Government Information Quarterly 31 (2014) 84–89

Contents lists available at ScienceDirect

Government Information Quarterly

j ourna l homepage: www.e lsev ie r .com/ locate /gov inf

Zero-based budgeting: Does it make sense for universal service reform?

Victor Glass ⁎, Stela Stefanova, Joseph PrinzivalliNational Exchange Carrier Association, 80 South Jefferson Road, Whippany, NJ 07981, USA

⁎ Corresponding author.E-mail address: vglass@neca.org (V. Glass).

0740-624X/$ – see front matter © 2013 Elsevier Inc. All rhttp://dx.doi.org/10.1016/j.giq.2013.05.022

a b s t r a c t

a r t i c l e i n f o

Available online 20 November 2013

Keywords:Universal service reformConnect America FundRLECTelecommunicationsTimes Interest Earned RatioFree cash flowNational Broadband Plan

The FCC's proposal for Universal Service Fund (USF) reformproposes to replace current universal service fundingprograms with the Connect America Fund (CAF), a fund devoted to using government support to make broad-band service available in un-served or under-served geographical locations. The goal is to maximize the bang-per-buck from available funds by looking at incremental upgrades to existing infrastructure in under-servedareas and incremental infrastructure build-outs in un-served areas. The strategy is to encourage competitionbetween broadband providers whether they are wireless providers or wireline providers.In essence this is a zero-based budgeting strategy inwhich the FCC begins with a clean slate and takes no accountof existing allocation of funds. This strategy, while it sounds reasonable in theory, threatens to bankrupt rurallocal exchange carriers (RLECs). As a result, it threatens not only universal broadband availability but also basicvoice service. We show the devastating financial effects of this zero-based budgeting plan on RLECs using reve-nue, demand and cost data from a survey sent to RLECs, as well as data available from the FCC economicmodel, which projects the likely technology winner at the county level and the amount awarded.The FCC has acknowledged that a flash cut in funding is not feasible. Instead, it has proposed freezing support peraccess line during a transition period.We show that this supposed glide path to the new supportmechanism alsoproduces devastating financial results for RLECs.Tests for RLEC financial distress include comparisons of support under current programs compared to supportunder the new CAF, change in RLEC free cash flow and RLECs' eligibility for Rural Utility Service loans.

© 2013 Elsevier Inc. All rights reserved.

Local Revenue(Including EUCL)

30%

Federal USF36% State USF

State AccessRevenue

12%

Interstate AccessRevenue

17%

1. Introduction

In its Connecting America: The National Broadband Plan, the FCC pro-poses to replace current universal service funding programs with theConnect America Fund (CAF), a fund devoted to making broadbandservice available in un-served or under-served geographical locations(FCC, 2010). The strategy is to maximize the bang-per-buck from avail-able funds by looking at incremental upgrades to existing infrastructurein under-served areas and incremental infrastructure build-outs in un-served areas. The FCC believes competition for support dollars betweenbroadband providers whether they are wireless or wireline providers,should reduce the cost of broadband deployment.

In essence the FCC adopts a zero-based budgeting approach for allo-cating support dollars to show the effectiveness of its plan. Using amodel, the FCC identifies which technology will likely extend broad-band availability in targeted areas at the lowest support funding cost.Afterwards, the FCC acknowledges that a flash cut in funding fromcurrent to new recipients is not feasible. Instead, the FCC proposes atransition period to allow current receivers of support enough time toprepare in case they lose funding.

ights reserved.

NECA and others filed comments strongly criticizing the FCC's zero-based budgeting strategy, whether it includes a transition period or not(NECA et al., 2010). The claim is that the FCC's plan fails to recognizethat existing USF recipients depend on this support for their very sur-vival. NECA quantified potential RLEC financial distress from loss of sup-port by projecting the size of the shift of funding away from RLECs to

5%

Fig. 1. Regulated revenue sources of RLECs.

Table 1Percent of current high cost support lost by state ($ in millions).

