|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 TH E EFFECT OF EMINENT DOMAIN ON PROPERTY VALUES By NATHAN VAN STEENBERGEN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIE NCE UNIVERSITY OF FLORIDA 2009
2 2009 Nathan Van Steenbergen
3 To Brian Wade your memory will live on forever
4 ACKNOWLEDGMENTS The author wishes to acknowledge several important individuals who made the completion of this thesis possible. The support and guidance from family, friends and professors helped to make this experience a very memorable one All of their sacrifice and dedication led to a finished thesis in a timely manner. Their efforts are greatly appreciated, and the following persons deserve special recognition. Dr. Rodney Clouser, supervisory committee chair, cont ributed opinions and constructive criticism to the writing of all drafts of the thesis. His commitment a nd time were valuable resource s throughout the writing process. His attention to schedules and detail kept the thesis precise and timely. His patience, fairness, and opinions are highly respected and appreciated. Dr. Charles Moss, supervisory committee cochair, assisted with the econometric analysis in this thesis. His understanding of indexes, index construction, and index manipulation led to the final results of this thesis. His diligence in explaining mathematical and statistical concepts was important to the understanding of the results With his guidance the quality of this thesis was improved. Dr. David Ling, committee member and minor chair, contri buted to the methods used in this thesis. His knowledge of real estate issues helped build a foundation for the thesis. His guidance in matters of real estate was invaluable. His assistance with the index method enhanced the overall quality of the thesis. Dawn Jourdan, committee member, assisted in all legal matters of this thesis. Her knowledge of the law and urban and regional planning helped develop a strong legal base for the thesis. Her opinions of the legal issues of eminent domain allowed for a well rounded thesis. Her help in the legal section of this thesis was a necessity.
5 Mr. and Mrs. Van Steenbergen, father and mother of the author, are recognized for giving their support and guidance throughout the writing process. Their examples of hard work and compassion served to inspire this achievement. Without their unconditional love none of this would be possible. Laurie, Jacob, Seth and Sarah, brothers and sisters of the author, without their confidence, and encouragement this thesis would never have be en completed. They served as an inspiration to the author during the writing process. Last and most importantly is Rachel. Without her confidence, compassion, patience, and love the thesis would never have been completed. Moreover, her understanding and en couragement throughout the entire process was invaluable. Without her help, this thesis would not have been possible.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 ABSTRACT ...................................................................................................................................10 CHAPTER 1 INTRODUCTION ..................................................................................................................12 Introduction .............................................................................................................................12 Problem Statement ..................................................................................................................14 Specific Problem .....................................................................................................................16 Obj ectives ...............................................................................................................................18 Description of Study Area ......................................................................................................18 2 EMINENT DOMAIN LAW AND AREA DESCRIPTION ..................................................21 Federal Takings Law ..............................................................................................................21 From Public Use to Public Purpose .................................................................................22 Berman v. Parker ......................................................................................................22 Hawaii Housing Authority v. Midkiff ......................................................................23 Kelo v. New London ................................................................................................24 Project Influence ..............................................................................................................25 Scope of the Project Rule .........................................................................................25 United States v. Miller .............................................................................................25 New Jersey Eminent Domain Law .........................................................................................26 Eminent Domain Act of 1971 ..........................................................................................27 Township of West Windsor v. Yevette Nieremberg .......................................................29 Asbury Park Redevelopment Area .........................................................................................30 Asbury Park Economic Conditions ........................................................................................33 Asbury Park Housing Characteristics .....................................................................................38 3 ECONOMIC THEORY ..........................................................................................................41 Land Market Theory ...............................................................................................................41 The Interaction of Supply and Demand ...........................................................................43 Unique Characteristics of the Land Market .....................................................................44 The Concept of Land Rent ......................................................................................................45 Ricardian Rent Theory ....................................................................................................47 Von Thunens Location Theory ......................................................................................50 Relation of Land Rent to Land Values (Income Capitalization Theory) ................................52
7 Appraisal Theory ....................................................................................................................52 4 LITERATURE REVIEW .......................................................................................................55 Price Index Introduction .........................................................................................................55 Repeat Sales Method for House Price Index Construction ....................................................59 Abnormal Local Market Appreciation ....................................................................................67 5 METHODS .............................................................................................................................78 6 RESULTS AND CONCLUSION ...........................................................................................87 Results .....................................................................................................................................88 Conclusion ..............................................................................................................................93 Potential Problems ..................................................................................................................94 Further Research .....................................................................................................................97 REFERENCES ..............................................................................................................................98 BIOGRAPHICAL SKETCH .......................................................................................................103
8 LIST OF TABLES Table page 21 Asbury Park and Monmouth County Decennial Population Trends .................................34 22 Asbury P ark and Monmouth County Age Profile ..............................................................35 23 Asbury Park and Monmouth County Household Income ..................................................37 24 Asbury Park Housing Data ................................................................................................39 25 Asbury Park Housing Values .............................................................................................40 61 Regression Results of Market Area ...................................................................................88 62 Regression Results of Submarket Area (WRA) .................................................................90 63 Yearly and Cumulative Abnormal Appreciation ...............................................................92
9 LIST OF FIGURES Figure page 21 Asbury Park Waterfront Redevelopment Area Project Map .............................................32 22 Areas Subject to Condemnation.........................................................................................33 31 Supply and Demand of Residential Land ..........................................................................44 32 Use of Value Products and Cost Curve Diagrams to Illustrate the Concept of Land Rent ....................................................................................................................................47 33 Presentation of Ricardos explanation of land rent. ...........................................................49 34 Effect of Transportation Costs on Land Rent at Various Distances from the Central Market ................................................................................................................................51 41 Comparison of real estate price indexes for a small area in St. Louis estimated by the chain and regression methods.. ..........................................................................................63 42 Sj t is the ratio of distance A to distance a, i.e., ( A / a). Sjt+k is ( B / b), etc. Since ( B / b) is less than ( A / a), Sjt+k< Sjt. This indicates a negative excess appreciation during that time interval.. .....................................................................................................................70 51 A map of the eleven towns in Monmouth County used for the construction of the market wide index. Asbury Park is highlighted in yellow, and the surrounding towns are out lined in red. Source: Monmouth County Planning Board, Monmouth, County, 1997. .....................................................................................................................84 52 An example of a repeat sales matrix used to create an index ............................................86 61 A Comparison of the Market wide index and Waterfront Redevelopment Area index. ...91 62 Median Sale Price of Repeat and Non Repeat Sales per Year ..........................................96
10 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 THE EFFECT OF EMINENT DOMAIN ON PROPERTY PRICES By Nathan Van steenbergen December 2009 Chair: Rodney Clouser Cochair: Charles Moss Major: F ood and Resource Economics Eminent Domain is the governments right to take private property for public use upon payment of just compensation. The threat of eminent domain can affect real estate by artificially inflating or deflating property values. This effect is known as project influence. Any increase or decrease in property value caused by the governments use of eminent domain is not allowed when determining a propertys condemnation value. When faced with the problem of project influence, an apprais er must determine the amount of appreciation or depreciation caused by a public project. This economic thesis applies a methodology which can be used by appraisers to quantify project influence. Reviews of economic theory and previous studies assisted in t he formulation of the methods used in this study. Sales data for 11 towns in Monmouth County, New Jersey were used to create a market wide hous e price index. This index was used to capture the movement of property prices in Monmouth County. Sales data was also used to create a house price index for a redevelopment area of Asbury Park, New Jersey affected by the use of eminent domain. By comparing these two indexes before and after the City of Asbury Park committed itself to the redevelopment project this study was able to quantify the amount of abnormal appreciation caused by the public project.
11 The results show that before the City of Asbury Park committed itself to the redevelopment project the redevelopment area appreciated 1.6% more than the market per y ear. However, after the government committed itself to the public project the redevelopment area appreciated 5% more than the market per year. This increase in appreciation in the redevelopment area may have been caused by the public project The method that was applied in this study proved to be a way to quantify abnormal appreciation caused by the use of eminent domain for a public project.
12 CHAPTER 1 INTRODUCTION Introduction Eminent domain is the right of the sovereign government to take private prope rty for public use upon payment of just compensation  T he Takings Claus e in the Fifth Amendment to the Constitution states, nor shall private property be taken for public use, without just compensation.  This provision is also applicable to the various states through the Fourteenth Amendment to the Constitution. This Takings Clause allows federal, state, and local governments to acquire private property as long as two conditions are met: 1) the private property must be taken for a public use, and 2) the property owner must be compensated justly for his or her loss. Just compensa tion is defined as the amount of loss for which a property owner is compensated when his or her property is taken; [it] should put the owner in as good a position pecuniar ily as he or she would be if the property had not been taken; gene rally held to be market value[3 ]. This market value is the most probable price which a property should bring in a competitive and open market under all conditions requisite to a fair sale, the buyer and seller each acting prudently and knowledgeably, and assuming the price is not affected by undue stimulus. [1 ] Compensation is a monetary value ultimately determined by the courts. It is the value of the property which would have been decide d in a free market environment given an arms length transaction between a willing seller an d a willing buyer. Interpretations of the phrase public use have been divided between a narrow view and a broad view. Originally, condemnation was used by public and quasi public entities such as public utility companies, development commissions, and railroad companies. Its use was always based on the narrow view that public use meant the property taken would be publically owned,
13 controlled, and accessed . H owever, Supreme Court decisions in ca ses such as Berman v. Parker [10 ] and Hawaii Housing Authority v. Midkiff  have broadened this interpretation of public use to justify takings for public purposes. A taking for a public purpose does not have to provide a public service or utility. Instead, eminent domain is warranted if the land taken is eventually used to provide benefits to society [ 39]. This expansion of the guideline for public use led many municipalities to use the power of eminent domain in urban renewal projects to combat slums and blight. An example of this is the use of eminent domain in the Inner Harbor and the Fells Point sections of downtown Baltimore in the 1970s [ 40] On June 23, 2005 the United States Supreme Court embraced a broader view of public use in the eminent domain case, Kelo v. The City of New London . This case expanded the interpretation of public use when it ruled that a citys decision to take property for the purpose of economic development satisfies the public us e requir ement of the Fifth Amendment [36 ]. The court observed that the governmental taking was meant to revitalize the local economy by creating temporary and permanent jobs, generating a significant increase in tax revenue, encouraging spinoff activities and maximizing public access to the waterfront [36 ]. The results of this court decision set a legal precedent; this was the first U.S. Supreme Court decision to allow the taking of nonblighted property by eminent domain solely for economic development purposes [ 20]. The Supreme Courts decision did not force states to obey their ruling. In writing the opinion, the majority wrote, this decision does not preclude any state from placing further restrictions on its exercise of the takings power [ 36]. In fact, many states already imposed public use requirements at the time of the decision. After this decision, many state legislatures either proposed, enacted or improved laws aimed at curbing the use of eminent domain for economic development. In a repor t by the Castle
14 Coalition forty two states have passed new laws aimed at curbing the use of eminent domain for private use [ 14]. Some states enacted stricter laws than others. In Florida and the Dakotas eminent domain is not allowed for private use, or pri vate purposes . However, in Idaho, Illinois, Kentucky, Maine, Nebraska, Alaska, Connecticut, Maryland, Missouri, Montana, Ohio, California, Delaware, Tennessee and Vermont eminent domain can still be used for blight removal and private use. . However, eight states failed to pass legislative reforms after the Kelo decision. The states that failed to pass legislative reforms were Arkansas, Hawaii, Massachusetts, Mississippi, New Jersey, New York, Oklahoma, and Rhode Island . These eight states are the most likely to use eminent domain for the purposes of economic development and blight removal. Problem Statement Eminent domain proceedings are cumbersome and time consuming [ 21]. It is typical that proceedings for the acquisition of property for publ ic use are begun with months, or even years of time spent in preliminary discussion and tentative plans [ 44]. During this time these discussions and plans may become public knowledge to owners and other persons in the area of the proposed improvement. The knowledge of a proposed public project involving condemnation can have an influence on the market value of the properties affected. This is known as project influence. Project influence can have a positive or negative effect on prope rty values in a condemned area. Condemnation blight is a decrease in the market value of a property due to pending condemnation action [ 22]. This is possible because there is a time lag between the governments commitment to a public project and the condemnation proceedings to t ake pr operty needed for that project. Lewis Orgul described the cause of condemnation blight when he wrote, it usually happens that the very institution of the condemnation proceeding puts the property under blight.
15 [ 53] In a forced sale situation the ow ners of improved property have no incentive to make alterations or improvements, nor do they have any incentive for large expenditures to maintain the property. If the owner has an income producing property, tenants have no incentive to stay in the propert y, leading to high vacancy rates. Either of these scenarios can substantially reduce the value of a property under the threat of eminent domain. Project enhancement is defined as an increase in a propertys market value in anticipation of a public project requiring condemnation action [ 22]. This can be caused by market driven speculation, or a change in the zoning of the property. Lewis Orgul explained, In some instances, owners have resorted to wash sales and to fictitious transactions designed to crea te a false impression of market activity. [ 53] This is an example of project influence driven by speculation. However, the anticipation of a public project may cause a zoning change. If this affects a propertys value it also may b e considered project inf luence. By law, an appraiser cannot include project enhancement and condemnation blight in valuing the subject property. Most courts agree that the United States cannot be charged in condemnation proceedings for values which it has created in constructing the project for which the property is taken; nor can the owner be charged for any diminution in value attributable to the project [ 4]. According to the Uniform Relocation Assistance and Real Property Acquisition Policy Act of 1970 the appraiser shall di sregard any increase or decrease in the fair market value of real property, before the valuation date, caused by the project for which the property is to be acquired. [ 64] An appraiser must comply with the scope of the project rule when confronted by enha ncement and blight. The scope of the project rule holds that the effects of a proposed project cannot be considered in valuing a property to be acquired for the project when it was
16 clear that the parcel under appraisal would be acquired for the project.[ 22 ] The two most important aspects of applying the scope of the project rule are determining when the government was committed to the project and whether it was probable that the appraised parcel wo uld be taken for the project . This is an important da te because it is at this point in time that the market reacts to the public project and property values are affected. If the property was within the scope of the project, then the date at which the government committed to the project is the date of valuati on. For compensation to be considered just, any project influence, whether positive or negative, accumulating from the date the government committed to the project to the condemnation valuation date must be separated fr om the propertys market value. Speci fic Problem If the condemner and seller cannot come to an agreement regarding property value, then the property is condemned and fair market value is determined in court. Valuation testimony from expert appraisal witnesses is heard by a Trier of Fact who d ecides the amount of just compensation awarded. Depending on the states condemnation proceedings this arbiter is either a court with a jury, a judge, or a tribunal of commissioners [ 58]. Each of these three arbiters is subject to their own criticisms. There is widespread belief that a jury is unsatisfactory in determining matters of a technical nature, such as condemnation [ 58]. This is because a typical juror does not have the background, experience, or education to understand highly scientific valuati on methods. In many respects judges and commissioners reveal the same weaknesses. All of these arbiters lack the experience and know how necessary to decide just compensation in the presence of project influence. The just compensation paid is considered to be the propertys fair market value. The compensation payment is designed to leave the property owner as well off as they would have been had there never been a taking. The courts decide this market value because the buyer and
17 seller cannot find a monetar y value which they both consider to be fair. Therefore, the courts compensation in condemnation cases is designed to make the buyer and seller equally content. Project influence forces the buyer or the seller to share in the losses or gains to the propert y unequally. It is for this reason that the amount of project influence is not allowed in condemnation valuation. Appraisers in condemnation proceedings must ultimately quantify this project influence in order to determine market value. Orgul describes the power of an appraisers opinion of value in condemnation proceedings Opinions of qualified witnesses as to value constitute the form of evidence most often introduced in condemnation proceedings. [ 53] However, there is a maze of conflicting testimony ca used by the expert appraisers, who can, and usually do, vary their final values depending which party they are representing [ 58]. Even though there is a discord between opposing appraisals, expert testimony still counts for more in the final valuation of just compensation than almost any other type of evidence. [ 53]. Courts often accept an appraisers opinion of value because they lack the experience in real estate valuation. This can lead to high variations in the opinion of a condemned value by opposing appraisers. When appraisers value properties they use a time adjustment factor to compare their subject property to similar properties in the area selling in different time periods. It may be simple for an appraiser to document the change in market prices between two dates in time. However, appraisers do not have a methodology that can separate out the different influences, such as a public project, that affect a propertys value for a given time period. An appraiser confronted with project influence may n eed to make a time adjustment and a project influence adjustment to compare two or more properties that have recentl y sold in the same market.
