TheCostofVacantandBlightedPropertiesinAtlanta:

AConservativeAnalysisofServiceandSpilloverCosts

FinalReport

DanImmergluck,PhD1

September1,2015

Acknowledgements:IwanttothankSaraToering,GeneralCounseloftheCenterfor
CommunityProgressforinvaluableassistance,withoutwhichthisreportwouldnothavebeen
possible.IalsowanttothankJordanWilliamsattheCityofAtlantaOfficeofHousingwhotook
theleadattheCityinhelpinggatherthedatafromdifferentagenciesthatareusedinthis
report.Finally,IwanttothankDerrickJordanfromtheOfficeofHousingandTerriLeeinthe
DepartmentofPlanningandCommunityDevelopment,andthemanyindividualsinvarious
agenciesthatassistedbyprovidingdataandfeedback.

ThisreportwaspreparedfortheCityofAtlantaaspartofacontractbetweentheCityofAtlantaandCommunity
Progress,datedMay13,2015,forservicesincludingthecompletionofaCostofBlightStudy.Contacttheauthorat
dan.immergluck@coa.gatech.edu.


TableofContents

Page

Introduction

Section1.ServiceCosts:CodeEnforcement,Police,andFire

1.1 CostEstimatesforCodeEnforcementandtheDepartmentof
CorrectionsCleanandCloseProject

11

14

23

1.2 CostEstimatesforPoliceDepartmentDispatchesAssociated
withVacantProperties
1.3 CostEstimatesforFireDepartmentServicesAssociatedwith
VacantProperties
Section2.SpilloverCostEstimatesonPropertyValuesandPropertyTax
Revenues
Conclusion:AggregatingtheServiceandSpilloverCostsDuetoVacant
PropertiesinAtlanta
CitedandRelevantLiterature

Appendix:SuggestionsforFutureDataCollection

25

27

Introduction
Theproblemofdistressedvacantpropertiesandtheblightthataccompaniesthemhasbeena
continualconcernincommunitydevelopmentandneighborhoodplanningintheU.S.
(AccordinoandJohnson,2000;Mallach,2006;SternliebandIndik,1969).Therootsofvacancy
andabandonmentattheneighborhoodlevelhaverangedfromdecliningemploymentand
population,tometropolitansprawl,toespeciallyrecentlysubprimelendingandits
accompanyingforeclosures.
Vacantpropertiesespeciallythoseinpoorconditionhavenegativeimpactson
neighborhoodsandcities.2Forexample,avarietyofstudieshavefoundnegativespillover
impactsofvacantand/orabandonedhomesonneighboringpropertyvalues.Inastudyof
Columbus,Ohio,Mikelbank(2008)foundthatvacantpropertiesreducedthepriceofnearby
homesbymorethan$4,000.InasimilarstudyofFlint,Michigan,GriswoldandNorris(2007)
determinedthateachvacantstructurewithin500feetahomereducedthehomevaluebyover
2percent.InastudyofBaltimore,Han(2014)alsofoundthatvacanthomeshadanegative
effectonnearbypropertyvalues.Vacantpropertiesarealsoassociatedwithhighercrimerates.
Cui(2010)analyzedcrimeandforeclosuredatainPittsburghandfoundthatviolentcrimes
within250feetofaforeclosedhomeincreasedbymorethan15percentoncetheforeclosed
homebecamevacant,withsimilareffectsonpropertycrime.Branas,Rubin,andGuo(2012)
foundthatvacantpropertyisamongthestrongestpredictorsofassaultamongadozen
demographicandsocioeconomicvariables.
Thenegativeeffectsofvacantpropertiestendtotaketwogeneralforms.First,vacant
properties,especiallythoseinpoorcondition,imposedirectservicecostsoncodeenforcement
units,policedepartments,firedepartments,courtsystems,andothergovernmentalagencies.
Second,vacantpropertiesespeciallypoorlymaintainedonescanimposenegative
spillovercostsonnearbyneighborhoods,includinglowerpropertyvaluesandhighercrime
rates.
Inthisanalysis,Iformulateconservativemeasuresofsomeofthechiefcostsimposedbyvacant
propertiesintheCityofAtlanta.Theanalysisisorganizedintotwomainsections.Section1
addressesdirectservicecostsintermsofcodeenforcement,police,andfirecosts.ThenSection
2estimatesthespillovercostsofdistressedvacantpropertiesonresidentialpropertyvalues,
andassociatedpropertytaxrevenues.

Vacantpropertiesincludevacantbuildingsorstructures,andvacantlots.Thefocusofmostofthisstudyisonvacant
buildings,withparticularattentiontotheeffectsofvacantresidentialbuildings,includingsingleandmultifamilybuildings.In
somecases,costsareconservativelyestimatedonlyforasubsetofvacantproperties.Forexample,duetoalackofresearchon
theeffectsofcommercialpropertiesonpropertyvalues,onlythespillovercostsofvacant,residentialbuildingsareincludedin
costestimates.

Theanalysisinthisreportyieldsaconservativerangeofbetween$1.67to$2.96millionin
annualCodeEnforcement,Fire,andPolicedirectservicecosts,incurredbytheCityofAtlanta,
thatarerelatedtovacantproperties.3Inaddition,theanalysisyieldsabest,reasonable
estimateoflossesinsinglefamilypropertyvaluesintheCityduetodistressed,vacant
propertiesof$153million,withaconservative,lowerboundof$55million.Suchadeclinein
propertyvaluestranslatesintoabest,reasonableestimateofpropertytaxrevenuedeclineof
$2.7millionperyear,withalowerboundof$985,000peryear.
Thisanalysisisnotcomprehensiveandoffersaconservativeestimateofthecostsofblightin
thecity.Forsomecosts,datacouldnotbeobtained.Amongthesearethoserelatedtounpaid
propertytaxes.UnfortunatelytheFultonCountyTaxCommissionerwasunabletoprovide
adequatedataformeasuringtheseeffects.Dataoncourtsystemcostswerealsonotavailable.
Thespillovercostsofvacantpropertiesonmultifamilyrentalandcommercialpropertieswere
notabletobeestimatedbecausetherearenoreliablestudiesavailableoftheeffectsofnearby
vacantpropertiesonthevalueofthesesortsofbuildings.Onlytheeffectsonthevaluesof
singlefamilypropertiesareconsideredinthisanalysis.
Thisstudyisalsoconservativebecause,ineachstepoftheanalysis,estimateswerecalculated
inaconservativefashion.Forexample,inthespilloverestimatesinSection2,onlyspillover
effectsoutto500feetfromadistressedvacantpropertywereconsidered,eventhoughsome
researchfindssmalleffectsoutto1,000feetormore.Moreover,onlytheeffectsduetovacant
propertiesinpoorordeterioratedconditionwereincludedinthecostestimatesinSection2,
despitethefactthatevenvacantpropertiesinfairorgoodconditionareexpectedtohave
some(albeitsmaller)negativeimpactonpropertyvalues.
ANoteonTimingoftheData
Theanalysesinthisstudywerenotallabletobeconductedforthemostrecentyear(2014).
DirectservicecostsforCodeEnforcementandtheFireDepartmentwereexaminedforcalendar
year2014,becausethesedataincludedsomeinformationonthevacancystatusofaffected
properties.ThiswasnotthecasefortheanalysisofPoliceservicecostanalysis,becausethe
PoliceDepartmentsadministrativedatadoesnottrackthevacancystatusofproperties
associatedwithdispatches.Inordertodeterminedispatchlocationsassociatedwithvacant
properties,datafrom2012werematchedviageographicinformationsystem(GIS)softwareto
thecityswindshieldsurveydatafromtheStrategicCommunityInvestment(SCI)study(APD
Solutions,2013),whichwascollectedin2012.Moreover,thespillovercostanalysisrequired
vacancyandconditiondataonlyavailableintheSCIdataset,andsoitwasconductedfor2012.

Datawerenotavailableatthetimeofthisstudytoestimateadditionaldirectcostsincluding,forexample,courtcosts
(solicitorsoffice,publicdefendersoffice,andmunicipalcourt),andcostsassociatedwithtaxdelinquencyandenforcementon
vacant/abandonedproperties.