State PV of investment gap in RLECareas awarded to wireline

PV of investment gap in RLEC areasawarded to wireline and wireless

Annual gap payments in RLECareas awarded to wireline

Annual gap payments in RLEC areasawarded to wireline and wireless

Current highcost support

% Currentsupport lost

AK $1043 $1043 $133 $133 $71 87%AL $0 $44 $0 $6 $32 −100%AR $0 $113 $0 $14 $37 −100%AZ $131 $138 $17 $18 $40 −58%CA $117 $129 $15 $17 $56 −73%CO $105 $146 $13 $19 $24 −43%FL $0 $36 $0 $5 $15 −100%GA $0 $96 $0 $12 $66 −100%HI $0 $0 $0 $0 $24 −100%IA $0 $171 $0 $22 $70 −100%ID $284 $314 $36 $40 $27 34%IL $0 $91 $0 $12 $37 −100%IN $0 $58 $0 $7 $63 −100%KS $6 $234 $1 $30 $131 −99%KY $0 $59 $0 $8 $47 −100%LA $0 $25 $0 $3 $49 −100%MA $0 $0 $0 $0 $1 −100%MD $0 $0 $0 $0 $1 −100%ME $1 $30 $0 $4 $12 −99%MI $0 $58 $0 $7 $21 −100%MN $3 $400 $0 $51 $72 −100%MO $0 $194 $0 $25 $63 −100%MS $0 $88 $0 $11 $22 −100%MT $787 $958 $101 $122 $62 63%NC $0 $28 $0 $4 $32 −100%ND $1 $502 $0 $64 $59 −100%NE $23 $203 $3 $26 $43 −93%NH $1 $1 $0 $0 $8 −99%NJ $0 $0 $0 $0 $1 −99%NM $706 $748 $90 $95 $44 103%NV $169 $169 $22 $22 $11 88%NY $2 $21 $0 $3 $24 −99%OH $0 $14 $0 $2 $23 −100%OK $16 $216 $2 $28 $96 −98%OR $136 $141 $17 $18 $38 −54%PA $1 $54 $0 $7 $44 −100%SC $0 $75 $0 $10 $97 −100%SD $29 $451 $4 $58 $64 −94%TN $0 $123 $0 $16 $47 −100%TX $120 $504 $15 $64 $162 −91%UT $280 $284 $36 $36 $17 106%VA $1 $52 $0 $7 $15 −100%VT $6 $6 $1 $1 $11 −93%WA $32 $38 $4 $5 $23 −82%WI $1 $107 $0 $14 $72 −100%WV $0 $15 $0 $2 $6 −100%WY $205 $227 $26 $29 $25 6%Total $4206 $8406 $537 $1073 $2008 −73%

85V. Glass et al. / Government Information Quarterly 31 (2014) 84–89

other entities, especially wireless carriers. NECA also showed theresulting shift from positive to negative free cash flow1 for manyRLECs and their loss in ability to borrow funds from Rural Utilities Ser-vice (RUS) loans, a prime source of RLEC funding, because of a sharpdrop in their Times Interest Earned Ratio (TIER).2

Using revenue, demand and cost data from a new survey sent toRLECs in April 2011, we confirm NECA's original dire projections. Wego on to suggest that any reform plan for universal service shouldtreat as a priority preserving the financial viability of carriers of last re-sort because the outside plant used for telephone service also providesdigital subscriber line (DSL) service.3 Many RLECs fall into this category.

1 Free cash flow = earnings before interest, taxes, depreciation, and amortization(EBITDA) less change in telephone plant in service, less expense on long term interest, lessincome tax.

2 TIER equals earnings before interest, taxes, depreciation and amortization (EBITDA)divided by interest expense on long term debt. RUS requires a TIER of at least 1 by theend of the fifth year of a loan feasibility study.

3 A carrier of last resort is a local exchange company that accepts a set of obligations inreturn for economic benefits. Among the obligations are duties to extend retail service toany retail customer within its franchise area and provide interconnection and wholesaleservices to other carriers (NRRI, 2009).

Without them, rural customers nowwell-servedwon't be served broad-band at all. Controls and incentives for broadband deployment makesense that financial viability is reasonably assured.

