Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
ITEM IV-A
crrY of Memorandum * WASHINGTgN Planning and Development Departrnent To: Downtown Redevelopment Committee From: Elizabeth Chamberlain, Planning Manager CC: Kevin Snyder, Interim Planning & Development Director Daryl Faber, Parks Arts & Recreation Director Date: June 4, 2010 Re: Economic Benefits of Parks and Open Space Included in your packet are a couple of journal articles written by John Crompton, a professor at Texas A&M University, who is a leader in parks and open space and has studied the economic benefits of these facilities within jurisdictions. Professor Crompton recently was the key note speaker at the Washington Recreation and Park Association conference and his presentation discussed the economic benefits of parks and open space. City Parks Arts & Recreation staff attended this conference. Staff has brought these forward to have a dialogue at Tuesday's Downtown Redevelopment Committee meeting about downtown park and open space. A~TBURN ~k MORE THAN YOU IMAGINED i Managing Leisure ~0, 203-218 (4ctober 2005} ~ RoUt~edge Taylor&Fran~isCraup Y i The im act of arks on ro ert value: ~ A ~ p y ' ' ' n fr m the ast two em ~r~cal evade ce o p ~ decades in the United States John L. Crompton Depar~men~ of Recrea~ron, Park and Tourism Sciences, Texas A&~! Urrrverslty, TX, USA The notion that parks have a positive impact on proximate property values was recognized in the debates surrounding the pioneering of large urban parks in England in the first half of the nineteenth century, and subsequently in the spread of this movement to the US in the latter half of that century. The empiMCal basis for these early assertions was rudimentary and nave. This paper reviews con- temporaryresearch using the more advanced analytical procedures now available to social scientists that has examined this issue. The findings confirm the initial rationale and suggest that a positive impact of 20 % on property values abutting or fronting a passive park is a reasonable starting point guideline for estimating such a park's impact. ry 'Y [NTRODUCT[UN investment of tax funds in the world's first The premise that parks have a positive publicly funded park in Birkenhead ~ in ' impact on proximate property values 1847 (Crompton, 2004}. derives fram the observation that people After touring Birkenhead Park in 1850, frequently are willing to pay a laxger Frederick Law Olmsted was responsible for amount of money for a home located transitioning both its picturesque design close to a park, than they are for a compar- principles and its proximate principle able home. In effect, this represents a `capi- funding rationale to Central Park in talization' of park land into increased New York City, and from there to urban property values of proximate Iand owners. park systems across the U5. Thus, from the The increments of enhanced value attribu- earliest days of urban park development in . _ tates fxom~ fhe 1S50s through table to a park were used to fund'' early the United S parks; just as such increments are used to the 1930s, there was an insistent, almost fund golf courses in community golf level- inviolate conviction among park, advocates opments in contemporary US society. The of the legitimacy of the proximate principle. premise of the proximate principle under- It was conventional wisdom among them, girded the earliest parks in England. It but it was also espoused by city planners was the central principle in ,lohn Nash's and elected officials. development of London's Regent's Park Olmsted and others undertook studies which was commenced in 1512; it was the that appeared to confirm the intuitive val- core rationale for Richard Vaughan Yates' idity of the proximate principle. Thus, for investment in Prince's Park, Liverpool, in example, his studies at Central Park con- 1542; and it provided the rationale for eluded that the annual debt charges incurred Managing ~ersure ISSN 2360-6719 print/ISSN 1466-~50x online ®2005 Taylor & Francis http:~~www.tandf . co.uk~j ournals v~ DQI:10.1080.~13606710500348060 ti 10 4~;.`0 1 S i 1 r ~ t 204 C Q p o ~"he ~ ~ b New York Cit for ac uisition and Bevel- Structural Attributes im c y y q E.g., numbers o~f bedrooms, bathrooms, p opment of the park were $534,000 annually, fireplaces, garages; square footage ofhause; lot Of th ~ size; age of structure; existence of pool the increase in property tax revenue sion received by the city as a result of the idenl enhanced value of ro erties around the. Neignbaurh°ad Attributes mod p p E,g., socio-economic characteristics of park amounted to $5.24 million annually; So neighboring residents; quality ofneighbaring structures; ownershiplrental; ethnic the net annual income accruing to the city composition from its investment in the park was $4.4 million (Fox, 1990}. Ca~nmunityAtfributes The This study and others like it were fairly E.g., school and tax districts Property Value of tl~ rudimentary and naive, reflecting the under- earl Locational Attributes developed nature of the statistical tools and E,.g,, proximity and accessibility to various grew research desi ns available at the time. All (dis)amenities including waste sites, power g lines, highways, shopping centers, churches, 1nCrf property value increases were attributed to schools, cultural opportunities, airport, public ~ ~ Earl transportation the proximity of a park and the potential of in PI other factors were ignored. As new tools A Environmental Attributes evolved, the quality of the studies investi- E.g., view from property, noise levels, pollution SL'V~ ]evc[s, stormwater gating this issue improved. and During the past couple of decades, there 19'l~ Time-Related Attributes have been three developments that have E.g,, month and year of sale, number of days vn Stud facilitated the emergence of studies which mar[cet In tl are much stronger from a science perspective a p that have addressed the impact of parks on Fig. ~ ~ Factors influencing property value. Source: eacl property values. First, the. increased sophisti- Nicholls (2002} 12 cation of hedonic analysis and the statistical regr tools associatedwith it have enabled the values. A GIS is a computer-based system iron array of other factors that may contribute to that stores and facilitates manipulation of vale changes in property values to be considered. geographic information. GIS.enables Multiple `thee These factors are identified in Figure 1.Listing Service data to bemapped by individ- to The second development was the evol- ual street addresses. It permits accurate cal- adv. ution in the ~980s of Multiple Listing Services culation of distances between locations such the in electronic form. In the US, it is standard as a house and a park and can delineate ~97~ practice for all real estate agents in a city areas within which affected properties are T1 to report the sales price of each transaction located. - _ ~ to t~ to a central data base that is accessible to These three developments have facilitated late all of them with details of the structural. the efficient use of hedonic price modelling. on ~ and physical attributes of the property. ,The theoretical foundation of hedonic it m This data base is called the Multiple Listing pricing techniques was laid down by ties Service. Now these data are available in elec- Lancaster X1966} who proposed that utility ject tropic form. They can be transposed on to was derived from the characteristics or attri~ and maps that are f ormulated as part of a city's butes of goods. For example, the character- to 1 geographic information system and spatially istics of a car from which utility is derived goo integrated with the location of parks. may include engine size, speed and acre- site Geographic information systems consti- Ieration, fuel economy, number of seats, may tute the third development that has comfort, luxury, colour, style and sfiatus. By 600 enhanced the quality of the science investi- estimating the contribution of each characM A gating the impact of parks on property teristic to the purchase decision, its~relative IysF i ~`he impact of parks an property values 205 importance can be identified. Thus, the role properties in the vicinity of Pennypack Park of the factors shown in Figure 1 in the deci- (Hammer et a! , 1974}. This 1,294-acre sion to pay a given price for a home can be stream-valley park is in northeast Philadel- identified through the use of regression phia. It was surrounded by residential areas models. developed at a density of approximately ten dwelling units per acre. The area around REVIEW QF `MQDERN' STUDIES` the park was comprised of `unimaginative housing, heavy inscale with natural ~and- The era of `modern' studies for the purposes scaping losing out to concrete and stone' of this review commenced in the 1970s and (p. 275}. Based on their subjective evaluation early 1980s, when the availability and of the area, the researchers hypothesized greater capacity of computing stimulated an that `the residents do not consider natural . . increased i nterest ~n investigating the issue. amenity to be very important so publYc Early studies in the 1970s were .conducted open space would be expected to have a rely in Philadelphia, and in Boulder, Colorado. Lively low effect on Iand values compared to~ A 1972 study in Philadelphia focused on other neighborhoods' (p. 275}. seven sites, at three parks, three schools, Despite the authors' pessimistic pragno- and one school-park combination .(Cyan, sis, regression analysis indicated that the 1972}. During the sample years of ,the park accounted for 33% of land value at 40 study, 1,725 property sales were recorded feet. This dropped to 9% at 1,000 feet .and in the neighbourhoods araund the sites. As 4.2% at 2,500 feet which was the peripheral a percentage of total housing units in limit set for the study. From these data, the each area, the sample size ranged from authors concluded that a net increase in 12 % to 25.5%. In all seven neighbourhoods real estate value of almost $3.4 million regression analyses indicated that distance X1974 values} was directly attributable to from the site had an impact on property the park. values, enabling the author to conclude, The most frequently cited study in the `there appear to be locational advantages literature of this era examined the effect of to school and park facilities, and these greenbelts on property values in three differ- advantages have been capitalized in ent areas of Boulder, Colorado (Correll et ai., the sale price of nearby pro ert p y' (Lyon, 1978}. A total of 1,382 acres of greenbelt had 1972, p. 126}. ~ been purchased adjacent to residential The Philadelphia study was one of the few developments in the ~ 10 ears rior to the Y p to test fora `net effects' curve which postu- 1978 study. The sample consisted of proper- aces that while there is a positive impact ties from--each-area that =sole- in a selected on the value of properties abutting a park, calendar year which were located within it may be lower than the impact on proper- 3,200 feet of the greenbelt n = 82 , ( } ties a block or two away which acre not sub- Variables in the regression model that a jetted to any nuisance created by access were believed likely to influence the sales M and egress. The polynomial equation used price of these single family homes were: i to test for this eff ~ } ect was found to be a walking distance zn feet to .the greenbelt; good fit on one site - a junior high school (ii} age of each house; (iii} number of site with an athletic field with the rooms in each house; (iv} square footage of maximum impact on property occurring each house; (v} lot size; (vi} distance to the 600 to S00 feet from the site. city centre; and vii distance to the nearest . Another Philadelphia study in 1974 ana- major shopping centre. Th.e ~ regression lysed the impact on sales price of 336 results showed that, other ~ thin s ~bein g g . w~ i , 1 '\M1 +~FY~ i I 4r ply ~y t 240 Crompton ~'he e ual there was a $4.20 decrease in the price pay the costs incurred by the city in purchas- Tl q of residential property for every foot one zng the greenbelt. This creates a major policy prop moved away from the greenbelt. This issue. However, it should not inhibit the pur- Day1 suggested that if other variables were held chase of park and open space areas because in 1 constant, the average value of properties overall economic benefits accrue to tax Arb, adjacent to the greenbelt was 32 % higher payers whose revenues fund all the govern- spat than those located 3,200 walking feet away. mental entities. tal a These results are shown in Table 1. Resolution of this conundrum requires one adja f)ne of the three neighbourhoods had been of two actions. The first requires that a city's ertiE able to take much greater advantage of the elected officials be prepared to accept the par] open space amenity in its planning than the inevitable criticism that is likely to occur trail other two neighbourhoods, so the authors when it raises taxes to purchase the land. gars initiated further analyses on it. In this neigh- This selfless, `statesman-like' position is den bourhood, price decreased $10.20 for every adopted because they recognize that in the app foot one moved away from the greenbelt. long-term the city's taxpayers will benefit the This resulted in: when return on the investment is viewed in ~ T . th the broader context of total tax payments eve e aggregate property value far the neigh- to all governmental entities. The alternative wa~ borhood berng approximately $5.4 m~ll~on strafe is to ersuad greater than it would have been in the gY p e the other taxing ent~- Wh~ absence of greenbelt. This increment ties to jointly fund purchase of the open $3.f resulted in an annual addition of approxi- space areas, since all will reap proximate disl mately $500,000 to the potential neighbor- tax revenue increments deriving from them, wee hood property tax revenue. The purchase A study undertaken in 'Worcester, Massa- tun ~ price of this greenbelt for the city was chusetts, in the early 1950s examined the the j ~ approximately $1.5 million and thus, the relationship between four parks and the min potential property tax revenue alone would values of all properties sold within a 4,000 ave allow a recovery of initial costs in only foot radius of each ark durin th ~ three years. gyp. 215) p g e preceding res 5 years ~n =170} More et a~l., 1982, 1985; res ~ ~ p p Ha ert et ~1., 1952 The multi 1 g There is an im ortant caveat to these osi- g Y p e listin 5.1' five results in that 56% of the $500,000 prox service from which the study's data were 7.3 imate increment of property tax revenue derived recorded actual sale price of a In ~ accrued to taxing entities other than the house, along with information onother charac- ho, city, i.e., county, school district and other teristics that might affect the sale price includ- pry ~ independent districts. Thus, the incremental ing lot size, number of rooms, age, garage, any return to the cif alone was not sufficient to taxes paid and condition. Distance to the wh park m feet was added to this set of variables. by The results showed that, on average, a ~ Table ~ Value of the average house related to house located 20 feet from a park sold for Or f greenbelt prox~m~ty $2,675 more than a house located 2 000 feet in ~ ' Walking distance Average value away. However, 50% of the aggregate the from greenbelt of house increase in value was derived from proper- url ties located within 500 feet of the parks. frc 30 ~ $54,379 Effects could not be traced beyond 2,000 (N x,000 50,348 feet from the parks: Using these data, it was the ~ ,283 49,1 l2 estimated that th x e aggregate property value e 2,000 46~~92 increase attributable to thes $1 3,200 41,206 e parks was ~ $3.5 m~llzon. ~ ~ 1,( I ton 7"he impact of parks on property values 207 rip has- The impact of two parks on the values of Washington County, Wisconsin, is located c alicy proximate residential developments in 40 minutes northwest of Milwaukee and is e pur- Dayton and in Columbus, Ohio, was reported growing rapidly. The impact of two parks in .cause in 1955 Kimmel, 1985}. The 170-acre Cox the county on property values was studied tax- Arboretum in Dayton was a wooded open ~Sielski, 2002}. Jackson Park is a 25-acre o~e~n- space containing specialized herb, ornamen- park located in the Village of Germantown. tal and other plant gardens. lts impact on an The study was provoked by two common sane adjacent fairly new sub-division of 300 prop- concerns: ~i} property owners adjacent to a city's erties was assessed. The 152-acre Whetstane proposed county park were concerned it ,,t the Park in Columbus, contained ballfields, would have a negative impact on their prop- accur trails, natural areas and a 13-acre rose erty value; and iii}taking the property off the . aid, garden, and it was adjacent to an alder resi- tax roles would put an undue burden on the a~ is dential area. In both cases, samples of rest of the residents. the approximately 100 residences were used in The study used assessed values and e~efit the study. measured the parks' impacts within ahalf- ed in The regression analyses indicated that for mile (2,640 feet} radius. It controlled for ~~ents every additional foot of distance a property structural variables. The results for Jackson . five was located away from Cox Arboretum and Park are illustrated in Figure 2. Properties enti- Whetstone Park, the selling price decreased within 200 feet increased by $113.36 in open $3.83 and $4.87, respectively. The average assessed value for each foot a property was E ate distance of properties in the study areas closer to the park. Aggregated incremental em, were 814 f eet and 973 f eet from Cox Arbore- assessed valuation attributable to the park 'ssa~ tum and Whetstone Park, respectively, .and was $1.58 million which generated $30,128 `:the these properties yielded proximate pre- in annual tax revenues.19.2 % of the assessed ,.the miums of $3,100 and $4,700. Given the value of properties within 200 feet of the park ,f~a4 average selling prices of properties in the was attributable to the park. For example, if a tiding residential areas were $58,.800 and $64,000 property located outside the influence of the X88; respectively, the park premium represented park was valued at $120,000, it would have a ~s~ting 5.1 % in the Cox Arboretum subdivision and value of $143,000 if it were located within 200 ere 7.3% at the Whetstone Park residential area. feet of Jackson Park. of a In neither case was an assessment made of At Homestead Hollow County Park, ~'ac- how this average premium varied between assessed value decreased by $4.96 for each clod- properties immediately. abutting the parks foot of distance from the park up to the age, and those located say} 2,000 feet away, half-mile radius. These results were similar the which presumably were much less impacted to the Boulder greenbelts study reparted ;es. ~ by the parks. earlier in the paper: Aggregate value attri u- a An empirical investigation in Salem, table to the parks was $880,000, generating for Oregon, in 198fi reported that open space annual tax revenues of $18,100. Feet in the form of greenbelt at the fringe of A county wide analysis of 6,598 single ate the urban area exerted an influence on family residences sold in a 2-year period in e'er- urban land values that extended inward Leon County, Florida, reported that homes ks. fxom the urban boundary about 5,000 feet within 200 feet of the nearest park were D,D~ (Nelson, 1986). The researcher concluded worth an extra $6,015, while the premium gas that urban land adjoining farmland zoned for those between Z00 feet and 1,320 feet ue exclusively for agriculture was worth X0.25 mile} was $1,773 Cape Ann Economics, as $1,200 per acre more than similar land 2003}. There was some evidence of the 1,000 feet away. impact of a restricted supply since. when . . . w~'y "~5""~ i~'1y"~ wlyp y 208 Crompton ~ a ~ : X120.00 $113.3b ~ C ~1oo.0a )r x ~8o.ao ~ c ~i v w,~~ . e r' ~ , Decrease of total ~ $G0.00 . assessed value per r, ~ foot t~ ,l. ~ $40,00 a; u ~ ~ ~ 14.53 v $20.00 it j ~ ~ $b.4 ~ rE .