Section1.ServiceCosts:CodeEnforcement,Police,andFire

1.1CostEstimatesforCodeEnforcementandtheDepartmentofCorrectionsCleanandClose
Project
UsingdataprovidedbytheCodeEnforcementunit,therewereover23,000codeenforcement
inspectionsin2014.Ofthese,almostall,22,862,werefromcasesinitiatedafter2010,forwhich
thedetailsofthecodeenforcementcasewereavailable.
Unfortunately,theCodeEnforcementdatadonotexplicitlyidentifyallvacantproperties,but
thecasedescriptionsdoindicateincidentswherethestructurewasclearlyvacant(cases
classifiedasopenandvacant).Thedataalsoindicateburntstructuresandvacantlots.
AnycaseincludedinanyofthesethreecategoriesisconsideredaKnownVacantproperty.
Additionally,manyinspectionswereclassifiedasjunk,trashanddebris,overgrowth,and/or
exteriorstructural.InspectionsclassifiedbyCodeEnforcementintwoorthreeofthese
categories,butnotclassifiedinthisanalysisasKnownVacant,werecategorizedhereas
LikelyVacant.AllotherpropertieswereclassifiedasLikelyOccupied,howeversome
portionofthesepropertiesisalmostcertainlyvacant.
The2014codeenforcementinspectionsbreakoutasinTable1.1.1.Justover23percentof
inspectionsareclassifiedhereasKnownVacant,withanother14.4percentclassifiedasLikely
Vacant.Thisleaves62.4percentcategorizedasLikelyOccupied.

Table1.1.1.CodeEnforcementInspectionsin2014byEstimatedVacancyStatusofProperty
(1)
EstimatedVacancyStatus

NumberofInspections

PercentofInspections

KnownVacant(2)
LikelyVacant(3)
LikelyOccupied

5,307
3,287
14,268

23.2%
14.4%
62.4%

TOTALINSPECTIONSIN2014(1)

22,862

100.0%

(1) Forcasesinitiatedafter2010;over95percentofall2014cases.
(2) Thisincludes3,992inspectionsindicatedasopenandvacant,1,037indicatedasvacantlot,and493asburntstructure.
Numbersdonotaddto5,307becausesomeinspectionsfallintomorethanoneofthesecategories.
(3) ThisincludespropertiesthatarenotKnownVacantbuthaveatleasttwoofthefollowingthreefeaturesindicated:a)junk,
trash,anddebris;b)overgrowth,andc)externalstructuralissues.

ToprovideacomplementarycheckonthenumberofCodeEnforcementcasesthatinvolve
vacantproperties,CodeEnforcementwasaskedtoprovidearoughestimateofthepercentage
ofunitsthatinvolvedvacantproperties,andestimatedarangebetween40and50percent.
Themoreconservativeestimateofthisrange(40percent)isjustslightlyhigherthanthe
portionofcasesthatfallintotheKnownVacantandLikelyVacantcategories.Therefore,8,594,
or37.6percentofthe2014inspections,providesareasonablyconservativeestimateofvacant
propertyrelatedinspections.
Table1.1.2providesarangeofestimatesforCodeEnforcementcostsbasedontheresultsin
Table1.1.1.CodeEnforcementpersonnelindicatedthatthetypicalinspectiontookonehalf
hourtoexecute.Thus,directcostsofinspectionsareequaltothenumberofinspectionstimes
onehalfhourtimestheaveragehourlyrateforcodeenforcementofficers,whichequals
$28.33,includingsalaryplusfringebenefits.4Inordertodetermineanestimatedindirectcost
chargeforcodeenforcementactivities,itisnecessarytoidentifytheassociatedcostsofnon
fieldinspectorpersonnelassociatedwithcodeenforcementactivity.Todothis,thebudgetfor
theCodeEnforcementUnitwasobtained.Aloadingfactorwasdevelopedtoreflecttheindirect
costsofnonfieldpersonnelaswellasequipment,suppliesandothermiscellaneousexpenses.
Thetotalpersonnelbudget(salariesandfringebenefits)fortheunit,excludingovertime,extra
helpandretentionbonuseswas$3,056,212.Ofthis,$1,296,506wasforthe22code
enforcementofficersinthefield.Thus,aloadingfactorof135.7%forsupervisory,
administrative,andsupportpersonnelcostswasestimated.Then,$110,380inequipment,
suppliesandothermiscellaneousexpensesweredividedbythetotalsalaryandfringeofthe
unittogetanindirectchargerateof3.6percent.Together,thesesumtoaloadingfactor,or
indirectchargerate,of139.3percent.
AsshowninTable1.1.2,thisanalysisyieldsarangeofannualcostsbetween$449,726and
$1,587,813peryearforvacantpropertyinspections,withareasonable,conservativeestimate
of$728,274annually.Thisfigureincludesonlythecostsforinspectingpropertiesanddoesnot
includeanycostsofamelioratingtheproblemsassociatedwiththevacantparcels.

The28.33perhourcostsisbasedonanannualbudgetfortotalsalaryplusfringebenefitsfor22codeenforcement
officersof$1,296,506.ThisfigurewasprovidedbytheCodeEnforcementUnit.

Table1.1.2.AnnualCodeEnforcementCostsforThreeScenariosofVacantProperty
InspectionCounts
ScenariosofVacantPropertyCases
KnownVacant
Known+LikelyVacant
Known+Likely+PossibleVacant

Inspections
5,307
8,594

Hours
6,634
10,743

DirectCosts
$187,934
$304,335

FullyLoaded
Costs
$449,726
$728,274

18,737

23,421

$663,524

$1,587,813

TheDepartmentofCorrectionsCleanandCloseProject
Inadditiontothesecosts,outsidetheCodeEnforcementunit,theCorrectionsDepartment
managesaprogramcalledInmateCleanandClose,whichisaprojectthatworkstosecure
vacantandopenbuildings.TheDepartmentprovidedcostsfortheprogramthroughMayof
2015ofjustover$107,000forpersonnelandsupplies.Adjustingthisupwardforatotalfiscal
yearresultsinanestimateof$128,400inannualcosts.

NotcountedinanyofthesecostsarethoseaffiliatedwiththeDepartmentofPublicWorks,
whichisfrequentlyaskedtocleanandcutpropertieswhoseyardsarepoorlymaintained.
Estimatesforeffectivecoststothecitywereestimatedtoexceed$200,000overaoneyear
periodfromJuly2013toJuly2014;howeverlienswerefiledfortherecoveryofthelarger
amountsandrecoverymayreducetheeffectivecoststothecity.

1.2CostEstimatesforPoliceDepartmentDispatchesAssociatedwithVacantProperties
DataonpoliceincidentswereobtainedfromtheAtlantaPoliceDepartment.Tobeconservative
andeliminatethosecallsthatmightbeviewedaslessthansubstantive,allincidentswherethe
durationfromcallreceivedtoendwaslessthan10minutesweredropped.Thiseliminated
approximately35percentofallincidentsfromfurtheranalysis.
Unfortunately,APDrecordsdonotindicatewhetheralocationaffiliatedwithanincident
involvedavacantproperty.Thus,inthiscase,GISsoftwarewasusedtomatchthelatitudeand
longitudecoordinatesintheAPDdatawithrealestateparcelsonthecityofAtlantaparcelmap.
Todothis,theAPDincidentrecordswereimportedintoArcGISandplottedusingthelatitudes
andlongitudesprovidedbythePoliceDepartment.(Averysmallshareabout2percentof
theoriginalrecordsprovidedbytheDepartmentdidnothavelatitudeorlongitudecoordinates;
theseincidentswerealsodroppedfromtheanalysis.)ThemapwasthenprojectedinArcGISin
thesamecoordinatesystemusedintheparcelshapefilefortheCityofAtlanta.(Figure1.2.1
belowshowstheincidentsplottedontopoftheparcelmapforthecity.)Oncethesetwofiles
werelinedup,itwaspossibletojointhemspatially,sothatforeachincident,anynearby
parcels(within40feet)wereidentified.Formost(79%)oftheseincidents,theincidentlocation
overlappedwith(wascontainedwithin)apropertyparcel.However,toaccountforsome
expectederrorinrecordingthelocationoftheincident,additionalparcelswithin40feetofthe
reportedincidentlocationthelocationwerealsoconsideredaspossiblybeingincidentsites.
Aspatialjoinidentifiedtheparcelsthatwerelocatedatthesamelocationasthepolice
incident.Forincidentswheretherewasnopreciselycolocatedparcel,thepresenceofavery
closeparcelwascheckedincasethereportedpoliceincidentlocationwasimmediately
adjacenttotheparcel(e.g.,onthestreetinfrontoftheproperty).Ofthe2012incidentslasting
atleast10minutes,154,235wereidentifiedaslocatedpreciselyonaparcel.Another9,965
incidentswereidentifiedaslocatedwithin20feetofthenearestparcel.Finally,another31,072
wereidentifiedaslocatedbetween20and40feetofthenearestparcel.
Table1.2.1providesdataontheincidentsassociatedwithpropertyparcelsin2012.Itprovides
threelevelsofsensitivityformeasuringthenumberofincidentsandincidenthoursassociated
withpropertiesinthecity.Itbreaksthesenumbersoutbythevacancystatusofthebuilding,
whichwasobtainedfromthe20112012citywidewindshieldsurveycompletedfortheCitys
StrategicCommunityInvestmentstudy(CityofAtlanta,2013).ThePoliceincidentdatawere
obtainedfor2012tomatchthetimingoftheSCIwindshieldsurvey.