18 Objectives The general objective of this study is to apply a methodology which could be used to quantify project influence. This methodology should be able to separate abnormal appreciation caused by the announcement of a public project from normal market driven appreciation over a specific time period. In doing so, this procedure could be used by appraisers to adjus t the value of properties affected by project influence. This will be achieved through the following specific objectives. Determine the appreciation rate for a market wide area both before and after a government announcement of a public project. Determine the appreciation rate for a sub market area directly affected by the government announcement of a public project. Compare the trend in appreciation rates for the two areas before the government was committed to the project and afterwards. Determine the dif ference between the appreciation rates for the market wide and submarket areas before and after the government announcement of a public project. Create a time adjustment factor that could be used to adjust for project influence over a given time period. D escription of Study Area This study deals with the redevelopment area in Asbury Park, New Jersey. It is necessary to understand this areas past in order to understand the problem this study is addressing. Asbury Park is a city in Monmouth County, New Jers ey located on the Central New Jersey Sea Shore. In the early to mid 1900s Asbury Park was known as a summer seaside tourist destination [ 7 ]. However, Asbury Park began to decline after the introduction of the Garden State Parkway in the 1950s, which gave t ourists easier access to less crowded beaches farther south . During the late 1960s and 1970s the city experienced economic and social deterioration. In the 1970s the city experienced race riots. The riots caused the more affluent families to relocate out of the city
19 to the emerging suburbs, where they shopped at newly developed suburban malls instead of merchants of the Asbury Park Central Business District [ 43]. By the early 1980s, due to the increase of tax exempt properties as well as the demolition of buildings in the area, the city witnessed a lowering of the economic tax ba se. [ 6]. Due to these conditions, on August 1, 1984, the Mayor and City Council found the waterfront area between Grand Avenue and the beach to be blighted or in need of redevel opment pursuant to the New Jersey State Redevelopment L aw [ 6]. This blight designation allowed the city to use the power of eminent domain to redevelop the area. The ori ginal Waterfront Redevelopment P lan was adopted on November 7, 1984 [ 6]. In 1986, the c ity selected Carabetta/Vaccaro Developers to be the prime developer of the area within the redevelopment zone where construction was planned. However, for various reasons, by 1992 Carabetta had bought out Vaccaro and was bankrupt . At that point all de velopment in the Waterfront Redevelopment area stopped. In the summer of 2001, M.D. Sass Municipal Finance Partners got city approval to buy Carabetta's redevelopment rights and waterfront properties for $7.4 million [ 61]. M.D. Sass also paid $6.5 million to the city for the tax liens the city placed on Carabettas beachfront properties. In August of 2001, M.D. Sass, in a joint venture with Ocean Front Acquisitions of Lakewood, signed a memorandum of understanding (MOU) with the city, detailing much of a fo rmal agreement to redevelop the beachfront that would be negotiated later [ 61]. The joint venture named the new development company Asbury Partners, LLC. The signing of the MOU committed both the City and Asbury Partners to a public private redevelopment p roject in the same area as the previous Waterfront Redevelopment Plan. The City would, at the request of the developer, acquire, through eminent domain, any lot between
20 Grand Avenue and the beach [ 6]. After the MOU was signed Asbury Partners began to assemble properties. The process began with Asbury Partners purchasing properties from willing sellers. However, it was not able to purchase all the properties it needed to assemble; and ultimately some properties were condemned. New Jersey law does not allow the increase or decrease in property values which may result in the marketplace following the announcement of a potential public project. New Jersey courts addressing this subject have found that project influence should not be considered when a property i s appraised for possible condemnation [ 62]. New Jersey eminent domain law states, Just compensation shall be determined as of the date on which action is taken by the condemne r which substantially affects the use and enjoyment of the property by the conde mnee [ 48]. The objectives of this statute are to protect the condemnee from a diminution in value resulting from the cloud of condemnation placed on the property by a potential condemne r and to insulate the condemne r from the ravages of an inflatio nary spiral.[ 46] Therefore, any project enhancement or condemnation blight recognized in New Jersey must be separated from a propertys market value when an appraiser is performing a condemnation valuation.
21 CHAPTER 2 EMINENT DOMAIN LAW AND AREA DESCRIPT ION A large aspect of the eminent domain issue involves what is legally permissible. In order to explain how eminent domain affects property prices, this study must first explain eminent domain law and the legal ramifications involved in condemnations. The eminent domain power rests with the people. Its allowance in the United States is based in the Fifth Amendment of the U.S. Constitution. Its use extends to state and local governments through the Fourteenth Amendment. However, different court decisions ha ve given differing interpretations of the law. These interpretations of the law influence what is legally per missible in condemnation cases. A description of the study area is also necessary in order to understand the blight that was affecting Asbury Park and the scope of the redevelopment project. Case and Mayer explain that population age and the presence of minority population can have an effect on appreciation rates [ 11]. Therefore, Asbury Pa rks characteristics in 2000 will be described in this chapter The methodology for quantifying project influence is dependent on the scope of the project and the date of the governments commitment. Therefore, the Waterfront Redevelopment Pla n will also be described in detail. Federal Takings Law The power to take p rivate property by eminent domain is viewed as an inherent political right of the sovereign [ 49]. It is limited only by constitutional and statutory laws . The Fifth Amendment constricts the power of eminent domain by two conditions. The property taken must be taken for a public use and the property owner must be compensated justly. The due process clause of the Fourteenth Amendment has been construed to require the payment of just compensation for the taking of private property by the states [ 17]. Acco rding to established rules and principles a governments necessity and expediency for a specific taking are considered
22 discretionary questions left for legislative resolutions [ 21]. These legislative issues include the decision to use eminent domain, the e xtent of the property taken, the need for particular areas, and the amount of land affected [ 50]. Therefore, the judicial issues that are determined in eminent domain proceedings involve a legislatives purpose for condemnation and the amount of just compe nsation awarded. From Public Use to Public Purpose While it is no t apparent that the Public Use Clause of the Fifth Amendment to the United States Constitution is a constraint, most scholars assume it is a limiting clause and the court has treated it as su ch [ 36]. Originally, the condemnation of private property was for government ownership of public infrastructure, such as roads, schools and government buildings . Over time this broadened to allow the property taken to be used by highly regulated priva te companies, such as railroad and utility companies. Since such uses were essentially open to the public these takings were considered a public use . However, the decisions of three Supreme Court cases broadened the interpretation of the Fifth Amendments Takings Clause to include the condemnation of private property solely for the purpose of transferring it from one landowner to another. These cases were Berman v. Parker , Hawaii Housing Authority v. Midkiff [ 32], and Kelo v. City of New London [ 36]. These decisions show that the Supreme Court has consistently embraced a broad interpretation of the public use requirement. The Supreme Court has concluded that the Fifth Amendment authorizes the use of eminent domain in any project that is undertake n for the benefit of the public, even if the actual use is not open to the public . Berman v. Parker Berman v. Parker was a Supreme Court decision that interpreted the Fifth Amendments Taking Clause, nor shall private property be taken for public use without just compensation
23 [ 65]. The plaintiffs in Berman v. Parker challenged the condemnation of their profitable retail establishment for transfer to another private owner . This condemnation was the first phase of the District of Columbias urban renewal plan, which targeted the area southwest of Washington, D.C. . Although the area s outhwest of Washington, D.C. was declared blighted, some of the property was being profitably used in a productive manner . Under the renewal plan some of the property taken would be kept in public ownership and used to build roads, schools, and parks . However, the rest would be assembled according to a comprehensive plan for various residential and commercial uses . This land would be resold to privat e owners and developers to accomplish the purposes of the renewal plan. Berman objected to being forced to sell his profitable business to the District in order to transfer it to another private landowner. The United States Supreme Court upheld the Distric ts authority to use its power of eminent domain for this public purpose . In this ruling Justice Douglas expanded the definition of public use to include public purpose based on physical, aesthetic, and monetary benefits and stated that the purpos e of the redevelopment plan was to address the blight issues in the area . After this Supreme Court decision a broader definition of public use was allowed in eminent domain cases. In fact, as David Mathues reports in his note, [a]ll reported federa l appellate decisions between 1954 and 1986 in which the definition of public use was contested upheld the challenged use of eminent domain. [ 42 ] Hawaii Housing Authority v. Midkiff Hawaii Housing Authority v. Midkiff was another Supreme Court case deal ing with the forced transfer of private property from existing owners to other private owners [ 32]. The Hawaii Land Reform Act of 1967 authorized the use of eminent domain to force landowners to sell their land in fee simple to their lessees [ 32]. Accordin g to the Hawaii legislature, the forced sales were necessary to break up the oligopolistic land ownership which was skewing the states residential
24 fee simple market, inflating prices, and injuring the public tranquility and welfare [ 32]. Midkiff objected to being forced to sell his real property to his lessee, and the Supreme Court ruled in favor of the Hawaii Housing Authority stating that the takings to correct concentrated property ownership was a legitimate public purpose. The Hawaii Housing Authority v. Midkiff decision reiterated the previous Supreme Court decision that eminent domain could be used to serve a public purpose not just a public use. These two Supreme Court decisions set a legal precedent, which later S upreme Court Justices followed. Kel o v. New London Susette Kelo, v. The City of New London, Connecticut was a Supreme Court case that dealt with the condemnation of nonblighted property. The City of New London, an economically distressed city, with a high unemployment rate, sought to rev italize an area of town left substantially underused when the federal government closed the Naval Undersea Warfare Center in the Fort Trumball area [3 6]. The comprehensive revitalization plan called for the integrated redevelopment of ninety acres, which r equired the New London Redevelopment Corporation to acquire all of the privately owned property within the plans boundaries . Susette Kelo and several of her neighbors owned residential property in this area. The properties were neither blighted nor i n poor condition, but were subject to condemnation because they were located within the area covered by the redevelopment plan [3 6]. The landowners objected to the transfer of their land to another private landowner for redevelopment purposes. The Supreme Court determined that a citys decision to take property for the purpose of economic development satisfies the public use requirement of the Fifth Amendment. The court observed that the governmental taking was meant to revitalize the local economy by creating temporary and permanent jobs, generating a significant increase in tax revenue, encouraging spinoff activities and maximizing public access to the waterfront [3 6]. This decision has
25 prompted municipalities to make use of their eminent domain authority. Since the Kelo decision, thousands of properties have faced the threat of eminent domain for private redevelopment, and many more projects are in the planning stages [ 15]. Project Influence One of the key components in determining what constitutes just compensation is the date of valua tion of the property taken This date is important because of the influence a project can have on property prices in the area. Project influence in a condemnation case refers to a positive or negative change in the market value of property as a result of the public project for which all or part of th e property is being taken [ 22]. Scope of the Project Rule When confronted by this problem apprais ers are required to follow the scope of the project r ule. This rule stat es that any increase or decrease in market value directly attributable to the project should not be considered in determining market value [22 ]. The United States Supreme Court case United States v. Miller dealt with the issue of project influence and the scope of the project rule. This court case explains how project influence has been dealt with in the past and how the scope of the project rule is applied in specific instances. United States v. Miller United States v. Miller [ 63 ] involves the enhancement of property values caused by the construction of t he Shasta Dam and Reservoir in N orthern California. In 1936, alternative routes for the relocation of a railroad right of way that would be flooded by the construction of the dam and reservoir were surveyed and staked [ 63]. One of these routes passed through the Miller property. Congress authorized the project in 1937 and property values escalated in proximity of the proposed project. In 1938, the United States filed in the District Court for Northern Califo rnia a complaint in eminent domain against the respondents and others whose lands were needed for
26 the relocation of the railroad [ 63]. The federal district court ruled that the appraisers had to estimate the value of the property excluding any enhancement in value from the proposed project after its date of authorization in 1937. The Supreme Court upheld this decision, stating: If the public project from the beginning included the taking of certain tracts but only one of them is taken in the first instance the owner of the other tracts should not be allowed an increased value for his lands which were ultimately to be taken. [ 63] The Supreme Court Justices questioned whether the lands taken were probably within the scope of the project from the time the go vernment was committed to it [ 63 ]. They concluded that if lands eventually taken were not within the scope of the project, but merely adjacent, the property owner should not be deprived any value added in the meantime by the proximity of the improvement [ 63]. However, if the lands eventually taken were within the scope of the project, the government should not have to pay any increase in value arising from the known fact that the lands proba bly would be condemned [ 63]. New Jersey Eminent Domain Law The stat e of New Jersey allows the taking of private property to clear blighted areas under the state constitution and state statutes. Article VIII, Section III of the State Constitution states: the clearance, replanning, development, or redevelopment of blighte d areas shall be a public purpose and public use, for which private property may be taken or acquired. Municipal, public or private corporations may be authorized by law to undertake such clearance, re planning, development or redevelopment.[ 45] The New J ersey Local and Redevelopment Housing Law [ 47] provides the framework to designate an area blighted or in need of redevelopment; and the Eminent Domain Act of 1971 [ 48] provides the steps to be taken by the government to condemn property. These guideline s provide direction for a municipality attempting to redevelop property within their jurisdiction.
27 The New Jersey Local and Redevelopment Housing Law provides the framework for municipalities seeking to designate an area in need of redevelopment. This de signation acknowledges an area as blighted and allows for steps to be taken to remediate the blighting conditions. An area can be determined in need of redevelopment if it is found to have: substandard, unsafe, unsanitary, dilapidated, or obsolescent buildings. Also, if property is abandoned or vacant land not likely to be developed through private capital. [ 47 ]. Redevelopment law requires substantial public input and process before an area can be determined to constitute an area in need of redevelopment. An investigation must be undertaken and a report prepared and reviewed by the Planning Board . The Planning Board must conduct a public hearing and make a written recommendation to the municipal governing body. A comprehensive redevelopment plan mus t then be prepared by the Planning Board and a public report made to the municipal governing body. The municipal governing body must then conduct a public hearing and adopt the redevelopment plan by ordinance . If any property is being considered for c ondemnation, it must be specifically identified in the redevelopment plan being considered and made public at that point. This process allows for transparency and takes months and even years to complete . Eminent Domain Act of 1971 If an area is deemed to be in need of redevelopment and a redevelopment plan is adopted, then the municipality is authorized to purchase or condemn property pursuant to the provisions of the Eminent Domain Act of 1971 [ 48]. The Eminent Domain Act of 1971 provides the steps to be taken by the government to condemn properties once they have been declared blighted. These steps are taken only if the condemner is unable to acquire title or possession through negotiations because of a disagreement on compensation [ 48]. A municipal ity following these steps must first file a complaint in form and content to demand judgment that a condemner
28 has duly exercised its authority to acquire the property being condemned [ 48]. Then, a condemner shall issue, and with due diligence, cause proces s to be served or published [ 48]. Within 14 days after the filing of the complaint, a notice of the pendency of the action must be filed. Finally, the issue proceeds to court for the dete rmination of just compensation. The Eminent Domain Act grants jurisdi ction on all condemnation matters to the courts. The courts determine the authority to exercise the power of eminent domain, compel exercise of that power, determine compensation, and determine title to all property affected by the action [ 48]. Once it is determined that the condemner is authorized to exercise its power of eminent domain the court appoints three commissioners to determine the compensation to be paid. Commissioners are residents of the county where the property is located. One must be an att orney who is engaged in the prac tice of law in the county [ 48]. Just compensation is determined as of the date of the earliest of the following dates: the date possession of the property being condemned is taken by the condemner in whole or in part; the da te of the commencement of the action; the date on which action is taken by the condemner which substantially affects the use and enjoyment of the property by the condemnee; or the date of the declaration of blight by the governing body [ 48]. This date is determined by the courts. Then, commissioners hold hearings at which the involved parties appraisal witnesses provide testimony of comparable sales. Within 4 months the commissioners make a decision of value to be awarded to the condemnee. If there is an a ppeal by either party as to the award determined by the commissioners the matter goes to a trial with or without a jury [ 48]. Possession of the property is given to the developer when compensation is determined and paid to the property owner.
