Skip Navigation

Federal Communications Commission

English Display Options

Commission Document

Response to Professors Reed and Sirbu

Download Options

Released: July 25, 2013

ResponsetoProfessors ReedandSirbu

WearegratefultoProfessors ReedandSirbufortheircommentsandsuggestionson
version2oftheConnectAmericaCostModel(abbreviatedeitherCACMorCAM).1
Wearepleasedthattheyfoundthat"[t]heCACMdoesa reasonablejobofproviding
anestimateofthecostofdeployingabroadbandnetworkacrossawidenumberof
deploymentscenariosfoundthroughoutthenation,"2 andfurtherthat,"[t]he
methodologytodevelopandestimatethecostofthenetworktopologyimplemented
inCACMshowthatreasonableassumptionsarebeingmadetosimulateactual
networkdeploymentcoststothebestabilityofcurrentestimationmethods."3

Background

TheCAM estimates,foreverycensusblock,the cost ofprovidingavoice- and
broadband-capablenetwork.Thatcostislevelized,thatis,isturnedintoaconstant
(orkindofaverage)monthlycostassociatedwithservingthecensusblockoverthe
network'sentirelife.4 Givencostsforeachblock,themodelcanbeusedtocalculate
support.5 Inparticular,afuture WirelineCompetitionBureau(Bureau) orderwill
determinealower"benchmark"thatidentifiesareasthatcouldbeeconomically
viableabsentsubsidies("thefundingthreshold"),andanuppercostthresholdthat
delineates thosedeemedtobeextremelyhighcost.Thetotalsubsidywillbeequal
tothecostofsupplyingeligibleCensusblocksthataremoreexpensivetoservethan
thefundingthreshold,butarenotmoreexpensivethantheextremelyhighcost
threshold.Thetotalbudgetforsupportinpricecapareaswassetbythe
Commissiontonotexceed$1.8billion.6Thuschangesincostorsupport
calculationschangetheobligationsthatcarriershavetoaccepttoreceivefunding,
butwill notchangethetotalamount offunding.
Professors ReedandSirbufocusedonsixtopics:documentation,network
technologies,voiceassumptions,networksizing,servicescope,andcostand
supportmodelinteractions.Foreaseofreference,ProfessorsReedandSirbu's
commentsareprovidedinitalics,andwehave,asmuchaspossible,respondedto
thosecommentsintheordertheyappearintheirreview.

1 WhilethepeerreviewreferstoCACMthroughout,thisresponse,consistentwith
Connect AmericaFundetal.,WCDocketNos.10-90,05-337,ReportandOrder,28
FCCRcd5301(WirelineComp.Bur.2013)(CAMPlatform Order), willrefertothe
modelasCAM.
2 LetterfromDavidP.ReedandMarvinSirbu,CarnegieMellonUniversity,toJulie
Veach, Chief,WirelineCompetitionBureauat1(Feb.18,2013)(R&S).
3 R&Sat5.
4 CAMPlatform Order,28FCCRcd5301,5308,para.15.
5 SeeConnectAmericaFundetal., WCDocketNo.10-90etal.,ReportandOrderand
FurtherNoticeofProposedRulemaking,26 FCCRcd17663,17715-16,paras.134-
135(2011) (USF/ICCTransformationOrder).
6 Seeid. at17764,17729,17738,paras.25,169,193.
1

1.Documentation
....thereisroomforimprovementtothemodel,mostimportantlythroughbetter
documentationthatwouldinvitemoreeffectivereviewofthemethodologyand

assumptionsemployedinordertomoreeasilyverifytheaccuracyoftheresults.7
...modeldocumentation... was...insufficienttoallowonetooperateorunderstand
instraightforwardfashionthekeyassumptionsineachmodelcalculationandreport

