Federal Communications Commission FCC 13-16 STATEMENT OF COMMISSIONER AJIT PAI APPROVING IN PART AND CONCURRING IN PART Re: Connect America Fund, WC Docket No. 10-90, High-Cost Universal Service Support, WC Docket No. 05-337. In the Universal Service Transformation Order, the Commission adopted benchmarks for the High-Cost Loop Support (HCLS) component of the Universal Service Fund to create “structural incentives for rate-of-return companies to operate more efficiently and make prudent expenditures.”1 The Commission proposed benchmarks based on a quantile regression analysis (QRA) of the correlation between the characteristics of various rate-of-return carriers and their historical capital and operating expenses but left the details to be decided by the Wireline Competition Bureau. Five months later, the Bureau adopted the Benchmarks Order, which established separate QRA benchmarks for each rural carrier’s capital expenses and operating expenses and phased in those benchmarks over 18 months.2 Like my colleagues, I believe that establishing limits on the universal service support a carrier can receive is a good thing. In this era of fiscal restraint, no one can expect the government to continue to fund their expenses without question. And with respect to HCLS, the Commission instituted such limits almost twenty years ago when it capped the program at the recommendation of the Federal-State Joint Board.3 It is important to recognize that the limits established in the Benchmarks Order and that we address here do not reduce the size of the HCLS program.4 They instead determine how universal service funding will be distributed among rural carriers, not how much will be spent overall. With that background, I turn to today’s order. I join it to the extent that it mitigates the harm of the QRA benchmarks and provides short-term relief for rural carriers. But I am disappointed that we do not go further to reform these benchmarks. As I have said before, universal service support should be stable and predictable and distributed consistent with the law and common sense.5 I am not so sure that the QRA benchmarks that we reaffirm today pass that test. First, the QRA benchmarks are unpredictable. As it stands, the Bureau is scheduled to recalculate the QRA benchmarks each year to reflect the investment and operating decisions of not just each individual rural carrier but its 737 brethren as well. Because this determination is relative, a carrier 1 Connect America Fund; A National Broadband Plan for Our Future; Establishing Just and Reasonable Rates for Local Exchange Carriers; High-Cost Universal Service Support; Developing a Unified Intercarrier Compensation Regime; Federal-State Joint Board on Universal Service; Lifeline and Link-Up; Universal Service Reform— Mobility Fund, WC Docket Nos. 10-90, 07-135, 05-337, 03-109, CC Docket Nos. 01-92, 96-45, GN Docket No. 09- 51, WT Docket No. 10-208, Report and Order and Further Notice of Proposed Rulemaking, 26 FCC Rcd 17663, 17742, para. 210 (2011) (Universal Service Transformation Order). 2 Connect America Fund; High-Cost Universal Service Support, WC Docket Nos. 10-90, 05-337, Order, 27 FCC Rcd 4235, 4259, para. 70 (Wireline Comp. Bur. 2012) (Benchmarks Order). 3 See Amendment of Part 36 of the Commission’s Rules and Establishment of a Joint Board, CC Docket No. 80-286, Report and Order, 9 FCC Rcd 303, 304, para. 11 (1993); see also Federal-State Joint Board on Universal Service; Multi-Association Group (MAG) Plan for Regulation of Interstate Services of Non-Price Cap Incumbent Local Exchange Carriers and Interexchange Carriers, CC Docket Nos. 96-45, 00-256, Fourteenth Report and Order, Twenty-Second Order on Reconsideration, and Further Notice Of Proposed Rulemaking in CC Docket No. 96-45, and Report and Order in CC Docket No. 00-256, 16 FCC Rcd 11244, 11259–68, paras. 31–53 (2001) (reviewing the history of the HCLS cap and modifying it). 4 Sixth Order on Reconsideration, para. 19. 5 Remarks of FCC Commissioner Ajit Pai, “Unlocking Investment and Innovation in the Digital Age: The Path to a 21st-Century FCC” (July 18, 2012), available at http://go.usa.gov/4QJG; Opening Remarks of Commissioner Ajit Pai at the Rural Broadband Roundtable in Oswego, Kansas (Sept. 6, 2012), available at http://go.usa.gov/4QJz. Federal Communications Commission FCC 13-16 2 has little ability to predict what its benchmarks will be in one, three, or five years unless it somehow guesses what the entire rural industry will do in the future. The order suggests that QRA benchmarks in one year may be “strongly predictive” of benchmarks the next.6 But this suggestion is based on 2010 and 2011 cost data, the two years before the QRA benchmarks were adopted and implemented. As a stockbroker might say: Past performance does not necessarily indicate future