|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 By CURTIS R. DUBBERLY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTER OF ARTS IN URBAN AND REGIONAL PLANNING UNIVERSITY OF FLORIDA 2011
2 2011 Curtis R. Dubberly
3 To all who have sacrificed to help me achieve this milestone
4 ACKNOWLEDGMENTS I thank the chair and members of my thesis committee for their guida nce. I thank my parents and grandparents for their love, support and encouragement. I thank my brother, Travis, for always believing in me, and Sheri for her unwavering love and patience
5 TABLE OF CONTENTS page ACKNO WLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIA TIONS ................................ ................................ ............................. 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 L ITERATURE REVIE W ................................ ................................ .......................... 16 Evolution ................................ ................................ ................................ ................. 16 American Zones ................................ ................................ ................................ ...... 17 The Influence of Community I nstitutions ................................ ................................ 18 Focusing on Small Business ................................ ................................ ................... 19 Comparison of British and American Concepts ................................ ...................... 20 Background ................................ ................................ ................................ ............. 20 ................................ ........................... 21 ................................ ............................... 23 Pilot Project Findings ................................ ................................ .............................. 23 Results of the Pilot Project ................................ ................................ ...................... 24 Go vernment Agencies ................................ ................................ ............................ 25 3 METHODOLOGY MATTERS ................................ ................................ ................. 30 Shift Share Analysis ................................ ................................ ................................ 32 Regression Analysis ................................ ................................ ............................... 33 Other Studies ................................ ................................ ................................ .......... 37 4 METHODOLOGY ................................ ................................ ................................ ... 41 H ypothesis ................................ ................................ ................................ .............. 41 Data ................................ ................................ ................................ ........................ 41 Statistical Method ................................ ................................ ................................ ... 42 Stating the Null Hypo thesis ................................ ................................ ..................... 43 Limitations ................................ ................................ ................................ ............... 44 5 RESULTS ................................ ................................ ................................ ............... 50
6 Findings ................................ ................................ ................................ .................. 50 Discussion ................................ ................................ ................................ .............. 53 Recommendations ................................ ................................ ................................ .. 55 APPENDIX : DATA USED IN METHODOLOGY ................................ ............................ 58 LIST OF REFERENCES ................................ ................................ ............................... 61 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 64
7 LIST OF TABLES Table page 2 1 Comparison of British and American Concepts ................................ .................. 29 4 1 Descriptive Statistics For First Regression ................................ ......................... 47 4 2 Descriptive Statistics For Second Regression ................................ .................... 47 4 3 Dummy Variable Schema ................................ ................................ ................... 47 5 1 Model summary for first r egression ................................ ................................ .... 57 5 2 Model summary for second regression ................................ ............................... 57 A 1 Establishment Growth, Employee Growth, and Annual Payroll Growth fro m 2004 to 2008. Data collected from US Census Bureau County Business Pattern by Zip Code 2004 2008. ................................ ................................ ........ 58 A 2 Demographic data from shapefile created by Shimberg Center for Housing Studies at t he University of Florida 2004 ................................ ............................ 59
8 LIST OF FIGURES Figure page 4 1 Establishment growth 2004 2008 by zip code for Jacksonville Florida ................... 48 4 2 Test zones for regression ................................ ................................ ........................ 49
9 LIST OF ABBREVIATION S Adjusted R Square H ow well the model would explain the total variation if the model was run on the po pulation rather than the samples. ETI Enterprise Florida, Inc. R R Square Helps to determine the variation in (Y) given (X) SIC Standard Industrial Classification OPPAGA ysis and Government Accountability OTTED Development
10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree o f Master of Arts in Urban and Regional Planning By Curtis R. Dubberly May 2011 Chair: Andres Blanco Major: Urban and Regional Planning Jacksonville, like many cities across the U.S ., has experienced blight in the downtown area. A popular antidote to this problem is the implementation of enterprise zones. In theory, enterprise zones reduce blight through revitalization by offering incentives in the form of tax cuts and reduced regu lation. It is thought that incentives like these will encourage businesses to locate in the area. However, recent studies have shown that the majority of enterprise zone programs evaluated produce little or no effect on job growth. My study uses regress enterprise zone is generating jobs at a higher rate than the surrounding region. There are limitations in the data however, due to availability and geographical considerations. The results suggest that the ente rprise zone is losing jobs, while other areas of the region are gaining jobs. Jacksonville appears to be undermining the enterprise zone in downtown by subsidizing sprawl. It is clear that an alternative strategy is needed to revitalize the urban core.
11 CHAPTER 1 INTRODUCTION Many cities across the nation have experienced blight in the urban core. A popular scenario for American cities is the citizen exodus of the traditional downtown in search for new, cheaper homes in the suburbs. When those who can afford to escape downtown leave, blight and crime often take their place. So what can planners, governments, or legislatures do to reverse this trend? Many states, and the federal government, have embraced economic development as a tool to revitalize blig hted areas. One economic development strategy used by governments is the creation of an enterprise zone. Enterprise zones are geographically designated areas that offer tax incentives and reduced regulation to businesses in order to incentivize firms to locate in the enterprise zone. Elvery (2008 ) explains the theory behind enterprise zones: Economic theory suggests that enterprise zones can increase zone resident employment. In zones where subsidies are provided for hiring zone residents, businesses ha ve an incentive to hire zone residents rather than nonresidents because the net cost of compensation for zone residents would be lower than for nonresidents. In addition, the subsidies might be used to increase the wages of existing jobs for zone resident s, perhaps them into employment. Even if enterprise zones do not have specific incentives for hiring zone residents, one would still expect a rise in zone ities if the policies led to addit ional jobs locating in the area. (p45) In theory, the incentives offered by the zone will create new jobs in the blighted area, thus slowly revitalizing the urban core. (Butler, 1991) Jacksonville, Florida, my birth place, has experienced a similar scenario as described above. The State Comprehensive Plan designates revitalization of distressed communities as a local responsibility. (Section 187.201 (16), F.S. ) So how do local
12 governments know if enterprise zones are effec tive? My study will examine the effects enterprise zone will effectively revitalize the urban core. First, my study will review the evolution of the enterprise zone. The rev iew will show that the original concept of enterprise zones, originally proposed in Britain, has been tweaked to fit American goals and objectives. The review will also examine the s green lighted the creation of a zone based on less than convincing evidence provided by a pilot project. Second, my study will review the base of literature on enterprise zones in order to reveal what scholars have learned about the effectiveness of ente rprise zones. The review will examine multiple methodologies and determine which methodologies are most effective in measuring impacts on employment and income. The literature review will show that empirical studies conducted on the effects of enterprise zones produce mixed results and very few studies attribute any growth in employment or income to the creation of enterprise zones. Third, using 2004 2008 county business pattern data by zip code, from the US Census, and Florida Geographical Data Library d ata used in conjunction with ArcGIS, significant growth in employment. The 2004 2008 period was chosen because the Florida Enterprise Zone Program was renewed in 2005. Stu dying this period will allow us to see how Jacksonville has developed since the renewal of the program. Furthermore, data for evaluating enterprise zones is hard to come by and the data used
13 for my study must come from the same time period. 2004 2008 is the most recent period that all the data needed was available. My study will use this data to run two significantly higher rate than the surrounding region. Finally, my study will examine the findings of my statistical tests and try to rate than non zone areas. The findings will be followed by a discussion on what my study revea ls. Jacksonville : A profile Jacksonville, Florida is a port city located at the mouth of the St. Johns River in north Florida. It is the largest city in land area in the United States. This is due to the consolidation of the city and county governments and the extension of the city limits to cover nearly the whole county. Jacksonville has become an important city in the region due to its location at the mouth of the St. Johns River. eater Jacksonville The city of Jacksonville has a population of 802,843 with 426,592, about 69%, in the labor force. 7.5% of the civilian labor force is unemplo yed. The industries that suppor t the most employ ees are the retail trade, finance, insurance, real estate, education, healthcare, and professional and administrative. The 2009 median household income for Jacksonville was $48,514, which is below the median household income for the US ($51,425). Jackso nville also has a large African American population, about 30% of the total population. Hispanics make up a smaller portion of
14 the population, around 6%. (US Census Bureau, 2005 2009 American Community Survey) th of the river has allowed the city to grow both in size and in economic terms. The first two big industries to take advantage northern Florida had been relatively untouched in terms of logging. These forests would provide a booming business for Jacksonville. By 1900, Florida was producing 30 50 percent. Abundant rainfall and easy access to shipping lanes allowed the industry to could be transported to multiple markets throughout the country. Although timber production has dropped off, shipp ing continues to be a very important aspect of The St. Johns River has also given the military an ideal harbor in which to conduct its military operations. Along with Mayport Naval Station, Jacksonville is home to Na val Air Station Jacksonville, Blount Island Command, and The Florida Air National Guard. Camp Blanding and Naval Submarine Base Kings Bay are also nearby. With so many military installations around, a citizen soldier culture has developed. The military has dispersed to other births. If this keeps happening the city of Jacksonville could pay a heavy price.