•/y/ , G. W a-too ft, 201-40o ft, 401-~00 ft, G01-soo ft. A 841-1000 W Fig, 2. Decrease of Iota! assessed value per foot Jackson Park ~ ~ p. al ~ ar ' the analysis focused on the most densel in y surrounding woodlands. Over the 2-year w populated parts of the county hover 2,500 period 442 sin le-famil ~ ~ people per square mile, primaril within the g Y residential proper- m Y ties were sold in the census blocks immedi- ~ ' city of Tallahassee} the vE premiums for ately surrounding the gardens. Regression parcels within 200 feet of a park rose to anal ses Indic approxrmatel $14 000, Y ated that properties physically ac Y ~ abutting the park had a premium of $47 000 In addition. to the county wide anal sus n ~ 1 ' ~ ' Y ~ ~ 04}, while for those not abuttin studies were undertaken on the s ecific wit • g but p hen 200 feet the premium was $21,000 impact of two parks. Myers Parkin Tallahas- ~n = 70}• These remiums see is a 47-acre natural area ark. Data fro p when applied to ! P m all properties within the 200-foot zone 58 single family home sales in the previous added $6.3 mill' ' 2 years were used i ion to the property tax base. n the anal . szs wh' Bch co - Y r n A stud of the rm cIuded that those within 200 feet of . Y pact of 14 neighbourhood the _ _ _ _ p~'k parks on suburban areas.- of -the -Dallas-Fort " sold for $24,fi00 more than they would have VL~orth metro lei was brought had they not been close p based on 3,200 res~- to the dental sales transactions recorded over a 2 park. Since there were 75 properties within 1 year eriod filler 2 this 200-foot zone the ~ p ~ ~ 001}• The parks were enhanced value all between 2.5 acres and 7.3 acres in size attributable to the park was $1.$45 million. exce t for two which w ~ Macla p ere 0.5 and 0.3 of an ~ y State Gardens on the fringe of Tal- acre. They were `intermittent lahassee is a Florrda State Park embr ly maznta~ned acing and were selected because of their ordinari- rolling hills, a picturesque lake and spectacu- ness rather than thei ~ Iar and extensive floral gardens featurin r excellence, The g author described them as `a standard of both native lants and P exotic fl ora. It park quality well w~thzn the range of an includes the 877 acre Lake Overstreet evenly mar inall com g Y mated developer, ~ addition which also features a lake and National monuments these are nat' gyp. lfi9}, ~ Fig :x~ E y, I The rmpac~ of parks on property values 209 The selected parks were in neighbourhoods small parks, the premium is small relative to of single-family houses. As far as possible, that of proximity. All else equal, then, more parks near arterial or collector streets, shop- value will be created by a series of small ping or commercial centres, or abrupt parks, which permit more total houses in changes in demographic characteristics their vicinity, than by a single large park of i were excluded from the study to clarify the equivalent area. effect of the park. The comprehensive Figure 4 demonstrates the outcome if this regression model incorporated 29 variables principle is applied to a 50 acre park illus- that could potentially influence sales price. tration. It suggests that the tax base enhance- Travel distances using a GIS program were ment emanating from six 8.33 acre parks with used as the distance variable. dimensions of 400 yards x 100 yards, and The price effects compared against home. non-overlapping impact zones, will be sub- values ahalf-mile from the parks are shown stantially greater in aggregate than the in Figure 3. Homes adjacent to parks premium generated by the 1210 yards x 200 received an approximate price premium of yards, 50 acre park. However, such a 22% relative to properties ahalf-mile away. revenue benefit is likely to be partially Approximately, 75% of the value associated offset by higher initial development and with parks occurred within fi00 feet of a construction costs, and more expensive park and 85% within 800 feet. This distance operations costs over time. Nevertheless, approximates a two to three minute walk the scenario of a set of smaller parks rather and delineated the parks' principal areas of than one large park may be appealing to influence. The price effects of the parks developers because they do not have to were insignificant at a distance of approxi- incur the" additional ongoing maintenance mately 1;300 feet ~a quarter mile), the con- costs whereas governmental entities do. • ; ventional ~ estimate of a 5 minute walk. A large data set to measure the impact of This study also found that while large parks the proximate principle was assembled for ` add more valuable to residents' property than the city of Portland, Oregon. It was comprised 20% f+ i v s. ~ IO% W j a t 0% © °oo/~ ©o~ ~ © o © o r"w ~ Y 1 ~ ~ n A ~ ~ M1 A ~ A A Travel Distance to Park, in feet Fig. 3. Impact of proximity to parks ~~4 neighbourhood parks, Dallas-Fort Worth Metroplex~ 4 ~ ~ -w .'i;~ilw"i~';.+~: r; i~~w~~l:~ _z~; ..~~Ir~s.a~wly`!y r 1~~ ~ a~wr; ~,~tir1~ 1 ~~~,y~1w I I € 2 I 0 Crom ton TJ I ~ ~ T; ~I ~ 1. A I U r ~ ~I i j' i. rw..r.w.rwrw~~w.w..'«ww..•ww.x~~w~x.r.nw~wrr Nww.wr..rTM.w••i•.nrwrr.y.....r.rwrr.ww.•r..n~wrwwN.....n•.•~-.w«.x.N 1 I~ ~ t k'k WN wq'fp'Y~NP N NNM'n NNN'N'IrNM NN« ? € ~~WI• R ~NNaMn~M.M:w Fxltl l.IkxfmM WYM Mi ! M Y~9-%fr:MMttlp Y.IV yP~R.W'Ii .1~r~.Y`Qia..k N'Pd ~ A I' Nt. Y. N~' ~ ~ ~ M r y, ~I i wax I. W. N:'p.:' wn ib NlN-M.q:,w ~ R.! ~Nw NP'r lNNw Nr wNr w.rN!K Y~ M. ; h'~raa M: 'AN ;x~da~Q. rr; i € I ! ~ ~ n~ wv w 'M s. ~ " ' i E i i y„...,-.n..M~_. N ♦ .nP W ..-~...„..-...N Nu.._. ! M P, ~ M R y.,.w....w..~..»...,.,.,,«,......,M....,...w~~.,w......,. a ~ ...~f ~ • N k X, N M t w I; N 4r ~ I ~ e M 'I i ~ iN N; N 1 € P .N WI N m t ~ l M 41 F A N w Y y. b€ ~1 .A N I ~ I N M ~ P E Y Na N ~ ~ P IW C,JIItx I N.~Nx t♦}xAW r+M ~RW kNlMMAmV kAr IN NON 111r iVl.~, r N a r t I ~ ~ N: 'N. i x N N' w: y.. ~ i~ M ~ I s ~ i • t i In 'n r n ! a N e~ ,~~71 ~Ii~•J,F4y~• 1~F.9'Yw N•Y, MAi A.N ^«+Ifiu.•F N N .f Y Y•M M I. I M Nv l ' , s 'M ~ I LL 1: W f 4 G I , A F W q ; I ; m y, P k I 6 ♦ I~x E ~ r i H.,,.-,«.E,~N...~„,.~I.,„... Lam. ..K .,~.'w,., ,.,..ww..r..~.... ~1. ~ i ~ N ~M~ P I r....- w .,.-.rn J ! Jk.. ~ ~ ~ N. ~M„t e~ ° w, r.....N i ..M „ql...,, ....w.~~.... i~ . ,a. ,a ~._,""~X._,„,N,.. ....M m W Y N w rN, X~ a m wti M ~Y tw Ic N M~ N.. I~i W.: E M N ~ Ii: M1 . w M' M Ntl NIY Y h I Y it T N N' Mi i N 1 w ox A III ~ W i R fp w: n n~ 1~~155II E MI (N qE a< M: ~ M ' l4 N 1 Y „ I ry ~ ; + M..€~"J~ .,..r . P w Ni w' M j W i N. pi ~ M. a, ~ / IN fl 'N' ~ 3 ' ~n y~-...... .,,,.y~.. t I ! m ~a ye i f' --w.1. I , Y ..w.. Y.. . w. ..wM.,. 4„Y, ~ N...w.. I iYN,eNM wNNNMY NJI q Mn PNNq MMpkM p.MNNYNNMW S~` .I. M N N NW.. M w, ~ W * v j ~ ~ x¢ qa ~ yl r. Y~{ 'gk.PP.NIN NYYN.YNNNPN YNN N. n r~~' ~ N N ?qw MM9 w.Y~N gYtlq Qq'NNp C:NN M~YYMEi.YlNYMN NNN P'NM1N N~''! ~ Ni-. skq FN NP'Y'q~'Y:n'!f tll Mn~WN' P' .-M Y'. N. p, n ~ n y m I • ~ N:.~ N u~E•..~d MM »'r •jl'r M~x iC1 MF3a~;r~ N~ M q' N..' .N'.... 1 iM N E ~ ~rINAwN•P:~:~gpNN'N~~WH•MpJ~~LY~~r~Rb~MmM' wN!rxrriNk'N~'~,,NMRY NYP'N,NN'NN'PNNPN ~ • ~ i na w~ Y ~ M~ ~ .MN-~•~.ruY~>~Y.: N.~,q~MY'.Nri,:~N~N°' .E . _ . _ ~ ~ ~1~ • I ~ NkP NrP N,N NN.N kPNw NRMNMMNYNM~NJI ~ . 1 NI IY~NA'~ANiMi N.NRi NYM'M W~{~ N Ot, Nwx NNXNNkM MMN NMNMNN NM MYNw NM PNNNNP W,NNYMN PY' i IM % EE ~ ~ N I N 1w M i p•,. ^~'^••""w'p`N'!f N WWIWM~W Vq"N"W'N 1 N M x NW'r'.N w'MN N W M N`~IFM`NTN~•„„~~^^•••^'.•• .:,,n~.m-~•«~-..»~......T...µ~ NiN'il IA N NM 1t \ NI.N..~L~..M.x.fI.N.N kM MM'N w'Y.N M N YI 11 Y MN it Y N M Mry^ N^^^"•'M.-.,^ -nwyF~..~...r...~..~.......•.,.-.-„.. N k i ~ ~ ! N_«. 1 ~ N,„.....,M 1 ""^"„`~'"w.~Mr«w.......,.~...•».,.,,.ww,••. ur N y N. 'I ~ I ~ ~ ~ i .4 N IA M Y w N M N M M> rM NNM M~friM'N:M'N'-0ifN'w.MN'IF. M~M'.N N'G f.~LL.M N'M N'M'Y I~. r 3 '.0. } ~ M .N.•.• N. N .,~w..W.. ...r.... Y ■ N• F 1 ~ ~ N. 4 t m N w. N . N. N Ir N. ~ I 'N N , NI H ti M',.. ' N p I N N Y N. N: • N P. I ~ {V 'iw I :~......p....w.+kr... ! P N Ni N .a-. ...«.W.,~. '.N. '.,i~ N I tll W wE M ! .R ~ N..,.~, ~ ~ I~H Y N.., Y N P N.«.... ~.~W.,.«..,..Y M. 1 N M M N ~ M [~fl i N:. ~ N M Y Y i x Y w E N N N E x N ~ N t fog ~ » rifN~lriM~.'r<~I.r.. «.p.+1f'SM EinY1•: i.fi+, ~~»-*lir. r. N: N Wr+.wwlw.wyr5!YM~ N. . .ao y 1e "k.. ~V I "^•~P•riN'Y"N`N`N tI M;M'kr Y~x~Ny NM NNNPMNN~MMMkNNM. N'N rNY r•qM. ~I 'b P !"N.~~. I,r M.N~~~riY.r+ N ..yW ,w }:F`.S'iM N W y 3 : ~ M i pa . IF ~ M: 0 fi M ~ •I I: b ' ~M.ti~ha<i.~Y.h~M•i~f.!iT~'hM'NW,UN kH N4 NF.W ~:I WNN!NN~i:Gp 1'P>•M:w i ..n t m Y E I i ! M M. ' t~,.,.,......M....,...,.,..,,,, ; ! i ~ II A M p'~!:r, 3 e B! I! M ~y:Mi. a~~l" N ~ I ,N ~ x•. ' ....MM,....,..,..«,..„..M«..,..wr ................M...ww.~M,.~»~...~ Y t %a~i~ Y N' w .nw ..~...,,t.N~Arp 1~'Y~wM.N~ ~ ..w w N"N ~M YI'N~iY'N~11"li ~'d ~u~N~YI Y"M'N A"11« the Fig, 4. Implications for proximate preminus of distributing 5o acres of parkland among six smaller parks re 1 rather than aliocatin ~ t to a sin le lar a ark g g 9 p ex; I aw ha of 16,636 single family home sales during a value by $2,262 X3.5%} ar $845 X1.2%} aV' I three year time period. The mean home depending on the model used, compared to uri sale price was $66,198 X1990 dollars} and property outside the 1,500 foot area. When the a~era a size was 1396 s uare feet. The the im I g q pact of different d-istanees within the impact of parks on property within a 1,500- 1,500 foot radius was evaluated by the two Tal foot radius was measured. It was estimated models, the premium values ranged from that a block was 200 feet, so the 1,500 foot $5,023 X7.6%} and $3,527 5.3% for ro er- Vai ~ } p p X0.28 mile} radius reflected an average dis- ties within 100 feet of a park, to $2,109 Dis Lance of approximately 7.5 blocks. X3.8%} and $1,004 X1.5%} for properties that Dis I• Results from these analyses were reported were located 1,301 to 1,500 feet awa . Dis Y I in two different papers, In the first paper the In the second paper using this same data Dis 193 public parks were not differentiated by set, the authors classified the ublic arks Dis p N p Dis type ~Bolitzer and Netusil, 2000}, Two scat- into three different cate ories: urban arks g p Dis i st~cal models were applied to the data set. natural area parks, and specialty parks/ ~ The authors concluded that homes within facilities ~Lutzenhiser and Netusil, 2001 . N } Nur 1,500 feet of a public park increased ~n These are defined ~n Table 2. The results iI r i~ i he rm act of arks on property values 211 i' tan ~ p p ~ Table 2 Definition of open space categories ppen space type Definition Urban Park More than 50% of the park is manicured or landscaped and developed far nannatural resource dependent recreation (e.g., swimming pools, ballfields, sports courts}. Natural area park Mvre than 50% of the park is preserved in native and/or natural vegetation. Park use is balanced between preservation of natural habitat and natural resource-based recreation (e.g., hiking, wildlilre viewing, boating, camping}. ` I` This definition includes parcels managed for habitat protection only with no ' public access or improvements}. '4r Specialty park/facility Primary use at the park and everything in the park is related to the specialty category (e.g., boat ramp facilities}. r~ ' ~.I °~.w;' 'Y. I~ ~1~ showed that bein within 1,500 feet . of a do suggest there are relative disadvantages *.1 g natural area park accounted for $10,648 to being located next to the facilities, since ;a ~ (16.1 of a home's sale price holding all the largest premiums for the urban park, w. other factors constant. The impacts of natural area park and specialty park/facili- urban arks and s ecialt arks facilities ties were in the 201-400, 401-600, and P p yp / were $1,214 (1.S%~ and $5,657 (8.5%~, 401-600 foot distance bands, respectively. respectively. The relatively low premium Another technically strong study was for the urban parks may be attributable in reported on the impact of the Barton Creek part to urban parks often having greater Greenbelt and wilderness Park in Austin, variations in quality. Texas (Nicholls and Crompton, 2005. This The impact of distance from each of the is a linear 171-acre natural area to the west three types of area on home values is of downtown that includes 7.5 miles of ~S: reported in Table 3. This shows; for multi-use trails. The authors examined its example, that a home located 401-600 feet impact on three neighbourhoods that bor- n away from a natural area park on average dered this amenity: Barton, Lost Creek and had a $12,621 premium (19.1 while the Travis. Single-family home sales over a average premium for a house adjacent to an three .year period constituted the data urban park was $1,926 (2.9%). These data source, The sample sizes of home sales for a, Table 3 Variations in proximate values at different distances for each open space`rype (1990 dollars 0 Variable Urban Park Natural Park Specialty Park/facility 9:~ Distance X200 $1,926 $11,210 $7,396 Distance 201-400 2,061 10,216 5,744 Distance 401- 600 1,193 12,621 10,283 Distance 601--800 817 11,269 5,661 Distance 801-1,000 943 8,981 4,972 Distance 1, 001-1,200 1,691 8,126 4,561 Distance 1,201-1,500 342 9,980 3,839 Number of observations 16,747 Y . . . 2 ~ 2 ,Crompton the Barton, Lost Creek and Travis neighbour- than of the greenspace. In the Travis area hoods were 224, 240 and 236, respectively. where the proximate premium was relatively Results of the study are summarized in low, the topography of the land did not allow Table 4. The table shows that the premium for non-adjacent properties to enjoy agreen- for adjacency to the greenbelt was highest in belt vista, so the premium was primarily a the Barton neighbourhood and that zt rep- reflection of the value accorded proximate resented 20% of the average price o~ alb homes access. rn that ner~hbourhood. The comparison cri- A study conducted in a 1,350 square mile tenon is important because all the homes suburban and exurban region in central impacted by the greenbelt are included in the Maryland used a sample of 55,7D9 arms- average price. If the comparison criterion had length single transactions of owner-occupied been with houses beyond the direct impact of residential properties that occurred in a 5 the greenbelt say 1,500 feet or more away), year time period Irwin, 2002. It measured then it is likely that the premiums shown in the proportions of areas within 400 meters Table 4 would have been substantially longer. of houses that were in different land uses. The last column in Table 4 shows the decline The study recognized that o ens ace ishet- P p ~n value with each foot of distance away from erogeneous and measured the impact on the greenbelt. These figures are substantially house sales price of different categories of higher than those reported earlier in the open space. The author reported that incon- ~ ; paper for the greenbelts in Boulder, Colorado, frost to residential, commercial or industrial ~ ; the two parks in Dayton, and for the parks in uses, open space had a positive im act on a p VL~ashington County, Wisconsin, although in residential property's value. However, the the first two cases the different values may premium for proximit to rivatel owned Y p Y be attributable to inflation in the two decade open space protected by a er etual ease- . p p time difference between the two studies. meat was $4,503 or 2.6%, while that on prop- F The ' lack of positive rmpact in the Lost erties close to government purchased open Creek area was attributed to the different space was $2,038 or 1.2%. It was su ested gg character of the greenway at that point, that the privately protected land yielded a ~i Homes dir ectly adjacent to the greenway ~n higher premium than the publicly owned ~ ~ Lost Creek were located on the edge of land because the 1 after is available to deep, thickly vegetated ravines which people from .outside the local ar ea. They offered neither recreational access nor may generate a spillover nuisance cost b attractive views. The vegetation inhibited reducin riva Y g p cy and increasing congestion recreational access and the views were of which is not present at privately .owned other properties across the ravines rather open space. Table 4 Results from three Austin neighborhoods proximate to the Barton Creek reenbelt and Wilderness 9 Area Decline in value Home sales prices ~$'s~ Adjacency Adjacency er #oot #rom p Nei premium premium the Greenbelt `ghborhood High Low Mean s} percenta a °s 9 ~ ~~j Barton 550,000 105,000 220,000 44,000 20% 13.51 Lost Creek 899,000 179,D00 356,000 D 0% 3 Travis 392,000 130,000 233,000 16,DD0 6°/° 10.61 I l i~ , w -a. .i i i The im act of arks ors ro er vairres 213 ~ P p PP~Y ~ A similar study was undertaken in Berks had benefited from substantial capital County in southeastern Pennsylvania investment in renovation during this time (Ready and Abdalla, 2003}. The data base period. a was 8,090 residences sold over a 4 year The five parks were Prospect (Brooklyn}, a period in the suburban~exurban areas of Crotona (Bronx), Clove Lakes (Staten e the county. Again, the amount of land that Island}, St. Albans (Queens}, and Serrano was in open space, residential, commercial (Bronx}. The graphs in Figure 5 compare and industrial use within 400 meters of the sales prices per square foot for single each house was measured. The authors con- family homes and multi-family units (where chided that within the 400-meter area, open these were present} over the 1992.2001 