results. That’s especially true here, when the whole point of the QRA benchmarks is to induce rural carriers to reduce their spending, which will necessarily feedback into QRA benchmarks for future years.7 Moreover, just this month the U.S. Department of Agriculture asked the Commission for “confidential access to the regression model to assist [the Department] in managing its lending program,”8 presumably because the Department, like rural carriers, thinks that predictability is a weakness—not a strength—of the QRA benchmarks. Indeed, as a recent six-month study of the QRA benchmarks found, “the QRA is a complicated statistical analysis that cannot be predicted—at least at the present—with any degree of confidence.”9 This unpredictability does not promote certainty. And it appears the investment environment has cooled as a result, impeding the deployment of next-generation technologies and broadband services to rural Americans.10 Indeed, the Department of Agriculture recently told the Commission that demand for Rural Utility Service loans for broadband build-out has plummeted this year due in part to the uncertainty created by the QRA benchmarks.11 Communications infrastructure requires not a one- or two-year investment, but a ten- or twenty-year commitment. If a rural carrier does not know whether or how much universal service funding will be available, it will not invest in the higher-cost regions of America. Indeed, it may not even have the choice, as banks and investors are unlikely to lend to rural carriers with uncertain funding. Unpredictability also undermines the stated purpose of the QRA benchmarks, which is to incentivize efficient and prudent decision-making by rate-of-return carriers. The HCLS mechanism is a backward-looking mechanism within the Universal Service Fund, meaning that support received today is based on—and is payment for—historical, two-year-old expenditures. Thus, the 2012 QRA benchmarks were based on 2010 expenses, and this year’s 2013 QRA benchmarks are based on data the National 6 Sixth Order on Reconsideration, para. 15 & note 89. 7 See, e.g., Vincent H. Wiemer & Michael J. Balhoff, CFA, White Paper: Lessons from Rebuilding the FCC’s Quantile Regression Analysis at 28 (Feb. 2013) (Rebuilding QRA White Paper), available at http://go.usa.gov/4he4 (“[T]he effect of the use of the model . . . is to create a much higher degree of unpredictability and to incent very conservative levels of spending by an individual carrier so that it does not risk shortfalls in recovery on its high-cost spending. Then, if most carriers take this approach each year as would be rational, each subsequent year becomes more conservative and there is a potential ‘race to the bottom.’”). 8 See Letter from John Charles Padalino, Acting Administrator, Rural Utility Service, to Marlene H. Dortch, Secretary, FCC, WC Docket Nos. 10-90, 07-135, 05-337, GN Docket No. 09-51, CC Docket Nos. 01-92, 96-45, WT Docket No. 10-208, at 2 (Feb. 15, 2013) (Dep’t of Agriculture Ex Parte Letter). 9 Rebuilding QRA White Paper at 28; see id. at 68 (“The new QRA is duck hunting when the winds are high, the distance is farther, and, for sport, there is no light.”). 10 See US Telecom Application for Review at 4–5 (arguing that the “perception of significant inaccuracy and unpredictability . . . whether correct or not, is causing widespread fear in the cost-company rate-of-return community which is having a chilling effect on broadband investment plans”). 11 See Dep’t of Agriculture Ex Parte Letter at 1–2 (reporting on a meeting between Secretary Vilsack and Chairman Julius Genachowski and noting that “demand for [Rural Utility Service] loan funds dropped to roughly 37% of the total amount of loan funds appropriated by Congress in [fiscal year] 2012”). Federal Communications Commission FCC 13-16 3 Exchange Carrier Association collected last July about calendar year 2011 costs.12 In other words, a rural carrier seeking to adjust its operating expenses today needs to know what the QRA benchmarks will be two years from now—in 2015—because those are the benchmarks that will apply to today’s spending. And because the costs of capital investment are generally spread over several years, not just one, a rural carrier making investments this year would likely need to know the QRA benchmarks for 2015, 2019, and even thereafter. Yet a rural carrier cannot even guess its 2015 QRA benchmarks because they will be based on an entirely new model and may or may not (today’s order leaves that to the Bureau’s discretion) bear some relation to the 2014 benchmarks.13 Unpredictability also unnecessarily imperils the financial situation of rural carriers. Consider a small business whose expenditures are just above the 2013 QRA benchmark. Because the support it receives in 2013 reimburses money already