29 Township of West Windsor v. Yevette Nieremberg Township of West Windsor v. Yvette Nierenberg [ 62] was a New Jersey court decision that identified the date of valuation for a property subject to condemnation. On July 8, 1987, fifty acres owned by Yvette Nierenberg wer e sold to Princeton Manor Associates (PMA) for $4,320,700 for the purpose of development. On July 29, 1988, Yvette Nierenberg received a letter notifying her that West Windsor Township may acquire her property for the purpose of establishing a Community Pa rk [ 62]. PMAs attorney advised PMA that it was unlikely to obtain the proper approval for development on Yvette Nierenbergs property due to the pending public project. From 1988 until 1991, negotiations between PMA and West Windsor Township continued [ 62]. On January 22, 1991, PMA filed a complaint in the law d ivision, seeking, among other relief, just compensation for the Township's alleged destruction of the value of its property The trial court concluded that the date of the Townships letter to Yvett e Nierenberg on July 29, 1988, substantially affected the value of the property sold to PMA. The court stated : A substantial effect upon the use and enjoyment of property is occasioned when the condemne r takes action which directly, unequivocally and imme diately stimulates an upward or downward fluctuation in value and which is directly attributable to future condemnation. [ 62]. New Jersey Statute 20:3 30(c) of the Eminent Domain Act of 1971 states: the date at which an action is taken by the condemner which substantially affects the use and enjoyment of the property by the condemnee is one possible date used to value property subject to condemnation. [ 62] Due to the letter from West Windsor Township, Princeton Manor Associates could not develop the property they had purchased from Yvette Nierenberg. The value of the fifty acres was determined as of July 29, 1988, the day at which the condemners letter notified PMA of future condemnation. Any diminution of value from that date until the date of the taki ng was not considered when determining compensation.
30 Asbury Park Redevelopment Area In 1978 the City of Asbury Park realized the need to begin a program of redevelopment for 230 acres of its waterfront area. The city drafted a Master Redevelopment Plan in order to inform the public of the citys plans. This Master Plan would improve the area using two separate actions: 1) concentrated code enforcement to improve conditions of residential areas which could be preserved and 2) assemblage of underutilized property for major reconstruction [ 6]. This Master Plan acknowledged the need for redevelopment, but it took time to implement. By 1984 the master plan had not begun and the 230 acres between Grande Avenue and the beach continued to deteriorate. The existing housing stock continued to fall far short of meeting modern standards, businesses were severely depressed (compared to former times), and more land was becoming vacant with each passing year [ 6]. An increase in tax exempt properties and the demolition of buildings inside the 230 acres lead to a decrease in the tax base. This caused an increase in the tax burden for the remaining res idents of Asbury Park. Due to the blighting conditions in Asbury Park in 1984 the city declared the 230 acre area between Grande Avenue and the beach blighted. On November 7, 1984 a Waterfront Area Redevelopment plan was adopted [ 6]. In 1986 the city chose Carabetta & Vaccaro Developers to be the main developer of the 230 acres where new construction was planned. It was understood that any property which could not be purchased by the developer would be condemned by the city and sold to Carabetta & Vaccaro In 1989, Carabetta bought out Vaccaro and two years later filed for bankruptcy. At that point all development ceased and the are a continued to deteriorate. For ten years the town could not find a deve loper willing to resurrect the waterfront p lan and the Asbury Park Waterfront Redevelopment Plan laid dormant. From 1991 until 2001 the city searched for a new developer willing to pur chase all debts and liens created by the de mise of the original developer.
31 I n August 2001 the city of Asbury Park selected a new developer for the Waterfront Redevelopment P lan, Asbury Partners, LLC. A Memorandum of Understanding was signed which stated: 1) Asbury Partners would be required to purchase all outstanding debts created by the bankruptcy of Carabetta & Vaccaro Developers; 2) Asbury Partners would be the prime developer; 3) any property owner who was unwilling to sell the ir property to Asbury Pa rtners would be subject to condemnation. The signing of this memorandum began a process of assemblage and redevelopment by Asbury Partners on the same 230 acres that was designated for redevelopment on November 7, 1984. The Waterfront Redevelopment program consists of three subareas, the renovation and infill area, the prime renewal area, and the boardwalk area. In the renovation and infill area buildings were to be renovated and empty lots developed. This area was not subject to condemnation. In the prime renewal area land was to be assembled into larger parcels for new construction. Property within this area was subject to condemnation. In the boardwalk area buildings were to be restored and/or adapted for reuse. Certain vacant areas were to be filled wit h oceanfront resort type uses. The land in the boardwalk area was public property, and was sold to the developer [ 6]. Figure 2 1 is a map of the three areas of eastern Asbury Park that were part of the Waterfront Redevelopment Program. The city informed th e citizens of the plans by holding several public meetings. During these meeting property owners were asked to express concerns and goals for the redevelopment area. Planning and design consultants working for the city and the prime developer presented the ir analysis of conditions in the area [ 6]. The main goals of the redevelopment were to generate new tax revenue for the city and spur economic growth through employment opportunities. After
32 these meetings the city and developer amended the old redevelopmen t plan and drafted a new design for the area. Figure 2 1. Asbury Park Waterfront Redevelopment Area Project Map Source: Asbury Park Planning Board, Waterfront Redevelopment Plan, pp. 8, 15 March 2002 By March 15th, 2002 a comprehensive waterfront redevelopment plan describing the new amendments was finished. It informed the residents which city blocks may be taken for assemblage. The city of Asbury Park stated t hat they would, at the request of the developer, acquire, through the power of eminent domain, any lot within blocks: 118, 127, 128, 129.01, 129.02, 130, 131, 132, 142, 143, 144.01, 144.02, 144.03, 145, 146, 159, 160, 161, 162, 175, 177, 178, 192, 193, 206, 207, 208, 209, 219, 221, 222, 227; and perhaps block 145, lot 1; block 160, lot 8; block 192, lot 1; and block 159, lots 18, 19, 20, 21, 22, 23, and 24 [ 6]. However, the City of Asbury Park could amend, revise, or modify the plan at any time in order to condemn more
33 properties. Figure 22 is a map of the Waterfront Redevelopment Area. All blocks shaded gray were subject to condemnation. Figure 2 2. Areas Subject to Condemnation Source: Asbury Park Redevelopment Board, Waterfront Redevelopment Plan, pp. 74, 15 March, 2005 Asbury Park Economic Conditions By 2000, the economic conditions and community profile of Asbury Park were very different from the rest of Monmouth County or New Jersey as a whole. Asbury Park was more populous and densely populated, had a younger population, had a larger minority population, had a higher proportion of nonfamily households, had a lower per capita income, had a larger population of families below the poverty level and had a lower level of education than most towns in Monmouth County. These statistics show what the environment was like in Asbury Park when redevelopment began. The statistics emphasize the fact that Asbury Park ranked lowest economically among all municipalities in Monmouth County by 2000.
34 In 2000, the City of Asbury Park had a population of 16,930. This was the most densely populated municipality in the county, with 11,842 persons per square mile. Asbury, much like the county, witnessed a large population growth from 1920 until 1950. However, its population remained relatively constant after 1950, peaking in 1960 at 17,336 persons (Table 2 1) Unlike Asbury Parks population, Monmouth Countys population has i ncreased every year since 1920. Table 2 1. Asbury Park and Monmouth County Decennial Population Trends Asbur y Park City and Monmouth County Decennial Population Tre nds from 1910 2000 Year Asbury Park Number % Change Monmouth County % Change Number 1910 10,150 94,724 1920 12,400 2,250 22.20% 104,925 10.80% 10,801 1930 14,981 2,581 20.80% 147,209 40.30% 42,284 1940 14,617 364 2.40% 161,238 9.50% 14,209 1950 17,094 2,477 16.90% 225,327 39.70% 64,089 1960 17,336 242 1.40% 334,401 48.40% 109,074 1970 16,533 803 4.60% 461,849 38.10% 127,448 1980 17,015 482 2.90% 503,173 8.90% 41,324 1990 16,799 216 1.30% 553,124 9.90% 49,951 2000 16,930 131 0.80% 615 ,301 11.20% 62,177 Source: Heyer, Gruel & Associates, Asbury Park Master Plan, Community Profile, pp.12, May 2006. Between 1990 and 2000 the median age in Asbury Park decreased from 32.3 to 30.6 years of age. This trend runs counter to the general graying of Ameri ca as the Baby Boom Generation population continues to age [ 34]. This is due to a high concentration of persons under the age of 19. According to the U.S Census in 2000, 32.6% of the c itys population was 19 years and younger compared with 28.25% of the c ounty (Table 2 2) Eighteen percent of the Citys population was 9 years and younger compared with 14.5% of the County
35 Table 2 2. Asbury Park and Monmouth County Age Profile Age Profile for Asbury Park and Monmouth County in 2000 Asbury Park Monmouth County Age Cohort Number Percent Number Percent Under 5 1,531 9.10% 42,231 6.90% 5 to 9 1,579 9.30% 46,996 7.60% 10 to 14 1,284 7.60% 46,312 7.50% 15 to 19 1,123 6.60% 38,109 6.20% 20 to 24 1,366 8.10% 29,297 4.80% 25 to 34 2,666 15.70% 75 ,308 12.20% 35 to 44 2,380 14.10% 111,681 18.20% 45 to 54 1,895 11.20% 92,239 15.00% 55 to 59 626 3.70% 32,655 5.30% 60 to 64 581 3.40% 23,580 3.80% 65 to 74 904 5.30% 40,084 6.50% 75 to 84 682 4.00% 27,025 4.40% 85 years and older 305 1.80% 9,814 1.60% Source: Heyer, Gruel & Associates, Asbury Park Master Plan, Community Profile, pp.20, May 2006. In 2000, Asbury Park was much more racially diverse than either Monmouth County or the State of New Jersey. At that time seventy five percent of the population in Asbury Park was a minority. The Citys percentage of minorities was much higher than Monmouth County (15.5%) or the state of New Jersey (27.4%). Asbury Parks population wa s mostly comprised of A frican Americans (62.1%). Asbury Park had a much hi gher percent age of African American s than Monmouth County (8%) or the state of New Jersey (13.6%). Correspondingly, the percentage of white residents in the City was significantly less (24.7%) than the County (84.3%) or the State (72.6%) . Slightly les s than half (46.9%) of all households in Asbury Park in 2000 were nonfamily households. That meant nearly half of all households living in a dwelling in Asbury Park were comprised of a single individual or individuals living with nonfamily members. This percentage
36 of nonfamily households in Asbury was much higher than t he rest of the County (28.5%) . Of the family households in Asbury Park, half of them consisted of a primary female householder, without a husband present. This wa s much higher than in Monmouth County (14%). Thirty eight percent of family households in Asbury were marriedcouple households, which wa s much lower than the rest of the County (81.4%) [3 4]. While Asbury Park experienced a modest increase in per capita income (20%) between 1990 and 2000, the increase was significantly less than that experienced by Monmouth County (51.5%) or the state (154.3%) over the same period . In 2000, the per capita income in Asbury Park was $13,516, which is less than half of the County per capita i ncome of $31,149. The average hous ehold income in the City was $23,081 which wa s less than half the County average of $64,271 and the State average of $55,146. More than one in every five households (21.8%) in Asbury Park reported an annual income of less than $10,000 compared with one out of every twenty (5.5%) for the County (Table 2 3). More than half (53.8%) of all households in the city reported annual incomes of less than $25,000 compared with less than a fifth (17.5%) of households in the County. In 2000, nearly one third (29.3%) of all families in Asbury Park reported income that was below the poverty level . The County reported 4.5% of all families below the poverty level in 2000. While Asbury Park accounted for only 2.2% of all families in the County, the City accounted for a proportionally much higher number of persons below the poverty level (14.7%) . The poverty in Asbury Park can be best shown through an example. Housing costs of owner and renter occupants as a percentage of total incom e were high in Asbury Park (Table 24). Forty five percent of home owners and fifty one percent of home renters were spending over 30% of their incomes for housing costs in 2000 . The s tate affordability threshold for
37 housing as a percent of income sta ted no more than 28% of gross income was to be allocated for housing goods or rent . Table2 3. Asbury Park and Monmouth County Household Income Household Income Distribution of Asbury Park & Monmouth County in 2000 Asbury Park Monmouth County Income Number of Households Percent Number of Households Percent Less than $10,000 1,477 21.80% 12,292 5.50% $10,000 $14,999 802 11.80% 9,194 4.10% $15,000 $24,999 1,375 20.30% 17,684 7.90% $25,000 $34,999 896 13.20% 19,394 8.60% $35,000 $49,000 866 12.80% 28,030 12.50% $50,000 $74,999 759 11.20% 43,074 19.20% $75,000 $99,999 345 5.10% 32,229 14.40% $100,000 $149,000 196 2.90% 35,533 15.80% $150,000 or more 70 1.00% 27,017 12.00% Total 6,786 100% 224,447 100% Source: Heyer, Gruel & Associates, Asbury P ark Master Plan, Community Profile, pp.31, May 2006 In 2000, Asbury Parks population was less educated than the Monmouth County. Eighty five percent of the C itys population 25 or older had not attained a bachelors de gree. Sixty four percent had a high s cho ol level of education or less . This wa s in great contrast to the rest of Monmouth County where fifty eight percent of its population had not attained a bachelors degree and forty percent had a high sc hool education or less. Between 1994 and 1999 Asbury Park added a total of 376 jobs, an increase of 16.3% However, unemployment in the c ity was close to 10% for each of those years. By 2000 the
38 unemployment rate was 7.6%. The c ity ha d a history of high unemployment rates. The population to jobs ratio in Asbury was 7.34 in 2000. That was much higher tha n the c ounty ratio (1.37) . Asbury Park Housing Characteristics In 2000, there were 7,744 housing units in Asbury Park. Most of these units were occupied; however, nearly 13% stood vacant. The highes t vacancy rates in the city were located in the Waterfront Redevelopment Area [3 4]. Also, Asbury Park ha d a low rate of home ownership. Of the occupied units in the c ity in 2000, four out of every five (80.5%) were renter occupied (Table 2 4) More than ha lf of the housing stock was built prior to 1960 (64.9%). The composition of the Citys housing stock wa s unusual. A quarter (23.2%) wa s comprised of single detached units and a third (32.1%) wa s comprised of multifamily structures co ntaining 20 or more uni ts [3 4]. In 2000, the City of Asbury Park had a very transitory population. Nearly a third (30%) of all housing units in the City had someone move in between 1999 and 2000 . Another third (31%) became occupied between 1995 and 1998. This high transitor y population wa s understandable with such a high renter population. The median house value in Asbury Park in 2000 was $92,800, which was less than half the County median house value of $203,100 (Table 25). In 2000, the median monthly rent in Asbury Park w as $615, which was less than the countywide average rent of $759 .This shows that the median property values in 2000 in the City of Asbury Park were much less than the county as a whole. Also, the median monthly rent in Asbury Park was nearly $150 less than the county in 2000.
39 Table 2 4. Asbury Park Housing Data Housing Unit Data of Asbury Park in 2000 Characteristics Number Percent Tenure of Occupied Units Owner Occupied 1,316 19.40% Renter occupied 5,438 80.50% Total Occupied 6,754 100% Year Structure Built 1999 March 2000 0 1995 1998 33 0.40% 1990 1994 49 0.60% 1980 1989 478 6.20% 1970 1979 961 12.40% 1960 1969 1,201 15.50% 1940 1959 2,538 32.80% 1939 or earlier 2,484 32.10% Units in Structure One detached 1,794 23.20% One attached 151 1.90% Two 969 12.50% Three or Four 914 11.80% Five to Nine 615 7.90% Ten to Nineteen 770 9.90% Twenty or More 2,507 32.40% Total 7,744 100% Number of Rooms One 673 8.70% Two 919 11.90% Three 1,914 24.70% Four 1,679 12 .10% Five 937 8.80% Six or More 1,622 33.80% Source: Heyer, Gruel & Associates, Asbury Park Master Plan, Community Profile, pp.37, May 2006 The socio economic conditions in the City of Asbury Park, New Jersey were very different than the rest of the co unty in 2000. Asbury Park was more populous and densely populated, younger, poorer, and less educated than the rest of the county in 2000. These conditions show the disparity between the City of Asbury Park and Monmouth County when development began in 2001.