space was the most desirable land use but period. The sample sizes (n} of sales trans- ~ the premiums on house prices were very actions from which the value data are small, even lower than those in the Maryland derived are shown underneath each graph. study. The results show that f he positive impacts The rejatively low premium values of renovation at the first two parks were sub- I, reported in these two studies may be a func- stantial; for the other three parrks the results tion of three factors: (i} the self-cancelling showed moderate enhancement of property effect of aggregating open space because values. both high quality amenity open space and Olmsted and Vaux considered Prospect al dispirited open space that leads to Park to be their masterpiece. In the 1992 - a decreased proximate property values are 2001 period, $103 million worth of capital included in the mean averages; (ii) averaging renovation was undertaken in the 526 acre the proximate premium over 400 meters park, restoring it to its status as a first- . because most proximate value is likely to class, signature park. The PIA and CA zones ~s` be captured within x50 meters and the selected for comparison were in the Park value decay beyond that distance is substan- Slope neighbourhood. In the most recent 4 tial so that at 400 meters it is likely to be zero; year period, single family homes sold for a and (iii} some parts of the study areas were between 32% and 153% more per square rural with zoning ranging from 1 to 5 acre foot in the PIA than in the CA (Figure 5a}. ~o~ minimum lot size; so the supply of private The same trend was apparent in the compari~ .y open space was relatively plentiful. son of multi-unit properties b.ut the diff er- ~y In 2003, comprehensive detailed studies of ence was nat as dramatic, ranging from 20% the impact of major renovations in five to 84% over the most recent four years ~d selected parks in New York City were under- Figure 5b}. In the case of multi-unit pro- taken (Ernst and Young, 2003}. The authors perties, the prices were similar before the did not use hedonic analysis. Rather, they renovations commenced and as more compared the values of property sales trans- improvements were made the value gap ss actions within Park Impacted Areas (PTAs}, between the PIA and CA zones became which consisted of 1-2 blocks immediately more accentuated. However, the per square adjacent to the parks, with associated foot values of both the PIA and CA properties Control Areas (CAs} comprising the next 3- increased markedly, Some of this may be 4 blocks beyond the PIAs. The CAs were attributable to inflation and the vibrant used to hold constant the influence of the economy at that time, but it is likely that other factors that might impact real estate some of the CA added value also is attribu- values. The comparisons were made over table to the renovations since being 3-6 the time period from 1992 to 2001. All of blocks away from the park suggests the the five parks selected for the case studies park exercises some proximate impact. ~~$ru~~~;rt~c1.. i. I i 214 Crompton Prospect Park (5~G acres}, $1Q3 million,1.993--ZQOI (a) Single Family ]-tomes (b) Mufti-unit Dwellings i0tl ..~..,.„....~...................~............,.........T.....~,....w~.r.. 450 ' D00 100 950 ~ 300 400 . . na 350 ~ ....µm. ...........M ow...- 300 ......,..,..r.. „d 300 ~r 150 . &.r---4? _ ~ .~,,:.;~a~:_ . 200 :r....__....... t~ W t.a Ati w 100 100 rJP.:., ~ 0 0 7992 1990 189a 1895 SBAB 1091 1005 1989 2000 3001 1992 1993 t99d 1995 1996 1987 1908 7499 2000 200t --~PIA(n781 •,.~._.GyIn~100y i ~-~pIA(nn575).....,.._..~.__~CA(n•869)..~ ii,i r , Cratona Park (~2$ acres), $1.2 million,1993-2Q01 I~' (c) Single Family t~Iomes (d) Multi-~rnit'Dwelling 90 - - BO ! ~ 50 ~ . 80 50 h~r- - SO , . ~ 7D ~ Sa A 0 ~ 1992 1993 189a 1095 t99G Nxt99'T 7988 1895 2nC0 ^14921. 1983 1994 t995 1998 1997 1896 1899 2000 2001 APIA Jn~30) • ^ ^ ^CA (tA11 ~W~~P1AJn~97}W • ^91' • ^nA(nC~74n} i Im~,...~..X~.~.~~.....x~........~....~..~y...~..~......~......~....~...._~~...~I L~.~.......,......,,......,..,.,..,......~..,~...,. - j Clove Lakes (198 acres}, $5.b million,1993-199G (e) Single Family Homes 250 j~, I;I 150 _.~t ~ 'i A ,092 ,993 198E ,995 1098 t497 1898 1990 2000 2001 ~ . •^~'^^h~^^PIA(n•a8) . ^~...~CA(na643J it St. Allaans Park (~1 acres), $1.7 million, .1999-ZaOQ I East: Single Famil Homes g Northwest: Single Family Homes I stsD stab ~ j stab ' __~.M ~ StSD r S12D / Y St40 ~ ~ I 1 5100 ~_M.~ 5170 SBa I ~ u ' S?0 , 380 i SOA - - - s4D -=j 520 ~ , Stn 3D ~ sa 1893 198a 1905 ^1998 1807~x 1998 1999 2000 3901. r 1 9 t893 194a 1995 t995 1987 1008 t999 loan 2001 93 i , ~-PIA(nM9.......~.~.~.~...~....., ~ X33} I III 'i , ; Serrano Park (2~/ acres}, $450,OU0,1995 (h) Single Family Homes (i) Multi-unit.Dwelling i Stab -m-~•-^r st30 - ~wf 'll I ; Stan ~ • a5n ,a" 580 ~ c • . SOD ` ~ i SBA i' i 540 Sao ~ t 520 - i 520 u y' SO t992 1995 199a 1995 199G t897 t998 1998 790D 2001 } 1892 1993 7994 1995 1806 1997 9998 1990 2000 2007 ) ~ ~ I~-PIAJn~S~•,,,....^~...CA(n~1111..+ r--~--'-P1AJn~C5)w..-.~.~~. •CA(nM321 I I, - i 1 i,.: Fag. 5. Comparison of the sales price per square foot of properties within the impact area APIA} of five parks with those of their controls areas ~CA~ IIIIIi i ~I; The impact of parks ort property values 215 Renovation of Crotona Park took place impact of the park. Data from its east side, from 1993 to 2001 at a cost of X12 million. summarized in Figure 5f, show no substantial During the 1970s and 1980s, the 128 acre difference between the sales value ofproper- park was situated within a decaying urban ties in the two areas. The second PIA was on neighbourhood in the South Bronx, charac- the park's northwest side. This is a more terized by burned-out vacant buildings, extensive residential area so the sample drug dealers and crime. Efforts were made size was larger. The PIA values historically to upgrade the neighbourhood, but invest- were slightly higher than the CA values, but meat in the park only came later. However, this gap increased dramatically to 19% in in a few years it was transformed from a 2001 after the improvements were com- place to beignored and avoided, to an attrac- pleted (Figure 5g}. Since the renovation tive asset. Figure 5c shows that values in the took place in 1999 and 2000, if there is PIA for the most part are higher for single impact on the market price of praperties, it family homes than in the CA, but the rela- was likely to become more obvious in the tively small number of sales transactions period beyond the timeframe~ of the study. means there is some volatility in the graph. Again, both the PIA and CA values increased Among multi-unit dwellings, the CA values substantially from the time the renovations were substantially higher than those in the commenced in 1999. PIA in the early years reflecting the blighted Serrano Park is a 2.5 acre playground and status of the park, but in the Iater years the park located in the Castle Hill section of the situation was reversed (Figure 5d~. There Bronx in a densely populated area. Although was a trend showing an increase in PIA $650,000 was invested in 199$ to renovate its values after the renovation work com- structures; it remains aesthetically unap- menced in 1995. pealing since the maj ority of it is `avast con- Clove Lakes Park is a 19S acre natural area Crete field'. It is heavily used, so there is surrounded almost exclusively by single noise and congestion. The graph in family homes. Between 1993 and 1996, X5.6 Figure 5h and 5i reflect these unattractive million was invested in renovating it. Since qualities in that the facility appeaxs to have that time, it has become a weekend destina- no proximate impact on property values. Lion for Staten Island's residents as well as In addition to the proximate value data a staple of the community. Single family reported in Figure 5, the authors empirically house prices in the PIA were higher than addressed other impacts in their case those in the CA before the renovation and studies. Thus, they were able to conclude: that trend subsequently continued. `Single family turnover rate was generally Figure 5e shows that in the last 3 years of lower near well improved parks as compared the study's time period, the value gap to adjoining ones. Quality parks serve to ranged from 36% to SO%. Although the gap stabilize local communities and area cata- s has generally not widened, the values of Iyst for the redevelopment of adjacent real properties in both the PIA and CA increased estate' gyp. 10}. markedly, as they did in Prospect Park; again ~ . suggesting the CA experienced some positive CONCLUDING COMMENTS proximate increment. The variability of the PIA sales price across years may be attribu- The evidence that has emerged from rela- table to the relatively small sample size. tively sophisticated analyses in ~ the past St. Albans Park X11 acres} was renovated in two decades, essentially endorses the legiti- 1999 and 2000 at a cost of ~ 1.7 million. Two rnacy of the proximate principle which was PIAs were used to measure the proximate demonstrated in the early "Englishr urban i ~ i li i ;i ' 216 Crompton ~ parks and later disseminated in the U5 by such as level of maintenance, maturation Olmsted in the nineteenth century. The level of the park, ratio of supply and t e of use. i evidence from these studies unequivocally demand and yp supports the contention that parks and Level of maintenance relates to quality. A . . Halo ~s with awell- roomed front ~ li open space contributes to rncreas~ng proxz- useful a gy g mate property values. garden which is likely to increase the value i i. ! ~ It is not possible to discern a generalizable of a home, whereas if it is overgrown with answer with regards to the magnitude of the weeds and littered with trash then the prop- ! proximate effect, given the substantial vari- erty value is likely to be diminished. Adverse ation in the size, usage and design of park impacts also may emanate from nuisances ~ lands in the studies, and disparities in the such as congestion, street parking, litter i i' residential areas around them. However, and vandalism, noise and ballfield lights i I i,~ ~ some point of departure based on the find- intruding into adjacent residences, poorly ~ ings reported here is needed for decision- maintained facilities, or anti-social makers in communities who try to adapt behaviours. I! ` these results to their local context. To meet Maturation level recognizes that it may this need, it is suggested that a positive take 30 to 40 years for new or renovated impact of 20% on property values abutting parks to mature. In the beginning trees are . or fronting a passive park area is a reason- small and spindly, plantings are scattered ! able starting point guideline. and immature, shade xs scarce, and the land- i The diversity of the study contexts makes scaping often is not aesthetically pleasing. 1`zable Thus the remium in the earl ears is ~ zt feasible to offer a tentative genera z p Y Y j answer regarding the distance over which likely to be less than in later years. I,il,j ~ the proximate impact of park land and open Like all other goods, the premiums that ~ space extends. There was consensus among people are prepared to pay to be proximate ! ~'I the studies that it has substantial impact to a park or open space are influenced by ~ u to 500-600 feet. In the case of community the available supply. If such amenities are p ~I; sized parks it tended to extend out to 1,500- relatively abundant, then the premiums will 2,000 feet, but after 500-600 feet the likely be relatively small or non-existent ' uses in an '~~II 2002 Sim~larl if ho premium. was small. Few studies tried to (Nicholls, y, j ~ identify impacts beyond that distance area have large private gardens, then it ~s ~ I'` because of the com ounding complexity likely that premiums will be lower than in ~~I~ p i ~ created by other potentially influencing vari- areas with little private space because pri- j I'~ ables, which increases as distance from a vately owned space may act as a partial sub- ark increases. However, es eciall rn the stitute for ublrc arks ace rller, 2001 . p p y p__ P P ~ j case of larger parks, it is likely there are Parks serving przmar~ly active recreation I I'~ additional economic benefits not captured users are likely to show much smaller proxi- i I ~ ~ by capitalization into increased property mate value increases, than those accommo- i ~ , values beyond this peripheral boundary, dating only passive use (Sainsbury, x964; ~I ~I since the catchment area from which users Hendon et al., 1967; More et aL, 1982. The i .'jI ~ come frequently extends beyond it. superiority of passive parks in enhancing i ,~~'i This type of work is useful in that. it pro- the tax ~ base presents local governments ~ ~~i~ ' vides a measure of the value of parks, with a conundrum because frequently they j 1i whereas elected officials tend to think only are under considerable pressure to give ~I,li of their cost. However, the focus is myopic priority to creating facilities fox active rec- ; since the value of parks to a community reational use. This is often the more attrac- 1i j;;. involves many factors other than proximity five option to conventional leisure. services ill~~ ~1 ~I ~ ~f , i ton 7"he impact of parks on property values 217 p `lion agency thinking in that it responds to an the analyses, Business property tends to ~d overt and highly visible user need, accom- turn-over less frequently than residential modates a relatively large number of partici- propeirty so this threshold is rarely available. A pants and generates revenues. Organized Second, business properties often are not tY rout recreational sports groups are especially entered into the Multiple Listing Service aloe effective in politically lobbying for facilities. data bases used for the analyses. Third, ith In contrast, users of passive parks, changes in annual rents, rather than prop- rop~ occasional users, and non-users of parks erty sales, could be used, but this infor- yexse who are the primary beneficiaries of mation tends to~ be proprietary and not ces passive facilities rarely offer a counterorga accessible to researchers. litter nized lobbying force. There is a growing recognition among lights Finally, it should be noted that appre- developers of the legitimacy of the proxi- ,QOrly ciation of property values is not always mate principle and of its utility for develo- ,QCial perceived by homeowners to be positive. pers. Thus, in a careful, comprehensive and lts corollary is that their property taxes are technically strong study that was commis- `may higher. Some residents who have lived in a sinned by a developer the author concluded: ated location for a long time and have no interest es are in selling their property, may see no personal Parks have traditionally been considered a ~tered benefits accruing to them from development cost center in neighborhood planning, an .land- or major renovation of a nearby park. Never- amenity that must be provided by local gov- .asing. theless, they are required to pay higher taxes ernment or required of private developers by ~s is because the appraised value of their prop- statute in order to be feasible. This research ert has increased. in contrast, suggests that providing parks ~n y new neighborhoods offers clear financial , ~~s~that In a broader context than a local neigh- benefits to developers, that those benefits ~~mate bourhood, it should be noted that these are predictable using objective research ~ed by types of analysis fail to capture the; `public' methods, and that they can be captured ~~s are benefits of parks beyond those that accrue through careful design and development ~~s will to proximate properties through the `capita practice. (Miller, 2001, p. lol} j~stent lization' captured by hedonic techniques. 'n an These public benefits include such factors Despite its limitations, the empirical findings it is as reduced soil erosion, water supply protec- reviewed in this paper are important an in Lion, wildlife habitat etc., and secondary because they provide park advocates with se pri- benefits that may be attributed to parks' legitimate monetary indicators of value. a~l sub- role in attracting visitors, businesses or Such indicators appear to be central in the ~D~01~. retirees to a community. decision-making paradigms used by many eatiori A further limitation of the studies reported senior bureaucrats;`-and to 'be demanded by roil- to this point is that they are confined to elected officials anxious to demonstrate mo- single family homes and do not address the `accountability' for public expenditures. 1964; impact of parks on proximate retail or In contemporary times, the power of the . The other commercial properties. These proper- proximate principle appears to have been ~ncing ties often constitute the major elements overlooked by park advocates since it has ~~ments around downtown parks. The lack of good rarely been part of the political debate. The they empirical work in this context is attributable evidence reported here suggests that the 4' give to three factors. First, hedonic. analysis principle should be resurrected. There are ~ rec- requires a threshold number of property encouraging signs that this is occurring. For attrac- sales to have occurred around apark to gen- example, the .city of Houston recently ~~ices erate the market data needed to undertake announced the construction of a ~13 acre ~ A i' I 218 Crompton i ~ ~ i downtown park to be completed by 2007 for Kimmel, M. M. X1985} Parks and property values: ~ ~ ' I' j $40 million. Mayor Bill White stated, `Much of An empirical study in Dayton and Columbus, a, ; . i ~ ~ the city's investment will be recouped over Ohio, Oxford, OH: Miami University, Institute ~ ' time by increased tax revenues from the of Environmental Sciences, MS thesis. ~ ~ enhanced value of ro ert around the Lancaster, K. J. X1966) A new approach to ~ , p p ~ consumer theory, The Journal of Political ~ , park that the park wzll create . Economy, 74,132-157. Lutzenhiser, M. and Netusil, N. R. X2001) The effect of open spaces on a home's sale pace, Contemporary Economic Policy, x9, REFERENCES 291-298. Bolitzer, B. and Netusil, N. R. X2000) The impact of Lynn, D. W. X1972) The spatial distribution and ~ ~ ! ~ o en s ace n n impact of public facility expenditures, p p s o property values m Portia d, ~'I~ ~ Berkeley, CA, Universit of California, De art- ' Oregon, Journal of Environmental ~Vlanage- y p ~ ment, 59,185--193. ment of City and Regional Planning, PhD is dissertation. Cape Ann Economics X2003) Land values and open I ~ s ace-beon Coun .San Francisco CA Trust Miller, A. R. (2001} Valuing open space: Land econ- > > P ,'',I for Public Land. omits and neighborhood parks, Cambridge, MA, MIT Center fvr Real Estate. ' Correll, M. R., Lillydahl, J. H. and Singell, L. D. ~ ; ~ ~ X1978} The effect of greenbelts on residential More, T. A., Stevens, T. H. and Allen, P. G. (1982} The property values: Some findings on the politi- economics of urban parks: Abenefit/cost cal economy of open space, Land Economics, analysis, Parks and Recreation, August, 31-33. 