ofresults.... Anotherproblemisthatthemodeldocumentationisfragmentedacross
differentdocumentsandsources.8
Themostimportant improvementProfessors ReedandSirbucalledforwasbetter
documentation.9 Weagreewiththeprofessors,andsubstantialprogresshasbeen
madeonthisfront sincetheyreviewedthedocumentation.
Thefollowingimprovementstodocumentation are currentlyavailableand
centralizedinonelocation,onthemodelwebsite:
1. TheCAMhomepage(cacm.usac.org)displaysa"systemupdatespage"link.
Thisprovidessummaryleveldetailsonmodelchangesbyversionnumber
andreleasedate.
2. TheCAMhomepage displaysa"Resources"buttontoprovideusersa
consolidatedlocationfordocumentationandadditionalresources.
3. Currentdocumentationlistedunderthe"Resources"button:
a. BackgroundInformationonConnectAmericaCostModel Providesa
summaryoftheConnectAmericaCostModelanditsrolewithinthe
ConnectAmericaFund.
b. CACMMethodology Comprehensivedocumentation providingdetails
onthemodel'smethodologyandthemethodologyusedtoderive
variousinputvalues (updatedaseachnewversionisreleased).
c. OpexOverview DescribesthedevelopmentoftheOpexinputsforthe
ConnectAmericaCostModel.
d. CapitalCostModel Derivesannualchargefactorsfordepreciation,
costofmoney,andincometaxesassociatedwithcapitalinvestments,
usedasinputsinthemodel.
e. CapexTutorial Alinktoatutorialvideoexplainingthecapital
expendituresworkbooktohelppartiesbetterunderstandthe
structureandinputscontainedintheworkbook.
f. UserGuide Provides helptousers withinformationonhowto work
withandanalyzetheConnectAmericaCostModel.
g. FAQ - FrequentlyAskedQuestionssenttoCACMSupportdesk
(CACMsupport@costquest.com).

7 R&Sat1.
8 Id. at 3.
9 Id. at1,3-5.
2

h. TileQueryFieldDefinitions Liststhefielddefinitions fordatafields
withinthetilequeryresults.
4. Additionalresourceslistedunderthe"Resources"buttontoassistusersin
analyzingmodelresultsinclude:
a. TelcoMasterTable Providesholdingcompanynameassociatedwith
servingwirecentersandincludesstate,companyname,studyarea
code,classificationasrate-of-return orpricecap,companysize, and
otherdata.
b. CoverageData Identifies censusblockspresumptivelyservedby
unsubsidizedcompetitors.
5. CACMsupport@costquest.com Linklocatedonthehomepageforusersto
submitquestionsrelatedtoaccess,administrationandoutputgeneration.
6. TheCACMhomepagealsodisplaysa"PostedDatasets"buttontoprovide
userswithaccesstomodelinputsandmodelresultsfromvariousmodel
runs.
7. SystemEvaluator Additional documentationisavailableviaemailwitha
linkfordownloadwithaspecificlicenseagreement.Assists userstotakea
closerexaminationofthesourcecode andtheoperationoftheCACM.
These improvementsin documentationaddress ProfessorsReedandSirbu's
suggestionforacomprehensiveusermanual,includingdocumentationofthe
derivationofthevaluesofinputassumptions.
TheBureaucontinuestowork withUSACandits contractortoimprovethe
documentationavailable inacentralizedlocation, themodelwebsite,reflecting
changesinmodelversionsandprovidingmorecompleteexplanationsaswarranted.
2.

NetworkTechnologies

Evenfortelephonecompanies,acceptedengineeringpracticesofthenetwork
technologiesconsideredtodayfordeploymentofbroadbandservicesina"greenfield"
includesHybridFiberCoaxnetworks. TheCACMshouldincludethisasanoption.10
Atrue"greenfield"calculationwouldconsiderHFCapproachsinceitcanbedeployed
fromtheexistingwirecentersoftheILECsasrequiredbythe"scorchednode"
assumptionofthemodel.11
Professors ReedandSirbusuggestthathybridfibercoax(HFC)deploymentshould
beincludedasanoptioninthegreenfieldmodel,onthebasisthatthiswouldbe
consistentwithincluding"thelowest,potentiallydisruptivetechnicalstrategies
availableinthemarket."12Weagreethat theleast-cost, mostefficient, and
reasonablewirelinetechnology capableofprovidingtheservicesspecifiedinthe