15 houses are a common site and the houses that are not abandoned are in disrepair. As a child I remember my parents rolling up the windows and locking the do ors whenever we drove through the downtown area because they were afraid. Crime has steadily been well above the national average in the past decade. This fact has been a contributing factor for the exodus of the downtown area. The United States, includ ing Jacksonville, experienced a boom after the end of the Second World War. Suburbs sprang up along the country side to house all the soldiers and sailors coming back home from the war. The development of these suburbs added from the city and caused the massive urban sprawl that is familiar to Jacksonville today. the mass exodus of white families from traditional downtowns to commuter towns. With a depleting revenue source the city could not afford to keep up with funding have any municipal services. These factors contributed to the consolidation o f Jacksonville and its surrounding suburbs. This was an attempt by the city to get back the revenue from the tax payers it had lost. The 1964 public school fiasco only made things worse for the city. All fifteen public high schools lost their accreditat ion at the clear that reform was needed in order to improve the city. In 1968 the county and city governments were consolidated in order to bring the badly needed refo rm.
16 CHAPTER 2 LITERATURE REVIEW Evolution The term enterprise zone first appeared in 1978 in a speech about urban blight by Sir Geoffrey Howe, a politician in the British House of Commons. At a time when the creation of government programs were the de fault solution for solving social problems, his proposal of creating an enterprise zone was viewed as radical. This is because Sir Geoffrey Howe envisioned enterprise zones as being nearly completely devoid of government regulation in the most dilapidated unhindered private enterprise would prevail where government programs had failed. Therefore, he proposed designating specific geographic locations as virtually tax and regulation free havens for free enterprise t o flourish. (Green, 1991) Instead of hindering business with regulation, Howe wanted to instead create conditions that would allow private enterprise to prosper. In essence, Howe was areas back to their former glory. Howe modeled his enterprise zone concept after the free trade zones of the world, such as Hong Kong. He observed extremely rapid economic growth in these areas and attributed it to low taxes, low regu lation, and few land use restrictions. Bendick and Rassmussen (1986) accredited the rapid growth among newly industrialized third world high level of individual work e of Hong Kong are slightly different but the main idea is the same; uninhibited free trade brings rapid economic growth.
17 er to bring economic activity to virtually abandoned areas that were devoid of any significant population. This is different from the American view of enterprise zones, which viewed the new strategy as a way to revive specific small communities. Butler e quates the British vision with inner catalyst for further economic development in the metropolitan area, which may or may not lead to an influx of migrant citizens. The British vision was never i ntended to reinvigorate slums. (Papke, 1991) The British enterprise zones waved property taxes for commercial and industrial buildings (not residential buildings), implemented a one year tax write off for building and machinery expenses, and exempted cust oms duties on imported materials used for the manufacturing of exports. This tactic was really only useful to large companies with deep pockets and large tax dues. Thus, the small business owner did not directly benefit from the British enterprise zone l egislation, which was never the primary objective in the first place. American Zones The American objective of using enterprise zones to revitalize blighted urban neighborhoods is based on the idea that poor neighborhoods and poor people can revitalize th emselves if given help. Proponents of enterprise zones believe there are potential is smothered in red tape, excessive taxation, and a culture of welfare 991, p32). There is some disagreement among supporters of enterprise zones however. Some believe that incentives alone will not awaken the supposed sleeping giant of human capitol in poor neighborhoods. These critics think
18 that in order to trigger growt h, more focus on neighborhood improvement is needed in conjunction with financial incentives. Other critics believe that tax cuts and reduced regulation should not be the means to revitalize blighted areas; rather governments should focus on providing ser vices and infrastructure. (Fisher & Peters, 2000) The Influence of Community Institutions Rather than emphasizing physical redevelopme nt, as was the case in the 1950s and 60 s, American proponents of enterprise zones stressed the importance of preserving a nd strengthening the community institutions that already exist. Jane Jacobs, a vocal supporter of community institutions, saw a problem with conventional urban development. While most planners emphasized separation of uses, larger scale planned developme nt, and large streets and blocks, Jacobs maintained that the complex social relationships of a community were at the heart of successful urban development. She realized that the large scale developments ignored the sense of place, thus removing people fro m their established context and historical identity. main ideas were: 1) The facilitation of mixed use buildings and preservation of old buildings 2) Dense populations 3) Small block sizes 4) Decentralized ownership work heavily influe nced the way Americans viewed enterprise zones. The idea of place became important. Enterprise zones were seen as an anti poverty tool that addressed the context of the people living in distressed neighborhoods. (Jacobs, 1961)
19 Focusing on Small Business There are two main reasons why proponents of enterprise zones support the promotion of small business. Enterprise zone advocates argue that small firms are the biggest generator of new jobs, and targeting small firms will help with the promotion of local economic activity. Around the time enterprise zones first appeared, David Birch (1979) of MIT conducted a study of the job generation process. He was interested in finding how firm behavior causes change and what policies to implement based on the info rmation found. He conducted a survey of 5.6 million U.S. firms using data from a credit rating company. He confirms that small enterprises (less than 20 employees) are the biggest net generators of new jobs, while very large firms have a tendency to be a net destroyer of jobs. According to Birch, the rate of job loss is about 8% across the nation, both in growing and declining areas. Therefore, the defining characteristic between growth areas and declining areas are the creation of new firms. (Birch, 1 979) Another reason why American enterprise zones focus on small business is that small firms make a better fit in the local economy. Butler points out that large start up firms tend to need new facilities, unlike small firms that can move into existing s tructures. Furthermore, small firms are more likely to hire local, unskilled labor than larger firms. Since the local, unskilled labor population is the primary focus of neighborhood improvement, in makes sense to encourage the growth of businesses that hire th e distressed population. (Green 1991)
20 Comparison of British and American Concepts The British concept of enterprise zones differs slightly from the American concept. Bipartisan legislation and the influence of community activists have modified t he original concep t of enterprise zones. Table 2 1 illustrates these differences. If the original concept of enterprise zones has changed, does this mean that the effects of the enterprise zone will change as well? To determine if these changes have aff ected enterprise zone performance in the United States we must look at how the program is administered, implemented, and how it performs. Background The Florida Legislature revised the existing Enterprise Zone Program in 1994 with the adoption of the Florida Enterprise Zone Act. The act repealed existing zones and Empowerment Zone/Enterprise Commu nity programs. ( Florida Enterprise Zone Program Annual Report, 2010) In 1995, local governments competed for zone designation and nineteen new enterprise zones were formed. Shortly after, in 1996, the program received another ation responsibilities were transferred to the OTTED (Use of Enterprise Zone Incentives Has Increased, but Challenges Continue, 2000 p5). Since that time the number of enterprise zones in F lorida has risen to a total of 56. ( Florida Enterprise Zone Program Annual Report, 2010) The program was set to expire in 2005, but the Florida Legislature extended the program for another ten years. During this time, the local zones could amend their
21 bou ndaries and be re designated to be in effect until December 31, 2015. (Florida Enterprise Zone Program Annual Report, 2010) The City of Jacksonville started the Enterprise Zone Pilot Project in 1999 in a distres sed inner city neighborhood. The pilot project area was designated in the Gateway Shopping Center, an inner city mall that has seen many of its tenants vacate in search for greener pastures. Only one business qualified and took advantage of the pilot proj ects incentives. Other chain based businesses located in the pilot project area but did not qualify for the incentives. However, these businesses were able to take advantage of state Enterprise Zone incentives. The pilot project has had a limited effect but the area has had some success with revitalization. (Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Made in Revitalization Project Area, 2003) The Florida State Legislature decided to designate one test area, or pilot project, i Development (OTTED) was given authority over the project and intended to provide employment for the local area citizens. The OTTED selected the pilot project area based o n the following set of criteria: The area must be contained within an enterprise zone that is composed of one contiguous area in a community with a population of 150,000 or more persons. The local government must grant economic development ad valorem tax e xemptions and electrical energy public service tax exemptions in the enterprise personal property. In other words ad valorem taxes can be property tax or even duty on import ed items. Property ad valorem taxes are the major source of The local government must develop a plan and commit at least $5million for revitalizing the pilot project area or an area that includes the pilot project area.