The RLECs and larger carriers produced their own plans for universalservice reform. We recommend that each plan should be evaluatedbased on the likelihood that it will not causefinancial distress to carriersof last resort, and on its ability to fund broadband more efficiently TheFCC has made it clear that it believes companies facing no limits onsupport have the incentive to increase costs without regard to efficiency(FCC, 2012). This belief has support from the academic community.Recent academic studies suggest that current support programs mayincent companies to maximize payouts from support funds instead ofmaximizing operational efficiencies (see Berg, Jiang, & Lin, 2010, 2011;Zolnierek, 2008). The FCC's OPEX/CAPEX model is one of several newapproaches for allocating support more efficiently.4 Although funding

4 The FCC set limits on overall capital and operating expenditures, but declined to setlimits on eleven separate cost categories (USFBenchmarkingOrder, FCC, 2012) as original-ly proposed in the USF/ICC Transformation Order and FNPRM (FCC, 2011b). The FCC rec-ognized that a larger number of caps could limit carriers' ability to optimize spendingthrough tradeoffs among different categories.

Fig. 2. Note: Map generated at http://www.broadband.gov/maps/availability.htm.

86 V. Glass et al. / Government Information Quarterly 31 (2014) 84–89

efficiency is a critical issue, this paper focuses on the impact of the long-term strategic plan for shifting funding to entities that can providebroadband service at least cost.

The remainder of the paper is organized as follows: Section 2 de-scribes the Connect America Fund; Section 3 describes the estimatedfunds flow shifts caused by the CAF plan and the resulting loss of sup-port to RLECs and translates them into twomeasures of financial health,free cash flow and TIER; and Section 4 presents our conclusions.

2. The connect America fund

The current high cost fund of the Universal Service Fund (USF) isdesigned to meet the requirements of section 254 of the Telecommuni-cations Act of 1996, which requires that telecommunications and infor-mation services in high cost, rural and insular areas must be reasonablycomparable to those available in urban areas. In 2010, the five compo-nents of the high cost fund disbursed $4.3 billion to rate of return andprice-cap companies, with $2.0 billion going to rate of return studyareas (FCC, 2011a).

Concerned that “networks that provide only voice service areno longer adequate for the country's communication needs,” the FCChas proposed sweeping changes to the USF and the intercarrier

Table 2Percent of current high cost support lost by line size ($ in millions).

Number ofaccess lines

PV of investment gap inRLEC areas awarded towireline

PV of investment gap in RLECareas awarded to wirelineand wireless

AnnualRLEC arwirelin

Less than 500 $228 $307 $29Between 500 and 1000 $396 $606 $51Between 1000 and 2000 $328 $873 $42Between 2000 and 4000 $600 $1250 $77Between 4000 and 6000 $418 $1180 $53Greater than 6000 $2236 $4191 $285Total $4206 $8406 $537

compensation (ICC) system (FCC, 2011a). The FCC acknowledged thatthe USF and ICC systems “have not been comprehensively reassessedin more than a decade,” despite dramatic changes in telecommunica-tions technology (FCC, 2011a). The Commission's goal is to reorientthese programs from their focus on voice communication carried oncircuit-switched networks to ubiquitous broadband carried on IP net-works, while at the same time making the programs more efficient.

In the National Broadband Plan presented in March 2010, the FCCidentified 14million people in 7million households without broadbandavailable at the FCC's National Broadband Availability Target of 4Mbpsdownstream/1Mbps upstream. A business case would not justify pri-vate investment to close this “broadband availability gap” because ofthe high cost of deployment and ongoing operations in these unservedareas. The FCC estimated that the net amount of funding needed toclose the gap is $23.5billion (FCC, 2010).