10 Id. at1.
11 Id. at7.
12 Id. at6-7.
3

USF/ICCTransformationOrder shouldbe usedintheCAM,13 buttheBureauhas
concludedforpolicyreasonsthatitisreasonabletofocusonafiber-to-the-premises
(FTTP)inthe CAM.14
TheBureauisonlyawareofonepricecapcarrierdeployinganHFCnetwork,and
thatcarrierrepresentsmuchlessthan0.05%ofallpricecaplines.15Inotherwords,
HFCisatechnologythatnocompanyeligiblefortherightoffirstrefusalhas
deployedatascalecommensuratewiththenumberoflocationseligibleforConnect
AmericaFundPhaseIIsupport.
Inanyevent,thelevelizedgreenfieldcostsofan HFCnetworkmayexceedthoseofa
fibernetworkintheareasofinterestforthereasonslaidoutbelow.Itispossible,
thoughfarfromcertain,thatanHFCnetworkcouldprovidesomenear-termsavings
relativetoaFTTPnetwork.16However,giventheextremelylonglifeofmanyofthe
assetsdeployedinnetworkbuilds,thismustbeweighedagainstthehigher
operatingcostsofmaintainingactiveelectronicsinthefieldinanHFCnetwork;and
thepossibleupgradecostsofanHFCnetworkovertime(e.g.,toupgradeelectronics,
tohandleincreasingcustomerbandwidthrequirementsand"tighten"thesystemto
utilizemorespectrumintheHFCsystem).Particularlyinruralareas,wherethere
arefewerpotentialcustomerspermileofoutsideplant,itisunclearwhetherthe
higherCPEcostoroperating-costsavingswouldprevail;theimpactofCPEcost
wouldalsodependheavilyonassumptionsmadeabouttakerate.
Wenotethatseveralfirmsthataredeployinghigh-capacitybroadbandare
deploying FTTPrather thanHFC,evenwhencompetinghead-to-headwithcable
companiesthathavedeployedHFC,suggestingthatthey viewFTTPasmorecost-
efficientthanHFC.17

13 USF/ICCTransformationOrder,26FCCRcd at17736,para.189.
14 CAMPlatform Order,28FCCRcdat5314-16,para. 33(adoptingFTTP);seealso
ConnectAmericaCostModel(CACM)version3.1.4.,at17,40,sections5.2.2.and8.3.
(2013)(CAMv.3.1.4Methodology),availableat
http://transition.fcc.gov/Daily_Releases/Daily_Business/2013/db0624/DOC-
321774A1.pdf.
15 Companiesmayrelyoncoaxialin-homewiringtodistributetelco-providedvideo
andbroadband,butthisisnotthesameasanHFCnetwork.Tothebestofstaff's
understanding,onlyVitelcoamongpricecapcarriershasprovisionedanHFC
network.See CAMPlatform Order,28FCCRcdat5314,para.33n.59.
16 Forexample,thecustomerpremisesequipment(CPE)inanFTTPnetwork,the
opticalnetworkterminal(ONT),ismoreexpensivethancomparableequipmentin
anHFCnetwork.
17 GoogleandVerizon'sdeploymentofitsFiOSnetworkaretwoexamples.See also
CedricLam,GoogleNetworkArchitect,HighPerformance,LowCost,ColorlessONU
forWDM-PON (2012)at3,availableat
http://static.googleusercontent.com/external_content/untrusted_dlcp/research.go
ogle.com/en/us/pubs/archive/37746.pdf (concludingthat"WDM-PONisthemost
4

3.Voiceassumptions
Thepersubscriberorvariablecostestimationmethodologyappliedforthecostof
voice-specificnetworkcomponentsisnotconsistentwiththeaccessnetworkcosting

approachwhichincludesfixedcosts.18
WedonotknowhowsignificantthepercentageofhardwarecostsforIMS-basedvoice
isrelativetototalcosts(thebasecaseCACM estimatesthispercentagetoberoughly

30%,whichisnotinsignificant). Ataminimum,thispersubscribercostestimation
methodologyisnotconsistentwiththeaccessnetworkcostingapproachbyreplacing
thesefixedcostswithvariablecosts.19
TheBureauhasnotmadeafinaldecisiononhowtomodelthecostofvoice
capability,andhassoughtcommentonthattopicinthevirtualworkshop.20We
note,however, thatallcostsarelevelized,sothecostofanyequipmentrequiredfor
servicewillbespreadovertimeregardlessofwhetheroneassumesthecostsare
incurredasaninitialcapitalexpenseoranongoingexpense.Furthermore,per-
subscribervoiceoverInternetprotocol (VoIP)equipmentcostsaremuchlessscale-
dependentthancircuit-switchedequipmentcosts.Instead,VoIPsystemscanbe
scaledfromafewhundred lines tohundredsofthousandsoflines.Consequently,
eventhesmallestpricecapcarriercanpurchaseVoIPwithper-linecoststhatare
notradicallydifferentfromtheper-linecostsfacedbycarriersmanytimestheir
size.