22 The pilot project area must be contiguous and no more than 70 acres so that state assistance can be effectively concentrated on revitalizing the acute area of economic distress. The pilot project area must contain a d iverse group of facilities or space for a mix of retail, restaurant, or service related businesses necessary to an overall revitalization of surrounding neighborhoods. Source: Office of Program Policy Analysis and Government Accountability, Report No. 03 6 4 The following eligibility requirements were specified for businesses to access the Enter into a contract governing the lease of commercial space in a facility located within the pilot project area; Begi n operations in a facility located in the enterprise zone between July 1, 1999, and July 1, 2000; and Be engaged in activities classified under three Standard Industrial Classification (SIC) codes 5311 (department stores), 5399 (general merchandise stores ), and 7832 (movie theaters). Source: Office of Program Policy Analysis and Government Accountability, Report No. 03 64 Under Florida State Law, no more than four businesses could be eligible for the incentives offered under the pilot project. The limit to the amount of tax credits to be awarded was $1 million annually. Eligible businesses would receive $5,000 in tax credit for every full time worker and $2,500 in tax credit for every part time worker. These credits could be applied against corporate in come or sales and use tax in a taxable year. (Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Made in Revitalization Project Area, 2003) Along with the pilot project incentives, businesses could also receive tax incentives from the s
23 sales tax incentives, corporate tax incentives, sales tax exemptions for businesses that produce motion pictures (for theaters or television), television series, commercial advertisi ng, music videos or sound recordings, tax credits for contributions to nonprofit scholarship funding organizations, and renewable energy tax incentives. (Florida Department of Revenue, 2010) Out of five potential app licants (Jacksonville, Miami, Tampa, Orlando, and St. Petersburg) the City of Jacksonville was the only city to apply for the pilot project. The City of Jacksonville proposed that the Gateway Shopping Center be designated as the pilot project area. This inner city commerce center was an ideal candidate because in a ten year span from 1989 to 1999 the shopping center lost most of its major tenants to suburban malls. The Gateway Shopping Center went from an occupancy rate of 95% (employing 1,500) in 1989 t o an occupancy rate of just 29% (employing fewer than 300) in 1997. The Gateway Shopping Center was designated as the pilot project area in 1999. (Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Made in Revitalization Project Area 2003, p3) Pilot Project Findings According to the Office of Program Policy Analysis and Government Accountability, only one applicant out of a total of three was approved by OTTED to receive program incentives. The two businesses that did not receive de signation did not qualify because the OTTED had determined the businesses could not be classified under the specified Standard Industrial Classification (SIC) code. The business that did qualify for program incentives received $40,000 in tax credit in the year 2000 to be applied towards corporate income tax. There is some evidence that the incentives did
24 not have the desired effect. A fact that lends credence to this notion is, according to the under the pilot program in (Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Made in Revitalization Project Area, 2003, p3). Results of the Pilot Project The Office of Program Policy Analysis and Government Accounta bility reported in 2003 that some revitalization progress has been made in the pilot project area, though only one business received any form of tax credit from the pilot project. By then, the arly 1,200 people. Clearly, the tax incentives offered to one business was not the factor that influenced this rise in employment. (Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Made in Revitalization Project Area, 2003) It is unc lear whether the revival of the Gateway Shopping Center can be attributed of the pilot project did not reapply in following years. It seems that if the incentives were effective, the business would have jumped at the chance to receive more credits. It increased from 29% in 1997 to 92% in 2003 because of the tax incentives received by one business in one year. Clearly, other factors contributed to the increase in occupancy rate. The possibility of receiving tax credits may have factored into the th development incentives are ineffective in attracting new businesses because of the limited weight businesses give to such incentives in location and investment
25 (Rubin & Wil der, 1989, p 419). The Department of Revenue does not give out information about the specific stores that receive tax credit, so it is impossible to know if Shopping Ce nter. There is not enough evidence to conclude the magnitude of the effect of the Enterprise Zone Pilot Project. This is a common occurrence when organizations or state o ffices have kept records on job development, this information lacks the level of detail necessary to identify the component of job generation that is due It is very likely that the growth i n the area was due to factors other than enterprise zone incentives. It appears that Jacksonville implemented its enterprise zone based on an unsuccessful trial. The revitalization of the area cannot be attributed to the incentives offered by the zone, so why continue with the program when the economy appeared to fix itself? This suggests that part of the problem with enterprise zones are the government agencies charged with administering them. Government Agencies It is important to consider who is admi nistering a program that you are evaluating. The performance of a given program may be a reflection of the program administrators. rise zones. OTTED has many responsibilities inherent to administering the Florida Enterprise Program, including: (Florida's Enterprise Zone Program Is Similar to Those of Other States, 2004)
26 Approving enterprise zone designation applications Approving enterprise zone boundary changes at the behest of the Florida State Legislature Conducting annual reports on the program to be submitted to the Governor and the Legislature Providing technical assistance to local program administrators Assisting businesses Monitoring local enterprise zones through pe r iodic site visits. OTTED also works closely with the Florida Department of Revenue and Enterprise Florida, Inc. (EFI). It is the responsibility of the Florida Department of Revenue to businesses in the enterprise zones. EFI is a public private partnership created in 1992 by the Florida Legislature. The r of high technology jobs in Florida by promoting small and mid size businesses within the state. It received further responsibilities in 1996 when the Legislature abolished the Department of Commerce development and international trade programs (Concerns Over Enterprise Florida's Performance, Servi c es to Distressed Areas Point to a Need to Consider Several Alternatives for Its Future Role, 2001) EFI also handles the marketing for the Enterprise Zone Program by marketing to businesses seeking designation within an enterprise zone. (About Enterprise Zones, 2010) Overseeing OTTED, EFI, and the Department of Revenue is The Florida nalysis and Government Accountability evaluative research and objective analyses to promote government accountability and
27 About OPPAGA, 2010). OPPAGA has conducted several program reviews of government agencies. It has submitted 56 such reports, dating back to 1994, on the Executive Office of the Governor, including Enterprise Florida programs and OTTED. (Concerns Over Ent erprise Florida's Performance, Servi c es to Distressed Areas Point to a Need to Consider Several Alternatives for Its Future Role, 2001) The multiple program reviews conducted by OPPAGA revealed some short comings in the Florida Enterprise Zone Program. In March 2000, OPPAGA reported that the use of incentives offered by Enterprise Zones had increased since the revision did not prepare the (Use of Enterprise Zone Incentives Has Increased, but Challenges Continue, 2000, p1) Thus, OPPAGA was unable to assess any changes in socio economic conditions. In December 20001 OPPAGA reported that EFI had met established performance (OPPAGA,2001, p1). The report also notes that EFI is not providing certain types of services needed by rural economic development organizations. Spending The 2010 Florida Enterprise Zone Program Annual Report shows that from 2004 through 2009 the program added a state wide total of 54,000 new jobs a t a cost of $229,045,943. That is about $4,241.59 per job. The total cost is the sum of local
28 incentives ($70,122,628) and state incentives ($158,923,315) offered. This does not take into account the man hours it takes to staff the multiple agencies tha t deal with enterprise zones, so the total cost of running the state wide program is probably much total amount that state and local governments have spent on the progra m must be huge. The report shows that Jacksonville provided $5,350,000 in tax incentives between October 1, 2008 and September 30, 2009. The money used to provide tax incentives is money that the local government could have used elsewhere and in other wa ys. So by implementing and sustaining an enterprise zone, the local government has forgone other options for economic development. With this in mind, it becomes important to make sure that the programs funded by tax payer dollars are worth the money. So the area? The original enterprise zone concept was molded by politics and activism into the form states use today. Will this modified concept produce the same effects as ori ginally envisioned by the original architects of enterprise zones? The findings of the Jacksonville Enterprise Zone Pilot Project and the reports issued by OPPAGA, both unable to attribute any socio economic changes to enterprise zone incentives, suggest money spent on enterprise zones, both locally and statewide, the question become s even more important.