Figure1.2.1.PoliceIncidentsPlottedonParcelMapfortheCityofAtlanta

ThetwolefthandsidecolumnsinTable1.2.1indicatetheincidentandhourcountsbyvacancy
statusasreportedintheSCIdata.However,formanyparcels,therewasnovacancystatus
availablefromtheSCIsurvey.Theseaccountforapproximately28percentoftheincidents
linkedtoproperties.Sincethereisnovacancyinformationavailablefortheseincidents,the
associatedincidentsandincidenthourswereassignedavacancystatusinthesameproportion
asthoseincidentswithknownstatus.Theseadjustedestimatesareshowninthetworight
handcolumnsinTable1.2.1.Thesecolumnsshouldprovideamoreaccuratecountofincidents
andincidenthoursassociatedwithvacantproperties.Theadjustedvacancyestimatesare
boldedinthetable.
Table1.2.1providesthreedifferentestimatesofincidentsandincidenthoursassociatedwith
vacantproperties.Thefirstestimate(inthetopsegmentofthetable)isthemostconservative
andincludesonlyincidentsthatarepreciselyidentifiedascollocatedwithparticularparcels.
Themiddlesegmentofthetableincludestheseincidentsbutalsothoseidentifiedaslocated
within20feetofpropertyparcels.Similarly,thebottomsegmentofthetableidentifies
additionalincidentslocatedwithin20to40feetofpropertyparcels.Thus,Table1.2.1identifies
arangeofincidentcountsandincidenthoursassociatedwithvacantbuildings.Thenumberof
7

incidentsfor2012rangesfrom12,919to16,756,andthenumberofincidenthoursrangesfrom
14,824to17,392.Theseincidents,then,averagejustoveronehoureachfromcallreceiptto
thecleartimefortheincident.

Table1.2.1.IncidentCountsandHoursAssociatedwithPropertiesbyVacancyStatus,2012

Vacancy Status

Incident Hours and Counts

Incidents
Incident Hours

Including only property locations identical to incident

95,913
Occupied
9,396
5,886
981
42,059
154,235

Least Conservative

Most Conservative

Vacant
No Structure
Not Visible
Unknown/blank
TOTAL

Incident Hours and Counts Assuming

Proportionate Reallocation of Incidents
w/ Unknown Vacancy Status
Incidents
Incident Hours

95,907

131,874

143,696

9,894
8,066
884
57,179
171,930

12,919
8,093
1,349

14,824
12,085
1,325

154,235

171,930

139,205

154,764

Including identical property locations and those within 20 feet

99,765
103,023
Occupied
9,684
10,121
Vacant
7,035
8,984
No Structure
1,194
1,090
Not Visible
46,522
61,884
Unknown/blank
164,200
185,102
TOTAL

13,512
9,816
1,666

15,204
13,497
1,637

164,200

185,102

Including identical property locations and those within 40 feet

116,151
113,433
Occupied

162,657

170,271

16,756
14,005
1,854

17,392
16,527
1,776

195,272

205,966

11,965
10,001
1,324
55,831
195,272

Vacant
No Structure
Not Visible
Unknown/blank
TOTAL

11,587
11,010
1,183
68,754
205,966

EstimatingDollarCostsAssociatedwithPoliceCallsRelatedtoVacantProperties
Inordertoestimatethecostsassociatedwiththeincidenthoursassociatedwithvacant
propertiestwofiguresareneeded.Firstistheaveragehourlycostofanincidenthourinterms
ofdirectstaffingcosts.Assumingthatatypicalincidentrequiresoneofficer5,andwithan
averagehourlyrateforofficers,includingfringebenefits,of$27.21,wecancalculatethedirect
staffcostsbymultiplyingthenumberofincidenthoursbythishourlycost.
Then,fullyloadedperstaffhourcostswerecalculatedusingthePoliceDepartmentsFY2015
budgetanddetailedinformationonpersonnelcostsfromtheDepartment.TheDepartment
providedsalaryandfringebenefitsbyorganizationalunitforallswornofficers.Afterexcluding
supportandadministrativeunits,thetotalsalaryandfringeforallswornofficersinthe
Departmentlikelytoworkinthefieldwasestimatedat$131,199,746forthefiscalyear.6Then,
thetotalsalaryandfringefigurefromtheDepartmentsFY2015budget($212,822,369)was
dividedbythisfigure,toyieldanindirectloadingfactorfornonfield(administrativeand
support)personnelof62.2%.Inaddition,theFY2015budgetshowedsupplies,contracted
services,andothercostsforthedepartmenttotaling$18,327,332.Dividingthisfigurebythe
totalpersonnelbudgetgivesasuppliesandmiscellaneousloadingfactorof8.6%.Addingthese
twoloadingfactorstogethergivesatotalloadingfactorof70.8%,whichwillbeappliedto
directsalaryandfringecosts.
Table1.2.2takestheincidenthourestimatesfromTable1.2.1andtranslatestheminto
estimatedcoststothecity.Itmultipliesincidenthoursbytheassociatedhourlypersonnelcosts
(salaryplusfringebenefits)forofficersinthefield.Thelastcolumnthenmultipliesthesecosts
by1.708toreflectthe70.8%loadingfactorestimateforAPD.Theresultisthatfullyloaded
costsforvacantpropertyrelatedpoliceincidentsin2012rangedfrom$688,941to$808,288.
AsinthecaseoftheFireDepartmentanalysis,thereisnoattemptheretoaccountforany
injuriesorfatalitiesresultingfromanyincidentsincludedinthisanalysis,includingany
associatedhealthcare,lostproductivity,oremotionalcosts.Thereisalsonoattemptmadeto
includeanycostsassociatedwithactionssubsequenttothedispatchperiod,includingany
courtorincarcerationcosts.

TheAPDincidentdataindicatethatover98.7%ofdispatchesinvolvejustoneofficer.Thus,oneofficerperdispatchis
assumed.
6
Officersinthefollowingorganizationalunitswereassumedtobeprimarilyadministrative,supervisoryorsupport:Chiefof
Police,AssistantChiefofPolice,PoliceAdministration,PoliceBackgroundandRecruitment,PoliceTraining,andCode
EnforcementChief.

Table1.2.2.ServiceCostEstimatesforPoliceforVacantPropertyRelatedCalls,2012
Estimate of Vacant
Property Incidents

Incidents

Most Conservative
Mid-Range Estimate
Least Conservative

12,919
13,512
16,756

Incident-Hours
14,824
15,204
17,392

Direct Staff Costs (1)

$403,361
$413,701
$473,236

Fully Loaded
Staff Costs (2)
$688,941
$706,601
$808,288

(1) Basedonaveragesalaryandfringeperhourof$27.21.
(2) BasedonindirectcostsforotherFireDepartmentpersonnel,supplies,etc.of70.8%ofdirectstaffcosts(salaryplus
fringe).

10

1.3CostEstimatesforFireDepartmentServicesAssociatedwithVacantProperties
ThecostsimposedontheCityofAtlantaintermsofFireDepartmentcostswereanalyzed
throughcloseexaminationoftheFireDepartmentsincidentreportdata,whichtrackawide
varietyofcharacteristicsnotonlybyincident,butalsobyunitrespondingtotheincident.The
FireDepartmentsdatahadtheadvantageofincludingafielddescribingthetypeofproperty
andwhetherthepropertywasoccupiedorvacant,althoughthisfieldwasleftblankina
sizeablenumberofcases.Asubstantialproportionoffireincidentresponsesoccuratparking
garages/facilities,andthesewereexcludedfromtheanalysis,aswerefiresatsiteswherethere
wasnobuildingorstructure.Datawereanalyzedfortheyear2014.TheFireDepartmentdata
showedthat,in2014,therewere584incidentsatresidentialorcommercialbuildings
(excludingparkingstructures)resultingin5,000dispatchesfromdifferentunits.Table1.3.1
providesinformationontheseincidents.