promisinglong-term,scalablesolutionfordeliveringhighbandwidthtotheend
user;MarkWeigleitner,Verizon,VerizonFiOSArchitecture,presentationtothe
MultimediaoverCoaxAlliance(2007),availableat
http://www.mocalliance.org/industry/presentations/2007_11_14_TechConference
/docs/MarkWegleitner.pdf (showingFiOSFTTParchitecture).Dr.Lam'sresearch
canbeviewedonhiswebpage.CedricLam,ResearchatGoogle,
http://research.google.com/pubs/author38397.html (lastvisitedJuly1,2013,4:44
PM)(showingGoogle'sFTTParchitecture).Similarly,TDSTelecomisdeploying
FTTPincertainareas.See,e.g.,PressRelease,TDSTelecom,TDS expandsfiber
network,bringsTDSTV Fiberville concepttoConcordandpartsofKnoxville (July2,
2013)http://www.tdstelecom.com/mediaroom/Article.aspx?id=69de9870-175f-
4612-9734-97c13a6c73dc.
18 R&Sat1.
19 R&Sat6.
20 See LetterfromMichaelJ.Jacobs,LegalAdvisortotheChief,WirelineCompetition
Bureau,toMarleneDortch,Secretary,FCC,WCDocketNo.10-90,Attach.at15-17
(filedFeb.6,2013)(WCBFeb.6,2013VirtualWorkshopSubmissionLetter);Letter
fromJamieSusskind,LegalAdvisortotheChief,WirelineCompetitionBureau,to
MarleneDortch,WCDocketNo.10-90,Attach.at1-2(filedMarch28,2013)(WCB
March28,2013VirtualWorkshopSubmissionLetter).
5

ManyVoIPsolutions(fromAvaya,Brocade,Cisco,Nortel,Siemensandothers) scale
fromsmall to large.21 A modular and/orcluster-basedframeworkallowstheseVoIP
solutionstomeetthespecificsizerequirementsforcarriers ofvaryingsizes.
Additionally,manyVoIP callcontrolsystemsmayberemotelylocated,providing
additional flexibilityandcapabilitiestoallowsmallercarriersandlargercarriersto
tailorthesolutionsneededtotheirrequirementsonacost-effectivebasis. Manyof
thecarriergradeVoIPsolutionsand/orgatewaystothepublicswitchedtelephone
network(PSTN) arelicensedsolutions,withfeesoftenbeingbaseduponthe
numberoflinesgoingtoandfromthePSTN amongotherparameters which
supportsamorelinearorvariablecostmodel.22
4.Networksizing
Thedescriptionofthemethodologyorengineeringpracticesappliedfordrivingthe
networksizinginCACMdeservefurtherattentionasithassignificantimplicationsfor
futurecostsofthenetworktomeetgrowingdemand.23

A.BandwidthCapacity

Professors ReedandSirbu notethattheamountofcapacityforwhichtheCAM
calculatescosts "isanimportantparameterforboththecostandviabilityofthe
broadbandnetworkinthefuture."24Inthissection,we provideadditional
explanationoftheassumptionsandmethodologyusedinthecurrentversionofthe
CAM.
Insummary,giventhenetworksizingattheCAMaggregationpoints,andthelikely
rangeofreasonabletakerates,thenetworkinthecurrentversionofCAMprovides
muchmorethan5.4Megabitspersecond(Mbps) ofcapacitypersubscriberinrural
areas.Additionally,thecurrentversionoftheCAM allows userstospecifyeven
greaterbusy-hourofferedloadstoallowfurtherexaminationoftheimpactofthis
variableoncostestimates.WenotethattheBureauhasnotyetadoptedaninput
valueforassumed capacitypersubscriber.
In the CAMPlatform Order,theBureauadoptedanFTTPapproach.25 In themodeled
GigabetPassiveOpticalNetwork(GPON),therearethreebasic aggregationpointsor
bottlenecks.Thefollowingisa listofcapacityattheaggregationpoints:

21 Asanexample,see CiscoUnifiedCommunicationsSystemRelease9.0SRND
(SolutionsReferenceNetworkDesign),availableat
http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/srnd/9x/uc9xsrnd.pdf
22 Forexample,see CiscoSystemsPWG200license agreement,availableat
http://www.cisco.com/en/US/docs/voice_ip_comm/pgw/9/feature/module/9.7_3_
/FlexLM.html.
23 R&Sat1.
24 R&Sat8.
25 CAMPlatform Order,28FCCRcdat5314-16,para. 33;see CAMMethodologyat17,
40,sections5.2.2.and8.3.
6