29 Table 2 1 Comparison of British and American Concepts British Tactics (original concept) American Tactics (modified concept) Deep tax incentives intended to spur major physical development. Provisions aimed at modest improv ements of existing commercial structures. Aimed at bringing business to virtually abandoned areas of the city. Aimed at generating economic activity in poor neighborhoods. Intended to generate general economic growth. Intended to reinvigorate slums i n specific geographic locations. Helping poor communities is a perceived byproduct of program, not primary objective. Primary objective is to revitalize poor communities. No provisions for existing houses. Incentives for improving existing housing. Tax incentives to employ machines rather than people. Tax incentives to encourage the hiring of low skilled or disadvantaged labor. No evidence of focus on community institutions. Community institutions are crucial to economic development. Target lar ge businesses. Small businesses should be favored over large ones. Source: Butler, S. (1991). Enterprise Zones: New Directions in Economic Development. Newbury Park: Sage.
30 CHAPTER 3 METHODOLOGY MATTERS Since states started implementing ent erprise zones in the early 1980 s, a number of empirical studies have been conducted to determine the effects of enterprise zones on local development. A January 2005 Minnesota House of Representatives Research uncover little net benefit to enterprise zones. However, studies and their results vary widely, delivering mixed (Enterprise Zones: A review of the Economic Theory and Empirical Evidence, 2005, p2) Many types of methodologies are used to det ermine the effects of enterprise zones. Not surprisingly, these various methodologies deliver various results. Some methodologies are better than others at dealing with the two main problems of measuring the effect of enterprise zones. Rubin and Wilder examined multiple empirical studies and found, The modest amount of empirical research is due to two basic constraints: (1) the lack of reliable quantitative data to evaluate zone performance, and (2) the difficulty of isolating the effects of zone design ation and incentives from those of other economic deve lopment factors and initiatives. (Rubin and Wilder, 1996, p2) Other than obtaining enough reliable data, the main problem is that researchers have difficulty accurately measuring the benefits deliv ered by an enterprise zone. It is difficult to measure these benefits because there is no way to know what would have happened but for the zone designation. In other words, it is difficult to isolate the effects of the zone i.e., employment growth and income growth from effects caused by demographics, local characteristics, economic factors, and so forth. The 2005 Policy
31 they still cannot completely separate the effe cts of a zone from growth in the community (Enterprise Zones: A review of the Economic Theory and Empirical Evidence, 2005, p9). One of the earliest and most publicized evaluations of enterprise zones was completed in 1989. Rubin and Armstrong ( and concluded the program had created 9,193 jobs between 1985 and 1988. Later, a study by Boarnet and Bogart (1996) examined the same program for the same time period. The later study revealed no job growth in seve n of the zones. How could two studies examining the same program during the same time frame come up with such different conclusions? The difference in conclusions is due to variations in methodology. (Boarnet, 2001) s program based on survey methodology. A survey was given to businesses located in the zone in 1988 and asked them to assess the influence of zone incentives on their businesses location decisions. One obvious problem with this methodology is that there is no control group. Since the survey was only given to businesses in the zone, the results cannot be compared to non zone employment rates. Thus, it is difficult to gauge what might have happened had the zone not been there. (Boarnet, 2001) The second study, Boarnet and Bogart (1996), did use a control group in a regression analysis. Since the second study controlled for the overall health of the regional economy, we can be more confident that the conclusions are accurately representative. (Boarnet, 2 001)
32 Survey studies have many limitations, reducing the confidence that the results are accurately representative. Surveys must rely on data provided by businesses in enterprise zones and thus may contain bias. Also, surveys cannot account for outside fa (Enterprise Zones: A review of the Economic Theory and Empirical Evidence, 2005, p10) Surveys also cannot determine whether a business would have located in the area had an enterpris e zone not been there. Because of these limitations, my study will focus on more sophisticated analysis. Shift Share Analysis Rubin and Wilder (1989) studied the Evansville, Indiana, Enterprise Zone impacts on employment using shift share analysis. They relatively cost effective job The study used data collected from surveys on firm characteristics, job development, foregone tax revenues, and employment levels before zone designation in 1983, until the end of the study in 1986. The shift share analysis revealed a statistically significant difference between the employment growth of the zone and the metropolitan area. Rubin and Wilder also conducted an analysis to de termine the cost per job and found that the average annual cost per job was $1,372. According to Rubin and Wilder, this cost per by local governments to stimulate econom ic activity. Shift share analysis is better than survey method in that it controls for some effects. However, shift share analysis has short comings as well Shift share analysis breaks down economic growth into three components: National share, indust ry mix, and
33 regional shift. The analysis will show how much each component contributes to local economic growth. Shift share analysis attempts to isolate job growth due to enterprise zone designation, but its main problem is that it assumes that job grow th in an enterprise zone is proportional to job growth in the region. Rubin and Wilder state that, the larger region and shared ). This assumption is a serious limitation because conditions in the enterprise zone may be different from overall conditions in the region. The growth inside the zone maybe related to growth in the region, but not in the neat, proportional way that shif t share analysis assumes. (Enterprise Zones: A review of the Econooic Theory and Empirical Evidence, 2005) Regression Analysis A more sophisticated method of analysis of enterprise zone benefits is regression analysis. A regression model can show how muc h of the growth occurring in enterprise zones is attributable to incentives offered by the zone and how much of the growth is attributable to other factors i.e. demographics, local characteristics, economic factors and so forth. Dabney (1991) studied h ow enterprise zone incentives affect business location nontraditional location factors causing enterprise zone incentives to be unable to offset the deficiencies of major loc ). The study examined enterprise zones in Maryland, Ohio, Illinois, Kentucky, Connecticut, Pennsylvania, Missouri, and Kansas. Dabney used annually updated data on
34 observations of the percentage rates of change in the number of establishments in the eight case study zones. The study used observations from two separate time periods; 1980 to 1982 (before zone designation) and 1982 to 1984 (after zone designation). Dabney tested whether zone designation affected the rates of change in number of establishments using an analysis of variance statistical technique. The data showed, influential in determining rates of change in the num (Dabney, 1991) Dabney concluded that the benefits of the tax incentives offered by enterprise zones are often outweighed by the higher economic costs of locating in the area; costs associated with insurance, transportation, and the acc ess to raw material. (Dabney, 1991) The Minnesota House of Representatives Research Departments policy brief on enterprise zones reviewed several studies that used regression analysis. The results are mixed; while some studies show a temporary gain in job s and income, 5 studies found no increase in jobs. (1991) policy brief noted that the study found a reduction in unemployment claims Papke used regression models to predict the effec ts of zone designation on labor markets. To do this, Papke used data on unemployment claims, machine and equipment capitol, and inventories from 1980 to 1988. The regression revealed that zone designation produced a 9.8 % decline in machinery and equipme nt, and an 8.3% increase in inventories. Designation also produced a 25% decrease in unemployment claims at the nearby claims office.