542}, 207-217. More, T.'A., Stevens, T. H. and Allen, P. G. (1988) Crompton, J. L. X2004} The proximityprinciple: The Valuation of urban parks, Landscape and i I impact of parks, open spaces and water Urban Planning, 15, 139-152. Nelson, A. C. 1986 Usin land markets to evaluate j ' features on residential property values and the ~ ) g property tax base, Ashburn, VA, National urban containment programs, American Plannin Association Journal, 52,156-171. Recreation and Park Association. g ~il~ Ernst & Young and New Yorkers for Parks X2003} Nicholls, S. X2002}Does open space pay? Measuring ~ Analysis of secondary economic impacts of the impacts of green spaces on propertyvalues IVew York City parks, New York. and the property tax base, College Station, TK, I: Texas A&M Universit PhD dissertation. 1 I,~ Fox, T. (1990} Urban open space: An rnvestmentthat y~ pays, New York, NY, The Neighbourhood Nicholls, S. and Crompton, J. L. (2005) The impact of ~I' Open Space Coalition. greenways on property values: Evidence from Austin, Texas. Journal of Leisure Research, ,,j Hagerty, J., Stevens, T. H., Allen, P.G. and , ~ ~ More, T. A. X1982) Benefits from open space 373), 321 ~-341. ; and recreational parks: A case study, Journal Ready, R. and Abdalla, C. X2003} The impact of ~ I of the ,northeastern Agricultural economics open space and potential Local disamenities on it ~ - Cor~ncil; 11.1), 13-~20. . - resde~nttal property values rn Berks County, i ' iii I Hammer T. R. Cou hlin R. E. and Horn Pennsylvania, State College, PA, Department E. T.N, (1974) Research report: The effect of Agricultural Economics and Rural Socio- of a large park on real estate value, logy, The Pennsylvania State University. ' ~ ' Journal of the American Institute of Planners, Sainsbury, J. C. X1964} The impact of urban parks j~ 40, 274-277. ~ on surrounding residential areas: A case ~'i I stud Seattle, WA, Universit of Washin ton, Henden, W. S., Kitchen, J.W. and Pringle, B. (1957} y~ y g I, ; The sociological and economic impact of urban MS thesis. ~a parks in Dallas, Texas, Lubbock, T~, Texas Sielski, D. M. (2002) The impact of parks on resrden- Tech University Press. tial property values: An analysis of two parks ~i'. Irwin, E. G. X2002) The effects of open space on within Washington County, Wisconsin, Wl, ~ residential property values, Land Economics, University of Wisconsin, Master of Urban 78, 465-480. Planning Program. ~~i .k ~Jaurnat o~ f Leisure Research (.rr~iy~r~ght ~(lUl 2001, YaI. 33, Na. 1, I-31 ,'V'atrnnud Rer~reatinn and Park Associat~an • r~i~ es The Impact of Parks on Property values: A Review of the Empirical Evidence Jahn L. ~arompton Department of Recreatic~r~, Park and Tourism Sciences, Texas A&M ~lniversity The real estate market ccfr~sistc~ntly demonstrates that ~z~ajiy people are willing to pay a larger amount fc~r a property located close to a park than for a house that does not offer this amenity. The higher value of these residences means that their owners pay }nigher propert~~ taxes. In many instances, if the incre- mental amount of taxes paid by each property which is ~~ttributahle tca the pres- ence of a nearby park i~ agl;~ egated, it is 5uffir_1er~t to pa}' the ar~~~l:al debt charges required to retire the bonds used to acq~~ire and develop the park. This process of capitalizatio~~ of lark land ia~to the value of nearby properties is termed the "proximate print ipEe." Results of approximately 3i1 studies which have empircrally investigated the extent arad legitimacy of the 1}roximate principlE? are reported, starting with Frederick Law C~lmsted's study of the impact of New Fork's (~entral Park. Only five studies were not sul~porti~~c~ cif the proxirr~ate prinriplf~ a~~d analysis of thez~~ suggested th~lse atypical results may he attributable to methodological deficien- cies. As a point of departure, the studies' results suggest that a positive impact of 2()% on property values al~lrtting or f~anting a passive park area is a reason- able starting point. If it is a }~eavily ►ised park catering tc} large n~aml~ers of artive recreation users, tlaen the proximate' value ~17cIrATTIe~~t may be minimal on abutting properties, bc~t ~~~~cy reach l ~3`%, can proper~t~es ts~~c~ car the ee blocks away. KEY'Wt7RD5: Pans, a~P~r1 s~ru•~~, ~iro~ert~y r~rllu~s Introduction The difficult fiscal f'nvit~onment that prevails in many cities, and the escalation of urban land value's, have made the economic justification of park land and open space incx~easirxgly necE~sSary~ in c~rdet- to rebttt thc~ persuasive rhetoric of those who say: "l am in Iit~°or of parks and open space beet we cant7ot afford the calaital costs of acc~taisitioYt at~d tlevelopt7tent becattse of more pressing prioritieti, or the loss of operational reven~xe that will accrue if the land is removed from the tax r~c~lls." C►overr~tllent officials often seek to enhance the tax bases of tltc~ir comm~inities by encouraging dc~velopmer~t. There is a widespread belief that this stt~ategy raises additional revenues from property taxes, which tht'r~ r<c~~ be used to itnprc~ve c~otYlnlurtity services with- out increasing, the taxe4 cif existing retiidents. 7'he nutioix that development brings prosperity is def'ply et~tbedded in the ArYtet icon psyche. Its c~ot~trast 1 Copyright ~ 200 All Rights Reserved ~ c,ROMPTa~v to the enhanced tax revenrxes accruing from development, contemporary conventional wisdom among many elected officials and decision makers is that open space and park land is a costly investment from which a community receives no economic return. The social merit of such investment is widely accepted, but social merit amenities frequently are regarded as being of sec- ondary importance when budget priorities are established. Advocates of park and open space provision view this economic concep- tualization of parks as flawed. Tlrey exhort the adage that much of the value of properties on the tax roll is acquired from amenities that are off the tax roll, and that the contributions of these amenities to the tax base are likely to be at least as substantial as those forthcoming from residential real estate developments. This paper reviews empirical e~ridence in the literature relat- ing to three key questions: ~ 1 } Do parks and open spaces contribute to in- creasing property values the proximate principle}? What is the magni- tude of this effect? and ~3} How does distaYlce effect the proximate principle? The Basin Principle The premise that parks and open space have a positive impact on prop- erty values derives from the observation that people frequently are willing to pay a larger amount of~ money for a home located close to these types of areas, than they are fc~r a comparable home further away. If this observation is empirically verified, then owners of the enhanced property are likely to pay higher property taxes to goverlllllf ~1tS heL'aLlse C)f the increase in the property's appraised value. In effect, this represents a "capitalization" of park land into increased propertyT values fc~r proximate land owners. Conceptually, it is argued that the competitive market will bid up the ~ralue of propert~~ just equal to the capitalized value of the benefits that property owners perceive they receive from the presence of the park or open space. F~conomists refer to this approach as "hedoxric pricing." It is a means of inferring the value of anon-market resource ~a park} f~ronl the prices of goods actually traded in the market place ~surroundirlg residential properties) . Iri some instances if the incremental amount of taxes paid by each prop- erty that is attributable to the presence of the park or open space is aggre- gated, it will be sufficient to pay the annual debt charges required to retire the bonds used to acquire arld develop the park. Irl these circumstances, the park is obtained at na long-term cost to the jurisdiction, This principle is illrzstrated by the hypothetical .~0 acre park shown in Figure ] . It is a natural, resorxrce oriented park with some appealing topog- raphy and vegetation. The cost cif acglziring and developing it fencing, trails, supplementary planting, some landscaping) is ~20,OOt) an acre, so the total capital cast is million. The annual debt charges far a ~l) year general obligation bond on ~1 r~lillion at are approximately $90,000. A projected annual income stream to service the bold debt was calcu- lated using the following assumptions: Copyright ~ 200 All Rights Reserved THE IMPAGT aF PARKS UN PROPERTY VALUES 3 Zone C Zone B Zane A I ,Z I ~ yds 50-acre Park a Figure 1. Layout of a 54 acre Natural Park and the Proximate Neighbr~rhood Area • If properties around the park are ~,U04 sq ft homes can half acre lots ~4D yd ~ fiD ydl with 4D yd frontages on the park, then there would be 70 lots in Zone A ~3D lots along each of the 1,2~D yd perimeters and 51ots along each of the 2a4 yd perimeters}. • Assume total property taxe4 payable to city, county, and school district are ~%a of the market value af~ the property. • Assume the market value cif' similar properties elsewhere in the jurisdiction beyond the immediate inlluence of~ this park is $2DD,DDD. • Assume th~• desire to live c~lase to a large natural park creates a willingness to pay a premium af~ ~D%~ for properties in lone A; l D% in lone B; and 5%, in Zone C, and that there are also 7D 105 itl lanes B and Table 1 shows that, give~z the above assumptions, the annual incremental property tax payments irz the three zones from the premiums attributable to the presence of the park amount to ~98,D(1Q. This is sufficient to pay the $9D,DDD annual bond debt c harges. The flows of this in~restrnent cycle are shown in Figure 2: ~i} the council invests ~9D,Dl)D a year f~~r eta years ~ar~nual debt charges an a $1 million bond} to construct ar renovate a park; iii} which causes the values of~ prop- erties proximate to the park to increase; (iii} leading to higher taxes paid by the proximate prope~~ty c~~vners to the council; (i~~} that are sufficient to fully reimburse the $9~~,(~00 annual financial investrrle~~t Ynade by the cc~lcncil. There are five adciiticjilal points worth ~~otil~g which nay further strengthen the economic casrt~. First, this illustration assumes no state f7r fed- eral grants are available to aicl in the lark's acquisition and development. If they were available to reduce the community's capital outlay, tht~n the incre- mental property tax il~come stream would greatly exceed that required to Copyright ~ 200 All Rights Reserved 4 CR~MPTUN Table 1 Praitrerty Taxes Pay the Annual Deft for Arquisitaons and the Uevelapment of the Park Ag~re~;ate Amount cif Incremental [ncrernental Property Tax Market ~'dluc~ Total Property Taxes Increments Value of attributed tca }'r-c~perty Attr~l~uted to [given Zone Each Hc~~aae the Park Taxes at ~%r tl~r~ Park Horrre Sites t)utsrde the park's X20{},{?{ll} X11 ~~,{}{1(} ~{l rnfluenre A premium) $24{},{)(){} ~,5(}{} ~S[)0 $~6,ooU B ~i(I% premium} ~2~f},()(1{) ~2{),[)(}[) $x,40{} $4{}4 ~~5,000 (S~In premium) $2I(~,{){}{) $I{),[){}0 $4,2[?~ ~I4,OOU ~98,D{)~ C1TY C+QUNCIL Council is fully reimbursed its Council invests $90,DOO per $9Q,UQ0 annual financial investment by the incremental yew to service construction or renovation of a park increases Annual property taxes paid by proximate properties to the council Values of properties proximate to incrementally increase the park increase ~igrcre Z. The Investment Cycle r1,5sr~caated with a Local (Tovernment's Investment in a Park Copyright ~ 200 All Rights Reserved THE IMI'A(~T PARK,S UN PRUPERT~' ~.ALi'ES 5 service the del}t paymel~t5. Second, the incremental propc~rry tax Income will continue to accrue to thc• carrlmunityr after the 20-year pf~riod during which the debt charges will be repaid, at Which time the net return tc~ the com- munity will be substantially el~l~anced. Third, there is e~~idel~ce tt~ suggest that investrnellt in parks affects the comparative advantage c~l~ a ~ community in attracting future businesses and desirable residential relocatol-s such as retirees ~Cromptc~I~ et al, 1997}, How- ever, the proximate c,rpit~~liz~ition apprclach does r~c►t capture thFA sec•c~ndary economic berefits attrlhutalrle to park provisiU« that accrue frr~m such sources. Fourth, a park ol~ the sire shown iii Figure 1 is likely to improve the quality t~f life and, thus, Dave some ecc~Iaomic value to txrlaan resident,5 living beyond Zone C. 1n all she studies revif~wed in this paper', the capitalization of benefits ceased at a sr~lect~~d distance, lxsually son~ewl~ere l~etWt~en 50() feet and 300(} feet away frolzl the bark per•irlleter in tlrb~in ccfntexts. Hawe~~er, it is unlikely that park user's alicl beneficiaries will he rests tcted only to thane individuals located within Sue°li a narrowly defined sc~r•vice area ~l.yt~I~, 197}. Finally, there is corlvit~cillg evidel~cf~ that tl~e public costs asscyciated with residential de~~elopment F~xcc~r~d the public revenues that accrue from it by, on average, approximately 1 r~`Io ~Crarnpton, iI~ press}. ~I'hlis, if the annual tax yield to a c•ommttniry was ~1 nlillioa~ f•r•am a resirlelltial develcApment, the median cost of sei~Ticing it is likely to lie X1.15 million. 1rI this case, i#~ the operation and maintenance ~ cysts associ~rted with Ilsillg the land tin a park or open space were less than $ l ~i0,000, then it would he mclre cast effective use of the land for the ccjrrlnruniry thala residel~tial develc~pnlel~t. A determining factc~►~ of the magl~itude cif a park's ilnpact on the prop- erty tax base is the extent of the park's r•ircumferencc ur c~dl;e (I.cttle, 1980}, if a 100 acre 17ark is circular irr shape, then it has a relatively small circum- ference. if thc~ 1 all ac rc~s is distributed mare linearly, tllerl the amount of edge increases substantially. The principle is illustrated by the calculations in Figure 3. The increased amount of edge mearxs that mare property can be sited adjacent to thc~ bark and the aggregate ellharlcement Value of the property tax base is likel~~ to he larger. This edge prinrilrle has been widely A circular park that iti 10(} acres in area will have a radius cif 1,177.3 feet. Given that the circumference of a circle is two times pi, times the radius ~~~rr}. the arnotrnt of edge will be 7,39fi.7 f eft. Assume this park is t~ril~eelf~d into a lc~ri~; strip raf grt~c+n which is one• square acre wide ~~U9 feet}--~in effect, layi~i~, one acre next tc~ another in a line. To had the length of the edge of ]O() ~ICres in this rcar~figr~ration Z(}{~ feei is triultiplied by 1Q0 times two, since there are two sides to this strip. The result is 91,S11f1 linear feet, 5.fi5 times as much edge cc~mparryd with a circular park of~ the same ii~~mher ref awes. That is the edge effect. Source: Little, (a. E. (1990 . Fi~'ure 3 illustrating the Edge Ef°fect Copyright ~ 200 All Rights Reserved G cROM~ra~v embraced in the design of golf courses which are incorporated into residen- tial real estate developments. It is important to recognize that some types of parks are more desirable than others as places to live nearby. For example, there is convincing evi- dence that large flat open spaces which are used primarily for athletic activ- ities and large social gatherings, are much less preferred than natural areas containing woods, hills, ponds or marsh Kaplan & Kaplan, 1990}. Further, it must be recognized that there are contexts in which parks exert a negative image on property values. A useful analogy is with awell-groomed front lawn which is likely to increase the value of a Name, but if it is overgrown with weeds then the property value is likely to be diminished Fox, 199x} . This point was made by the deputy director of the Parks Council, a nonprofit advocacy organization in New York City when she observed: "we have many poor neighborhoods in the South Bronx near parks. But the parks are not helping them. If you put money into} a park, chances are that you will improve one portion of the neighborhood. But if the park does not have proper security and maintenance, it becomes a liability for nearby homes" ~Tibbets, 1998, p. 9}. Adverse impacts may result from nuisances such as: congestion, street parking, litter and vandalism which may accompany an influx of people coming into a neighborhood to use a park; noise and ball- field lights intruding into adjacent residences; poorly maintained, or blighted derelict facilities; or undesirable groups congregating in a park en- gaging in morally offensive activities. In rural contexts, the proximate presence of undeveloped public park or open space is likely to be regarded by many landowners as an asset. How ever, in same contexts it may be viewed negatively because of trespass con- cerns. Hence, many proximate landowners in rural areas post and fence their land against trespassing Gartner, Chappelle & Giraud, 1996} . A final negative impact is that appreciation of property values results in higher property taxes. Residents who have lived in a location for a long time and have no interest in selling their property, may see no personal benefits accruing to them from development or major renovation of a nearby park. Nevertheless, they are required to pay higher taxes because the appraised value of their property has increased. The conceptual outcomes discussed in the previous paragraphs are sum- marized in Figure ~ which recognizes that both positive and negative impacts an property values are possible. The top half of Figure 4 suggests that prop- erty value benefit increments associated with proximity and accessibility will decay as distance from the park increases. The lower half of Figure 4 suggests that any negative values are likely to be limited to properties in close prox- imity to the park and these will decay more rapidly than positive impacts as distance from the park increases---that is, the positive curve is likely to be flatter than the negative curve ~Li & Brown, 1980) . Thus, in the negative scenario property in the park's service area but beyond say) 500 feet is still likely to experience an increase in value, since same benefits of access to the Copyright ~ 200 All Rights Reserved THE IMPACT ClF PARKS UN PROPERTY VALUES 7 Increased Market Value of Property Increase in property valu~~ due to proximity of park ~ Location of Park ~ ' "r y ~ , ,r~`"'~~-~~~~.. 