Thefirstaggregationpointisthefibersplitter whichsupports upto32lines/
subscribers with2.5 Gigabitspersecond(Gbps) ofbackhaul;26
Thenextaggregationpointistheopticallineterminal (OLT),wherefiber-
opticsignalsfrommultiplesplitter fibersareaggregatedandshiftedontoa
10GbpsconnectiontoEthernetswitchesandrouters (10Gbpstotal
bandwidthforeachOLT).27TheCAMassumesthateachOLTcanhandleas
manyas58splitter feederfibers,or1,856 subscribers (assumingall
locationspassedsubscribe) (10,000Mbps/1,856subscriber =5.39Mbps
persubscriber).AtfullcapacityofthesplitterandtheOLT,eachsubscriber
willreceivemorethan 5.4Mbpsofcapacity;
Towardthecorenetwork,aggregationpointsareEthernetswitchesand
routers,whosecapacities(numberoflinecards)increasewiththenumberof
subscribers assumedtobeonthenetwork.Inotherwords,theCAM
capturestheneedforincreasedcapacityintheEthernet(backhaul)network
according tothesupportednumberofsubscribers.Unitcostsduetothese
capacityincreasesfallwithgreater capacity.
Thecapacitypersubscriber isultimatelydrivenbythenumberofsubscribers per
splitterfiber,whichdependsonlocaldemand:indenserareas,the splitteris
typicallyfullyconsumed, whileinmoreruralareasitislikelythattherewillbe
fewerthan32subscribers persplitterfiber.Inaddition,insmallerserviceareas,the
OLTswillnothaveashighutilizationasthoseinlargerserviceareas.28 Denser,less-
ruralareas areaslesslikelytohavehighcosts aremorelikelycapacity
constrained attheOLT.Regardless,inthemostcapacity-constrainedareas,each
subscriber willreceiveatleast 5.4 Mbpsofcapacity.29Inruralareas,withfewer

26 Infact,eachsplitter mayhaveonlyafractionofthe32possible locationsas
subscribers,meaningtheCAMallowsforevenmorecapacitypersubscriberthan
describedinthisexample.Thatmeansthateachlocationhasaminimumof75
Mbpsofdownstreamcapacity.
27 Forsimplicity'ssake,theFCCassumes1Gbps=1,000Mbps.
See,e.g.,FCC,2013MeasuringBroadbandAmerica--FebruaryReport,Conclusionand
Next Steps,availableat http://www.fcc.gov/measuring-broadband-
america/2013/February#Conclusion ("Googleoffers1Gbps(1000Mbps)servicein
KansasCity,MO").
28 OLTsareoften,thoughnotnecessarily,locatedincentralofficesandcantherefore
aggregatedemandfromanentireservicearea.Whilethisreducesthelikelihoodof
anOLTbeingonlypartlyfilled,itdoesnoteliminatethatproblementirely.
29 Thiscorrespondstoabusy-hourofferedloadof5.4Mbps. Maximalusageofa4
Mbpsconnectionwillresultinabusyhourofferedloadof4Mbps.Thereforethe5.4
Mbpsbusy-hourofferedloadcorrespondstomorecapacitythantheamount
requiredifallsubscriberswereusingeverybitofavailablebandwidthona4Mbps
networkatonetime.And,asnotedabove,moreruralareas,withfewersplittersper
7

than32subscribers persplitterfiberandfewerthananaverageof58splitters
connectedtoeachOLT,eachsubscriber willhavemanytimesthis5.4Mbps capacity
bydefault,withtheexactamountdeterminedbylocalconditions.
Lastly,severalparametersoftheCAMcanbeadjustedtoinfluencethebandwidth
requirements.Thenumberofsubscriberscanbeadjustedtoinfluencethe
bandwidthcapacityrequirementsfortheprojectedsolutionset.Additionally,the
solutionset can beadjustedby changing thebusy-hourofferedloadandthus
increasingordecreasingthecapacityavailabletosubscribers.

B.Residentialvs. BusinessSizing

ProfessorsReedandSirbustatethatthemodel"alsodoesnotdifferentiateinthe
bandwidthinputassumptionsforresidentialandbusinessuserseventhoughthesetwo
setsofcustomerstypicallypurchasedifferentservicetiers."30Theynotethat"an
importantbusinessconsumeroffibertrunking wirelessbackhaul isnotspecified."31
TheCAMdifferentiates thetypeofservicesassumedtobepurchasedbyresidential
andbusinesssubscribers,asnotedinthedocumentation.TheCAM'snetwork
topologyassumescertainbusinesslocationsareservedusingadedicated,special
access/privatelinefiber.32
Themodelestimatesthepotentialdemandforservicesbasedonthetypeof
business;businessesareclassifiedas"technologyoriented"or"allotherbusiness"
basedontheirNorthAmericanIndustryClassificationSystem (NAICScode)andthe
numberofemployeesateachlocation.Themodelprovisionsdedicatedfiber
connectionstotechnologyorientedbusinesslocationsthathave10ormore
employeesandallotherbusinesslocationsthathave50ormoreemployees.33 The
modeldoesnotincludeanycostfortheassociatedelectronicsnecessarytolightthe
dedicatedfiberservicetobusinesslocationsthatareassumedtopurchasesuch