spent, the unexpected decrease in support may force the carrier to cut its current costs dramatically—not because doing so is efficient but because it may be the only way to make payroll and keep the lights on. If the carrier then falls below the 2014 QRA benchmark, it may suddenly feel flush with cash . . . only to see another large drop in 2015. This whip- saw effect is itself an impediment to long-term financial planning and efficient investment, and the unpredictability of the QRA benchmarks from one year to the next exacerbates this problem. This is not just a matter of good policy. Promoting predictability is a congressional command: Alongside the basic universal service principle of serving consumers “in all regions of the Nation,” section 254 requires that the Commission run the Universal Service Fund on the principles that there be “specific, predictable and sufficient . . . mechanisms to preserve and advance universal service.”14 To be sure, this predictability principle can yield to countervailing statutory principles,15 but none of the latter pertain here. And though this principle certainly does not require that we promise a particular carrier a particular amount,16 surely Congress intended that we give carriers at least some guidance—at least enough so that the QRA benchmarks can serve the purpose we set for them. Second, although the 2013 QRA benchmarks make more sense as a result of today’s order (more on that later), I still have my doubts about their utility. The 2013 QRA benchmarks do not, for example, incentivize efficient investment since they apply to expenses incurred in calendar year 2011—and the Universal Service Transformation Order was not released until November of that year. They do not plausibly redirect support from low-cost areas to high-cost areas because, even after today’s order, carriers like Copper Valley Telephone and Arctic Slope Telephone will have lower caps merely because they serve Alaska.17 They do not target inefficient carriers (only “outliers”), nor do they encourage broadband deployment.18 In short, I am concerned that the 2013 QRA benchmarks may just end up arbitrarily redistributing support from one group of carriers to another. 12 See 47 C.F.R. § 36.611. 13 Sixth Order on Reconsideration, para. 16 (“We direct the Bureau, as it updates the benchmarks for 2014, to consider whether these benchmarks should be held constant for multiple years.”). 14 47 U.S.C. § 254(b)(2), (3), (5). 15 See Alenco Communications v. FCC, 201 F.3d 608, 621 (5th Cir. 2000). 16 Id. at 623. 17 Indeed, several studies have shown that costs of infrastructure deployment in Alaska are significantly higher than costs in the lower 48 states. See Rebuilding QRA White Paper at 21. 18 Despite post-adoption language in today’s order suggesting that the QRA benchmarks “provid[e] additional support for carriers below their caps to extend broadband to rural consumers,” Sixth Order on Reconsideration, para. 2, there is no connection between the additional support given to rural carriers below their caps and broadband (continued...) Federal Communications Commission FCC 13-16 4 Moreover, the data underlying the benchmarks are themselves flawed.19 The Benchmarks Order forthrightly admitted that the Commission does not keep data on carrier study area boundaries.20 Instead, the Benchmarks Order relied on a commercial database for 2012 and 2013 while it put in place an 18- month process to correct errors and address inaccuracies in that database’s study area boundary information.21 The new boundary information hopefully will benefit those subject to the 2014 QRA benchmarks. But the 2013 QRA benchmarks are tainted by the inaccuracies and errors inherent in the admittedly flawed mapping data. And the only comprehensive study of the benchmarks conducted to date found significant problems with fourteen of the sixteen variables used to produce them.22 I nevertheless approve today’s order in part because it provides some short-term relief for rural carriers that will be (and have been) affected by the 2013 QRA benchmarks. For example, one problem with the Benchmarks Order was that it analyzed the capital expenditures (capex) and operating expenses (opex) of carriers separately. So if a small carrier like Council Grove Telephone in Kansas exceeded its capex limit by $13, it would be capped even though its opex was more than $300 below par. The same is true for Hemingford Cooperative in Nebraska, whose capex exceeded its limit by $6 but had operating expenses $40 below the benchmark. The Benchmarks Order established these separate limits even though that item recognized that carriers should trade off capital investments and operating expenses to minimize the total cost of the network.23 Today’s order recognizes that problem24 and corrects it by summing the