40 Table 2 5. Asbury Park Housing Values Housing Values, 2000 Asbury Park Owner Occupant Housing Units By Value Value Range Number Percent Less than $50,000 84 6.40% $50,000 $99,000 701 53.30% $100,000 $149,999 372 28.30% $150,000 $199,000 109 8.30% $200,000 $299,999 50 3.80% $300,000 $499,000 0 0 $500,000 $999,999 0 0 $1,000,000 or more 0 0 Total 1316 100.00% Median Value $92,800 Renter Occupant Units by Contract Rent Contract Rent Number Percent Less than $200 542 10.30% $200 $299 268 5.10% $300 $499 935 17.80% $500 $749 2406 45.80% $750 $999 780 14.80% $1,000 $1,499 199 3.80% $1,500 or more 16 0.30% No Cash Rent 107 2.00% Total 5253 100.00% Median Contract Rent $615 Source: Heyer, Gruel & Associates, Asbury Park Master Plan, Community Profile, pp.41, May 2006
41 CHAPTER 3 ECONOMIC THEORY The previous chapters introduced the court cases involving eminent domain and project influence, the problems afflicting the City of Asbury Park, and some government tools used to fix those problems. This chapter focuses on the underlying economic theory of land markets. In order to apply a methodology to quantify project influence it is necessary to understand the determinants of property value in the property market. D ue to the unique characteristics of real property, the property market is different than the market for other consumer goods. This chapter focuses on (1) land market theory, (2) the concept of land rent and Ricardian rent, (3) von Thunens location theory, (4) income capitalization theory and (5) property appraisal theory. Land Market Theory Neoclassical microeconomic theory states that the market for a good consists of rational buyers and sellers, motivated by maximizing their self interests [ 66]. Market p rice is influenced by a buyers willingness and ability to pay for a good at a certain price and a sellers willingness and ability to supply a good at a certain price. Typically in a market, the higher a goods utility and scarcity then the higher its pri ce. Land also derives value from its utility and scarcity. The price and quantity of land exchanged is determined at the point where supply and demand interact in the market. The overall supply of land resources is comprised of its physical and economic supply. The physical supply of land is the physical existence of land resources available on the planet. The physical supply of land is considered to be fixed, making it a scarce resource. There can be no more physical land supplied to the market than alread y exists. The economic supply of land is the quantity of land that is offered for sale at a given price. The economic supply of land is the portion of physical supply that man uses [ 9]; therefore, it is fixed and limited only by the total
42 physical supply. By clearing unused forest or brushland the economic supply of land can be increased, but the physical supply does not changed. Raleigh Barlowe, a land economist, explain ed how supply and price are related in the land market, The economic supply of land r esources is responsive to price and demand factors, and it reflects the scarcity or abundance of physical land resources, their relative accessibility, and their general use capacity. [ 9]. Microeconomic theory states that consumers want to maximize their utility subject to their income constraint. This concept is paramount in the understandi ng of demand. Barlowe writes that it is necessary for us to emphasize effective demand to buy the determination of prices and the movement of products in the market [ 9]. It is not sufficient for a consumer to desire a good or service. The consumer must be willing and able to purchase their wants and/or needs in order for their demand to influence the market. Barlowe describe d th e demand for land resources as representing the summation of its direct and derived demands [ 9]. The direct demand for land is the desire for the land itself. The derived demand for land comes from the demand for the products the land produces such as a fa rms commodities, or an apartments rental services. Most people want land resources as a means to an end. Land resources are desired because they offer opportunities for income, employment, food and housing. The sum of the economic supply and demand of va rious types of land equals the overall supply and demand in the land market [ 9]. When describing the supply and demand for property it is usually meaningful to distinguish between different types of land resources. By distinguishing the type of land resour ce one can describe the overall supply and demand of different types of land (i.e. farmland, cropland or residential land). These submarkets of land
43 can exist independently of one another, with different market participants operating in each submarket. T he Interaction of Supply and Demand Under perfectly competitive assumptions the point at which the demand and supply curves intersect determines the market price and quantity demand of a product (P and Q in F igure 3 1). A change in the demand for residenti al land, brought on by population growth, would shift the demand curve out (to the right) from D to D. If the supply of residential land remains unchanged the price of residential land rises from P to P. Shifts in the overall supply situations for the va rious land uses are usually prompted by changes in demand [ 9] Barlowe described these market forces, Generally speaking, these changing demand conditions reflect the current situation regarding population numbers, income levels, individual needs and choi ces, and the impact of technology both in stimulating additional demand and in providing possible substitutes. [ 9 ] However, in the short run the supply of land is fixed due to the length of time it ta kes to bring new land into use. The differences in the slopes of supply and demand curves represent how elastic (responsive) individuals are to price changes relative to quantity changes. For a change in price that results in a less than proportionate change in quantity supplied or produced, the economic suppl y is s aid to be inelastic. Barlowe wrote When the prices of goods produced on the land are high relative to costs and the market outlook is favorable, grazing lands are plowed for food production, subdivisions are developed, and new lands are put into pr oduction. [ 9] Due to the time it takes to buy and sell property the supply and demand of most land resources tend to be inelastic in the short term.
44 Figure 3 1. Supply and Demand of Residential Land Unique Characteristics of the Land Market Despite th e similarities between land markets and perfectly competitive markets, the market for land resources has a number of characteristics which make it unique. First, land is fixed in its location. Second, land is heterogeneous in nature, differing in fertility climate, vegetation, location, and access. Third, land being sold at any particular time represents only a small proportion of the total supply. Fourth, land is generally sold and financed in large units that are not easily divisible. Finally, while the economic supply of land for any one use changes, the
45 physical supply of land is fixed. These imperfections in land markets cause the supply and demand for land to be very localized and fragmented [ 9]. For example, the market for housing in Gainesville, Flo rida tends to have market participants located exclusively in town, making it highly localized. This housing market can be divided into the markets for owners and renters, and the rental market can be further sub divided into the markets for apartment, condo, or house rentals. This is an example of how fragmented the land market can be. The Concept of Land Rent Land rent is a key concept in economic land theory. Land rent provides a theoretical base for explaining the value associated with real estate resou rces and the incentive given to land ownership [ 9]. Market prices, under free market conditions, will allocate scarce resources to higher valued uses, which command higher market prices and generate larger economic rents. Therefore, land rent can be treat ed as a residual economic surplus, equaling the portion of the total value product or of the total returns that remain after payment is made for the total factor costs or total cost, respectively.[ 9]. Land rent represents a return on invested capital; in this case the invested capital is the land resource. Consider the long run profit maximization situation for a producer, where p = price of the output, q = (x1, x2) is a production function, c ( v z ) is a total cost function with w1 and w2 as the prices o f inputs v and z, respectively [ 51]. The following profit maximizing conditions in terms of inputs : 1212max()(,)xxpfxxcvz (3.1) which has first order conditions
46 ** 12 1 1(,) 0 () fxx pw kfx (3.2) ** 12 2 2(,) 0 () fxx pw lfx (3.3) If ** 12, xx is the optimal profit maximizing choice for each factor, then the output price times the marginal product of each factor should equal the price of the factor. In other words, the value of the marginal product, MP, of a factor should equal t he price of that factor: ** 1121(,) pMPxxw (3.4) ** 2122(,) pMPxxw (3.5) One can also consider the long run profit maximizing conditions in terms of outputs : max(,),ypycvz such that 0 y (3.6) which has first order conditions (,) 0 cvz p yy (3.7) or, p = MC The competitive firm will maximize profits at the level of output, y where the marginal cost, given the level of output and input prices, is equal to the price of the output. The profit maximizing conditions above show that the demand for land is a derived demand, which is derived from the final goods and services that the land produces [ 57]. Figure 3 2 illustrates the concept of land rent with the uses of value product and cost curves [ 9]. Land rent is the surplus illustrated by the shaded rectangle ABED that remains after the variable input costs (rectangle CBEF) is deducted from the total product produced (represented by rectangle CADF). In the value product diagram ( figure 2a), land rent is equal to
47 AVP AFC times the number of variable inputs utilized. In the cost curve diagram (figure 2b), land rent is equal to AR AC times the units of output produced. Differences in the rent paying capacity of parcels may be related to factors of soil fertility or location. Figure 3 2. Use of Value Products and Cost Curve Diagrams to Illustrate the Concept of Land Rent Source: R. Barlowe, Land Resource Economics: The Economics of Real Estate, PrenticeHall, New J ersey, (1978), pp.165. Ricardian Rent Theory David Ricardo, a 19th century economist, set forth what is now considered the classical view of rent theory. He attributed agricultural land rent to soil fertility. Ricardo started his analysis by assuming a newly settled town would put the most fertile soil, Grade A, into production first. On farm A, no rent would accumulate because the farm would operate at the
48 break even point. This is the point at which the revenue from selling the goods would just cover the costs of production. As the population expanded more agricultural goods would be demanded and less fertile land, grade B, would be put into production. Assuming farm B uses the same amount of inputs as farm A, it would produce fewer total agricultural products due to the fertility of the Grade B soil. Therefore, farm B would have a higher cost per unit of output, and would have to sell its goods at a higher price. The products from farm A would also sell for a higher price. Therefore, the grade A land would yield considerable land rent due to the difference between its relatively low costs of production and the new, higher price of the good sold. The Ricardian rent theory can be explained using an example taken from Barlowe [ 9]. The productivity of four pa rcels is viewed in terms of costs and returns per unit of output. Assume the four parcels represent four grades of land (Grades A D) with yield capacities of 50, 40, 30, and 25 units of output. It is assumed that the units of capital and labor used on each parcel have a cost of $100 and that use of the four grades of land generate minimum costs of $2.00, $2.50, $3.33, and $4.00, per unit output, respectively. If all of the output needed was supplied by grade A land, the market price of the good would corres pond with the $2 per unit cost of production. When the demand for these goods increases it raises prices to $2.50. At this point grade B land can be put into use because this price will cover the cost of production at the new parcel. At the higher product price, grade A land receives an economic surplus of $.50 per unit of output. This surplus is unnecessary from the standpoint of continued production; but since it exists, it becomes an economic return, or land rent, to the owners of the grade A lands. Pro duct prices must rise to $3.33 in order for grade C lands to be put into production. This price generates an economic surplus or land rent equivalent to $.83 for every unit of output on grade B lands. This would also generate an additional land rent equal to $.83 cents for each unit
49 of output produced on the grade A lands, equaling a total land rent of $1.33 for each unit of output. A price of $4 per unit of output is needed to bring the grade D lands into production. At this point, a n additional land rent equal to $.67 for each unit of output arises on the grade A, B, and C lands. At a price of $4 per unit output, grade C lands earn a total land rent of $.67 per unit of output; grade B lands earn a total land rent of $1.50 per unit of output and, grade A la nds earn a total land rent of $2 per unit of output. Figure 3 3. Presentation of Ricardos explanation of land rent Source: R. Barlowe, Land Resource Economics: The Economics of Real Estate, PrenticeHall, New Jersey, (1978), pp.170.
50 Von Thunens Location Theory Another 19th century economist named Johann Heinrich von Thunen attributed land rent to the distance from a central market. He hypothesized that if all lands surrounding a central market had the same fertility, the difference i n land rents would be attributed to the distance from that central location. Lands closest to the market would enjoy a rent advantage over those located at a greater distance. This is because land acquired further away from the central market causes net re turns to decrease as higher transportation costs are incurred. This concept is best understood using a diagram from Hartwick and Olewiler [ 51]. Assume n equals the number of workers, and that w is the prevailing wage. Also, y equals the number of boxes of citrus, p equals the price per box paid at a central market and, tx equals the transportation costs per box at distance ( x ). The revenue used to pay labor and land per crop yield at distance x from the central market is ( ptx ), which equals the labor per c rop yield w ( n/ y ), where ( n/ y ) is the workers per crop yield, plus land rent. Therefore, the revenue per box to be allocated to the land at x is: n ptxw y (3.8) Multiplying by y (crop yield per acre) yields: () n yptxw y or (()) pytxywn (3.9) which is the land rent per acre at distance x A bid rent function is derived from Equation 4.9, which reflects the land rent produced at each distance away from the central market (Figure 3 4). When production take s place at the central market, where x equals zero, the equation reduces to reflect the rent per acre as the revenue per acre minus labor costs per acre ( py wn ). Land rents are dissipated by transportation
51 costs as production is transferred to distant parc els. Transportation costs c ontinue to consume rents until x where ( py tx ( y ) wn ) is zero and all rents are exhausted. This point is referred to as the norent margin. The bidrent function approach is useful in examining the allocation of land rents among competing land uses at various distances from the central market. Figure 3 4. Effect of Transportation Costs on Land Rent at Various Distances from the Central Market Source: J. M. Hardwick and N. Olewiler, The Economics of Natural Resource Use, AddisonWesley, New York, (1985).