Table1.3.1SummaryStatisticsonFireIncidentsConcerningResidentialandCommercial
Buildings2014*
Occupancy

Number of Incidents

Occupied
Reported Vacant
Unreported

276
83
225

Total

584

47.3%
14.2%
38.5%

Number of Unit
Dispatches

2,683
833
1,484
5,000

53.7%
16.7%
29.7%

Number of Staff
Responding

9,135
2,907
4,951
16,993

53.8%
17.1%
29.1%

Number of Staff Hours

10,101
4,138
2,175

61.5%
25.2%
13.3%

16,414

*excludes parking structures, vacant lots, and other locations without identified residential or commercial structure

Table1.3.1showsthatwhilereportedvacantpropertiescomprise14.2percentoffiresat
residentialandcommercialbuildings,almost40percentofpropertieshadunreportedvacancy
status.Whatmaybemoreimportant,however,isthateventhisconservativeestimateof
vacantbuildingfireincidentsaccountedforadisproportionatepercentageofstafftimeand
costscomparedtooccupiedbuildingfires.Overall,justover25percentoffirefightingstaff
hoursdevotedtobuildingfireswereassociatedwiththesereportedvacantstructurefires.

ThefiguresinTable1.3.1werederivedfromFireDepartmentrecordswhichindicatevarious
characteristicsforeachdispatch,includingincidentnumber,unitnumber,location,thenumber
ofstaffinvolvedinthedispatch,thelengthoftimeofthedispatchuntiltheincidentwas
resolved,andsomeadditionalfeatures.Inmorethan15percentofincidents,thenumberof
staffonthedispatchwasnotreported.Inthesecases,theaveragestaffnumberfordispatches
fromthesameunitduringtheyearwasassignedtothatdispatch.
11

AccountingfortheLargeNumberofBuildingFireswithUnreportedOccupancyStatus
AsseeninTable1.3.1,almost40percent(38.5%)oftheincidentsatresidentialorcommercial
buildingswereatbuildingswherethevacancystatuswasnotreported.Ifweassumethatthe
shareofthesepropertiesthatwerevacantisthesameastheshareofreportedoccupancy
statusproperties,thenthatshareissimplyequalto(14.2%/(14.2%+47.3%)),whichequals
23.1%.Table1.3.2showsthechangeinthevacantbuildingtotalsunderthisassumption.While
thenumberofincidentsincreasessignificantlyto135,andthenumberofdispatchesandstaff
increase,thelengthofthesecallsareshorter,sotheydonothaveaslargeasaneffectonstaff
hours,ortherefore,ondirectcosts.

Table1.3.2.AdjustingVacantBuildingFigurestoAccountforUnreportedOccupancyStatus,
2014
Incidents

Dispatches

Staff

Staff-Hours

Reported Vacant Buildings

83

833

2,907

4,138

+ 23.1% X Unreported

52

343

1,144

503

135

1,176

4,051

4,641

Likely Vacant

EstimatingtheServiceCostsofVacantBuildingFires

Directcostsofdispatcheswerebasedonstaffandequipmentcostsforatypicaldispatchof24
firefightersofdifferentranksandatotalofsevenvehiclesofdifferentsize.Thisscenariowas
basedoninformationfromFireDepartmentpersonnel.Directstaffcostsperhour(salaryplus
fringebenefits)werecalculatedat$920.87perhourforatypicalmixof24staff,foraperstaff
hourdirectcostof$38.37perhour.Equipmentcostswereestimatedat$681perhourfor
these24staff(on7vehicles).Thusequipmentcostsperstaffhourwereestimatedat$28.38.
Then,fullyloadedperstaffhourcostswerecalculatedusingtheFireDepartmentsFY2015
budgetandinformationonfieldpersonnelcostsfromtheDepartment.TheDepartment
providedaveragesalaryandfringebenefitsbycategoryoffirefighterforallfieldpersonnel.The
totalsalaryandfringeforsuchpersonnelwasestimatedat$57,528,064.Thenthetotalsalary
andfringefigurefromtheDepartmentsFY2015budget($91,919,992)wasdividedbythis
figure,toyieldanindirectloadingfactorfornonfieldpersonnelof59.8%.Inaddition,theFY
2015budgetshowedsupplies,contractedservices,andothercostsforthedepartmenttotaling
$10,366,807.Dividingthisfigurebythetotalpersonnelbudget($91,919,992)givesasupplies

12

andmiscellaneousloadingfactorof11.3%.Addingthesetwoloadingfactorstogethergivesa
totalloadingfactorof71.1%,whichwillbeappliedtodirectsalaryandfringecosts.
Table1.3.3usesthesefigures,togetherwiththedatafromTable1.3.2,tocalculatethe
estimatedcostsforReportedVacantandLikelyVacantBuildingFires.TheresultsinTable3
indicatethatFireservicecostsforvacantbuildingsinthecityrangedfrom$388,000to
$436,000in2014.
Itisimportanttopointoutthatthesecostsdonotincludeanycostsorharmassociatedwith
fatalitiesorinjuries(andassociatedemotionalcosts,lostproductivity,orhealthcarecosts)and
donotincludeanydamagetotheproperties.Noattemptismadeheretocalculatewhatcould
besizeablemonetaryandnonmonetarycostsfromsuchoutcomes.

Table1.3.3.EstimatedServiceCostsofFiresatVacantBuildings,2014
Staff-Hours

Dire ct
Staff Costs(1)

Fully Loaded
Staff Costs(2)

Equipment
Costs(3)

Total Costs

Reported Vacant Buildings

4,138

$158,775

$271,505

$117,436

$388,942

Likely Vacant

4,641

$178,075

$304,509

$131,712

$436,220

(1) Basedonaveragesalaryandfringeperhourof$38.37.
(2) BasedonindirectcostsforotherFireDepartmentpersonnel,supplies,etc.of71.1%ofdirectstaffcosts(salaryplus
fringe).
(3) Basedonestimatedequipmentcostsof$28.38perstaffhour.

13

Section2.EstimatingtheSpilloverCostsofDistressedVacantPropertiesonSingleFamily
HomeValuesandPropertyTaxRevenueinAtlanta
InSection1ofthisstudy,thedirectservicecostsofvacantpropertieswereestimatedfor
variouscityservices,includingpolice,fire,andcodeenforcement.Inthissecondpartofthe
study,thecoststhatareimposeduponneighborhoodsandtaxpayersintheformofreduced
propertyvaluesandtheassociateddeclineinpropertytaxrevenueareestimated.Thesecosts
aretypicallyreferredtoasspillovercostsintheresearchliterature.
Theapproachhereistoutilizethesignificantamountofrecentstudiesfromothercities,
combinedwithlocaldataonvacantpropertiesindifferentconditions,todevelopestimatesof
thesespillovercosts.Thedataandtimerequiredtodirectlymeasurethepercentageeffectof
vacantpropertiesonnearbypropertyvaluesusingprimaryrealestatedataisquitesubstantial,
andanyparticularmeasurementofsucheffectsissubjecttothelimitsoftheavailabledata.
Theapproachusedheretakesadvantageofanowsubstantialliteratureontheeffectofvacant
anddistressedpropertiesonpropertyvalues.Thisstudyconductsametaanalysisofthehigh
qualitystudiesthathavebeendoneacrossdifferentcitiesanddifferentyears,andestimates
thespillovercostsonnearbypropertyvaluesduetodistressedvacantproperties,usingthe
centraltendenciesofthesefindings.
Then,actualdataonvacantproperties,brokenoutbytheirphysicalcondition,werecombined
withthesespillovereffectpercentagestoestimatethecumulativeeffectsofvacantproperties
inAtlantaonpropertyvalues.These,inturn,wereusedtoestimatepropertytaxrevenue
effects.Sensitivityanalysiswasperformedusingparticularlyconservativeestimatesfromthe
literatureinordertodevelopalowerboundonthelikelypropertyvalueimpacts.

