District (Jacksonville) for the years 1973 through 2000 s how significant changes in residential and commercial growth (predominantly an in crease in both) near the major thoroughfares and beltways near the military bases. Whereas the lack of area for growth in the MacDill AFB study area allows growth of commercial ar ea at the expense of residential area and vice-versa. According to Christian and Ha rper (1982) agglomeration economies are enhanced from beltways that are proximal to railroads, airports, a nd seaports as they allow for the clustering of linked industries Both of these citie s have specialized transportation functions because trade is a major activity in both urban centers. Christian and Harper (1982) explained the input-output model as a method of forecasting manufacturing (i ndustrial) forces on economic growth. Mayer and Hayes (1983) basically described the model as a series of input-output matrices and can be used to account for all sector inputs and outputs of a city. In the case of a military-base, the inputs and outputs could be income or revenue. Previous Economic Impact Studies from Military-base Closures The RAND study (Dardia et al. 1996) was mentioned the most in the research conducted concerning BRAC and was based on thr ee California bases that were closed in the early rounds of BRAC from 1988 to 1994. The study used three benchmarks to gauge the changes of economic impact. The three be nchmarks are: (1) expert projections of what would take place in each community, (2 ) the experience of a matched set of California bases that had not closed, and (3 ) the experience in th e broader regions in which the closed bases were located. The study had mixed results based on the variables

PAGE 40

29 that were selected. The vari ables that were used incl uded: population, housing units, vacancy rate, unemployment, labor force, K-12 school enrollment, and retail sales. The variables considered the change s in each after the bases clos ed and the region’s proximity to the bases. The study revealed the gloom -and-doom scenario as being an extreme prediction. Initial results showed that ove rcoming the negative effects were not as difficult as had been predicted. The study also revealed that spillover into other areas from BRAC was negligible as distance from the base increased. According to the RAND study, the impact affected the unemployed worker s and their families and the revenue lost by individual businesses more than the co mmunity as a whole. This allowed the community to overcome the impact through pr oper planning and land use. If there is good indication that the economic climate in th e region is favorable, then impact from BRAC will be swift. Smaller and less diverse economies will require substantial longer recovery times. Thus it stands to reason that the growth and development patterns of a region be studied when a base is targeted for BRAC action. The conclusion given in the RAND study simply stated that predicting th e effects of economic impact are difficult. The RAND study mentioned distance, but the study did not use distance as a variable. An example of vari ables that could explain dist ance in a spatial relationship between the bases and commercial nodal activit y are the straight-lin e distances between the bases and other nodes or the CBDs. Accessi bility of the base to the CBD and other nodes is also vital to understa nd the economic impact of BRAC. US Senator Dianne Feinstein’s webpage had the text of a letter (2004) she sent to Peter Potochney, the director of BRAC. She emphasized three issues need to be added to the military value section of the BRAC crite ria. The second issue she outlined includes accessibility considerations. However, Senator Feinstein did not include the proximity to

PAGE 41

30 the CBD an important node in most urban economies. Another as pect that Senator Feinstein failed to mention is the populat ion of California. Ca lifornia has roughly onesixth of the United States total population. Thus it would be easy to pr esume that if there are more bases in California and more pe ople, then California should have a higher proportion of the numbers in base closur es and personnel unemployed. Feinstein’s arguments are lost to the BRAC Commission b ecause previous studies (Dardia, et.al. 1996; Hooker and Knetter, 2001) did not suppor t her. Senator Feinstein did have one advantage with her argument: mo st of the bases closed in pr evious rounds have left very few choices for the BRAC Co mmission in future rounds. Certain factors determine the economic recovery from BRAC actions. The GAO explained the factors in deta il (Holman 2001c) and supplied a visual representation of those factors. According to the GAO, eight factors significantly affect economic recovery: (1) reuse of base property, (2) government assistan ce, (3) public confidence, (4) leadership and teamwork, (5) natural and la bor resources, (6) diversified local economy, (7) regional economic trends, and (8) nationa l economic conditions. Th e last two factors listed possibly played the strongest role s in previous BRAC actions and economic recovery for those areas. The important issu e to keep in mind with the next round of BRAC is that national recovery is not the only factor that needs to be studied for the economic recovery of BRAC actions. Hooker and Knetter (2001) studied th e economic effect of employment and personal income effects that occurred from BRAC. They explained the importance of the reduction in defense spending from 1986 to 1998 and the need for base closure; however, they also mentioned the difficulties in decidi ng the bases selected for closure. The most singular factor, as mentioned in the previous studies, involved in the fight against BRAC

PAGE 42

31 is the economic effect on the local comm unity. Hooker and Knetter approached the problem using a newly constructed dataset to study the employment and personal income effects at the county level. They explained that military bases ar e a major employer of most counties in which the base is located (up to 30% in some cases according to the study); and thereby accounted fo r a larger share of income and tax revenue in the area. Hence, transfer payment, rela ted income played a vital role in these local community’s economic stability. Government transfer paymen ts helped pay for certain activities within the local community, especially in the operation and maintenance of the local communities infrastructure and services. Normally local taxes are utilized in education, road repairs, community development, mainte nance of parks, etc. The importance of the transfer payment differs in one major asp ect; the federal government (in the form of property tax, sales tax, and othe r taxes that are owed) pays the taxes for the operation and maintenance of the base. Normally the taxe s collected by the civilian sector (sales, property, and in most states a state income tax) pay for the operation, services, and maintenance of the local community. Commun ities with a military-b ase enjoy the luxury of receiving additional tax s upport from the federal government; thus they attempt to overcome the decision for BRAC to occur in their communities. Hooker and Knetter mentioned an interest ing point concerning BRAC and the local community: the opportunity cost of the resources the base a ffords the community after it is closed. In particular, they cited the av ailable land the community received and the possible use of the land after it was released to the community as the most important consideration. Hooker and Knetter gave two ex amples of scenarios that can assist the recovery of local economies af ter the bases closed (in a st udy of the Presidio of San Francisco Army Base and Moffett Field Nava l Air Station in the Silicon Valley, both

PAGE 43

32 bases in California). Primarily using employme nt and personal income indicators in their study and measuring the responses from th e counties by comparing the results to counterfactual scenarios. The first scenario assumed the county’s employment and per capita personal income growth rate equal to the state’s growth rate The second scenario assumed the difference between the county and th e state’s growth rate in the years before the time of base closure would have persisted after the base closed. In the case of their study they chose to use a two-year period be fore base closure to measure growth. The results of their research showed that nonbase employment grew faster in closure counties than it did in the counterfactual model. Th e study proved that spillover from job loss on bases did not affect the surrounding areas as assumed from the impact analysis. Instead, the study proved that if the base s’ resources are properly used in alternative ways, then an increase of job creation coul d occur if industries with higher multiplier effects are brought in to substitute for jobs lost under the base closure. Hooker and Knetter explained the findings are similar to recent st udies that were conducted by Davis et al. (1996) based on the dynamics of labor mark ets in larger regions involving basic industries. Hooker and Knetter (2001) also found similar results to Aschauer’s (1990) “below-unity estimates of output multipliers for government consumption and military investments from aggregate data.” The personal income results from Hooker and Knetter’s study also revealed very little im pact from BRAC. According to Hooker and Knetter there were no statistically significan t impact on per capita income from BRAC. Furthermore, the study indicated a slight grow th in per capita personal income in the county compared to the state’s growth after BRAC. Hooker and Knetter gave two explanations for the results: (1) genera lly outgoing military personnel have belowaverage income in comparison to income of employees working in other sectors and (2)

PAGE 44

33 the older and more experienced civilians who lose jobs tend to gain employment at higher salaries. According to Hooker and Knetter, economic impacts have traditionally been projected instead of estimated and measure d. They argued that projections from inputoutput models tend to ignore th e capacity of regional economies to adjust to closures. They further argued that the main issue to measuring economic impacts was the estimation of impacts that would occur without base closure. The type of base is another factor Hooker and Knetter mentioned that tends to assist the comm unity in recovering from base closure. Bases that require more highly skilled workers, utilize more methods of transportation for shipping and receiving of supplies, personnel, and equipment, and provide more resources for future developm ent tend to assist the local community’s recovery after they are closed. Air Force base s normally fit the criteria described and in past BRAC actions were usually the majority of the larger bases selected. Hooker and Knetter concluded that future studies should at tempt to assess obscure d results instead of simply project results of economic impact They also emphasized the importance of refuting the negative impact predicted and concentrate on establishing the positive aspects that can occur with the proper and well-developed planning of the use of the base’s resources after closure. The preceding research suggested several main issues for further study of base closure impact: (1) do not predict, estimat e the impact on economic growth rates from base closures, (2) include distance variables and accessibility to transportation corridors, (3) consider the percentage of total population that is employ ed at the base in question, (4) research should consider using a smalle r geographical scale to estimate impact at local levels, (5) include a larger number of social and economic inde pendent variables to increase the variability and random pattern of the model, (6) and apply the criteria for

PAGE 45

34 base closure as defined by the Department of Defense (DoD). Ou r study considered the previously aforementioned research and was c onsistent with similar methods used in the previous research. However, the six issues discussed were included in our study to determine the base’s negligible impact on the economic growth rates at a smaller geographical scale (census-tract level), whereas previous resear ch is conducted at a larger geographical scale (county, regional, or greater). New Approach to Predicting Economic Impact from BRAC Base Realignment and Closure (BRAC) 2005 and Florida's Bases The next round of BRAC was scheduled for January 2005. Since the next round was announced, many politicians from the federa l to the local level have organized to prevent BRAC from occurring in their state or local community. Florida is the home of many military bases. Past BRAC rounds have resulted in the closure of several installations in Florida; howe ver, the upcoming round has affected the state more than it has in past rounds. Governor Jeb Bush has already authorized th e state to organize grassroots activities to raise over $200 million dollars to fight BRAC in Florida. The main reason for the governor’s action may be b ecause of statements given by Secretary of Defense Rumsfeld concerning the relocati on of Central Command from MacDill AFB in Tampa, Florida to the Middle East and possible closure of MacDill AFB. Although MacDill AFB has been the base most often i nvolved in the rumors of BRAC, Florida is concerned about the possibility of other base closures. The largest bases in Florida are Pensacola Naval Air Station (NAS), Eglin AFB, Mayport Naval Station (NS), Jacksonville NAS, MacDill AFB, Patrick AF B, and Key West NAS. Other bases in Florida that do not have a large permanen t military personnel presence include Camp Blanding and Avon Park Bombing Range becau se of their mission as training bases.

PAGE 46

35 Jacksonville and Tampa, Florida Jacksonville and Tampa were chosen fo r our study because they are the largest MSAs with bases in Florida. The economic ac tivities in the MSAs, the size of the CBDs, the transportation corridor s between the bases and CBD from the surrounding communities and the diversity of factors invol ved if bases are chosen for closure allow for the development of mode ling the estimates for economic impact. The bases chosen for the study included MacDill AFB in Tampa and Jacksonville NAS (Jax NAS) and Mayport NS in Jacksonville. An important f actor concerning our study compared to past studies involved the area being studied. Instead of incorpora ting an entire MSA, county or region, our study emphasized the censustracts that surround and connect the transportation corridors between the CBD a nd base. The previously mentioned studies practically stated that spill over from base closures was ne gligible. Instead of focusing on the projected impact of a potential base closure, our study assessed if there was a discernible statistical relationship between distance to a military-base and urban economic growth rates taking into account th e locational accessibility to the CBD and other prominent urban nodes. According to the Jacksonville Chambe r of Commerce the total number of employees at Jax NAS was 24,648 and Mayport NS was 15,001 in the year 2000. However, when aggregating the total number of military members and federal government employees in the census-tracts th at were selected for the study areas the numbers were smaller. In 1980, Jax NAS ha d 15,484 military members out of a total population of 249,362 that resided in the st udy area and by 2000 the military population had declined to 5,032 out of a total po pulation of 303,909. The federal government employees that resided in the study area for Jax NAS (south central Jacksonville) in 1980

PAGE 47

36 were 6,500 and in 2000 the number of federal government employees was 6,802. An explanation that could account for the cha nge in military members was the addition of homes on the base itself and the improved pay military members experienced over the last twenty years. However, the number of civilian employees increased. The closure of Cecil Field NAS in west Duval County ma y account for the increase in civilian employees near Jax NAS. One other factor could account for the reduction of military members living near Jax NAS: Jax NAS supplie s the aviation units for the carrier group stationed at Mayport NS on the mouth of the St John’s River and Atla ntic Ocean in east Jacksonville. Mayport NS study area had a military population of 11,541 out of a total population of 197,768 in 1980 and 9,190 out of a total population of 272,590 in 2000. The reduction in population thus rejects the th eory that personnel had migrated from Jax NAS. The only other conclusions could be th e growth of military housing on the bases, which may not be included in the census or the increase of average salary allowed military members to move further from the bases. Finally, many service members may have been out to sea when the census wa s being taken. Jax NAS and Mayport NS have been a part of Jacksonville for the most part of last century (at least since World War II). During the past thirty years, the requir ed number of service members has been approximately 20,000 for Jax NAS and 15,000 for Mayport NS. Additional Federal employees for the Mayport NS study ar ea numbered 3,960 in 1980; the population was 4,802 in 2000. Basically an increase was observed in civilian employment by the federal government while a reduction was observed in military member population. Of greater importance is the fact that military population has been reduced in the local area of both study areas while there has been significant economic and urban growth. However, the

PAGE 48

37 significant growth in civilian employees work ing on the base in que stion may have an adverse effect from BRAC. The significance of the civilian and military population of persons working on the base is approximately one percent of each study area’s total population. The small percentage of the total population working on the bases in each study area suggested that the impact on the economic growth rates from the base may prove to be negligible. Jacksonville NAS and Mayport NS Study Areas and Diversity of Industrial Employment: 1980 and 2000 Industry employment data gathered from the 1980 and 2000 census showed the overall changes in employment profiles with the study regions. The categories of employment were broken down into thirteen employment groups or sectors for each of the three study areas (Tables 2-1, 2-2, 23). The employment figures for 1980 and 2000 for the residents of Jax NAS are seen in Table 2.1. The employment of residents in the Ma yport NS study area was compared using the same categories that were used for Jax NAS. The employment populations for 1980 and 2000 for the residents of Mayport NS are shown in Table 2-2. The Mayport study area was similar to Jax NAS because there were a majority of increases in employment in most of the cat egories; however, the increases were not as significant as they were in the Jax NAS st udy area. The reduction in military residents was not as significant in the Mayport study ar ea. Over twenty y ears the reduction was approximately 2,000 residents in the Mayport study area, where Jax NAS saw a reduction of over 10,000 military residents. One other fa ctor that was not considered for the reduction of military personnel was the DoD’s reduction in force since 1986. The reduction in force could be a significant factor, however, according to the Jacksonville

PAGE 49

38 Chamber of Commerce, Jax NAS has over 24,000 employees in 2001 and approximately 6,000 of those employees were civilian employ ees. The tendency of growth employment from the basic activities in the Jax NAS study area indicated that ba se closure should not have a negative economic impact accor ding to the study conducted by Hooker and Knetter (2001) and percentage of the study area’s total population employed by the military. However, the DoD may decide that the mission of the base prevents the selection of the base for BRAC. Jax NAS s upplies the air support for the carrier group that is stationed at Mayport. More impor tantly, Mayport is the only port other than Norfolk, Virginia on the eastern seaboard that is home to carrier groups, thus meeting vital criteria for the mission of the Department of the Navy and the defense of the Eastern United States. However, if relocation of th e carrier group from Ma yport to another base with a similar mission occurs; then it is al most assured that Jax NAS will be closed. MacDill AFB and Tampa, Florida The same industrial employment variables us ed in Jacksonville were also used in Tampa. Armed Forces personnel that lived in the study area in 1980 were 10,624 out of a total population of 249,646 and in 2000 were 2,172 out of a total population of 263,580. The number of federal employees that re sided in the study area numbered 3,651 in 1980 and 3,091 in 2000. Again, explaining the significant reduction in military personnel in the study area may involve several factors, which are not known, but may have possible causes. The important issue for our study was the significant reduction of military personnel residing in the study areas. Since the reductions of military personnel have occurred, then BRAC should not have as much of a significant impact from the loss of military salaries. Unlike the Jacksonville st udy areas, the number of federal employees also decline.

PAGE 50

39 Table 2-1. Employment populations for Jack sonville Naval Air Station (Jax NAS) study area Economic Activity Employment 1980 Emplo y ment 2000 Difference Percentage Difference Transaction Type Agricultural, Forestry, Fishing, and Mining 1,153 569 -584 -50.65 BE Construction 6,300 9,827 3,527 55.98 BE Manufacturing 10,203 9,222 -981 -9.61 BE Transportation, Communicatio ns, and Public Utilities 9,098 14,490 5,392 59.27 NBE Wholesale Trade 5,228 5,604 376 7.19 NBE Retail Trade 17,233 17,936 703 4.08 NBE Financial, Insurance, and Real Estate 10,338 17,725 7,387 71.45 NBE Business and Repair Services 5,073 14,695 9,622 189.67 NBE Personal Services, Entertainment, and Recreationa 5,465 11,265 5,800 106.13 NBE Health Services 8,004 15,203 7,199 89.94 NBE Education Services 7,383 8,526 1,143 15.48 NBE Other Professional Services 5,097 6,514 1,417 27.80 NBE Public Administrationb 7,444 7,618 174 2.34 NBE Armed Forces 15,484 5,032 -10,452 -67.50 NBE a Recreation is both a basic and nonbasic activity. b Includes civilian employees on military base. According to past studies, especially Hooker and Knetter (2001), if the employment in the area is higher than the re gional or state growth rates, then base closures should not have a significant negative impact on the lo cal community. Reductions in the number of military personnel residing in the study areas have been noted, and the changes in federal employees have seen increases in J acksonville and reductions in Tampa.

PAGE 51

40 Table 2-2. Employment popula tions for Mayport NS study area Economic Activity Employment 1980 Emplo y ment 2000 Difference Percentage Difference Transaction Type Agricultural, Forestry, Fishing, and Mining 937 513 -424 -45.3% BE Construction 5,881 8,982 3,101 52.7% BE Manufacturing 8,326 7,838 -488 -5.9% BE Transportation, Communication s, and Public Utilities 8,171 13,487 5,316 65.1% NBE Wholesale Trade 4,271 4,804 533 12.5% NBE Retail Trade 15,910 15,905 -5 0.0% NBE Financial, Insurance, and Real Estate 8,482 16,730 8,248 97.2% NBE Business and Repair Services 4,205 14,107 9,902 235.5% NBE Personal Services, Entertainment, and Recreationa 4,701 11,281 6,580 140.0% NBE Health Services 5,377 12,689 7,312 136.0% NBE Education Services 6,366 8,342 1,976 31.0% NBE Other Professional Services 4,240 6,126 1,886 44.5% NBE Public Administrationb 5,281 6,211 930 17.6% NBE Armed Forces 11,451 9,190 -2,261 -19.75 NBE a Recreation is both a basic and nonbasic activity. b Includes civilian employees on military base. However, federal employment was only one factor in determining economic impact. Establishing an overall view of em ployment in the study area must be achieved as previously done in the Jacksonville area. The employment data by sector for 1980 and 2000 for the residents of Tampa are shown in Table 2.3. The changes in military personnel that re side in the study area are similar to Jacksonville. However, unlike J acksonville the percentage of the total population in the study area employed at the military-base was significantly smaller (less than 0.5%).

PAGE 52

41 Although the percentage of th e total population in the study area of military employees was significantly small, the impact on economic growth rates from base closure should be consistent with the percentage of military employees seen in the total population of the study area (a negative impact of less than 1% on the economic growth rates in the study area). There was a marked and noticeable difference between employment changes in Jacksonville and those in Tampa. The categorie s that show an incr ease in employment did not show a dramatic increase, while there was a substantial decrease in military residents in the area and the decreases in em ployment in the categor ies was greater than those categories with increases. Consider ing the study of Hooker and Knetter, the reduction in employment may enhance the pos sibility of improvement of industry and employment in those industries with base cl osure. First, the base in Tampa is an Air Force base and is associated with more extensive resources an d skilled employment. Second, the lack of space for growth proxima l to MacDill AFB creat es possible scenarios for further growth with available space create d with base closure. Finally, the location of the base is prime real estate with obvious advantages in terms of transportation and residential growth. Furthermore, the location ha s several notable qualities: (1) the base is accessible to the CBD by several major roads, (2) it covers about one-fourth to one-third of the lower portion of a penins ula, thus it is acces sible to sea transportation, (3) it is home to a large airfield, theref ore it is accessible to air transp ortation, (4) infrastructure is in place to support all of the tr ansportation routes, and (5) prim e real estate for residential development. Given the characteristics, MacD ill might be construed as a prime candidate for BRAC selection. There is on e other factor that the Do D may consider for MacDill

PAGE 53

42 AFB and that is encroachment. According to census data and reports, the Tampa area is one of the fastest growing MSAs in the United States. Table 2-3. Employment populatio ns for MacDill AFB study area. Economic Activity Employment 1980 Emplo y ment 2000 Difference Percentage Difference Transaction Type Agricultural, Forestry, Fishing, and Mining 1,825 320 -1,505 -82.5% BE Construction 6,141 7,593 1,452 23.6% BE Manufacturing 13,078 9,607 -3,471 -26.5% BE Transportation, Communicatio ns, and Public Utilities 8,675 11,811 3,136 36.1% NBE Wholesale Trade 6,025 5,291 -734 -12.2% NBE Retail Trade 19,929 14,921 -5,008 -25.1% NBE Financial, Insurance, and Real Estate 9,007 14,981 5,974 66.3% NBE Business and Repair Services 5,975 19,124 13,149 220.1% NBE Personal Services, Entertainment, and Recreationa 6,829 11,780 4,951 72.5% NBE Health Services 7,847 13,009 5,162 65.8% NBE Education Services 7,247 8,281 1,034 14.3% NBE Other Professional Services 5,236 5,827 591 11.3% NBE Public Administrationb 4,956 5,087 131 2.6% NBE Armed Forces 10,624 2,172 -8,452 -79.56 NBE a Recreation is both a basic and nonbasic activity. b Includes civilian employees on military base. The growth of Tampa allows the DoD to consider the possibility of the growth to enhance the economic recovery from base clos ure and the fact that Tampa’s growth has encroached on the AFB in the last twenty year s causes problems because of the air traffic

PAGE 54

43 from the base. The best possible solution for bot h sides is to consider how to assist in proactive planning in the event MacDill AFB is closed. Recovering from Economic Impact Assuming Tampa or Jacksonville Bases Are Selected for BRAC 2005 Undoubtedly, the local community’s economy will be affected because of BRAC. Proper planning for BRAC should be started im mediately to assist the recovery from economic impact because of BRAC. Hardest hit might be the areas juxtaposed to the base (in the short run), if there are economic spi llovers that are high ly localized. Our study examined the extent to which proximity to a base affects local economic growth rates and the degree to which variability in urban growth is explained by distance to a base (which accounts for the locational accessibility to other prominent nodes within the urban economy). Software Used for Our Study The dependent and independent variable s were taken or created from 1980, 1990, and 2000 US Census Bureau STF3A Files us ing Microsoft Access and Microsoft Excel programs. The 1980 STF3A Files were distribut ed in text format and Microsoft Access was used to create the 1980 database of so cial and economic (independent) variables and dependent variables used for our study. Once the 1980 database was created, the 1980, 1990 and 2000 database files were converted to Excel files for further use. ESRI ArcGIS 8.0 was used to create the distance and acces sibility variables required for our study. ArcGIS was also used for the spatial analyses which is discussed in greater detail in the next chapter. NCSS was the statistics soft ware program used for the stepwise and multiple regression analyses that are discussed further in chapter four.

PAGE 55

44 CHAPTER 3 MAPPING URBAN GROWTH AND CHANGE Chapter 2 discussed previous research concerning study areas for BRAC and economic impact. Most studies were concer ned with a large area surrounding the base, normally an entire city, county, MSA, or region. The previous studies discussed in Chapter 2 did not, however consider the incl usion of the growth patterns between the CBD and base. The previous studies also minimized the factors concerning economic impact. The RAND study utilized factors that addressed popul ation, housing units, vacancy rates, unemployment, labor force, K-12 enrollment, and retail sales. Hooker and Knetter addressed employment and per capita personal income changes. The majority of reports that addressed the i ssues and factors concerning base closure were centered primarily on population, employment or unemp loyment, and income and not on spatial patterns or urban spatial structures. The importance of base closure not having a significant negative impact did not explain whether the base had an immediate impact on urban and economic growth patterns in areas with close proximity to a base. Spatial analysis and statistical evidence was needed to support the hypothesis that military bases influence the areas near or c ontiguous to the base’s perimete r. If there was evidence that the military-base enhances the local economic growth, then the negative impact of base closure may decline with increa sing distance from the base. The purpose of our study was to assess th e local impact of military bases. The previous studies hold that local spillover eff ects were negligible, yet these studies did not

PAGE 56

45 actually include distance variables nor did they consider variability in relation to the locational accessibility of areas to the base and other prominent urban nodes. Previous research has also been couched from a small-scale perspective, encompassing entire counties, regional areas or entire MSAs. By contrast, our study focused on intra-MSA variability and inco rporated distance measures for a more restricted study area. Figure 3-1 is an image of the Jackso nville MSA including both study areas. Of greater importance was the impact on the imme diate areas that border or lie in close proximity to a military-base. Software Used to Create the Images for the Study Areas The software package used to create images for our study was ArcGIS 8.0. The data was gathered from several sources. Th e data used in the image processing of the Jacksonville study areas came from The Univ ersity of Florida’s Geoplan Center ( http://www.geoplan.ufl.edu./ ) under the Florida Geographi c Data Library (FGDL) and the map data was available with the St. John’s River Water Management District ( http://www.sjrwmd.com/programs/index.html ). The data for image processing of the Tampa – MacDill AFB study area also came from the FGDL and land use data came from the Southwest Florida Water Management District ( http://www.swfwmd.state.fl.us/data/ gis/libraries/physical_dense/lu95.htm ). Many processes and tools were used to develop th e images in ArcGIS (ESRI, 2001). The most commonly used tools were the Spatial Anal yst tool, Editor tool, and Xtools Pro. US Census TIGER Line files were used to es tablish the 1980 census-tract borders by taking the maps included in the 1980 US Census catalogues and editing the 2000 US censustracts in ArcMap using the editor tool.

PAGE 57

46 Figure 3-1. Jacksonville Metr opolitan Statistical Area (Image created using FGDL and Census Bureau Data). The edited census-tracts were then correct ed using Xtools Pro to calculate the area in square feet. The census-tracts were the unit of analysis in our study. The dependent

PAGE 58

47 and independent variables were calculated using census-tract data Florida Geographic Data Library files, and land use files from St. John’s River and Southwest Florida Water Management Districts. Furthermore, the di stance variables were calculated by using spatial data (specific land use variables) and the distance between each of the nodes, CBD, and military bases. Another important aspect of the census-tract’s utilization as the unit of analysis was because of decentralization of each study area. Decentralization was revealed through the creation of multinodal and polycentric patterns within the area. A ccording to Christian and Harper (1982) decentralization of employ ment and industry was a pattern of urban growth that has been recorded since th e 1940s. Christian and Harper described the process of decentralization by explaining th e vital role of multiple nodes (multinodal) within a region and their impact on urban economic growth patterns. Christian and Harper basically stated that the outward growth from the CBD led to more prominent roles of the nodes on economic growth pattern s within the area. Thus, if holding to Christian and Harper’s work, the Jacksonvi lle and Tampa study areas are decentralized and the CBDs of both areas are reliant on th e strength of the nodes within the area for further economic growth. Furthermore, Christ ian and Harper explai ned the significance of the development of the multinodal system seen in both the Jacksonville and Tampa study areas. The phenomena of suburban grow th after World War II led to outlying suburban centers that interacted with the CBD in a manner that greatly enhanced the economic growth of the region. The examples that were revealed in the Jacksonville study areas were the beach communities in the Mayport NS study area and Orange Park in the Jax NAS study area. Unlike Jacksonville MacDill AFB is not proximal to the outlying nodes for Tampa, which are St. Pete rsburg and other major metropolitan areas

PAGE 59

48 that are connected to the megalopolis of Central Florida. Centers of industry and commerce are also nodes within an urban center’s area of influence described by Christian and Harper (1982) as another as pect of the multinodal system explained. The commercial nodes used in our study played a v ital role along with di stance variability measures in assessing the military-base’s impact on economic growth. Christian and Harper also explained the importance of pol ycentric spatial stru ctures’ role in the economic growth of an area. According to th e polycentric spatial structures’ roles given by Christian and Harper, commerce and industr y are structured along hierarchical lines that influence decision-making functions by directly or indirectly determining new industry or commercial locatio ns, thus influencing the impact of economic growth rates within an area or region. The connectivity of major transportati on corridors to the multinodal systems (keeping in mind the roles of polycentric spatial structures, and decentralization of the MSAs) in the census-tracts of the study areas was the determining factor in the decision for census-tracts bei ng the units of analysis. The ArcGIS Spatial Analyst tool was used to calculate distance measures to the base and the CBD from the centroids of census-tracts. The purpose of the spatial analysis was to attempt to find a possible spatial relati onship between the base and the su rrounding area. Particularly, the commercial interaction between the base, CBD and key commercial areas within the study area distance factors that may play a ro le in the economic relationship the base may have with the local community. Creating Shapefiles for the Study Areas Since the unit of analysis fo r the study areas was the census-tract, the first step taken in creating the shapef iles for our study areas was determining the census-tract boundaries for the earliest time period be ing used in our study. Since our study’s

PAGE 60

49 temporal limits were 1980 to 2000, 1980 was the ideal time for establishing the censustract boundaries. 1980 census-tract boundaries were created using ArcGIS software programs, 1980 Census-tract maps included in US Census catalogues, and the 2000 US Census TIGER Line files (see Figures 3-2, 3-3, and 3-4). Establishing Study Area Boundaries Using Shapefiles The Jacksonville Metropolitan Statistical Area Jacksonville MSA encompasses several countie s in northwest Florida: Duval, Clay, St. John’s, and Nassau Counties. Our study wi ll involve two bases in the Jacksonville MSA: Mayport Naval Station and Jacksonville Naval Air Station. Mayport NS is situated along the southern side of the mouth of the St. John’s River as it empties into the Atlantic Ocean and east of the CBD. The Mayport NS area of study is entirely in Duval County. Jacksonville NAS is on the wester n side of the St. John’s River before the river turns to the east of the CBD. The Jacksonville NAS study area was situated along the southern border of Duval County and also encompassed the portion of Clay County that contains the Orange Park city limits. The means of creating a localized study area for the two bases was accomplished by using US Ce nsus-tracts from the US Census TIGER Line Files. By using census-tracts an area can be created for studying the economic impact between the bases and the CBD. The importance of the CBD is simple: The CBD is the heart of the area’s economic activity. The area between the CBD and the base defined the most active economic corridor.

PAGE 61

50 Figure 3-2. Mayport NS study ar ea census-tract boundaries. Figure 3-3. MacDill AFB study area census-tract boundaries.

PAGE 62

51 Figure 3-4. Jacksonville NAS st udy area census-tract boundaries. Census-tract boundaries in the study areas ch anged over time with increase in population. To overcome the problems associated with tract boundary change, th e 1980 census-tracts

PAGE 63

52 were used and the data from the 1990 and 2000 census were changed to fit within the boundaries of 1980. Mayport NS Study Area The Mayport NS study area was unique because there were only two major accessibility routes to the base from the CBD A large portion of the Jacksonville MSA lies to the north of the base; however, there is only one highway spanning the St. John’s River to the base and it was built during th e time of the study and did not open until recently. The other bridges cross the river afte r most of the major transportation routes enter the CBD (see figure 3-5). The inability to reach the ba se by road from the northern portion of the Jacksonville MSA limits encr oachment and reduces the likelihood of spillover effects. The important factor was that the base may impact only those areas that were accessible between the base and the CBD. Since there were no direct routes to the north of the base, the need to test for econom ic impact in those areas may be irrelevant. Another important factor concer ning urban growth was the lack of railroad activity near the base. Figure 3-5 illustrate s the lack of rail accessibili ty to the beaches and base. Although there is a lack of railroads near the base, there is an abundance of air activity. Craig Field is located in the center of the st udy area and the military-base has an airfield for the purpose of transferring aircraft from Jax NAS to the carrier group. Future urban growth is possible because of the new hi ghway construction connecting the northern regions to the area near the military-base. Ho wever, the possibility of increased growth could be improved with connecting railroad s to the industrial areas in the region, especially if the base is consid ered in future base closures.