52 Relation of Land Rent to Land Values (Income Capitalization Theory) Most land resources produce a predictable future flow of reoccurring land rents. From a theoretical standpoint, land and real estate resources have a current val ue that is equal to the present value of their expected future land rents [ 9]. Current values are determined by estimating the expected flows of land rent and determining their present value. This can be described by the equation below: 23... (1)(1)(1)(1)naaaa V rrrr (3.10) where V is the value of the property, a = the expected average annual land rent, and r = the capitalization interest rate, or the rate needed to convert a given periodic payment into a given cash value, and n = the number of years in the analysis [ 9]. When n approaches infinity, the geometric series reduces down to V = a/ r For example, if one can assume an expected average annual land rent of $10,000 and use a 5% capitalization rate, the land resource in question has a value of $10,000 divided by 5%, or $200,000. Appraisal Theory An appraisal is an opinion of value. When real estate is sold a professional real estate appraisal may be needed to determine a propertys value. The professional real estate appraiser uses standard techniques t o develop their opinion of value. Due to conscious efforts over the years these techniques have become more scientific [ 9]. There are three specific appraisal procedures for determining real property values in the U.S. They are: (1) the income capitalization approach or income approach, (2) the market comparison or sales approach, and
53 (3) the replacement cost or cost approach. These three procedures are founded on the basic economic principles of income capitalization, substitution, and production cost, res pectively. The income approach to valuation determines a propertys value by capitalizing the land rent (profit) that the property produces. In theory, the market value for a property should be worth the discounted present value of all expected future land rent streams [ 2]. This is based on the economic principles of producer surplus, land rent and income capitalization, which were discussed earlier in this chapter. Producer surplus is determined by the market price, a firms marginal and average cost curve s and the quantity supplied by the firm. The producers surplus is known as economic rent, or land rent, depending on what the firm produces. The future land rent must be forecasted and that future forecasted income must be capitalized at some rate to determine the propertys value. This method of valuation is used when the property being sold is generating some sort of income. Therefore, the property is usually leased. It is more difficult to use this technique on properties which are owned because it is hard to quantify the use value of a particular owner. The sales comparison approach to valuation determines value based on closed sales, listings and/or pending sales of properties similar to the appraised property. Barlowe explained This approach finds i ts rational in the economic principle of substitution [ 9]. In the free market, informed buyers are not willing to pay any more for a particular property then it costs them to purchase a similar property. Appraisers study the current local land market of t he subject property in order to determine value. The sales comparison approach is applicable to all types of real property interests when there are sufficient recent, reliable transactions to indicate value patterns or trends in the market [ 2]. Typically t he sales comparison approach is reserved for owner occupied properties, i.e., properties that are not income producing. It is most useful when
54 a number of similar properties are sold or are currently for sale in the appraised propertys market. Barlowe exp lain ed This valuation technique works well with properties such as urban and suburban residences, apartment houses, and farms that are somewhat standardized. [ 9] The cost approach to valuation determines value based on what it would cost to reproduce th e subject property or a property with the same utility, based on current construction costs. The estimate of production cost is adjusted for losses in value caused by the age, condition, and utility of the subject property. The value of the land is then added to determine a final cost to reproduce the property. This approach is based on the economic principle of production costs. It is rooted in the belief that there is a close relationship between cost and value. The cost approach is useful when there is a lack of market activity, which would limit the usefulness of the sales comparison approach. The cost approach may be used to develop an opinion of market value and is frequently applied to proposed construction, special purpose or specialty properties, and other properties that are not frequently exchanged in the market such as public buildings [ 2]. In order to apply a methodology to quantify project influence it is necessary to understand the determinants of value in the land market. Land market theory su ggests that the price of land is based on free market economic principles like supply and demand. Land rent theory suggests that properties with higher utility (such as fertility and distance from the central business district) command a higher economic land rent. Using capitalization theory it is shown that economic land rent can be converted to land value. Appraisal theory takes the principles of land market theory, land rent theory, and capitalization theory into consideration when determining land value
55 CHAPTER 4 LITERATURE REVIEW This chapter contains a literature review of house price indexes. These indexes are used to capture house price appreciation over time. Since project influence affects property values, it is necessary to quantify house price movement in this study. One way of approximating project influence is by creating an index for affected areas and comparing it to a market wide index. By capturing the movement of house prices, this study can quantify abnormal appreciation caused by the a nnouncement of a public project requiring the use of eminent domain. The house price indexes created in chapter 5 are constructed using the same methods as those introduced in this chapter. The model is a local repeat sales house price index based on the m ethods created by Bailey, Muth and Nourse (BMN) [ 8] and further developed by Case and Shiller (CS) [ 13]. Finally, this chapter introduces two studies that explain how to capture abnormal appreciation in specific areas compared to the larger market wide area using the repeat sales house price index method [ 59] [5 ]. Price Index Introduction Irving Fisher, in The Making of Index Numbers described a price index number as an average of the percentage change of prices over time. Fisher explained that the great majority of index numbers are actually used to indicate price movements in time. [ 23 ] The percentage change in the price of something through time is found by dividing the price in the future period, Pf, by the price in current period, Pc. The ratio betw een these two prices is called the price relative, i.e., price relative = Pf /Pc [ 23 ]. Therefore, a price index is an average of all price rel atives between periods of time. Fisher described four relevant types of indexes: the simple arithmetic index by th e fixed base system, the simple arithmetic index by the chain system, the simple geometric index and the
56 median index. In 1967, Fisher wrote, the simple arithmetic average of relative prices by the fixed base system is familiar to most people. In fact the very word average means, to most people, only the simple arithmetic average [by the fixed base system]. [ 23] All of these indexes differ based upon the type of mathematical calculation pe rformed to average the numbers. There are two steps in obtaini ng the simple arithmetic price index number. First, determine the price relatives for the terms. Second, average the two price relatives. For example, if two separate acres of land in a particular location both sell for $100,000 in 2000 then both sell for $104,000 in 2001. The price relatives would be (104,000/100,000), or 1.04 for each property. The average of the price relatives, would be (1.04+1.04)/2 = 1.04. This number represents the percentage of appreciation for these two houses compared to the base year, which is always equal to 1.00. In this example the two houses have appreciated 4% over the one year period, (1.041.00) .04 1.00 If there was another property which sold in the same time period we would find the price relative for that propert y and then average all three price relatives to get a simple arithmetic average. Construction of a simple arithmetic price index calculated using the chain system uses changing base years to average the price relatives. The first step for constructing this type of index is to calculate the price relatives for each period. For example, if we have a set of four houses: two selling in the base year (year zero) and again in year one and two selling in year one and again in year two, then we calculate the price relatives for the two separate periods. The first price relative is like a chain connecting year zero to year one (with year zero as the base year). The second price relative is like a chain connecting year one to year two (with year two as the base year). These two price relatives (chains) are then multiplied to link the price relatives and
57 create a price index which represents the appreciation rate from year zero to year two, where year zero is the base year, i.e., base year = 1.00. The simple geometric price index number is calculated using the geometric average. First, the price relatives of two numbers are determined based on two years. Then, instead of adding the price relatives together and dividing by the number of terms, one finds the product of the price relatives and then extracts the nth root of the product. For example, the simple price relatives for the two houses are both 1.04, as they were above. The price relatives are then multiplied and the square root of that product is found, e.g. 21.041.041.04 This index is 1.04, which is equal to the arithmetic average in the above example. The simple median index number is calculated by selecting the middle number from a set of price relatives. The price relatives are determined from on e year to the next. Once the median price relative is chosen then the chain method is used to link all of the yearly price relative chains together and create a median index number. Three economists formulated different methods for constructing price index es in the 19th and 20th centuries. These indexes were created to calculate the change in commodity prices over time. These indexes are named the Laspeyres, Paasche and Fisher price indexes. They are named after their respective creators. The Laspeyres pric e index, named after the German economist Etienne Laspeyres, is computed as 0 00,, ,,nctct l ctctpq P pq (4.1)
58 where Pl is the Laspeyres Index, ,nctp is the price of some commodity in year n, 0, ctp is the pric e of some commodity in the base year, and 0, ctq is the quantity sold of that same commodity in the base year. The current year price is the year the index is constructed. The base year can be any year before that. The Laspeyres index answers the question, How much is the price of a commodity today compared to the price of that same commodity in some base year? The base year is normalized to 1.00 in order to easily describe the index number as a percentage change in price. A Laspeyres index of 1 would state that a consumer in the current period can purchase the same bundle of commodities as they did in the previous period. Hence, the Laspeyres index can be thought of as the inflation rate when the numerator is a bundle of goods using current prices and current quantities. The Paasche price index, named after the German economist Hermann Paasche, is computed as 0,, ,,() ()nn nctct P ctctpq P pq (4.2) w here Pp is the Paasche price index, ,nctp is the price of a commodity at time n, 0, ctp is the price of a commodity in the base year, and ,nctq is the quantity sold of some commodity in year n. The Paasche index answers the question, What is the difference in price in todays commodities versus the price for the same commodities in base year dollars? A Paasche index of 1 would state that a consumer could have purchased the same bundle of commodities in the base period as they are consuming in the current period, assuming a constant income. Hence, one can think of the Paasche index like the inflation rate when the numerator is a bundle of goods using base year pric es but current year quantities.
59 The Laspeyres price index tends to overstate inflation, and Paasche price index tends to understate inflation. This is due to the fact that the consumer is a price taker. The consumer will react to price changes by changing the quantities that they consume. Therefore, Irvin Fisher created an index which was between these two indexes. The Fisch er price index, named after the American economist Irving Fischer, is computed as FPLPPP (4.3) where PF is equal to the Fischer index, PP is equal to the Paasche index and PL is equal to the Laspeyres index. The Fisher Price Index is the geometric mean of the Laspeyres and Pasche price indexes. It is a price index number which is located between the other two indexes. This is done in order to adjust for the over stated and under stated inflations of the Laspeyres and Pasche indexes respectively. The biases associated with each index are minimized by calculating the geometric average. This index answers the question, What is the unbiased price of todays commodities in constant dollars? Repeat Sales Method for House Price Index Co nstruction In 1963 Martin Bailey, Richard Muth and Hugh Nourse wrote an article for the Journal of the American Statistical Association titled, A Regression Method for Real Estate Index Construction [ 8 ]. In this article they noted that the differences in physical quality of real property created difficulties in making estimations of house price indexes. They believed these difficulties could be avoided by basing an index on sale prices of a particular property at different times. Bailey, Muth and Nourse w rote, The problem of combining price relatives of repeat sales of properties to obtain a price index can be converted into a regression problem, and standard techniques of regression analysis can be used to estimate the index.[ 8 ] They believed
60 the regres sion method of estimation would be more efficient than other methods (arithmetic, geometric, and median) for combining price relatives because it utilized information about the price index for earlier periods contained in sales prices in later periods. Als o, by using this method, the standard errors of the estimated index numbers could be computed to see how statistically accurate they were. They explained that the major problem when creating house price index numbers using a median sales index or a chain i ndex was the great variation in quality among properties used. They explained that indexes based upon the average sales prices of all properties of some particular kind in any given time period will be deficient for two reasons: (1) variation in the quali ty of properties sold from period to period will cause the index to vary more widely than the value of any given property, and (2) if there is a progressive change in the quality of properties sold at different times, the index will be biased over time. [ 8] They found that a way to avoid these problems was to eliminate the quality difference by using the sale and resale of an unchanged property over time. Bailey, Muth and Nourse were the first to create a repeat sales house price index using the regression method (referred to here as the BMN method). Below is a description of their methodology. Let: '' t itt itt tB RU B or (4.4) ''' ittttittrbbu (4.5) where lower case letters stand for the logarithms of the corresponding capital let ters. Ritt is the ratio of the final sales price in period t to the initial sales price in period t for the i th pair of transactions with initial and final sales in these two periods. Bt and Bt are the true but unknown
61 indexes for period t and t respectively, where t = 0,1, T 1, and t = 1, T They assumed that the residuals in log form, uitt, ha ve 2, and were uncorrelated with each other. If these conditions are met these estimators will be BLUE, also known as the best linear unbiased estimators. Estimation of the unknown B s can be treated as a regression problem. Let xt take the value 1 if period t is the period of initial sale, +1 if the period of final sale, and 0 otherwise for every propertys pair of transactions. In order to normalize the index, let B0 = 1 or b0 = 0. Using the above assumption, Equation 4.5 becomes: '' 1 T itt jjitt jrbxu (4.6) or, in matrix notation: rbxu (4.7) In Equation 4.7 r and u are ndimensional column vectors, where n = ',' ,' tttt ttnn is the number of pairs of transactions with initial s ale in period t and final sale in period t ; b is a T dimensional column vector of unknown logarithms of the index numbers to be estimated; and x is an n by T matrix. For a pair of sales whose initial period of sale is other than the base period the corres ponding row of x has a 1 in the t th column; for the final sale date, the corresponding row has a +1 in the t th column; all other elements of x are zeros. Given the BMN assumptions made above about u, the least squares estimator: 1 (')(') bxxxr (4.8) i s the minimum variance linear unbiased estimator of b. The t th diagonal element of ( x x ), which is a T by T matrix, is the number of pairs of transactions with initial period of sale in t plus the number of pairs with final sale in period t The t, t th off diagonal element of ( x x ) is ntt.
62 Finally, ( x ,r ) is a T dimensional vector whose t th element is the sum of all price relatives for which period t is the period of final sale minus the sum of all price relatives for which t is th e period of initial sale. Bailey, Muth, and Nourse believed there were advantages and disadvantages to the repeat sales regression method. One advantage to the repeat sales method is it could be easily modified to eliminate the effects on value of certain changes in a property between the period of initial and final sale. Some examples of certain changes that could affect values are: properties that have remodeling done or have built an addition to the structure, the change in the number of dwelling units of an apartment building, or a change in socioeconomic variables in the area. One disadvantage to the regression method is that a depreciation adjustment cannot be readily estimated along with a price index using the BMN regression method. Correlated errors could be another possible problem with the BMN regression method. In many cases there may be data on more than two sales of a given property for the time period covered by the index [ 8]. If so, there is no unique way to reduce these sales to price relati ves, and any way of computing price relatives is likely to have a problem of correlated errors. However, if a property sells three times in three successive years, the first and second years sales should be used as one price relative and the second and third years sales as another price relative. This would avoid correlated residuals . In any case, the estimator (Equation 4.8) is still unbiased if the errors ( uit) have zero mean, regardless of their inter correlation [ 8]. This method was first used i n Hugh O. Nourses Ph.D. dissertation, The Effect of Public Housing on Property Values in St. Louis [ 52]. The advantages of using the repeat sales method over the chain method for real estate price index construction were illustrated in his research.
63 Th e reason for constructing the regression method was caused by the erratic behavior of the chain method estimated. Figure 4 1. Comparison of real estate price indexes for a small area in St. Louis estimated by the chain and regression methods. Source: M. J. Bailey, R. F. Muth, and H. O. Nourse, A Regression Method for Real Estate Index Construction, Journal of the American Statistical Association, 58, pp.941, (1963). In Nourses study it was desired to create a property price index for a small area of nor thwest St. Louis. He obtained his data from tax stamps affixed to warranty deeds. Values estimated from the tax stamp values were then adjusted on the basis of other information in the
64 warranty deed. For instance, if any transaction conveyed partial interest, then the sale price was adjusted by the percentage of interest conveyed . A transaction was not included if more than one property was conveyed in the deed. Figure 4 1 is a chart of Nourses results. It shows the chain and repeat sales property pri ce indexes estimated from the data collected. The two indexes diverge greatly in the years prior to 1943. Most of the difference in the levels of the two indexes is accounted for by the change in the index from 1938 to 1939. In addition, the year to year c hanges in the index estimated by the chain method are very erratic from 1938 to 1943, much more so than the index using the regression method. For 1944 and the following years, however, the indexes estimated by the two methods are very similar apart from the higher level of the chain estimate. Bailey, Muth and Nourse believed that the reason for the divergence between the two estimates was due to the fact that there were relatively few final sales upon which to base the chain estimates for the earlier years In 1989 Karl E. Case and Robert J. Shiller wrote an article for the American Economic Review titled, The Efficiency of the Market for Single Family Homes . The purpose of the paper was to perform tests of efficiency of the singlefamily housing mar ket using data from the Society of Real Estate Appraisers tapes for the years 1970 to 1986 for Atlanta, Chicago, Dallas, and San Francisco/Oakland. These tapes contained sales prices and other information about the homes. Case and Shiller extracted from t hese tapes for each city data on houses sold twice for which there was no apparent quality change and for which only conventional mortgages applied. They created quarterly repeat sales indexes for each city. Their regression was based on the Bailey, Muth, and Nourse (BMN) method. However, Case and Shiller modified this method in order to address the house specific component of the change in log price being heteroskedastic.
65 They believed that the house specific error increased with the interval between sales As the time between sales increased the variance in the errors did also. They named their method of price index construction the Weighted Repeat Sales (WRS) method. Their motivation for the WRS method was the assumption that the log price Pit of the i th house at time t is given by ittititPCHN (4.9) w here Ct is the log of the citywide level of housing prices at time t Hit is a Gausian random walk (where itH has zero mean and variance 2 h ) t hat is uncorrelated with Ct and Hit; i is not equal to j for all T and Nit is an identically distributed normal noise term (which has zero mean and variance 2 N ) and is uncorrelated with Ct and Hit for all j and T and with Nit unles s i = j and t = T Here, Hit represents the drift in individual housing value through time, and Nit represents the error in price due to imperfections in the market for housing. These errors could occur because the price a house sells for depends on random variables such as interested buyers and the specific real estate broker/agent selling the house. Case and Shiller used a three step weighted (generalized) least squares procedure. In the first step, the (BMN) method was followed exactly, and a vector of r egression residuals was calculated. In the second step, the squared residuals in the first step regression were regressed on a constant and the time interval between sales. The constant term was the estimate of 2 N and the slope term was the estimate of 2 H In the third step, a generalized least squares regression (a weighted regression) was run by first dividing each observation in the stepone regression by the square root of the fitted value 2 H in the second stage regression and running the regression again.