WhatDoExistingStudiesSayabouttheEffectofVacantPropertiesonNearbyHomeValues?
Agooddealofresearchhasexaminedthespillovercostsofvarioustypesofdistressedhousing
onnearbyhomevalues,includingtheeffectsofforeclosedproperties,theeffectsofvacant
properties,andtheeffectsoftaxdelinquentproperties.Theprecisedefinitionsofvacancy,
foreclosure,andtaxdelinquencyvaryacrossstudiesduetothenatureofthedataavailableand
differencesinlocaldefinitionsoftheseterms.7Inrecentyears,thegreatestvolumeofsuch
workhasconcernedtheimpactofforeclosuresonnearbyhomevalues.However,while
foreclosuresmaycatalyzeanincreaseinvacantorphysicallyneglectedhomes,mostofthese
studiesdonotdirectlymeasuretheimpactofthevacancyorphysicalconditionofnearby

Essentiallynorecentliteraturehasexaminedtheeffectofvacantnonresidentialpropertyonhomevalues,ortheeffectof
vacantpropertiesonnonresidentialormultifamilypropertyvalues.Thus,anysucheffectsarenotaccountedforinthisstudy.

14

values.(Afewofthesestudiesdoseparatelymeasuretheimpactofvacant,mortgage
distressedproperties,andtheyareconsideredhere.)
Thefocushereisonstudiesthatmeasuretheeffectofdifferentsortsofvacant,residential
propertiesonnearbyhomevalues.Whilemanycostofblightstudiesclaimtoincludethe
spillovereffectsofvacancyorblightedpropertiesonnearbyhomevalues,asetof8studies
conductedoverthelast10yearswereidentifiedthatwereviewedassufficientlystrongto
includeinthismetaanalysisofspilloverimpacts.Someotherstudiesexaminedtheeffectsof
onlyvacantlotsonnearbyproperties,ordidnotdistinguishbetweenvacantstructuresand
vacantlots.Othersexaminedtheeffectsofparticularinterventions,suchastargetedcode
enforcementorthegreeningoflots,thatdidnotdirectlyidentifythespillovercostsofvacancy
orblight.(Someofthesestudiesmaybereferencedinotherpartsofthisreportwheretheir
implicationsarerelevant.)
Notsurprisingly,someofthestudiesexaminedhereoccurredinthesamecities.Thisispartly
becausesomecitieshavedevelopedbettersetsofdataondistressedproperties,homevalues,
andotherrelevantinformationthatareneededtoconductstrongstudies.Whiletheprecise
magnitudesofthespillovereffectsareexpectedtovarysomewhatbasedonthelocationofthe
study,thegenerallyconsistentfindingsamongthestudiesandthestudiesinothercities
suggestthattheseeffectsaresimilaracrossdifferenttypesofcities.Moreover,oneofthe
studiesiscarriedoutacrossfifteenmetropolitanareas.8
Forthepurposeshere,thekeyfindingofinterestinthesestudiesistheextenttowhichnearby
distressedvacantpropertiesaffecthomevalues.Thestudiesgenerallymeasurethedegreeto
whichadistressedpropertywithinacertainradiusofahomereducesthevalueofthehome.
Theradiiatwhichtheseanalysesaredonetendtorangebetween250and1,000feet,withall
ofthestrongstudiesidentifiedhereincludingameasurementintherangeof500to660feet
(about1/10thto1/8thofamile).Whilesomestudiesfindnegativeeffectsasfaroutas1,000
feetormore,theeffectstendtogetquitesmallbeyondthe500660footdistanceandare
ignoredhere.Thus,anyspillovercostsestimatedinthisanalysiswillbeconservatively
measuredbyignoringeffectsbeyondthisrange.Forsimplicity,wewillconsiderallestimatesin
the500660footrangeas500footestimates,anotherconservativeassumption.
Table2.1summarizesthespilloverestimatesfromtheeightstrongstudiesidentified.These
studiesusedstrongeconometricmethodstoidentifythemagnitudesofspillovereffects.Most
ofthemusedwhatarecalledspatialhedonicmethods,usingadvancedeconometricmethods

Mostofthesestudiesoccurwithinonecityoronecountybecausethesortofdatarequiredonvacantpropertiesisoften
highlylocalizedandnotgenerallyavailableacrosscountiesormetropolitanareasinaconsistentfashion.

15

Table2.1.FindingsonNegativeSpilloverPriceEffectswithin500FeetofDistressed
ResidentialStructuresinUrbanMarkets,20072015(1)

City
Whitaker&Fitzpatrick,2014
Almetal.,2014
GriswoldandNorris,2007
Whitaker&Fitzpatrick,2013
Griswoldetal.2014
Mikelbank,2008
Han,2014
Gerardietal.,2012
Mean
Range

Effectsoutto500feet
TaxForeclosed
orDelinquent
Vacant

Cleveland
Chicago
Cleveland
Cleveland
Cleveland
Columbus
Baltimore
15Metros

5.20%
3.40%
2.26%
1.80%
3.07% (2)

3.15%
1.8%to5.2%

1.80%
0.83%
1.35%
0.32%
1.30%

(2)
(3)
(3)
(4)

1.12%
0.32%to1.8%

Notes:

(1)
(2)

(3)
(4)

A few of these findings are actually measured out to distances of 660 feet, so that the effects here are conservative estimates at 500 feet.
These factors are averages of the effects found in 3 of the 4 submarkets used in this study: extremely weak, weak, and moderately functioning;
these are the sorts of neighborhoods where most tax delinquent properties exist in Atlanta. The effect in highly functioning markets is substantially
larger in magnitude (more negative) and is excluded here for the sake of being conservative in estimating spillover costs.
This is a spatially weighted average of the magnitude of the effect found within 250 feet and that found from 251 to 500 feet. The 250-foot effect is
given weight, and the 251-500-foot effect is given weight, reflecting the difference in spatial areas surrounding the distressed property.
This is an average of the magnitude of the effect found for vacant homes with seriously delinquent mortgages and lender-owned homes in belowaverage condition.

Figure2.1.RangeofNegativeSpilloverEffects(as%ofPropertyValue)to1/8mile

Spillover % on Properties
within 1/8 mile

0.0%

-0.32%

-1.0%
-2.0%
-3.0%

-1.12%
-1.80%

-1.80%

-3.15%

-4.0%
-5.0%
-6.0%

-5.20%
Tax Delinquent/Foreclosed

Vacant

16

tocontrolfordifferencesamongpropertiesandpropertylocationsotherthanthenumberof
nearbydistressedproperties.Thesestudiescontrolfordifferencesinthesize,structure,
numberofbathroomsandbedrooms,andotherqualitycharacteristicsamongdifferenthouses.
Theyalsocontrolfordifferencesinneighborhoodandlocationcharacteristics.Someuseda
hybridhedonicmethodutilizingthechangeinsalepriceasthedependentvariable(repeat
sales).9Whilenostudyisperfect,thestudiesheregotosignificantlengthstoisolatethe
spillovereffectsofdistressedpropertiestothegreatestextentpossibleusinghighqualityand
detaileddata.
Table2.1distinguishesfindingsacrosstheeightstudiesbetweenthosepertainingtovacant
propertiesandthosepertainingtaxdelinquentortaxforeclosedproperties,withthislatter
categoryoftenrepresentingprimarilyvacantproperties.Taxdelinquentortaxforeclosed
vacantpropertiesareexpectedtobe,onaverage,moredistressedthantheaveragevacant,
nondelinquentproperty,becauseownersofvacantpropertieswhoarecurrentontheirtaxes
aremorelikelytomaintaintheproperties.Conversely,taxdelinquentownersmaybeinthe
processofabandoningtheirproperties.Figure2.1illustratestherangeofthesespillovereffects
at500feet.Forvacant(nontaxdelinquent)propertiestheyrangefrom0.32percentinone
studyto1.8percent,withanaverageof1.12%.Fortaxdistressedproperties,thespillover
effectsrangefrom1.8percentto5.2percent,withameanof3.15percent.Thus,the
distressed,taxdelinquentpropertieshaveamarkedlylarger,negativeeffectonnearby
propertyvalues,whichisexpectedbecausetheseproperties,onaverage,aremorelikelytobe
physicallydistressed.
Theserangesofimpactareconservativeinatleasttwoways.First,asexplainedabove,some
studiesfindnegativeeffectsofvacancyorneglectbeyondthe500footradius.Butthese
measuresarelesscommonandthemagnitudesarequitesmall,sowhiletheymaybematerial
innature(especiallybecausemorepropertiesliewithin1,000feetofavacantstructurethan
withinthe500footradius),theyarenotcountedforthesakeofreliabilityandbeing
conservativeinestimatesofspillovercosts.Second,someofthelargestestimatesofnegative
impact(intheGriswoldetal.2014study)werenotincludedinthemetaanalysisheredueto
theiroccurringonlyinhighlyfunctioning,(thatislowerpovertyandhigherpropertyvalue)
neighborhoods.BecausethegreatmajorityofdistressedpropertiesinAtlantaarelocatedin
lowerincomeandlowervalueneighborhoods,includingsuchlargemagnitudespillover
measuresherewouldnotbeappropriateandriskoverestimatingthespillovercostsofblight.