PAGE 64

53 Figure 3-5. Modes of transportati on in the Mayport NS study area. Another unique feature of Ma yport NS is the site and situation of the base. The military-base is situated at the mouth of the St. John’s River emptying into Atlantic Ocean (Figure 3-5). The site is located on very marshy land and was unsuitable for development when the base was first built. Recent advances in urban development have made development of most of the land surrounding Mayport NS an easy and profitable task. The rapid development of land near Ma yport NS has become an encroachment issue with the DoD since BRAC rounds began to ta ke place in 1988. The most important issue with encroachment involves bases with airf ields; Mayport NS has an airfield for the purpose of outfitting the carrier group before maneuvers. If urban growth continues at its present rate in this area, the base could become a prime target for future BRAC rounds (See Figure 3-6).

PAGE 65

54 Figure 3-6. Urban and economic growth in Mayport NS study area (1973-2000). Note: Urban and economic growth combines the commercial and residential land use areas for the given year. The true indicator of economic impact di d not rely on images that have been represented thus far, but in the data that was represented in those figures. Jacksonville Naval Air Station (Jax NAS) Jax NAS has greater accessibility to th e Jacksonville CBD and possibly plays a greater role in economic development in the local area. Jax NAS is accessible to three major highways. Two of those highways (Interstate 95 and US Route 17) lead to the heart of the CBD, the third highway (Interstate 295) leads to Jacksonville’s entire periphery locations (See Figure 3-7). Finally Jax NAS is situated between Jacksonville’s CBD and Orange Park. The Jax NAS location is almost in the center of the Jacksonville MSA (See

PAGE 66

55 Figure 3-1). The location of the base may a llow for greater nonbasi c activity between the base and the MSA, which was the basis of the studies mentioned in the previous Chapters, and possibly influence the ur ban and economic growth in the local communities surrounding the base. Of greater importance is the underlying potential of land use if the base is close d. Because of the base’s location, availability of diverse transportation routes, and the pot ential for industrial, resident ial, and commercial growth, the opportunity to improve the economic impact after a base closure has greater potential than a base closure in Jacksonville, Nort h Carolina or Manhattan, Kansas. Camp LeJeune Marine Corps Base in Jacksonville, North Carolina and Fort Riley, Kansas, which are located in smaller towns and are the predominant source of income for those communities and play a greater role in the economy. Bases in large MSAs have only a small function in the economic structure with the ability to overcome economic impact from base closure. However, the vital role of the base’s mission to the overall goal of the DoD establishes the DoD’s criteria for base closure. There has been a tremendous amount of ur ban growth in Jax NAS study area since 1973. Jacksonville has shown a tremendous amount of growth to the south of the CBD, just as is seen to the east of the CBD. Ma yport NS had the advantage of being near three smaller towns on the beaches; Jax NAS also has an advantage of being near a fast growing community to the south: Orange Park (See Figure 3-7). Access to major highways to the south of Jack sonville has assisted the loca l communities in this area to prosper and grow, in addition the presence of the base and access to the Intercoastal Waterway via the St. John’s River has provide d additional means of growth in the area. Unlike the Mayport NS area, tourism is not as much a factor on the local economy.

PAGE 67

56 However, an advantage the Jax NAS study ar ea has over the Mayport NS study area is the access to major Figure 3-7. Transportati on routes to Jax NAS.

PAGE 68

57 Figure 3-8. Residential growth in the Jax NAS study area (1973-2000).

PAGE 69

58 Figure 3-9. Commercial growth in the Jax NAS study area (1973-2000). highways, particularly the Interstate system Interstate 295 connects Jax NAS to most of the MSA’s industrial areas. The importance of the transportation corridors to Jax NAS

PAGE 70

59 could be used as an argument for base closur e because of the prime real estate that would be made available if the base is sel ected for closure (Figures 3-8 and 3-9). Tampa-MacDill AFB Study Area The Tampa MSA has similar and different attributes with the Jacksonville MSA. The similarities for both cons ist of four counties, have a coastal boundary, and have a population over one million. The Tampa MSA differs from the Jacksonville MSA because of more CBDs of closer cities, more diverse modes of transportation within the Tampa MSA, and greater spatial diversity. Th e Jacksonville MSA cons ists primarily of the city of Jacksonville, whose city limits is the entire c ounty of Duval; and then Clay, Nassau, and St. John’s counties make up the re st of the MSA. After Jacksonville-Duval County, the next most populated county is Clay County w ith a population of approximately 150,000. The Tampa MSA has a huge economic advantage because of population: the next most populated county after Tampa-Hillsborough County is Pinellas County with a population at approximately 900,000. Jacksonville-Duval County’s population is smaller than the population in Pinellas County. The smallest population by county in the Tampa MSA is Hernando County and the population is approximately the same as Clay County. The extreme differe nce in population provides more opportunity for urban and economic growth and a better means to overcome the negative economic impact that may result from a base closur e. The greatest difference between the two MSAs is the Tampa MSA is the furthest left boundary of a Megalopolis: the Central Florida Corridor, which extends from the Ta mpa MSA in the west through the Lakeland and Orlando MSAs in the centr al portion of the megalopolis to Daytona on the Atlantic coast.

PAGE 71

60 The proximity of major metropolitan CBDs to the Tampa CBD is an advantage that Tampa has over many MSAs or urban areas with military bases. The CBDs of Clearwater and St. Petersburg, Florida are not very far from Tampa’s CBD (See Figure 3-10). MacDill AFB is separated from the St. Peters burg CBD (the next closest CBD to the base after the Tampa CBD) by Tampa Bay. The base does have access to St. Petersburg’s CBD by Gandy Bridge, which sp ans Tampa Bay. However, any economic impact from base closure should affect the Tampa CBD be fore any affect would occur on the other CBDs in the Tampa MSA. Also, any localized effect of urban and economic growth from the base would not include any areas outsi de of the study area because of natural boundaries mentioned previously. Income a nd taxes generated by the base and base personnel predominantly affects the commer cial and service activities within the immediate area. Another byproduct of income and property taxes paid by the government and personnel affect the immediate area in the form of income being produced for schools, police, fire, and emergency serv ices, and some medical services. More importantly, urban and economic growth ma y depend on improvement and expansion of these types of services in the local area. Th e development of these services do not depend on the presence of the base alone, normally re sidential growth has more of an impact on the increase or decrease of these services within the local area. Failing to reject a discer nable significance of the ba se on urban and economic growth in our study area may be difficult. Th e basic principles of Central Place Theory explained the enormous effects a CBD has on urban and economic grow th, especially if the CBD has a strong spatial and economic re lationship with outlying nodes. The Tampa CBD has very densely populated cities as nodes whose CBDs are pos sibly as strong as

PAGE 72

61 Tampa’s. The proximity of MacDill AFB to the Tampa CBD may cause difficulties because of the CBD’s economic strength a nd past growth trends from the CBD. Figure 3-10. Tampa MSA.

PAGE 73

62 Figure 3-11 illustrates the distinct adva ntage MacDill AFB has because of its relative location. The proxim ity of MacDill Air Force Base to the CBD and the availability of all major modes of transpor tation may support the sp atial relationship the base has with the CBD. The availability of ra ilroads, highways, air transportation and a port may allow some possibility of the base having discernable significant impact of economic growth on the localized area. Economic -growth should exhibit a rather healthy growth trend because of the movement of goods and services within the area between the base and the CBD. A spatial relationship in terms of co mmercial, industrial, and residential growth should support a positive impact of economic growth from the presence of the base. Empirical Approach to Study Areas Dependent Variables and Each Study Area The dependent variables created for the three study areas assessed the military-base impact on urban and economic growth at the localized level. Three dependent variables for analyzing economic growth were chosen to test the variabili ty of military-base impact. The tables of dependent variables’ valu es selected for the regression analyses of our study areas are found in Tables 3-2 through 3-4 with a key explaining the dependent variables created for regression analyses found in Table 3-1. The three dependent variables used in regression analyses were percentage change in commercial land use, percentage change in residential land use, and percentage change in median incomes. Three periods of time were included to warra nt a progression of growth in our study areas: 1980 to 1990, 1990 to 2000, and 1980 to 20001. The period of time that produced 1 St. John’s River Water Management District 1973 landuse data was employed, the landuse data for 1980 was not available and Southwest Florida Water Management District 1999 landuse data was employed, the landuse data for 2000 was not available.

PAGE 74

63 the best analytical results was 1980 to 2000. Th e dependent variables were tested in a multiple regression analysis at 95% confidence level. Figure 3-11. Modes of transporta tion accessible to MacDill AFB. Independent Variables and Each Study Area Several types of independent variables are created for each study area (Appendix A, Table A-1). The first sets of variab les are created using socio-economic and demographic data taken from the US Cens us Bureau between the years 1980 and 2000.

PAGE 75

64 Another set of variables is created from land use data (commercial and residential areas) taken from the St. John’s River and Southw est Florida Water Management Districts. Variables representing percentage changes between 1980 and 2000) are created from the socio-economic and demographic vari ables and land use variables. Table 3-1. Key to dependent variables. Name Database ID Description Percentage Change in Commercial Land use Area PCT CMR80_ CMR99 The percentage change of area by square feet of commercial land use, the numbers following identify the total commercial area for that year. Percentage Change in Residential Land use Area PCT RES80_ RES99 The percentage change of area by square feet of residential land use, the numbers following identify the total residential area for that year. Percentage Change in Median Household Income PCT MD_HHLD_ INC The percentage change calculated for the median household income for the given years. Distance variables were also created measuri ng straight-line and road distances between the military-base, nodes, and the CBD. Variab les for accessibility indices were created from straight-line and road distances between the nodes, CBD, with distances to the bases, as well as accessibility indices between the nodes a nd CBD without distances to the bases. The majority of independent variables were similar for each study area, specifically the socio-economic, demographic, land use, and percentage change variables. However, the accessibility and distance vari ables were significantly different for each study area. The differences in the number of independent variable s per study area were because of the varying amount of distan ce measures between nodes, the CBDs, and bases. The Jacksonville study areas had more independent variables because there were more nodes in the study areas.

PAGE 76

65 Choropleth maps suggested the possibility of military-base impact by observing the percentage of change of the dependent variab le and the proximity of the base. However, choropleth maps did not discer n the impact of a base on the economic growth rates for the localized area. Results of Choropleth Maps for the Jax NAS Study Area The percentage change in commercial area of the Jax NAS study area illustrated the possible impact that a near by nodal CBD may have on urban and economic growth of the study area (Figure 3-12). The census-tract immediately below the military-base is the Orange Park CBD; a tremendous amount of growth occurred in the census-tracts surrounding the Orange Park CBD and th e military-base. Although the nodal CBD possibly influences the factors fo r growth in the localized area there is a possibility that the military-base may also serve as an impacting force on the urban and economic growth of the localized area. Also of importance is the growth occurring on the opposite side of the St. John’s River from the military-base. The major highways (I-95 and US 1) that travel from the south of the CBD can be redirected to the military-base by connecti ng with I-295, this provides a transportation corridor from the base to the CBD via a diffe rent route and thus may explain the growth from the CBD. There is a great deal of grow th seen, but major highways lead to another nodal CBD in the MSA (St. Augustine), which is approximately 30 miles further south. The certainty seen in figure 3-12 was s ubstantial commercial growth between 1973 and 2000; however, commercial growth was only one of the variables involved to explain possible economic growth. Figure 3-13 illustrate d the residential changes in the Jax NAS study area.

PAGE 77

66 The significance seen in the figure was sim ilar to the commercial change seen for the same areas. Several assumptions can be made for the significance in change of the residential area. First, the availability of open land for development as the CBD’s urban growth expanded was more prominent to th e south and west of the CBD between 1973 and 2000. Second, highly accessible transportation ro utes leading from the CBD were found in the localized areas to the south and l ead to major nodal CBDs (Orange Park and St. Augustine). Finally, many areas near military bases were normally reserved for planning because of the airfields located on such bases; air traffic tended to act as a negative factor on land values, especially residential areas. The residential growth in the study area illustrated a strong possibility that the base may have some impact on the urban and economic growth of the localized area. The percentage change in median househol d income showed some greater growth immediately south of the Orange Park CBD The growth rates for median household income were apparent in figure 3-14. Unlik e the observations for the land use changes, the greatest growth was between the base and the CBD. Possible assumptions were similar to those given for land use changes. A nother indicator may be seen in real estate values over the time periods studied or the area may be more attractive to one-person households. Results of Choropleth Maps for the Mayport NS Study Area The percentage change in commercial area for the Mayport NS between 1973 and 2000 illustrated an interesting occurrence betw een the CBD and the three beach CBDs. Figure 3-15 showed the greatest amount of ch ange in commercial area occurred in census-tracts proximal to the beaches a nd base, particularly census-tracts 143.01, 143.02,

PAGE 78

67 and 146. Thus reflecting a great amount of change occurring between the CBD, major nodes and the base. Overall, the greatest amount of growth occurred near the three beach nodes to the south of Mayport NS. The base ma y have some influence in the growth of the area because of its location; however, the growth tends to follow the transportation corridors in the study area. Figure 3-15 illustrated growth along the transportation corridors, which follow the principles of Central Place Theory and Econom ic Base Theory in th e spatial relationship of the growth compared to the proximity of the CBD. The figure also illustrated a return in growth toward the CBD possibly because of interaction between the CBD and the beach nodes and base. The percentage of change in the resi dential area between 1973 and 2000 was quite different than the changes seen in the commercial area in the Mayport NS study area. Figure 3-16 illustrated the greater changes in re sidential growth occurred closer to the beach nodes. The majority of census-tracts near the CBD actually experienced a decline in residential area, particul arly those census-tracts that underwent or experienced a change in commercial area. The explanation c ould be because of the development of lowlying land in the largest census-tract in our study area (tract 143.02) The development of newer techniques to reclaim land that is nor mally unsuitable for building such as swamp and marshland has allowed the further developm ent of the areas closer to the base and beaches. The expansion of residential grow th in those areas may have allowed the expansion of residential area in the outer areas of our study area and making land in the CBD readily available for commercial development. Figure 3-17 illustrated the percentage of median household income having a similar pattern of growth that was s een with percentage change of residential area. The CBD and

PAGE 79

68 other prominent urban nodes showed strong tren ds in growth. Of gr eat interest was the strongest trend in growth that was obse rved in the areas closest to the base. Results of Choropleth Maps for the MacDill AFB Study Area The first aspect to remember concerning the differences between the two bases in Jacksonville and MacDill AFB in Tampa is MacDill AFB does not lie near a nodal CBD. MacDill AFB is situated approximately 6 m iles south of the CBD and is separated from the remainder of the MSA because of its lo cation on the southern tip of the Interbay Peninsula. The spatial relationship to the CBD leads to the assumption that any growth in the study area will be influen ced primarily by the CBD. Howeve r, the base may still have an impact on the economic growth in the lo calized area because of several factors: proximity to the CBD, isolation of the base from most of the MSA because of its location, the accessibility of dive rse modes of transportation that are located near the base, and the small amount of land available for commercial, reside ntial, and industrial development. Of greater importance is the ava ilable land resulting from the closure of the base could open up more opportunities for economic growth because of the other factors. Before the base can be selected for closure, the impact the base ha s on the area should be assessed even if the DoD and the BRAC commis sion select the base for closure. If a significantly substantial impact can be prove n, the argument against base closure has stronger support. The census-tracts representing the greatest percentage change in commercial area are close to the CBD (figure 3-18). The major ity of commercial gr owth is located in tracts surrounding the CBD. Of greater importa nce is the number of census-tracts that show a negative value in commercial growth. Fi ve of the eight census-tracts immediately to the north of the base show negative growth and three of the five have substantial

PAGE 80

69 negative growth. An assumption can be made that the base may have no impact or a possible negative impact on the economic growth of the localized area. Several factors may explain the negative growth within th e immediate area; such as, gentrification, creating available land for rezoning and developm ent, and increases in real estate values. Only the three census-tracts reflecting substantial negative commercial growth (figure 3-18) have a positive va lue in residential growth (f igure 3-19). The remainder of the census-tracts may have an extremely sm all amount of positive residential growth (maximum value of 0.02), but mostly negative growth values are represented. The values seen in residential and commercial growth support the earlier assumptions concerning the limited amount of space available fo r any growth in the study area. The greatest amount of percentage change seen in median household income is along a major transportation corridor between the CBD and the base (Figure 3-20). Although there is a substantia l amount of growth observed in the census-tracts, the impact provided by the base may be in conjunction with impacts because of the CBD. Review of Choropleth Maps Results The choropleth maps suggested that the bases in the Jacksonville study areas may have discernible impact on the economic-gr owth of the localized areas. Although the spatial analyses illustrate the possibility of discernible impact from the base on the economic-growth of the localized area, the nodal CBDs may actually have more of an impact on the economic growth than the ba ses. However, the bases may increase the effects of the nodal CBDs’ impact on economic-growth of the localized area. Unlike the study areas in the Jacksonville MS A, the spatial analyses for the Tampa study area suggested that the CBD is the pr edominant force in the economic growth of the study area

PAGE 81

70 Table 3-2. Jax NAS dependent variables. Censustract Commercial Area Percentage Change Residential Area Percentage Change Median Household Income Percentage Change 303 1444.79 118.04 124.59 304 826.01 -18.62 129.15 305 100.00 4.99 117.67 306 -9.40 0.00 111.98 307 100.00 127.21 194.23 308 61.17 20.19 129.13 309 100.00 240.73 162.36 2 100.00 -53.21 149.28 3 100.00 -17.44 106.81 4 288.99 -46.05 152.75 5 100.00 -59.37 304.45 6 151.06 -31.02 126.81 7 137.62 -11.94 226.29 8 51.05 -20.99 169.12 10 59.52 -75.87 63.62 11 -56.97 -1.21 219.53 12 -36.53 3.64 256.21 13 6.28 -17.85 127.51 15 -12.75 -5.50 156.50 16 90.95 -40.11 110.14 17 -23.95 -76.63 136.38 18 -23.62 -76.86 77.39 19 -21.97 56.81 121.34 20 -67.26 -6.45 196.17 21 46.13 -21.29 180.33 22 256.67 -22.27 146.67 23 -10.04 -25.76 162.50 24 1060.14 -7.53 212.75 25 12.62 -11.39 150.82 26 -53.87 -44.27 171.59 27 109.60 -4.98 92.38 28 1221.60 -9.85 162.97 29 504.54 -7.88 199.83 123 105.52 -20.91 125.00 124 -18.69 -16.12 123.60 130 -32.50 -3.71 146.94 131 77.96 -11.37 181.87 132 100.00 -69.52 112.64 133 479.11 98.24 149.35 134.01 888.72 10.86 147.91 134.02 100.00 -11.46 115.89

PAGE 82

71 Table 3-2. continued 135.01 100.00 80.59 147.46 135.02 114.36 2.60 149.67 162 -16.81 -17.75 129.19 163 -47.57 -44.06 142.25 164 672.49 -5.31 168.36 165 26.34 8.47 114.89 166.01 248.11 12.01 81.72 166.02 100.00 22.67 91.51 167.01 464.52 77.18 109.04 167.02 1240.45 294.80 137.00

PAGE 83

72 Table 3-3. Mayport NS dependent variables. Censustract Commercial Area Percentage Change Residential Area Percentage Change Median Household Income Percentage Change 1 -58.74 -27.68 164.01 2 100.00 -53.21 149.28 3 100.00 -17.44 106.81 4 288.99 -46.05 152.75 5 100.00 -59.37 304.45 6 151.06 -31.02 126.81 8 51.05 -20.99 169.12 10 59.52 -75.87 63.62 11 -56.97 -1.21 219.53 12 -36.53 3.64 256.21 13 6.28 -17.85 127.51 15 -12.75 -3.68 156.50 16 90.95 -40.11 110.14 17 -23.95 -76.63 136.38 18 -23.62 -76.86 77.39 19 -21.97 56.81 121.34 138 100.00 100.00 309.91 139.01 100.00 101.82 200.79 139.02 73.49 28.20 180.83 139.03 102.42 137.15 230.43 140 231.25 89.88 198.46 141 29.29 11.13 171.50 142 307.93 24.56 244.12 143.01 245.05 56.47 148.68 143.02 4722.05 1033.70 157.47 145 124.89 2.03 108.35 146 251.59 39.12 149.21 147 100.00 191.87 129.11 148 772.98 46.73 139.14 149.01 1620.57 6.86 102.62 149.02 100.00 68.46 126.96 150.01 100.00 -4.06 92.70 150.02 -20.12 -3.69 87.35 151 0.36 3.33 135.88 152 21.63 -3.56 148.59 153 -3.21 -1.17 108.94 154 15.68 -27.92 90.57 155 167.11 -24.61 123.89 156 474.52 -3.32 123.06 157 122.59 -40.00 97.84 158.01 41.21 2.99 101.17 158.02 365.16 34.96 155.30

PAGE 84

73 Table 3-4. MacDill AFB dependent variables. Censustract Commercial Area Percentage Change Residential Area Percentage Change Median Household Income Percentage Change 18 119.27 -36.41 123.51 19 71.80 -36.68 203.76 20 93.93 -20.45 112.43 21 451.09 -15.33 143.53 22 32.46 -10.54 195.15 23 -47.80 -6.93 139.88 24 -21.57 -17.62 60.53 25 -18.17 -12.59 104.85 26 -48.77 -92.23 207.38 27 131.19 -22.11 122.68 28 9312.35 -10.26 134.52 29 23.85 -13.85 197.13 30 233.70 -18.75 198.8 31 459.58 -24.30 239.2 32 171.96 -11.77 156.91 33 2092.39 -10.14 269.18 34 6313.90 -25.77 163.13 35 125.69 -21.53 163.93 38 84.20 -3.89 212.7 39 4.15 -19.92 110.96 40 421.46 -43.77 130.48 41 2834.95 -50.46 91.8 42 -0.53 -22.21 161.39 43 -79.57 -30.79 126.23 44 100.00 -39.21 174.82 45 616.53 -19.43 123.32 46 30.20 -27.68 105.38 47 -1.32 8.77 165.96 48 -29.63 12.70 124.33 49 45.26 -19.53 326.89 50 -31.70 -29.43 136.28 51 -19.94 1096.21 1218.98 53 -38.92 -40.54 181.68 54 -76.16 -4.91 212.8 55 1091.21 -95.04 457.69 57 35.40 -17.76 218.18 58 -9.93 -9.23 155.09 59 -3.58 2.88 185.02 60 60.09 -8.52 255.38 61 -16.91 -25.76 294.62 62 145.40 -18.80 403.87 63 374.84 -5.89 246.24 64 746.65 -1.36 180.37

PAGE 85

74 Table 3-4. Continued 65 -55.14 7.89 150.27 66 -46.96 3.23 100.01 67 -12.55 -6.29 205.83 68 34.27 -15.95 196.19 69 3184.58 -14.48 189.75 70 -17.60 -13.87 119.26 71 -69.55 9.16 105.76 72 100.00 -12.71 197.03 117.01 0.00 0.00 221.28 117.02 246.88 43.50 171.22 240.01 396.79 -25.55 131.99 240.02 -61.31 -9.12 239.02 240.03 0.00 14.61 241.27 241 -80.24 -11.83 208.10 244.03 165.56 -13.16 225.84 244.04 100.00 -26.03 225.04 244.05 100.00 0.02 203.78 244.06 21.41 -17.09 110.04 244.07 -37.08 29213.21 157.67 245.02 364.41 57.75 203.35

PAGE 86

75 Figure 3-12. Commercial area percentage change 1973-2000 for Jax NAS study area. Note: 1973 land use data used since 1980 land use data not available.

PAGE 87

76 Figure 3-13. Residential area percenta ge change 1973-2000 for Jax NAS study area.

PAGE 88

77 Figure 3-14. Median household income pe rcentage change 1980-2000 for Jax NAS study area.

PAGE 89

78 Figure 3-15. Commercial area percentage change 1980-2000 for Mayport NS study area. Note: 1973 land use data used since 1980 land use data not available.

PAGE 90

79 Figure 3-16. Residential area percentage change 1973-2000 for Mayport NS study area. Note: 1973 land use data used since 1980 land use data not available.

PAGE 91

80 Figure 3-17. Median househol d income percentage change 1980-2000 for Mayport NS study area.

PAGE 92

81 Figure 3-18. Commercial area percentage change 1980-1999 in the Tampa study area. Note: 1999 land use data used since 2000 land use data not available.

PAGE 93

82 Figure 3-19. Residential area percentage change 1980-1999 in the Tampa study area. Note: 1999 land use data used since 2000 land use data not available.

PAGE 94

83 Figure 3-20. Median househol d income percentage change 1980-2000 of the Tampa study area.

PAGE 95

84 CHAPTER 4 REGRESSION ANALYSES OF THE STUDY AREAS Multiple regression analyses were co nducted using NCSS software to model variations in percentage change of median household income, commercial area and residential area from 1980 to 2000 for each of the three study areas. The units of analysis chosen for our study were census-tracts of the study areas. The socio-economic and demographic data used for the dependent a nd independent variables were taken from STF3 files of the US Census Bureau. Di stance and accessibility index variables were created from the Florida Geographic Data Li brary (FGDL) files. Additionally, several accessibility indices were created to account for variations in locational of various census-tracts to selected commercial nodes, including the Central Business District (CBD), retail hubs, and the military-base. More over, straight-line and road distances were calculated between both these nodes to the cent er points of each census-tract. Distance variables were also included to access the impact of indi vidual nodes on growth rates in the selected study areas. A forward stepwise regression was employe d to determine the variables that are relevant in explaining percen tage change in the dependent variables. The variable criterion was set at the 95% confidence level. A total of 8 multiple regression models were tested to explain variation in growth rates by census-tract for the time period examined. Note that the stepwise regression for perc entage change in commercial area in the Tampa study area did not select any independent variables, thus a multiple regression

PAGE 96

85 analysis of the study area could not be c onducted. The 8 remaining multiple regression models will be discussed in turn. The key components examined in the mu ltiple regression analyses were the regression (or beta) coefficients, the t-Values and identified transa ction types defined by Economic Base Theory concepts (the vari ables had one or more of the following functions: basic function, nonbasi c function, and/or a transfer of income function). The sign of the beta coefficient determined the na ture of the independent variable’s impact on the dependent variable. The greater the t-Valu e associated with an independent variable, the greater impact on the dependent variable (g rowth rate). The identifier listed in the transaction type column explains the inde pendent variable’s role on the study area’s economic base. The adjusted R-square was an indicator of the m odel’s goodness of fit. Finally, normality tests and plots of error term s were examined to determine if the error structure was normally distributed, and whet her there were any random outliers. If outliers were found, they were removed and the models re-ran without them. However, if the new results caused greater difficulties the original model was retained and the outlier highlighted in the discussion section. Jacksonville NAS Study Area Regression Analyses Results Three multiple regression analyses were conducted for the Jax NAS study area, one for each of the dependent variables. Each model will be briefly discussed. The stepwise regression results for the de pendent variable per centage change in commercial area selected 10 inde pendent variables with beta coefficients that were significantly different from zero. Table 4-1 li sts the results of the multiple regression analysis for percentage change in commercial area. The initial regression analyses revealed an outlier in census -tract 304, which is adjacent to the Orange Park commercial

PAGE 97

86 node in census-tract 306. Before the removal of the outlier, the model failed to meet the normality assumptions. On removal of the outlier the model met the standard criterion for normality. The estimated coefficients for selected independent variables rejected the hypothesis that the beta values were not statistically diffe rent from zero at the 95% confidence level. An examination of the resu lts revealed that lo cational accessibility variables and distance variables did not play a role in explai ning variation in commercial growth rates as these variables were not se lected. Thus, the results of the regression analysis explained the role of certain demographic variables on the impact of commercial growth rates. The positive va riables suggested that an increase in the demographic variability resulted in a d ecrease in commercial growth in our study area. An opposite effect was observed with negative variables, which suggested a decrease in demographic variability result in an increase in commer cial growth for the same study area. Three independent variables with positive beta coeffi cients and very high t-Values (greater than 4.2) reflected an impact on commercial gr owth rates through nonba sic economic activity and/or transfer income (Table 4-1). The positive beta coefficients suggested a positive association between various demographic variables and commercial growth rates. However, two independent variables had negati ve beta coefficients with high negative tValues (-3.7) suggesting a nega tive association on commercial growth rates. The negative beta coefficients suggested decrease in comm ercial growth rates as either the number of persons born out of state, service employment, or vacant housing increases. The adjusted R-squared value was high, sugge sting that the nine variables selected accounted for approximately 83% of the variat ion in commercial growth rates by censustracts. The three normality tests were accepte d. The scatterplot revealed a random pattern

PAGE 98

87 between residuals and predicted values. The pr obability plot also revealed the remaining variables lie within range of normality. The regression model revealed no discernible evidence of the military-base having a significant impact on the commercial gr owth rates in our st udy area for the period examined. If the base had any impact on the commercial growth rates in our study area, the impact was negligible or very minor. The stepwise regression for the dependent variable percentage change in median household income for the Jax NAS study area se lected five independent variables for the multiple regression analysis. Of greater in terest, a distance variable (Straight-Line Distance to Commercial Area in Census-tract 167.01) was selected for this model indicating that a possible rela tionship in percentage change for the dependent variable and distance in the localized area may be present. However, extreme outliers were present in the model and after the removal of the outliers (ass ociated with census-tracts 7, 5, 11, and 12) the model was left with thr ee variables and the distance variable was removed (Table 4-2). Interestingly, the outlier s were in census-tracts that are proximal to the CBD (census-tracts 11 and 12 are part of the CBD). The multiple regression model produced relatively weak results in comparison to the model reexamined earlier with an adjusted R-Square value that falls below 0.30. An interesting variable select ed was the commercial area in 2000. The negative beta coefficient suggested the decrease in co mmercial area had a positive impact on the percentage change of median household in come growth rates. The positive beta associated with median gross rent between 1980 and 2000 suggested a positive impact on the growth of median household in come during time period examined.