OLT,eachsubscriberwillhavemorecapacitythanthat.Thecurrent versionofthe
modelcalculates costsbasedona"greaterof"approach,usingeitherthedefault
capacityimpliedbytheFTTParchitectureoraspecifiedbusy-hourofferedload,
whicheverisgreater.Thisenables userstospecifycapacityusesfarinexcessof
whata4Mbpsdownstreamnetworkcanprovide totesttheimpactoncost
estimates.
30 R&Sat9.
31 Id.
32 AdditionalinformationabouthowCAMmodelsdemandforhigherbandwidth
servicesandtreatsthecostsassociatedwithsuchservicescanbefoundintheCAM
Methodology.CAMMethodology at19-21,sections5.2.3.2and5.2.3.3.
33 Businesslocationswithfeweremployeesareprovisionedthesameasvoiceand
broadbandservicesatresidentiallocations.CAMMethodologyat20,tbl.3.
8

services,norarethoselocationsincludedwhenunitizingtotalcostwithinacensus
block.34
TheCAMdetermineshowmanyfiberstrandsareusedbythevariousdemand
locationsandallocatesthecostoffiberandstructurebetweenspecial
access/privatelinelocations(includingcelltowers,business locationswithmore
employees, andcommunityanchorinstitutions)andotherlocations(i.e.,residential
locationsandthosebusiness locationsassumedtobepurchasingresidential-type
services),withsupportcalculatedbasedonlyoncostsrelatedtothelattergroupof
locations.Themodelsimilarlycapturesthesharingofthe middlemilenetworkby
estimatingthat50percentofthecostsofaninterofficerouteareattributableto
specialaccess/privatelinedataservicesprovidedtocertainbusinesslocations,
communityanchorsandwirelesstowers,andthosecostsareexcludedfromcost
calculations.Locationsservedbysuchspecialaccess/privatelineservices(which
includesdirectInternetaccess)arealsoexcludedfromtheunitizationoftotal
middlemilecostofacensusblock,i.e.,whenthetotalmiddlemilecostofservingthe
censusblockisdividedbyalllocationspassed,thelocationspassedonlyinclude
residentialandthosebusinesslocationsassumedtoreceivethesametypeofvoice
andbroadbandservicesasresidentialsubscribers.

C.CACMProjectionforDemandGrowth

Severalofthereviewers'questionsdealwithhowtheCAMhandlesdemandwhich
growsovertime;they suggestedthat"itwouldbemoreintuitiveifthemodel
allowedfordirectinputofaprojectedtrafficgrowthrate,andgeneratedreasonable
fillfactorstobeusedinordertominimizecostlyfuturereinforcement."35Coststhat
arecalculatedbasedonlyonnear-termdemandmayexcludetheadditional
investmentrequiredtohandlefuturedemand.
Webelievetheapproach thattheCAMtakesprovidestheBureauwithareasonable
meansoftestingalternativeapproaches.First,thenumberofsubscribersusedto
calculatecostscanbechosentoincludethetotalnumberoflocationsconnected
overtime,ratherthanjustthenumberoflocationsconnectedatanyonepoint.36
Second,asdescribedabove,themodelcanbeadjustedtomakeassumptionsabout
busy-hourofferedloadattheendofthetimeperiodofinterestratherthanthe

34 CAMMethodologyat51-53,app.2.
35 R&Sat9.
36 TheBureauhassoughtcommentonareasonabletakerateassumptionforthe
CAMandwilladoptatakeratefortheCAMinafutureorder.See WCBFeb.6,2013
VirtualWorkshopSubmissionLetter,Attach.at39-43.
9

beginning.37 WenotethatthecostscalculatedbyCAMarenotparticularlysensitive
totheseassumptions.38