ThestudiesutilizinghybridrepeatsalesapproachesincludeHan(2014)andGerardietal.(2012).Therepeatsalesapproach
suffersfrompotentialbiasduetoalackofinformationonimprovementstopropertiesbetweensubsequentsales(theHan
studyattemptstoomitpropertiesthatmayhavebeenflippedbutmaybelimitedinitsabilitytodoso).Thespatialhedonic
methodssufferfrompotentialomittedvariablebiasaswell,althoughofadifferentsort,althoughthesmallareaspatial
controlsminimizethisproblem.

17

Usingthisanalysis,inordertoprovideforareasonablerangeofsensitivityanalysis,the
spillovercostsofdistressedvacantpropertiesonhomevalueswillbeestimatedusingtwo
differentmagnitudesofspillovercosteffects.GIStechniqueswillbeusedtoidentifythe
numberofdistressedvacantpropertiesthatliewithin500feetofeach14unithomeintheCity
ofAtlanta.Thenusingthespillovereffectestimatesandtheappraisedvaluesofthehomes
(fromcountypropertytaxassessors),thedecreaseinvaluesofallhomeswithin500feetofa
vacanthomewillbecalculatedandsummed.10Thiswillyieldtheaggregatedecreasesinvalue
duetovacanthomes.Then,usingestimatesofassessedvalueandmileageratesfortheCity
fromFultonCounty,lossesinmarginaltaxrevenuewillbeestimated.

IdentifyingtheNumberofVacantPropertieswithin500feetofSingleFamilyHomes
Inordertoidentifythenumberofvacantpropertieswithin500feetofsinglefamily(14unit)
homes,datafromacomprehensivewindshieldsurveyofresidentialparcelsintheCityof
Atlanta,whichwasconductedbetweenDecember2011andAugust2012fortheCitys2013
StrategicCommunityInvestmentReport(APDSolutions,2013).TheSCIsurveynotonly
indicatedthevacancystatusofresidentialbuildings,italsoindicatedtheconditionof
residentialbuildings.Buildingswereclassifiedasbeingindeteriorated,poor,fair,or
goodcondition.Forthepurposeshere,vacantpropertiesclassifiedintheSCIsurveyas
deterioratedorpoorarecalleddistressed,vacanthomes.
Thefocusinthisanalysisisonthespillovereffectsofvacantbuildingsindistressedcondition.
The2012SCIwindshieldsurveyclassified2,411vacantresidentialbuildingsasbeingin
deterioratedorpoorcondition(labeleddistressedinthisanalysis),andanother5,606vacant
buildingsasbeinginfairorgoodcondition.Thelocationsofthedistressedvacantresidential
propertieswereplottedusingtheirparcelnumbersandaparcelmapshapefilefortheCityof
Atlanta.UsingArcGIS,500footbuffersaroundeachofthedistressedvacantpropertieswere
calculated.TheseareplottedagainstaparcelmapforthecityinFigure2.2.Whileallsortsof
vacanthomesaredisproportionatelyconcentratedacrossaswathofthecityrunningfromthe
northwestsidetothesouthside,thedistressedvacanthomesareevenmorespatially
concentratedintheseneighborhoods.
ByusingaspatialjoininArcGIS,the500footbufferswereintersectedwithallparcelsinthe
city.Inthisway,thenumberofbufferstouchingeachoftheapproximately80,000singlefamily
(14unit)homesinthecityofAtlantawerecalculated.Thiscalculationprovidedthenumberof

10

Thetaxappraisedvaluesofhomesmaybehigherorlowerthanthehomestruemarketvaluesandaregeneratedonan
annualbasis.Thesevaluesaregenerallycreatedwiththeuseofacomputerizedautomatedmassappraisal(CAMA)systems
utilizedbycountytaxassessors.

18

Figure2.2.500footBuffersaroundDistressed,VacantResidentialBuildings

Table2.2.SingleFamilyHomesbyNumberofDistressedVacantResidentialBuildingswithin
500Feet,2012
Rank of S-F Home by Number of Nearby Distressed
Vacant Properties

Number of Distressed Vacant Buildings

within 500 feet

Lower Quartile (25%ile)

Median (50%ile)

Upper Quartile (75%ile)

90th percentile

19

distressedvacantpropertiesthatarelocatedwithin500feetofeachhome.Duetothespatial
clusteringoftheseproperties,thesenumbersvarysignificantly,asmightbeexpected.Table2.2
showsthathomesintheupperquartileintermsofthenumberofnearbydistressedvacant
propertieshaveatleast2suchpropertieswithin500feet,andthetopdecileofneighborhoods
has5ormoresuchpropertieswithin500feet.Thismeansthat10percentofsinglefamily
homesinthecityhave5ormoredistressedvacantresidentialbuildingswithina500foot
radius.Yetmanyhomesinthecityhavenodistressedvacanthomeswithina500footradius.
Morethanhalfofthehomesinthecityfallintothiscategory.AsindicatedbyFigure2.2,these
homestendtobelocatedinthemoreaffluentnorthernandnortheasternpartsofthecity.
Someoftheliteraturereviewedforthisstudysuggeststhatthespillovereffectsofadditional
nearbyvacantpropertiesonpropertyvaluesarenotentirelylinear.Inparticular,asmoreand
moredistressedvacantpropertiesexistnearahome,thenegativeeffectsonhomevaluewill
eventuallydeclineandreachalimit.Forexample,ifhavingonedistressedvacantproperty
within500feethasa3%effectonahomesvalue,thenhavingthreesuchpropertiesnearby
mayaccumulatetoa9%cumulativeeffect.However,itislesslikelythatgoingfrom3nearby
distressedvacanthomesto9nearbydistressedvacanthomeswillincreasetheeffectby
anotherthreefold,from9%to27%.Whiletheresearchonsuchnonlineareffectsis
somewhatscarce,someworkintheforeclosureliteraturesuggeststhattheseeffectswilltend
tohitaplateauafterreachingsomewherearound10distressed,vacanthomes.Tobe
conservative,Ilimitthenegativeeffectsofdistressedvacantpropertiesto5.Forexample,if
theeffectofhavingonedistressedvacanthomewithin500feetis3%,thentheeffectof
having5isestimatedas15%,buttheeffectofhaving6isalsoestimatedat15%,asisthe
effectofhaving10vacanthomeswithin500feet.11

SummingUptheSpilloverCostsonSingleFamilyHomeValuesinAtlantaduetoVacant
ResidentialProperties
Inordertoestimatethecumulativeimpactofdistressedvacantresidentialpropertiesonsingle
familypropertyvalues,themagnitudeofthespillovereffect(expressedasapercentofvalue
pervacanthomewithin500feet,uptoalimitof5vacanthomes)mustfirstbeidentified.To
dothis,wedrawontheresultsofthemetaanalysissummarizedinTable2.1andFigure2.1
above.ThefirstrowinTable2.3presentswhatisviewedfromareadingoftheliteratureasthe
best,reasonableestimateofcumulativespillovercostsonsinglefamilypropertyvaluesdueto

11

Forsensitivitypurposes,theeffectsinTable3belowwerereestimatedassumingthatthelimitoftheeffectsisnotreached
untilthenumberofnearbyvacanciesreaches10,insteadof5.Thesizeofthecumulativespillovereffectswasnotsubstantially
largerbecause,asshowninTable2,90percentofhomesaresurroundedby5orfewerdistressedvacanthomes.Moreover,
thehomessurroundedbyhigherlevelsofvacantpropertiestendtohavelowervalues,sothedollaramountoftheeffectis
oftennotverylarge.