35 Moore (2003) used regression analysis to evaluate the effects enterprise zones have on the number of firms. The study examined several zones within the state of California from 1987 to 1991. Moore used County Business Pattern data from the US Census Bureau to conduct a Two Way Fixed Effects Model. The data includes Standard Industrial Classification codes, number of est ablishments, number of zone program revealed an increase in jobs in some employment classes and job losses in others. The sectors that showed growth were retail, real est ate, finance, and insurance. Moore notes that while enterprise zones did create some firms and thus some new jobs, the research has limitations that can create doubt as to whether enterprise zones improve socio economic conditions. The limitations includ e controlling for the migration effect (most people do not live in the zone but commute to work), geographic mismatches (Enterprise Zones do not geographically correspond to data tracts like zip codes, census tracts, etc.), and lack of a cost benefit analy sis. (Moore, 2003) temporary increase in employment, but no positive effect upon earnings. The study h at the propensity score matching model to compare enterprise zone census tracts to non enterprise zone census tracts. The study used annual establishment level employment data to conduct the model. The results show that while enterprise zones increase employment by about 3% a year, the increase is not permanent. (O'Keefe, 2004)
36 (2001) study showed that employment within enterprise zones grew at a faster rate than the surrounding county. Beck used Census data on job growth and establishment growth to conduct a regression that examined 51 enterprise zones across the nation. In the study, Beck tested three types of causal variables. The first type included the number of different tax, financial, infrastructural, and quality of life improvements offered by each zone. The second type of causal variable measured the effort put into a zone, while the third type is spatial factors. He notes that the only significa nt predictors in the regression model were quality of life factors, which could either be offered in the zone or offered outside the zone. This study adds credence to grow th are things that are not exclusive to the enterprise zone itself. Hence, the evidence suggests that it is not the zone incentives that are influencing business location decisions, but rather the amenities offered by quality of life improvements. (Beck, 2001) births, while another study of many state enterprise zone programs showed no increase in per capita employment. Engberg co authored a study with Bondonio (Bondonio & Engbe rg, 2000) covering five state enterprise zone programs and found no changes in employment, income, or poverty rates. showing no impact on employment. The study used municip al level data to conduct an econometric analysis on enterprise zones from 1982 to 1990. The data includes
37 regression revealed that the enterprise zone program had no effect on munici pal level employment. (Boarnet & Bogart, 1996) (Fisher & Peters, 2000) study of enterprise zones results is uncertain. It might be par tly related to the statistical method, the data used for (Enterprise Zones: A review of the Economic Theory and Empirical Evidence, 2005, p11). Other Studies Peters and Fisher (2 004) take a position along the same line as Dabney. Based to micromanage economic development is limited; governments should instead focus on providing quality of lif e improvements and let the economy regulate itself. Peters and Fisher see traditional incentive based economic development as a waste of 10% of the time, and are simply a w 2004, p32). iew of the research on enterprise zones reveals that while the number of programs, the number of zones designated, and the land area 66). According to the authors, the body of
38 research conducted on enterprise zones is enough to merit no confidence in the program. Elvery (2008) used employment probability models along with neighborhood level propensity score matching to evaluate the ent erprise zone programs of Florida and California. Elvery used a three stage estimation strategy for the study. The first stage used employment probability models to determine the correlation between employment and neighborhood characteristics. These chara cteristics include race, education, marital status, age, immigration status, and number of children. The next stage, (Elvery, 2008) The third stage matches the propensity scores to estimate the effect of enterprise zone policies on resident employment. He found, In California, the estimated effect on the raw employment rate of all men ranges is 1.69 percentage points, and the estimated effect on the neighborhood component of em ployment probability is 2.55 percentage points. In Florida, the estimated effect on the employment rate of all men is 3.47 percentage points, whereas the effect on the neighborhood component of employment probability is 1.87 perce ntage points. (p56) T hese findings suggest to the author that enterprise zones reduce the likelihood that a man living in an enterprise zone is employed. Business Location Factors The concept of enterprise zones is based on the premise that tax incentives offered by the zone will influence the business location decisions of new firms to the extent that they decide to locate within the zone. But what are the factors that businesses consider when deciding where to locate? Is the tax structure of a location really the deciding factor when determining where a firm is to be located?
39 It is difficult to measure what factors influence business location decisions for two reasons. Data is not readily available and it is difficult to measure the interaction between location character istics like land prices, tax structure, labor force, and the presence of unions. Bartik (1985) studied the location decisions of a new firm using a conditional logit model and found some evidence that contradicts traditional location theory literature. C onventional wisdom on the subject maintains that state taxes do not influence business location decisions. However, Bartik found that state taxes did play a study showed that a 10% increase in state corporate tax caused a 3% decline in new resulted in a 2% decline in new firms. Other factors that had significant effects on location decisions were the provision of infrastructure, wages, existing activity, and the presence of unions. Factors that had no significant effect were education, construction costs, population density, and energy prices. (Bartik, 1985) Interestingly, as state d earlier, Dabney also promotes infrastructure improvements as a way to attract new firms. From this evidence, it seems that a better strategy for enterprise zones would be to provide increased public investment in quality of life factors instead of provi ding tax incentives. Many economists doubt the significant effect tax incentives have on location decisions. Even Bartik admitted that his study showed little effect. However, there does seem to be a consensus that infrastructure improvements will incre ase the birth of new firms.
40 Methodology matters when assessing the impact of enterprise zones. Some studies are more sophisticated than others and they sometimes contradict each other. Even though the more sophisticated studies attempt to control for external factors, they still cannot prove what would have happened had the zone not been there. But we can be sure that the more sophisticated studies, those using regression analyses, are better at measuring changes in socio economic conditions. The more sophisticated studies have been able to attribute very little growth to enterprise zone incentives. The majority of the literature suggests that enterprise zones have no permanent effect on socio economic conditions in the zones. Furthermore, a common t heme among researchers is that public investment in infrastructure is a significant factor in business location decisions, while the findings of effects of tax incentives on location decisions appear to be mixed. Yet, every enterprise zone is different. Zones vary in size, economic and demographic conditions, location, incentives, and many other factors. While the literature is convincing in dispelling the illusion that enterprise zones will be the cure all for local governments, further study is needed
41 CHAPTER 4 METHODOLOGY Hypothesis am began in 1999. Since that until 2015. It is now 2011 and there are very few visible signs that downtown Jacksonville is better off from having an enterprise zone. However, that does not mean that progress has not been made. In order to measure the success of the enterprise zone, I will use ArcGIS mapping software to map the growth of businesses. This should demonstrate where new firms are locating; if enterprise z one incentives really work, we should see the majority of establishment births happen in the enterprise zone. Furthermore, I will compare the average annual growth rate of employees from the enterprise zone with the rest of the region. Again, if Jacksonv working then it should have a higher average annual employee growth rate than the rest of the region. Furthermore, I will conduct two regression models to predict what factors are influencing the average annual employee growth ra te. Data The data used in my study primarily comes from the US Census. Using 2004 through 2008 County Business Pattern by Zip Code data from the annual economic survey, I compiled figures for number of establishments, number of employees, and annual p ayroll totals for all Standard Industrial Classification codes for fifty zip codes in the Jacksonville Metropolitan Statistical Area. The totals do not include figures for government or rail establishments. I used this data to calculate growth in
42 establi shments, employees, and annual payroll for the four year pe riod between 2004 and 2008 ( Table A 1). I also used these figures to calculate the annual growth rate of employees for the Enterprise Zone and the surrounding Urban Zone and the outer Suburban Zon e. The annual growth rate is calculated by dividing the rate of change by the number of years you are observing; in this case 4 (2004 2008). The rate of change is calculated using the following formula: Rate of Change = ((value at end of period value a t beginning of period)/value at beginning of period) 100 Using ArcGIS mapping software I mapped all the zip codes for the Jacksonville Metropolitan Statistical Area using maps created by the University of Florida Shimberg Center for Housing Studies in 2 004 and downloaded from the Florida Geographic Data Library. The shapefile I downloaded from the Florida Geographic Data Library also contained figures on black population, Hispanic population, number of housing units and number of households for each zip zones over the zip code map to determine which zip codes corresponded to After entering the establishment growth from 2004 to 2008 into the Zip Code layer attribute table I was able to determine which zip codes had experienced the mos t growth in establishments. ( Figure 4 1) Statistical Method For the first regression I used a dummy regression to determine if the annual employment growth rate in enterprise zone zip codes was significantly different than non zone zip codes. The dependant variable in my regression is annual employment growth rate. The predictors (independent variables) for my regression are annual establishment growth rate and annual payroll growth rate. To control for area