D.AdditionalQuestions

Professors ReedandSirbuaskseveralspecificquestionswhichweaddressbelow:
Howmuchofthenetworkisinstalledinitially?Asdiscussedabove,consistent
withthepolicydecisionsmadein theUSF/ICCTransformationOrder,the
currentversionoftheCAMdeterminesthetotalinvestmentsneededtobuild
thegreen fieldnetwork atultimatecustomerdemand andcalculates
levelizedannualcostsofthatinitialinvestment.TheBureauexpectsthat
networktobeconsistentwithend-of-perioddemand.
Howmuchworkingcapacityisdeployedforeachuserinthenearterm?The
versionofCAMreviewedbyProfessors ReedandSirbuassumesatleast
5.4 Mbps ofdownstreamaccessforeachsubscriber.Asnotedabove,the
currentfutureversionofthemodelincludes theabilitytospecifyagreater
capacity.
WhatinvestmentismadeintheplantupfronttosupporttheFCC goalthat
mostlocationsreceivea6Mbps/1.5Mbpsorfasterserviceattheendof5
years?TheCommissionsaidthat6Mbps/1.5Mbpsserviceneedstobe
availableto"anumberofsupportedlocationstobespecified."39 TheBureau
hasnotyetspecifiedthatnumber.WenotethatbecausetheBureauhas
adoptedaFTTPnetworkarchitectureinthe CAMPlatform Order, thereisno
incrementaladditionalcostinthemodelforlocationstoreceive 6Mbps/1.5
Mbpsservice.40
Howdoesthemodelincorporatetheannualprojectedincreasesinbroadband
trafficperuser? Asnotedabove,byassuming acapacityconsistentwiththe

37 Again,theinputvaluesforcapacitywill besetbytheBureauinafutureorder.See
WCBFeb.6,2013VirtualWorkshopSubmissionLetter,Attach.at20;WCBMarch
28,2013VirtualWorkshopSubmissionLetter,Attach.at3.
38 AllPhaseIIsupportedlinesmustofferatleast4Mbpsdownstream,withsome
supplyingatleast6Mbps.USF/ICCTransformationOrder,266FCCRcdat17702,
17726-27,17735,paras.105,160-163,187.Themaximumbusy-hourloadfrom
linesofsuchspeedislikelylessthanthecapacityasdescribedabove,sothereisno
costimpactforgreatercapacityassumptions.Evenwithassumptionsabouthigher
busy-hourusage(i.e.,assuminghigher-speedconnections),theimpactoncostsis
likelytobelimited.
39 USF/ICCTransformationOrder,266FCCRcdat17726,para.160.Thedirection
that"themostlocationspossible"receiveserviceissubjecttotheconstraintofbeing
"consistentwiththeCAFPhaseIIbudget."Id. at17735,para.187.
40 Thisisnottosaythatprovidersmaynotincursomeincrementalcashcostsin
providingthehigher-speedservice,onlythatthereisnoimpactonthecalculationof
costsinthemodel.
10

endofthemodelingperiod,theBureaubelievesthecurrentversionofthe
modelallowsforreasonablebandwidthgrowth.
5.Servicescope
ThescopeofCACMincludesvoiceandbroadbandservices,butnotvideoservices.
Giventhatvideoisrapidlyemergingasanapplicationinthebroadbandspace,this
distinctionisnotlogicallyconsistentwithcurrenttechnicalandmarkettrendsin
broadband. Theimplicationsofthisomissionarethateconomiesofscaleandscope

betweenbroadbandandvideoservicesmaynotberealizedorproperlyreflectedinthe
resultsoftheCACM.41
ProfessorsReedandSirbusuggestthatvideocostsandpresumablyrevenuesshould
be accountedforindeterminingtheefficientsubsidy.Attheoutset,wenotethatthe
Bureauhasnotyetmadeadeterminationofwhattypesofrevenuesshouldbe
includedwhendeterminingthefundingthreshold.Thedecisionofwhetherto
include videorevenuesindeterminingthesupportthresholdisapolicydecisionnot
subjecttopeerreview.
Moreover,evenifvideorevenuesarerelevant,theywouldonlybesototheextent
thattheyexceedtheincrementalcostofvideoprovision,andhencecontribute
towardsharedcosts. Itispossiblethatthecontributionfromvideoserviceswould
berelativelysmall,andthereforeunlikelytodistorttheCAMsubsidyestimates.42
Beinglateentrants,incumbentlocalexchangecarriers(ILECs) havearelatively
smallshareofallvideosubscribers.43Becauseoftheirvideomarketshare,ILECs
typicallyfacehigherpersubscriberprogramming,consumerpremiseequipment,
andcustomeracquisitioncoststhantheir cableand satelliterivals,44 reducingtheir