'n ~ r e Distance from Park Decrease in property varue due to proximtry to Decreased Market highly deveCaped park with nuisance factors Value of Property Figure 4. The P+~sitive and Negative Impacts of Parks on Residential Property Values. Source. Li, N. M. and Brown, H. J, X1980}. park's amenities accrue to these homeowners but they avoid the nuisance costs inflicted on those who live close to it. The early Empirical Studies The legitimacy of the proximate principle was conventional wisdom that prevailed among park professionals, landscape architects and urban planners in the early years of the twentieth century. Given his legendary, inspirational role in the architecture, design and popularisation of parks in the United States, it should come as no surprise that this conventional wisdom emerged from the work. of Frederick t.aw Olmsted. Before funding for Central Park was committed, Ulmsted explained how the proximate principle would result in the park being self-financing and his argument convinced key decision-makers. Thus, the New York City Comp- troller, writing in t85b shortly after the city acquired title to tl~e land for Central Park, said, "the increase in taxes by reason of the enhancement of values attributable to the park would afford more than sufficient means for the interest incurred for its purchase and improvement without any increase in the general rate of taxation" ~Metropolita.n Conference of City and State Park Authorities, 192fi, p. 12}. Olmsted consolidated the initial conceptual acceptance of the proxi- mate principle for Central Park by subsequently providing empirical verifi- cation of iL He was responsible for the earliest documentation of the rely Copyright ~ 200 All Rights Reserved $ CRaMPTUN tionship between public parks and real estate values (Fox, 1994} . His data are slzmmari~ed in Table 2. This documentation was widely disseminated and was a powerful weapon in the armory of early public and open space advo- cates seeking to persuade communities tc} commit new investments into these amenities. So{~n after Caentral Park was completed, the New York Parks Commission was able to assert that before the park was developed, the three wards adja- cent to the park paid one dollar in every thirteen the city received in taxes; but after its development they paid one-third of the entire expenses of the city, even though acquiring the land for [ventral Park removed 10,044 lots from the city's tax roll ~Metrc}poll tan (ac»lf erez~ce of City and State Park Authorities, 1926} . Attributi~~g all the high increase in the property values in these three wards to the park, as Qlmsted aid the New York Parks Commission claimed, was probably inappropriate and an exaggeration of~ the park's influence. It is likely that natural growth i~~ the qty's population which caused a northerly movement of people wo~~ld have created increased property values in these wards without the park. Indeed, the average values i~a other parts of the cit~l increased approximately 104% during this time period. However, if this av- erage rate of increase had beeXa applied to the three wards contiguous to Central Park then their property value wo~ilcl have l~een about X53 million; Table 2 Fred~rirk Lary Olmsted's ~ocument~lt2o72 of the Im~bact of Central Park on the Pro~berty fax Base of the "Three P~oxamate Wards Assessed value in 1$73 ~236,U81,515.(){) Assessed value in 185b ~6,429,565.UU Showing an increased valuation of $~{}9,b51,95U.(){1 The tcatal expenditlu-e for ccfn5trtictiuri, frorn May 1~', l 1357 to January 1", 1$74, i5 X8,873,671.5{) The cost of land of the Park to the c~tY~ tw 5,U28,844.1{) Thc~ cast of the Park tc~ the ctt4T is ~13,9U2,515.06 The rate of tax for the year 1$'73 is ~ 5{), yielding; ott the increase of valctatiorl as above titated, increatie of tax artzoutiting to $5,~41,29H.7Ft. Total increase of tax in three wards $5,41,298.75 The annual interest on the cost c~# laird and tml~rt~vemex~t of t1~e Pack, up to this tone, at 5~x percenl ~834,15U.94 lleduct one percent, on X399,3{){} of stock, isstu~d at five pc rc~ent 3,933.U(} 83{1,157.94 Excess of increase of tax, in three wards, over interest o~~ $4,411,14(),81 cast of land and improvements Source: Fox, "T. (199(1} Copyright ~ 200 All Rights Reserved THE tMI'ACaT t)F PARKS [)N PRC~PERT~' VALUES 9 whereas it was actually $~36 rziillion. Thus, even when this is ranstdered, the park's influen~ e remained considerable. The highly publicized hr~ancial success of~ Central Park generated calls far the scenario to be replicated elsewhere in the New York City area. For example, in a letter to the New York Tirrres in 1SS1 a ror~respandet~t noted that Central Park "has not o~lly paid, f~rtt it has been a mast prc~fita.ble in- vestme~lt, and regarded irr the light of a real estate transactic}n alo~ae, it has been a great success" ~Nc~w ~c~rk Tirries, 1532, ~Jarr. p. 3}, He went can to observe that ``those who want a reductio~r in the tax rate ar~d those who favor the movement for its e~fFact c~~r real est~rte" were nc~w ``certain" to sul7part development of future l~~trkw. As a resrclt of• the Central Park success, the letter writer advocated a l~rolac~sal to acquire and develop two rle~~ 2,000 acre parks orl the periphery crt the city before its expanding poprilati+~n reached those areas. He argued: Fou+ or five rnillians of dollars at the rrtrnc}st will b~ suffirie+,t a~~d, as experience has proved, the f~ity will a~ot gar+ly be reimbursed for the outlay, b}rt will receive in the increased tax incc~n~e collected on the enhanced vall~e of land cnntigu+>us to the proposed parks much more than will be required for maintenan+'e and c3ther accounts, leaving, as in the case of (central Park, a ~randscarl~c• prc►fit can the invest- ment gyp. 3). Similar arguments were used in many other lc7cales, as local govern- ments realized that large public parks encotrragecl ne~~~ residential develop- ment on the periphery cat a city which they believed cxpar~ded ar~d strength- ened the tax base (Fax, 1190}.The dacrimerrted evidence from Central Park established the proxinlity pri~rc:iple as conventional wisdom among plaarners and park advocates, and res~~lted iY~ it being used to ju4tify major park in- vestments in many otlrer cortununities, most notably irl r~earhy Brooklyn, in Boston and in Kansas (city, 1~1 Brooklyn, for example, it was a p~•irx~e factor in stimLrlating developr7rF~nt +~1~ the 52h acre Prospcyct Park, which C~lr~lsted and his partner Calvert V~rux also desigt~ed and built, since one raf the ~~zaitl purposes of that park w~~s tc; stimulate new real estate develop~rieilt Fox, 1990} . The #irst county park system irl the U.S. was the Essex County Park Commission i~a New ,]ersfyy which was established in 1 ~i95. Much of its early justification for park investment was based o~~ the proxi~~late prc+perty prin- ciple. rn 191 the CaotnrYlissio~X engaged a cansrritant to assess tllc~ impact on land values of four Newark parks--1~:~rstside, westside, weequahic, and Branch Brook weir, 1+.~~$}. ~1~he results are sumtl~ari~:ed in Tal~Ie 3. They showed that over a 1 ~ year period, the increased taxes Maid to the county by adjacent property owl1~'rs, wllich were attributable to the four parks, were sufficient to pay all debt charges and almost all cif the nlai~rtena~~rce casts. Similar results were reported in ~i study undet•takc°rl by a firm +~f ac- countants for the rleighl~urii~g Union ~ ~c~trrlty Park Systerll irr New Jersey in 192 The Playground, 1925) . The study focused on property adjacent to Warinarrco Park in bc~tll the (:ity of Elisabeth and the Borough of Roselle, Copyright ~ 200 All Rights Reserved 10 CRaMPT~N Table ~ ~'he Im~act of Four Ner~ark Parks on Adjacent Property Values Rate of Increase in Property Values Property Adjacent Rest of Same Adjacent Taxing Park to Parks Taxing District Districts Eastside 9 times times 2~/2 tames Westside I5 times 3 times 3 times Weequahic I4 times 7 times 3 times Branch Brook 5 times 2'/~ times 32/.~ times apart adjoins park) Source: Weir, L. H. (1928) . for the years 1922 and 1927. For comparative purposes, the study reported assessed values of the City of Elizabeth; the Tenth Ward of that city in which the park was located; and of the balance of the taxing district of Roselle, for the same years. Results of the study are summarized in Table 4. The consultants reported that the increase in assessed values in the Eliz- abeth Tenth ward outside the area adjoining the park in this period was 64.1%, 1f the area adjoining the park had increased in value at that rate since 1922, then its assessed value would have increased by only X450,000, giving a total for 1927 of $1.15 million instead of the $3.77 million shown in Table 4. The difference of $2.62 million they believed was attributable directly to the influence of the park. A similar situation was evident on the Roselle side of the park where the rate of increase for the Borough property beyond the park area was fable 4 ~'he Influence a~ f ~arinaneo Park on Adjacent Eand Values in the City of Elizabeth and the Borough oj~ Moselle ~ 92Z-I 92 7 Adjacent to Adjacent to City of Tenth Ward Park on Borough of Park in Elizabeth in Ela~abeth Elizabeth Side Roselle Roselle 1922 Assessed 83.90 16.10 x.703 7.10 1.0'1 Value* 1927 Assessed 125.13 29.05 3.770 11.5'7 x.65 Value* % Increase 49.1%a $0.4°0 436.1% 62.$% I47.0°~0 *Values are an ~ millions. Source: County parks increase property values. The Playground, March I928: 633-634 Copyright ~ 200 All Rights Reserved THE IMPACT [3F PARKS C}N PROPERTY VALUES 11 34.5%. If this rate were applied to the park area property, then the increase in assessment values from 1922 to 1927 would have been $370,000 giving a total of only $1.44 million instead of the actual total of X2.65 million shown in Table 4. Again, the differe~lce of $1.21 million was attributed by the con- sultants to the influence of the park. A subsequent update of this study reviewed the 17 year period from 1922 to 1939 Herrick, 1939} . It reported that there was a 632% increase in assessed valuations on properties adjacent to warinanco Park during this period. This was nearly 14 times the average increase of 4fi%a for the entire city during the same period cif years. The property in Elisabeth adjacent to the park which was assessed at ~7Q3,000 in 1922, rose to $5.1 million in 1939. A similar, though less spectac~~~lar, increase was shown curl lands adjacent to the park in Roselle where valuations on Land adjacent to the park increased by 257%. In the first third of the twentieth century, developments of parkways and playgrounds were considered to be as central economic, social, a~ld political issues, as the development of parks. Development acid maintenance of park- ways was a major respa~~sibility of some urban park department~~, and their positive impact on proximate land values was a primary justif~catirm for their development. The prev~~iling mind-set was that parkways were analogous to linear parks aa~d, thus, a similar premium attributak~le to their aesthetic ap- peal would be present. Fmpix~ical studies appeared to co~~firm thzs premium Nolen Sc Hubbard, 1937}. I-Iowever, it was not passible to u~~tangle the myriad of infl~zences ac~~c~until~g for the increases, and historical perspective suggests that much of the value increase was attributalale to more effective and efficient access for traffic and transit, rather than to the parkways' aes- thetics. In most communities tc~clay, the distinction between parks and play- grounds has disappeared. Typically, playground equipment is one of multiple features incorporated i~~ta the design of parks. Playgrou~~ds as independent entities are confined primarily to inner city neighborhoods where they are vestiges of a previous planni~~g era. However, in the first third c~~f the twen- tieth century, independent playgrounds were a common feature in the urban landscape. These entities were defined as, "spaces wholly designed for play, and having little or no park-like qualities" Stoney, 1927, p. 3241. It had been claimed that playgrr~unds were likely to depreciate Iand values in their vicinity, but the empirical evidence suggested this concern was generally unfounded, especially in proximate rather than abutting properties ~5toney, 1927; Feldman, 1929 .The cases investigated indicated that, for the most part, playgrounds did not retard the natural rise Qf land values. In residential neighborhoods, playgrounds tended to increase the value of prox- imate property at a greater rate than in neighborhoods where business and industry were present. These conclusions were based on the results from only two studies. However, both studies were carefully executed and were com- prehensive involving 2`~ different sites in three different communities, and Copyright ~ 200 All Rights Reserved X ~ (~R() V[PTC?N they reached similar conclusions. These characteristics suggested that a rea- sonable level of~ canfiderrce could be placed in the generalizability of their findings. The relatively small number of early studies relating to the impact of parks an property values was supplemented by many subsequent studies in later years, These reflected the car~tin~recl central role of urban parks in communities throughout tlae century. Irr contrast, the role of parkways and stand-alone playgrounds diminisheel considerably in later years, which ex- plair~s the subsequent absence of studies measuring their impact. Throughout the tirrle period cif the studies reviewed here-from the earliest days of urban park development in the 1~354s, through the 1930s-- there was an insistent, almost inviolate ccrnvictic~~r among park and open space advocates of the legitimacy of~ the proximate principle. It was conven- tional wisdom among tlre~n and was also espoused by elected c}fficials. How- ever, in many ways, these early studies creating this conventional wisdom were naive, reflecting the Lrnderdevelapecl nature crf the statistical tools and re- search designs in the early years of the field. They were limited to simple calculations of increased tax r-ec~eipts accruing from properties in praxirnity to parks, park~vays and playgrounds Fox, 19J0~ . This approach ignored the necessit~T of ~rnraveling the complicated plexus crf factors that may influence property values irr addition to parks. It was noted that these "are not rrrerely additive, but react oar each ether ar~d gray react in opposite directions in different cases" ~Noletr & Hubbard, 1 X37, p. 1 . In subsequent eras, substantial improveiner~ts were made in methods used for quantifying the impact erf parks and open space an real estate values. Statistical techniques, such as regression analysis, made it possible to identify the relative influence c~Yr property values of factors other than parks. The emergence of these analytical tools defined the er~d of the era cif "early" empirical studies rather than any specific: date, brit tlris tended to occur in the 193~s, The tooter Empirical Studies The review of later empirical studies is divided into three main sections. The first section chronologically reviews studies reporting results in urban areas, with the exceptiurr of a pioneering, pathfinding study completed in the late 1930s ~I-derrick, f93~}, these studies were all undertaken after lgfi(l, The growth in their number after this time was coincident with the increas- ing capability] of computing. Almost all of the later studies used least squares regression analysis as their primary statistical tool. Typically, property prices or assessed valuations were regressed against a measure of distance and a set of control variables which measured the contributions of other potential influences on property value as well as parks and open space. The increased sophistication of computing made feasible more complex analyses contain- ing agreater number of control variables. The key questions these analyses addressed were: Copyright ~ 200 All Rights Reserved THE IMPA[a'T PARKS ()N PRUPERTY VALI..~ES 13 (i) Did ~~arks and open space contribute to increasing property values when other potenti;~l influences on those values were also taken rnto ;account? (ii} How large was tl~e proximate effect? (iii} fiver what distance does the effect extend? A subsection re~~iews studies that did riot treat parks and open spaces as being homogeneous, but which recogl~ired there are qualitative differences among them that are likely ttr ~~esult in cliffere~~t i[~~pacts on proximate prop- erty values, Findings emerging lrom studies of parks and open sp;rces in urban areas may nc~t be generalizable to non-urban or to large state and national level parks because of differences in context, scale crr nrissio~a. For this reason, results from studies ~rnrlertakeil in thc~sc~ cu~~text5 also are reviewed in sep- arate sub-sectrons. Resr~lts f~~c~~n water based p;u•ks are ~~ot reviewed here because they add a level crf~ cr►~~zplexity to tl~e discussion that was deemed to be outside the scope of this paper. In tl~c° final section, studies are re~~iewed whose findings did not eirdor se the proximate principle. .Results f ram the Ur6c~n Strcr~ces The shift from the rudimentary early empirical studies to stranger meth- odological approaches was initiated by Herrick (I939) . His primary purpose was "to show the possibilities of a simple metlrc~ci of analysis applied to avail- able data" (Herrick, 19409 p. 9fi}. It was `~.a years bef~}re c}ther5 emulated his approach which highlighted the pioneering nature of the study. Pioneers of new methods lay definition expose ther~~selves to criticisna. Cc~lle;igues iden- tified what they believe~~ to lae signific4~rrt weaknesses in the mathematical models he de~~eloped, but at the samt~ time they ac•krrowledged, "Mr. Her- rick's paper is an interesting first apprcyach~ (~~ckerrnarl ~ Goodrich, 1940, p. 5~i} . He was the first to rise st~~tistical techniques to try at~d isolate the unique contribution of parks to property value increases vis-a-ti~is c►ther factors, 1t was an attempt to rectify the fundamental weakness inherent in the early studies of ascr~il~ing all incr~east~s to the existence of a park and disregarding the array of other factors that may have contributed to the increases, such as differences in the size, agf~ arld quality cif residences erected on lets; lat. size; proximity to a Caentral Eiusiness District, schools, or shoppirag cc~rrters; and access to other facilities d~rd amenities which ge~rerate real estate value. Her- rick (1939) used regression ~r~~alysis to identify the impact of park acreage and populatia~~ density ~~z~ real estate value in Washington, DC fc~r the 1911- 1937 period. Herrick concluded that his analyses suggested: "Mast cities could afford to have twenty to thirty percent of their areas in parks. The ten percent rule, which has been suggested, is much too low" (p. 92) . However, the