20

Table2.3.EstimatesofCumulativeSpilloverEffectsonSingleFamilyPropertyValuesand
PropertyTaxesDuetoDistressedVacantResidentialBuildings,2012
Cumulative Effect of
Distressed Vacant
Properties on SF
Values

Average Effect
per Distressed
Vacant
Property

Potential Cumulative
Impact on
Annual Property
Tax Revenue (2)

Assumptions

Effect of Distressed Vacant

Properties within 500 Feet
on Single-Family Values (1)

Best Reasonable Estimate

-3.15% per vacant bldg

- $153,222,604

- $63,551

- $2,723,133

More Conservative

-1.12% per vacant bldg

- $55,420,545

- $22,987

- $984,956

(1) All estimates assume no further effect when count of properties within 500 feet exceed 5. (Sensitivity analysis with limit of 10 showed only marginally larger
total effects.)
(2) Estimated as 40% X decline in value X 44.431 mileage rate (2012); ignores exemption effects that may result in smaller or zero on taxes on some low-value
properties.

distressedvacantproperties.Thiseffectis3.15%foreachdistressedvacantpropertywithin
500feet,whichistheaverageoftheresultsfromthestudiesestimatingtheeffectsoftax
delinquentproperties.
ThesecondrowinTable2.3presentsamoreconservativesetofassumptions,whichleadto
smallerspillovercostestimates.Thisrowassumesthatdistressedvacantpropertieshaveonlya
1.12%effectonhomevalueswithin500feet.Thismagnitudeistheaveragefromthestudies
inTable2.1thatestimatetheimpactofvacant(butnottaxdelinquent)propertiesonhome
values.
Thebestreasonableassumptionresultsinestimatedcumulativespillovercostsofdistressed,
vacantresidentialpropertiesonsinglefamilyhomevaluesintheCityofAtlantaof
$153,222,604.Suchalossinvalue,inturn,couldleadtoadeclineinannualpropertytax
revenuesbyasmuchas$2,723,133,althoughthiseffectmaybemitigatedinthosecaseswhere
someverylowvaluesmaynotexceedexemptionlevels.Onaperpropertybasis,thisestimate
meansthateachofthe2,411distressedvacantpropertiesreducestheaggregatevalueof
homeswithin500feetbyatotalof$63,000.
AmoreconservativeassumptionisusedinthesecondrowofTable3.Here,theaverageofthe
findingsonvacant(asopposedtotaxdelinquent)propertiesisused,withthespilloverestimate
of1.12%perdistressedvacancy.Underthisassumption,thecumulativeeffectonsinglefamily
homevaluesis$55,420,545,withaneffectof$22,987perdistressed,vacantproperty,anda
cumulativeeffectonannualpropertytaxesofasmuchas$984,956.
Theresultsofthisanalysis,summarizedinTable2.3,indicatethatthetotalcostsofdistressed
vacantpropertiesintheCityofAtlantarangefrom$55millionto$153millioninlostproperty
21

values.Thistranslatesintolostpropertytaxrevenuesontheorderof$1millionto$2.7million
annually.Atanaveragepropertyvaluecostrangingfrom$23,000to$63,000pertroubled
property,abenefitcostperspectivesuggeststhatbasedonthesecostsalone,substantial
investmentinremediationordemolitionofsuchpropertiesmaybewarranted.Combining
thesecostswiththesubstantialcostsavingsthatmightbeobtainedbyreducingtheservice
costsdetailedinSection1,theargumentforpublicinvestmentinremediatingordemolishing
distressedvacanthomesbecomesevenstronger.

AMajorCaveat:MitigatingAnyNegativeEffectsofPoorlyMaintainedVacantLotsFollowing
Demolition

Whileanumberofrecentstudies(GriswoldandNorris,2005;Griswoldetal,2014;Whitaker
andFitzpatrick,2014)havefoundthatthatdemolitionprogramsinFlint,Michiganand
Cleveland,Ohiohaveresultedinsignificantreductionsinspillovercostsonlocalproperty
values,theexperienceofsomecitiessuggeststhatifthevacantlotsresultingfromdemolition
arenotaddressedadequatelytheycancreatetheirownsetofspillovercosts.TheCityof
Philadelphia,inparticular,afterengaginginmajordemolitioncampaignsinearlieryears,has
foundthatlargenumbersofpoorlymaintainedvacantlotscreatetheirownsetsofproblems
forcommunities(EconsultandUniversityofPennsylvania,2010).Moreover,recentresearchon
greeningprogramsaimedatgreeningandmaintainingtheselotsshowlargepositiveimpactson
neighboringpropertyvalues(Buchianeri,G.,K.Gillen,andS.Wachter,2012).Theseeffectsare
duebothtotheeliminationofthenegativeimpactsontheneighborhoodofaneglectedvacant
lot,butalsoduetothepositiveamenitiesprovidedbywellmaintainedgreenspace.
Therefore,iftheCityofAtlantaincreasesitseffortstowardsdemolishingdistressed,vacant
homes,especiallythoseposingthegreatestnegativeimpactsonlocalcommunities,itshould
planforgreeningandmaintenanceactivitiesandcostsgoingforward.Otherwisethe
investmentindemolitionmaynotresultinasubstantialrateofreturnintermsofincreased
propertyvaluesandtaxrevenues.

22

Conclusion:AggregatingtheServiceandSpilloverCostsDuetoVacantPropertiesinAtlanta
Thepurposeofthisstudywastoestimatethecostsimposedbyvacantpropertiesinthecityof
AtlantaonthepublicandontheCityofAtlanta.Section1gatheredandanalyzeddataoncosts
tothecityintermsofservicecostsindealingwithvacantpropertiesthroughcode
enforcement,publicsafety,andfireprotectionservices.Section2identifiedthespillovercosts
ofdistressedvacantpropertiesonsinglefamilyhomevaluesinthecity,andonassociated
propertytaxrevenues.
Itisimportanttopointoutthatcostsidentifiedinthisstudyarebynomeanscomprehensive.
Manylikelycostsarenotincludedinthestudy.Forexample,becausethereislittletono
researchoftheeffectsofvacantpropertiesonthevaluesofmultifamilyorcommercial
properties,theseeffectsarenotcapturedhere,andthesecostsarelikelytobesignificant.
Moreover,wheneveruncertaintyofcostswasencountered,effortsweremadetobe
conservative.Therefore,thefindingshereshouldbeviewedasalowerboundonthecosts
imposedbyvacantpropertiesonthecity,andonlocalgovernment.
Notwithstandingthiscaution,thislowerboundoncostsofvacantpropertiesacrossSection1
and2ofthisstudyaredescribedinTableC.1.Therangeofquantifiable,knownannualcosts
associatedwithvacantpropertiesinthecityisveryconservativelyestimatedatbetween$2.6
and$5.7milliondollars.Thesefiguresdonotincludemanyunmeasuredcosts.Examples
includecourtcosts,unrecoveredboardingordemolitioncosts,costsofinjuryfromfires,and
thespillovercostsonmultifamilyrentalorcommercialbuildings.Beyondannualcosts,the
best,reasonableestimateofonetimecoststosinglefamilypropertyvaluesthatareestimated
at$153million.Thisestimateisbasedonusingthestudiesthatappearmostappropriatefor
estimatingtheimpactofphysicallydistressedanddisinvestedpropertiesonnearbyhome
values.Forthepurposesofprovidingaminimumestimateofthemagnitudeoftheseimpacts,
TableC.1alsoprovidesamuchmoreconservativeestimate,basedonstudiesthatexaminein
theimpactofvacantandnotjustdistressedortaxdelinquentpropertiesonnearby
propertyvalues.Whilethesecostsaccruemostlytopropertyowners(includinghomeowners)
andnotdirectlytolocalgovernment(otherthanthepropertytaxportion),theyshouldbe
consideredaspartoftheoverallcostsofvacancyandblight.