41 R&Sat1.
42 ThechoicetomodelaFTTPnetwork,withorwithoutvideorevenues,doesnot
implycarriersacceptingmodelsubsidiesmustdeployanFTTPnetwork andoffer
(ornotoffer)videoservices.
43 Thecombinedsharesofalltelephonemultichannelvideodistributors(MVPDs)
accountedforapproximately8.4 percentofMVPDsubscribersattheendof2011,
comparedto6.9percentattheendof2010.Thelargestofthese,Verizon,wasthe
7th largestMVPD,whilethe2nd largest,AT&T,wasthe9th largest. Annual
AssessmentoftheStatusofCompetitionintheMarketfortheDeliveryofVideo
Programming
, MBDocketNo.12-203,FifteenthReport, FCC13-99at12-13,paras.
28,30(rel.July22,2013)(15th VideoCompetitionReport).Accordingtooneanalyst,
asofMarch2013,telephoneMVPDsharehasrisentoclosertotenpercentofthe
market(LeichtmanResearchGroup,pressrelease,May20,2013,
http://leichtmanresearch.com/press/052013release.html,viewedMay31,2013).
44 15th VideoCompetitionReport,FCC13-99at34-35,paras.69-72.See Steve
Donohue,Verizonproposespayingcablenetworksbasedonviewership,FierceCable
(Mar.18,2013),availableat http://www.fiercecable.com/story/verizon-proposes-
paying-cable-networks-based-viewership/2013-03-18 (notingthatVerizon's
11

pervideosubscribermargins.Infact,thepervideosubscribercontributionfrom
videocouldbenegative,buttheoperatormightstillprovidetheservicebecauseit
preventsfurtherlossofvoicecustomerstomobileandotherrivals.45Thefailureto
includeanetcontributionfromvideothereforemaybeunlikelytoleadtoan
overstatementofthenecessarysubsidyrequiredforavoiceandbroadband-capable
broadbandnetwork.
6.Costandsupportmodelinteractions
Arrivingatconsistentresultsbetweentheassumedtakeratesandthelevelof
universalservicesupportrequiresatimeconsumingiterationprocessofrunningfirst
thecostmodelatalowertakerate,andthenthesupportmodeltoseeifthatlevelof
takeratecanbesatisfiedwithintheuniversalservicebudgetcapinordertoendup

withconsistentresultsforboththecostandsupportmodels.46
Professors ReedandSirbu argue thatamodelthatdoesnotconsistentlymodeltake
ratesanduniversalsupportassumptionssimultaneously "isultimatelynotsufficient
forevaluatingpolicyoptions,andtheCommissionwouldbenefitfromamodel
whichmadeconvergenceonconsistentnumberseasiertoachieve."47
Weagreethatamodelthatcouldsimultaneouslytakeassumptionsaboutprice,
servicestandardsanddemandasinputsto determinesubscriptionratedecisions,
whichinturnwouldbeusedtoidentifytheoptimal CAMtakerate,andultimately
determinethe upperthresholdfortheCAFsubsidygiventheCommission's budget,
wouldbeausefultool.However,suchamodelingexercisewouldrequire
substantialadditionalprogrammingandcomputationalresources unavailabletous.
Giventheseconstraints,weareusingmanualiterationsoastoensureconsistency
acrossourassumptionsaboutthevariables inplay.

attempttonegotiatefeespaidforcontentisunlikelytosucceedgiventhemarket
powerandmostfavorednationclausesofComcast andDirecTV).SNLKagan
reportsthatlargerMVPDsfacelowerprogrammingcosts.SeeTelcoTVOutlook:
CompetitiveanalysisofUStelcovideodeployments,2007edition,SNLKagan,June
2007("volumediscountsandmost-favored-nation(MFN)clausesensuring
multichannelgiantssuchasComcastandTimeWarnerCablealwayspayrates [for
programming] lessthanorequaltothosechargedtheircompetitors. ... Wepolled
somesmalloperatorsandfoundtheywerepayingabout30%morethanthe
averageforadozennetworksrandomlychosenforoursurvey.").
45 Forexample,thiswouldbetrueifpackagingvideowithvoice(andperhaps
broadband)makes,holdingotherthingsconstant,customerslesslikelytoswitch
theirvoiceservicestoanotherprovider.
46 R&Sat11.Seealso id. at1.
47 Id. at11.
12

Note: We are currently transitioning our documents into web compatible formats for easier reading. We have done our best to supply this content to you in a presentable form, but there may be some formatting issues while we improve the technology. The original version of the document is available as a PDF, Word Document, or as plain text.

close
FCC

You are leaving the FCC website

You are about to leave the FCC website and visit a third-party, non-governmental website that the FCC does not maintain or control. The FCC does not endorse any product or service, and is not responsible for, nor can it guarantee the validity or timeliness of the content on the page you are about to visit. Additionally, the privacy policies of this third-party page may differ from those of the FCC.