66 The results were an estimated WRS index, WRS(t). They explain that one way of describing how well these variables were measured was to compute the ratio of the standard deviation of a variable to the average standard error for that variable. In their study they concluded that for the log index levels this ratio was 13.87 for Atlanta, 24.52 for Chicago, 9.94 for Dallas and 28.03 for San FranciscoOakland. Therefore, they write that they can make very accurate statements about both the monthly and yearly levels of house prices in the cities. Case and Shiller adjusted their estimated WRS index for inflation in order to make a real WRS index in each city. This was done by subtracting the quarterly Consumer Price Index from the Weighted Repeat Sales Index: W(t) = WRS(t) ln(CPI(t)); where W(t) is the real weighted repeat sales index in each city, deflated by the city specific consumer price index, ln(CPI(t)). The advantage of the Weighted Repeat Sales index over the original Bailey, Muth, and Nourse regression index was it took into account the possibility that the variance of the error term was not constant across houses. C S believed that this variance was likely to be related to the inte rval of time between sales, and they showed some evidence that this was true [ 12]. They also believed that homes sold after long time intervals had a greater influence on the index relative to homes sold over short time intervals. The repeat sales index is not free from criticisms. Dean Gatzlaff and David Ling described the issues with the repeat sales house price index method, The repeat sales method is also subject to several criticisms: (1) upgrading of the property between sale dates may be ignored, (2 ) relative property characteristic prices may change over time due to changing preferences, (3) estimation efficiency is reduced because all single sales are dropped from the analysis, and (4) the repeat sales sample may not be representative of the stock of housing leading to sample selection bias. [ 27] Also, as stated above, depreciation cannot be separated from house price
67 change, leading to an index which can be biased upward. Haurin and Hendershot described sample selection bias by explaining that ho uses that sell frequently may be starter homes bought by individuals with a short expected duration of stay. This type of house would therefore appear relatively frequently making the repeat sales data non random [ 31]. In a study by Clapp, Giaccotto and Tirtiroglu they found properties that had sold twice had average sales prices about 15% lower than those that sold only once [ 19]. However, they explained that while short term price trends may differ between full samples and repeat subsamples; there were no systematic differences among the samples in their study of Hartford Connecticut over periods of three or more years. They believe that arbitrage typically forces prices for the repeat sample to grow at the same rate as those for the full sample period [ 19]. Abnormal Local Market Appreciation In 1995 Wayne Archer, Dean Gatzlaff, and David Ling wrote an article in the Journal of Urban Economics titled, Measuring the Importance of Location in House Price Appreciation [ 5]. This study examined the variation in rates of price appreciation within an individual metropolitan market. They explained that the existence of variation in appreciation within metropolitan areas had not been formally documented at the time of the study. Therefore, they developed a methodology to examine the location variation in house price changes in Dade County (Mi ami) Florida, from1971 to 1992. They explained that traditional models of urban rents assume that changes in housing values are largely driven by changes in current and expected future rents. Thus, the formation of expectations and how quickly and completely changes in expected rents are capitalized are important determinants of housing appreciation patterns [ 56] [55 ] [24 ]. They explained, to the extent that market participan ts anticipate any of the differential effects on rent, then their effects should be immediately capitalized into property values in a competitive housing market.[ 5] This
68 would imply the absence of an effect on appreciation from an event after the period i n which it is revealed. Therefore, the only observable appreciation differences should be from unexpected differential shocks in the value of a homes physical or locational characteristics [ 5 ]. Their example supposed that decisions to add or expand roads and highways may result in significant relative price changes. Such anticipated shocks may affect ex post price appreciation, especially over short time periods. The effects of these shocks may even be discernable over longer time (holding) periods to the extent that anticipated shocks are (1) not immediately capitalized or (2) not uniformly distributed across the urban area [ 5]. Their model supposes: ijttjtijtPMSR (4.10) where Pijt is the unit price of the i th residence in submarket j in period t ; Mt is the unit price index for the metropolitan area as a whole (market wide) during period t ; Sjt is the unit price index for submarket j relative to the market wide price index, in period t ; and Rijt is the unit price level for the i th res idence in submarket j relative to the submarket price level, in period t In essence Sjt is the amount by which the market wide index at time t must be scaled (up or down) to adjust for the location of the residence in submarket j Rijt represents the por tion of the unit price that is not explained by the market wide index and the location of the property in a particular submarket. In effect, Rijt, is a residual scale factor that captures the idiosyncratic price changes of the individual home. Appreciation over any k number of previous periods would be revealed in the following ratio of prices: () ()() ijtktkjtkijtk ijt tjtijtPMSR PMSR (4.11)
69 C onverting Equation 4.11 to natural logs gives () () ()ln(lnln)(lnln)(lnln).ijtk tk t jtkjt ijtkijt ijtP MMSSRR P (4.12) This expression is similar to that used in constructing house price indices by the method of repeat sales (e.g., Bailey, Muth and Nourse and Case and Shiller). However, it is generalized here to account for cross sectional abnormal appreciation appreciation from the market wide appreciation rate. Their study was concerned with inter temporal changes in the values of Sijt, which represented subm arket (cross sectional) relative variation in the housing price index. If Sjt+k, the scalar to the market wide index for submarket j at time t+k equals Sjt, the scalar at time t then price appreciation in submarket j has equaled that of the overall marke t over those k periods. However, if Sjt+k differs from Sjt, then submarket j has appreciated at a rate different than the overall market. In F igure 4 2, Archer, Gatzlaff and Ling showed both of these situations. The top curve is the price index for submarket j ; the bottom curve is the market wide index. The two are equal at time zero. Over the entire T year study period, submarket j has experienced more appreciation than the market (distance D >distance d ). However, relative appreciation rates are sensitive to the specific time interval. In their example, from time period zero to time t submarket j appreciated more quickly than the market ( A > a ). From time t to time t+k submarket j experienced negative appreciation relative to the market (the ratio B / b is le ss than A / a). The submarket and the market appreciated at identical rates from time t+k to time t+2k ( C / c = B / b). In terms of Equation 4.11, identical appreciation implies Sjt+2k = Sjt+k.
70 Figure 4 2. Sjt is the ratio of distance A to distance a, i.e., (A /a). Sjt+k is (B/b), etc. Since (B/b) is less than (A/a), Sjt+k
71 () ''() '1lnT ijtk titijtkt t ijtP mDe P (4.13) where Pij( t+k )/ Pijt is the pricerelative of property i as defined earlier in Equation 4.10; Dit is a dummy variable which equals 1 at the time of the initial sale ( t = t ), +1 at time of the second sale ( t = t+k ), and 0 otherwise; mt is the logarithm of the cumulative (marketwide) price index in period t ; and ei j( t + k ) t is a disturbance term. The value of m1 is set at zero to normalize the index of beginning period sales to a value of 1.0, and the T 1 subsequent coefficients are estimated by an Ordinary Least Squares (OLS) regression. Deviations in individual house price relatives from the estimate of the market index are captured by the vector of disturbance terms. In this formulation any submarket (locational) effects are captured in the disturbance term. Archer, Gatzlaff, and Lings estimation procedure general ized the standard model represented by Equation 4.13 to estimate the relative price for each submarket j Similar to Dit, a matrix of submarket time dummies, Ij, is constructed. For each submarket, Ijt is set equal to 1 at the time of the initial sale; + 1 at the time of the second sale; or otherwise zero. This model can be written as () ''''() '1 '1lnTT ijtk titjtjtijtkttt ijtP mDsIe P (4.14) T he coefficients on Ijt, sjt represent the marginal difference between the index submarket j and the market wide index. By differencing the estimated coefficients on the interacted terms (i.e., sjt+k sjt), one can yield an estimate of the cumulative abnormal appreciation for submarket j, relative to the metropolitan area as a whole, over a k holding period. Equation 4.14 is estimated in two steps. In step 1 the market wide coefficients, mt, are estimated for the entire metropolitan sample, using Equation 4.13, and their effect is removed. In step 2 the residuals from step 1 are regressed on the time dummies, It, for submarket j Step 2 is
72 repeated for each submarket group, obtaining the respective sjt for each submarket. In the twostep procedure, tests of significance for the sjt test the difference between the overall metropolitan house price index and the house price index of each submarket. Tests of differential appreciation are conducted in two ways: unrestricted and subject to a holding period restriction. Archer, Gatzlaff and Ling initially hypothesized that the index level of the census area was equal to the index level of the ove rall market in each year. That is, ln Sjt (in Equation 4.12) was equal to zero for all time periods. They tested that hypothesis for all of their submarkets by evaluating the standard statistical hypothesis; that the estimated coefficients on the submarket s (interaction) dummy variables (the sjt values from E quation 4.14) were jointly equal to zero. They use Equation 4.14 to test the submarket interaction coefficients in the absence of constraints. This way a standard F test then indicated whether controll ing for submarket location (as in Equation 4.14) significantly increases the explanatory power, relative to Equation 4.13. Their second set of tests examined whether appreciation in submarket j was different from the average appreciation for the overall market, assuming a particular holding period. For this purpose the estimate of cumulative abnormal appreciation over a k year holding period is given by (')' '11 (lnln)Tk jtkjt tss Tk (4.15) where sjt are the coefficients from estimating Equation 4.14. With 21 years of historical data and a five year holding period, T k equals 16 in Equation 4.15. That is there are 16 sequential 5year appreciation rates (years 1 to 6, 2 to 7, 3 to 8, etc.). Dividing through by T k provides the average ex post 5year appreciat ion rate. Note that the annualized 5year appreciation rate will differ from the annualized appreciation rate over the entire 21 year study period. The appropriate F test then tests if Equation 4.15 is different from zero rather than whether the 21 individual
73 coefficients are different from zero. Note that the resulting F tests are being applied to a linear combination of submarket coefficients. If unexpected valuation shocks are not random over space and/or if the effects of shocks on price are transitory (perhaps due to supply responses), the second set of tests (which impose linear restrictions) will produce fewer rejections of the null hypothesis of no abnormal returns. Archer, Gatzlaff and Ling used repeat sales data from the Florida Department of Reven ues 1992 property tax records to estimate house price indices for the Miami MSA (Dade County). The data included information on the most recent selling prices and dates (year and month) for all properties sold between 1971 and 1992, as well as other prope rty specific characteristics such as lot size, living area, age of structure, and the owners mailing address. Archer, Gatzlaff and Ling found that slightly over one half of the 79 census tract groups examined experienced rates of appreciation that were s ignificantly different from the overall Miami marke t 5]. In 2000, Greg Smersh and Marc Smith wrote an article in the Journal of Housing Economics titled, Accessibility Changes and Urban House Price Appreciation: A Constrained Optimization Approach to Determining Distance Effects [ 59]. In their study they explored the construction of the Dames Point Bridge over the St. Johns River in Jacksonville, Florida, to examine the impact of changes in a ccessibility on housing values. They explain that public decisions regarding transportation networks are much debated and are important public policy concerns. They wrote that recent debates had shifted the focus from measuring the degree of property value change to responses to the value change. They explained One response had been to seek to capture the benefits that new and existing developments derive from government provided facilities through the adoption of impact fees, special
74 assessments, and related fees and charges imposed on beneficiaries. Another response had been the private proper ty rights initiatives that provided for property owners to receive compensation for reductions in the value of their property that resulted from government actions [ 5]. They believed any of these actions required measures of the effects of change of prope rty prices. They explained that an examination of a bridge presented a unique opportunity to examine the effects of a change in the transportation network on property prices because it presented the possibility of both positive and negative effects resulti ng from a single change. They explained that, these [positive and negative] effects result if there is improved access to destinations on the opposite side of the river provided for those on one side and result in increased congestion on the opposite side [ 5]. They investigated both benefits to the north of the bridge due to increased accessibility from the urban perimeter and negative repercussions to the south of the bridge due to increased traffic congestion and crime. The Dames point bridge was built e ast of downtown Jacksonville in 1989. It connected northeast and southeast Jacksonville for the first time. Smersh and Smith hypothesized that properties to the north would experience abnormal positive appreciation because of gained accessibility to the de velopment southeast of downtown, including employment and shopping, as well as access to the beach. They believed that the areas southeast of the bridge may be negatively affected due to the increased congestion, potential zoning changes, and resultant inc reased intensity of development, cri me, and other negative effects. They explained that the impact of the bridge on property values would not be immediate because it would be a large public works project with a long construction period leading to effects o n property values that would precede the completion of the project due to anticipation of the change [ 5]. Therefore, the beginning of the impacts would precede the 1989 opening of the
75 bridge and might be traced back to the announcement of its construction (or even before then for those land owners aware of the potential construction). They believed that the benefits of improved accessibility would therefore be felt over an anticipatory period as the market determined the impact of the bridge [ 5]. However, a fter 1989 the expectation was that the effect of the change would have been captured and appreciation rates on opposite sides of the bridge would be similar, albeit starting from a different price level than before construction. Therefore, their hypothesis for the study was that through the construction of the bridge there would be positive appreciation to the north, relative to the city as a whole, because of gained accessibility to downtown Jacksonville [ 5 ]. They also hypothesized that there would be nega tive appreciation relative to the city as a whole due to increased congestion and other factors. Smersh and Smith employed the repeat sales methodology to test for abnormal (positive and negative) appreciation in geographic submarkets. This technique allow ed for the estimation of inter temporal market price indices for quality adjusted or standardized properties. This technique was originally developed by Bailey et al. [ 8] and was more recently discussed by Gatzlaff and Ling [ 27]. The repe at sales estimatin g equation is ln(/)T itiT titit tPPcDE (4.16) where Pit/ PiT is the ratio of sales price for property i in time periods T and t ; Dit is a dummy variable which equals 1 at time of initial sale, +1 at the time of second sale, and 0 otherwise; ct is the logarithm of the cumulative price index in period t To clarify ct = ln(1+ A )t, where At is the cumulative appreciation rate for year t [ 59]. The implicit assumption in the repeat sales approach is that the quality of these houses has remained constant over time.
76 The repeat sales model can be expanded to test for differences in cumulative appreciation between the market and a submarket. Following Archer et al. [ 5] the repeat sales equation is extended from Equation 4.16 as ln(/) ''TT itiT tit titit ttPPcDcDE (4.17) where Dit is a dummy variable which equals 1 at the time of initial sale or +1 at the time of second sale if the property is in an area of (predicted) abnormal appreciation, and 0 otherwise. Now, ct is the logarithm of the cumulative price index in per iod t for the general market and tc is the logarithm of any additional (positive or negative) cumulative appreciation due to being in a particular submarket. Their study tested two submarkets: all houses within x miles to the north of the bridge and all houses within x miles to the south of the bridge. A spline regression technique was applied where multiple iterations of the model were run to estimate the distance effects of any abnormal appreciation [ 59]. The spline regression is a methodology which tests many distances of x (containing varying numbers of houses in the particular submarket) to determine the threshold distance at which the difference between the market and submarket is most evident. In their s tudy, Smersh and Smit h tested one tenth of a mile intervals. The optimum model, based on coefficient t statistics, converged on the distance that contained a minimum number of observations (to reach statistical significance) and captured the greatest difference (in appreciatio n) between market and submarket. Appreciation rates for Duval County, the area north of the Dames Point Bridge, and the area south of the Dames Point bridge were calculated for the period 1980 to 1990, a period encompassing the 1989 opening of the bridge a s well as the construction period. For the geographic submarket Dames Point North, their results suggest that these houses appreciated
77 nearly 8.7% more than the market appreciation of 52% for the 10year period. Much of the divergence occurs between 1985 a nd 1990, consistent with a period of anticipation of the opening of the bridge [ 59]. For the geographic submarket Dames Point South, their results suggest that these houses appreciated over 5% less than the market appreciation of 52% for the ten year perio d [ 59]. The studies discussed in this chapter provide the methods for creating repeat sale house price indexes. This methodology was created by Bailey, Muth and Nourse to capture house price movement using a regression technique. Case and Shiller expanded upon this process to address the issue of heteroskedasticity. Archer, Ling and Gatzlaff used the repeat sale house price index methodology to compare local appreciation rates to the larger area market appreciation rate in order to capture abnormal local appreciation rates. Smersh and Smith then used the Archer, Ling and Gatzlaff study as a foundation for their study, which captured the abnormal appreciation in the areas surrounding a newly construct ed bridge in Jacksonville, Florida.