PAGE 99

88 Similar to the regression results for the pe rcentage change in commercial area, the normality assumption was verified. Unlike the previous regression model discussed for our study area, the three normality tests were not accepted until after the removal of the outliers. Once again, this regr ession model failed to uncover a discernible impact of the base on the economic growth rates as neither the distance to base variable nor the locational accessibility variables were se lected by the stepwise procedure. The stepwise regression results for the percen tage change in residential area for the Jax NAS study area selected 21 independent vari ables including the in tercept. Of greater importance was the selection of an accessibil ity index variable and a distance variable. The initial regression results identified five outliers (census-tracts 8, 124, 167.02, 304, and 309). Interestingly, two of the outliers were proximal to the CBD (8 and 124); two were proximal to the Orange Park commerc ial node (304 and 309); and one was proximal to the military-base (167.02). The multiple regression results (Table 4-3) revealed that the distance variable had a positive beta coefficient suggesting that as distance from the CBD increases residential growth rates increase. Thus, the residential gr owth rates tend to be lower near the CBD. Also, the t-Value (20.0136) was substantially higher than the t-cri tical value (|2.06|), which suggested that the distance variable ha d a substantial role in the positive impact on residential growth rates. The accessibility index variable also had a positive beta coefficient, but the t-value was substantially smaller than the di stance t-Value (4.6755). The inclusion of the accessibility variable suggested that there was a discernible impact of the prominent nodes (including the base ) on the economic grow th within our study area. Specifically, as locationa l accessibility to the promin ent nodes increased so did the residential growth rate. Note that the calcu lation of the locationa l accessibility index

PAGE 100

89 included the location of the base, suggesting that the base was an important node in explaining residential growth rates. The m odel revealed a large F-ratio and a better overall goodness of fit (with an adjusted R-S quare of 0.992), with a low root mean square error. The normality tests indicated that the error terms were normal. Overall, the results supported the hypothesis that th e base had a statistically significant impact on the residential growth rates, although its’ individual impact was unknown given that the locational accessibility bundl es the impact of all prominent nodes in the study region. While it was shown that the base and the CBD had a significant impact on the residential growth rates, the result did not imply that th e base was the most influential node in our study region. Furthermore, in all three cases examined distance to the base, as a standalone variable (distance to the base) did not test to be signifi cant. Only when distances to all prominent nodes were considered in a bundled accessibility index did it prove significant. In short, the base played a role in explaining va riation in residential growth rates, however, its’ overall impact was nonseparable from the other prominent nodes. Mayport NS Study Area Regression Analyses Results Three multiple regression models were run for the Mayport Naval Station study area. The same three dependent variables th at were analyzed for the Jax NAS study area were also used for the Mayport NS study area. The stepwise regression conducted for perc entage change in commercial area for the Mayport NS study area selected four inde pendent variables including the intercept (Table 4-4). The initial regression results re vealed two outliers (associated with censustracts 149.01 and 149.02), however, the rem oval of both outliers created severe complications with the ongoing model’s result s, particularly c ontinual removal of

PAGE 101

90 variables until a model was non-existent. The removal of the resi dual outlier (149.01) did not complicate the model’s abil ity to meet goodness of fit cr iteria and did not hinder the acceptance of the assumptions of normality (t he scatterplot still shows the predicted outlier in census-tract 149.02). The model results suggested th at the base did not have a discernible impact on the commercial grow th rates in the Mayport area. Unlike the previous model discussed (Table 4-3), ther e were no distance or locational accessibility variables selected, implying that the location of the base offers no explanatory power in accounting for variation in commercial growth rates. However, the variables that were selected suggested that a positive impact was observed because of two demographic variables with positive beta values and signi ficant t-Values. The re sults suggested that population and residential growth were the f actors which account for commercial growth in the area. The stepwise regression for percentage change in median household income produced seven significant independent vari ables (Table 4-5). Unlike the previous regression models, the multiple regression anal ysis for percentage change in median household income did not reveal extreme outliers. This regression model included a distance variable. Although the distance variable had both a positive beta coefficient and significant t-Value (6.4893), the distance variab le was not directly associated with the military-base, but represents the distance of census-tracts to a specified commercial node in census-tract 143.01. Also, census-tract 143.01 was adjacent to the census-tracts containing both Jacksonville Beach and Ne ptune Beach (census-tracts 140 and 141). Thus, the beta coefficient indicated that as distance increases from the commercial node an increase could be observed in median hous ehold income. More importantly, the results

PAGE 102

91 revealed that the location of the military-b ase did not have a discernible impact on the growth rate of median household income. Other implications were seen in the mode l’s remaining demographic variables. The percentage change in population between 1980 and 2000, the dependent ratio in 2000, and the percentage change in persons residing in the same county between 1980 and 2000 had high negative t-Values. The beta coeffi cients for these variables suggested the decrease in each variable resulted in an in crease in the growth rates of median household income. Whereas the betas for the percentage change in self employed persons between 1980 and 2000 and the percentage change in median gross rent between 1980 and 2000 suggested the increase in each variable result ed in a decrease in the growth rates of median household income. However, the impli cations seen in these variables did not reveal any evidence that would suggest an impact on the growth rates of median household income from the base. Overall th e model was very good, with an adjusted RSquare of approximately 0.84. The stepwise regression for percentage change in residential area selected 10 independent variables including the intercept. Only one distance variable was selected (Table 4-6). The initial regression results revealed two outliers (census-tract 19 and 143.02) in the model. After the removal of th e two outliers the model kept the distance variable for census-tract 143.02. However, the mo st significant variables observed in the model were employment variables (educationa l services in 2000 and percentage change in employment in the production, repair, a nd labor industries between 1980 and 2000). Of great importance was the location of the cen sus-tract 143.02: proximal to the base and beach. The positive coefficients suggested that as distance increases from the base and beach, there was a decrease in percentage change of residential area growth rates (Figure

PAGE 103

92 3-16). The employment variables suggested that increases in those industries had a significantly positive impact on the residential growth rates. The model revealed that a significantly positive impact had occurred on the residential growth rates, but the base played a small and limited role with the commercial node in census-tract 143.02. Furthermore, the residential pattern of grow th observed near the base and CBD may be related to other factors such as real estate values near the beach areas and CBD. Thus, there was virtually no statistical evidence that the base had an impact on the residential growth rates during the period in question. No te that the model met the assumption of normality and independence of the error terms. The results for the Mayport NS study area suggested the base has a negligible impact on the economic growth in the study area. MacDill AFB Study Area Regression Analyses Results Only two multiple regression models were analyzed for the Tampa study area. As stated earlier, the stepwise regression for th e dependent variable percentage change in commercial area did not select any independent variables; th erefore, the model was not considered. The stepwise regression for percentage change in median household income selected 20 independent variab les including the intercept. Of great importance was the selection of the accessibility variable (Table 4-7). The initial regression results revealed an extreme outlier in census-t ract 51 located in the heart of the Tampa CBD. After the outlier was removed and the stepwise re gression conducted agai n, only 12 independent variables were selected. However, the accessi bility variable remained in the model. The accessibility variable (without the Ai r Force Base) suggested once again that the base did not have a significant impact on commercial growth rates. The CBD and other commercial nodes did have significance in the impact on commercial growth in the

PAGE 104

93 area. The beta coefficient associ ated with this variable sugge sted that the variable had a positive impact on the median household income growth. That is, increasing accessibility to prominent nodes was associated with higher growth rates. Moreover, since the variable excluded the base, the impact was predom inantly the result of accessibility through transportation corridors betw een the CBD, commercial nodes, and the census-tracts. Thus, the positive coefficient suggested that as locational accessibility of census-tracts to the CBD and commercial nodes increased, so did median household income. However, the base may have had a minor role in the impact from the CBD because of the proximity of the base to the CBD. Also, the increase in distance from the CBD and base should reveal a decrease in growth (Figure 3-20), a nd lesser growth on the fringes of our study area. The difference may be because of the other major commercial nodes in the region, such as the St. Petersburg and Clearwater CBDs. Other independent variables selected with significant positive beta coefficients were Persons Employed in the Wholesale Trade Industry in 1990 Persons Employed by the Federal Government in 2000 Percentage Change in Median Gross Rent Between 1980 and 2000 and Percentage Change in Mean Housing Values Between 1980 and 2000 The demographic variable for federal employees suggested that the base may have ha d a contributing role to growth rates in the area. Negative beta coefficients were asso ciated with the commerc ial area variable and renter occupied housing. The most significant negative beta coefficient observed was the Percentage Change of Depe ndent Ratio Between 1980 and 2000 variable. Not only was this variable the most significant negative vari able, it had the largest t-Value in the model (-10.7063). The variable suggested that the “d ependency ratios” had a negative impact on the percentage change of median household income growth rates. The negative beta coefficient for commercial area also s uggested a negative im pact on the median

PAGE 105

94 household income growth rates. Furthermore an examination of the transaction types of variable producing negative co efficients suggested a trend in impact of basic economic activity. The model results revealed that the ba se did not have a discernible impact on the median household income growth rates. Th e model indicated a strong goodness of fit with an adjusted R-square value above 0.84. The assumption of normality and independence of error were also accepted. Once again, there was no evidence to suppor t the contention that proximity to the base played a significant role in explaining variation in pe rcentage change in median household income. The proximity of the base to the CBD, however, made it difficult to assess its underlying im pact or importance. The stepwise regression for the dependent va riable percentage change in residential area selected 15 independent variables incl uding the intercept. Unlike the previous model, both a distance and accessibility index variable were not selected, and the one predicted outlier (census-tract 51) eventual ly caused more problems with its removal. However, the outlier can be explained because of its location in the heart of the CBD. Once again the distance between the base a nd the CBD was minimal and could account for the non-selection of these variables. Th e limited amount of land for land use change between the base and CBD could also account for the non-selection of these variables. Although distance and accessibility index variables were not sele cted, travel time to work variables were selected (Table 4-8). Of great significance was the very high positive beta coefficient observed for the Percentage Change of Persons Residing in the Same House Between 1975 and 1995 Furthermore, the t-Value was the highest amongst the positive demographic variables. The variable revealed positive impact on residential growth rates within the localized

PAGE 106

95 area. Three other variables also supported a positive impact on residential growth rates; Persons with a Household Income of $50,000 or More in 1980, Percentage Change in Persons with a Household Income of $50,000 or More Between 1980 and 2000 and Persons Paying Gross Rent of $500.00 or More in 1990 Several indicators could suggest the positive impact on the residential growth rates within the study area. The first indicator was the gentrification of the CBD may have forced an increase in the cost of rent as well as an increase in real estate values. The second indicator was the proximity of the base to the CBD. Finally, th e third indicator was the possibi lity of high turnover with military members and their families transferring in and out of the area. Finally, as mentioned before, the limited amount of space prevented large increases in commercial and industrial growth. Therefore, any new deve lopments in residen tial area would occur with the redevelopment of existing residentia l areas. Thus, an increase in relative costs for new housing or rent may occur. Further explanation for this occurrence was seen in the negative beta coefficients. Particularly, the variable e xplaining median owner’s cost. Median owner’s cost had the highest negative t-Value (-11.2738) and suggested that the up keep cost of housing had a negative impact on residential growth rate s. An assumption can be made that the redevelopment of some areas had forced cu rrent housing to “keep up with the Joneses” and thus the explanation, the subsequent in crease in the costs of housing maintenance. However, there were more coefficients w ith a negative value than variables with a positive value. A review of the spatial analysis for percentage change in residential area revealed that the trend in growth was obs erved along a transporta tion corridor between the base and the CBD. The majority of censu s-tracts that were located on the Interbay Peninsula revealed very little or no growth. More significantly, the model revealed that

PAGE 107

96 the base did not appear to have any impact on the residential growth rates in the area. Overall, the model was better than the median household income model according to the goodness of fit criteria. Most of which wa s explained by economic and demographic variables. Once again, the regression results suggest that the base has no discernible impact on percentage change in median household income. Also, the mode l that includes an accessibility variable excludes the base. Howe ver, the spatial relationship between the base and the CBD (i.e., their proximity) may explain why the base is insignificant in terms of accounting for variati ons in residential growth ra tes within the study region. Table 4-1. Jacksonville NAS study area regression results for per centage change in commercial area between 1973 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept -255.7974 -2.4514 0.018694 Reject Ho n/a Persons Born Out of the State in 1990 -0.1065 -3.9158 0.000342 Reject Ho NBE & Transfer Income Persons Residing in the Same County in 1975 0.1978 3.3902 0.001582 Reject Ho NBE & Transfer Income Percentage Change in Persons Residing in the Same County Between 1975 and 1995 2.3996 5.6203 0.000002 Reject Ho NBE & Transfer Income Persons Using Public Transportation in 1980 0.772 4.1696 0.000159 Reject Ho NBE & Transfer Income Persons Employed in the Agricultural, Fishing, Forestry, and Mining Industries in 2000 5.9254 2.5746 0.013842 Reject Ho BE, NBE, & Transfer Income

PAGE 108

97 Table 4-1. Continued Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Persons Employed in the Services, Entertainment, and Recreation Industries in 2000 -1.3965 -3.7934 0.000493 Reject Ho BE, NBE, & Transfer Income Vacant Housing in 1980 -0.8899 -2.8986 0.006057 Reject Ho NBE & Transfer Income Persons Paying Gross Rent of $500 Per Month or More in 1990 1.3535 10.3618 0.000001 Reject Ho NBE & Transfer Income Median Gross Rent in 2000 0.643 3.1518 0.003071 Reject Ho NBE & Transfer Income T-Critical |2.021075| Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 2158367.4898 2158367.4898 Model 9 5585325.054 620591.6727 28.3795 0.000001 Error 40 874703.3843 21867.5846 Root Mean Square Error = 147.876923847963 R-Squared = 0.8646 Coefficient of Variation = 0.711742196868125 Adjusted R-Squared = 0.8341 Normality Tests Section Assumption Value Probability Decision (5%) Skewness 0.7307 0.464951 Accepted Kurtosis -0.8299 0.406605 Accepted Omnibus 1.2227 0.542630 Accepted Plots Section Histogram Probability Plot Scatterplot 0.0 3.0 6.0 9.0 12.0 -300.0-150.00.0150.0300.0Histogram of Residuals of PCTDCMR73_CMR2Residuals of PCTDCMR73_CMR2K Count -300.0 -150.0 0.0 150.0 300.0 -3.0-1.50.01.53. 0 N ormal Probability Plot of Residuals of PCTDCMR73 Expected NormalsResiduals of PCTDCMR73_CMR2K -300.0 -150.0 0.0 150.0 300.0 -500.00.0500.01000.01500. 0 Residuals vs PredictedPredictedResiduals Note: 1973 land use data used since 1980 land use data not available. The acronyms in the regression equation section are: BE = Ba sic Economic Activity and NBE = Nonbasic Economic Activity.

PAGE 109

98 Table 4-2. Jacksonville NAS study area regression results for pe rcentage change in median household income between 1980 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept 110.197 7.9473 0.000001 Reject Ho n/a Percentage Change in Median Gross Rent between 1980 and 2000 0.3091 3.1782 0.002711 Reject Ho NBE & Transfer Income Commercial Area in 2000 -1.64E-06 -2.6915 0.010019 Reject Ho BE, NBE & Transfer Income T-Critical 2.015368 Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 903660.7186 903660.7186 Model 2 14711.6239 7355.8119 9.3755 0.000406 Error 44 34521.2664 784.5742 Total (Adjusted) 46 49232.8903 1070.2802 Root Mean Square Error = 28.0102523574346 R-Squared = 0.2988 Coefficient of Variation = 0.202005525255716 Adjusted R-Squared = 0.2669 Normality Tests Section Assumption Value Probability Decision (5%) Skewness -0.7370 0.461099 Accepted Kurtosis -1.7723 0.076346 Accepted Omnibus 3.6843 0.158480 Accepted Plots Section Histogram Probability Scatterplot 0.0 3.0 6.0 9.0 12.0 -60.0-30.00.030.060.0Histogram of Residuals of PCTDMD_HHLD_INC_80 Residuals of PCTDMD_HHLD_INC_80_2K Count -60.0 -30.0 0.0 30.0 60.0 -3.0-1.50.01.53.0 m al Probability Plot of Residuals of PCTDMD_HHLD_IExpected NormalsResiduals of PCTDMD_HHLD_INC_80_2K -60.0 -30.0 0.0 30.0 60.0 60.090.0120.0150.0180.0 Residuals vs PredictedPredictedResiduals

PAGE 110

99 Table 4-3. Jacksonville NAS study area regression results for per centage change in residential area between 1973 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept -91.3334 -18.4307 0.000001 Reject Ho n/a Persons Under the Age of 16 in 2000 -7.61E-03 -3.8322 0.000761 Reject Ho NBE & Transfer Income Percentage Change in Persons Residing in the Same House Between 1975 and 1995 -0.0206 -3.7323 0.000982 Reject Ho NBE & Transfer Income Persons Working at Home in 1980 0.4728 7.4418 0.000001 Reject Ho NBE & Transfer Income Persons Employed in Retail Trade Industry in 2000 0.103 8.9234 0.000001 Reject Ho NBE & Transfer Income Persons Employed in Public Administration Industry in 1980 -0.1695 -9.0124 0.000001 Reject Ho NBE, & Transfer Income Percentage Change in Persons Employed in the Public Administration Industry Between 1980 and 2000 7.84E-02 6.1238 0.000002 Reject Ho NBE, & Transfer Income Percentage Change in Self Employed Persons Between 1980 and 2000 0.22 26.7689 0.000001 Reject Ho NBE, & Transfer Income Percentage Change in Median Household Income Between 1980 and 2000 -8.41E-02 -3.7235 0.001004 Reject Ho NBE, & Transfer Income Median Income in 1990 -2.37E-03 -10.4601 0.000001 Reject Ho NBE, & Transfer Income Median Income in 2000 2.23E-03 8.3239 0.000001 Reject Ho NBE, & Transfer Income Vacant Housing in 2000 0.1256 12.1922 0.000001 Reject Ho NBE & Transfer Income

PAGE 111

100 Table 4-3. Continued Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Housing Used for Recreation in 1990 -0.9578 -9.9349 0.000001 Reject Ho NBE, & Transfer Income Percentage Change in Housing Used for Recreation Between 1980 and 2000 -2.59E-02 -4.8793 0.000051 Reject Ho NBE, & Transfer Income Persons Paying Gross Rent of $500 Per Month or More in 1980 0.9866 11.4547 0.000001 Reject Ho NBE & Transfer Income Persons Paying Gross Rent of $500 Per Month or More in 2000 -0.0115 -2.9341 0.007069 Reject Ho NBE & Transfer Income Median Gross Rent in 1980 0.108 4.5607 0.000116 Reject Ho NBE & Transfer Income Mean Housing Values in 2000 -9.3E-05 -5.9772 0.000003 Reject Ho NBE & Transfer Income Accessibility Index with Naval Air Station in 1990 3.7501 20.0136 0.000001 Reject Ho NBE & Transfer Income Distance to Commercial Area in Census-tract 309 0.7546 4.6755 0.000086 Reject Ho NBE & Transfer Income T-Critical |2.059539| Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 1221.1959 1221.1959 Model 20 96293.3891 4814.6695 282.6090 0.000001 Error 25 425.9126 17.0365 Total (Adjusted) 45 96719.3017 2149.3178 Root Mean Square Error = 4.12753006517063 R-Squared = 0.9956 Coefficient of Variation = 0.801081102540034 Adjusted R-Squared = 0.9921 Normality Tests Section Assumption Value Probability Decision (5%) Skewness 0.4743 0.635308 Accepted Kurtosis 1.4450 0.148451 Accepted Omnibus 2.3130 0.314581 Accepted

PAGE 112

101 Table 4-3. Continued Plots Section Histogram Probability Plot Scatterplot 0.0 3.5 7.0 10.5 14.0 -10.0-5.00.05.010.0Histogram of Residuals of PCTDRES73_RES2Residuals of PCTDRES73_RES2K Count -10.0 -5.0 0.0 5.0 10.0 -3.0-1.50.01.53.0 rmal Probability Plot of Residuals of PCTDRES73Expected NormalsResiduals of PCTDRES73_RES2K -10.0 -5.0 0.0 5.0 10.0 -100.0-37.525.087.5150.0 Residuals vs PredictedPredictedResiduals Note: 1973 land use data used si nce 1980 land use data not available.

PAGE 113

102 Table 4-4. Mayport NS study area regression results for percentage change in commercial area between 1973 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept 254.1904 5.7252 0.000001 Reject Ho n/a Percentage Change in Population Between 1980 and 1990 4.2206 -3.3820 0.001711 Reject Ho NBE & Transfer Income Percentage Change in Persons Under 16 Years Old Between 1980 and 2000 5.9165 5.5476 0.000003 Reject Ho NBE & Transfer Income Percentage Change in Residential Area Between 1973 and 2000 1.8013 3.5362 0.001112 Reject Ho NBE & Transfer Income T-Critical |2.026192| Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 2091038.0809 2091038.0809 Model 3 20340538.9603 6780179.6534 181.4573 0.000001 Error 37 1382510.4771 37365.14803 Total (Adjusted) 40 21723049.4374 543076.2359 Root Mean Square Error = 193.30066742981 R-Squared = 0.9364 Coefficient of Variation = 0.855941948217211 Adjusted R-Squared = 0.9312 Normality Tests Section Assumption Value Probability Decision (5%) Skewness 0.3881 0.697911 Accepted Kurtosis 0.0538 0.957086 Accepted Omnibus 0.1535 0.926098 Accepted Plots Section Histogram Probability Plot Scatterplot 0.0 3.0 6.0 9.0 12.0 -600.0-300.00.0300.0600.0Histogram of Residuals of PCTDCMR73_CMR 2 Residuals of PCTDCMR73_CMR2K Count -600.0 -300.0 0.0 300.0 600.0 -3.0-1.50.01.53.0 r mal Probability Plot of Residuals of PCTDCMR7 3 Expected NormalsResiduals of PCTDCMR73_CMR2K -600.0 -300.0 0.0 300.0 600.0 -1000.0500.02000.03500.05000. 0 Residuals vs PredictedPredictedResiduals Note: 1973 land use data used si nce 1980 land use data not available.

PAGE 114

103 Table 4-5. Mayport NS study area regression results for percentage change in median household income between 1980 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept 55.1629 2.7273 0.009914 Reject Ho n/a Percentage Change in Population Between 1980 and 2000 -2.5716E-01 -5.3313 0.000006 Reject Ho NBE & Transfer Income Dependent Ratio in 2000 -103.4834 -5.7125 0.000002 Reject Ho NBE & Transfer Income Percentage Change in Persons Residing in the Same County Between 1980 and 2000 -0.220632 -5.3487 0.000006 Reject Ho NBE & Transfer Income Percentage Change in Self Employed Persons Between 1980 and 2000 0.365847 6.3257 0.000001 Reject Ho NBE & Transfer Income Percentage Change in Median Gross Rent Between 1980 and 2000 0.542015 5.1965 0.000009 Reject Ho NBE & Transfer Income Distance to Commercial Area in Census-tract 14301 7.69655 6.4893 0.000001 Reject Ho NBE & Transfer Income T-Critical |2.030108| Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 946418.8213 946418.8213 Model 6 112159.8406 18693.3068 35.9677 0.000001 Error 35 18190.3883 519.7254 Total (Adjusted) 41 130350.2288 3179.2739 Root Mean Square Error = 22.797486 R-Squared = 0.8604 Coefficient of Variation = 0.151869 Adjusted R-Squared = 0.8365 Normality Tests Section Assumption Value Probability Decision (5%) Skewness -0.1604 0.872574 Accepted Kurtosis 0.013 0.989609 Accepted Omnibus 0.0259 0.987136 Accepted

PAGE 115

104 Table 4-5. continued Plots Section Histogram Probability Plot Scatterplot 0.0 3.0 6.0 9.0 12.0 -60.0-30.00.030.060.0Histogram of Residuals of PCTDMD_HHLD_INC_8 0 Residuals of PCTDMD_HHLD_INC_80_2KCount -60.0 -30.0 0.0 30.0 60.0 -3.0-1.50.01.53 m al Probability Plot of Residuals of PCTDMD_HHLExpected NormalsResiduals of PCTDMD_HHLD_INC_80_2K -60.0 -30.0 0.0 30.0 60.0 50.0125.0200.0275.0350. 0 Residuals vs PredictedPredictedResiduals Note: 1973 land use data used since 1980 land us e data not available. The acronyms in the regression equation section are: BE = Ba sic Economic Activity and NBE = Nonbasic Economic Activity.

PAGE 116

105 Table 4-6. Mayport NS study area regression results for percentage change in residential area between 1973 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept -107.7941 -6.8176 0.000001 Reject Ho n/a Percentage Change in Persons Using Other Types of Transportation Between 1980 and 2000 0.417 4.9898 0.000024 Reject Ho NBE & Transfer Income Persons Employed in the Educational Services Industry in 2000 0.3217 11.8797 0.000001 Reject Ho NBE & Transfer Income Percentage Change in Employment in Productivity, Repair, and Labor Industries Between 1980 and 2000 0.6865 8.7656 0.000001 Reject Ho BE, NBE, & Transfer Income Persons with an Income Less Than $25,000 in 1980 -2.39E-02 -2.7402 0.010240 Reject Ho NBE & Transfer Income Percentage of Persons with an Income Less Than $25,000 Between 1980 and 2000 -0.8151 -4.9575 0.000026 Reject Ho NBE & Transfer Income Percentage of Persons with an Income Between $25,000 and $50,000 Between 1980 and 2000 0.011 4.1303 0.000267 Reject Ho NBE & Transfer Income Total Commercial Area in 1973 4.62E-06 4.1508 0.000252 Reject Ho BE, NBE, & Transfer Income Total Commercial Area in 2000 -7.23E-06 -6.6718 0.000001 Reject Ho BE, NBE, & Transfer Income Straight-Line Distance to Commercial Area in Census-tract 14302 7.6038 4.6693 0.000059 Reject Ho NBE & Transfer Income T-Critical |2.042272|

PAGE 117

106 Table 4-6. continued Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 2144.4933 2144.4933 Model 9 122534.7924 13614.9769 56.1112 0.000001 Error 30 7279.2835 242.6429 Total (Adjusted) 39 129814.0759 3328.5661 Root Mean Square Error = 15.576995317206 R-Squared = 0.9439 Coefficient of Variation = 2.12741069368163 Adjusted R-Squared = 0.9271 Normality Tests Section Assumption Value Probability Decision (5%) Skewness -1.6673 0.095450 Accepted Kurtosis 0.7077 0.479121 Accepted Omnibus 3.2808 0.193899 Accepted Plots Section Histogram Probability Plot Scatterplot 0.0 2.5 5.0 7.5 10.0 -40.0-22.5-5.012.530.0Histogram of Residuals of PCTDRES73_RES 2 Residuals of PCTDRES73_RES2K Count -40.0 -22.5 -5.0 12.5 30.0 -3.0-1.50.01.53.0 N ormal Probability Plot of Residuals of PCTDRES73_Expected NormalsResiduals of PCTDRES73_RES2K -40.0 -22.5 -5.0 12.5 30.0 -100.0-25.050.0125.0200.0 Residuals vs PredictedPredictedResiduals Note: 1973 land use data used si nce 1980 land use data not available.

PAGE 118

107 Table 4-7. MacDill AFB study area regression results for percentage change in median household income between 1980 and 2000. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept 113.9096 7.8188 0.000001 Reject Ho n/a Percentage Change of Dependent Ratio Between 1980 and 2000 -1.7222 -10.7063 0.000001 Reject Ho Transfer Income Persons Employed in the Wholesale Trade Industry in 1990 0.4067 3.8235 0.000360 Reject Ho BE, NBE, & Transfer Income Percentage of Persons Employed in the Financial, Insurance, and Real Estate Industries Between 1980 and 2000 -5.13E-02 -2.2037 0.032085 Reject Ho BE, NBE, & Transfer Income Persons Employed in the Business and Repair Industries in 1990 -0.4018 -4.2155 0.000102 Reject Ho BE, NBE, & Transfer Income Persons Employed by the Federal Government in 2000 0.3353 3.6652 0.000589 Reject Ho NBE & Transfer Income Percentage of Persons Below the Poverty Level Between the Age of 15 and 64 Between 1980 and 2000 -8.04E-02 -5.8722 0.000001 Reject Ho NBE & Transfer Income Renter Occupied Housing in 1990 -2.92E-02 -3.2146 0.002267 Reject Ho NBE & Transfer Income Percentage Change in Median Gross Rent between 1980 and 2000 0.3389 4.9597 0.000008 Reject Ho NBE & Transfer Income Percentage Change in Mean Housing Values between 1980 and 2000 0.2219 5.5003 0.000001 Reject Ho NBE & Transfer Income Commercial Area in 1980 1.88E-06 -2.4517 0.017681 Reject Ho BE, NBE, & Transfer Income Straight-Line Distance Accessibility Index without Air Force Base 6.8677 3.2286 0.002178 Reject Ho NBE & Transfer Income T-Critical |2.007584|

PAGE 119

108 Table 4-7. continued Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 2102905.3596 2102905.3596 Model 11 261398.5491 23763.5045 31.9065 0.000001 Error 51 37984.1042 744.7864 Total (Adjusted) 62 299382.6533 4828.7525 Root Mean Square Error = 27.2907742199947 R-Squared = 0.8731 Coefficient of Variation = 0.149374432129979 Adjusted R-Squared = 0.8458 Normality Tests Section Assumption Value Probability Decision (5%) Skewness 0.9835 0.325363 Accepted Kurtosis -1.2070 0.227428 Accepted Omnibus 2.4241 0.297580 Accepted Plots Section Histogram Probability Plot Scatterplot 0.0 5.0 10.0 15.0 20.0 -60.0-30.00.030.060.0Histogram of Residuals of PCTDMD_HHLD_INC_ 8 Residuals of PCTDMD_HHLD_INC_80_2K Count -60.0 -30.0 0.0 30.0 60.0 -3.0-1.50.01.53.0 Probability Plot of Residuals of PCTDMD_HHL D Expected NormalsResiduals of PCTDMD_HHLD_INC_80_2K -60.0 -30.0 0.0 30.0 60.0 50.0150.0250.0350.0450. 0 Residuals vs PredictedPredictedResiduals Note: The acronyms in the regression equation s ection are: BE = Basic Economic Activity and NBE = Nonbasic Economic Activity.

PAGE 120

109 Table 4-8. MacDill AFB study area regression results for percentage change in residential area between 1980 and 1999. Regression Equation Section Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Intercept 8276.7958 13.1567 0.000001 Reject Ho n/a Persons Under the Age of 16 in 2000 -2.1512 -4.3659 0.000065 Reject Ho Transfer Income Percentage Change in Persons Residing in the Same House Between 1975 and 1995 44.6239 13.1549 0.000001 Reject Ho NBE & Transfer Income Percentage Change in Persons Working at Home Between 1980 and 2000 2.0663 3.7363 0.000488 Reject Ho NBE & Transfer Income Persons Employed in the Manufacturing Industry in 2000 -4.8495 -3.1012 0.003193 Reject Ho BE, NBE, & Transfer Income Percentage Change in Persons Employed in Transportation, Communications, and Public Utilities Industries Between 1980 and 2000 -6.235 -11.2613 0.000001 Reject Ho BE, NBE, & Transfer Income Persons Employed in the Wholesale Trade Industry in 1980 -12.6776 -4.2096 0.000109 Reject Ho BE, NBE, & Transfer Income Percentage Change in Persons Employed in Personal Services, Entertainment, and Recreational Industries Between 1980 and 2000 -4.5535 -4.1513 0.000131 Reject Ho BE, NBE, & Transfer Income Percentage Change in Persons Employed in the Educational Services Industry Between 1980 and 2000 -2.2925 -3.6725 0.000594 Reject Ho NBE & Transfer Income

PAGE 121

110 Table 4-8. continued Independent Variable. Beta Coefficient. T-Value (Ho: B=0) Prob. Level Decision (5%) Transaction Type Percentage Change in Persons with a Household Income Between $30,000 and $50,000 Between 1980 and 2000 -1.4696 -3.2359 0.002176 Reject Ho NBE & Transfer Income Persons with a Household Income Between of $50,000 or More in 1980 14.6673 7.5936 0.000001 Reject Ho NBE & Transfer Income Percentage Change in Persons with a Household Income of $50,000 or More Between 1980 and 2000 0.3703 5.1271 0.000005 Reject Ho NBE & Transfer Income Persons Paying Gross Rent of $500 Per Month or More in 1990 4.75 6.5706 0.000001 Reject Ho NBE & Transfer Income Percentage Change in Aggregate Housing Values Between 1980 and 2000 -1.1971 -4.3005 0.000081 Reject Ho NBE & Transfer Income Median Owner's Cost with Mortgage in 1980 -18.8658 -11.2738 0.000001 Reject Ho NBE & Transfer Income T-Critical |2.009575| Analysis of Variance Section Source DF RSS Mean Square F-Ratio Prob Level Intercept 1 13555368.1609 13555368.1609 Model 14 802134182.08 57295298.7198 72.0250 0.000001 Error 49 38979100.418 795491.8453 Total (Adjusted) 63 841113282.5 13351004.484 Root Mean Square Error = 891.903495483583 R-Squared = 0.9537 Coefficient of Variation = 1.93799295632215 Adjusted R-Squared = 0.9404 Normality Tests Section Assumption Value Probability Decision (5%) Skewness -0.5101 0.609950 Accepted Kurtosis -0.6830 0.494601 Accepted Omnibus 0.7267 0.695326 Accepted

PAGE 122

111 Table 4-8. continued Plots Section Histogram Probability Plot Scatterplot 0.0 3.8 7.5 11.3 15.0 -2000.0-1000.00.01000.02000. 0 Histogram of Residuals of PCTDRES80_RES9 9 Residuals of PCTDRES80_RES99 Count -2000.0 -1000.0 0.0 1000.0 2000.0 -3.0-1.50.01.53.0 o rmal Probability Plot of Residuals of PCTDRES8 0 Expected NormalsResiduals of PCTDRES80_RES99 -2000.0 -1000.0 0.0 1000.0 2000.0 -5000.03750.012500.021250.030000 Residuals vs PredictedPredictedResiduals Note: 1999 land use data used si nce 2000 land use data not available.