dramatic findings and conclusions of this study have to be tempered by the reserva- tions expressed by critics about the applieatior~ of the regression analysis Copyright ~ 200 All Rights Reserved 14 CROMPTON Ackerman & Goodrich, 1940} . In the long term, the study's main contri- bution was its pioneering illustration of the role of statistical tools in inves- tigating this issue. Although no additional work evaluating the proximate principle was reported after Herrick's study for 25 years, the principle retained its status as the prevailing conventional wisdom through the 194Qs and 50s. For ex- ample, in their Haute builders' Nfanual for Land Develo~rment, the National As- sociation of Home Builders noted: "In the vicinity of park and recreation areas, enhanced values of building sites up to 15% to 20%a, with a high level of sustained value over the years, are not uncommon experiences" Little, 1960, p. 85}. However, in 1961 the lack of convincing scientific evidence to support such anecdotal and experiential conclusions caused William Penn Mott Jr., who at that time was Superintendent of Parks for the city of ~ak- land, to write a letter to the Caro Foundation in San Francisco stating the "need for concrete evidence to indicate that parks are good business and that the purchase of park lands for future use is good business for a city" Wonder, 1965, p. 3} . As a result of that letter, the Caro Foundation sponsored a study focused on two parks in Gakland Wonder, l 965 .The samples were relatively small, but they confirmed the positive impact of parks on the assessed values of proximate properties. The results are summarized in Table 5. Clinton Park was in a relatively affluent area, while the San Antonio Park neighborhood property values were substantially Iower, In both loca- tions, the mean assessed values which were supplied by the Tax Collector's office} of properties fronting the park were dramatically higher than those of properties located one or two blocks away from the parks. A third neigh- borhood relatively close to the San Antonio Park was used as a control area. It mirrored the San Antonio neighborhood in size, type of dwelling units, ethnic composition, median family income, and education level, but was not subject to the influence of a park. Thus, its first zone fronted on to other Table 5 The Impact of T~vo Parks in Dakland on the Assessed Values of Properties in the Surrounding Neighborhoods Properties Fronting Properties Qne Block Properties Two Blocks Name of Park the Park from the Park from the Park Clinton Park $3,41fi $2,390 $2,355 San Antonio Park $1,489 $940 $932 Control Area* $$7G $932 $1 195 *In the control area, the first zone fronted an to other houses rather than a park, so these values were not subject to the influence of a park. Source: Wonder, R. L. (I965~ Copyright ~ 200 All Rights Reserved THE IMPA(:T OF PARKS 4N PROPERTY yALUES 15 houses rather than a park. Its aggregate assessed values were substantially lower than those of the San Antonia neighborhood, but all the difrerence was attributed to properties on the block that immediately fronted the San Antonio Park. The wider availability and greater capacity of camptatizlg in the 1970s and 19HOs stimulated an increase in the number of empirical studies inves- tigating the issue. A Philadelphia study focused on seven sites, at three parks, three schools, and one school-park combination ~l~yon, 1972}. Iluring the sample years of the study, 1,75 property sales were recorded in the neigh- borhoods around the sites. As a percentage of total hauling units in each area, the sample size ranged from 12% to 25.5%a. In all seven neigl~bc~rhoods regression analyses indicated that distance from the site had an impact on property values, enabling the author to conclude, "them appea~~ to be lo- cational advantages to school and park facilities, and these advantages have been capitalized in the sale pace of nearby property" gyp. 126}. Another Philadelphia study in 1974 analyzed the impact on sales price of 336 properties in the vicinity of Pennypack Park (1-lammer, Coughlin & Horn, 1974}.This 1,294 acre stream-valley park is in Irc~rth-east Philadelphia and was surrounded by residential areas developed at a density cif approxi- mately ten dwelling units per ;rare. The area around the park was comprised of "unimaginative housing, heavy in scale with natural landscaping losing out to concrete and stotre" gyp. 275}. Based on their subjective evaluation of the area, the researchers hypothesized that "the residents do nc~t consider natural amenity to be very important" so "public open space would be ex- pected to have a relatively low effect on Land values c+~mpared to of her neigh- borhoods" (p. 275}. Despite the authors' pessimistic prognosis, regression analysis indicated that the park accounted for 33% of land value at 4d feet, This dropped to 9% at 1,000 feet and 4.2°~~ at 2,500 feet which was the peripheral limit set for the study. From these data, the authors concluded that a net increase in real estate value of X3.3 millic~~r was directly attributable to the park. The most frequently cite~~l study in this literature examined the effect of greenbelts on property values in three different areas of boulder, Colorado Correll, Lillydahl & Singell, 1978} . A total of 1,382 acres of greenbelt had been purchased adjacent to residential developments in the 10 years prior to the 1978 study. The sample consisted of properties from each area that sold in a seleci.ed calendar year which were Located within 3,200 feet of the greenbelt ~n = 82). Variables in the regression model that were believed likely tc~ influence the sales price of these single family homes were: ~i} walking distance in feet to the greenbelt; iii} age of each house; 4iii} number of roams in each house; Div} square footage of each house; ~v} lot size; Zvi} distance to the city center; and vii} distance to th~~ nearest major shopping center. The regression re- sults showed that, other things being equal, there was a X4.20 decrease in the price of residential property for every foal one moved away from the Copyright ~ 200 All Rights Reserved ~ b CR(]MPT~N greenbelt, This suggested that if other variables were held constant, the av erage value of properties adjacent to the greenbelt was 32% higher than those located 3.200 walling feet away. These results are shown in Table f~. Cane of the three neighborhoods had bee~~ able to take much greater advantage of the open space amenity in its planning than the other two neighborhoods, so the authors initiated further analyses an it. Ii1 this neigh- borhood, price decreased $10.20 for every foot one moved away frame the greenbelt. This resulted in: the aggregate property value fc~r the neighborhood being approximately $5.4 mil- lion greater than it would have been in the absence of greenbelt. This increment res~ilted in an annual addition of approximately $SUU,UUU to the potential neigh- borhood property tax reventt~~. The purchase pride of this greenbelt for the city was approximately X1.5 million, and thus, the potential property tax revenue alone would allow a recovery of initial costs in only three years (p, 215}. There is an important caveat tc~ these p~~sitive results in that Sfi%a of the $500,000 proximate increment of property tax revenue accrued to taxi~ag entities other than the city, i.e, county, school district, and other independent district~S. Thus, the incremental rettarra to the cite alone was not sufficient to pay the cost~S inclxrred by the city its purchasing the greenbelt, This creates a major policy issue. However, it should ~~ot inhibit the purchase of park and open space areas because overall economic benefits accrue to taxpayers whale revenues fund all the governmental entities. Resolution of this co- nundruYYl requires one o#' two actions. The first requires a city to he prepared to accept the inevitable criticism that is likely to occur when it raises taxes to purchase the land, k~~owing that its taxpayers indeed will benefit when rettirrl ors the invest~r~ex~t is viewed il~ the broader context of total tax pay- ments to all governmental entities. The alternative strategy is to persuade the other taxing entities to ,jointly bind purchase cif the open space areas, since all will reap proximate tax revenue increments deriving from them. A study undertaken in 'Worcester, Massachusetts, in the early 1980s ex- amined the relationship l~etweea~ four parks and the values of all properties sold within a 4,000 foot radi~~s cif each park during the preceding five years ~~r, = 170} ~Hagertyr, Stevens, Alle~l & Mare, 19f~2; Nlore, Stevens & Allen, Table 6 [value of the ,grlera~e ~Io7~se r~nr~ ~xreenhe~l Proximity walking Distance from Clreenbelt A~~e~ age Value of House 3t} $54,:373 1,()O[) 59,34 1,2f~3 49,17 2,UU() 4F~,19~ 3,~()() 41,~f1~i Sotkrce. Correll, M. R., I,illydal~l, ,J I I., & 5irigell, 1.. D. (I t~7~) . Copyright ~ 200 All Rights Reserved THE tMPAta'T Ut' PARKS CAN PROPERTY VALUES 1'7 19$x; More, Stevens & Allen, l9$$}.The multiple listing service from which the study's data were de3•ived recorded actual sale price of a house, along with information on other characteristics that might effect the sale price including lot sire, number of rcfoms, age, garage, taxes paid and condition. Distance to the park in feet was added to this set of variables. 7'he results showed that, on average, a hoarse located ~0 feet from a park sold for X2,675 more than a hcause located 2,()f}0 feet away. However, SO°~, cif the aggregate increase iI~ value derived fi•urrl properties located wirhil~ 50(} feat af` the parks. Effects could not be traced bc~}fond 2,0[}0 feet frond the parks, Using these data, it was estimated that the aggregate property value increase attributable to these parks was X3.5 Illtllr031. The impact of two 1}arks on the values of proximate residential devel- opments in Dayton al~d in Colrlmbus, Ohio was reported in I985 (Kimmel, 19$5). The 170 acre (aox Arboretum in Dayton was a wooded open space containing specialized herb, ornamental and other plant gardens. Its impact on an adjacent fairly new subdivision of 300 properties was assessed. The 152 acre Whetstone Park In f aulumbus, c~or~tairled ball-fields, trails, natural areas and a 13 acre rose gardr3~, and it was adjacent to dIl older ~~esidential area. In both cases, samples caf approximately 1(}0 I•esidences we~•e used in the study. The regression analyses i3~dicated that for every additiunal loot o#' dis- tan~e a property was located away from ~jox Arboretum and Whetstone Park, the selling price decreased X3.$3 and X4.$7, respectively. The average dis- tance of properties in the study areas wc~I•e $1 ~ feet and y73 feet from [;ox Arboretum and whetstone P~tr~k, respectively, and these properties ~rielded proximate prercliums of ~:3,10~} and $4,7t}0. Givers the average selling prices of properties in the residential areas were X5$,800 and ~64,Oa0, the park premium represented 5.1:1% iIl the Gox Arboretum subdivision ~rnd 7.35% at the 'VI+'hetstone Park Iesidential area. 11l neither case was an ~rssessment made of how this average premium varied between properties immediately abutting the parks and those located say} 2,(}00 feet away, which presumably were much less impacted by the parks. An empirical investigation in Salem, t.)regoli, in 10$0 reported that open space in the form of greenbelt at the fringe of the urban area exerted an influence on urban land v~rlues that extended inward from the urban bound- ary about .x,00(} feet ~Nc~lson, 19$6) . The researcher concluded that urban land adjoining farmland a~,oned exclusively for agriclrhure was worth $1,200 per acre mare than similar la~lc~ I,000 feet away. The Influence of Different Park !)esi~~rt dn~ Ise C~harr~cteri.stir,s While the above studies ct}rlsistently reported that parks and open space had a substantial positive i mpac t on proximate property values, other studies have refined this conclusion l~~r identifying differences in the magnitude of this impact based on a bark's ;rttributes. These differences pertained to ~i) whether a park was desi~;3xed to service active rec•rcxatiorl users ~~r to offer Copyright ~ 200 All Rights Reserved 1 S C~RUMPTUN users a more passive, contemplative experience; and iii} whether a park was easily visible from adjacent streets or was sufficiently obscured from public view that it encouraged anti-social behavior. Results from an early study undertaken in the city of Spokane, Wash- ington, are shown in Table 7 Sainsbury, 1964}. This was a relatively naive study devoid of sophisticated statistical controls, but it was the first to em- pirically verify a continuum of effect between active and passive parks. Parks were classified into the three categories of active, combined active and pas- sive, and passive. The values ~~f residential properties adjacent to ar sur- rounding parks were positively impacted regardless of the type of park, and magnitude of the impact declined with distance from the parks. However, there were substantial differences in impact along the active/passive cantiz~- uum with active parks exercising the [east positive impact and passive parks the most positive impact. A more detailed study with better controls pertaining to this issue was undertaken soon after in Dallas Hendon, Kitchen & Pringle, 1967}. Ten parks were selected for study. The impact on properties within 500 feet of each park was compared with that on properties which were beyond 5a0 feet but still within the park's service area and gone of influence. In half of the parks the main feature was a playground, while the other five parks were larger and featured community playing fields. The data in Table 8 show that properties within 500 feet of a playground park were of lesser value than other properties beyond 500 but within the park's service area. However, the inner area values were higher than those of properties that were outside the playground parks' service areas. In con- trast, properties around the larger playing field parks were of higher value than properties that were more distant in the service area. The authors of the study stated: "In conclusion, it appears that the community playfield Tabte 7 The Impact of Different Types of RQrks an Residential .Property ~adues [;ombined Active Active Recreat~or~ and Massive Passive Recreation Areas Recreation Area~S Areas % change in adjoining + 1 d% + 33% + 7D% lots relative to average value of their census tracts % change in residential + 14% +63% blocks surrounding the parks relative to the average value cjf their census tracts Source: Sainsbury, C. (1964} . Copyright ~ 200 All Rights Reserved THE 1MPAC"I' 4F PARKS ~N PROPERTY VALUES 19 Tahle N A Comparison of 1~iean Assessed ~'alz~es of Properties Within 500 feet and Beyond 500 feet of ~'0 Parks ire Dallas, Texas Prt,pc~rties Within 50() Properties fiver 504 Feet 1~'eet Ratio: Mea~~ Mean Under 540 Assessed Number of Assessed Number of Type of Park Value Properties Value Properties {fiver 500 Playground Parks Casa View 3,637 40 128 3,773.00 485 .96 Beckley Heights 3,394 44 141 4,197.00 760 .81 Hattie Rankin Moore 1,37200 179 1,528.(}0 `341 .90 Sleepy Hollow 2,683.00 39 2,556.0(} 55 1.05 Preston Hollow 9,4.`39 04 154 11,247.40 51 fi .81 Playfield Parks Harry stone 5,0.8 (}4 195 5,040.00 707 1.40 Pleasant Oaks fi,980 00 171 5,879,x() i05 1.19 Beckley-saner 3,4`36.00 250 2,742,(10 -194 1.25 Martin Weiss 3,3`i5 40 262 3,258.110 741 1.02 Exline 2,382,00 113 2,254.41(1 594 1.Ofi Source: Hendon, S., Kitchen, W., ~c Pringle, B. ~ 19fi7). park, because of its large size, generally acts to increase property values of properties immediately adjacent to it while the playground generally de- creases the values of similar px-aperties" gyp. 74}. The authors attributed the reasons for the adverse impact can nearby property of the playground parks not only to noise and the flow of additional people into the area, but to their quality. For example, in the Preston Hollow neighborhood, the park's adverse impact was relatively strong ~2U%~. In this area property values wert~ high, $9,439 within 50q feet compared to ~I1,2o7 in the rest of the service area Table 8}. 'The authors offered the following explanation for the adverse effect: The detrimental character of the park appears to lie in its appearance relative to the rest of the neighborhood. Probably if the appearance were improved, by plant- ings or some form of redesign, the adverse erect would be diminished. It seemed to be true in all cases, that the aesthetically pleasing park done which had an attractive design, was we11 maintained, and highly Landscaped} caused an increase in property values of properties arnund the park, relative to other properties...The parks which were well shaded, well designed and were of pleasing appearance had a positive impact, while those which were poor•Iy designed had an adverse effect upon property values gyp. 74} , Added dimensions to these findings were reported in a study which employed sophisticated statistical controls (welcher & Zerbst, 1973. It fo- Copyright ~ 200 All Rights Reserved 20 GROMPT4N cured on five parks in Columbus, ahio; Audubon, Kenlawn and Linden parks were on the north side of the city, while Hauntz and Westgate were on the west side. All were located in neighborhoods comprised predarni- nately of single Family homes. However, the spatial relationships between the parks and adjacent residential properties differed in two ways. First, at Hauntz, Linden and ~~estgate, houses faced the park with a street betwee~l them; while at Audubon axrd l:~enlawn, houses backed on to the parks sep- arated from them only by a lerrce. Second, most houses had a view of open space, trees, grass etc., but those around [_,inderl Park, and part of Auduban Park looked out on intensively used recreation facilities. Prices of properties which had been sold in the previous five years that were immediately adjacent to these ~reighborhood parks constituted the de- pendent variable. The regression analysis controlled far house age, number of rooms, year of sale and lc~t size. The stlydy differentiated between property ~1} facing a park across a street; iii) backing {}rr to a park; ar~d viii} facing a heavy recreation use area or park building. ~l~'he first category was com- prised of properties facing Westgate and Hatr~ltz Parks. These homes sold far approximately more than identical properties located away from the park. In contrast, there was rro proximate premium associated with homes in the second category around Alxdubon and l~enlawn which backed on to the parks, since they sold for a similar' price to these beyond the parks' view zones. Further investigatic~r-r seeking an explanation of this finding revealed that the city's parks department received freclarent complaints from neigh- borhood residents of drinkir7g and other disturbing activities at night in Kerz- lawn and Audubon Parks. K.enlawrl Park was almost completely surrounded by private residences, so it was almost invisible from the street. Therefore, it was an excellent gathering place fc~r people who wanted to be undisturbed whether for legal or illegal purposes. Audubon Park contained aheavily--used baseball diamond, which meant that homeowners had strangers very close to their backyard for substantial time periods. This lack of privacy may have accounted for the lack of positive izxrpact on property values. Properties around Linden Park fell into the third category since the park consisted mainly of heavily used recreation facilities, such as baseball diamonds and a children's playground, rather than cif passive open vistas. These homes sold for approximately S% less than identical properties away from the park. Another study reported in 1973 sc~t~ght to identify the differential effects of four kinds of open space an property values: ~ 1 } public open space with recreation facilities ~e.g. playgrounds, athletic fields; public open space without recreational facilities (e.g. parks, arboretums, cemeteries}; ~3} pri- vate open space ~e.g. large estates}; and ~4} institutional open space ~e.g. colleges, private schools, country clubs) Coughlin & Kawashima, f973}.The analysis was undertaken irZ a large area of northwest Philadelphia. The study compared the value of properties irr census hlocks that adjoined one of these Copyright ~ 200 All Rights Reserved THE 1MPAt~ 1' Oli PARKS OIti PROPERTY ~A.I~tJ ES ~ 1 open space categories with c}ther census blocks. A total of 1.955 census blocks were included in the analysis aid they contained 300,000 inhabitants. The regression analysis included a large number of other variables that could influence property values, and it identified separately the park impacts on blacks comprised mainly 1~# homeowners and those on which renters predominated..