78 CHAPTER 5 METHODS This chapter describes the application of a methodology which can be used to quantify project influence caused by the announcement of a governments public project using eminent domain. The conceptual model is based on a repeat sale house price index model. This type of price index is based on transaction prices of the same properties at different time periods. Using a repeat sales house price index model, house price appreciation can be quantified over a designated time period. By using this type of i ndex method the difficulties associated with other house price index methods can be avoided [ 13] [54 ]. These methods can be used to show abnormal appreciation in a designated submarket area by expanding the repeat sales index model to include a designated submarket. In order to quantify the movement in property prices in a specific area over a designated time period a house price index model must be constructed. This model will quantify price movements over time for those properties included. Constructing a market wide house price index is the first step to quantifying abnormal (positive or negative) appreciation caused by a governments public project. The following conceptual model is a repeat sale house price index. The origins of this index construction technique can be traced back to Bailey, Muth and Nourse [ 8] and Case and Shiller [ 12 ]. Let: '' t itt itt tD VE D or (5.1) ''' ittttittvdde
79 where lower case letters represent the natural logarithms of the corresponding capital letters. Vitt is the price relative (ratio) of the second sale price in period t to the initial sales price in period t for the i th pair of transactions with first and second sales in these two periods. Dt and Dt are the true but unknown indexes for period t and t where t cannot be the last year of the index and t cannot be the first year of the index (i.e., t = 0,1, T 1; and t = 1,2, T ). It is assumed that the residuals in log form eitt 2, and are not correlated with each other. Estimation of the unknown D s may be found using linear regression. Let: '' 1 T itt jjitt jvdxe (5.2) where xt is a dummy variable which equals 1 if period t is the pe riod of initial sale, +1 if period t is the period of final sale, and 0 otherwise for each pair of house transactions. In this model, dt represents the logarithm of the cumulative price index for year t ; dt = ln(1+A )t, where At is the cumulative appreciati on rate for year t In order to normalize the index let D0 equal 1 or d0 equal 0. In matrix form Equat ion (5.2) becomes: vdxe (5.3) where v and e are ndimensional column vectors. In this instance n = '' ,',tttt ttnn is equal to the number of pairs of house transactions with initial sales in period t and final sales in period t In matrix form d is a T dimensional column vector of unknown logarithms of the index numbers to be estimated; and x is an n by T matrix. For a pair of transactions whose first sale is not the base period, i.e., t = 1, 2, T 1, the corresponding row of x has a 1 in the t th column; and for
80 that same row has a +1 in the t th column; all other columns of that row are 0. Given all e s have a mean of zero, the same variances, and are uncorrelated with each other, the ordinary least squares estimator: xr d xx (5.4) is the best linear unbiased estimator of d. The t th diagonal element of the numerator, a T by T matrix ( xx), is the sum of the number of pairs of house transactions with initial sale in period t plus the number of pairs of house transactions with final sale in period t The t t th off diagonal element of ( xx) is ntT. The denominator is a T dimensiona l vector whose t th element is the sum of all price relatives for which period t is the period of final sale minus the sum of all price relatives for which period t is the initial sale. Case and Shiller [ 12] argue that the variance in eitt in Equation (5.2) increases as the length of time between a propertys two sales increases. Therefore, this study will test for heteroskedasticity. The methods for this test are based on the methods performed by Case and Shiller . They explain that their Weighted Rep eat Sales (WRS) method assumes that the log price, vit, of the i th house at time t is ittititvcHN (5.5) where ct is the log of the citywide level of housing prices at time t Hit is a Gaussian random walk variable representing the drift in individual housing value through time, and Nit represents the noise in price due to imperfections in the market for housing. In the first step, the repeat sale house price index procedure is followed exactly as above. In the second step, the squared re siduals from the regression in step one are regressed on a
81 constant and the time interval between sales. The constant term is the estimate of 2 N and the slope term is the estimate of 2 H In the third step, 2 N is tested for statistical significance. If the estimate for 2 N is not statistically significant at a specified confidence level there is no evidence of heteroskedasticity. A market wide index can be construct ed using the method described above. However, in order to quantify project influence a sub market index must be created to test for abnormal (positive or negative) appreciation in the geographic submarket affected by the announcement of the governments public project. The repeat sales model can be expanded to test for differences in cumulative appreciation between the market and submarket. Following Archer, Gatzlaff and Ling. [ 5] the price of an individual property may be represented as ijttijtijtPDDH (5.6) w here Pijt is the unit price of the i th residence in submarket j in time period t ; Dt is the unit price index for the market area as a whole during time period t ; ijtD is the unit price index for submarket j rel ative to the market wide price index, in period t ; and Hijt is the unit price level for the i th house in submarket j relative to the submarket price level, in period t In this instance, jtD is the amount the market wide index a t time t must be changed to adjust for the location of the residence in submarket j ; and Rijt represents the portion of unit price that is not explained by the market wide index and the location of the property in a particular submarket. Rijt is a residual scale factor that captures the individual houses peculiar price change. Appreciation over any k number of previous periods is determined by the ratio of prices:
82 () ()() ijtktkjtkijtk ijt tjtijtPDDH PDDH (5.7) Converting Equation 5.7 to natural logs gives () '' () ())lnlnlnlnlnlnlnijtk tkt jtk jt ijtkijt ijtP DDDDRR P (5.8) This equation is similar to Equation 5.1 above. However, it is generalized to account for abnormal appreciation in the submarket; which is defined as a deviation in the appreciation rate of the submarket from the market wide appreciat ion rate. In this study the submarket is the area affected by the influence of the public project. Equation 5.8 can be rewritten to: '' '' 11 TT itt jjjjitt jjvdxdxe (5.9) which is also similar to Equation 5.1. However, the difference between Equation 5.1 a nd Equation 5.9 is due to the addition of a submarket index in Equation 5.9. In this expanded version x is a dummy variable which equals 1 at the time of the initial sale and +1 at the time of the second sale if the property is located in an area of (predicted) abnormal appreciation, and 0 otherwise. In this instance, dt is the logarithm of the cumulative price index in period t for the general market and td is the logarithm of any additional (positive or negative) cumulative appre ciation due to being in the particular submarket. The model will test one submarket: the area most affected by the use of eminent domain. This area will differ on a case by case basis. However, in order for a municipality to use the power of eminent domain there must be a plan drafted before properties can be assembled [ 36].
83 In this plan the area that is scheduled to be condemned is specified before the public project begins [ 6]. Equation 5.8 is estimated in two steps. In the first step the market wide coef ficients, dt are estimated for a designated area including the submarket and the surrounding areas. In step 2 the residuals from step one are regressed on the time dummies, x for the designated submarket. Sales data was obtained for residential properties (1 4 family houses) in 11 towns in Monmouth County, New Jersey, from the Monmouth County Taxation Board from 1997 to 2008. Sales not considered arms length transactions were removed from the sample. There were 11,429 arms length transactions collected in these 11 towns from 1997 to 2008. Of the 11,429 arms length transactions collected there were 2,367 repeat transactions collected which represented either an initial or a final sale in a price relative. These 2,367 repeat sales represented 20.7% of the total population of arms length transactions for the 11 towns. There were 1,233 price relatives constructed from the repeat sales data. Of the 1,233 price relatives constructed, 312 were located within the City of Asbury Park representing 25.3% of all price relatives. Of the 312 price relatives located within Asbury Park, 148 were located within the Waterfront Redevelopment Area representing 12% of all price relatives constructed and 47.4% of all price relatives from Asbury Park. The data collected included property information on the location, property class, assessed value, sale date, sale price, buyer name, seller name, and the location of the property by book and page number in the Monmouth County tax records. Figure 51 is a picture of Asbury Park (in ye llow) and the 10 towns surrounding Asbury Park used for the market wide index (outlined in red).
84 Figure 5 1. A map of the eleven towns in Monmouth County used for the construction of the market wide index. Asbury Park is highlighted in yellow, and the surrounding towns are out lined in red. Source: Monmouth County Planning Board, Monmouth, County, 1997. The first step in quantifying abnormal appreciation known as project influence is to construct a market wide price index. The market wide index includes price relatives from repeat sales in Asbury Park and the 10 surrounding towns. This market wide index represents the true or normal cumulative appreciation for the area between the years 1997 and 2008. This market index is compared to the submarket index over these 13 years to capture the difference in price movement during this period.
85 One property ( i ) represents one row of the matrix used to estimate the index. This can be written 12 1122334455667788 2 1 99101011111212 1lnii jdxdxdxdxdxdxdxdx P e dxdxdxdx P (5.10) where ln( P2/ P1) represents vitt, the pr ice relative for property i P2 is the final sale price for property i and P1 is the initial sale price for property i ; the ds are the estimates from the regression. They represent the logarithm of the cumulative price index for a specific year. The inde x constructed in this particular study has 12 estimates, one for every year. When estimated they are in log form and must be anti logged in order to create the price index. To start the index at 1.00 the first estimate, d1, is set equal to zero. When d1=0 is raised to the exponential function it becomes equal to 1.00, (i.e. 1d e = 01.00 e ) which is the first number in the price index. Figure 5 2 is an example of a matrix containing nine properties used to create an index from 1995 until 2003. The column on the left is each propertys price relative. Each row after the title row represents the years of the initial and final sales. For example, property one initially sold in 1996 and again in 1999; property two initially sold in 1997 and again in 2001, etc. The index is constructed by regressing column one, the column of price relatives, on each yearly column except the first (1995). Again, this is done so that the index starts at 1.00. Using these methods and the price r elatives collected, a market wide index is created for the Monmouth County region. The second step in quantifying abnormal appreciation known as project influence is to create a sub market price index for the area affected by the use of eminent domain. In this study the area affected by the use of eminent domain was the Waterfront Redevelopment Area of Asbury Park. This area of Asbury Park was designated to be assembled by the developer Asbury
86 Partners not long after the Memorandum of Understanding was sign ed (August, 2001). The signing of the Memorandum of Understanding was considered to be the date at which the city committed to the public project. 1 2 3 4 5 6 7 8 995'96'97'98'99'00'01'02'03' 010010000 001000100 000100010 000010100 010000001 100010000 000001100 000000011 000100001 relative v v v v v v v v v Figure 5 2. An example of a repeat sales matrix used to create an index The proc ess of creating a sub market index is similar to creating a marketwide index. First, the submarket area must be specified. Then, all arms length repeat sale transactions must be collected for that area. Finally, price relatives must be created for all re peat sales and a matrix of time dummies must be created. The submarket index created for Asbury Park also begins in 1997 and ends in 2008. This submarket index is then compared to the market wide index to determine if any abnormal appreciation has occur red between 1997 and 2008.
87 CHAPTER 6 RESULTS AND CONCLUSION The objective of the study was to develop a methodology to quantify project influence caused by the governments use of eminent domain for public projects. Repeat sales house price indexes were constructed for a market wide area and a submarket area. In the market wide area there were 1,233 price relatives created using 2,367 arms length residential house transactions located in 11 towns in Monmouth County, New Jersey. A submarket index was created using 148 price relatives from properties located within the Waterfront Redevelopment Area (WRA) of Asbury Park. An empirical model used to create the indexes included the property price relatives as the dependent variable and binary variables as the independent variables. These indexes are constructed yearly from 1997 til l 2008. The date of the governments commitment to the Waterfront Redevelopment Project wa s August 2001. However, it is necessary to compare the abnormal appreciation rate in the WR A both before and after the influence of the public project. By comparing the abnormal appreciation before to the abnormal appreciation after, this study can quantify the project influence in the area. Appraisal and law literature suggests that project in fluence causes areas affected to appreciate or depreciate due to the threat of eminent domain. Therefore, by comparing the appreciation rate of an area threatened by eminent domain to a market wide appreciation rate one can test for abnormal appreciation. Any difference between the market wide index and the submarket index between 2002 and 2006 is considered to be caused by the influenc e of the redevelopment project. These years are chosen because this is the time when Asbury Partners was assembling propert ies in the area.
88 Results Appreciation rates for Monmouth County and the Waterfront Redevelopment Area in Asbury Park, New Jersey were calculated for the period 1997 to 2008. This time period includes the August 2001 signing of a Memorandum of Understanding between the City of Asbury Park and its primary developer Asbury Partners. The signing of the memorandum in late 2001 committed the City to a large redevelopment project which involved assembling properties using eminent domain. The application of this me thodology allows an examination of both the magnitude of the difference in appreciation rates and the period of time at which change became apparent [ 59]. Table 6 1. Regression Results of Market Area Year Estimate Std. Err. t P>t Index 1997 0 1.00 19 98 0.0789003 0.0260867 3.02 0.003 1.082096 1999 0.1520559 0.0250457 6.07 0 1.164225 2000 0.3058472 0.0244328 12.52 0 1.357775 2001 0.5182998 0.0234202 22.13 0 1.67917 2002 0.6062958 0.0233778 25.93 0 1.833627 2003 0.8285716 0.0245175 33.8 0 2.29 0045 2004 1.016757 0.0251939 40.36 0 2.764216 2005 1.249689 0.0281825 44.34 0 3.489258 2006 1.341431 0.0306171 43.81 0 3.824512 2007 1.281649 0.0341155 37.57 0 3.602575 2008 1.275221 0.0366428 34.8 0 3.579492 Market area regression results are presented in Table 6 1. Regression results and the index created using these results begin in 1997 with the first year normalized at 1.00. Determining appreciation rates from index numbers requires simple mathematics. The difference in index numbers betwee n any year and the base year represents the appreciation rate over that time
89 period, i.e. from 1997 to 2002 there was 83.3% positive appreciation (1.833 1.00). However, if the base year isnt the beginning year the two index numbers must be normalized fir st, i.e. from 2002 to 2006 there was 109% positive appreciation (2.0861.00). To normalize, divide both index numbers by 1.833. T tests were performed to test whether these estimates are statistically significant at the 5% significance level. The null hypo thesis is that each estimate is equal to zero. H0: dn= 0 vs. H1: dn n, where n = 2, 3, 14) (6 1) The two sided alternative hypothesis is that each estimate is statistically different than zero. Thirteen separate tests were performed. The results show that all of the estimates from 1998 to 2008 are statistically different than zero at the 5% significance level. The regression had an adjusted R2 equal to .76, meaning 76% of the sample variation in the observations is explained by ( or predicted by) the 11 regressors. A submarket index for the redevelopme nt area was created u sing the estimates from Table 62. By taking the antilog of the submarket estimates in Table 62 and multiplying them by the anti log of the market estimates in Table 6 1 a new index is constructed. This is the Waterfront Redevelopment Area submarket index (another method to create the same submarket index is to add the estimates in Table 6 1 to the estimates in Table 6 2, then take the anti log of these sums either method produces the same submarket index ). The ratio of the WRA submarket index to t he m arket wide index is equal to jtd The last column in Table 62 is the submarket index for the Waterfront Redevelopment Area in Asbury Park, New Jersey created using the methods developed by A rcher, Gatzlaff, and Lin g [ 5].
90 Tabl e 6 2. Regression Results of Submarket Area (WRA) Year Estimate Std. Err. t P>t Index 1997 0 1.00 1998 0.033731 0.1213464 0.28 0.781 1.046205 1999 0.093438 0.1115481 0.84 0.404 1.06037 2000 0.0496833 0.1113592 0.45 0.656 1.291964 2001 0.1564624 0.1032344 1.52 0.132 1.963566 2002 0.075314 0.1417332 0.53 0.596 1.977058 2003 0.21766 0.145474 1.5 0.137 2.846903 2004 0.2781025 0.1445369 1.92 0.056 3.650483 2005 0.232888 0.1460512 1.59 0.113 4.404281 2006 0.2462478 0.1467306 1.68 0.096 4.89238 2007 0.1365828 0.1586144 0.86 0.391 4.129812 2008 0.0544083 0.1677073 0.32 0.746 3.779642 An F test with 11 restrictions was performed to see if the estimates in Table 6 2 are jointly equal to zero. The null hypothesis was that the index l evel of the submarket area is equal to the index level of the market area in each year. That is, jtd = 0 for all time periods. H0: ''' 230,0,,0kddd vs. H1: '0jd at least one j,j = 1,,k (6 2) The results show that the null hypothesis is rejected at the 95% significance level, finding that the market index level and the submarket index level are statistically different from each other each year. Figure 6 1 is a diagram showing the results of the two indexe s. Movement in the index from period to period represents appreciation (positive and negative). This diagram shows that from 1997 to 2002, the Market wide index and the Waterfront Redevelopment Area index moved together in relative unison. However, from 2002 to 2006 the two indexes diverge and the Waterfront Redevelopment Area ap preciates more than the market. This was the period of time when Asbury Partners was assembling property in the WRA.