PAGE 123

112 CHAPTER 5 SUMMARY AND CONCLUSION Summary Our study revealed certain obstacles th at had been problematic for Base realignment and Closures (BRACs) in the pa st 20 years. Consistent obstacles were congressional in-house arguments over BRAC, grass-roots efforts in prevention of BRAC, political interference out side of Congress’ control, congressional and Department of Defense (DoD) debates over selected base s and/or areas of impact, and legislators gloom-and-doom debates over economic impact. The ability to overcome these obstacles appeared to be problems that face future BRACs. However, severa l possibilities could assist BRAC implementation, the most obvious would be educating the public about the results of previous impact studies of ec onomic growth rates after BRAC has occurred. Ideally, an impact study at the smallest ge ographical scale may possibly reveal the greatest impact possible for the economic growth rates for a local area. Thus far, previous research has been conducted that revealed th e results of impact for economic growth rates at larger geographical scales (county, regiona l, and greater scales). Our study was also consistent with previous research and the ge ographical scale (censustract level) applied may also be considered equivalent to the regi onal level. Future rese arch should consider a smaller geographical scale (census block groups). Urban population size and labor force consid erations should be a consideration in future research. The percentage of military personnel and civilian contracted employees out of the total population in previous resear ch was normally between a range of less than

PAGE 124

113 1% to approximately 3%. The percentage was normally equivalent to the percentage of the initial negative impact on economic growth rates observed in our study areas. If the percentages of population were found to be higher at a sma ller geographical scale (census block groups), then perhaps the impact at a more localized level may be observed. Unlike previous results found in resear ch at greater geographical sc ales, one advantage that can be used in future research that was utilized in our study was the inclusion of distance and accessibility variables. The selection of areal units should include distance and accessibility to major transportation corridor variables. For example, the study areas chosen for our study could be the basis for a new study with changes only to the geographical scale in the research being c onducted. The spatial analyses and regression models may change because of the change in geographical sc ale. A possible indication in the change in results could be explained by a si mple change in variables. For instance, the stepwise regressions never selected milita ry population variables for the regression models but military population variables may be selected in some models at a smaller geographical scale. Particularly if census bl ock groups have a tendenc y to reveal higher percentages for military employment out of the total population within the census block groups. A significant result in previous rese arch may be explained because of the extremely low percentage of military employees out of the total population at the geographical scale chosen. Again, the results of our research were negl igible for all cases, just as the percentage of total populati on employed by the military appeared to be negligible. Another observation revealed in our study that should be mentioned in future research is the gross waste of taxes and money used to prevent BRAC. Government offices spend millions to hundreds of millions of dollars in efforts to find a method for

PAGE 125

114 preventing BRAC. Unfortunately once the decisi on to close a base has been rendered, the money spent is lost and any efforts were wa sted. Instead of using the money to prevent BRAC, the money should be used in assisti ng planning commissions in those areas that may be affected in preparation to overcome an impact in the most expedient and efficient means possible. Also, educational tools s hould be implemented to prepare the local population for the upcoming possibility of cl osure and expedite the transition from military employment to future possibilities if needed. The Florida grass-roots movement in our study may explain the importance of re direction of funds. Our study mentioned the possibility of BRAC for M acDill AFB, since our study was completed the DoD has announced some BRAC activity. Th e most significant effect th at will be felt by Florida will actually take place in Ma yport NS: by the end of this year (2005) the carrier USS John F. Kennedy (JFK) will be decommissione d. The JFK is stationed at Mayport NS. Several possibilities may occur with this deci sion: (1) a new carrier will be reassigned to Mayport NS and any other changes that occur with the reassignment will be a non-factor on the current economy, (2) a new carrier wi ll not be assigned and only the remaining naval groups will remain leading to the possi ble closure of Jacksonville NAS (Jax NAS) because of the change in its mission status: no carrier – no need of aircraft, (3) the worst scenario would be a closure of Mayport NS and the resulting closure of Jax NAS. The possibility of MacDill AFB closing is still a po ssibility, and the most likely choice in the 2005 round of BRAC. The results found in our study revealed that the possibility of MacDill AFB being selected for BRAC could be profitable for the city of Tampa, however, a study at a smaller ge ographical scale could show a significant negative impact on the area if MacDill AFB is closed. The most significant outcome of all previous research showed a negligible impact on th e economic-growth rates at the geographical

PAGE 126

115 scales chosen, including our study. The importa nce of the results found in the previous research could be used as an advantag e by planning commissions through implementing planning tools to overcome initial impact on economic growth rates. This could be accomplished by zoning for industry and reside ntial areas that are made available by BRAC that would improve the local areas ec onomic-growth rates and thus the economy of the area as a whole over time. Eventually a diverse economic base for any metropolitan statistical area will include a number of factors that could explai n the rate of economic growth. A list of the factors that could explain th e rate of economic growth in clude those mentioned in the previous paragraphs in this section; geographical scale, population growth and density, and development along transportation corridors. Other factors that c ould explain the rate of economic growth are: gentri fication, the relative location of the CBD, local impact of commercial and financial nodes, zoning, and economic diversity to name a few. The latter factors listed play an integral role in the MSAs ec onomic growth. Gentrification has recently been a necessary action with many majo r cities in improving real estate values of land located proximally to the major city’s CBD. Improvement in land use includes the removal of decrepit and condemned buildings and replacing the buildings with newer and more aesthetic structures or new zoning of land use for parks and recreation. The relative location of the CBD and the local impact of commercial and financial nodes are selfexplanatory. The CBD should impact the economic-growth rate because the CBD is the economic heart of any community. The local impact seen from the commercial and financial nodes should reflect the growth that is revealed by the activity of the nodes; i.e., if the nodes have a healthy and robust grow th because of increas ing economic activity, then the economic growth rates should reveal a positive impact. However, if a negative

PAGE 127

116 growth trend is seen in th e nodal activity in the local community, a negative impact should be observed. Economic diversity may e xplain the results for mo st of the previous research conducted for BRAC impact on econom ic growth rates. Research has shown that economic diversity plays a vital role in regional econom ic growth and development. It may be reasonable, therefore, to assume that areas with little economic diversity, especially places that are heavily reliant on a military-base as an income generating industry, could be devastated economically from a base clos ure. Eventually, zoning also has a function on economic growth rates. Fo r instance if a zoning commission were to select an area for commercial zoning in which there are only tertiary roads, then growth in the commercial area may suffer because of insufficient traffic volume. The most pressing concern for zoning is to benefit the local area by cr eating zones for land use that will improve real estate values and the pot ential for future economic-growth. More importantly, zoning could be one of the factor s that could play a major role in future BRAC; preplanning by zoning land use of av ailable land from BRAC could reduce and possibly remove any initial ne gative impact felt by the local community in the event of BRAC. Concluding Remarks The literature concerning base impact on economic growth at the regional level revealed that a base’s influence on a lo cal economy might be negligible when the economy is large and diverse. Concern of futu re base closures is continuously debated amongst the Department of Defense (DoD) a nd Congress, there are conflicting opinions on the likely impact of closure. Both side s have legitimate arguments concerning base closures. The DoD is continuously looking for methods to reduce budget costs because of budget cuts (before 2001) the DoD experienced Congress has argued th at base closures

PAGE 128

117 are detrimental to a healthy local economy. Th e latter may be true in some cases where bases in small communities are the pre dominant industry. Although small communities such as Jacksonville, NC and Manhattan, KS rely solely on the base for economic growth, bases in large cities or metropolitan areas have a diverse economy and the base in those areas have a negligible impact on the economic growth in those areas. Previous research rejected the hypothesis that military bases have a negative impact on economic-growth rates at regional levels. Although the initial impact had negative connotations, economic recovery was usually quick and positive economic growth rates eventually were observed. Our study was consis tent with previous research results and also rejected the hypothesis that military bases had a discernible impact on economic growth rates at a geographical scale (censustract level) because of proximal distances and accessibility along transpor tation corridors between the base and major commercial and financial nodes at the 95% confidence level. Thus, the geographical scale is the first factor that shoul d be considered in future research concerning BRAC. Howe ver, the rate of economicgrowth could be explained by a number of factors, such as population growth and density, the relative location of the Central Business District (CBD), developm ent along transportation corridors, local impact of commercial and financial nodes, gentrification, economic diversity, zoning, and other factors that are essentially non-base re lated. Spatial analysis should reveal the impact on economic growth rates because of the diversity of f actors aforementioned. Although the observed results from BRAC impact was negligible in all regression models of our study, a smaller geographical scale ( census block groups) may reveal different results.

PAGE 129

118 The spatial analyses revealed that economi c growth occurred in most of our study areas. However, the impact on economic growth could have been caused by a number of factors, such as the CBD, transportati on corridors between the CBD, base, and commercial nodes, gentrification, and other certain elements that can impact the economic-growth of our study area. Of great importance was the understanding that spatial analysis did not reject or fail to reject theory on economi c growth indicators. Spatial analysis only offers a visual represen tation of spatial data. In the case of our study, the spatial analyses offered visual repr esentations of the depe ndent variables used for determining the possibility of base impact on economic growth rates of our study areas. Although the spatial analyses revealed certain growth patterns near the bases in some cases, there was no discernible eviden ce that the bases were responsible for the impact on economic-growth rates. Instead, th e bases may have played a minor role in economic-growth rates through inclusion w ith the CBD and other commercial nodes. The regression analyses revealed the result s of data testing to determine the causes for either a positive or negative impact on th e economic-growth of our study areas. The significance of base impact on economic growth rates in our study areas was found to be negligible in all cases. Only one case reve aled the base having a minor role and the dominant force on economic growth rates in the study area involved the locational accessibility between the base, CBD, and commercial nodes. Thus, the impact on economic-growth rates was not by the base, but by the CBD. Whereas, transfer income from the base assisted the CBD and co mmercial nodes impact on the local economy. Our study was consistent with previous re search and studies with similar results that supported the contention that the impact of a military-base may be negligible or nondiscernible in economies that are large and di verse. The implications (if assertion is

PAGE 130

119 correct): base closures in la rge and diversified economies mi ght have a limited negative impact and would likely be replaced with land use activities that w ould lessen the overall impact of base closure and the associated loss of transfer income. Overall, all cases rejected the hypothesis that military bases ha ve a discernible impact on economic growth rates at a geographical scale (census-tract level) because of proximal distances and accessibility along transportation corridors betw een the base and major commercial and financial nodes at the 95% c onfidence level. The explana tion for the rejection of the hypothesis may be explained through the geogra phical scale and/or the relatively small percentage of the total population of each study area employed at the military bases. The research results should only be a pplied to the study area s in our study. The results of our study revealed that the three military bases in the Tampa and Jacksonville MSAs did not have a discernible impact on local economic-growth rates over the periods examined. This was possible because of the si ze and diversity of th ese economies and/or the geographical and temporal scales at which the analysis was conducted. Furthermore, the results of our study revealed no eviden ce of localized spillover effects from the military bases to census-tracts surrounding th e three bases included in our study. Future research should consider a comparative study between geographical scales (regional through census block groups) and greater temporal scales (greater than ten to twenty-year periods over possibly thirty or forty years minimum) in de termining discernible impact on economic-growth rates at each level.

PAGE 131

120 APPENDIX KEY TO INDEPENDENT VARIABLES Table A-1. Key to Independent Variables. Name Description Census-tract Identification The census-tract identification fo r the Jacksonville study areas in Duval and Clay Counties and the Tampa study area in Hillsborough and Pinellas Counties. Total population The total po pulation of the tract for that set of census data. Percentage Change in Population The percentage change between the given populations by years (e.g., 80_90 is the percentage ch ange of population between 1980 and 1990). Total population Under Age 16 The sum of all persons under the age of 16. Percentage Change in Persons Under Age 16 Between 1980 and 2000 The percentage change in populat ion of persons under the age of 16 between 1980 and 2000. Total population Age 16 to 64 The sum of all persons between the ages of 16 and 64. Percentage Change in Persons Between the Ages 16 and 64 Between 1980 and 2000 The percentage change in populatio n of persons between the ages of 16 and 64 between 1980 and 2000. Total population Over Age 64 The sum of all persons over the age of 64. Percentage Change in Persons Over Age 64 Between 1980 and 2000 The percentage change in populatio n of persons over the age of 64 between 1980 and 2000. Dependent Ratio The dependen cy ratio for given years. Percentage Change in Dependent Ratio Between 1980 and 2000 The percentage change in de pendent ratios between 1980 and 2000 Total Households The total number of hou seholds reported in the census-tract. Percentage Change in Households Between 1980 and 2000 The percentage change in tota l number of households between 1980 and 2000. Born In State The number of people re ported that were born in Florida. Percentage Change of Persons Born in State Between 1980 and 2000 The percentage change of the num ber of people reported that were born in Florida between 1980 and 2000. Born Out Of State The number of people reported that were not born in Florida but were born in the United States.

PAGE 132

121 Table A-1. continued Percentage Change of Persons Born Out of State Between 1980 and 2000 The percentage change in the num ber of people reported that were not born in Florida but were bor n in the United States between 1980 and 2000. Resided In Same House The number of people reported that lived in the same house during 5 years before the census data. Percentage Change of Persons that Resided In Same House Between 1975 and 1995 The percentage change in the num ber of people reported that lived in the same house during 5 years before the census data between 1975 and 1995. Resided In Same County The number of people reported th at lived in the same county during 5 years before the census data. Percentage Change of Persons that Resided In Same County Between 1975 and 1995 The percentage change in the num ber of people reported that lived in the same county during 5 years before the census data between 1975 and 1995. Resided In Florida The number of people reported that lived in Florida during 5 years before the census data. Percentage Change of Persons that Resided In Florida Between 1975 and 1995 The percentage change in the num ber of people reported that lived in Florida during 5 years before the census data between 1975 and 1995. Worked In County The number of people that worked in the county of their residence. Percentage Change of Persons that Worked In County Between 1980 and 2000 The percentage change in the number of people that worked in the county of their residence between 1980 and 2000. Drive Alone The number of people that drive a car, truck or van to work without other passengers. Percentage Change of Persons that Drive Alone Between 1980 and 2000 The percentage change in the number of people that drive a car, truck or van to work without other passengers between 1980 and 2000. Carpool The number of peopl e that carpool to work. Percentage Change of Persons that Carpool Between 1980 and 2000 The percentage change in the num ber of people that carpool to work between 1980 and 2000. Public Transportation The number of people that use public transportation as a means of going to and returning from work. Percentage Change of Persons that Used Public Transportation Between 1980 and 2000 The percentage change in the number of people that use public transportation as a means of goi ng to and returning from work between 1980 and 2000. Other Transportation The number of people that use a bicycle, motorcycle, moped, or other means than previously mentioned as a means of travel to and from work.

PAGE 133

122 Table A-1. continued Percentage Change of Persons that Used Other Transportation Between 1980 and 2000 The percentage change in the numbe r of people that use a bicycle, motorcycle, moped, or other means than previously mentioned as a means of travel to and from work between 1980 and 2000. Work At Home The number of people that work at home. Percentage Change of Persons that Worked At Home Between 1980 and 2000 The percentage change in the numbe r of people that work at home between 1980 and 2000. Travel Less Than 10 Minutes The number of people that require less than 10 minutes of travel time to their work. Percentage Change of Persons that Travel Less Than 10 Minutes Between 1980 and 2000 The percentage change in the num ber of people that require less than 10 minutes of travel time to their work between 1980 and 2000. Travel Between 10 to 19 Minutes The number of people that requir e between 10 and 19 minutes to travel to their work. Percentage Change of Persons that Travel Between 10 to 19 Minutes Between 1980 and 2000 The percentage change in the number of people that require between 10 and 19 minutes to tr avel to their work between 1980 and 2000. Travel Between 20 to 29 Minutes The number of people that requir e between 20 and 29 minutes to travel to their work. Percentage Change of Persons that Travel Between 20 to 29 Minutes Between 1980 and 2000 The percentage change in the number of people that require between 20 and 29 minutes to tr avel to their work between 1980 and 2000. Travel 30 Minutes or More The number of people that require more than 30 minutes to travel to their work. Percentage Change of Persons that Travel 30 Minutes or More Between 1980 and 2000 The percentage change in the num ber of people that require more than 60 minutes to travel to their work between 1980 and 2000. Veterans The number of people reported that they are military veterans. Percentage Change in Veterans Between 1980 and 2000 The percentage change in the num ber of people reported that they are military veterans between 1980 and 2000. Non-Veterans The number of people reporte d that they are not military veterans. Percentage Change in NonVeterans Between 1980 and 2000 The percentage change in the num ber of people reported that they are not military veterans between 1980 and 2000. Armed Forces The number of people reported that are currently serving in the United States military and living in the census-tract.

PAGE 134

123 Table A-1. continued Percentage Change in Armed Forces Between 1980 and 2000 The percentage change in the num ber of people reported that are currently serving in the United States military and living in the census-tract between 1980 and 2000. Civilian Employees The number of people reported that are currently employed but not with the military. Percentage Change in Civilian Employees Between 1980 and 2000 The percentage change in the num ber of people reported that are currently employed but not with the military between 1980 and 2000. Agriculture, Fishing, Forestry, and Mining Industries The number of people reported working in the Agriculture, Forestry, Fishing, or Mining industries. Percentage Change in Agriculture, Fishing, Forestry, and Mining Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Agriculture, Forestry, Fishing, or Mining industries between 1980 and 2000. Construction Industry The number of people reported working in the Construction industry. Percentage Change in Construction Industry Between 1980 and 2000 The percentage change in the number of people reported working in the Construction industry between 1980 and 2000. Manufacturing Industry The number of people reported working in the manufacturing industry. Percentage Change in Manufacturing Industry Between 1980 and 2000 The percentage change in the number of people reported working in the manufacturing industry between 1980 and 2000. Transportation, communications, and Public Utilities Industries The number of people reported wo rking in the Transportation, Communications, or Public Utilities industries. Percentage Change in Transportation, communications, and Public Utilities Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Transportation, Communicat ions, or Public Utilities industries between 1980 and 2000. Wholesale Trade Industry The number of people reported wo rking in the Wholesale Trade industry. Percentage Change in Wholesale Trade Industry Between 1980 and 2000 The percentage change in the number of people reported working in the Wholesale Trade industry between 1980 and 2000. Retail Trade Industry The number of people reported wo rking in the Retail Trade industry.

PAGE 135

124 Table A-1. continued Percentage Change in Retail Trade Industry Between 1980 and 2000 The percentage change in the number of people reported working in the Retail Trade industry between 1980 and 2000. Financial, Insurance, and Real Estate Industries The number of people reported working in the Financial, Insurance, or Real Estate industries. Percentage Change in Financial, Insurance, and Real Estate Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Financial, Insurance, or Real Estate industries between 1980 and 2000. Business and Repair Services Industries The number of people reported work ing in the Business or Repair Services industries. Percentage Change in Business and Repair Services Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Business or Repair Services industries between 1980 and 2000. Personal Services, Entertainment, and Recreation industries The number of people reported work ing in the Personal Services, Entertainment, or Recreation Services industries. Percentage Change in Personal Services, Entertainment, and Recreation industries Between 1980 and 2000 The percentage change in the number of people reported working in the Personal Services, Entert ainment, or Recreation Services industries between 1980 and 2000. Health Services Industries The number of people reported work ing in the Health Services industry. Percentage Change in Health Services Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Health Services industry between 1980 and 2000. Education Services Industries The number of people reported working in the Educational Services industry. Percentage Change in Education Services Industries Between 1980 and 2000 The percentage change in the number of people reported working in the Educational Services industry between 1980 and 2000. Other Professional Services Industries The number of people reported wo rking in other professional services not mentioned in th e other reported industries. Percentage Change in Other Professional Services Industries Between 1980 and 2000 The percentage change in the number of people reported working in other professional services not mentioned in the other reported industries between 1980 and 2000. Public Administration The number of people reporte d working in the Public Administration industry.

PAGE 136

125 Table A-1. continued Percentage Change in Public Administration Between 1980 and 2000 The percentage change in the number of people reported working in the Public Administrati on industry between 1980 and 2000. Employment in Management, Business or Financial Professionals, and Sales Occupations Employment in Management, Business or Financial Professionals, or Sales. Percentage Change in Employment for Management, Business or Financial Professionals, and Sales Occupations Between 1980 and 2000 The percentage change in th e employment for Management, Business or Financial Professi onals, or Sales between 1980 and 2000. Employment in Private or Protective Services and Other Professional Services Occupations Employment in the Private or Protective Services, or Other Personal Services. Percentage Change in Employment for Private or Protective Services and Other Professional Services Occupations Between 1980 and 2000 The percentage change in th e employment for Private or Protective Services, or Other Pe rsonal Services between 1980 and 2000. Employment in Production, Repair, and Labor Occupations Employment in Production, Repair, or Labor. Percentage Change in Employment for Production, Repair, and Labor Occupations Between 1980 and 2000 The percentage change in the employment for Production, Repair, or Labor between 1980 and 2000. Wages and Salary Employees The number of people reported working for a salary or wages. Percentage Change in Wages and Salary Employees Between 1980 and 2000 The percentage change in the number of people reported working for a salary or wages between 1980 and 2000. Federal Government Employees The number of people reporte d working for the Federal Government. Percentage Change in Federal Government Employees Between 1980 and 2000 The percentage change in the number of people reported working for the Federal Government between 1980 and 2000.

PAGE 137

126 Table A-1. continued State Government Employees The number of people reported work ing for the State Government. Percentage Change in State Government Employees Between 1980 and 2000 The percentage change in the number of people reported working for the State Government between 1980 and 2000. Local Government Employees The number of people reported working for the Local Government. Percentage Change in Local Government Employees Between 1980 and 2000 The percentage change in the number of people reported working for the Local Government between 1980 and 2000. Self Employed The number of people reported that are self-employed. Percentage Change in Self Employed Between 1980 and 2000 The percentage change in the num ber of people reported that are self-employed between 1980 and 2000. Household Income Less Than $10000 Per Year The number of people reported having a household income less than $10,000 per year. Percentage Change in Household Income Less Than $10000 Per Year Between 1980 and 2000 The percentage change in the number of people reported having a household income less than 10,000 dollars per year between 1980 and 2000. Household Income Between $10000 and $30000 Per Year The number of people reported having a household income between 10,000 and 30,000 dollars per year. Percentage Change in Household Income Between $10000 and $30000 Per Year Between 1980 and 2000 The percentage change in the number of people reported having a household income between 10,000 and 30,000 dollars per year between 1980 and 2000. Household Income Between $30000 and $50000 Per Year The number of people reported having a household income between 30,000 and 50,000 dollars per year. Percentage Change in Household Income Between $30000 and $50000 Per Year Between 1980 and 2000 The percentage change in the number of people reported having a household income between 30,000 and 50,000 dollars per year between 1980 and 2000. Household Income Greater Than $50000 Per Year The number of people reported havi ng a household income greater than 50,000 dollars per year. Percentage Change in Household Income Greater Than $50000 Per Year Between 1980 and 2000 The percentage change in the number of people reported having a household income greater than 50,000 dollars per year between 1980 and 2000. Median Household Income The median house hold income reported for the census-tract.

PAGE 138

127 Table A-1. continued Percentage Change in Median Household Income Between 1980 and 2000 The percentage change in the median household income reported for the census-tract between 1980 and 2000. Median Family Income The median family income reported for the census-tract. Percentage Change in Median Family Income Between 1980 and 2000 The percentage change in the me dian family income reported for the census-tract between 1980 and 2000. Income Less Than $25000 Per Year The number of people reported having an income less than $25,000 per year. Percentage Change in Income Less Than $25000 Per Year Between 1980 and 2000 The percentage change in the num ber of people reported having an income less than 25,000 dollars per year between 1980 and 2000. Income Between $25000 and $50000 Per Year The number of people reported having an income between 25,000 and 50,000 dollars per year. Percentage Change in Income Between $25000 and $50000 Per Year Between 1980 and 2000 The percentage change in the num ber of people reported having an income between 25,000 and 50,000 dollars per year between 1980 and 2000. Household Income Greater Than $50000 Per Year The number of people reported havi ng an income greater than 50,000 dollars per year. Percentage Change in Household Income Greater Than $50000 Per Year Between 1980 and 2000 The percentage change in the num ber of people reported having an income greater than 50,000 dolla rs per year between 1980 and 2000. Median Income The median income reported for the census-tract. Percentage Change in Median Income Between 1980 and 2000 The percentage change in the median income reported for the census-tract between 1980 and 2000. Above Poverty Level Ages 15 to 64 Years The number of people between the ages of 15 and 64 years reported as being above the poverty level. Percentage Change in Above Poverty Level Ages 15 to 64 Years Between 1980 and 2000 The percentage change in the num ber of people between the ages of 15 and 64 years reported as being above the poverty level between 1980 and 2000. Above Poverty Level Ages Over 64 Years The number of people 65 years and older reported as being above the poverty level. Percentage Change in Above Poverty Level Ages Over 64 Years Between 1980 and 2000 The percentage change in the num ber of people 65 years and older reported as being above the poverty level between 1980 and 2000. Below Poverty Level Ages 15 to 64 Years The number of people between the ages of 15 and 64 years reported as being below the poverty level.

PAGE 139

128 Table A-1. continued Percentage Change in Below Poverty Level Ages 15 to 64 Years Between 1980 and 2000 The percentage change in the num ber of people between the ages of 15 and 64 years reported as being below the poverty level between 1980 and 2000. Total Housing Units The total number of housing units in the census-tract. Percentage Change in Total Housing Units Between 1980 and 2000 The percentage change in the tota l number of housing units in the census-tract between 1980 and 2000. Occupied Housing Units The number of o ccupied housing units in the census-tract. Percentage Change in Occupied Housing Units Between 1980 and 2000 The percentage change in the number of occupied housing units in the census-tract between 1980 and 2000. Vacant Housing Units The number of vacant housing units in the census-tract. Percentage Change in Vacant Housing Units Between 1980 and 2000 The percentage change in the num ber of vacant housing units in the census-tract between 1980 and 2000. Housing Units For Recreational Use The number of housing units that are only used seasonally or for recreation purposes in the census-tract. Percentage Change in Housing Units For Recreational Use Between 1980 and 2000 The percentage change in the number of housing units that are only used seasonally or for recreation purposes in the censustract between 1980 and 2000. Owner Occupied Housing Units The number of housing units that occupied by the owner. Percentage Change in Owner Occupied Housing Units Between 1980 and 2000 The percentage change in the number of housing units that occupied by the owner between 1980 and 2000. Renter Occupied Housing Units The number of housing units that are occupied by renters. Percentage Change in Renter Occupied Housing Units Between 1980 and 2000 The percentage change in the number of housing units that are occupied by renters between 1980 and 2000. Gross Rent Greater Than $500 Per Month The number of people reported that their gross rent is 500 or more dollars per month. Percentage Change in Gross Rent Greater Than $500 Per Month Between 1980 and 2000 The percentage change in the num ber of people reported that their gross rent is 500 or more do llars per month between 1980 and 2000. Median Gross Rent The median gross rent reported for the census-tract.

PAGE 140

129 Table A-1. continued Percentage Change in Median Gross Rent Between 1980 and 2000 The percentage change in the me dian gross rent reported for the census-tract between 1980 and 2000. Aggregate Value of Housing Units The aggregate value of all housing units in the census-tract. Name Description Percentage Change in Aggregate Value of Housing Units Between 1980 and 2000 The percentage change in the a ggregate value of all housing units in the census-tract between 1980 and 2000. Median Owner Costs with Mortgage The median owner cost with mortgage per month reported for the census-tract. Percentage Change in Median Owner Costs with Mortgage Between 1980 and 2000 The percentage change in the median owner cost with mortgage per month reported for the cen sus-tract between 1980 and 2000. Mean Value of Housing Units The mean value for housing units reported in the census-tract. Percentage Change in Mean Value of Housing Units Between 1980 and 2000 The percentage change in the mean value for housing units reported in the census-tract between 1980 and 2000. Total Square Feet of Commercial Land use by Designated Year The total area in square feet calculated for the commercial land use designated area(s) in the census-tract. Percentage Change in Total Square Feet of Commercial Land use Between 1980 and 2000 The percentage change in the tota l area in square feet calculated for the commercial land use between 1980 and 2000. Total Square Feet of Residential Land use by Designated Year The total area in square feet calculated for the residential land use designated area(s) in the census-tract. Percentage Change in Total Square Feet of Residential Land use Between 1980 and 2000 The percentage change in the tota l area in square feet calculated for the residential land use between 1980 and 2000. Road Distance with Accessibility Index to AFB The road distance Accessibility Index with the value of the distance to the AFB added. Road Distance without Accessibility Index to AFB The road distance Accessibility Index without the value of the distance to the AFB added. Road Distance with Accessibility Index to NAS The road distance Accessibility Index with the value of the distance to the NAS added.

PAGE 141

130 Table A-1. continued Road Distance without Accessibility Index to NAS The road distance Accessibility Index without the value of the distance to the NAS added. Road Distance with Accessibility Index to NS The road distance Accessibility Index with the value of the distance to the NS added. Road Distance without Accessibility Index to NS The road distance Accessibility Index without the value of the distance to the NS added. Straight-Line Distance with Accessibility Index to AFB The straight-line distance Accessibi lity Index with the value of the distance to the AFB added. Straight-Line Distance without Accessibility Index to AFB The straight-line distance Accessibi lity Index without the value of the distance to the AFB added. Straight-Line Distance with Accessibility Index to NAS The straight-line distance Accessibi lity Index with the value of the distance to Jax NAS added. Straight-Line Distance without Accessibility Index to NAS The straight-line distance Accessibi lity Index without the value of the distance to Jax NAS added. Straight-Line Distance with Accessibility Index to NS The straight-line distance Accessibi lity Index with the value of the distance to Mayport NS added. Straight-Line Distance without Accessibility Index to NS The straight-line distance Accessibi lity Index without the value of the distance to Mayport NS added. Britton Plaza Britton Plaza in Tampa (census-tract 67). Sports Complex Location of Legend’s Field and Raymond James Stadium, Tampa (Census-tract 26). Westshore Mall The Westshore Mall in Tampa (census-tract 46). Center Mall The location of Tampa Bay Center Mall (census-tract 27). Commercial Area with Census-tract ID and the Year CMR designates the commercial area. The numerical id in parenthesis identifies the census-tract that the commercial area is located. The number following th e underscore is the year that the data is given for the commercial area.

PAGE 142

131 LIST OF REFERENCES Allen W., Liu D., Singer S., 1993, “Accessibility Measures of U.S. Metropolitan Areas,” Transept. Res., Part B: Methanol. 27B(6), pp.439–449. Aschauer D., 1990, “Is Government Spending Stimulative?” Contemporary Policy Issues, 8(4), 1990, pp.30-46. Banister D., Berechman J., Transport I nvestment and Economic Development, UCL Press, London, 2000. Business Executives for Nati onal Security, May 1997, “The Case for Military-base Closures,” BENS Hill Advisory, (Business Executives for National Security 0597), http://www.bens.org/releases_0597.html May 27, 1997. Berechman J., 2001, “Transport Investment and Economic Development: Is There a Link?” Prepared for the European Confer ence of Ministers of Transport, ECMT, 119th Round Table on Transport and Economic Development. Black J., Conroy M., 1977, “Accessibility Meas ures and the Social Evaluation of Urban Structure”, Environmen tal Planner, A 9, pp.1013–1031. Boarnet M., 1996, “The Direct and Indirect Economic Effects of Transportation Infrastructure”, University of Califor nia Transportation Center Working Paper 340, 1996, Univ. of California, Berkeley, Calif. Brett J, March 2004, “Base Closures W ould Weaken Regional Economy,” Boston Business Journal, http://boston.bizjourn als.com/boston/stories/ 2004/03/22/editorial5.html March 19, 2004. Burgess E., 1924, “The Growth of the City: An Introduction to a Research Project,” American Sociological Society, 18, p. 86. Burt J., Barber G., Elementary Statistics for Geographers, The Guilford Press, New York, 1996. Christaller W., Die Ze ntralen Orte in S ddeutschland, Gustav Fischer Verlag, Jena, Germany, 1933. Baskin C., Translated, Central Places in Southern Germany, Prentiss-Hall, NJ, 1966.