The CAF the FCC proposes would directly subsidize broadband de-ployment. The current high cost program would be transitioned to theCAF in two stages. In the first stage, there would be immediate reformswhich the FCC believes would “start rate-of-return carriers on the pathtowards market-driven, incentive-based regulation” (FCC, 2011a).These reforms include reducing support amounts and limiting reim-bursement for capital and operating expenses. In the second stage, all

gap payments ineas awarded toe

Annual gap payments in RLECareas awarded to wireline andwireless

Currenthigh costsupport

% Currentsupportlost

Totalnumber ofaccess lines

$39 $47 −38% 39,800$77 $109 −53% 140,033

$111 $248 −83% 343,024$160 $328 −77% 610,386$151 $294 −82% 601,826$535 $982 −71% 3,397,522

$1073 $2008 −73% 5,132,591

5 Available at http://www.broadband.gov/maps/availability.htm. Accessed on October21, 2011.

6 As there is not a one-to-one correspondence between counties and study area bound-aries, the percent of overlap between study areas and counties was computed. If a countyfell entirelywithin the boundary of a study area, the county's investment gapwas assignedto the study area. If a county fell only partially within the boundary of a study area, thecounty's investment gap was estimated as the percent of its geographic area within thestudy area multiplied by the county's investment gap. Finally, the investment gap wassummed across all counties either wholly or partially served by the RLEC.

-$4

-$3

-$2

-$1

$0

$1

$ in Millions

2011

2011

2012

2012

2013

2014

Free Cash Flow per study area per year (transition in effect)

Free Cash Flow per study area per year (status quo)

2014

2015

2015

2016

2017

2018

2018

2019

2020

2021

2021

2022

2022

2023

2024

2024

2025

2025

2026

2027

2028

2028

2029

2030

2031

2031

2032

Fig. 3. Free cash flows per study area per year.

87V. Glass et al. / Government Information Quarterly 31 (2014) 84–89

remaining high cost programswould be transitioned to the CAF, provid-ing ongoing support for broadband service in areas that are uneconomicto serve.

Nevertheless, after the transition to the CAF, the FCC would employ“company- and technology-agnostic” criteria to award funding to oneprovider, making it a zero-based budgeting approach. Under this ap-proach, companies that already have operations in areas that can't sus-tain a network provider could see sharp cuts in funding. The NationalBroadband Plan acknowledges that as high cost funding is shifted tothe CAF, the “FCC…should take into account the impact of potentialchanges in free cash flows on providers' ability to continue to providevoice service and on future broadband network deployment strategies”(p. 165, FCC, 2010).

To estimate the cost of closing the broadband availability gap,the FCC analysis proceeded in two broad phases using a “BroadbandAssessment Model.” In the first phase, the FCC estimated the currentavailability of broadband. Themodel characterized broadband coveragein the 8.2million census blocks of the country by download speed andtype of broadband technology, using a combination of commerciallyavailable and predicted information. The result was a list of servedand unserved census blocks, with a block falling into the unservedcategory if it did not have DSL, cable, or wireless service meeting theFCC's target of 4Mbps downstream/1Mbps upstream. As noted above,7million households were classified as unserved.

In the next phase of the analysis, the FCC determined the cost ofbringing the unserved areas up to the targeted service level by usingengineering models to compare the costs of upgrading or extendinginfrastructure with a variety of terrestrial technologies, taking into ac-count wireless 4G build-out. The FCC estimated the cash flows frombuilding and operating the network over a 20 year period, consideringaverage revenue per user, initial, ongoing, and replacement capital ex-penses, operating expenses, sales, general, and administrative expenses,depreciation, cost of money, and taxes. The investment gap of$23.5billion is the net present value of these cash flows.

3. Projected RLEC support flow reductions

In 2011 NECA collected regulated revenue, demand and costs datafrom RLECs for 2008, 2009 and 2010. The data request was sent to allRLECs participating in NECA revenue pools. Six hundred and fortyeight companies (57%) responded with partial or complete data. Afterexcluding submissions from companies which left most of the fieldsblank, we use the remaining 634 responses to evaluate the potentialimpacts of FCC proposals on RLECs to cap and phase out existing costrecovery mechanisms.

To emphasize the extent towhich rural telephone companies rely onsupport mechanisms for cost recovery we provide information oncurrent sources of revenue for RLECs. Fig. 1 shows that over 40% of

regulated revenues come from federal or state USF. Almost 30% of reve-nues are associated with Inter- and Intrastate Access and the remaining30% of revenues come from end user revenues.