91 These two indexes capture appreciation rates in the market and su bmarket over time. From 1997 to 2002 the Waterfront Redevelopment Area appreciated 7.8% more than the market appreciation rate of 83.3% for the 5year holding period. This amounts to a 1.6% higher appreciation rate per year from 1997 to 2002. From 2002 to 2006 the Waterfront Redevelopment Area appreciated 20% more than the market appreciation rate of 109% for that 4year holding period. This amounts to a 5% higher appreciation rate per year from 2002 to 2006. Figure 6 1. A Comparison of the Market wide index and Waterfront Redevelopment Area index. The results indicate that before the City of Asbury Park signed the Memorandum of Understanding the Waterfront Redevelopment area appreciated 1.6% more than the market every year. Ho wever, after the MOU was signed the redevelopment area appreciated 5% more than the
92 market every year until 2006. By 2006 the redeveloper had assembled all of the properties it needed in the WRA. The finalization of the assemblage and the downturn of the housing market coincide. This can be seen in the narrowing of the two indexes from 2006 to 2008. The indexes show a slight separation beginning in 2000. This may be caused by the markets anticipation for the redevelopment. This wa s due to news articles abo ut Asbury Partners purchasing all of the old developers liens in the City of Asbury Park  However, the city was not committed to the public project until August of 2001. Table 6 3. Yearly and Cumulative Abnormal Appreciation Time Period Abnormal App reciation Year Cumulative Abnormal Appreciation since 1997 97' 98' 3.32% 1998 3.32% 98' 99' 5.60% 1999 8.92% 99' 00' 4.07% 2000 4.85% 00' 01' 21.78% 2001 16.94% 01' 02' 9.11% 2002 7.82% 02' 03' 16.49% 2003 24.32% 03' 04' 7.75% 2004 32.06% 04' 05' 5.84% 2005 26.22% 05' 06' 1.70% 2006 27.92% 06' 07' 13.29% 2007 14.63% 07' 08' 9.04% 2008 5.59% Applying the repeat sales house price indexes methodology was a way to quantify abnormal appreciation in locationspecific areas. Table 6 3 present s the yearly and cumulative abnormal appreciation for the submarket relative to the market wide area. The numbers in the first two columns represent how much more or less the WRA appreciated over the market as a whole over a given year. For example, from 1997 to 1998 the redevelopment area appreciated 3.32% less than the market as a whole, and from 1998 to 1999 the WRA appreciated 5.6% less than the market. The numbers in the second two columns represent the cumulative abnormal
93 appreciation of the WRA over the market from 1997 to a given year. For example, from 1997 to 2002 the WRA appreciated 7.8% more than the market as a whole. These percentages do not represent appreciation or depreciation, but a comparison of market to the submarket appreciation rates o ver time. Conclusion The scope of the project rule states that the effects of a proposed project cannot be considered in valuing a property to be acquired for the project when it was clear that the parcel under appraisal would be acquired for the project. [ 22]. The two most important aspects of applying the scope of the project rule are determining when the government was committed to the project and whether it was probable that the appraised parcel would be taken for the project [ 22]. If the property was within the scope of the project, then the date at which the government committed to the project is the date of valuation. For compensation to be considered just, any project influence, whether positive or negative, accumulating from the date the government committed to the project to the condemnation valuation date must be separated from the propertys market value. For this study applying the scope of the project rule begins by determining when the government was committed to the project. This date is dete rmined to be when the Memorandum of Understanding was signed between the City of Asbury Park and the primary developer Asbury Partners, in August of 2001. The next step in applying the scope of the project rule is determining whether it was probable that the appraised parcel would be taken for the project. For this study the properties needed for assembly were designated in a specific location known as the Waterfront Redevelopment Area. The main objective of this study was to apply a methodology to quantify project influence which could be used by appraisers and the courts confronted by this problem. The results show
94 that from 1997 to 2002 the Waterfront Redevelopment Area of Asbury Park appreciated 7.8% more than the overall market. However, from 2002 to 2006 the Waterfront Redevelopment Area appreciated 20% more than the overall market. The difference between these two is 12.2%. This means from 2002 to 2006 there was an average of 3% more abnormal appreciation in the WRA than before the MOU was signed. By l aw, when appraising a condemned property, any abnormal appreciation caused by the redevelopment project must be separated from normal market appreciation. Any property located within the waterfront redevelopment area condemned between 2002 and 2006 must be discounted in order to determine its unaffected fair market value. A way to discount a property would be to multiply it by, 1 1.03n where n represents the number of years between 2002 and the year the property was condemned. The results of this study could be useful to property appraisers and the courts. It could aid property appraisers attempting to value condemned property. An appraiser may find this methodology useful when attempting to quantify project influence. If project influence is discovered, this methodology may provide a useful way for an appraiser to make a time factor adjustment to a condemned property in order to determine a more accurate market value. The courts could also benefit from this study. This empirical model coul d be used by the courts to help in their final decision of just compensation. Potential Problems A test for heteroskedasticity was performed based on a procedure created by Case and Shiller in 1989. Case and Shiller argued that the house specific component of the change in log price may not be homoskedastic. They believed that the variance of the error term increased as the time interval between sales increased. The residuals for the price relative observations were regressed on a constant and the time int erval between the observations. A significance test was
95 performed at a 95% significance level. The results of this test showed that there was no sign of heteroskedastici ty in the Monmouth County data. A test for sample selection bias was also performed. Ga tzlaff and Haurin indicated in 1994 that using only information from a sample of sold homes to estimate value movements for the entire housing stock may be subject to substantial bias [ 26]. Those properties selling more than once may not be a good representation of the total stock of properties sold. To test for sample selection bias the median sale price of repeat and non repeat homes are compared yearly. The results show that the homes selling repeatedly are on average 9% less than those selling once betw een 1995 and 2006. However, from 1995 until 2001 the median sale price of repeat and nonrepeat sale properties is nearly the same. The difference between the total stock of sold properties and those selling more than once becomes evident after 2001. Figur e 6 2 is a graph of the median sales price for repeat and non repeat sale properties per year. The results show that from 1997 till 2001 the two are very similar. However, between 2001 and 2006 the two diverge and the median sales price of nonrepeat trans actions are on average 16% higher than repeat transactions. Clapp and Giacotto explain some of the reasons for this, there are some plausible sources for sample selectivity in the repeat sales method. First, properties are bought to be repaired for resale ; second, there may be a lemons phenomenon in which properties that do not meet buyers expectations are repeatedly sold; and third, starter homes sell repeatedly as owners move to better housing. [ 18] This result shows that the repeat sales index may not be a good representation of the population of all properties sold from 2001 to 2006. These results may be understating the project influence because of lower prices paid for homes with more than one sale.
96 Figure 6 2. Median Sale Price of Repeat and N onRepeat Sales per Year The submarket index created in this study may have been influenced by sampling bias. The reason for this is due to the small submarket sample size. The sample of price relatives used to create the submarket index had 148 observati ons. These 148 observations were regressed on 11 estimates ranging from 1998 to 2008. While this would usually average to 15 observations per year, there were less observations in the beginning and ending years. Most of these price relative observations oc cur between 2002 and 2006. An ideal case would be for at least 30 observations per year ffrom 1997 to 2008 for a total of 3 30 price relative observations. This was not accomplished in this study. Therefore, a price relative outlier could be more influentia l in the beginning and ending years of this study biasing the index.
97 Further Research In order to the test the effectiveness of this study further research must be done using the repeat sales methodology described in this study. Arms length transaction s fo r areas affected by the use of eminent domain need to be collected and this model needs to be applied to that data. By doing so the methodology used in this study can be tested on other areas. The use of the repeat sales index method may be one way to capt ure abnormal appreciation in a specific area. The assessed value house price index method and the hedonic house price index method are two alternatives to the repeat sales method. By using the assessed value index method, price relative observations could be more than doubled. This could be useful for submarket areas that have very little repeatsales data. While the hedonic method requires significantly more data than either the repeat sale and assessed value methods, it may be a more accurate index metho d. The argument can be made that the appreciation of a submarket is influenced by many factors that contribute to a propertys price. For instance, the appreciation in a submarket may be due, in part, to the introduction of a new casino just outside of the submarket. Using the repeat sales (and even the assessed value method) to capture the submarkets appreciation cannot capture the appreciation due to the casino, and the appreciation due to the threat of eminent domain at the same time. This is one weakne ss of the repeat sales method for co nstructing house price indexes. Therefore, further research should be done utilizing the assessed value and hedonic index methods with the Monmouth County and Asbury Park data in order to quantify abnormal appreciation i n this submarket. By doing so each method could be compared to find the most effective way to capture abnormal appreciation in the Waterfront Redevelopment Area.
98 REFERENCES 1. The Appraisal Foundation, Uniform Standards of Professional Appraisal Practice The Appraisal Foundation, Chicago ( 1994). 2. The Appraisal Institute, The Appraisal of Real Estate The Appraisal Institute, Chicago, (2008). 3. The Appraisal Institute, The Dictionary of Real Estate Appraisal, The Appraisal Institute, Chicago (2002). 4. The Ap praisal Quality Board, Uniform Appraisal Standards for Federal Land Acquisitions, U.S. Government Printing Office, Washington (1992). 5. W. R Archer, D. H. Gatzlaff, D. C. Ling, Measuring the Importance of Location in House Price Appreciation, Journal of Ur ban Economics, 40, pp. 334353, (1996). 6. Asbury Park Planning Board, Waterfront Redevelopment Plan, pp.3, 15 March 2002. 7. S. Ayres Asbury Park, New Jersey (Postcard History Series), Arcadia Publishing, Charleston, South Carolina, (2005). 8. M. J. Bailey, R. F. Muth and H. O. Nourse, A Regression Method for Real Estate Index Construction, Journal of the American Statistical Association, 58, pp. 933942, (1963). 9. R. Barlowe, Land Resource Economics, Prentice Hall, Englewood Cliffs, New Jersey, (1978). 10. Berman v. Parker 348 U.S. 26 (1954). 11. K. E. Case and C. J. Mayer, Housing price dynamics within a metropolitan area, unpublished, ( 1995) 12. K. E. Case and R. J. Shiller, Prices of Single Family Homes since 1970: New Indexes for Four Cities, New England Economic Rev iew, September/October, pp. 4556, (1987). 13. K. E. Case and R. J. Shiller, The Efficiency of the Market for Single Family Homes, The American Economic Review, 79, pp. 125137, (1989). 14. The Castle Coalition, 50 State Report Card Tracking Eminent Domain Reform Legislation since Kelo, The Castle Coalition, August ( 2007) 15. The Castle Coalition, Flood gates open tax hungry governments & landhungry developers rejoice in green l ight from U.S. Supreme Court Citizens Fighting Eminent Domain Abuse, Online, http://www.castlecoalition.org/index.php?option=com_content&task=view&id=44&Item id=144, 24 June ( 2009)
99 16. C Chesler A mid economic slowdown, a legendary s hore town awaits the fruits of redevelopment New Jersey Monthly, published 5 May 2005, Online, http://njmonth ly.com/articles/jerseyshore/greetings from asbury park again.html 27 August ( 2009 ) 17. Chicago B. & Q. R.R. v. Chicago, 166 U.S. 226, (1897). 18. J. M. Clapp and C. Giacotto, Estimating Price Indices for Residential Property: A Comparison of Repeat Sales and As sessed Value Methods, Journal of the American Statistical Association, 87, 300306, (1991). 19. J. M. Clapp, C. Giacotto and D. Tirtiroglu, Housing Price Indices Based on All Transactions Compared to Repeat Subsamples, AREUEA, 19, pp. 270285. (1991). 20. K. Darde n, Eminent Domain in the Aftermath of Kelo: A Research Guide Spring 2007, The Georgia State Law Library, 08 August ( 2009) 21. Duke University School of Law, Urban Renewal: Acquisition of Redevelopment Property by Eminent Domain, Duke Law Journal, 1, pp. 123138, (1964). 22. J. D. Eaton, Real Estate Valuation in Litigation, Appraisal Institute, Chicago, ( 1995) 23. I. Fisher, The Making of Index Numbers A Study of Their Varieties, Tests, and Reliability, Augustus M. Kelly Publishers, New York, (1967). 24. M. Fujita, A Rational Expec tations Model of Urban Growth and Land Markets, Journal of Real Estate Finance and Economics, 4, pp. 225265, (1991). 25. B A. Garner editor Black's Law Dictionary West Group, Minneapolis Minnesota, (2004). 26. D. H. Gatzlaff and D. R. Hauri n, Sample Selection Biases in Local House Value Indices, Journal of Urban Economics, 43, pp. 199222, (1998). 27. D. G. Gatzlaff and D. C. Ling, Measuring Changes in Local House Prices: An empirical Investigation of Alternate Methodologies, Journal of Urban Ec onomics, 35 pp. 221244, (1994). 28. D. W. Glaves, Date of Valuation of Eminent Domain: Irreverence for Unconstitutional Practice, The University of Chicago Law Review 37, pp. 319357, (1969) 29. A. C. Goodman, Hedonic Prices, Price Indices, and Housing Markets Journal of Urban Economics, 10, pp. 471484, (1978). 30. J M. Hartwick, N. Olewiler The Economics of Natural Resource Use, AddisonWesley Educational Publishers New York, N.Y., (1985).
100 31. D. R. Haurin and P. H. Hendershot, House Price Indexes: Issues and Res ults, AREUEA Journal, 19, pp. 259269, (1991). 32. Hawaii Housing Authority v. Midkiff 467 U.S. 229 (1948). 33. Hays v. Port of Seattle 251 U.S. 233 (1920). 34. Heyer, Gruel & Associates Asbury Park Master Plan Community Profile, May ( 2006) 35. A. J. Hosios and J. E. Pesando, Measuring Prices in Resale Housing Markets in Canada: Evidence and Implications, Journal of Housing Economics, 1, pp. 115, (1991). 36. Kelo v. The City of New London, 545 U.S. 469, (2005) 37. L. A. Kochin and R. W. Parks, Vertical Equity in Real Estate Assessment: A Fair Appraisal, Economic Inquiry, 20, pp. 511531, (1982). 38. J E. Krier and C. Serkin, Public Ruses, Volume 2004, Issue 4 Michigan St ate Law Review, (2004). 39. J.J. Lazzarotti, Public Use or Public Abuse, University of Missouri at Kansas City Law Review, 68, 4975 (1999). 40. M. Leitner and B.J. Schumacher, Spatial crime displacement resulting from large scale urban renewal programs in the city of Baltimore, MD: a GIS modeling approach presented at the IV International Conference on GeoComputation, 2528 July ( 1999). 41. L aura M ansnerus, At the Shore ; 15 Years Later, Rebuilding Asbury Park, New York Times, published 5 Dec. 1999, Online, http://www.nytimes.com/1999/12/05/nyregion/at the shore 15 years later rebuilding asbury park.html?pagewanted=1 12 June ( 2009) 42. G. D Mathues, Note, shadow of a bulldozer? : RLUIPA and eminent domain after kelo, 81 Notre Dame Law Review, pp. 16531662 (2006). 43. Monmouth County Planning Board and STV Incorporated, Asbury Park Transportation Improvement Study, pp. I 2, September 2005. 44. Elmer A. Mullen, Consultant, An Analysis of Condemnation, The American Association of Certified Appraisers, Inc., Cincinnati, ( 1980). 45. New Jersey Exposition Authority v. Giant Realty Associates 143 N.J. Super. 338 ( 1976). 46. New Jersey State Govern ment, New Jersey Local and Redevelopment Housing Law 47. New Jersey Government, New Jersey State Constitution, Article VIII, Section III 48. New Jersey Statute Authority, Eminent Domain Statute 20:3 20 49. Nichols, Eminent Domain, Sackman & Van Brunt, Chicago, Il. ( 1950 )
101 50. Nichols, Extent of Judicial Inquiry into Power of Eminent Domain, 28, Southern California Law Review (1955). 51. W. Nicholson, Microeconomic Theory Basic Principles and Extensions, SouthWestern, Indiana, ( 2005 ) 52. H. O. Nourse, The Effects of Public Housing on Property Values in St. Louis, Land Economics, 39, pp. 433441, (1963). 53. L. Orgul, The Law of Eminent Domain, Columbia University Council for Research in the Social Sciences, Charlottesville, North Carolina, ( 1953) 54. R. B. Palmquist, Alternative Techniques for Developing Real Estate Price Indexes, Review of Economics and Statistics, 62, pp. 442480, (1980). 55. J. Poterba, House Price Dynamics: The Role of Tax Policy and Demography, Brookings Papers on Economic Activity, vol. 2, pp. 143203, (1991). 56. M B. Rachlis and A. M. J. Yezer, Urban Location and housing price appreciation, Papers of the Regional Science Association, 57, pp. 155164, (1985). 57. A. Regaldo, The Presence of Wetlands and their Effects on Rural Land Values, University of Florida Thesis, (1995). 58. W.P. RokesAn Analysis of Property Valuation Systems Under Eminent Domain, Bureau of Business and Economic Research, School of Business Administration, Montana State University (1961). 59. G. T. Smersh and M. T. Smith, Accessi bility Changes and Urban House Price Appreciation: A Constrained Optimization Approach to Determining Distance Effects, Journal of Housing Economics, 9, pp. 187196, (2000). 60. S.Z. Searles, A Practical Guide to Condemnation, American Law Institute, Philadel phia,Pa., (1969). 61. Staff Report, How a New Developer Took Over and Redevelopment Facts, Asbury Park Press, written 23 January 2005, online, http://www.app.com/apps/pbcs.dll/article?AID=/20050123/NEWS/70827015, read 27 January ( 2009 ) 62. Township of West Windsor v. Nierenberg 150 N.J. 111, (1997). 63. United States v. Miller 317 U.S. 369, (1942). 64. U.S. Government, The Uniform Relocation Assistance and Real Property Acquisition Policy Act of 1970, Public Law 91646, .11(c)(1). 65. United States Government, U.S. Const itution, Amendment V.
102 66. H. R. Varian, Intermediate Microeconomics, W.W. Norton & Company, New York, N.Y., (2003). 67. R. Wenzlick, As I See the Fluctuations in the Selling Prices of Single Family Residences, The Real Estate Analyst, 21, pp. 541548, (1952). 68. H. W yngarden, An Index of Local Real Estate Prices, Michigan Business Studies, Volume 1, Number 2, Ann Arbor, University of Michigan, (1927).
103 BIOGRAPHICAL SKETCH Nathan Van Steenbergen was born on September 13th 1983, in Douglasville, Georgia. In 1985, Natha n and his family settled in Tampa, Florida. Nathan received all of his primary and secondary education in Tampa. Upon graduation from Sickles High School, Nathan went to Florida State University in Tallahassee. His interest in the study of economics prompt ed him to pursue a Bachelor of Science in t raditional economics, with a specialization in applied, international and monetary economics. He graduated in the spring of 2007. In the fall of 2007, Nathan began the applied economics program at the University of Florida in Gainesville. His interest in the field of real estate led him to work on a thesis related to property valuation. He graduate d in the fall of 200 9 with a Master of Science and graduate minor of real e state. Upon the completion of his graduate studies, Nathan seeks to become an appraiser in Tampa. He will begin working as a trainee appraiser at CB Richard Ellis in September 2009. He hopes to receive an MAI appraisal designation from the Appraisal Institute.