PAGE 143

132 Clay J., Stuart A., Walcott W., 1988, “Jobs Highways and Regional Development in North Carolina”, Institute for Transporta tion Research and Education, Charlotte, N.C. 1988 Cockrell C, December 1998, “Lo cal Military-base Closures Leave Civilian Workers with Re-employment Difficulties, Says New UC Berkeley report,” University of California-Berkeley (Public Affairs), http://www.berkeley.edu/news/medi a/releases/98leg acy/12-8-1998a.html December 8, 1998. D'Agostino, R.B., Belanger, A., D'Agosti no, R.B. Jr. 1990."A Suggestion for Using Powerful and Informative Tests of No rmality.", The American Statistician, November 1990, Volume 44 Number 4, pages 316-321. This tutori al style article discusses D'Agostinos tests and tells how to interpret normal probability plots. Dardia M, McCarthy K F, Malkin J, Vernez G, February 1996, “The Effects of Militarybase Closures on Local Communities: A S hort-Term Perspective,” Rand Research Brief (MR-667-OSD), http://www.rand.org/publications/RB/RB7511/RB7511.html February 1996. Davis S, Haltiwanger J, Schuh S, Job Crea tion and Destruction, Cambridge and London: MIT Press, 1996. Deakin E., 1991, “Jobs, Housing, and Tran sportation: Theory and Evidence on Interactions between Land Use and Tran sportation”, Transportation, Urban Form, and the Environment, Spec. Rep. 231, Tran sportation Research Board, Washington, D.C., 1991. Dunbar S, Evolution of the Base Closure Pr ocess: The Struggle to Keep “The Sticky Fingers of Politics” Out, Course Paper submitted to faculty of the National Defense University and National Wa r College (EComp 2000), http://www.ndu.edu/library/n2/n005603E.pdf 2000. ESRI, Getting to Know ArcGIS Deskt op, ESRI Press, Redlands, CA, 2001. Feinstein D, January 2004, “Wants California’ s Past Contributions in Previous Base Closures to Be a Factor,” Senator Fein stein Seeks New Criteria in Base Closure Decisions (US Senator Dianne Feinstein’s Webpage), http://feinstein.senate .gov/04Releases/r-brac1.htm January 29, 2004. Fik T, The Geography of Economic Developmen t: Regional Changes, Global Challenges, McGraw-Hill, Boston, 2000 Forkenbrock D., Road Investment to Fo ster Local Economic Development, Iowa University Public Policy Center, Iowa City, Iowa, 1990 Garreau J., Edge Cities, New Yo rk Times Books, New York, 1992.

PAGE 144

133 Hansen W., 1959, “How Accessibility Shapes Land Use”, Journal of American Institute Planners 25, pp. 73–76. Harris C., Ullman E., November 1945, “The Na ture of Cities”, Annals of the American Academy of Political and Social Science 142, pp. 7-17. Hogan G, February 1997, “Evaluation of Military-base Closure Alternatives,” Thesis submitted to faculty of Virginia Polyt echnic Institute and State University (ETDComp 1997), http://scholar.lib.vt.edu/theses/available/etd23261942972620/unrestricted/etd.pdf February 1997. Holman B, Project Director April 1995, “Military Bases: Analysis of DoD’s 1995 Process and Recommendations for Clos ure and Realignment,” Report to the Congress and the Chairman, Defense Ba se Closure and Realignment Commission (GAO-95-133), http://archive.ga o.gov/t2pbat1/154008.pdf April 14,1995. Holman B, Project Director, July 1996, “Military Bases: Potential Reductions to the Fiscal Year 1997 Base Clos ure Budget,” Report to Congressional Committees (GAO/NSIAD-96-158), http://www.gao.gov/archive/1996/ns96158.pdf July 15, 1996. Holman B, Project Director, November 1998, “Military Bases: Re view of DoD’s 1998 Report on Base Realignment and Closure,” Report to Congressional Committees (GAO/NSIAD-99-17), http://www.gao.gov/ar chive/1999/ns99017.pdf November 13, 1998. Holman B, Project Director April 1999, “Defense Reform Initiative: Organization, Status, and Changes,” Report to the Chairman, Subcommittee on Military Readiness, Committee on Armed Serv ices, House of Representatives (GAO/NSIAD-99-87), http://www.gao.gov/ar chive/1999/ns99087.pdf April 21, 1999. Holman B, Project Director, May 2001a, “DoD Financial Management: Integrated Approach, Accountability, and Incentives Are Keys to Effective Reform,” Testimony Before the Subcommittee Government Efficiency, Financial Management, and Intergovernmental Relations, Committee on Government Reform, House of Representatives (GAO-01-681T), http://www.gao.gov/new.items/d01681t.pdf May8, 2001. Holman B, Project Director July 2001b, “Military-base Cl osures: DoD’s Updated Net Savings Estimate Remains Substantial,” Report to the Honorable Vic Snyder House of Representatives (GAO-01-971), http://www.gao.gov/new.items/d01971.pdf July 31, 2001.

PAGE 145

134 Holman B, Project Director, August 2001c, “Military-base Closur es: Overview of Economic Recovery, Property Transfer, and Environmental Cleanup,” Testimony Before the Subcommittee on Government Efficiency, Financial Management, and Intergovernmental Relations, Committee on Government Reform, House of Representatives (GAO-01-1054T), http://www.gao.gov/new.items/d011054t.pdf August 28, 2001. Holman B, Project Director, March 2004, “M ilitary-base Closures: Observations on Preparations for the Upcoming Base R ealignment and Closure Round,” Testimony Before the Subcommittee on Readiness, Committee on Armed Services, House of Representatives (GAO-04-558T), http://www.fpmi.com/lerpress/d04558t.pdf March 25, 2004. Hooker M, Knetter M, 2001, Meas uring the Economic Effects of Military-base Closures, Economic Inquiry 39(4): 583-598 http://ei.oupjournals.o rg/cgi/reprint/39/4/583.pdf October 2001. Hoyt H., The Structure and Growth of Resi dential Neighborhoods in American Cities, Federal Housing Administration, Washington, D.C., 1939. Ingram D., 1971, “The concept of accessibil ity: A Search for an Operational Form,” Regional Studies, 5, pp.101–107. Jackson C, Day F, 1993, Locational Concentrati ons of Military Retirees in the United States, Professional Geographer 45(1): 55-65. Lockwood D, August 1999, “Military-base Closures: Time for Another Round?” Congressional Research Service Report for Congress (CRS Report RL30051), http://www.ncseonline.org/ NLE/CRSreports/Public/pub7.cfm?&CFID=13717015&CFTOKEN=28863539 August 3, 1999. Lockwood D, February 2000, “Military-ba se Closures: Where Do We Stand?” Congressional Research Service Report for Congress (CRS Report RL 30440), http://www.ncseonline.org/ NLE/CRSreports/Public/pub18.cfm?&CFID=13717057&CFTOKEN=50907795 February 18, 2000. Mayer H., Hayes C., Land Uses In American Cities, Park Press, Champaign, IL, 1983. Nelson A., Moody M., December 2000, “Effect of Beltways on Metropolitan Economic Activity,” Journal of Urban Pla nning and Development, 126, pp. 189-196 O’Neill J, July 1998, “Review of the Department of Defense on Base Realignment and Closure,” Letters sent to Congress me mbers Strom Thurmond, Carl Levin, Floyd Spence, Ike Skelton, Thomas Daschle, Trent Lott, and Richard Gephardt (Congressional Budget Office 1998), http://www.cbo.gov/showdoc.cfm?index=647&sequence=0 July 1, 1998. O’Sullivan A., Urban Economics, 2nd Ed., Irwin, Homewood, Ill., 1993

PAGE 146

135 Ozbay K., Ozmen-Ertekin D., Berechman J., June 2003, “Empirical Analysis of Relationship between Accessibility and Ec onomic Development,” Journal of Urban Planning and Development 129, pp. 97-119. Palm R., The Geography of American Cities, Oxford University Press, New York, Oxford, 1981. Payne-Maxie and Blayney-Dyett, The Land Use and Development Impacts of Beltways, U.S. Department of Trans portation, Washington, D.C., 1980 Peregrino D, March 2004, “Regions Milita ry Bases Add Billions to Incomes and Business,” University of Texas El Pa so Horizons (University Office of Communications), http://www.utep.edu/horizons/Mar2004Issu e/Mar04Pages/Mar04MilitaryImpact.ht ml March 2004. Perry D., Watkins A., The Rise of the Sunbe lt Cities, Sage Publication, Beverly Hills, Calif., 1977. Ravid R., Practical Statistics for Educators, University Press of America, Lanham, Maryland, 1994. Rostow W., The Stage of Economic Growth, Cambridge University Press, Cambridge, 1960. Rugg D., Spatial Foundations Of Urbanism Wm. C. Brown Company Publishers, Dubuque, Iowa, 1972. Sale K., Power Shift, Random House, New York, 1975. Schwalbe S, June 2003, “An Expose on Base Realignment and Closure Commissions,” Air & Space Power Chronicles (Chronicles Articles 2003), http://www.airpower.maxwell.af.m il/airchronicles/cc/schwalbe.html June 10, 2003. Singer N, Cromwell J, Kohn S, Dunson M, Glass W, Williams C, et.al. December 1996, “Closing Military Bases: An Interim Assessment,” CBO Papers (Congressional Budget Office 1996), ftp://ftp.cbo.gov/46xx/doc4665/1996Doc33.pdf December 1996. Singer N, March 1997, “Military-base Closures and Realignment Procedures,” Testimony before the Subcommittee on Military Installations and Facilities Committee on National Security U.S. House of Repr esentatives (Congressional Budget Office 1997), ftp://ftp.cbo.gov/42xx/doc 4291/1997doc16-Entire.pdf March 18, 1997. Twight C, Spring/Summer 1989, “Institutional Underpinnings of Parochialism: The Case of Military-base Closures,” The Ca to Journal (Volume 9 Number 1, 1989), http://www.cato.org/pubs/j ournal/cj9n1/cj9n1.html Spring/Summer 1989.

PAGE 147

136 US Census Bureau, 1980, STF3 Files, US Government Press, Washington DC. US Census Bureau, 1990, STF3 Files, US Government Press, Washington DC. US Census Bureau, 2000, STF3 Files, US Government Press, Washington DC. Vickerman R., Spiekermann K., Wegener M., 1999, “Accessibility and Economic Development in Europe”, Regional Studies, 33, 1, pp. 1–15. Weinstein B., Firestine R., Regional Growth and Decline in the United States: The Rise of the Sunbelt and the Decline of the Northeast, Praeger, NY, 1978. Wiggins J, Project Director, April 1996, “M ilitary Bases: Closure and Realignment Savings Are Significant, but Not Easily Quantified,” Report to the Chairman, Subcommittee on National Security, Internat ional Affairs, and Criminal Justice, Committee on Government Reform and Ov ersight, House of Representatives (GAO/NSIAD-96-67), http://www.defenselink.mil/brac/docs/gaosave96-2.pdf April 8, 1996. Wu F., 1998, “Polycentric Urban Developmen t and Land Use Change in a Transitional Economy: The Case of Guangzhou,” Environmental Planner, A, 30, pp.1077–1100. Yeates M., Garner B., The North American City, Harper and Rowe Publishers, New York, 1976.

PAGE 148

137 BIOGRAPHICAL SKETCH Born in Elizabethtown, Kentucky, on Ma rch 8, 1962, I moved to Frankfort, Kentucky at age 15. I graduated from Fr anklin County High School in May 1980. I enlisted in the Kentucky Army National Guar d in 1981 and began an 11-year career in the military. In December 1982, I enlisted in the United States Army where I began my active military career. Over the next 10 year s I was stationed in various military installations beginning in Fo rt Carson, Colorado; West Germany (FRG); Sam Houston, Texas; Fort Drum, New York; and ending my mi litary career in Kansas City, Missouri in 1992. My duties in the Army encompassed severa l leadership positions with an emphasis as a medical specialist with f light training. While in the Army my first two children were born including Rachel Anne Hawkins (born on March 26, 1986 in Stuttgart, FRG); and Beatrice Rose (born on September 29, 1989 in Watertown, New York). After my divorce in 1994, I managed several retail establishments and returned to college to complete a degree that I had begun in 1980. During this time, I met my current wife ; and my third child, Quinton Zachary Vitelli-Hawkins, was born on Augus t 25, 1999. In May 2000, I graduated from Tennessee State University, Summa Cum Laude, w ith a B.A. in History with licensure to teach. In the summer of 2001, I was selected as a Teacher Educator for the US Space Program at the US Space and Rocket Center in Huntsville, Alabama. I was awarded a Master of Science degree in geography (w ith emphasis on geographic technologies) from the University of Florida in May 2005.


Permanent Link: http://ufdc.ufl.edu/UFE0010488/00001

Material Information

Title: Military-Base Impact on a Local Economy: A Case Study of Three Military Bases in Two Metropolitan Statistical Areas
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010488:00001

Permanent Link: http://ufdc.ufl.edu/UFE0010488/00001

Material Information

Title: Military-Base Impact on a Local Economy: A Case Study of Three Military Bases in Two Metropolitan Statistical Areas
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010488:00001


This item has the following downloads:


Full Text












MILITARY-BASE IMPACT ON A LOCAL ECONOMY: A CASE STUDY OF
THREE MILITARY BASES IN TWO METROPOLITAN STATISTICAL AREAS















By

KENNETH E. HAWKINS


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Kenneth Eugene Hawkins
































This thesis is dedicated to my three children; Rachel Anne, Beatrice Rose, and Quinton
Zachary with the hope of inspiring them to achieve higher goals than mine.















ACKNOWLEDGMENTS

I thank my wife Maria for her support, understanding, and help during the time of

separation and hardships, especially since she also earned her Doctorate while working

and keeping our son. I thank my mother and Aunt Faye for instilling in me the

importance of an education and the pursuit of my dreams. I thank my mother-in law, Dr.

Veronica Free, for the advice and support from a professional educator's perspective. I

thank my father-in-law and step-mother-in-law (Drs. Mario and Carol) for their support

and assistance during my years in the graduate program at the University of Florida.

I thank the professors in the Department of Geography and the office assistants for

their time and effort in helping me through the years in graduate school. I thank Dr. Paul

Zwick (a member of my supervisory committee) and Professor Stanley Latimer. Their

classes and wisdom in uses of geostatistical analysis and spatial analysis were

instrumental in my research. I thank Dr. Jane Southworth for her assistance and advice as

a member of my supervisory committee, and for her classes in remote sensing. Finally, I

thank my mentor, tutor, and inspiration in studying quantitative methods: Dr. Timothy

Fik. I hope to do him justice with my work and will remember the efforts and assistance

he gave me for many years.
















TABLE OF CONTENTS


page

A C K N O W L E D G M E N T S ............................................. .............................................. iv

L IST O F T A B L E S ......... ............................. .......... ... .............. .. vii

LIST OF FIGURES ............. ..... ........................ ......... ............... viii

A B STR A C T ................................................. ..................................... .. x

CHAPTER

1 INTRODUCTION AND LITERATURE REVIEW ....................................................1

Base Impact on Urban and Economic Growth of the Localized Area ...............1......1
Problems Associated with Base Realignment and Closures.................. ...........
Base Realignment and Closure, Economic Impact, Accessibility, Military Bases,
and the Central Business District ................................... ........................... ......... 5
Determining Base Realignment and Closure (BRAC)............... ...................5
Concern over Economic Impact ............. ...................... .. ............... 19
Alleviating Negative Economic Impact with Accessibility .............................20

2 M E T H O D S ......................................................... ................ 2 3

Economic Growth and Development because of Military Bases and the Central
Business District: Economic Base Theory and Accessibility ..............................23
Military Bases, Central Place Theory, and the Central Business District
(C B D ) ...................................... ............................................... 2 3
Economic Base Theory........................................... .. .......................24
Methods for a New Approach to Predicting Economic Impact from BRAC.............25
M military Bases Applied to Economic M odels ........................................ ..........25
Previous Economic Impact Studies from Military-base Closures.....................28
New Approach to Predicting Economic Impact from BRAC ..................................34
Base Realignment and Closure (BRAC) 2005 and Florida's Bases ....................34
Jacksonville and Tampa, Florida.................. .... .... ............................35
Jacksonville NAS and Mayport NS Study Areas and Diversity of Industrial
Em ploym ent: 1980 and 2000 ............................................... ........ ....... 37
M acD ill AFB and Tam pa, Florida......................... ....... ................... .... 38
Recovering from Economic Impact Assuming Tampa or Jacksonville Bases
A re Selected for BRA C 2005 ............................................... ............... 43









Softw are U sed for Our Study ........................................ ......................... 43

3 MAPPING URBAN GROWTH AND CHANGE .............................................. 44

Software Used to Create the Images for the Study Areas .......................................45
Creating Shapefiles for the Study Areas................................... ....... ............... 48
Establishing Study Area Boundaries Using Shapefiles ............... ....... ...........49
The Jacksonville Metropolitan Statistical Area.............. ...... .............. 49
M ayport NS Study Area ................................ .. .... ............... 52
Jacksonville Naval Air Station (Jax NAS) .................................. ............... 54
Tam pa-M acDill AFB Study Area ............................................ ............... 59
Empirical Approach to Study Areas..................... ....... ...................... ............... 62
Dependent Variables and Each Study Area............................................. 62
Independent Variables and Each Study Area.....................................................63
Results of Choropleth Maps for the Jax NAS Study Area..................................65
Results of Choropleth Maps for the Mayport NS Study Area...........................66
Results of Choropleth Maps for the MacDill AFB Study Area ........................68
Review of Choropleth M aps Results............................................... .............. 69

4 REGRESSION ANALYSES OF THE STUDY AREAS .......................................84

Jacksonville NAS Study Area Regression Analyses Results ...................................85
Mayport NS Study Area Regression Analyses Results...........................................89
MacDill AFB Study Area Regression Analyses Results................ .............. ....92

5 SUM M ARY AND CONCLU SION ................................... ................................... 112

S u m m a ry .......................................................................................1 12
Concluding Remarks .................. ........................................ .... ......... 116

APPENDIX KEY TO INDEPENDENT VARIABLES................................................120

L IST O F R E F E R E N C E S ......... .. ............... ................. .............................................. 13 1

BIOGRAPHICAL SKETCH .............. ........... .. ............. 137















LIST OF TABLES
Table page

2-1. Employment populations for Jacksonville Naval Air Station (Jax NAS) study
a r e a ............................................................................... 3 9

2-2. Employment populations for Mayport NS study area............................................40

2-3. Employment populations for MacDill AFB study area................ ..................42

3-1. K ey to dependent variables............................................................ .....................64

3-2. Jax N A S dependent variables .................................. ............... ............... 70

3-3. M ayport N S dependent variables. ........................................ ....................... 72

3-4. M acDill AFB dependent variables. ........................................ ...................... 73

4-1. Jacksonville NAS study area regression results for percentage change in
commercial area between 1973 and 2000. .................................... .................96

4-2. Jacksonville NAS study area regression results for percentage change in median
household income between 1980 and 2000................................... ............... 98

4-3. Jacksonville NAS study area regression results for percentage change in
residential area between 1973 and 2000. ..................................... ............... 99

4-4. Mayport NS study area regression results for percentage change in commercial
area betw een 1973 and 2000. ............................................................................ 102

4-5. Mayport NS study area regression results for percentage change in median
household income e between 1980 and 2000.................................... ............... 103

4-6. Mayport NS study area regression results for percentage change in residential
area betw een 1973 and 2000. ............................................................................ 105

4-7. MacDill AFB study area regression results for percentage change in median
household income e between 1980 and 2000.................................... ............... 107

4-8. MacDill AFB study area regression results for percentage change in residential
area betw een 1980 and 1999. ............................................................................ 109

A-1. Key to Independent Variables. ........................................ ......................... 120















LIST OF FIGURES


Figure page

2-1. Incom e flow s. .......................... ........................... .......................25

3-1. Jacksonville Metropolitan Statistical Area (Image created using FGDL and
C ensus B bureau D ata) .......................... ...................... ... ......... ........... 46

3-2. Mayport NS study area census-tract boundaries. .................... ............................. 50

3-3. MacDill AFB study area census-tract boundaries. .................................................50

3-4. Jacksonville NAS study area census-tract boundaries. ...........................................51

3-5. Modes of transportation in the Mayport NS study area ..........................................53

3-6. Urban and economic growth in Mayport NS study area (1973-2000). ...................54

3-7. Transportation routes to Jax N A S. ........................................ ........................ 56

3-8. Residential growth in the Jax NAS study area (1973-2000). ...................................57

3-9. Commercial growth in the Jax NAS study area (1973-2000). ..................................58

3-10. T am pa M SA ...................................................... ................. 6 1

3-11. Modes of transportation accessible to MacDill AFB. .............................................63

3-12. Commercial area percentage change 1973-2000 for Jax NAS study area. .............75

3-13. Residential area percentage change 1973-2000 for Jax NAS study area ...............76

3-14. Median household income percentage change 1980-2000 for Jax NAS study
area .................................................................................77

3-15. Commercial area percentage change 1980-2000 for Mayport NS study area.........78

3-16. Residential area percentage change 1973-2000 for Mayport NS study area...........79

3-17. Median household income percentage change 1980-2000 for Mayport NS
stu d y a re a ...................................... ................................. ................ 8 0

3-18. Commercial area percentage change 1980-1999 in the Tampa study area ............81










3-19. Residential area percentage change 1980-1999 in the Tampa study area. ..............82

3-20. Median household income percentage change 1980-2000 of the Tampa study
area. ................................................................................ 83















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

MILITARY-BASE IMPACT ON A LOCAL ECONOMY: A CASE STUDY OF
THREE MILITARY BASES IN TWO METROPOLITAN STATISTICAL AREAS

By

Kenneth E Hawkins

May 2005

Chair: Timothy Fik
Major Department: Geography

Military bases have had a profound impact on urban and regional economic growth.

Our objective was to assess the impact of military bases on local economic-growth rates

in three distinct metropolitan statistical areas (MSAs). We examined the extent to which

proximity to a military-base or the Central Business District (CBD) affects local

economic growth rates and the degree to which variability in growth is explained by the

distance to a base.

Our study population included three metropolitan statistical areas: Jacksonville

Naval Air Station (NAS) and Mayport Naval Station (NS) located in Jacksonville,

Florida, and MacDill Air Force Base (AFB) located in Tampa, Florida. We used spatial

analysis and multiple regression analysis to determine a discernable impact on economic-

growth rates of the localized areas from each of the three military bases at the 95%

confidence level. Our hypothesis was that military bases have a discernible impact on

economic growth rates at a geographical scale (census-tract level) because of proximal









distances and accessibility along transportation corridors between the base and major

commercial and financial nodes at the 95% confidence level.

Spatial analyses showed discernible impact on the economic growth rates of the

study areas; however, the cause of economic growth is not discernible among impact

from the base, commercial nodes, economic nodes, demographics, distance variables and

accessibility variables in the study areas. Regression analyses revealed possible positive

and negative causes for economic-growth rates of the study areas. However, the

significance of base impact on economic-growth rates was negligible in all cases. The

study areas showed no evidence of localized spillover effects.














CHAPTER 1
INTRODUCTION AND LITERATURE REVIEW

Base Impact on Urban and Economic Growth of the Localized Area

Military bases have had a profound impact on urban and regional economic growth.

Our objective was to investigate to what extent variability in urban and economic growth

rates are explained by proximity to important nodes in 3 distinct study areas. Specifically,

we investigated the degree to which variability growth ratios are explained by distance to

a military-base vs. other prominent nodes in the metropolitan areas (e.g. CBD,

commercial, residential, and industrial) by spatial relationships.

Our hypothesis was that military bases have a discernible impact on economic-

growth rates at a geographical scale (census-tract level) because of proximal distances

and accessibility along transportation corridors between the base and major commercial

and financial nodes at the 95% confidence level.

Problems Associated with Base Realignment and Closures

Base Realignment and Closures (BRAC) have occurred since the formation of the

Department of Defense (DoD). Before the 1980s, BRAC was a defense function handled

primarily by the DoD. The DoD did not have to answer to any governmental offices

when BRAC actions were considered. After the Vietnam War, Congress began to get

involved because they felt that BRAC was targeted toward those states that did not fall

into line with defense spending. Several congressional members began organizing to

prevent BRAC; and by the late 1970s, Congress passed regulations that prevented the

DoD from approving any BRAC actions without congressional authorization.









Congressional intervention prevented further BRAC actions until 1988. The events that

occurred between 1981 and 1988 concerning the United States and the Soviet Union led

to the eventual collapse of the Soviet Union and a change in the DoD's mission and

planning. The DoD had to consider the downsizing of personnel and the ability to rapidly

respond to changing military and strategic needs. The development of new weapon

systems and training led to a "modernization" of the DoD and a reevaluation of their

infrastructure.

The DoD had to deal with another problem from Congress while modernizing. By

1988, defense spending had been an integral part of the Reagan administration's plan on

crippling the Soviet economy. The intense defense development programs during the 8

years of Reagan succeeded in bringing an end to the 40-year-old Cold War. Problems

associated with the post-Cold War defense spending brought about a change in Congress

and cuts in defense budgets. The DoD argued that cuts in their budget would affect

development of training, and weapons, and the creation of rapid deployment forces.

Congressional leaders argued that downsizing military strategies would include a review

of weapons development programs and potential base closures. The DoD argued that

some bases were no longer vital to the new mission of post-Cold War defense, leading to

new rounds of BRAC. Eventually, Congress agreed that there was a need to consider

BRAC actions.

The new rounds of BRAC were considerably different than BRAC actions before

the 1980s. The new rounds of BRAC that Congress agreed to would be more of a

congressional function than BRAC actions of the past. Although the DoD would have a

say as to what bases should be considered, a commission would be established to review

the bases and make recommendations to Congress. BRAC would evolve over the next 7









years because of differences between Congress and the DoD; however, intervention in

the 1995 round by President Clinton prevented any further BRAC action until 2005. The

Government Accounting Office (GAO) and the Congressional Budget Office (CBO) are

the two primary organizations that review and report the progress of previous BRAC

actions and serve as part of a check-and-balance system between the DoD and Congress.

The problem is that both the DoD and Congress have alternative agendas concerning

BRAC. The DoD needs to close unnecessary bases; however, Congress is concerned

about economic impact from base closures. Although, there is a need to close bases,

political considerations come into play. Congress paints a gloom-and-doom picture of

base closures (which previous research has shown is not necessarily true); and the DoD

has a valid argument that base reduction has not matched the reduction in force, and that

reduction in costs for unnecessary bases can be effectively applied to other budgetary

concerns.

Preparation of BRAC in cities and metropolitan statistical areas (MSAs) with bases

can prevent or reduce the effects of economic impact. When a new round of BRAC is

announced, federal, state, and local government officials actually try to prevent base

closures in their respective areas. Unfortunately, the local population believes in the

gloom-and-doom scenario painted by Congress and will do anything they can to prevent

the closure of bases to protect their local economics. Much preparation and planning must

be done to prepare for the economic impact that occurs with base closure and/or

realignment. City planning and commerce officials, and other business and government

organizations can reduce the negative economic impacts with proper organization and

planning. The land use associated with the loss of a base can be replaced with new

businesses and industries by rezoning. If substitute industries are situated to move in,









when the military moves out, the impact should be substantially reduced. There is one

other useful alternative of proper planning and organization: the prevention of BRAC.

One of the criteria that the DoD and BRAC commissions consider when choosing bases

for BRAC actions is the urban growth of the local area approaching the outer boundaries

of the base (also known as encroachment). The DoD's major concern about

encroachment on a base's boundaries mainly arises with bases that have airfields.

Complaints of noise pollution and other civilian issues generally cause problems for the

military in the local area. The DoD has argued that local planning and commerce officials

should keep in mind the problems associated with encroachment and other civilian issues

when determining future land use in the local areas.

BRAC assessments need to (and usually do) consider the economic impact of the

local area. A key to preventing real dangers to local communities and economic impact is

to consider the size of the urban area and the diversity of its economy; most BRAC

actions in the past have involved large military bases in or near MSAs. Recent studies

(Dardia et al. 1996, Cockrell 1998, Hooker and Knetter 2001) that the economic impact

in past areas have not been as severe as predicted. Recovery of the areas has been swift

and their economies have typically recovered fairly rapidly. The areas affected were

typically in major MSAs with diverse industries that aided in economic recovery. Future

BRAC assessments need to consider the impact of bases in smaller urban areas (such as

Jacksonville, NC; and Manhattan, KS) whose industries are predominantly service-

related and thrive on the military presence. BRAC actions in these types of communities

can be devastating. Accessibility of bases to the central business district (CBD) may also

play a vital role in how urban planning and economic growth rates in areas near bases









may be potentially affected by base closures. Land use before, during, and after BRAC

needs to be considered if negative economic impacts must be reduced.

Base Realignment and Closure, Economic Impact, Accessibility, Military Bases, and
the Central Business District

Determining Base Realignment and Closure (BRAC)

The first BRAC Commission was formed in December 1988 and Congress

authorized a new round of closures. Reminded of the difficulties of base closures in

previous years, Congress acted by passing legislation that would allow the DoD to close

bases, yet allow Congress to keep control of the proceedings and actions involving these

base closures. The Defense Base Closure and Realignment Act of 1990 was passed to

correct deficiencies from the first round of BRAC in1988. Many congressional members

were not overly enthused with the selections made in 1988, particularly those from

California. Since 1988, economic impact of the urban areas affected by BRAC

recommendations have caused resistance from Congress and individual states' legislative

bodies in an era where it was necessary to reduce the military infrastructure which meet

the post Cold War goals of the Department of Defense (DoD). There are a great number

of government documents from the Government Accounting Office (GAO) and

Congressional Budget Office (CBO) that express in detail the concerns of the DOD and

Congress and economic impact from BRAC recommendations. The GAO publications

primarily discuss the importance of the reduction in military infrastructure. Yet the

concern of economic impact because of BRAC is discussed in the same GAO

publications. According to the GAO (Holman 1995, 1996, 1998, 1999, 2001a, 2001b,

2001 c, 2004, Wiggins 1996) the increases in the DOD budget and spending during the

1980s led to significant changes in the future of military planning and policy. The DoD









reevaluated the mission objectives and determined that excessive spending could be

reduced in some areas and redirected to meet the requirements for spending on research

and development (R&D) of better weapons systems, equipment and improved readiness

of units. Once the DoD determine what bases or functions on a base were unnecessary, a

list of those facilities and functions were drawn and proposed to BRAC. BRAC later

released those reports to Congress and the Congressional Budget Office (CBO) as well as

the Government Accounting Office (GAO). Recommendations were made to determine

the effect of cost reduction for the DoD and the negative impact on the economies of

urban areas in close proximity to the bases targeted for closure or realignment.

Of importance was the impact on the employment of the local community, real

estate values, and future use of the land returned to the urban region. The DoD is

interested in the cost to maintain, train and operate the facility and its personnel. Another

issue that DoD takes into consideration is the urban region's growth and the

encroachment of growth toward the base and the subsequent; particularly bases with

aviation units that require flight training and operations. Virtually no attention, however,

is paid to the spatial relationships between military bases and the Central Business

Districts and other important nodes in the urban area. Moreover, the economic impact of

military personnel leaving the urban region and the loss of linked income is of secondary

concern. The studies need to look much further than just the loss of jobs in the civilian

sector and changes in real estate values, as future business opportunities from the

availability of land may play a vital role in the recovery process once a base is closed.

Urban planners must control the potential growth in areas progressing into encroachment

on the fringe of military bases, and plan in accordance to remove the possibility of

encroaching land uses. The first step in planning is to determine the location of a base's









aviation component and second the base's training areas. By creating a buffer or no

growth zone between the urban land use and the base would improve a region's chance of

surviving possible base closure. A good example of land use encroachment can be seen in

the MacDill Air Force Base study area because the base is centrally located in the Tampa

MSA. Jacksonville Naval Air Station has not been as fortunate as urban growth from

Orange Park (south of the base) and Jacksonville (north of the base) have surrounded the

boundaries of the base. Fortunately, the base's airfield is located away from the urban

growth and near the St. John's River. While urban sprawl has occurred in the area

surrounding the base, the effects of this urban growth pattern has been beneficial to the

base because of the real estate values and the industrial and commercial activity near the

base. Thus, encroachment has not been a major problem even though the urban growth

reached the boundaries of the base. One final factor that prevents recommendation for

closure is the mission of the base: it is home to the air wings that support the carrier

group based at Mayport NS. Closure of Jax NAS could occur, however, the space

required for the base's mission is not sufficient at Mayport NS. Therefore, if the Jax NAS

closed, the mission of Mayport NS would be adversely affected. Without air support for

the carrier group stationed at Mayport NS, the mission of the base would be weakened

enough for base closure consideration.