23

TableC.1.EstimatedCostsDuetoDistressed,VacantPropertiesintheCityofAtlanta12
Annual Costs
More
Less
Conservative Conservative

One-Time Property Value Loss

Best
Very
Reasonable
Conservative

Service Costs
Code Enforcement
Department of Corrections' Clean and Close Project
Police Dispatch Costs
Fire Department Dispatch Costs

$449,726
$128,400
$688,941
$388,942

$1,587,813
$128,400
$808,288
$436,220

Spillover Costs
One-Time Loss in Single-Family Property Values
Annual Decline in Property Tax Revenues
TOTAL ESTIMATED COSTS

$984,956

$2,723,133

$2,640,965

$5,683,854

$153,222,604

$55,420,545

$153,222,604

$55,420,545

12

Asmentionedearlierinthisanalysis,thesecostsarenotcomprehensive.TheydonotincludessomeservicecoststotheCity
ofAtlanta,includingcleaningandcuttingcostsincurredbytheDepartmentofPublicWorksforyardmaintenanceorcourt
costs(solicitorsoffice,publicdefendersoffice,andmunicipalcourt).Costsassociatedwithtaxdelinquencyandenforcement
onvacant/abandonedpropertiesarealsonotincluded.

24

CitedandRelevantLiterature

Accordino,J.andJohnson,G.2000.AddressingtheVacantandAbandonedPropertyProblem.
JournalofUrbanAffairs22:301315.

Ahrens,M.2009.VacantBuildingFires.NationalFireProtectionAssociation.

Alm,J.,Z.Hawley,J.Lee,andJ.Miller.PropertyTaxDelinquencyanditsSpilloverBenefitson
NearbyPropertyValues(October7,2014).AvailableatSSRN:
http://ssrn.com/abstract=2507049.

APDSolutions.2013.TheStrategicCommunityInvestmentReport:AReportPresentedtothe
CityofAtlantaDepartmentofPlanningandCommunityDevelopmentOfficeofHousing.
November.

Branas,C.,D.Rubin,andW.Guo.VacantPropertiesandViolenceInNeighborhoods.
InternationalScholarlyResearchNetwork:PublicHealth2012:5.

Buchianeri,G.,K.Gillen,andS.Wachter.2012.ValuingtheConversionofUrbanGreenspace.
PennsylvaniaHorticulturalSociety.

CommunityResearchPartnersandRebuildOhio.$60MillionandCounting:TheCostofVacant
andAbandonedPropertiestoEightOhioCities.RebuildOhio.

Cui,L.2010.Foreclosure,VacancyandCrime.November1.AvailableatSSRN:
http://ssrn.com/abstract=1773706.

DeltaDevelopmentGroup.2013.FinancialImpactofBlightontheTriCOGCommunities.
September.Availableathttp://svcog.org/wpcontent/uploads/2014/08/BlightImpact
FullReport.pdf.

EconsultandUniversityofPennsylvania.2010.VacantLandManagementinPhiladelphia.

Gerardi,K.,E.Rosenblatt,P.Willen,andV.Yao,2012.ForeclosureExternalities:SomeNew
Evidence.PublicPolicyDiscussionPapersNp.125.FederalReserveBankofBoston.July25.

Griswold,N.,B.Calnin,M.Schramm,L.Anselin,andP.Boehnlein.2014.CaseWestern
University.February.
25


Griswold,N.andP.Norris.2007.EconomicImpactsofResidentialPropertyAbandonmentand
theGeneseeCountyLandBankinFlint,Michigan.Report#200705.TheMSULandPolicy
Institute.

Han,H.2014.TheImpactofAbandonedPropertiesonNearbyPropertyValues.HousingPolicy
Debate24:311334.

Heckert,M.andJ.Mennis.2012.TheEconomicImpactofGreeningUrbanVacantLand.A
SpatialDifferenceindifferenceanalysis.EnvironmentandPlanningA44:30103027.

Mallach,Alan.2006.BringingBuildingsBack:FromAbandonedPropertiestoCommunityAssets.
Montclair,NJ:NationalHousingInstitute.

Mikelbank,B.2008.SpatialAnalysisoftheImpactofVacant,AbandonedandForeclosed
Properties.FederalReserveBankofCleveland.

Sternlieb,G.andIndik,B.1969.HousingVacancyAnalysis.LandEconomics:45:117121.

TheReinvestmentFund.2014.StrategicPropertyCodeEnforcementandItsImpactson
SurroundingMarkets.August.

Whitaker,S.andT.Fitzpatrick.2014.LandBank2.0:AnEmpiricalEvaluation.WorkingPaper12
30R.FederalReserveBankofCleveland.

Whitaker,S.andT.Fitzpatrick.2013.DeconstructingDistressedPropertySpillovers:TheEffects
ofVacant,TaxDelinquent,andForeclosedPropertiesinHousingSubmarkets.Journalof
HousingEconomics22:7991.

Winthrop,B.andR.Herr.2009.DeterminingtheCostofVacanciesinBaltimore.Government
FinanceReview.June.

26

Appendix:SuggestionsforFutureDataCollection
Inordertobeabletobetteridentifythemunicipalservicecostsassociatedwithdifferenttypesof
properties,whichcouldalsobeusedtotargetthemostburdensomepropertiesforremediationor
demolition,anumberofstepsshouldbeconsideredbyvariousmunicipalagencies.
1) Police,Fire,CodeEnforcement,andMunicipalCourtSystemsshouldidentifytheparcel
identificationnumberforeachpropertyinvolvedinanincident.Noneoftheseagencies
currentlyincludeparcelIDnumbers(orPIN)numbersintheirdatasystems.Whilestreet
addressescanbeuseful,therecordingofstreetaddressescanbepronetofrequenterrorsand
estimationswhichmakethepreciselocationofapropertydifficult.Misspellingsandother
problemscausefrequentproblems.Forpoliceandfire,identificationofaparcelnumberwill
needtobedonebasedonGPScoordinatesorbasedonastreetaddress,orboth,butmatching
toaparcelnumberisgenerallyfeasibleifthecoordinatesoraddressesarerecordedaccurately.
Ofcourse,manylocationsarenotassociatedwithaparticularproperty,andthoselocationswill
notbeassociatedwithaparcelnumber.TheuseofPINnumberswillallowintegratingor
matchingincidentsassociatedwithpropertiesacrossdifferentagencies,aswellaswithmany
othersortsofdatasuchastaxassessordata,taxdelinquencydata,etc.

2) Similarly,alloftheseadministrativedatasystemsshouldincludealocationtypefieldto
identifyeasilywhetherthelocationisarealestateparcelornot.Otherlocationtypesmight
includeexpressway,arterialstreet,sidestreet,etc.Togetherwithdataonthetypeof
property(commercial,industrial,multifamilyresidential,singlefamilyresidential,etc.)thisfield
willbehelpfulinidentifyingthenatureofthelocation.

3) Alloftheseagenciesshouldincludecompleteandaccurateinformationontheconditionofthe
propertywhereanincidentoccurs.Ataminimum,theoccupancystatusofeachpropertyshould
berecorded.TheFireDepartmentistheonlyagencythatcurrentlytracks(atleastpartially)
whetherapropertywhereanincidentoccursisvacantornot,andthereisincompletedatafor
manyincidents.Beyondvacancy,usefuldatawouldincludeanestimateoftheconditionofthe
building,similartothetypologyusedintheStrategicCommunityIndicatorssurvey(good,fair,
poor,deteriorated).

4) TheCodeEnforcementUnitshouldtrackhoursspentinspectinganddealingwithpropertiesby
casenumber,togetherwiththeassociatedparcelnumber.Alsoincludedshouldbecostsdueto
boardingordemolition.Recoveryofsuchcostsshouldalsobetracked.

5) MunicipalCourtshouldutilizeexistingCodeEnforcementcasenumberstotrackthetimeand
costsofcases.Parcelnumbersshouldbeincludedinsuchdatasystems,again,toeasilylinkto
otherdatasetsbyproperty.Thecostsofmunicipalcourtinspectionsshouldbetracked,
includingtimeandanyexpenses,includingpublicdefendercosts,filingfees,etc.Allpersonnel
costsshouldincludefullyloadedcosts.
27


6) TheFultonCountyTaxCommissionershouldprovideparcelleveldetaileddataonalltaxactivity,
includingdelinquencies,taxliensales,taxforeclosures,etc.,onaregularbasis.Thisdatashould
indicatethestreetaddressandparcelnumberofeachproperty,theannualtaxbill,any
outstandingtaxbill,anypurchasedtaxliens(amount,whenpurchased,interest,etc.),anytax
foreclosureaction,andotherrelevantinformation.Suchdatashouldbeupdatedonanannual
basis.Relianceonvendorsthatmakeretrievalofsuchdatacostlyshouldbereconsideredto
enablereasonableresponsetimestodatarequestsandtheprovisionofgranulardatainreadily
understoodformats.

28