43 demographics and conditions, figures for black population, Hispanic population, number of housing units, and households were also tested as predictors of employment growth. My dummy variable was coded as 1 = enterprise zone, 0 = non enter prise zone. This will determine how much enterprise zone designation affects the dependant variable, in this case annual employment growth rate. Table 4 1 gives the descriptive statistics for the first regression. For my second regression I will test if the annual employment growth rate for enterprise zones is significantly different from other parts of the Jacksonville Metropolitan Statistical Area. Table 4 2 gives the descriptive statistics for the second regression. The metropolitan area was divided i nto three parts; Enterprise Zone, Urba n Zone, and Suburban Zone. ( Figure 4 2) These three zones are the dummy variables for my second regression. The schemas for the dummy variables are illustrated in Table 4 3. Stating the Null Hypothesis My regressions are testing the models ability to predict the variation in employment growth given the predictors. In other words, we are testing if factors like enterprise zone designation, establishment growth, Hispanic population, etc. are affecting employment growth Furthermore, the coefficient determination of the regression will determine the magnitude and direction (positive/negative) of the variation in employment growth given the predictors. Therefore the Null Hypothesis is: There is no difference in the pr ediction of (Y) given (X); Y = employment growth, X = predictors
44 Limitations Boarnet (2001) explains that while methodology plays a big role in determining the effects of enterprise zones, data also affects the outcomes. I deally, an enterprise zone e valuation should have outcome data for both zones and a well chosen control group. These data might include information on jobs and wages within zones and employment r ates among targeted populations (Boarn et, 2001, p248) Even though I have a control gro up, the data does have some limitations. For example, the data for establishment growth, employment growth, and annual payroll enterprise zone does not geographically co rrespond to zip code boundaries. Therefore, some zip codes are completely within the enterprise zone while others only contain parts of the enterprise zone. Because of this, the figures obtained from the US Census may be skewed. Some growth could have o ccurred in a zip code that contains parts of the enterprise zone without actually growing in the enterprise zone itself. Boarnet (2001) explains why this limitation is common to evaluating enterprise zones, Zones are often smaller than municipalities or other jurisdictions, such that zone boundaries do not correspond to any geographic unit for which data are regularly available. This was the case, for example, in New Jersey, where Boarnet and Bogart (1996) used municipal data in lieu of information on zo ne employment. Similarly in California, Dowall (1996) used geographic information systems (GIS) technology to apportion zi p code employment data to zones. (p248) Furthermore, the data contains figures for employment in a specific place in my case zip cod es. Therefore, we only know the employment figures for the zip code and not for individuals. So the enterprise zone may see growth in employment, but there is no guarantee that the zone is increasing employment for zone residents. People often
45 commute t o work, so the figures for growth may contain employment of non zone residents. (Boarnet, 2001). The data for establishment growth, employment growth, and annual payroll growth were obtained for 2004 through 2008. This time period is important because it encompasses the year the Florida Enterprise Zone Program was renewed (2005). This will help to evaluate program performance since the renewal. The 2010 Florida Enterprise Zone Program Annual Report shows that total state incentives have gone from $14,060 ,057 in 2004 to $23,433,535 after renewal of the program in 2005. From 2004 2008 the total state incentives increased by $31,291,384. The time frame does present some limitations however. The four years observed in my study also correspond to a nationa l recession. Downturns in the economy may affect zone performance. A bad economy could render successful zones unsuccessful. Thus, it is important to compare the performance of the enterprise zone to the performance of the rest of the region. This is b ecause the recession affected the region fairly equally, so even though an enterprise zone is not experiencing positive growth it could be performing better than the rest of the region. For this reason, I have compared data from the Enterprise Zone to dat a from the Urban Zone and Suburban Zone. Furthermore, four years may not be long enough to get a true sense of how the enterprise zone has changed socio economic conditions. Ideally, the data should include figures for years before zone designation all the way to the present However, limitations of locating such data are presently insurmountable. Also, because the data from the US Census corresponds to zip codes, I was forced to use predictors in my regression that contained f i gures that also correspo nd to
46 zip code. Very little data is available in this way so the number and type of predictors are reflections of the limitation of the data. Choosing predictors for the regression models proved difficult. The choice of predictors should reflect inferenc es into what factors are influencing the dependant variable. However, because of limited data I must use what is available. Ideally, I would like to test how property values influence the average annual employee growth rate in different zip codes. I hyp othesize that cheap land significantly influences business location decisions, and that the price of land will be more of an influence than enterprise zone incentives. However, I could not find data on property values that matched up with the time frame o f my study. Also, ideally I would test the amenities offered by quality of life improvements. Dabney and Bartik have determined that quality of life factors influence business location decisions. If data were available, I would test the amount of money invested in each zip code for quality of life improvements as a dependant variable in my regression. I chose to test black population data as a predictor because Jacksonville has a very large black population; about 30% of sted black population and Hispanic population to see if a certain race was receiving more benefit than another, or if race had a negative effect on job growth. I tested number of households and housing units as predictors to see if housing stock influence d job growth. It is my contention that businesses will locate where they have access to the most customers; especially in the retail trade, one of center, new bu sinesses will locate close to the new customer base.
47 Table 4 1. Descriptive Statistics For First Regression Variable Mean Std. Deviation N Job Growth Rate .2984 .15934 50 Enterprise Zone .1200 .32826 50 Establishment Growth 51.6800 75.17610 50 Annu al Payroll Growth 66662.6400 130796.13679 50 Black Population 4788.2000 7103.22937 50 Hispanic Population 824.3000 797.18905 50 HousingUnits 9175.9000 5926.16400 50 Households 8396.0800 5493.93522 50 Table 4 2 Descriptive Statistics For Second Regr ession Variable Mean Std. Deviation N Job Growth Rate .2984 .15934 50 Enterprise Zone .1200 .32826 50 Urban Zone .4800 .50467 50 Table 4 3 Dummy Variable Schema Zone Code Enterprise Zone 1 0 0 Urban Zone 0 1 0 Suburban Zone 0 0 0
48 Figure 4 1 Establishment growth 2004 2008 by zip code for Jacksonville Florida Enterprise Zone
49 Figure 4 2 Test zones for regression Suburban Zone Urban Zone Enterprise Zone
50 CHAPTER 5 R ESULTS Findings Figure 4 You can see from the map that the downtown area, including the enterprise zone, is experiencing negative growth while the outer regions are experiencing posit ive growth. Clearly, there is some factor that is influencing business location decisions to a greater degree than enterprise zone incentives. The map suggests that the city is supporting growth in the form of sprawl while simultaneously working to revit alize the downtown area. This strategy seems counterintuitive. Why would you support sprawl by building roads and other services when you are simultaneously trying to incentivize businesses to locate downtown? Being from Jacksonville, I have observed ne w developments, new business parks, and new strip malls being constructed in large numbers in the green regions of the map. All of this green field development is undermining the enterprise zone program that Jacksonville is spending millions of dollars on While the government continues to subsidize sprawl, the majority of new businesses will chose green filed development instead of infill development. Table A 1 shows the figures for establishment growth, employee growth, and annual payroll growth for a ll zip codes in the Jacksonville Metropolitan Statistical Area between 2004 and 2008. These figures were used to calculate annual growth rates. The average annual growth rate for employees will help us determine which areas are experiencing positive grow th. The average annual growth rate of employees in enterprise zones is 0.2%, while the average annual growth rate of employees in urban and suburban zones were 3.3% and 3.9% respectively. This suggests that the
51 enterprise zone is losing jobs at 0.2% a y ear, while non zone areas are gaining jobs at about 3% a year. The reason for this negative growth in the enterprise zone is still unclear. The results of the first regression will help determine what factors may be affecting job growth. The first regres sion, using IBM SPSS Statistics software, tested enterprise zone designation, establishment growth, annual payroll growth, black population, Hispanic population, number of household units, and number of households as predictors of employment growth. The M odel Summary (Table 5 1) shows the results of the Correlation value that helps determine if there is a relationship between the dependant variable and the predictors. The R Squ are value is 0.181. The R Square value helps determine how well the model explains the variation in (Y) given (X). In this case, the model explains 18% of the total variation in the data. The Adjusted R Square value is 0.04. The Adjusted R Square value tells us how well the model would explain the total variation if the model was run on the population rather than the samples. It also explains the diminished predictive capability based on the sample size and number of predictors. The Significant F Chan ge is 0.264 and is crucial to the models predictive capabilities. Using a 95% confidence interval, it is crucial to get a significance that falls into the rejection region in order to reject the null hypothesis. To do this, the significance must be equal to or less than 0.05. Because the significance is 0.264, we fail to reject the null hypothesis and conclude that there is no difference in the variation of employment growth given the predictors.