By 1995, the DoD had established a list of criteria that appeared to satisfy both

sides. The list of DoD criteria was published by the GAO (GAO-95-133, 1995) and is the

basis for the DoD's recommendation for future rounds of BRAC. According to the GAO

the list of DoD criteria is broken into three categories with individual criteria for base

selection in each category.

DoD Criteria for Selecting Bases for Closure or Realignment:










* Military Value (priority consideration is to be given to the four military value criteria)

Current and future mission requirements and the impact on operation
readiness of DoD's total force.

The availability and condition of land, facilities, and associated airspace at
both the existing and potential receiving locations.

The ability to accommodate contingency, mobilization, and future total force
requirements at both the existing and potential receiving locations.

Cost and manpower implications.

* Return on investment

The extent and timing of potential costs and savings, including the number of
years, beginning with the date of completion of the closure or realignment, for
the savings to exceed the costs.

* Impact

The economic impact on communities.

The ability of both the existing and potential receiving communities'
infrastructures to support forces, missions, and personnel.

The environmental impact.

Once the bases are recommended on these criteria, the recommendations are presented to

the BRAC Commission for further investigation. The final BRAC recommendations are

then presented to legislative and executive branches for approval.

Although the process seems rather simple, many factors can disrupt the process

beginning with the DoD selections. A GAO report (Holman 1995) stated that the DoD is

sensitive to the economic impact on affected communities; however, later GAO reports

(Holman 2001c, Holman 2004) discussed the importance economic impact has in DoD's

selection criteria, but military value has top priority in all selection processes. The GAO

reported (Holman 2004) a list of requirements that the DoD adopted to establish a

guideline for all services and DoD agencies to implement in the base selection process:









* All installations must be compared equally against selection criteria and a current
force structure plan developed by the Secretary of Defense.

* Decisions to close military installations with authorization for at least 300 civilian
personnel must be made under the BRAC process. Decisions to realign military
installations authorized for at least 300 civilian personnel that involve a reduction of
more than 1,000 or 50 percent or more of the civilian personnel authorized, also must
undergo the BRAC process.

* Selection criteria for identifying candidates for closure and realignment must be made
available for public comment before being finalized.

* All components must use specific models for assessing (1) the cost and savings
associated with BRAC actions and (2) the potential economic impact on communities
affected by those actions.

* Information used in the BRAC decision-making process must be certified-that is,
certified as accurate and complete to the best of the originator's knowledge and
belief. This requirement was designed to overcome concerns about the consistency
and reliability of data used in the process.

* An independent commission is required to review DoD's proposed closures and
realignments and to finalize a list of proposed closures and realignments to be
presented to the President and, subject to the President's approval, to Congress.

* The BRAC Commission is required to hold public hearings.

* The BRAC process imposes specific time frames for completing specific portions of
the process (see app. I for time frames related to the 2005 BRAC round).

* The President and Congress are required to accept or reject the Commission's
recommendations in their entirety.

* In addition to GAO's role in monitoring the BRAC process, service audit agencies
and DoD Inspector General (IG) personnel are extensively involved in auditing the
process to better ensure the accuracy of data used in the decision-making and enhance
the overall integrity of the process.

One requirement caused some previous complications with the BRAC process: the

acception or rejection of the BRAC Commission's recommendations in their entirety by

both the President and Congress. Apparently Congress gives the final endorsement for all

BRAC recommendations. Twight (1989) gave an insight to congressional actions

concerning BRAC recommendations. According to Twight (1989), congressional









members appear to be more concerned with their own political ambitions than addressing

the issue of needed BRAC actions, which according to the previously mentioned GAO

and CBO reports, Congress agreed that there is a need for BRAC. Twight (1989)

explained the question of congressional decision-making in the BRAC process through

several examples such as prior authority for base closures and realignment before 1988,

actions that occurred because of recommended base closures before 1988, and

congressional actions that occurred to refuse the recommendation of DoD proposed

BRAC actions before 1988. Schwalbe (2003) described the actions and attitudes of

Congress, the Presidency, and DoD during the recent BRAC rounds and supported

Twight's research.

Another problem that needs to be addressed is the education of the general public

concerning BRAC. As long as Congress defines the economic impact as being

devastating to local, regional, and state economies, the average citizen without a

grounded understanding of economic principle will believe that their own economic well-

being will be jeopardized with any BRAC actions in their community. Dardia et al.

(1996) conducted research for the RAND Institute that studied the effects of BRAC in

three communities in California between 1992 and 1995. When the bases in the study

(Fort Ord, George AFB, and Castle AFB) were selected for closure many California

legislators at both state and federal levels tried in vain to prevent the closures. The study

eventually concluded that predicting economic impact is extremely difficult, especially in

the before closure stages, and those predictions are not necessarily an end-all scenario as

they may appear for the communities affected, however, they also concluded that waiting

for long-term studies to be conducted is not reasonable. The answer to these problems is

very complex because the problem is very complex. The past history between the DoD









and Congress concerning BRAC is key to solving many of the problems that have arisen

in the previous four rounds of BRAC.

One goal was to meet the requirements of the Balanced Budget and Emergency

Deficit Control Act of 1985. The Senate bill S.1702, sponsored by Senators Gramm,

Rudman and Hollins, called for the elimination of the federal budget deficit by cutting the

domestic spending by one-half and the defense spending by one-half Another mission

the DoD was reevaluating was the situation that was developing in the Soviet Union. The

increased spending on defense by the Reagan administration had repercussions in the

Soviet Union. In an attempt to keep up with American advances in military technology,

the Soviet Union was quickly creating a hazardous national situation because of increased

spending in their own military efforts to keep up with the United States. BRAC was

created for the purpose of assisting the DoD in determining what course of action would

be feasible in reducing costs and making defense spending more efficient. Dunbar (2000)

conducted research concerning BRAC while attending the National Defense University

and National War College. In Dunbar's report, the past history concerning Congress, the

DoD, and base closures reflects similar actions reported in the previously mentioned

references: Congress' inability to trust the DoD's recommendations for base closures and

realignments.

An interesting aspect of Dunbar's research was the DoD's method of base closure

before 1988. According to Dunbar, most base closures and realignments were conducted

to the nuclear response from the Soviet Union. The bases were made larger and placed

farther from larger Metropolitan Statistical Areas (MSAs) to prevent the annihilation of

population centers in the event of a nuclear war. Unfortunately the economic

development between an urban area and the military bases increased the ability of urban









growth between the two locations. Many cities that were near a military-base may have

been a small city or mid-sized city when realignment occurred thirty or forty years ago.

Some examples are the two areas chosen for our study: Tampa and Jacksonville, Florida.

According to the U.S. Census Bureau in 1970, the population for Duval County Florida

(Jacksonville) was 528,865 and Hillsborough County Florida (Greater Tampa) was

490,265. Referring to Rugg's (1972) research on urban growth patterns and the economic

development of MSAs, the two MSAs were in early stages of economic growth patterns

that resulted in a megalopolis in the Tampa area and Jacksonville's MSA spilled over into

surrounding counties (Clay, Nassau, and St. John's). The difference between the two

areas is the location of the military bases. The 1980 Census reports the population for

Duval County Florida as 571,003 and Hillsborough County as 646,960.

At this time the Tampa-St. Petersburg MSA had become a small megalopolis,

whereas the growth in Jacksonville shows a slight increase in the MSA' growth. During

this phase of Tampa's growth, MacDill AFB could have developed a significant impact

on Tampa's economy. The population of Hillsborough County in 1990 was 834,054 and

almost matched the population of Pinellas County reported at 851,659. By 2000 the

population of Tampa was 998,948 and Pinellas was 921,482, thus the population

Hillsborough County's ability to grow toward the east permitted the population to pass

that of Pinellas County, which had almost grown to the full extent of the county's

borders. Although there is no direct connection between MacDill AFB and either St.

Petersburg or Orlando, the significance of their interaction with Tampa and its' economic

base should be consistent with previous regional studies concerning base closure impact.

The significance of military bases being situated near major accessibility routes

increases their nonbasic activity value to the economic development of the local









community. For example, Orange Park in Clay County was closer to Jacksonville NAS

and Mayport NS was situated near the beaches area of Duval County. The growth

between both bases and Jacksonville and the growth between the bases and the smaller

communities developed into the Jacksonville MSA. Another factor that enhanced the

growth involved the earlier BRAC rounds closing several smaller Naval bases in Florida

and realignment to the two bases in Jacksonville.

The GAO detailed the problems involved in preparing the land and facilities for

exchange to the local communities in many of their reports (Holman 1997, 1998, 2001a,

2001c). Normally the average time for transfer of land can be between 3 to 7 years. Most

of the time consuming factors involved environmental cleanup; however, parcels of base

property are turned over when they are ready for use.

The means of overcoming a base closure and keeping the economic impact at a

minimum would rely on several factors: preparation of the base for closure, alternative

methods of turning property over at a quicker pace, privatization in place, finding new

environmental cleanup technology, and preparing the local economies for transition to

newer nonbasic activities to name a few. The most important aspect of preparing the local

community for a base closure is explaining that the economy is not going to be destroyed

as most politicians would have them believe. Often Congressional members agreed that

there may be a need to redirect defense spending from unnecessary costs to R&D and

improved readiness of forces, until the closure or realignment of a base is situated within

their congressional district. On learning which facilities and functions are revealed many

congressional members begin grass root operations to "save" the base from closure. The

most common method of organizing the grass root campaign is through an official

website of a federal or state legislator or in some instances the governor's office.









Another method of starting a grass root campaign is the website of a local

university's political science, journalism, or public relations department. Every state that

has a military-base has a website that is designed to prevent the closures of their bases by

presenting different studies conducted by local universities, institutions, or research

facilities. State or federal legislators sponsor most of the websites. For example, U.S.

Senator Dianne Feinstein has a webpage that discussed practically any congressional

concern for citizens of California. Interestingly, Feinstein has a history of voting for

defense budget cuts; however, her webpage (Feinstein 2004) has many sites that argue

against any base closures in California.

The fact remains that the DoD has a need to determine the means to reduce costs

and improve budget efficiency. Congress needs to understand that some bases have little

or no military value and are wasteful spending of tax dollars or if the base and its

functions are that important to the urban region then increase defense spending. However,

the purpose of BRAC is to save tax dollars in the first place and increasing the defense

budget to save bases and functions of bases is detrimental in more ways to the economy

of the whole instead of a small part of the state or nation. The GAO (Holman 2001a)

discussed in great detail the DoD's budget, costs, and savings because of BRAC.

The importance of Congress's role in the process is they give the final approvals of

all BRAC commission decisions. Congress has been difficult in approving past BRAC

recommendations. Congress denied the last two rounds of DoD requests for BRAC

commissions because of previous discrepancies. However, one of Congress' concerns

with BRAC was possible interference after the recommendations had been approved by

any party and outside the requirements established by the 1990 act. Schwalbe (2003)

explained in explicit detail the interference that occurred in 1995. President Clinton









interfered with the approved closing of two bases, which clearly was prohibited by

Congress. Although the interference did create a new concept (privatization in place) of

possibly protecting military bases in future BRAC rounds, the Congressional response

was overwhelmingly strong against any future BRAC considerations. The political

weapon of Congress has been the economic impact that is created by base closures.

Although the short-term effect can be disruptive and possibly devastating to the urban

areas that are affected by BRAC, studies have shown otherwise.

The upcoming BRAC round in 2005 will be quite different than the previous four

BRAC rounds. BRAC 2005 is designed to improve DoD budget spending; however, it

will include some significant changes that make it different that the previous four BRAC

rounds. Schwalbe (2003) characterized this new round of BRAC as the "Mother of all

BRACs" and explained Secretary of Defense Donald Rumsfeld's plans to cut the same

amount of surplus that was cut in the BRAC round of 1988 to 1995 combined. The 2005

round will also cut at least 25 percent of the DoD's remaining real estate. Based on

Public Law 107-107, Section 3000, there are some important details that are required

before the next BRAC round can occur. Before any recommendations can be made, DoD

has to prepare a force-structure plan. The Secretary of Defense concerning any possible

national security threats between 2005 and 2025 must base the plan on an assessment.

DoD must then submit that assessment to Congress detailing the inventory of DoD's

infrastructure based on the force-structure plan. The greatest impact for DoD and BRAC

requests occurred with the law amending the Defense Base Closure and Realignment Act

of 1990 significantly changing the selection criteria. DoD has been directed by Congress

in the upcoming BRAC round to make military value the primary consideration for

recommendation for BRAC action. DoD is to assign the bases values as before, however,









when making the recommendation the values are to be based on the following:

preservation of training and staging areas; preservation of military installations

throughout a diversity of climate and terrain areas in the U.S. for training purposes; high

consideration of joint war fighting, training, and readiness; and high consideration for

contingency, mobilization, and future total force requirements at locations that support

operations and training. The selection criteria based on military value was not only

changed, but also mandated as the primary factor for BRAC consideration, however,

selection criteria is not only relegated to military value. Other selection criteria include

the extent and timing of potential costs and savings for DoD, the economic impact on the

local communities affected, the local communities being able to support additional

infrastructure and forces, and factors concerning environmental costs for cleanup,

restoration, and disposal.

Ten years have passed since the last BRAC round. Since 1995, Senator John

McCain and Senator Carl Levin have sponsored congressional action for two new rounds

of BRAC. In two Congressional Research Service Reports for Congress, Lockwood

(1999, 2000) reported the results of previous BRAC actions up to that time. Lockwood

also reported concerns of the DoD's request for two more rounds of BRAC and the

reasons given by the DoD for the new rounds and congressional action concerning the

need for new BRAC Commissions. Lockwood emphasized similar results that were

published by the GAO and CBO concerning the costs and savings of the DoD after the

previous four rounds.

Lockwood reported the estimated savings at $5.7 billion; however, the GAO

reported (Holman 2001b) the estimated annual savings had increased $5.6 billion in 1999

to $6.1 billion by 2001. The GAO reported (Holman 2004) the estimate is now nearly $7









billion in annual savings from the previous four BRAC rounds. The BRAC savings have

been substantial according to the reports. However, BRAC was not the only action by the

DoD designed to reduce defense spending. The Balanced Budget and Emergency Deficit

Control Act of 1985 called for the defense budget to be cut in half, BRAC alone would

not reduce the defense budget by half. Lockwood (1999) discussed an understanding of

the DoD's requests for BRAC. Lockwood explained that Secretary of Defense William

Cohen released the Quadrennial Defense Review (QDR) in 1997, which simply reported

a major review of the military's strategies and capabilities. According to Cohen's review,

the reduction in force had drastically surpassed the reduction of infrastructure. There was

a significant difference in Cohen's percentages: force structure was reduced by 33

percent; infrastructure was reduced by 21 percent. Cohen' s conclusion was a request for

two new rounds of BRAC in 1999 and 2001. Although Cohen's arguments for two new

rounds of BRAC were creditable, Congress was still stinging from the interference of the

1995 BRAC round and was in no hurry to appease the DoD's requests. Cohen continued

to emphasize the need for BRAC by declaring the significant savings from BRAC would

achieve the balance between force structure and infrastructure, thus supplying the

necessary funding for future readiness of force and weapons development to bring the

modern military up to speed with the military mission.

The 105th Congress neglected to give Cohen serious consideration because of their

concerns of political and economic fallout within their own districts. There are some

instances that BRAC can be difficult for smaller communities to overcome. Fort Riley

Kansas is a major Army base in the middle of the plains in Kansas. The city that would

be impacted economically from Fort Riley's closure if it were to occur is Manhattan

Kansas, a small city by most standards with a population of 44,831 in 2000 (US Census,









2000). Manhattan Kansas' economic activity is predominantly agricultural; however, it is

home to Kansas State University and Fort Riley. The university and military installation

are the two largest nonbasic activities within the small city. Farming is a way of life for

most of the people in the area surrounding Manhattan, but the agricultural industry relies

heavily on weather and climate. If the base were to close and several years of drought

occurred after the closing, the economic impact could possibly destroy the small city.

This is one reason Congress has a legitimate argument concerning BRAC. The fact is that

most military installations are located near areas, cities or metropolitan areas that have a

larger, diverse economy.

Finally, recent global events have given Congress another argument against future

BRAC actions. Since September 11, 2001, the military mission and goals have been

enhanced with the War on Terror and Operation Iraqi Freedom. Although there has been

a great deal of bipartisanship concerning America's military actions in both operations, a

majority of congressional members have used the military actions as an argument to

prevent further BRAC until the actions have been resolved. A GAO report (Holman

1998) emphasized the importance of reduction of operations and maintenance of

infrastructure (O&M) if the DoD is to meet the required expenses to modernize the force

structure. If the infrastructure costs are not met, then diverting the funds required for

force modernization to O&M jeopardizes the overall goals of the DoD. Another GAO

report (Holman 1998) discussed the Quadrennial Defense Review and the expected

savings from personnel reductions might not be achieved. According to this report an

expected $3.7 billion would be saved by 2003 if forces were reduced by 175,000

personnel.









At the time of these reports it is possible that no one knew of the future

consequences that occurred on September 11, 2001. Since that fateful day our forces have

been stretched to a breaking point. Schwalbe (2003) explained the Department of

Defense's position to counter the requirements of meeting the demands of today's

military. The DoD will assign some bases to inactive status that are selected for BRAC.

Once a base is relinquished to the private or civilian sector it cannot serve a military

purpose or be applied a military function. Because of the possibility of future need the

DoD will try to retain some of the properties in an inactive status similar to the status

assigned naval vessels when they are decommissioned, but may be needed for future

missions. If the DoD reduced the bases to meet the same percentage of personnel and

then a surge in personnel occurred to meet the military requirements for the War on

Terror, there may not be enough bases to house the rise in force structure.

Other methods could be considered for studying BRAC. Geographical Information

Systems (GIS) have advanced to a high level in the past twenty years. Particularly,

geospatial and imagery analysis have grown more advanced and many avenues for

research can be utilized to improve possible necessary scenarios that could create means

of preventing or accepting BRAC recommendations in the future. The best use of GIS is

the planning for land use of BRAC recommendations before the bases being

recommended. If each urban area that has a base were to plan for the closure or

realignment of the base, then planning for the reduction of economic impact could negate

the effect or at least lessen the economic impact.

Concern over Economic Impact

Gentrification plays a role in urban and economic growth. A recent trend seen in

many MSAs is a revitalization of Central Business Districts. The gentrification consists









of older buildings and structures being renovated or destroyed and new buildings and

structures being built in place of the old and the environmental cleanup of parks or green

spaces within the CBDs. Gentrification permitted large metropolitan areas to not only

reclaim areas within the CBD, but also to renew economic activity within the CBD that

was viable with the economic growth and activities that were continuing to spread from

the city's center.

A Thesis presented to Virginia Polytechnic Institute by Hogan (1997) discussed the

negative economic impact on local communities that experienced BRAC actions. Hogan

argued that the interference from Clinton in 1995 set precedence for future BRAC

considerations: privatization in place. The argument for privatization in place has merit

because it saves the DoD money in several ways without disrupting local economies. The

method would allow portions of a base or functions of a base to be maintained by private

industry and remove the operating cost from the DoD and place the burden of cost on

private industry. A strong argument for base closures is made using the data reported by

the Business Executives for National Security (BENS) Hill Advisory (1997) on

employment figures from base closures compared to the total U.S. employment and the

job loss of Fortune 500 companies: approximately 120,000 jobs from four BRAC and

over 250,000 from Fortune 500 companies in the first six months of 1996.

Alleviating Negative Economic Impact with Accessibility

Rugg (1972) explained the phenomenon as the "multiplier effect" using previous

research conducted by Hoyt (1961). Utilizing the economic base theory and defining the

military-base as a nonbasic activity of the Tampa MSA, MacDill AFB had become a

normal supplement to the new basic activity of Tampa MSA. The coalescence of Tampa

and St. Petersburg led to newer transportation and accessibility routes over the bodies of









water that separated the two MSAs. The new routes were also located closer to MacDill

AFB. The newer roads and the urban growth fed one another and in 1990 the growth of

Tampa-St. Petersburg continued until 2000 when the megalopolis of Tampa-St.

Petersburg-Orlando could be seen forming along Interstate 4 (1-4). Tampa MSA has

accessibility to other areas other than the Central Florida Megalopolis. Two Interstate

highway systems are significant to Tampa: 1-75 and 1-4. Interstate 75 gives Tampa

accessibility to areas on a north-south axis. Interstate 4 is the highway that connects the

Central Florida Megalopolis. Both interstates connect to other interstates (I-10 and 1-95),

which increases the basic economic activity for the Tampa MSA. Jacksonville is not as

fortunate as Tampa because its economic base is relegated to greater distances.

Jacksonville's basic activities require accessibility to MSAs at greater distances. Atlanta,

Georgia; Tallahassee, Florida; Savannah, Georgia; the cities along Florida's Atlantic

seaboard; and eventually the Central Florida MSA provide infrastructural exchanges to

Jacksonville's basic activities. The nonbasic activities related to military bases in

Jacksonville play an important role in the city's economic growth and development. The

presence of military personnel and bases provide an economic stimulus in terms of sales

and services supported by military related transfer income and consumption.

Another aspect with the military bases is their location within the Jacksonville

MSA. Both bases lie on important transportation routes into the heart of the Jacksonville

MSA. The urban growth pattern in Jacksonville has shown a tendency for growth toward

the bases rather than toward the major transportation routes leading to Savannah, Georgia

(1-95 to the North); Atlanta, Georgia (1-10); and Tallahassee, Florida (1-10). Industrial

growth patterns appear to follow the major transportation routes; however, residential and

commercial growth appears to follow the routes toward the bases. The relative locations









of the CBD to the bases are key to certain urban growth patterns. The need to study the

economic base of the region that immediately surrounds the CBD and the base and the

major routes that connect the two is an area that should be considered when studying the

economic impact of BRAC. This area will potentially absorb the greatest impact after

BRAC actions. Presence of a military-base is a nonbasic function and transfer of

payments are of great importance to the local economy, helping to boost the tax base and

increase the commercial and residential activity in the area. The closure of a base doesn't

necessarily impact the MSA as a whole, but as the RAND study (Dardia, et.al. 1996) of

three bases in California showed us, the impact on the local communities and areas

juxtaposed to the base are profound.














CHAPTER 2
METHODS

Economic Growth and Development because of Military Bases and the Central
Business District: Economic Base Theory and Accessibility

Military Bases, Central Place Theory, and the Central Business District (CBD)

Walter Christaller conducted a great deal of research revealing the relationship

between an urban center and economic growth. Central Place Theory (Christaller, 1933)

suggested that a city tends to decentralize as its economic base grows. Nonetheless, a

disproportional amount of growth was associated with more central locations. The more

recent literature (Rostow 1960; Rugg 1972; Palm 1981; Forkenbrock 1990; 0' Sullivan

1993; Boarnet 1996; Wu 1998; Vickerman et.al. 1999; Banister and Berechman 2000;

Nelson and Moody 2000; Berechman 2001; et.al.) highlighted other factors that influence

urban growth patterns and the role of commercial, financial, and residential districts in

the centralization of growth along prominent nodes or corridors. Military expenditures

and income also played a vital role in the growth pattern of an urban center. This is the

argument that is normally presented by congressional members when arguments for

selected base closures are recommended.

Since Christaller's research concerning Central Place Theory (1933) was published,

most research conducted concerning urban planning and growth place a great deal of

emphasis on the economic indicators within an urban area, its surrounding region, and the

nearby trade centers. The significance of the economic impact of industry loss and gain

may determine the scale in which an urban center may grow, the rate of growth, and the

pattern of growth. Moreover, employment mix and a city's function or economic profile









will also play a role in determining a city's ability to weather adjustments from the loss of

various industries.

Economic Base Theory

The spatial relationship of the urban and regional economy can affect urban growth

pattern. Economic base theory is one method of explaining the impact of industry loss or

gain. According to Yeates and Garner (1976) economic activity may be expressed by the

following equation:

Total activity in the city = Total in basic activity + Total in nonbasic activity.

(TA) = (BA) + (NBA)

Total income of the city = Total income derived from basic activities + Total
income derived from nonbasic activities.

The basic-nonbasic ratio (BA/NBA) represents the ability of basic (export-oriented)

activity to support nonbasic activities such as retail sales, consumer and producer

services, etc. Nonbasic activities are also supported by transfer payments and transfer

income (government expenditures on military bases fall into this category). According to

Fik (2000) the basic-nonbasic ratio creates a multiplier, which is applied to determine the

impact of transfer income on employment expressed:

Total Economic Activity = Basic Economic Activity + Nonbasic Economic Activity
Basic Economic Activity Basic Economic Activity Basic Economic Activity

(TE/ BE) = (BE/ BE) + (NBE/ BE)

Whereas TE / BE is the multiplier and is represented with the symbol m* the new
equation is:

m* = 1 +(NBE/BE)


In addition to the multiplier, Fik explains 4 different types of income flows or

transactions, which have an effect on the economic growth trends on an area. The type of









income flow that typically concerns the impact a military-base has on the local economy

is transfer of income. A visual representation of income flows is given in Figure 2-1. The

importance of this concept can be relative to military-base closures because transfer

payments to the base support industry at the metropolitan level.


Figure 2-1. Income flows.

Methods for a New Approach to Predicting Economic Impact from BRAC

Military Bases Applied to Economic Models

Military bases are a nonbasic function of urban centers and supply a transfer of

income from the federal government as well as from the military labor force. Federal

funding for military bases includes purchasing necessary goods and supplies from the

local urban economy. Since most urban growth models are based on economic

foundations, spatial characteristics and classifications, involve social and political factors,









and must rely on transportation networks; a strong argument can be given to reflect the

possibility of the military factors in recent years having an influence on economic

growth. Although the military impact has not been commonly used in previous research

of urban growth and planning or in economic models, military bases are usually lumped

in with one or more land use categories (most often institutional) or certain labor force

categories such as governmental employment. However with recent controversial issues

concerning base closures, studies are becoming more recognizant of the economic impact

an urban center may face on a base targeted for closure. Risa Palm (1981) explained in

great detail interesting points of the defense spending in previous years and the impact on

regional urban economic growth in the South and West regions of the United States.

Palm also referred to other research conducted concerning federal defense spending

(Rostow 1960; Sale 1975; Perry and Watkins 1977; Weinstein and Firestine 1978) and

the economic development of the regions in which the defense spending was concerned.

The importance of the research may be the foundation of current defense spending trends

in the same regions, and the greater impact it now has with higher military salaries.

Several factors in recent years also contribute the growing influence military bases

have in the nonbasic function of an urban center: increases in military pay, emphasis on

research and development, support services, civilian labor force requirements, and import

of perishable goods are just a few. In recent years the military pay increases have

normally been higher than the cost of living increases in the South, Southwest and

Midwest regions of the United States. Because of the restructuring of the military's

training operations and equipment (TO&E) has changed significantly since the Cold War,

many of the urban centers that contained bases have industries that are reliant on federal

contracts to supply goods and services required for the fulfillment of the military









obligation to the United States. A concern of many urban centers is the closure of bases

can relegate those industries to closure as well. The economic impact on the urban center

can be devastating to the urban economy if there is a significant reliability of the urban

center's industry on the nonbasic function of the base. However, according to Economic

Base Theory (Rugg 1972; Yeates and Garner 1976; Christian and Harper 1982; Mayer

and Hayes 1983; Hartshorn 1992; Fik 2000; et.al.) an urban center with a diverse

economic base should overcome the loss of one type of industry given the depth and

breadth of its other industrial linkages. Therefore, the argument of many DoD officials is

the restructuring of land use available after a base closure should be utilized efficiently to

compliment and/or enhance existing basic-nonbasic functions of the urban center. The

opposing view is the time it takes to replace the lost source of revenue can be difficult or

impossible to overcome in the short term and lead to the eventual destruction of certain

sectors of the urban center in the long term.

Urban transportation systems play a significant role to military bases. Highway

systems are vital to military bases. Mobility of military labor force is vital to location of

the base. Most employees of the military must be on the job earlier than the civilian

sector and thus having a reliable transportation network to travel to work is necessary.

Nelson and Moody (2000), Kim and Chung (2001), and Kim et al. (2003) discussed the

importance of transportation corridors and their effects on urban growth models. The

three study areas chosen for our study have major transportation corridors that connect

the military bases to other nodes including the CBD. Another important feature of the

transportation corridors is the availability to the residential areas to and from the military

bases. The access to transportation corridors outside the localized area may also allow for

some military members to live outside the study areas. Availability of transportation









corridors throughout the MSA may account for the spillover seen in previous research

(Dardia, et.al. 1996; Hooker and Knetter, 2001) concerning economic impact from base

closures. The land use classification given by St. John's Water Management District

(Jacksonville) for the years 1973 through 2000 show significant changes in residential

and commercial growth (predominantly an increase in both) near the major thoroughfares

and beltways near the military bases. Whereas the lack of area for growth in the MacDill

AFB study area allows growth of commercial area at the expense of residential area and

vice-versa. According to Christian and Harper (1982) agglomeration economies are

enhanced from beltways that are proximal to railroads, airports, and seaports as they

allow for the clustering of linked industries. Both of these cities have specialized

transportation functions because trade is a major activity in both urban centers.

Christian and Harper (1982) explained the input-output model as a method of

forecasting manufacturing (industrial) forces on economic growth. Mayer and Hayes

(1983) basically described the model as a series of input-output matrices and can be used

to account for all sector inputs and outputs of a city. In the case of a military-base, the

inputs and outputs could be income or revenue.

Previous Economic Impact Studies from Military-base Closures

The RAND study (Dardia et al. 1996) was mentioned the most in the research

conducted concerning BRAC and was based on three California bases that were closed in

the early rounds of BRAC from 1988 to 1994. The study used three benchmarks to gauge

the changes of economic impact. The three benchmarks are: (1) expert projections of

what would take place in each community, (2) the experience of a matched set of

California bases that had not closed, and (3) the experience in the broader regions in

which the closed bases were located. The study had mixed results based on the variables









that were selected. The variables that were used included: population, housing units,

vacancy rate, unemployment, labor force, K-12 school enrollment, and retail sales. The

variables considered the changes in each after the bases closed and the region's proximity

to the bases. The study revealed the gloom-and-doom scenario as being an extreme

prediction. Initial results showed that overcoming the negative effects were not as

difficult as had been predicted. The study also revealed that spillover into other areas

from BRAC was negligible as distance from the base increased. According to the RAND

study, the impact affected the unemployed workers and their families and the revenue lost

by individual businesses more than the community as a whole. This allowed the

community to overcome the impact through proper planning and land use. If there is

good indication that the economic climate in the region is favorable, then impact from

BRAC will be swift. Smaller and less diverse economies will require substantial longer

recovery times. Thus it stands to reason that the growth and development patterns of a

region be studied when a base is targeted for BRAC action. The conclusion given in the

RAND study simply stated that predicting the effects of economic impact are difficult.

The RAND study mentioned distance, but the study did not use distance as a

variable. An example of variables that could explain distance in a spatial relationship

between the bases and commercial nodal activity are the straight-line distances between

the bases and other nodes or the CBDs. Accessibility of the base to the CBD and other

nodes is also vital to understand the economic impact of BRAC.

US Senator Dianne Feinstein's webpage had the text of a letter (2004) she sent to

Peter Potochney, the director of BRAC. She emphasized three issues need to be added to

the military value section of the BRAC criteria. The second issue she outlined includes

accessibility considerations. However, Senator Feinstein did not include the proximity to









the CBD an important node in most urban economies. Another aspect that Senator

Feinstein failed to mention is the population of California. California has roughly one-

sixth of the United States total population. Thus, it would be easy to presume that if there

are more bases in California and more people, then California should have a higher

proportion of the numbers in base closures and personnel unemployed. Feinstein's

arguments are lost to the BRAC Commission because previous studies (Dardia, et.al.