We estimate that if CAF funding as determined by the FCC Broad-band Assessment Model completely replaces current USF mechanisms,rural telephone companies are to lose 73% of their current support. Toarrive at this estimate we use investment gap amounts and the lowestcost technology provided by the FCC5 at the county level to calculatethe present value of levelized CAF payments awarded to wireline tech-nology in RLEC serving territory.6 As we are interested in evaluating thelong term effect of the transition of USF to CAFwe assume for simplicitythat CAF payments will be spread equally over time for both operatingshortfalls and capital expenditures. Since wireless and wireline compa-nies compete for funds under the proposed NBP, only a portion of thetotal funding will be awarded towireline companies. In RLEC territorieswhere wireline is identified as the lowest-cost technology, pool mem-bers are assumed to supply all digital subscriber lines (DSL) and fiberto the premises (FTTP) infrastructure. In Table 1 we report total supportamounts awarded to RLEC study areas regardless of technology as wellas amounts awarded to wireline only. Over the next 20years, RLECs areestimated to receive annual CAF payments of $537 million in presentvalue dollars for both operating shortfalls and capital expenditures.Current annual USF support for these companies is $2 billion. Thus,CAF funds available to all RLECs are 27% of current universal servicepayments. The FCC also proposes to reduce and ultimately eliminateswitched access charges during the same period, thus eliminating anadditional source of network cost recovery for RLECs.

While RLECs inmost states lose all of their current support, there aresome companies that see an increased inflow of funds. We sum up theUSF and CAF amounts for RLEC territories within each state. At thestate level rural companies in Alaska, Idaho, Montana, Nevada, Utah,and Wyoming end up with higher support under the new mechanism.Except Alaska these are all Rocky Mountain states. Fig. 2 displays amap generated from the NBP website. It shows investment gap perhousing unit by lowest cost technology. Wireless technology is shownas the cheapest technology to provide broadband to unserved and un-derserved areas in all but the counties in the West region of theUnited States.

0%

10%

20%

30%

40%

50%

60%

Fig. 4. Percent of study areas in sample falling below TIER=1 (transition in effect).

88 V. Glass et al. / Government Information Quarterly 31 (2014) 84–89

Table 2 compares funding under USF and CAF by size of the compa-ny. We grouped RLECs into 6 categories by their number of access lines.The smallest rural companies, which in total account for about 40,000lines lose 38% of their current support and the largest RLECs, which ac-count for about 3.4million lines lose 71%.

So far we have shown the long term view of the impact of the pro-posed USF reform; that is, when CAF is the only funding mechanismleft to help RLECs recover their cost. For the short term, the FCC has al-ways stated that there will be no flash cuts. Current USF will betransitioned out over some period of time to the new funding mecha-nism. Next we model the impacts of the transition. As all details andthe exact length of this transition are still unknown, we assume equalreductions in USF funding over 20 years. These are conservative esti-mates as the transition period will likely be shorter.7 Using currentgrowth factors for revenues, demand and costs, we analyze the poten-tial impacts of reducing USF support on free cash flows for an averagecompany using a sample of 634 RLECs. An average company in the sam-ple has 4500 access lines.8 Free cashflow is calculated as earnings beforeinterest, taxes, depreciation, and amortization (EBITDA) minus capitalexpenditures (CapEx) minus interest expense on long term debtminus income tax. We calculate EBITDA using revenue data from thesurvey. Capital expenditures are computed by taking the difference ofthe value of telephone plant in service between two years. Interest ex-penses are reported by companies and income tax is calculated for tax-able RLECs using an income tax of 35%.