52 The second regression, again using IBM SPSS Statistic so ftware, yielded similar results. The second regression tested the difference in employment growth given the zip codes location, whether it be in the Enterprise Zone, Urban Zone, or Suburban Zone. The Model Summary (Table 5 2) shows the results of the reg ression. As in the first regression, the significance does not fall in the rejection region. Because the significance is greater than 0.05, we reject the null hypothesis and conclude that there is no difference in the variation of employment growth given the predictors. The regression models reveal that annual establishment growth rate, annual payroll growth rate, black population, Hispanic population, number of housing units, and number of households are poor predictors of employee growth. This allows us to eliminate these factors from the list of possible contributors to the variation in zone performance. The average annual growth rate for employees revealed that the enterprise zone was losing jobs while the urban and suburban zones were gaining jobs Coupled with two regression models that show no significance, there is enough actually losing them. It is impossible to go back in time and determine what would have hap pened had the zone not been there, but the evidence suggests that enterprise zone incentives are not enough to encourage businesses to open or expand in the zone rather than other areas. Furthermore, my findings coincide quite well with the literature. M ost empirical studies have determined that enterprise zones produce little to no benefit.
53 Discussion The enterprise zone concept has been around for decades Originally conceived in Britain, enterprise zones have evolved to encompass American objectiv es. States were the first to implement e nterprise zones in the early 80 s. The federal government got involved and reorga nized the program in the mid 90 s. Since that time, the uble discerning the effectiveness of the program. A number of empirical studies have been conducted on the effectiveness of enterprise zones to create positive changes in socio economic conditions. The earliest studies showed that enterprise zones creat e jobs, but because the methodology used in these studies were survey method, the studies have serious limitations. The most recent studies find either a temporary gain in jobs and income or no discernable impact at all. These later studies use a more so phisticated methodology so the results are most likely more accurate than the earlier studies. The majority of literature on the effects of enterprise zones suggest that zones do very little to improve socio economic conditions. Yet, governments continu e to increase spending on the program. The vast amount of money spent on the enterprise zone programs could be used for other services, so it is imperative that policy makers are able to measure the impacts of the programs they prescribe. The fact that e nterprise zones are so difficult to evaluate is a serious limitation. In an era of renewed government transparency and weary disposition to government spending, proving that programs are worth the money is more important than ever. Unless clear proof of the significant benefits of enterprise zones is obtained, the enterprise zone programs around the country is likely to be criticized in years to come.
54 Jacksonville has experienced large decentralization, much like other cities across the nation. Jacksonvi decentralization was the implementation of an enterprise zone. My study focused on Using the same type of methodology as the latest studies on enterprise zones, I determined that the enterprise zone is losing jobs at a rate of 0.2% a year while other parts of the metro area are growing jobs about 3% a year. The regressions could not attribute the variation in job growth to any of the predictors. Therefore, it is safe to say that black population, Hispanic population, number of households, and number of housing units do not significantly affect job growth. The regressions also revealed that a zip codes location i.e. enterprise zone, urban zone, or suburban zone; had no enterprise zone is not creating jobs at a significantly higher rate than the surrounding area. It is true that there are limitatio ns to my study. The amount and type of data available proved to be the deciding factor in choosing what area characteristics to control for. Furthermore, other factors that have not been discovered may be affecting the zones performance. However, my stu dy coupled with the evidence discovered in the literature review, suggests that my results closely reflect actual changes in socio economic conditions. Why are governments continuing to promote enterprise zones when evidence suggests that they do not work? We can speculate that the program has survived this long because it is politically popular. From the perspective of a politician, enterprise
55 zones sound wonderful. Enterprise zones are promoted as a way to help poor neighborhoods revitalize through the implementation of tax incentives. A politician who promotes helping the poor while implementing tax breaks can likely expect a lot of remember that the politicians make t he final decisions when it comes to government programs. So, as long as enterprise zones are politically popular and are promoted as viable economic development tools, enterprise zone programs will continue. Recommendations Figure 4 1 reveals the competin g strategies implemented by Jacksonville. The government cannot expect to see rapid revitalization of downtown while simultaneously subsidizing sprawl. Therefore, it is my recommendation that Jacksonville discontinue the provision of new infrastructure i n green field areas. Instead, Jacksonville should implement some kind of service boundary for new development. If the city only provides infrastructure for new development in areas that the city deems ripe for redevelopment, then more businesses would lo cate in those areas. But, if the city continues to provide infrastructure for sprawl development, businesses will continue to locate away from downtown. A service boundary would signal the intention to invest money downtown, rather than investing in spra wl. Businesses would soon realize that locating downtown is a better long term strategy because they will receive benefits from the enterprise zone and benefit from the heavy public investment in the area. If you are a business, you would chose to locate in an area that will receive much public investment over an area that will see very little public investment. Coupling the enterprise zone strategy with an urban service boundary would do much to revitalize
56 downtown. In fact, a better strategy would be to implement an urban service boundary while discontinuing the Enterprise Zone. Implementing an enterprise zone is not the only revitalization strategy available to governments. Community Redevelopment Agencies, Business Improvement Districts, and Main St reet programs have all had some success in revitalizing blighted areas. In in a holistic, comprehensive way that allows each strategy to benefit from the other. We have seen that Jacksonville has undermined its revitalization strategy by subsidizing sprawl at the same time. In order to successfully revitalize, Jacksonville must create policies that will strengthen its efforts; not compete with them.
57 Table 5 1 Mod el summary for first regression Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 .425 .181 .044 .15579 .181 1.323 7 42 .264 Table 5 2 Model summary for second regre ssion Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 .182 .033 .008 .15999 .033 .801 2 47 .455