1996; Hooker and Knetter, 2001) did not support her. Senator Feinstein did have one

advantage with her argument: most of the bases closed in previous rounds have left very

few choices for the BRAC Commission in future rounds.

Certain factors determine the economic recovery from BRAC actions. The GAO

explained the factors in detail (Holman 2001c) and supplied a visual representation of

those factors. According to the GAO, eight factors significantly affect economic

recovery: (1) reuse of base property, (2) government assistance, (3) public confidence, (4)

leadership and teamwork, (5) natural and labor resources, (6) diversified local economy,

(7) regional economic trends, and (8) national economic conditions. The last two factors

listed possibly played the strongest roles in previous BRAC actions and economic

recovery for those areas. The important issue to keep in mind with the next round of

BRAC is that national recovery is not the only factor that needs to be studied for the

economic recovery of BRAC actions.

Hooker and Knetter (2001) studied the economic effect of employment and

personal income effects that occurred from BRAC. They explained the importance of the

reduction in defense spending from 1986 to 1998 and the need for base closure; however,

they also mentioned the difficulties in deciding the bases selected for closure. The most

singular factor, as mentioned in the previous studies, involved in the fight against BRAC









is the economic effect on the local community. Hooker and Knetter approached the

problem using a newly constructed dataset to study the employment and personal income

effects at the county level. They explained that military bases are a major employer of

most counties in which the base is located (up to 30% in some cases according to the

study); and thereby accounted for a larger share of income and tax revenue in the area.

Hence, transfer payment, related income played a vital role in these local community's

economic stability. Government transfer payments helped pay for certain activities within

the local community, especially in the operation and maintenance of the local

communities infrastructure and services. Normally local taxes are utilized in education,

road repairs, community development, maintenance of parks, etc. The importance of the

transfer payment differs in one major aspect; the federal government (in the form of

property tax, sales tax, and other taxes that are owed) pays the taxes for the operation and

maintenance of the base. Normally the taxes collected by the civilian sector (sales,

property, and in most states a state income tax) pay for the operation, services, and

maintenance of the local community. Communities with a military-base enjoy the luxury

of receiving additional tax support from the federal government; thus they attempt to

overcome the decision for BRAC to occur in their communities.

Hooker and Knetter mentioned an interesting point concerning BRAC and the local

community: the opportunity cost of the resources the base affords the community after it

is closed. In particular, they cited the available land the community received and the

possible use of the land after it was released to the community as the most important

consideration. Hooker and Knetter gave two examples of scenarios that can assist the

recovery of local economies after the bases closed (in a study of the Presidio of San

Francisco Army Base and Moffett Field Naval Air Station in the Silicon Valley, both









bases in California). Primarily using employment and personal income indicators in their

study and measuring the responses from the counties by comparing the results to

counterfactual scenarios. The first scenario assumed the county's employment and per

capital personal income growth rate equal to the state's growth rate. The second scenario

assumed the difference between the county and the state's growth rate in the years before

the time of base closure would have persisted after the base closed. In the case of their

study they chose to use a two-year period before base closure to measure growth. The

results of their research showed that nonbase employment grew faster in closure counties

than it did in the counterfactual model. The study proved that spillover from job loss on

bases did not affect the surrounding areas as assumed from the impact analysis. Instead,

the study proved that if the bases' resources are properly used in alternative ways, then an

increase of job creation could occur if industries with higher multiplier effects are

brought in to substitute for jobs lost under the base closure. Hooker and Knetter

explained the findings are similar to recent studies that were conducted by Davis et al.

(1996) based on the dynamics of labor markets in larger regions involving basic

industries. Hooker and Knetter (2001) also found similar results to Aschauer's (1990)

"below-unity estimates of output multipliers for government consumption and military

investments from aggregate data." The personal income results from Hooker and

Knetter's study also revealed very little impact from BRAC. According to Hooker and

Knetter there were no statistically significant impact on per capital income from BRAC.

Furthermore, the study indicated a slight growth in per capital personal income in the

county compared to the state's growth after BRAC. Hooker and Knetter gave two

explanations for the results: (1) generally outgoing military personnel have below-

average income in comparison to income of employees working in other sectors and (2)









the older and more experienced civilians who lose jobs tend to gain employment at higher

salaries. According to Hooker and Knetter, economic impacts have traditionally been

projected instead of estimated and measured. They argued that projections from input-

output models tend to ignore the capacity of regional economies to adjust to closures.

They further argued that the main issue to measuring economic impacts was the

estimation of impacts that would occur without base closure. The type of base is another

factor Hooker and Knetter mentioned that tends to assist the community in recovering

from base closure. Bases that require more highly skilled workers, utilize more methods

of transportation for shipping and receiving of supplies, personnel, and equipment, and

provide more resources for future development tend to assist the local community's

recovery after they are closed. Air Force bases normally fit the criteria described and in

past BRAC actions were usually the majority of the larger bases selected. Hooker and

Knetter concluded that future studies should attempt to assess obscured results instead of

simply project results of economic impact. They also emphasized the importance of

refuting the negative impact predicted and concentrate on establishing the positive

aspects that can occur with the proper and well-developed planning of the use of the

base's resources after closure.

The preceding research suggested several main issues for further study of base

closure impact: (1) do not predict, estimate the impact on economic growth rates from

base closures, (2) include distance variables and accessibility to transportation corridors,

(3) consider the percentage of total population that is employed at the base in question,

(4) research should consider using a smaller geographical scale to estimate impact at

local levels, (5) include a larger number of social and economic independent variables to

increase the variability and random pattern of the model, (6) and apply the criteria for









base closure as defined by the Department of Defense (DoD). Our study considered the

previously aforementioned research and was consistent with similar methods used in the

previous research. However, the six issues discussed were included in our study to

determine the base's negligible impact on the economic growth rates at a smaller

geographical scale (census-tract level), whereas previous research is conducted at a larger

geographical scale (county, regional, or greater).

New Approach to Predicting Economic Impact from BRAC

Base Realignment and Closure (BRAC) 2005 and Florida's Bases

The next round of BRAC was scheduled for January 2005. Since the next round

was announced, many politicians from the federal to the local level have organized to

prevent BRAC from occurring in their state or local community. Florida is the home of

many military bases. Past BRAC rounds have resulted in the closure of several

installations in Florida; however, the upcoming round has affected the state more than it

has in past rounds. Governor Jeb Bush has already authorized the state to organize

grassroots activities to raise over $200 million dollars to fight BRAC in Florida. The

main reason for the governor's action may be because of statements given by Secretary of

Defense Rumsfeld concerning the relocation of Central Command from MacDill AFB in

Tampa, Florida to the Middle East and possible closure of MacDill AFB. Although

MacDill AFB has been the base most often involved in the rumors of BRAC, Florida is

concerned about the possibility of other base closures. The largest bases in Florida are

Pensacola Naval Air Station (NAS), Eglin AFB, Mayport Naval Station (NS),

Jacksonville NAS, MacDill AFB, Patrick AFB, and Key West NAS. Other bases in

Florida that do not have a large permanent military personnel presence include Camp

Blanding and Avon Park Bombing Range because of their mission as training bases.









Jacksonville and Tampa, Florida

Jacksonville and Tampa were chosen for our study because they are the largest

MSAs with bases in Florida. The economic activities in the MSAs, the size of the CBDs,

the transportation corridors between the bases and CBD from the surrounding

communities and the diversity of factors involved if bases are chosen for closure allow

for the development of modeling the estimates for economic impact. The bases chosen

for the study included MacDill AFB in Tampa and Jacksonville NAS (Jax NAS) and

Mayport NS in Jacksonville. An important factor concerning our study compared to past

studies involved the area being studied. Instead of incorporating an entire MSA, county

or region, our study emphasized the census-tracts that surround and connect the

transportation corridors between the CBD and base. The previously mentioned studies

practically stated that spillover from base closures was negligible. Instead of focusing on

the projected impact of a potential base closure, our study assessed if there was a

discernible statistical relationship between distance to a military-base and urban

economic growth rates taking into account the locational accessibility to the CBD and

other prominent urban nodes.

According to the Jacksonville Chamber of Commerce the total number of

employees at Jax NAS was 24,648 and Mayport NS was 15,001 in the year 2000.

However, when aggregating the total number of military members and federal

government employees in the census-tracts that were selected for the study areas the

numbers were smaller. In 1980, Jax NAS had 15,484 military members out of a total

population of 249,362 that resided in the study area and by 2000 the military population

had declined to 5,032 out of a total population of 303,909. The federal government

employees that resided in the study area for Jax NAS (south central Jacksonville) in 1980









were 6,500 and in 2000 the number of federal government employees was 6,802. An

explanation that could account for the change in military members was the addition of

homes on the base itself and the improved pay military members experienced over the

last twenty years. However, the number of civilian employees increased. The closure of

Cecil Field NAS in west Duval County may account for the increase in civilian

employees near Jax NAS. One other factor could account for the reduction of military

members living near Jax NAS: Jax NAS supplies the aviation units for the carrier group

stationed at Mayport NS on the mouth of the St. John's River and Atlantic Ocean in east

Jacksonville. Mayport NS study area had a military population of 11,541 out of a total

population of 197,768 in 1980 and 9,190 out of a total population of 272,590 in 2000.

The reduction in population thus rejects the theory that personnel had migrated from Jax

NAS. The only other conclusions could be the growth of military housing on the bases,

which may not be included in the census or the increase of average salary allowed

military members to move further from the bases. Finally, many service members may

have been out to sea when the census was being taken. Jax NAS and Mayport NS have

been a part of Jacksonville for the most part of last century (at least since World War II).

During the past thirty years, the required number of service members has been

approximately 20,000 for Jax NAS and 15,000 for Mayport NS. Additional Federal

employees for the Mayport NS study area numbered 3,960 in 1980; the population was

4,802 in 2000. Basically an increase was observed in civilian employment by the federal

government while a reduction was observed in military member population. Of greater

importance is the fact that military population has been reduced in the local area of both

study areas while there has been significant economic and urban growth. However, the









significant growth in civilian employees working on the base in question may have an

adverse effect from BRAC.

The significance of the civilian and military population of persons working on the

base is approximately one percent of each study area's total population. The small

percentage of the total population working on the bases in each study area suggested that

the impact on the economic growth rates from the base may prove to be negligible.

Jacksonville NAS and Mayport NS Study Areas and Diversity of Industrial
Employment: 1980 and 2000

Industry employment data gathered from the 1980 and 2000 census showed the

overall changes in employment profiles with the study regions. The categories of

employment were broken down into thirteen employment groups or sectors for each of

the three study areas (Tables 2-1, 2-2, 2-3). The employment figures for 1980 and 2000

for the residents of Jax NAS are seen in Table 2.1.

The employment of residents in the Mayport NS study area was compared using

the same categories that were used for Jax NAS. The employment populations for 1980

and 2000 for the residents of Mayport NS are shown in Table 2-2.

The Mayport study area was similar to Jax NAS because there were a majority of

increases in employment in most of the categories; however, the increases were not as

significant as they were in the Jax NAS study area. The reduction in military residents

was not as significant in the Mayport study area. Over twenty years the reduction was

approximately 2,000 residents in the Mayport study area, where Jax NAS saw a reduction

of over 10,000 military residents. One other factor that was not considered for the

reduction of military personnel was the DoD's reduction in force since 1986. The

reduction in force could be a significant factor, however, according to the Jacksonville









Chamber of Commerce, Jax NAS has over 24,000 employees in 2001 and approximately

6,000 of those employees were civilian employees. The tendency of growth employment

from the basic activities in the Jax NAS study area indicated that base closure should not

have a negative economic impact according to the study conducted by Hooker and

Knetter (2001) and percentage of the study area's total population employed by the

military. However, the DoD may decide that the mission of the base prevents the

selection of the base for BRAC. Jax NAS supplies the air support for the carrier group

that is stationed at Mayport. More importantly, Mayport is the only port other than

Norfolk, Virginia on the eastern seaboard that is home to carrier groups, thus meeting

vital criteria for the mission of the Department of the Navy and the defense of the Eastern

United States. However, if relocation of the carrier group from Mayport to another base

with a similar mission occurs; then it is almost assured that Jax NAS will be closed.

MacDill AFB and Tampa, Florida

The same industrial employment variables used in Jacksonville were also used in

Tampa. Armed Forces personnel that lived in the study area in 1980 were 10,624 out of a

total population of 249,646 and in 2000 were 2,172 out of a total population of 263,580.

The number of federal employees that resided in the study area numbered 3,651 in 1980

and 3,091 in 2000. Again, explaining the significant reduction in military personnel in the

study area may involve several factors, which are not known, but may have possible

causes. The important issue for our study was the significant reduction of military

personnel residing in the study areas. Since the reductions of military personnel have

occurred, then BRAC should not have as much of a significant impact from the loss of

military salaries. Unlike the Jacksonville study areas, the number of federal employees

also decline.









Table 2-1. Employment populations for Jacksonville Naval Air Station (Jax NAS) study


area
Economic
Activity
Agricultural,
Forestry,
Fishing, and
Mining
Construction
Manufacturing
Transportation,
Communicatio
ns, and Public
Utilities
Wholesale Trade
Retail Trade
Financial,
Insurance, and
Real Estate
Business and
Repair
Services
Personal Services,
Entertainment,
and
Recreation
Health Services
Education
Services
Other
Professional
Services
Public
Administration
b


Armed Forces


Employment Employment Diferene Percentage Transaction
1980 2000 Difference Type


1,153


6,300
10,203


9,098


5,228
17,233

10,338


5,073



5,465


8,004
7,383


5,097


7,444

15,484


569


9,827
9,222


14,490


5,604
17,936

17,725


14,695



11,265


15,203
8,526


6,514


7,618

5,032


-584 -50.65


3,527
-981


5,392


376
703

7,387


55.98
-9.61


59.27


7.19
4.08

71.45


9,622 189.67



5,800 106.13


7,199
1,143


1,417


89.94
15.48


27.80


2.34


-10,452 -67.50


BE


BE
BE


NBE


NBE
NBE

NBE


NBE



NBE


NBE
NBE


NBE


NBE

NBE


a Recreation is both a basic and nonbasic activity. b Includes civilian employees on military base.

According to past studies, especially Hooker and Knetter (2001), if the employment

in the area is higher than the regional or state growth rates, then base closures should not

have a significant negative impact on the local community. Reductions in the number of

military personnel residing in the study areas have been noted, and the changes in federal

employees have seen increases in Jacksonville and reductions in Tampa.









Table 2-2. Employment populations for Mayport NS study area
Economic Employment Employment Diffe e Percentage Transaction
Activity 1980 2000 r Difference Type


Agricultural,
Forestry,
Fishing, and
Mining
Construction
Manufacturing
Transportation,
Communication
s, and Public
Utilities
Wholesale Trade
Retail Trade
Financial,
Insurance, and
Real Estate
Business and
Repair Services
Personal Services,
Entertainment,
and Recreationa
Health Services
Education
Services
Other
Professional
Services
Public
Administrationb
Armed Forces


5,881
8,326


8,171


4,271
15,910

8,482


4,205


4,701

5,377
6,366


4,240


5,281
11,451


8,982
7,838


13,487


4,804
15,905

16,730


14,107


11,281

12,689
8,342


6,126


6,211
9,190


-424


3,101
-488


5,316


533
-5

8,248


9,902


6,580

7,312
1,976


1,886


930

-2,261


-45.3% BE


52.7%
-5.9%


65.1%


12.5%
0.0%

97.2%


BE
BE


NBE


NBE
NBE

NBE


235.5% NBE


140.0%

136.0%
31.0%


44.5%


NBE

NBE
NBE


NBE


17.6% NBE


a Recreation is both a basic and nonbasic activity. bIncludes civilian employees on military base.

However, federal employment was only one factor in determining economic

impact. Establishing an overall view of employment in the study area must be achieved

as previously done in the Jacksonville area. The employment data by sector for 1980 and

2000 for the residents of Tampa are shown in Table 2.3.

The changes in military personnel that reside in the study area are similar to

Jacksonville. However, unlike Jacksonville the percentage of the total population in the

study area employed at the military-base was significantly smaller (less than 0.5%).


-19.75


NBE









Although the percentage of the total population in the study area of military employees

was significantly small, the impact on economic growth rates from base closure should be

consistent with the percentage of military employees seen in the total population of the

study area (a negative impact of less than 1% on the economic growth rates in the study

area).

There was a marked and noticeable difference between employment changes in

Jacksonville and those in Tampa. The categories that show an increase in employment

did not show a dramatic increase, while there was a substantial decrease in military

residents in the area and the decreases in employment in the categories was greater than

those categories with increases. Considering the study of Hooker and Knetter, the

reduction in employment may enhance the possibility of improvement of industry and

employment in those industries with base closure. First, the base in Tampa is an Air

Force base and is associated with more extensive resources and skilled employment.

Second, the lack of space for growth proximal to MacDill AFB creates possible scenarios

for further growth with available space created with base closure. Finally, the location of

the base is prime real estate with obvious advantages in terms of transportation and

residential growth. Furthermore, the location has several notable qualities: (1) the base is

accessible to the CBD by several major roads, (2) it covers about one-fourth to one-third

of the lower portion of a peninsula, thus it is accessible to sea transportation, (3) it is

home to a large airfield, therefore it is accessible to air transportation, (4) infrastructure is

in place to support all of the transportation routes, and (5) prime real estate for residential

development. Given the characteristics, MacDill might be construed as a prime candidate

for BRAC selection. There is one other factor that the DoD may consider for MacDill









AFB and that is encroachment. According to census data and reports, the Tampa area is

one of the fastest growing MSAs in the United States.

Table 2-3. Employment populations for MacDill AFB study area.
Economic Employment Employment Diffe e Percentage Transaction
Activity 1980 2000 r Difference Type


Agricultural,
Forestry,
Fishing, and
Mining
Construction
Manufacturing
Transportation,
Communicatio
ns, and Public
Utilities
Wholesale Trade
Retail Trade
Financial,
Insurance, and
Real Estate
Business and
Repair
Services
Personal Services,
Entertainment,
and
Recreation
Health Services
Education
Services
Other
Professional
Services
Public
Administration
b


Armed Forces


1,825


6,141
13,078


8,675


6,025
19,929

9,007


5,975



6,829


7,847
7,247


5,236


4,956

10,624


320


7,593
9,607


11,811


5,291
14,921

14,981


19,124



11,780


13,009
8,281


5,827


5,087

2,172


-1,505


1,452
-3,471


3,136


-734
-5,008

5,974


-82.5% BE


23.6%
-26.5%


36.1% NBE


-12.2%
-25.1%

66.3%


NBE
NBE

NBE


13,149 220.1% NBE


4,951


5,162
1,034


591


131

-8,452


72.5% NBE


65.8%
14.3%


NBE
NBE


11.3% NBE


2.6% NBE

-79.56 NBE


aRecreation is both a basic and nonbasic activity. bIncludes civilian employees on military base.

The growth of Tampa allows the DoD to consider the possibility of the growth to

enhance the economic recovery from base closure and the fact that Tampa's growth has

encroached on the AFB in the last twenty years causes problems because of the air traffic









from the base. The best possible solution for both sides is to consider how to assist in

proactive planning in the event MacDill AFB is closed.

Recovering from Economic Impact Assuming Tampa or Jacksonville Bases Are
Selected for BRAC 2005

Undoubtedly, the local community's economy will be affected because of BRAC.

Proper planning for BRAC should be started immediately to assist the recovery from

economic impact because of BRAC. Hardest hit might be the areas juxtaposed to the base

(in the short run), if there are economic spillovers that are highly localized. Our study

examined the extent to which proximity to a base affects local economic growth rates and

the degree to which variability in urban growth is explained by distance to a base (which

accounts for the locational accessibility to other prominent nodes within the urban

economy).

Software Used for Our Study

The dependent and independent variables were taken or created from 1980, 1990,

and 2000 US Census Bureau STF3A Files using Microsoft Access and Microsoft Excel

programs. The 1980 STF3A Files were distributed in text format and Microsoft Access

was used to create the 1980 database of social and economic (independent) variables and

dependent variables used for our study. Once the 1980 database was created, the 1980,

1990 and 2000 database files were converted to Excel files for further use. ESRI ArcGIS

8.0 was used to create the distance and accessibility variables required for our study.

ArcGIS was also used for the spatial analyses, which is discussed in greater detail in the

next chapter. NCSS was the statistics software program used for the stepwise and

multiple regression analyses that are discussed further in chapter four.














CHAPTER 3
MAPPING URBAN GROWTH AND CHANGE

Chapter 2 discussed previous research concerning study areas for BRAC and

economic impact. Most studies were concerned with a large area surrounding the base,

normally an entire city, county, MSA, or region. The previous studies discussed in

Chapter 2 did not, however consider the inclusion of the growth patterns between the

CBD and base. The previous studies also minimized the factors concerning economic

impact. The RAND study utilized factors that addressed population, housing units,

vacancy rates, unemployment, labor force, K-12 enrollment, and retail sales. Hooker and

Knetter addressed employment and per capital personal income changes. The majority of

reports that addressed the issues and factors concerning base closure were centered

primarily on population, employment or unemployment, and income and not on spatial

patterns or urban spatial structures. The importance of base closure not having a

significant negative impact did not explain whether the base had an immediate impact on

urban and economic growth patterns in areas with close proximity to a base. Spatial

analysis and statistical evidence was needed to support the hypothesis that military bases

influence the areas near or contiguous to the base's perimeter. If there was evidence that

the military-base enhances the local economic growth, then the negative impact of base

closure may decline with increasing distance from the base.

The purpose of our study was to assess the local impact of military bases. The

previous studies hold that local spillover effects were negligible, yet these studies did not









actually include distance variables nor did they consider variability in relation to the

locational accessibility of areas to the base and other prominent urban nodes.

Previous research has also been couched from a small-scale perspective,

encompassing entire counties, regional areas, or entire MSAs. By contrast, our study

focused on intra-MSA variability and incorporated distance measures for a more

restricted study area.

Figure 3-1 is an image of the Jacksonville MSA including both study areas. Of

greater importance was the impact on the immediate areas that border or lie in close

proximity to a military-base.

Software Used to Create the Images for the Study Areas

The software package used to create images for our study was ArcGIS 8.0. The

data was gathered from several sources. The data used in the image processing of the

Jacksonville study areas came from The University of Florida's Geoplan Center

(http://www.geoplan.ufl.edu./) under the Florida Geographic Data Library (FGDL) and

the map data was available with the St. John's River Water Management District

(http://www.sjrwmd.com/programs/index.html). The data for image processing of the

Tampa MacDill AFB study area also came from the FGDL and land use data came

from the Southwest Florida Water Management District

(http://www.swfwmd.state.fl.us/data/gis/libraries/physical_dense/lu95.htm). Many

processes and tools were used to develop the images in ArcGIS (ESRI, 2001). The most

commonly used tools were the Spatial Analyst tool, Editor tool, and Xtools Pro. US

Census TIGER Line files were used to establish the 1980 census-tract borders by taking

the maps included in the 1980 US Census catalogues and editing the 2000 US census-

tracts in ArcMap using the editor tool.






















































Figure 3-1. Jacksonville Metropolitan Statistical Area (Image created using FGDL and
Census Bureau Data).

The edited census-tracts were then corrected using Xtools Pro to calculate the area

in square feet. The census-tracts were the unit of analysis in our study. The dependent









and independent variables were calculated using census-tract data, Florida Geographic

Data Library files, and land use files from St. John's River and Southwest Florida Water

Management Districts. Furthermore, the distance variables were calculated by using

spatial data (specific land use variables) and the distance between each of the nodes,

CBD, and military bases.

Another important aspect of the census-tract's utilization as the unit of analysis was

because of decentralization of each study area. Decentralization was revealed through the

creation of multinodal and polycentric patterns within the area. According to Christian

and Harper (1982) decentralization of employment and industry was a pattern of urban

growth that has been recorded since the 1940s. Christian and Harper described the

process of decentralization by explaining the vital role of multiple nodes (multinodal)

within a region and their impact on urban economic growth patterns. Christian and

Harper basically stated that the outward growth from the CBD led to more prominent

roles of the nodes on economic growth patterns within the area. Thus, if holding to

Christian and Harper's work, the Jacksonville and Tampa study areas are decentralized

and the CBDs of both areas are reliant on the strength of the nodes within the area for

further economic growth. Furthermore, Christian and Harper explained the significance

of the development of the multinodal system seen in both the Jacksonville and Tampa

study areas. The phenomena of suburban growth after World War II led to outlying

suburban centers that interacted with the CBD in a manner that greatly enhanced the

economic growth of the region. The examples that were revealed in the Jacksonville

study areas were the beach communities in the Mayport NS study area and Orange Park

in the Jax NAS study area. Unlike Jacksonville, MacDill AFB is not proximal to the

outlying nodes for Tampa, which are St. Petersburg and other major metropolitan areas









that are connected to the megalopolis of Central Florida. Centers of industry and

commerce are also nodes within an urban center's area of influence described by

Christian and Harper (1982) as another aspect of the multinodal system explained. The

commercial nodes used in our study played a vital role along with distance variability

measures in assessing the military-base's impact on economic growth. Christian and

Harper also explained the importance of polycentric spatial structures' role in the

economic growth of an area. According to the polycentric spatial structures' roles given

by Christian and Harper, commerce and industry are structured along hierarchical lines

that influence decision-making functions by directly or indirectly determining new

industry or commercial locations, thus influencing the impact of economic growth rates

within an area or region. The connectivity of major transportation corridors to the

multinodal systems (keeping in mind the roles of polycentric spatial structures, and

decentralization of the MSAs) in the census-tracts of the study areas was the determining

factor in the decision for census-tracts being the units of analysis. The ArcGIS Spatial

Analyst tool was used to calculate distance measures to the base and the CBD from the

centroids of census-tracts. The purpose of the spatial analysis was to attempt to find a

possible spatial relationship between the base and the surrounding area. Particularly, the

commercial interaction between the base, CBD, and key commercial areas within the

study area distance factors that may play a role in the economic relationship the base may

have with the local community.

Creating Shapefiles for the Study Areas

Since the unit of analysis for the study areas was the census-tract, the first step

taken in creating the shapefiles for our study areas was determining the census-tract

boundaries for the earliest time period being used in our study. Since our study's









temporal limits were 1980 to 2000, 1980 was the ideal time for establishing the census-

tract boundaries. 1980 census-tract boundaries were created using ArcGIS software

programs, 1980 Census-tract maps included in US Census catalogues, and the 2000 US

Census TIGER Line files (see Figures 3-2, 3-3, and 3-4).

Establishing Study Area Boundaries Using Shapefiles

The Jacksonville Metropolitan Statistical Area

Jacksonville MSA encompasses several counties in northwest Florida: Duval, Clay,

St. John's, and Nassau Counties. Our study will involve two bases in the Jacksonville

MSA: Mayport Naval Station and Jacksonville Naval Air Station. Mayport NS is situated

along the southern side of the mouth of the St. John's River as it empties into the Atlantic

Ocean and east of the CBD. The Mayport NS area of study is entirely in

Duval County. Jacksonville NAS is on the western side of the St. John's River before the

river turns to the east of the CBD. The Jacksonville NAS study area was situated along

the southern border of Duval County and also encompassed the portion of Clay County

that contains the Orange Park city limits. The means of creating a localized study area for

the two bases was accomplished by using US Census-tracts from the US Census TIGER

Line Files. By using census-tracts an area can be created for studying the economic

impact between the bases and the CBD. The importance of the CBD is simple: The CBD

is the heart of the area's economic activity. The area between the CBD and the base

defined the most active economic corridor.






































Figure 3-2. Mayport NS study area census-tract boundaries.


Figure 3-3. MacDill AFB study area census-tract boundaries.


I I Census Tract
117.01 Census Tract ID











































SI Census Tract

167.01 Census Tract ID
_/


2 1 0 2 4 6



Figure 3-4. Jacksonville NAS study area census-tract boundaries.

Census-tract boundaries in the study areas changed over time with increase in population.

To overcome the problems associated with tract boundary change, the 1980 census-tracts









were used and the data from the 1990 and 2000 census were changed to fit within the

boundaries of 1980.

Mayport NS Study Area

The Mayport NS study area was unique because there were only two major

accessibility routes to the base from the CBD. A large portion of the Jacksonville MSA

lies to the north of the base; however, there is only one highway spanning the St. John's

River to the base and it was built during the time of the study and did not open until

recently. The other bridges cross the river after most of the major transportation routes

enter the CBD (see figure 3-5). The inability to reach the base by road from the northern

portion of the Jacksonville MSA limits encroachment and reduces the likelihood of

spillover effects. The important factor was that the base may impact only those areas that

were accessible between the base and the CBD. Since there were no direct routes to the

north of the base, the need to test for economic impact in those areas may be irrelevant.

Another important factor concerning urban growth was the lack of railroad activity near

the base. Figure 3-5 illustrates the lack of rail accessibility to the beaches and base.

Although there is a lack of railroads near the base, there is an abundance of air activity.

Craig Field is located in the center of the study area and the military-base has an airfield

for the purpose of transferring aircraft from Jax NAS to the carrier group. Future urban

growth is possible because of the new highway construction connecting the northern

regions to the area near the military-base. However, the possibility of increased growth

could be improved with connecting railroads to the industrial areas in the region,

especially if the base is considered in future base closures.




































Figure 3-5. Modes of transportation in the Mayport NS study area.

Another unique feature of Mayport NS is the site and situation of the base. The

military-base is situated at the mouth of the St. John's River emptying into Atlantic

Ocean (Figure 3-5). The site is located on very marshy land and was unsuitable for

development when the base was first built. Recent advances in urban development have

made development of most of the land surrounding Mayport NS an easy and profitable

task. The rapid development of land near Mayport NS has become an encroachment issue

with the DoD since BRAC rounds began to take place in 1988. The most important issue

with encroachment involves bases with airfields; Mayport NS has an airfield for the

purpose of outfitting the carrier group before maneuvers. If urban growth continues at its

present rate in this area, the base could become a prime target for future BRAC rounds

(See Figure 3-6).






























Legend
Major Roads
S Ports
Railroads
W Census Tract Boundary
Airfields
a Urban and Economic Growth 1980
SUrban and Economic Growth 2000


1a
Miles
1050 1 2 3


Figure 3-6. Urban and economic growth in Mayport NS study area (1973-2000).

Note: Urban and economic growth combines the commercial and residential land use
areas for the given year.

The true indicator of economic impact did not rely on images that have been

represented thus far, but in the data that was represented in those figures.

Jacksonville Naval Air Station (Jax NAS)

Jax NAS has greater accessibility to the Jacksonville CBD and possibly plays a

greater role in economic development in the local area. Jax NAS is accessible to three

major highways. Two of those highways (Interstate 95 and US Route 17) lead to the heart

of the CBD, the third highway (Interstate 295) leads to Jacksonville's entire periphery

locations (See Figure 3-7). Finally Jax NAS is situated between Jacksonville's CBD and

Orange Park. The Jax NAS location is almost in the center of the Jacksonville MSA (See









Figure 3-1). The location of the base may allow for greater nonbasic activity between the

base and the MSA, which was the basis of the studies mentioned in the previous

Chapters, and possibly influence the urban and economic growth in the local

communities surrounding the base. Of greater importance is the underlying potential of

land use if the base is closed. Because of the base's location, availability of diverse

transportation routes, and the potential for industrial, residential, and commercial growth,

the opportunity to improve the economic impact after a base closure has greater potential

than a base closure in Jacksonville, North Carolina or Manhattan, Kansas. Camp LeJeune

Marine Corps Base in Jacksonville, North Carolina and Fort Riley, Kansas, which are

located in smaller towns and are the predominant source of income for those

communities and play a greater role in the economy. Bases in large MSAs have only a

small function in the economic structure with the ability to overcome economic impact

from base closure. However, the vital role of the base's mission to the overall goal of the

DoD establishes the DoD's criteria for base closure.