We show in Fig. 3 the trends in free cashflowunder a status quo andunder a transition scenario. RLECs experience negative free cash flowsin 2014 when USF is decreased. It is interesting to see that free cashflows become negative, granted it is three years later and much lessnegative, even if funding mechanisms do not change.9 The dire

7 The FCC established the CAF and began the transition of legacy support by eliminatingLocal Switching Support and creating CAF ICC. While the ultimate goal of the transition isto distribute all high-cost support through CAF, in the near term, rate of return carrierscontinue to receive the remainder of their universal service support through the reformedhigh-cost supportmechanisms. Using themost recently filed informationwhich is public-ly available from the USAC and FCCwebsites we calculate the impact of the reformed highcost supportmechanism for the first year of transition. The reduction due to the near-termreforms outlined in the USF/ICC transformation Order is about 4% in high cost loop sup-port. This decrease does not take into account the 5% reduction in switched revenue re-quirement per year mandated by the Order as well as the potential loss of support dueto the presence of an unsubsidized wireline competitor. It also does not make an adjust-ment to high cost loop support for the low rate benchmark.

8 To give a better sense for the diversity of RLECs, themedian size of RLECs in the sampleis 2143 access lines, with a range from 49 lines to 69,193 lines.

9 We have used current growth assumptions to project revenue and cost trends.Switched access demand – traditional landline voice service – is declining by about 13%per year. Access lines are declining by about 5% per year. Special access demand is increas-ing by about 9% per year, but is likely to slow down as the market matures and DSL de-mand and backhaul demand from wireless companies flatten out.

projections for RLECs suggest that to keep these companies in businessand keep rural America connected, reform beyond CAF is necessary.

If companies do not have any free cash available from normal oper-ations, they will need to borrow to be able to run their businesses andcontinue to invest in their networks. With negative free cash flowsthis will not be likely. To qualify for loans, RLECs usually have to meetstrict requirements regarding free cash flow and Times Interest EarnedRatio (TIER). The Rural Utilities Service (RUS) has noted that 99% ofthe borrowers in its Telecommunications Infrastructure Loan Program,the most rural of its portfolios, receive interstate Universal ServiceFund support (RUS, 2011). In reviewing a loan application, the RUS eval-uates a potential borrower's ability to service its debts by projectingTIER in a feasibility study. By the end of a five-year forecast period, theborrower must achieve a projected TIER of at least 1.0 to qualify for aloan. Moreover, the borrower must be able to maintain the TIER thatis projected at the end of the five-year forecast period. The Rural FinanceTelephone Cooperative, which plays an important role for RLECs bysupplementing RUS loans, requires applicants to achieve and maintaina TIER of 1.5 (RTFC, 2011).

Using our projections for revenue and interest expense we calculateTIER as EBITDA divided by interest expense for companies reportingpositive interest expense in 2010.We assume constant interest expensethrough the transition period. We found that only 5% of RLECs have aTIER less than 1 at the beginning of the period and close to 50% fallbelow the statutory requirements for TIER in 2025 when USF transitionis in effect (Fig. 4).

4. Conclusions

To address the potentially harmful financial impacts of FCC reformproposals, rural associations filed their own proposal to reform the Uni-versal Service Fund and Intercarrier Compensation (NECA et al., 2011).10

It recognized the decline in switched access revenues and the transitionto broadband. Although rules are still under development, the USF/ICCTransformation Order and FNPRM (FCC, 2011b) suggested a more cau-tious approach than initially proposed. While many of the details of theproposed reform are now known, they address only short-term issues re-lated to the current high cost fund and near-term funding as the FCCmoves switched access rates to bill and keep. It is clear from our analysisthat robust reform must address the larger issue of how to maintain thelong term financial health of RLECs and provide incentives to increasebroadband availability and subscription rates in a world of increasingcompetition, rapidly developing communication technologies and largernetwork capacity requirements.

10 Other parties also filed proposals (e.g. AT&T et al., 2011; State Members of the FederalState Joint Board, 2011).

89V. Glass et al. / Government Information Quarterly 31 (2014) 84–89

References

AT&T, CenturyLink, FairPoint Communications, Frontier Communications, Verizon, &Windstream (2011). America's Broadband Connectivity (ABC) Plan framework.

Berg, S., Jiang, L., & Lin, C. (2010). Incentives for cost shifting and misreporting: U.S. ruraluniversal service subsidies, 1991–2002. Working paper.