58 APPENDIX DATA USED IN METHODOLOGY Table A 1 Establishment Growth, Employee Growth, and Annual Payroll Growth from 2004 to 2008. Data collected from US Census Bureau County Business Pattern by Zip Code 2004 2008. Zip Code Establishment Growth 2004 2008 Employee Growth 2004 2008 Annual Payroll Growth 2004 2008 ($) 3220 2 EZ 56 3846 116620 32204 EZ 37 10033 765508 32205 UZ 9 1194 15431 32206 EZ 32 200 40072 32207 UZ 34 1690 264724 32208 EZ 3 1447 3264 32209 EZ 17 495 107987 32210 UZ 33 1165 9727 32211 UZ 93 2757 28285 32212 UZ 2 87 1736 32215 UZ 2 132 4900 32216 UZ 70 6 111111 32217 UZ 2 570 5202 32218 UZ 193 3623 80600 32219 UZ 26 579 52956 32220 UZ 38 299 10294 32221 UZ 38 1978 60923 32222 UZ 47 1197 18166 32223 UZ 142 1621 78877 32224 UZ 181 3205 238061 32225 UZ 112 1766 9311 0 32226 UZ 61 1121 45872 32234 UZ 8 256 19850 32244 UZ 40 186 12518 32246 UZ 264 4830 189355 32250 SZ 19 1113 29721 32254 EZ 44 792 66652 32256 UZ 275 2628 190175 32257 UZ 59 1701 109733 32258 UZ 153 8279 401625 32259 SZ 144 1773 40250 322 66 SZ 19 255 3507 32277 UZ 3 73 14543 32011 SZ 46 54 7059 32033 SZ 14 268 12915 32034 SZ 97 896 53910 32043 SZ 48 676 51851
59 Table A 2 Demographic data from shapefile created by Shimberg Center for Housing Studies at the University of Florida 2004 Zip Code Black Population Hispanic Population Household Units Households 32202 EZ 3310 112 1816 1408 32204 EZ 3922 176 3916 3287 32205 UZ 6159 886 14540 13323 32206 EZ 17439 391 10467 8280 32207 UZ 7069 1586 16283 15144 32208 EZ 25397 322 13738 12574 32209 EZ 38857 281 17526 15234 32210 UZ 14575 2531 24665 22736 32211 UZ 10075 1809 15003 13847 32212 UZ 689 304 348 346 32215 UZ 287 73 248 200 32216 UZ 5775 1623 12434 11774 32217 UZ 2780 1238 8956 8498 32218 UZ 14983 623 14801 13772 32219 UZ 4022 98 3950 3498 Table A 1 Continued Zip Code Establishment Growth 2004 2008 Employee Growth 2004 2008 Annual Payroll Growth 2004 200 8 ($) 32046 SZ 10 17 3414 32065 SZ 78 1369 30961 32068 SZ 93 862 22947 32073 SZ 1 1919 34102 32082 SZ 78 216 65938 32086 SZ 64 546 36546 32091 SZ 33 465 25336 32092 SZ 173 1848 91233 32095 SZ 12 169 8421 32097 SZ 88 574 12446 32145 SZ 1 72 139 32233 SZ 36 372 18612 32656 SZ 4 202 9083 Source: US Census Bureau, County Business Patterns 2004 2008 EZ = Enterprise Zone; UZ = Urban Zone; SZ = Suburban Zone
60 Table A 2 Continued Zip Code Black Population Hispanic Population Household Units Households 32220 UZ 438 195 4241 3876 32221 UZ 1803 510 6734 6347 32222 UZ 484 182 1732 1596 32223 UZ 1075 888 9873 9610 32224 UZ 2618 1847 13450 12608 32225 UZ 7152 2538 17236 16591 32226 UZ 130 149 3226 3013 32234 UZ 546 84 2458 2257 32244 UZ 10405 2533 18569 17118 32246 UZ 5762 2323 14343 13339 32250 SZ 1057 701 12053 10925 32254 EZ 7998 245 6094 53 43 32256 UZ 3512 1835 15247 14034 32257 UZ 3143 1865 15651 14856 32258 UZ 986 477 4428 4316 32259 SZ 369 432 6286 6000 32266 SZ 53 152 3462 3272 32277 UZ 8636 1242 11384 10827 32011 SZ 603 89 4385 4121 32033 SZ 332 46 963 854 32034 SZ 2351 533 133 36 10230 32043 SZ 1884 691 7643 7057 32046 SZ 557 64 3074 2851 32065 SZ 1588 1158 7345 7050 32068 SZ 1334 1276 13351 12590 32073 SZ 4540 2705 18950 18111 32082 SZ 286 702 13312 11866 32086 SZ 584 731 11167 9363 32091 SZ 3349 279 6261 5439 32092 S Z 249 93 2739 2393 32095 SZ 3841 586 9507 8480 32097 SZ 925 163 4096 3798 32145 SZ 1093 124 1515 1361 32233 SZ 4279 1516 10691 9846 32656 SZ 109 208 5302 4545 EZ = Enterprise Zone; UZ = Urban Zone; SZ = Suburban Zone
61 LIST OF REFERENCES ( 2000). Use of Enterprise Zone Incentives Has Increased, but Challenges Continue. Government Accountability. ( 2003). Enterprise Zone Pilot Project Incentives Not Widely Used, But Progress Mad e in Revitalization Project Area. Program Policy Analysis and Government Accountability. ( 2004). Florida's Enterprise Zone Program Is Similar to Those of Other States. The gram Policy Analysis and Government Accountability, Tallahassee, FL. (2005). Enterprise Zones: A review of the Econooic Theory and Empirical Evidence. Minnesota House of Representatives Research Department. ( 2010). Florida Enterprise Zone Program Annual Re port. Tallahassee: Executive Office Of The Governor. Bartik, T. J. (1985). Business Location Decisions in the United States: Estimates of the Effects of Unionization, Taxes, and Other Characteristics of States. Journal of Business and Economic Statistics 3 (1), 14 22. Beck, F. D. (2001). Do State designed Enterprise Zones Promote Economic Growth? Sociological Inquiry 71 (4), 508 532. Bendick, M. J., & Rassmussen, D. (1986). Enterprise Zones and Inner City Economic Revitalization. Youngstown State Univer sity: Urban Institute Press. Birch, D. (1979). The Job Generation Process. Cambridge: Program on Neighborhood and Regional Change. Boarnet, M. G. (2001). Enterprise Zones and Job Creation: Linking Evaluation and Practice. Economic Development Quarterly 1 5 pp. 242 254. Boarnet, M. G., & Bogart, W. T. (1996). Enterprise Zones and Employment: Evidence from New Jersey. Journal of Urban Economics 40 (2), pp. 198 215. Bondonio, D., & Engberg, J. (2000). Enterprise Zones and Local Employment: Evidence from th Regional Science and Urban Economics 30 (5), 519 549. Dabney, D. Y. (1991). Do Enterprise Zone Incentives Affect Business Location Decisions? Economic Development Quarterly 5 (4), pp. 325 334.
62 Elvery, A. J. (2008). The Impact of Ent erprise Zones on Resident Employment: An Evaluation of the Enterprise Zone Programs of California and Florida. Economic Development Quarterly pp. 44 59. Fisher, P. S., & Peters, A. H. (2000). The Cost of Enterprise Zone Incentives and an Alternative Use of the Money. National Tax Association: Proceedings of the 93rd Annual Conference (pp. 67 73). Green, R. E. (1991). Enterprise Zones: New Directions in Economic Development. Newbury Park: Sage. Greenbaum, & Engberg. Impact of State Enterprise Zones. Gree nbaum, & Landers. (2009). Why Are State Policy Makers Still Proponents of Enterprise Zones? Whay Explains Their Action in the Face of a Preponderance of the Research? International Regional Science Review 32 466. Jacobs, J. (1961). The Death and Life of Great American Cities. New York: Random. Miller, J. (1998). An Environmental History of Northeast Florida. University of Florida Press. Moore, W. S. (2003). Enterprise Zones, Firm Attraction and Retention: A Study of the California Enterprise Zone Program Public Finance and Management 3 (3), 376 392. O'Keefe, S. (2004). Job Creation in California's Enterprise Zones: A Comparison Using a Propensity Score Matching Mode. Journal of Urban Economics 55 (1), 131 150. Papke, L. E. (1991). Tax Policy and Urba n Development: Evidence from an Enterprise Zone Program. National Bureau of Economic Research. Peters, A., & Fisher, P. (2004). The Failures of Economic Development Incentives. Journal of the American Planning Association 70 (1), 27 37. Rubin, B. M., & W ilder, M. G. (1989, Autumn). Urban Enterprise Zones Employment Impacts and Fiscal Incentives. Journal of the American Planning Association 55 (4), pp. 418 431. Runin, M. M., & Armstrong, R. B. (1989). The new Jersey Urban Enterprise Zone Program: An Eval uation. New Jersey Department of Commerce, Energy, and Economic Development, Trenton. Sridhar, K. S. (1999). The Incentive Programs and Unempolyment Rate. Review of Regional Studies 30 (3), 275 298.
63 United Nations Development Programme. (n.d.). Millenniu m Development Goals Retrieved February 16, 2011, from United Nations Development Programme: http://www.undp.org/mdg/basics.shtml US Census Bureau 2005 2009 American Community Survey (n.d.). Retrieved February 27, 2011, from US Census Bureau American Fac t Finder: http://factfinder.census.gov/servlet/ACSSAFFFacts?_event=ChangeGeoContext &geo_id=16000US1235000&_geoContext=&_street=&_county=Jacksonville&_cit yTown=Jacksonville&_state=04000US12&_zip=&_lang=en&_sse=on&ActiveGeo Div=&_useEV=&pctxt=fph&pgsl=010&_su bmenuId=f Wilder, M., & Rubin, B. (1996). Rhetoric versus reality: A review of studies on state enterprise zone programs. Journal of the American Planning Association 61 pp. 473 491.
64 BIOGRAPHICAL SKETCH Curtis Ryan Dubbe rly was born in 1984 in Jackso nville, Florida. The younger of two boys, he grew up in Orange Park, Florida, a nearby suburb of Jacksonville. Curtis, an avid surfer, guitar player and athlete, graduated from Ridgeview High School in 2003. He attended Florida State University for two years before deciding to transfer to the University of Florida to pursue a BA in a rchitecture. Upon graduating in 2009, he entered into graduate school at UF and began working on his MA in urban and regional p lanning. Upon completion of his MAURP, Curt is would like to pursue a career in Florida, where he sees much potential for responsible and functional growth.