There has been a tremendous amount of urban growth in Jax NAS study area since

1973. Jacksonville has shown a tremendous amount of growth to the south of the CBD,

just as is seen to the east of the CBD. Mayport NS had the advantage of being near three

smaller towns on the beaches; Jax NAS also has an advantage of being near a fast

growing community to the south: Orange Park (See Figure 3-7). Access to major

highways to the south of Jacksonville has assisted the local communities in this area to

prosper and grow, in addition the presence of the base and access to the Intercoastal

Waterway via the St. John's River has provided additional means of growth in the area.

Unlike the Mayport NS area, tourism is not as much a factor on the local economy.










However, an advantage the Jax NAS study area has over the Mayport NS study area is

the access to major


Miles
1050 1 2 3
I'Ll"


-egend
- Major Roads
H Ports
Railroads
W Census Tract Boundary
Airfields


Figure 3-7. Transportation routes to Jax NAS.












aPE^TR


- Residential
Growth 1980
- Residential
Growth 2000


Miles
105 0 1 2 3
I'Ll"


Major Roads
P orts
Railroads
W | Census Tract Boundary
Airfields


Figure 3-8. Residential growth in the Jax NAS study area (1973-2000).











Growth
Legend
Commercial
Growth 1980
Commercial
Growth 2000


STra portation
legend
Miles Major Roads
1 0 0 1 2 3 Ports
Railroads
I I Census Tract Boundary
Airfields

Figure 3-9. Commercial growth in the Jax NAS study area (1973-2000).

highways, particularly the Interstate system. Interstate 295 connects Jax NAS to most of

the MSA's industrial areas. The importance of the transportation corridors to Jax NAS









could be used as an argument for base closure because of the prime real estate that would

be made available if the base is selected for closure (Figures 3-8 and 3-9).

Tampa-MacDill AFB Study Area

The Tampa MSA has similar and different attributes with the Jacksonville MSA.

The similarities for both consist of four counties, have a coastal boundary, and have a

population over one million. The Tampa MSA differs from the Jacksonville MSA

because of more CBDs of closer cities, more diverse modes of transportation within the

Tampa MSA, and greater spatial diversity. The Jacksonville MSA consists primarily of

the city of Jacksonville, whose city limits is the entire county of Duval; and then Clay,

Nassau, and St. John's counties make up the rest of the MSA. After Jacksonville-Duval

County, the next most populated county is Clay County with a population of

approximately 150,000. The Tampa MSA has a huge economic advantage because of

population: the next most populated county after Tampa-Hillsborough County is Pinellas

County with a population at approximately 900,000. Jacksonville-Duval County's

population is smaller than the population in Pinellas County. The smallest population by

county in the Tampa MSA is Hernando County and the population is approximately the

same as Clay County. The extreme difference in population provides more opportunity

for urban and economic growth and a better means to overcome the negative economic

impact that may result from a base closure. The greatest difference between the two

MSAs is the Tampa MSA is the furthest left boundary of a Megalopolis: the Central

Florida Corridor, which extends from the Tampa MSA in the west through the Lakeland

and Orlando MSAs in the central portion of the megalopolis to Daytona on the Atlantic

coast.









The proximity of major metropolitan CBDs to the Tampa CBD is an advantage that

Tampa has over many MSAs or urban areas with military bases. The CBDs of Clearwater

and St. Petersburg, Florida are not very far from Tampa's CBD (See Figure 3-10).

MacDill AFB is separated from the St. Petersburg CBD (the next closest CBD to the base

after the Tampa CBD) by Tampa Bay. The base does have access to St. Petersburg's

CBD by Gandy Bridge, which spans Tampa Bay. However, any economic impact from

base closure should affect the Tampa CBD before any affect would occur on the other

CBDs in the Tampa MSA. Also, any localized effect of urban and economic growth from

the base would not include any areas outside of the study area because of natural

boundaries mentioned previously. Income and taxes generated by the base and base

personnel predominantly affects the commercial and service activities within the

immediate area. Another byproduct of income and property taxes paid by the government

and personnel affect the immediate area in the form of income being produced for

schools, police, fire, and emergency services, and some medical services. More

importantly, urban and economic growth may depend on improvement and expansion of

these types of services in the local area. The development of these services do not depend

on the presence of the base alone, normally residential growth has more of an impact on

the increase or decrease of these services within the local area.

Failing to reject a discernable significance of the base on urban and economic

growth in our study area may be difficult. The basic principles of Central Place Theory

explained the enormous effects a CBD has on urban and economic growth, especially if

the CBD has a strong spatial and economic relationship with outlying nodes. The Tampa

CBD has very densely populated cities as nodes whose CBDs are possibly as strong as









Tampa's. The proximity ofMacDill AFB to the Tampa CBD may cause difficulties

because of the CBD's economic strength and past growth trends from the CBD.


Miles
4 2 0 4 8 12





































Legend
MacDill AFB StudyArea m Urban

CBD Census Tract Boundary
CBD11101 LE I


Figure 3-10. Tampa MSA.









Figure 3-11 illustrates the distinct advantage MacDill AFB has because of its

relative location. The proximity of MacDill Air Force Base to the CBD and the

availability of all major modes of transportation may support the spatial relationship the

base has with the CBD. The availability of railroads, highways, air transportation and a

port may allow some possibility of the base having discernable significant impact of

economic growth on the localized area. Economic-growth should exhibit a rather healthy

growth trend because of the movement of goods and services within the area between the

base and the CBD. A spatial relationship in terms of commercial, industrial, and

residential growth should support a positive impact of economic growth from the

presence of the base.

Empirical Approach to Study Areas

Dependent Variables and Each Study Area

The dependent variables created for the three study areas assessed the military-base

impact on urban and economic growth at the localized level. Three dependent variables

for analyzing economic growth were chosen to test the variability of military-base

impact. The tables of dependent variables' values selected for the regression analyses of

our study areas are found in Tables 3-2 through 3-4 with a key explaining the dependent

variables created for regression analyses found in Table 3-1. The three dependent

variables used in regression analyses were percentage change in commercial land use,

percentage change in residential land use, and percentage change in median incomes.

Three periods of time were included to warrant a progression of growth in our study

areas: 1980 to 1990, 1990 to 2000, and 1980 to 20001. The period of time that produced


1 St. John's River Water Management District 1973 landuse data was employed, the landuse data for 1980
was not available and Southwest Florida Water Management District 1999 landuse data was employed, the
landuse data for 2000 was not available.









the best analytical results was 1980 to 2000. The dependent variables were tested in a

multiple regression analysis at 95% confidence level.


Figure 3-11. Modes of transportation accessible to MacDill AFB.

Independent Variables and Each Study Area

Several types of independent variables are created for each study area (Appendix

A, Table A-i). The first sets of variables are created using socio-economic and

demographic data taken from the US Census Bureau between the years 1980 and 2000.









Another set of variables is created from land use data (commercial and residential areas)

taken from the St. John's River and Southwest Florida Water Management Districts.

Variables representing percentage changes between 1980 and 2000) are created from the

socio-economic and demographic variables and land use variables.

Table 3-1. Key to dependent variables.
Name Database ID Description
The percentage change of area by
Percentage Change in PCTACMR80 square feet of commercial land
Commercial Land CMR99 use, the numbers following
use Area identify the total commercial
area for that year.
The percentage change of area by
Percentage Change in PCTARES80 square feet of residential land
Residential Land RE use, the numbers following
use Area identify the total residential
area for that year.
Percentage Change in PCTAMDThe percentage change calculated
Median Household for the median household
INC
Income income for the given years.

Distance variables were also created measuring straight-line and road distances between

the military-base, nodes, and the CBD. Variables for accessibility indices were created

from straight-line and road distances between the nodes, CBD, with distances to the

bases, as well as accessibility indices between the nodes and CBD without distances to

the bases. The majority of independent variables were similar for each study area,

specifically the socio-economic, demographic, land use, and percentage change variables.

However, the accessibility and distance variables were significantly different for each

study area. The differences in the number of independent variables per study area were

because of the varying amount of distance measures between nodes, the CBDs, and

bases. The Jacksonville study areas had more independent variables because there were

more nodes in the study areas.









Choropleth maps suggested the possibility of military-base impact by observing the

percentage of change of the dependent variable and the proximity of the base. However,

choropleth maps did not discern the impact of a base on the economic growth rates for

the localized area.

Results of Choropleth Maps for the Jax NAS Study Area

The percentage change in commercial area of the Jax NAS study area illustrated the

possible impact that a nearby nodal CBD may have on urban and economic growth of the

study area (Figure 3-12). The census-tract immediately below the military-base is the

Orange Park CBD; a tremendous amount of growth occurred in the census-tracts

surrounding the Orange Park CBD and the military-base. Although the nodal CBD

possibly influences the factors for growth in the localized area, there is a possibility that

the military-base may also serve as an impacting force on the urban and economic growth

of the localized area. Also of importance is the growth occurring on the opposite side of

the St. John's River from the military-base.

The major highways (1-95 and US 1) that travel from the south of the CBD can be

redirected to the military-base by connecting with 1-295, this provides a transportation

corridor from the base to the CBD via a different route and thus may explain the growth

from the CBD. There is a great deal of growth seen, but major highways lead to another

nodal CBD in the MSA (St. Augustine), which is approximately 30 miles further south.

The certainty seen in figure 3-12 was substantial commercial growth between 1973 and

2000; however, commercial growth was only one of the variables involved to explain

possible economic growth. Figure 3-13 illustrated the residential changes in the Jax NAS

study area.









The significance seen in the figure was similar to the commercial change seen for

the same areas. Several assumptions can be made for the significance in change of the

residential area. First, the availability of open land for development as the CBD's urban

growth expanded was more prominent to the south and west of the CBD between 1973

and 2000.

Second, highly accessible transportation routes leading from the CBD were found

in the localized areas to the south and lead to major nodal CBDs (Orange Park and St.

Augustine). Finally, many areas near military bases were normally reserved for planning

because of the airfields located on such bases; air traffic tended to act as a negative factor

on land values, especially residential areas. The residential growth in the study area

illustrated a strong possibility that the base may have some impact on the urban and

economic growth of the localized area.

The percentage change in median household income showed some greater growth

immediately south of the Orange Park CBD. The growth rates for median household

income were apparent in figure 3-14. Unlike the observations for the land use changes,

the greatest growth was between the base and the CBD. Possible assumptions were

similar to those given for land use changes. Another indicator may be seen in real estate

values over the time periods studied or the area may be more attractive to one-person

households.

Results of Choropleth Maps for the Mayport NS Study Area

The percentage change in commercial area for the Mayport NS between 1973 and

2000 illustrated an interesting occurrence between the CBD and the three beach CBDs.

Figure 3-15 showed the greatest amount of change in commercial area occurred in

census-tracts proximal to the beaches and base, particularly census-tracts 143.01, 143.02,









and 146. Thus reflecting a great amount of change occurring between the CBD, major

nodes and the base. Overall, the greatest amount of growth occurred near the three beach

nodes to the south of Mayport NS. The base may have some influence in the growth of

the area because of its location; however, the growth tends to follow the transportation

corridors in the study area.

Figure 3-15 illustrated growth along the transportation corridors, which follow the

principles of Central Place Theory and Economic Base Theory in the spatial relationship

of the growth compared to the proximity of the CBD. The figure also illustrated a return

in growth toward the CBD possibly because of interaction between the CBD and the

beach nodes and base.

The percentage of change in the residential area between 1973 and 2000 was quite

different than the changes seen in the commercial area in the Mayport NS study area.

Figure 3-16 illustrated the greater changes in residential growth occurred closer to the

beach nodes. The majority of census-tracts near the CBD actually experienced a decline

in residential area, particularly those census-tracts that underwent or experienced a

change in commercial area. The explanation could be because of the development of low-

lying land in the largest census-tract in our study area (tract 143.02). The development of

newer techniques to reclaim land that is normally unsuitable for building such as swamp

and marshland has allowed the further development of the areas closer to the base and

beaches. The expansion of residential growth in those areas may have allowed the

expansion of residential area in the outer areas of our study area and making land in the

CBD readily available for commercial development.

Figure 3-17 illustrated the percentage of median household income having a similar

pattern of growth that was seen with percentage change of residential area. The CBD and









other prominent urban nodes showed strong trends in growth. Of great interest was the

strongest trend in growth that was observed in the areas closest to the base.

Results of Choropleth Maps for the MacDill AFB Study Area

The first aspect to remember concerning the differences between the two bases in

Jacksonville and MacDill AFB in Tampa is MacDill AFB does not lie near a nodal CBD.

MacDill AFB is situated approximately 6 miles south of the CBD and is separated from

the remainder of the MSA because of its location on the southern tip of the Interbay

Peninsula. The spatial relationship to the CBD leads to the assumption that any growth in

the study area will be influenced primarily by the CBD. However, the base may still have

an impact on the economic growth in the localized area because of several factors:

proximity to the CBD, isolation of the base from most of the MSA because of its

location, the accessibility of diverse modes of transportation that are located near the

base, and the small amount of land available for commercial, residential, and industrial

development. Of greater importance is the available land resulting from the closure of the

base could open up more opportunities for economic growth because of the other factors.

Before the base can be selected for closure, the impact the base has on the area should be

assessed even if the DoD and the BRAC commission select the base for closure. If a

significantly substantial impact can be proven, the argument against base closure has

stronger support.

The census-tracts representing the greatest percentage change in commercial area

are close to the CBD (figure 3-18). The majority of commercial growth is located in

tracts surrounding the CBD. Of greater importance is the number of census-tracts that

show a negative value in commercial growth. Five of the eight census-tracts immediately

to the north of the base show negative growth and three of the five have substantial









negative growth. An assumption can be made that the base may have no impact or a

possible negative impact on the economic growth of the localized area. Several factors

may explain the negative growth within the immediate area; such as, gentrification,

creating available land for rezoning and development, and increases in real estate values.

Only the three census-tracts reflecting substantial negative commercial growth

(figure 3-18) have a positive value in residential growth (figure 3-19). The remainder of

the census-tracts may have an extremely small amount of positive residential growth

(maximum value of 0.02), but mostly negative growth values are represented. The values

seen in residential and commercial growth support the earlier assumptions concerning the

limited amount of space available for any growth in the study area.

The greatest amount of percentage change seen in median household income is

along a major transportation corridor between the CBD and the base (Figure 3-20).

Although there is a substantial amount of growth observed in the census-tracts, the

impact provided by the base may be in conjunction with impacts because of the CBD.

Review of Choropleth Maps Results

The choropleth maps suggested that the bases in the Jacksonville study areas may

have discernible impact on the economic-growth of the localized areas. Although the

spatial analyses illustrate the possibility of discernible impact from the base on the

economic-growth of the localized area, the nodal CBDs may actually have more of an

impact on the economic growth than the bases. However, the bases may increase the

effects of the nodal CBDs' impact on economic-growth of the localized area.

Unlike the study areas in the Jacksonville MSA, the spatial analyses for the Tampa

study area suggested that the CBD is the predominant force in the economic growth of

the study area










Table 3-2. Jax NAS dependent variables.
Commercial Area Residential Area Median Household
Census-
trat Percentage Percentage Income Percentage
tract
Change Change Change
303 1444.79 118.04 124.59
304 826.01 -18.62 129.15
305 100.00 4.99 117.67
306 -9.40 0.00 111.98
307 100.00 127.21 194.23
308 61.17 20.19 129.13
309 100.00 240.73 162.36
2 100.00 -53.21 149.28
3 100.00 -17.44 106.81
4 288.99 -46.05 152.75
5 100.00 -59.37 304.45
6 151.06 -31.02 126.81
7 137.62 -11.94 226.29
8 51.05 -20.99 169.12
10 59.52 -75.87 63.62
11 -56.97 -1.21 219.53
12 -36.53 3.64 256.21
13 6.28 -17.85 127.51
15 -12.75 -5.50 156.50
16 90.95 -40.11 110.14
17 -23.95 -76.63 136.38
18 -23.62 -76.86 77.39
19 -21.97 56.81 121.34
20 -67.26 -6.45 196.17
21 46.13 -21.29 180.33
22 256.67 -22.27 146.67
23 -10.04 -25.76 162.50
24 1060.14 -7.53 212.75
25 12.62 -11.39 150.82
26 -53.87 -44.27 171.59
27 109.60 -4.98 92.38
28 1221.60 -9.85 162.97
29 504.54 -7.88 199.83
123 105.52 -20.91 125.00
124 -18.69 -16.12 123.60
130 -32.50 -3.71 146.94
131 77.96 -11.37 181.87
132 100.00 -69.52 112.64
133 479.11 98.24 149.35
134.01 888.72 10.86 147.91
134.02 100.00 -11.46 115.89









Table 3-2. continued
135.01 100.00 80.59 147.46
135.02 114.36 2.60 149.67
162 -16.81 -17.75 129.19
163 -47.57 -44.06 142.25
164 672.49 -5.31 168.36
165 26.34 8.47 114.89
166.01 248.11 12.01 81.72
166.02 100.00 22.67 91.51
167.01 464.52 77.18 109.04
167.02 1240.45 294.80 137.00










Table 3-3. Mayport NS dependent variables.
Commercial Area Residential Area
Census-
trat Percentage Percentage
Change Change
1 -58.74 -27.68
2 100.00 -53.21
3 100.00 -17.44
4 288.99 -46.05
5 100.00 -59.37
6 151.06 -31.02
8 51.05 -20.99
10 59.52 -75.87
11 -56.97 -1.21
12 -36.53 3.64
13 6.28 -17.85
15 -12.75 -3.68
16 90.95 -40.11
17 -23.95 -76.63
18 -23.62 -76.86
19 -21.97 56.81
138 100.00 100.00
139.01 100.00 101.82
139.02 73.49 28.20
139.03 102.42 137.15
140 231.25 89.88
141 29.29 11.13
142 307.93 24.56
143.01 245.05 56.47
143.02 4722.05 1033.70
145 124.89 2.03
146 251.59 39.12
147 100.00 191.87
148 772.98 46.73
149.01 1620.57 6.86
149.02 100.00 68.46
150.01 100.00 -4.06
150.02 -20.12 -3.69
151 0.36 3.33
152 21.63 -3.56
153 -3.21 -1.17
154 15.68 -27.92
155 167.11 -24.61
156 474.52 -3.32
157 122.59 -40.00
158.01 41.21 2.99
158.02 365.16 34.96


Median Household
Income Percentage
Change
164.01
149.28
106.81
152.75
304.45
126.81
169.12
63.62
219.53
256.21
127.51
156.50
110.14
136.38
77.39
121.34
309.91
200.79
180.83
230.43
198.46
171.50
244.12
148.68
157.47
108.35
149.21
129.11
139.14
102.62
126.96
92.70
87.35
135.88
148.59
108.94
90.57
123.89
123.06
97.84
101.17
155.30










Table 3-4. MacDill AFB dependent variables.
Commercial Area Residential Area
Census-
trat Percentage Percentage
Change Change
18 119.27 -36.41
19 71.80 -36.68
20 93.93 -20.45
21 451.09 -15.33
22 32.46 -10.54
23 -47.80 -6.93
24 -21.57 -17.62
25 -18.17 -12.59
26 -48.77 -92.23
27 131.19 -22.11
28 9312.35 -10.26
29 23.85 -13.85
30 233.70 -18.75
31 459.58 -24.30
32 171.96 -11.77
33 2092.39 -10.14
34 6313.90 -25.77
35 125.69 -21.53
38 84.20 -3.89
39 4.15 -19.92
40 421.46 -43.77
41 2834.95 -50.46
42 -0.53 -22.21
43 -79.57 -30.79
44 100.00 -39.21
45 616.53 -19.43
46 30.20 -27.68
47 -1.32 8.77
48 -29.63 12.70
49 45.26 -19.53
50 -31.70 -29.43
51 -19.94 1096.21
53 -38.92 -40.54
54 -76.16 -4.91
55 1091.21 -95.04
57 35.40 -17.76
58 -9.93 -9.23
59 -3.58 2.88
60 60.09 -8.52
61 -16.91 -25.76
62 145.40 -18.80
63 374.84 -5.89
64 746.65 -1.36


Median Household
Income
Percentage Change
123.51
203.76
112.43
143.53
195.15
139.88
60.53
104.85
207.38
122.68
134.52
197.13
198.8
239.2
156.91
269.18
163.13
163.93
212.7
110.96
130.48
91.8
161.39
126.23
174.82
123.32
105.38
165.96
124.33
326.89
136.28
1218.98
181.68
212.8
457.69
218.18
155.09
185.02
255.38
294.62
403.87
246.24
180.37










Table 3-4. Continued
65 -55.14 7.89 150.27
66 -46.96 3.23 100.01
67 -12.55 -6.29 205.83
68 34.27 -15.95 196.19
69 3184.58 -14.48 189.75
70 -17.60 -13.87 119.26
71 -69.55 9.16 105.76
72 100.00 -12.71 197.03
117.01 0.00 0.00 221.28
117.02 246.88 43.50 171.22
240.01 396.79 -25.55 131.99
240.02 -61.31 -9.12 239.02
240.03 0.00 14.61 241.27
241 -80.24 -11.83 208.10
244.03 165.56 -13.16 225.84
244.04 100.00 -26.03 225.04
244.05 100.00 0.02 203.78
244.06 21.41 -17.09 110.04
244.07 -37.08 29213.21 157.67
245.02 364.41 57.75 203.35



































US 1

1-95


Change
ercial Area
3 2000*
Major Roads
-67.26 --16.81
-16.80 -61.17

Miles ~61.18 90.95
1 05 0 1 2 3 90.96 -137.62

137.63 1444.79


Figure 3-12. Commercial area percentage change 1973-2000 for Jax NAS study area.

Note: 1973 land use data used since 1980 land use data not available.








































e Change in
dential Area
973 2000*
Major Roads
W -76.86- -31.02
W -31.01 --16.12
S-16.11 --5.50
S-5.49 12.01
12.02 294.80


Miles
1 050 1 2 3
I'Ll"


Figure 3-13. Residential area percentage change 1973-2000 for Jax NAS study area.


US 1

1-95




































US 1

1-95


ange in
hold Income
2000
Major Roads
63.62 114.89
114.90 129.15
Miles 129.16 149.35
1 05 0 1 2 3 149.36 171.59

171.60 304.45


Figure 3-14. Median household income percentage change 1980-2000 for Jax NAS study
area.








78







W 1E





















Percentage Change in Co clal rea 73 2000'
I/ 7 l .3 d .1i
-3.20 73.49
S73.50 90.95
90.96 151.06 Miles
2 1 0 2 4 6
161.07- 4722.05
Major Roads


Figure 3-15. Commercial area percentage change 1980-2000 for Mayport NS study area.


Note: 1973 land use data used since 1980 land use data not available.









































Percentage Change in Re de2nifaAreal 3 2000'
| 1.7686 -31 0;
1 31.01 --3.68
S-3.67 3.64
3.65 56.47
S56.4 8-1033.70
Major Roads


Miles
2 1 0 2 4 6
1 1 1 1


Figure 3-16. Residential area percentage change 1973-2000 for Mayport NS study area.


Note: 1973 land use data used since 1980 land use data not available.



































Figure 3-17. Median household income percentage change 1980-2000 for Mayport NS
study area.
















72


St. Pet rsburg i 1


81









U
1-4






Percentage Change in
Commercial Area
1980 1999*
MacDill -80.24 --29.63
Air Force Base
i3 -29.62 4.15
.4.16 100.00
S100.01 396.79
396.80 9312.35
5 Miles 2 3 City Limits
S I I I Major Roads


Figure 3-18. Commercial area percentage change 1980-1999 in the Tampa study area.

Note: 1999 land use data used since 2000 land use data not available.





























Percentage Change in
Residential Area
1980 1999*
1 -95.04 --25.77
-25.76 -17.76
2 4 -17.75 -10.54
-10.53 0.02
0.03-29213.21
MilesCity Limits
St. Pet rsburg 1 05 0o 1es 2 3 City Limits
_t I-- Major Roads


Figure 3-19. Residential area percentage change 1980-1999 in the Tampa study area.

Note: 1999 land use data used since 2000 land use data not available.

































Percentage Change in
Median Household Income
1980-2000
S60.52 123.51
1 123.52 161.39
161.40 197.03
197.04 -225.04
S225.05 1218.98
m I City Limits
-- Major Roads


Figure 3-20. Median household income percentage change 1980-2000 of the Tampa study area.














CHAPTER 4
REGRESSION ANALYSES OF THE STUDY AREAS

Multiple regression analyses were conducted using NCSS software to model

variations in percentage change of median household income, commercial area and

residential area from 1980 to 2000 for each of the three study areas. The units of analysis

chosen for our study were census-tracts of the study areas. The socio-economic and

demographic data used for the dependent and independent variables were taken from

STF3 files of the US Census Bureau. Distance and accessibility index variables were

created from the Florida Geographic Data Library (FGDL) files. Additionally, several

accessibility indices were created to account for variations in locational of various

census-tracts to selected commercial nodes, including the Central Business District

(CBD), retail hubs, and the military-base. Moreover, straight-line and road distances were

calculated between both these nodes to the center points of each census-tract. Distance

variables were also included to access the impact of individual nodes on growth rates in

the selected study areas.

A forward stepwise regression was employed to determine the variables that are

relevant in explaining percentage change in the dependent variables. The variable

criterion was set at the 95% confidence level. A total of 8 multiple regression models

were tested to explain variation in growth rates by census-tract for the time period

examined.

Note that the stepwise regression for percentage change in commercial area in the

Tampa study area did not select any independent variables, thus a multiple regression









analysis of the study area could not be conducted. The 8 remaining multiple regression

models will be discussed in turn.

The key components examined in the multiple regression analyses were the

regression (or beta) coefficients, the t-Values, and identified transaction types defined by

Economic Base Theory concepts (the variables had one or more of the following

functions: basic function, nonbasic function, and/or a transfer of income function). The

sign of the beta coefficient determined the nature of the independent variable's impact on

the dependent variable. The greater the t-Value associated with an independent variable,

the greater impact on the dependent variable (growth rate). The identifier listed in the

transaction type column explains the independent variable's role on the study area's

economic base. The adjusted R-square was an indicator of the model's goodness of fit.

Finally, normality tests and plots of error terms were examined to determine if the error

structure was normally distributed, and whether there were any random outliers. If

outliers were found, they were removed and the models re-ran without them. However, if

the new results caused greater difficulties the original model was retained and the outlier

highlighted in the discussion section.

Jacksonville NAS Study Area Regression Analyses Results

Three multiple regression analyses were conducted for the Jax NAS study area, one

for each of the dependent variables. Each model will be briefly discussed.

The stepwise regression results for the dependent variable percentage change in

commercial area selected 10 independent variables with beta coefficients that were

significantly different from zero. Table 4-1 lists the results of the multiple regression

analysis for percentage change in commercial area. The initial regression analyses

revealed an outlier in census-tract 304, which is adjacent to the Orange Park commercial









node in census-tract 306. Before the removal of the outlier, the model failed to meet the

normality assumptions. On removal of the outlier the model met the standard criterion for

normality.

The estimated coefficients for selected independent variables rejected the

hypothesis that the beta values were not statistically different from zero at the 95%

confidence level. An examination of the results revealed that locational accessibility

variables and distance variables did not play a role in explaining variation in commercial

growth rates as these variables were not selected. Thus, the results of the regression

analysis explained the role of certain demographic variables on the impact of commercial

growth rates. The positive variables suggested that an increase in the demographic

variability resulted in a decrease in commercial growth in our study area. An opposite

effect was observed with negative variables, which suggested a decrease in demographic

variability result in an increase in commercial growth for the same study area. Three

independent variables with positive beta coefficients and very high t-Values (greater than

4.2) reflected an impact on commercial growth rates through nonbasic economic activity

and/or transfer income (Table 4-1). The positive beta coefficients suggested a positive

association between various demographic variables and commercial growth rates.

However, two independent variables had negative beta coefficients with high negative t-

Values (-3.7) suggesting a negative association on commercial growth rates. The negative

beta coefficients suggested decrease in commercial growth rates as either the number of

persons born out of state, service employment, or vacant housing increases.

The adjusted R-squared value was high, suggesting that the nine variables selected

accounted for approximately 83% of the variation in commercial growth rates by census-

tracts. The three normality tests were accepted. The scatterplot revealed a random pattern









between residuals and predicted values. The probability plot also revealed the remaining

variables lie within range of normality.

The regression model revealed no discernible evidence of the military-base having

a significant impact on the commercial growth rates in our study area for the period

examined. If the base had any impact on the commercial growth rates in our study area,

the impact was negligible or very minor.

The stepwise regression for the dependent variable percentage change in median

household income for the Jax NAS study area selected five independent variables for the

multiple regression analysis. Of greater interest, a distance variable (Straight-Line

Distance to Commercial Area in Census-tract 167.01) was selected for this model

indicating that a possible relationship in percentage change for the dependent variable

and distance in the localized area may be present. However, extreme outliers were

present in the model and after the removal of the outliers (associated with census-tracts 7,

5, 11, and 12) the model was left with three variables and the distance variable was

removed (Table 4-2). Interestingly, the outliers were in census-tracts that are proximal to

the CBD (census-tracts 11 and 12 are part of the CBD).

The multiple regression model produced relatively weak results in comparison to

the model reexamined earlier with an adjusted R-Square value that falls below 0.30. An

interesting variable selected was the commercial area in 2000. The negative beta

coefficient suggested the decrease in commercial area had a positive impact on the

percentage change of median household income growth rates. The positive beta

associated with median gross rent between 1980 and 2000 suggested a positive impact on

the growth of median household income during time period examined.









Similar to the regression results for the percentage change in commercial area, the

normality assumption was verified. Unlike the previous regression model discussed for

our study area, the three normality tests were not accepted until after the removal of the

outliers. Once again, this regression model failed to uncover a discernible impact of the

base on the economic growth rates as neither the distance to base variable nor the

locational accessibility variables were selected by the stepwise procedure.

The stepwise regression results for the percentage change in residential area for the

Jax NAS study area selected 21 independent variables including the intercept. Of greater

importance was the selection of an accessibility index variable and a distance variable.

The initial regression results identified five outliers (census-tracts 8, 124, 167.02, 304,

and 309). Interestingly, two of the outliers were proximal to the CBD (8 and 124); two

were proximal to the Orange Park commercial node (304 and 309); and one was proximal

to the military-base (167.02).

The multiple regression results (Table 4-3) revealed that the distance variable had a

positive beta coefficient suggesting that as distance from the CBD increases residential

growth rates increase. Thus, the residential growth rates tend to be lower near the CBD.

Also, the t-Value (20.0136) was substantially higher than the t-critical value (|2.06|),

which suggested that the distance variable had a substantial role in the positive impact on

residential growth rates. The accessibility index variable also had a positive beta

coefficient, but the t-value was substantially smaller than the distance t-Value (4.6755).

The inclusion of the accessibility variable suggested that there was a discernible impact

of the prominent nodes (including the base) on the economic growth within our study

area. Specifically, as locational accessibility to the prominent nodes increased so did the

residential growth rate. Note that the calculation of the locational accessibility index









included the location of the base, suggesting that the base was an important node in

explaining residential growth rates. The model revealed a large F-ratio and a better

overall goodness of fit (with an adjusted R-Square of 0.992), with a low root mean square

error.

The normality tests indicated that the error terms were normal. Overall, the results

supported the hypothesis that the base had a statistically significant impact on the

residential growth rates, although its' individual impact was unknown given that the

locational accessibility bundles the impact of all prominent nodes in the study region.

While it was shown that the base and the CBD had a significant impact on the residential

growth rates, the result did not imply that the base was the most influential node in our

study region. Furthermore, in all three cases examined distance to the base, as a stand-

alone variable (distance to the base) did not test to be significant. Only when distances to

all prominent nodes were considered in a bundled accessibility index did it prove

significant. In short, the base played a role in explaining variation in residential growth

rates, however, its' overall impact was non-separable from the other prominent nodes.

Mayport NS Study Area Regression Analyses Results

Three multiple regression models were run for the Mayport Naval Station study

area. The same three dependent variables that were analyzed for the Jax NAS study area

were also used for the Mayport NS study area.

The stepwise regression conducted for percentage change in commercial area for

the Mayport NS study area selected four independent variables including the intercept

(Table 4-4). The initial regression results revealed two outliers (associated with census-

tracts 149.01 and 149.02), however, the removal of both outliers created severe

complications with the ongoing model's results, particularly continual removal of