~nong both cif these groups, access to public open space without recreation facilities was important. Accessibility to private and irlsd- tutional open space impac~tc~d hcwmeow~~er blocks licit z~ot rental blocks, while there was a positive relationship with open spate ca~~taining recreation fa- cilities ar~d rental blocks bttt nc~t homeowner blocks. Table 9 summarizes the implications of the study's findings relating to public open space with nca recreation facilities. Based on the average number of dwelling units per acre and the average housing rinit value given in the table fotatnote, t_he incre~~~c;ntal value attributable to three hypothetical dif ferent sized open space parks is computed ~~sing the analysis results. Com- putations are made for both individual dwelling units and for their aggre- gation in the four distant e zo~~es. The percentage increment attributable to the park, increases markedly ~~ith the size of the park. Th~~s, in the ease of a `~5 acre park, increments range from an average c~l~ 9,9% within 1,f~00 feet of the park, down to 4.17% in the 5,000 to 10,000 ~ec~t radil~s. Despite the low percentage inc~ ement in the outer bands, their aggregate incremental contribution to the tax base is substantial because the larger radi aYld greater width ol~ tl~e outer distance bands means that they eml~race a quantumly greater number of properties than the closer bands. The overall findings strongly supported the proximate principle, but there was one exception in that an anomalous negative impact occurred on properties which backed directly on to the park. The authors attributed this Table 9 E~fect on Pra~iert~~ Value of Public ~en Space with IVo .Hecre~~tion Far,~lities* '1"o~al Per 1hA~ell~ng ~ init S~LC~ of Park Sire of Pars. Distance tc~ Residence 1~Acre 5 AcrE• 25-Acre 1-Acre :~-Aare `~~-Acre (feet) Park Park }'ark Pay°k Park Park 0-1,0[}0 ~r~~,904 $2i75,71~8 ~498,5I3 $83.3) $:349.913 $1,2[17.05 1,x1}0-2,5(}q 43,057 215,2ri8 l,U7fi,29(} 12.9 t 64.86 324.28 2,511(1-5,0(}0 :37,148 185,7Gt() 92t~,~i99 31.E 15,117 78.34 5,000-10,0[111 :39,246 l~}[3,2513 981,292 41.83 4.14 20.69 $1'71,355 $8[13,049 $3,4134,794 *Assuming 13.8 dwelling units per ,acre, and base val~ie of average housia~g crnit is 2,185. Source: Coughlin, R. E.,& Kawa~h~ma, T (19731. Copyright ~ 200 All Rights Reserved 22 cROM~ro~v to: "abutting owners feeling vulnerable from park users, who may cross over their land and cause annoyance to the owners or even physical damage to their properties. In an attitude survey carried out concurrently with this study, 21 % of respondents rated the park poor or bad from the point of view of safety from crime, and an additional 45% rated it only fair" gyp. 277} . Finally, results from the study of four parks in Worcester, Massachusetts discussed earlier strongly supported the proximate principle Hagerty et al, 1985; More et a1,1982; More et al, 198$}. However, the authors also reported that parks with natural landscapes created the highest values in adjacent property, while property next to active recreation facilities had slightly lower values which were attributed to noise and pedestrian traffic. Following the models described in Figure 4, these negative influences quickly dissipated and property values one block away from the active parks showed a positive proximate increment. The empirical literature reviewed in this section offers evidence to sup- port the proximate value curves shown in Figure 4. Properties that face or directly abut parks which primarily serve active recreation users are likely at best to Shaw only a small positive value increment attributable to the park. This is attributable to the noise, nuisance and congestion emanating from the influx and egress of traffic and people. However, values are likely to rise substantially, and negative amounts are unlikely to be present, on properties located beyond the first block adjacent to the park. In contrast, the value of properties close to parks offering users a passive experience generally follow a classic distance decay curve with those closest to the park exhibiting the highest increments of value, There is same evidence in these studies that parks in which there is anti- social behavior may create a negative impact on properties facing or abutting them. The probability of this type of behavior increases if parks are not easily visible from nearby streets. Again, however, any negative impact is likely to dissipate beyond the first black. Findings from Non-Urban Studies Most studies measuring impact of the proximate principle have been undertaken in urban settings. Their findings may not be useful for those whose focus is at the state or national level. For this reason, studies that have been undertaken in those contexts are discussed in this and the following sub-sections of the paper. State and national parks typically are not estab- lished and operated primarily to provide benefits to local residents. Their mandate is much broader so their economic contributions are likely to arise from visito~~ expenditures in the area, rather than be captured i~~ proximate real estate values. Nevertheless, it seems likely that the proximate principle will apply, at least in some cases, even though such an impact may be per- ceived as incidental to the mission of these parks. Copyright ~ 200 All Rights Reserved THE IMPA(a PARKS UN PROPERTY VALUES 23 An empirical analysis of determinants of land values in the Adirondack Forest Preserve in New York Mate was reported ira 1978 ~Vrooman, 1975} . The Preserve is a region Krithin which privately-awned land and state-owned land are interspersed. ~f itls fi million acres, are owned publicly and one purpose of this study waw to test wlrether~ the state-awned land which will remain undeveloped irrrpac~ed the price o#~ privately-c~wraed land that was adjacent to it. T'he data cotlsisted of the sale prices of ~5~ vacant land parcels during a three year period which did ~~cft cor~tair~ l~~rrildings and were nc~t waterfront properties. The regression analysis indicated that being adjacent to state land had a large positive impact on price. The price cif such parcels was about $~0 f~er acre higher than similar parcels that were riot adjacent to state land. Given that the mean price f{~r all sites in the sample was $114 per acre, this represented a 1 ~.5~~1 incremental increase ire valrie. A 1983 study of the impact of six New Y(~rk State parks orl surrounding property values reported that iz~ foirr cases there wars no impact Brawn & Connelly, 19$3} .The authr~rs suggested two reasons which may explain these findings. First the areas lacked intense dc+velopmer~t and were characterized by predominantly mixed rural land uses, so proximate upc~r~ spacf~ had little additional appeal, Second, in areas that were developed arorind these four parks, the lots were largt~ incc~rporatirrg backyard pools and other amenities which effectively discountt~d or nullified the importance of recreational op- portunities offered by a ~rc~ark3y state park where the l~ousf~s were gold. At the rerrrair~ing tw°c~ parks, the ~rn~rlyses showed there was ~{n impact. At catkins Glen State Pi~rk for each 1f)~) feet closer tc~ thc3 park a residence was located, its selling pric e increased by $5~), while aI ~ec~waydir~ State Park the increase was per 190 feet. The authors used keewaydin Mate Park to illustrate the magr~ituclcxs of these i~rcr cmental increases un properties in the three local commurlilic~s of Town of Alexandria Bay, Villagcy of~~lexandria Bay are{1 Town of arleans where the irlcrement5 ref}resented ~%4 11~% ar~d 1 b% of the tax base reslrectiv~ly. Table 1(} shows tl~cx imp~rct of tl►ese incre- mental increases can the tax rc~v~~nues accruing to thf~ threcy communities din 1953 dollars} , A Maryland study reported in 1993 that the preser~~ation of a significant tract of forest land accounted firr at least 10%a of the value ref a house within one mile of the site in B~rltimc~re County: at least ~3% in [aarroll County; and at least 4% in 1-toward (aolrnty ~(aurtis, 1~~93}. When thcl r~rdir~s w~~s redtrcecl to a quarter mile, open space farm land accounted fi~r~ a minimum of 15% of the value of a hoErse in Baltimore (;crrrntyr ante ~i%, in (~arrc~ll ['ourlty, htrt it depressed hame values ~~y <<t least rr7 Hc~warcl t aorrn~~-. Csenerally, findings f~rr~m the nc~n-urban str~dies mirr~rr tllost~ from the urban studies in strppc~rtirrg tl~c~ proxirlrate pr~inc•iplca. l~ewlaite th{• concerns of rural landowners relating ~o adjacent plrblic l~rnds fat ilitatin~ access to trespassers Gartner et al, 199~i}, these frrrdrrlgs suggest thtit properties prox- imate to public park, forest Err open-sp~rr t~ land area likely to receive positive increments of ~ralrre. Copyright ~ 200 All Rights Reserved ~4 t.RUMPTUN Table IO The .Influence of Keewaydin State Park an the Property Tax Base and the Property Tax Revenue af' Three Local Communities* Town of Village ~f Alexaz~dria Bay Alexandria Bay Town of Urleans Average sale price af' $44,`?7`~ $41,27 $40,296 properties Number of properties "57 111]0 476 Average enhanced assessed $1,70`1 $6,780 $0,302 value of each property attributable to Keewaydin State Park Total enhanced assessed $948,482 X4,(}67,$20 $2,999,638 value Taxes paid attributable to $117,931 $63:1,237 X70,911 incremental park values Mown, village, fire/lrght district, school district, etc) * 1983 dollar values Source: Brown, T. L., & (aonnelly, N. A. X1983} The Impact of Large Federal ar ~~"fate Park ar Dpen Space Areas on the Laval Tax Base The conventional wisdom among many elected officials, especially in rural areas, is that public acquisition of land for outdoor recreation adversely effects the revenue generating capacity of local jurisdictions. The belief is that since publicly owned land is exempt from taxation, its removal from the tax rolls increases the burden on other taxpayers, and in some instances may lead to the demise of communities. A common context in which controversy on this issue arises is the acquisition anti development of new state park sites. The cumulative research findings of the studies reported in this paper to this point suggest that developing outdoor recreation amenities is likely to lead to a rise in proximate propertc~ values which will generate more rev- enue than is lost by removing the land from the tax base, Two empirical studies were identified which specifically addressed this controversial issue. In bath cases, the findings offered support ft~r the proximate principle and did y~ot support the conventional wisdom. A I971 study reported the impact c}f 15 park land acquisitions made is1 Pennsylvania by the U.S. Carps of Engineers or Pennsylvania State Parks Epp, 1971 ~ . The aggregate property values of the to~vnship in which each park was located were compared with the values of the rest of the county which were not s~~bject to the park's immediate influence. Data were derived from assessed values. The vall~es for both areas were tracked for an ~ I-year period, starting five years before acquisition of park land began. It was as- Copyright ~ 200 All Rights Reserved THE IMPA('.'T ~P PARKS C)N PROPERTY ~fAI,L`ES 25 sumed that the control sites, czrmprised of the rest of the county, gave a goad approximation of the land values that would have prevailed if the park sites had not been acquired. In I~ of the 15 park sites the total value of each township's taxable real estate was higher the year after acquisition began than it was in the previous year. At the other three sites, township land values recovered in the second, fourth and fifth years. The author concluded that these results indicated the increase in the value of l~irrd remaining curl the tax rolls more than offset the loss of taxable land caused by acquisition, so the revenue base of school districts and other local guverr~rnent entities was nc~t adversely aflected. To facilitate comparison l~etweera the park sites a~~d the contr~~l areas, a dollar value index was dr~vcaloped which established the market value in the year the land was acquired at 100. In the five years before acquisition com- menced the value index of~ land on average across the 15 park site townships was 84, while the value in the rest of• the counties was 91). For the five years after acquisition the aver~rge values for tlir~ park townships and coritral areas were 11,E and 108, respcartively. Thus, as a group, th~~ 15 park townships moved from fi~o below the cc }ritrol are~is values before acgrrisiticin, to above them after acquisition. The study's author concluded, "It seems likely that public acquisition of recreational laErd in amouirtti rrp tc~ fiU,0a4 acres does not reduce the real property tax base" gyp. ~fi}. Results of this study sirggested that the proximate principle is likely to apply to state acrd federal parks, even though much of the e~~idence reviewed in this paper refers to nilrrlicipal parks. However, in addition to proximate principle benefits, feder~rl and state lands after briirg additional revenue benefits to local governments because in same cases they receive payments ira lieu crf taxes from the fe~dei al and st<rte gavernr~rerrts. The compensatory impacts of such laayrnent5 c~ra local government rev enues were believed to explain the findings reported in a ~~370 study (Barron Sc ~ansma, 1070}, The aritlrors used multiple regression analysis to test the hypothesis that state or federal land ownership irl a fc~r~ested three county area ref north-vresterri Per~rrsylv,~nia adversely affected the fiscal capacity of local government through removal of Dart of they l~r•operty tax fiase. 'The hypothesis was rejected lrecause it was found that neither higher tax rates can private lands, nor reduced levels of per capita local government expen- ditures ~i.e. counties, townships and school districts were associated with large amounts of plrblic land, indicating that local governments were oat placed at an economic disadvfrirtage by public land prugrar~~s. Indeed, the data "appeared to indicate th~~ reverse" gyp. 370. In the three counties c ornprising the study area, tl~e proportions of state and federal land were ~iI `I , 48°~o and 17~r.. The consequences of the loss of local tax base were recognr~ecl by the ferleral government and the Pennsyl- vania State government tivlrich both provided payrxients iri lieu of~ taxes ora these lands to local jurisdictions, The arithars believed these payrxients ex- plained their results, cc~ric~lrrding that "tlre payments iri lieu of taxes effec- tively substitute for foregoire tax reverluf±s" gyp. 370}. Copyright ~ 200 All Rights Reserved 26 (;R~MPT~N ~'abie Z 1 ~'omparison of Nfean Value of Properties within 50Q Feet and fiver 500 Feet at Three Fort Worth Parks Mean Value Number c~f~ Mean Vallee X09 Number of Dit~erence C}ver 5{)i) Feet Properties beet and Clricier Properties Significant at .