Berg, S., Jiang, L., & Lin, C. (2011). Universal service subsidies and cost overstatement:Evidence from the U.S. telecommunications sector. Telecommunications Policy, 35,583–591.

Federal Communications Commission (2010). Connecting America: The National Broad-band Plan. (rel. March 16, 2010).

Federal Communications Commission (2011a). Notice of proposed rulemaking and furthernotice of proposed rulemaking, 26 FCC Rcd 4554.

Federal Communications Commission (2011b). Connect America Fund and intercarriercompensation reform order and FNPRM.

Federal Communications Commission (2012). Connect America Fund, high cost universalservice support order.

National Exchange Carrier Association, Inc., National Telecommunications CooperativeAssociation, Organization for the Promotion and Advancement of SmallTelecommunications Companies, & Western Telecommunications Alliance(2011). Comments.

National Exchange Carrier Association, Inc., National Telecommunications CooperativeAssociation, Organization for the Promotion and Advancement of SmallTelecommunications Companies, Western Telecommunications Alliance, &Rural Alliance (2010). Joint comments.

National Regulatory Research Institute (2009). Carriers of last resort: Updating a traditionaldoctrine.

Rural Telephone Finance Cooperative (2011). Ex Parte presentation of Rural TelephoneFinance Cooperative.

Rural Utilities Service (2011). Overview of telecommunications and broadband loan andgrant programs. Ex Parte presentation of Rural Utilities Service.

State Members of the Federal State Joint Board (2011). Comments by state members of theFederal State Joint Board on universal service.

Zolnierek, J. (2008). Loop density and telephone costs: Panel data evidence.Telecommunications Policy, 32, 262–272.

Victor Glass holds the position of Director of Demand Forecasting and Rate Developmentat theNational Exchange Carrier Association (NECA). At NECA since 1984,Mr. Glass is cur-rently responsible for forecasting demand and setting switched and special rates for morethan 1100 telephone companies. During recent years, he has been heavily involved in ac-cess restructure, universal service reform, and new access services. He is the lead author ofNECA's “Last Mile,” “Middle Mile,” and “Packet Train” broadband studies and NECA's“Rural Cost Trends” study. He has contributed to Legg Mason's quarterly Equity ResearchReport, to Prudential Equity Group's Telecom/Rural Wireline Report, and served as an ad-visor to the Telecommunications Group at Goldman Sachs. His articles have appeared inAmerica's Network, Rural Telecommunications, Xchange, Opastco Advocate, GovernmentInformation Quarterly, Journal of Business Forecasting, Energy Economics, American Econ-omist, and Journal of Regulatory Economics. He is a member of the Rural Policy ResearchInstitute's telecommunications panel and the Center for Research in Regulatory Industries(CRRI). Dr. Glass earned his MBA in marketing and finance, and Ph.D. in economics fromColumbia University.

Stela Stefanova is Manager, Common Line Rate Development and Access Reform Analysisat the National Exchange Carrier Association (NECA). She is responsible for rate develop-ment, statistical and quantitative analysis of data and policy proposals and GIS analyses.Before joining NECA she worked as an Economist with the Economic Research Service atthe U.S. Department of Agriculture (USDA) and taught Economics and Statistics coursesat the University of Delaware. She has published articles in the Journal of RegulatoryEconomics and the Journal of Natural Resources Policy Research. She has also authoredUSDA reports and a book chapter on Environmental Valuation. Stefanova has a doctoratein Economics from the University of Delaware.

Joseph Prinzivalli is Manager, Rate Development & Quantitative Analysis at the NationalExchange Carrier Association (NECA). He manages development of interstate access ratesand analysis of the financial impacts of intercarrier compensation reform. He contributedmaterial on Voice over IP and IPTV to NECA's report, “The Packet Train Needs to Stop atEvery Door.” Prior to his job at NECA, Joseph held a variety of positions at AT&T with afocus on business-to-businessmarket research, pricing policy, revenue analysis in supportof pricing recommendations, and customer satisfactionmeasurement. He holds amaster'sdegree in statistics from Columbia University and a master's degree in economics fromRutgers, The State University of New Jersey.