~1 Rc}sernont Park $5,79 l~i~ $h,562 59 Yes Marne Park 4,5Fi.i l kid 5,571 4~ Yes Eastover Park 7,355 lti5 Fi,4l{) ~9 Yes Scaurce: i-fenc~on, W. S. ~ 1970 These detailed findings were consistent with those reported by the Na- tional Park Service on the impact of two of its facilities National Park Ser- vice, 1961 } . In Dare [county, North Carolina, near Cape Hatteras National Seashore Area, the National Park Service reported that total assessed valua- tio~~ within the county nrore than do~zbled soon after the area was opened. At the same time, tax rates were red~rced from $1.00 to S6 cents per ~lOD. Similar cor~cl~isions were reported after the expansion of Grand Teton Na- tional Park in Teton Cou~aty, Wyornirig. Findin~:s IVat 5'u~~ortive of the Proximate Principle Five studies were located which reported findings that did not unequiv ovally support the proximate principle. A 1966 study used multiple regres- sion tc~ evaluate the relative influence of a combination of 14 independent variables ora urban growth patterns, irlc:lcrdirig distance to a playground or recreation area. However, this was not one of the fc~~rr variables that had a significant influence on land values Weiss, Donnelly & Kaiser, 1966}. Two studies undertaken in the late 1960s that were directed by the same researcher Hendon} rept~z~ted mixed resaxlt<s in that they offered only partial support for the proximate principle. The first site vas a two and a half block area of housing which equated tc~ a depth of five lots} around a 1[} acre park in Lubbock, Texas Kitchen & Hendon, 1967}. The area was character- ized as ``homogeneous" so the influence of other potential influencing var- iables was not measured. There were 556 properties within this zone of in- fluence of the park, and data were available for ~~30 of them. (correlation analysis explained their relationship betweej~ distance from the park and ~i} assessed value of the property; iii} sale price of properties that had been sold in the previous five years; and iii} assessed value of the land. There was a significant correlation only with tl~e last cjf these three measures, and it was a fairly small correlation ~ - .17 } . The second study focused on three parks in the city of Forth Worth Hendon, 1972}. They were: (i} Eastover Park, which was 13.5 acres sur- rounded by law to middle income residential property primarily occupied by African-Americans; ~~i} Marine Park, ~ti~hich ~~•as l~ acres with a srirround- Copyright ~ 200 All Rights Reserved THE [MPACT (?F PARKS QN PRQPER`IY VALU~~S 2'l ing population characterized ~h low to middle income and predominantly white; and (iii Rosemont Park, a community park of 30 acres bordering a large boulevard. Results are summarized in Table 11. In Marine and Rose- mont Parks, the mean values of properties within 500 feet of the parks were of significantly greater value than properties mare dista~xt from the park. However, this support fo~• the proximate principle was partially offset by the findings at Eastover Park where the direction of the significant relationship was the antithesis of that which was anticipated. Findings from a large scale study involving 18 park sites in 13 munici- palities in westc;hester Cciunty, New York were reported in 1986. Community parks of 25 acres or mare were selected through a systematic process based on a number of pre-established criteria (Yoegel, 1986}. Tlae neighborhoods around the selected parks were chararterized as being relatively homoge- neous. The 18 sites generated ~cpproximately 2,50(1 individual house price/ park relationship quantifiable data pointas. The impact of the park on three zones (termed tiers} was evaluated. Reside~~tial properties in Tier 1 were immediately adjacent to a park Tier comprised the next two rows of res- idential properties directly bel~cnd Tier 1. Tier 3 consisted of thcj two ~~ows of residential hcame plots lying behind Tier 2, that is, four and five rows from the park. Tiers 2 and 3 were pf~rceived tc~ be "cantrc~l areas." It was anticipated that thc~ findings would endorse the proximate prin- ciple, but the regression {analyses showed ~~o difference in value between those properties adjacent t~~ a + crmmrinity park and similar properties located in the other two tiers, Thc~ stl~ciy's design may account for the uc~expected result because it was different From the design used in most of the other studies reviewed. Give~a that f~~irly large community parks (at least 25 acres in size} were used in the stcrd;~, the lack o#~ a relationship r~iay havca reflected the proximity of all three tiers tc~ the park. It seems passible that the adjacent properties of Tier 1 may Dave c~xpenenced a nuisance factor which depressed any incremental value i~~crease to the level of that accr~zing to properties located 2-5 blocks away in Tiers 2 and This would be consistent with the lower curve in Figure 4, 't'here was no measr~re crf how well the prices of properties in these three tiers compared to those a greater distance away. Thus, it seems r~easonablc~ to postulate that if a control area lead been estab- lished fi-10 blo~vks away Dram tl~e parks, instead of 2-5 blocks away, then a distance decay impact oz~ residential properties may have enrergeci. Methodological limitations may also have acc~ou~lted f~~r the findings of a 1982 study which failed to validate thc~ proximate1 principle (Schroeder, 1982}. t]sing 5~iC~ randomly selected residential properties loc~cted in several communities in Du Pag~y (;ou~lty, Illinois, the study's ol~lectives were to test for a sigraifican t relationship hc~tween thc~ value c}f~ rc~sicle~xtial property and (i} per capita expenditu~ eti for parks and recreation in thcase communities; and (ii} the acreage ul~ (aa►d per 1,0()() pc~pulatiu~~. 'I"he cc~gressic~n analysis indicated no e~~idence off` ~r relationship in either case. It was sril~segt~ently suggested that ~~nappropc iate statistical pr•caceclurrs nay have c~or~tribrited to the findings of no relatic~nsl~il~ (Arthr~r, 1983}, hilt the aictl~or repected this criticism (Schroeder, 19t~3~. Copyright ~ 200 All Rights Reserved 28 c,ROM~TO~v Both variables used in this study are inadequate surrogates for capturing the value of parks in residential property values. The failure of any other researchers working in this area to adopt these operationalizatians suggests their fundamental weakness. Per capita expenditure is an input measure nat an output measure, whereas the proximate principle relates to quantity and quality of output in the form of parks and open space. It is the tangible output assets which influence the sale price of proximate properties, not dollar inputs. Both per capita expenditures and acres per 1,000 pop~rlation are gross aggregate measures which do not relate proximity of residence and park. Any evaluation of the effect of the proximate principle must by definition include a measure of distance decay between park and residence, and this is absent when these gross measures are used. In canclusiorr, ane ref the five studies reviewed in this section reported mixed results, but in two of the three parks which were investigated in it the proximate principle was supported. In three of~ the remaining studies, failure to verify the proximate principle may be attrib~~ted to unorthodox and flawed measurement measures that were used. These involved failure to con- trol for other influencing variables, are inappropriate control area against which proximate value irlcremerlt4 could be measrxred, and measures which failed to embrace the control element cif distance decay. f~oz~clusions Three key questions were posed in the introduction to the reti~iew of the later empirical studies. The first question asked whether parks and open space contributed to increasing proximate property values. Results from ~5 studies that investigated this issue were reviewed and in 2Q of them the em- pirical evidence was supportive. Examination of the five studies that did not suppart the proximate principle suggested that in four of those cases the ambivalent findings may be attributable to methodological limitations. The support extended beyond urban areas to include properties that were proximate to large state parks, forests and open space in rural areas. The rural studies offered empirical evidence to sE~pport not only the proxi- mate principle, but also to refute the conventional wisdom that creating large state or federal park or forest areas results in a net reduction in the value of an area's tax base. Six of the supportive studies further investigated whether there were differences in the magnitude of impact among parks with different design features and different types of uses. The findings demonstrated that parks serving primarily active recreation areas were likely to show much smaller proximate value increases than tht~se accommodating only passive ~~se. How- ever, even with the noise, nuisance and congestion emanating from active users, in most cases proximate properties tended to show increases ii7 value when compared to properties outside ~ park's service zone. Impacts on prox- Copyright ~ 200 All Rights Reserved THE I1~~PAC:T z)I~' PARKS C]1~' PRUPt;R"IY VAI.[.~ES 29 imate values were not likely tc) be positive in those cases where (i} a park was not well maintained; iii} park was rat easily visible from nearby streets acrd, thus, pro`~ided opportlir~ities for ~~nti-social behtrvrot; arl(t viii} the pri- vacy of properties backiri~; oar to a linear park was c()inprotnisc~d by park users. The second questian posed related to the ma~riitude al~ the proximate effect. A definitive ~erxeralizal~le answer is riot feasible ~iverr the substantial variation in both the sire, usage and design of~ park lands in the si udies, acrd the disparity in the resicier~tial areas aroitr~d them which were iris{estigated. However, same point of (lep~rrture base(] ora the firldirlgs reported here is needed for decision-rnal.ers in c()mrnrrrlities that try to adapt these results to their local context. Ta r~Yeet this need, it is srig~ested tl~at a l~ositive impact of 2()°~~ an property values ahuttin~ ()r lc-anti~r~ a l~assi~~e park area is a ~~ea- sonable starting point ~tiirielirrc~. If the park is lat•l;e ~sav uver• ~~cres}, well- maintained, attractive, and its rrse is mainly passive, their this f~gll~~e is likely to be lc)w. If it is small ara(1 e~nl~races so~rte active use, tlrerl this ~;-ui(telirle is likely to be hil;h. If` it is ~r heavily trse(l E~ark rrr(•()rp~rrtttrlr~ such i~ecreati(}~~ facilities as athletic fields crr ~c swimmirt~ pool, there th(z prcaxi~rrate ~~alue increment may be rrlir~icx~al t~r~ abutti~~~; properties 1)rrt ~rrav tea( lr lfl`,''c~ c)n properties twa or three krl()cks away, The diversity of the studtir contexts also makes it n(~rafeasihle t(~ offer a generalizable defrIrltrverl4lsjti'f`C to the (iota] grt(ssti()rt pos(~(l ire thc? irrtrodrrc- tion corlcerrrecl with thc° rlist~tr~ce over whi(~h the proximate irrlp~rct of` park land and ope11 spate extc7nds. Howevca~, there appeared tr) lie ~~Tide a~ree- rrrent that it had srrbstacr~ial irrrpact tier to 5()0 1•eet a~1(l that ir3 she case of commrrrrit~~ sized parks it extended c)tr1 to ~,f1{~~ feet. F(t~~T strrd~es tried to identify impacts bevorr(l that distance 1)ecause (~l~ tlrr cu~rrpour1di11~ corr~- plexity created by other lrotecrtially irrfltrerrc~izrg vari~rhles, wl~ich irrcrea5es as distance Pram park in(~reases. Nevertheless, in the c'asc u[ tlr(~se l~lr~er parks there was evidc~rice to sra~~;est irr~pact l)etiurrd this artificial peripheral haund- ary, since the catc•hmerxt are,r from which risers carne extenclec! be~~o~~d it Allen et al, l . Refer('~rces .~.ckermarl, F. L., & (~oc~dric h, E•. I'. 41~~(}).The rtfl'cts c}f parks upcar~ 1a11d arici read estate ~~alueti` Disc•u55rnn. 1'hN f'lrznn~ri Jr~~rF~rarrl, ti 5:~-~i~a. Allen, P [x,. Stc'vef~s, "T'. t{ , ~ V{~n~e, 1'. A. ~ l~~~i) I~~ea5urlr~~; the' r~c o~1c111i1( vtilue l~f l~rl~an parks. c auticrr~ LPZ sure Sr ~r~nr~~, "r (4) . ~fi i-471 Artllilr, L M, (19~3~) C:[am~nl•nf~ 1111 tin article, "'1'lil' Relat:1c,114h~t~ cat lc~ral perk ~u~d recreation ser~~l(1's to residentrtirl prl►pt•ity ~,tlues." Jo~trrtn! r~~ I.P15ZIi1' ~{PSerrrrh, ].i ~;i), `~~:~-~4~i. Barron, •J & ,~arrsm~l, J, Ir (l~i7i~~. lrxap~rc`t ref pr~hlil` l~r~ci k,ru~;rrir~~~ 1►~7 lc~l,~l ~~fver`~rmerrt lmanc`es rlm~rtr•arr ]ra~r7nril r,~ fI~►~rrr~l~urr~lli:rr~rrurrr~r5, a`~, :~ifii-:~i 1. Brel'hcr, R , & Br~~cller, 1~: (1!tb'~). 5,r.rrt' Il~y l~~~c~r~~h(x~}, ~1~at~rjrrr~r f ~ir~ar ~~4~tif~ral, Ck tc~lat~r, ~7t3-431, 4~3t~. I3rcawn, L., ~ ~aonnelly, N •1. ~i!1t~;~). Strzf~ ~irxrks r,[rZr~ r~srr~vrrt2ril ~~~~•t~' 71~lurs z7r ,'V~rc~ I''rlrit {?nlrlrhlished manu~l•rrpt~, (;crr«rll L~niver~~l~, {)t'partrnrr~t of Natural Rescyurl e~, lthala, N4' Copyright ~ 200 All Rights Reserved 30 [;ROMPTON Chandler, H. P. (1920}. Financing public parks, Park International, November, p. 250. Cook, E. 994) A Trust for Public Land memorandum to Rand Wentworth, April Cited in Steve Lerner and William Poole (I999} . The economic bPriefits of parks and npPn spares. San Francisco: Trust for Public Land. Correll, M. R., Lillydahl, H., & Singell, I~. D. (1975) . The effect [af greenbelts on residential property values: Some findings on the political economy of open space. Land I;ronomacs, 54 (2), 207-217. (;oughlin, R. E., & Kawashima, T. (19'73}. Property rlalues rind mien space in IVorthruest Philadelphia' An eznpiriral analysis. (Discussion Paper No, . Philadelphra: Lrniversity of Pennsylvania, Re- gional Science Research Instrtute. t:rompton, L. (in press}. Parks and Open Space: The highest and best use of public land Journal o- f' Park and Recreation Administration f;rompton, J. I~., Love, L. L., & More, T. A. (1997), Caharacteristres of companies that considered recreation /parks; open space to be rmpc~rtant rrY (re)location decisrons. Journal of Park and RecrPatiori Administration, 1 ~i (1) , 37-55 t;urtrs, R. E. (1993}. tiraluin~' UpP?J span in 1~~arylr~nd• An hedonir analtisis. Master's thesis, Unrversrty cif Maryland, [college Park. Diane, K E ~ I964}. 7hP econninir irnplirations of the I~~onal Park SystPnt irz ~Vlariraiba County. Tempe, Arizona: Bureau of Brasiness Servrres, Arr7cma State L~nrver5ity. Dr~~rsinn of Recreation Retiource Plannir~g, Natior~al Park Service (19fily 'fie eronarnirs of estalr lrshirtg ~'Vational Park Washrrrgtorr, D(~: t1.5. Government Printing Office. Duntr, H. (1912), T}re e[fe[t of park acid lxrulevard improvements on pr[perty values, Pro- rPP.din~'S Of the IfnurtePnth Annual (;onrlPntion n~ the Arr?PT~fan Associafiori of Park SupPrintPnderits, 3()-34. Editorial (1925}. (;aunty parks increase prolaerry values. ['hP Playground, March, 633-634. Epp, D..J (1971 } The effect u1 pr~blir land acquisrtron fur outdoor rerreatron on the real estate tax base. Journal of I P25urP Rcserrrrh, (1 17-27. Feldman, W. (1929) . The effect4 of playgrounci5 on land values of the "borderrng on" and "adjacent to" properties to the play~;rou~~ds icr Brouklyrr, New York and Orange, New Jersey. Playground and Rr'rrYatrori, September, 37a-354 Fox, T. (199()} (~'rhan open Spare Ari rn7~estrnPnt that days, New York. Thc~ Ncrghborhood ()pen Space (:oalrtron. [Iartner, 1~~4'. (ahappelle, D. E., ~C C;rraud, 'I'. (1996}, The influence of natural resource characterrstirs on property value: A [~as[? stirdy.,Jnurnal of 'I'ravPl Research, Summer, 64-7I. Hagerty, Stevens, T H., Allen, I'. G., & More, T. A. (195`2}. Benefit.5 f'r[~m open space and recreational parks: A cage study Journal rJ` the ~'~'orthPastPr-rr A~•rirultzrral L~ronornirs Coupril, [ 1 Hammer, T. R, Coughlrrr, R. E , & Hur n, E T., IV (1974) Researcl report: The effect of a large park on real estate value. Journal of tlae ~~rneriran IristitutP o~ Planners, 4[), ]uly, 274-?7i. 1lendo~7, VV. S (1972) The pax k as a deterrurarant of property val[tes. I,hP I~ercl Estate Appraitier September/Octcrher, 73-7~1 1lendcm, W. 5., Kitnc~yn,.J ~N, Prrnglc, B (l{1fi7) -11tP5oriolr}~lal and PfonofirrZ( 2mpaft OJurhlzn parks rri I)allac, IPxaS I.trbbock, "If xati: Texas 'l~~rh 1Jrr~vertirty Press. [1c~rr~[k, (19311). The [yffertti of parks rrpori lan[l an[l real etitate values. 1fiP Plarrrters Jourrlril, ~J (41, 59-94. Herrick, (194f1} 'I'h[~ c~ff~~cts of parlt~ upon land acrd real [xstat[~ valu[~s; (aon['lrzding discrrssron. `I'IaP I'lanrir~rs JourJial, fi (4), 9495. HuuS, R O (193?) l~inarzrJr2~- znurrirr~ial rerrPatrort. Meuasha,'~~'~sconsrn: George Barrie Publr5lrirrg (;o. kaplan. R., ~ l~rplan, S. ~199t)) "I'he r~xprrrcrrrrP n~ rtritzrre New Ycark: (aambridge University Press Copyright ~ 200 All Rights Reserved 'THE IMI'A[,T [)li' PARK ON PR~PERT~' ti'AL[~ES 31 Kimmel, M. M. (1985). Paarks r~rrd ~rr~rrYrty va~u~s: ,bra crn~artcaf ,ti~udy, art I)rryfr)ra rrrad (;adzcrrrbz~,i, f)har~ Master's thesis, Miami Uttiversit}', Oxford, [~hro Kitcheta, `N., & Herndc)n, W (1 X67) .Land ti"aloes adjacent tca all ur lyan Irel~hl)orhood park. Land Eronomtrs, •I 357-~►{~{). Lewis, N. P. X1923} Thr' ~ilannari~' o~ the rrtoderra racy. New York. ]c)lrn V4'rlev [ai, N Vl., & Brown, kl .I (198{)a , 'I~IIC rr}nei~hhcarht,od exter7ialittes at~d liedu~uc ldnu~tng prices I ar~d Erarznm~r,s, 5{i (2), 1`~`r-1~1 Little, E ~ 1969} f:iaal~n~P r,~ thr lrzr~d New Yc,rk• Pergamolt. Little, E;. ~199f~}. fyrcrnrr,rlye frfr 1'I rrr~~rrr. Balttntore: ]oho Hc)pklns L'nlti~ersity Press, Lyon, D. V4'. X1972} "I'hr s~atrrrl r~a5t,rhratarm and ,rra~art o~ f~u~rlrr ~rara[rCy r'xfir~rarlrtur(a5. Doctural dis- sertation, L~uiverslty of (:allfc►I rtia, $erkeley. Mart,n, D. X1994, Novemk)er 151 ~'t-~~il~~ r~ew ways tc) save decavlrtl; ltarks. TiiN ~~~r~~~ Yr,►-k Trna~S, p. Al {i Metrr~prrllta~l [;orlfererrce of (;tty arld State Park Aratl~urtties (1926} Parks as IrI~'estrnent5. New 'ark City. [sited in I,. Fl ~1~'eir (1928), Yrirk, A manurtl rr` rnunari~rrC aril ro2rnty darks New Fork: A.S. Barnes Mare parks for New Yc)rk ~1ti8>, ~alittary 9) [~rls~~ued letter tc) the eclitot. ~Verr~ t[,rk lirrar-s, p.:~. Mare, 'I..q., Stevens T. H., & ~,llen, N. G. (1982). rl'he eccmc)mic~ ctf ~irhan p~trkti ~ l)erre[it/cost analysls, f arks rrnd l4~rrealaorr, At~~;ust, 31-33. tilnre, `I'. A., Stevens, T. II , & <<~.Ileli, (l c388) Valrtatron of uI hall parrs I.and~rrr~x' and Urban ~'lannan~, 15, 139-152. Nelson, A. C:. (11186) L~srn~; land tnarkerAs to evaluate ttrhan colrtalrtl7lel~t pro~;ramy. ,9nar'raran Plannan~r A,ssrrcra[inn jra~rrrrrrl, Spr IrI~;, 156-171, Nolen, ] , & Hubharcl, li. ti' ~ 19'i7), F'rarkr~ay~ rrracl jrrnd rlalur'S. (;arubrici~;e. MA Har~~arcl Unt~erslty Press. Sainsbur~~, ~ 19{i4). ThY rrnpart r f r~rbrrra f~c~rk,5 ura ,ti~rrrnr~r,dara~ rr~Sartr~ntaal rrrr~a~' rrlsr titrrdti~. Master's thesis, ~?n Iversity crf Waslrrlr~;tor1, Seattle. Schroeder, T D. (198`?) Tllc• I elatiolrship of lnc al park arrc! recreatlc,tr set~~tce~ to reSldential prc►perty values. jnr~rnrrl of 1,Paslarr H~tiearch, 1~l (3), 22;x-`~3~1. Schroeder-, T D. (198;x) Use E ~1 ~rtrrl~ iple re~ressiuit in rec~re~ttrorr researc kI r1 ciiscltsslc,tr c)# sev'etdl issues ~ournr~l n~ f,r~r,tittr~ fi~~rarrlr, l5 `?4i-`'Ci{~, Slmr)n, D. E. (1976) A ptc}spe c t toy l~~trks 7,Far' f'rrblar Irat~r'e~l, ~4, Srtnlnlt~r, 2"r-39. Storey, (a ~1927~. `I"he rncrc•a~e o1 fared values alr,urld 1)lay7~IOtutds ~'i~r~PLrry~,~'r,~rrad, September, Tihbets,,]. (1998). C)pr~ra 5~lrlrr~ ~r~rtsr~r7~r~rararya. lrtrrrstrn~,r to varrr rr,rr~rrau,rlt~y 5 rrr►rrr,rnrr la~'aJth. [Aam1)rid~e, M~~: Lincoln lnstttute of Land 1'~,licv. ~'rc,c,man• D. 11.;1978}, A]1 errlptrrc al analysis cyf rfeter~mrnartts cyl lard r~altles In tlrry .~dircmdack Park. Arnrtracrrra Jonr~na~ of hrr,nr,rrtt+5 and Snrarllr)~r'y, `i7 (2}. l6~)-17i. Welcher, J. & lerhst, R. I I ~ 1 x;31, The erteE ualrtteti of Iieil;hhr)rhr)or! 1)ai k4 An ernprrical ~rr~estt~atlcan. land l;~'r,rar,rrrrrs, ~.~Y, 99-1{}~ Weir, L H. (1928) Prtrks• .~l rrarrrlrraJ rte' naz~nzrz~ial rr,ar~ rnunty ~ark~ New York: A S Bat'new Werss, S, 11 ,Donnelly, `1' ~~C 1~tisc•i F. ~196ti) Land valtte aitd land development rnlluertces tartars. Atr tlnalVtJC al)plu.tc•l~ ic>> eStahlltiltln~; policy alteruatne5 ~,rrrrd I;'ror~nrrirrs, ~12, 23{)- 23`3. Wc)ndel, R. I,. (1965). ~#ra anrxlyszs c►/ IhP a.SSPS,Sr~d T~rr~urrtaari o~J ~rrtraalr ~irr~ir~rtrr+ an ~r►nxarraaty Pn publar parks. ~lnpubltshed marrttwr rips ,San Prarlrlsc ca: [aarct Fc)rltrclatiotl. Yc)egel, A. {1986}, Art aragrrrry rnlc, the arla~ar[ of ~rrzrk [arar[ 1r,ratiorr rr~ir,ra irra~lr Jrrrrxrl~ rNSadPratra[ ~rrrr~ir~r-ty 7,u['rrr's 172 m2rtdle rryrrJ tr~,(~r xr,rr,mr' rr}rrtrrruraattY~ arc i1~rSf(I~YS~r~ (:rlrlrilti, ~'~'r'r~' I''rTrk rtr)Ctc)I~al dissertation, New Yor1~ [ rt~vers~tt, New York Copyright ~ 200 All Rights Reserved