separate QRA benchmarks for the remainder of 2013.25 This salutary measure alone should reduce the number of capped rural carriers from 159 to 70, sparing small carriers like Council Grove and Hemingford Cooperative. One other positive step merits attention. For capped carriers, the order adds a 15 percent backstop to limit the total amount of support a rural carrier can lose in 2013 based on the QRA benchmarks.26 With this step, rural carriers will have a slightly easier time managing their cash flow in 2013 and, hopefully, keeping their doors open. (Continued from previous page) build-out. Indeed, if a rural carrier now below the cap chooses to reinvest that support in broadband, it risks pushing itself over the cap in future years, thus mitigating any benefits it may have received. This, in part, may be why there has been almost no support in rural America for the current QRA benchmarks. 19 This concern was recently reinforced by both the Department of Agriculture, which has asked the Commission to “correct the structural and data integrity concerns of the rural carriers,” Dep’t of Agriculture Ex Parte Letter at 2, and an independent six-month study of the QRA benchmarks, see Rebuilding QRA White Paper at 17–19, 21. 20 Benchmarks Order, 27 FCC Rcd at 4245, para. 25 (relying on Tele Atlas study area boundary data instead). 21 Id. at 4246, paras. 27–28. 22 See Rebuilding QRA White Paper at 17. 23 See, e.g., Benchmarks Order, 27 FCC Rcd at 4265, para. 96 (Appendix A) (hypothesizing that “in locations where trenching is unusually expensive, an efficient carrier may install aerial plant (use poles rather than trench) . . . [which] would involve lower capital costs than trenching, but higher future operations costs”). 24 Sixth Order on Reconsideration, para. 26 (noting that “carriers often consider the trade-offs between capital costs and operating expenses when making investment decisions” and recognizing that the Benchmarks Order gave no credit to companies that did so to minimize costs). 25 Id., para. 29. 26 Id., para. 30. Federal Communications Commission FCC 13-16 5 But even with these modifications, many carriers will still feel the arbitrary impact of the 2013 QRA benchmarks. For example, more than 70 percent of the carriers that will be capped in 2013 will face lower QRA benchmarks than they did in 2012.27 And while the QRA benchmarks went down, costs per loop went up as line loss continued throughout rural America.28 This double whammy means that a carrier like Bridgewater Telephone in Minnesota, which was not capped in 2012, will be capped in 2013 even though it cut more than $125,000 in operating expenses and $135,000 in capital expenditures from one year to the next. Going forward, I remain hopeful that we will address the unpredictability of the caps that will apply in the years to come. Two thoughts come to mind on that front: For one, we should consider holding the QRA benchmarks constant for several years with annual adjustments for line loss.29 Had we done that this year, the 70 carriers now capped would have dropped to 54, and almost 80 percent of the capped carriers would have had higher 2013 QRA benchmarks than they will after today’s order. Moreover, holding the benchmarks constant for a series of years will allow rural carriers to make longer term investments and adjust their operations because they will be able to actually predict the benchmarks that will apply to them in the future. For another, we should consider phasing in new, reduced QRA benchmarks over the course of a year to ease the financial impact that unexpected changes may cause to rural operations. Such a phase-in would be easily administrable if, for example, the Commission simply averaged the old and the new benchmarks whenever recalculation caused the new benchmarks to fall.30 That simple step should address the unpredictability of the QRA benchmarks and the order’s observation that if the QRA “benchmarks are updated infrequently, each update could cause larger and more sudden changes in support levels, at least for a subset of carriers.”31 And phasing in the QRA benchmarks for 2014 is especially important since they will be based on an entirely new set of data and an entirely new model. It is my sincere hope that the Commission will, in the coming weeks, seek to address those concerns. I stand ready to work with my colleagues to help make that happen. 27 In other words, of the 70 carriers that will be capped in 2013, 50 carriers had lower capex-plus-opex limits in 2013 than in 2012. 28 The rural growth factor, tied in our rules to the annual growth or contraction of supported rural loops, is -2.5527 percent for 2013. See USAC, Federal Universal Service Support Mechanisms Fund Size Projections for Second Quarter 2013 at 12. 29 To illustrate, consider Bridgewater Telephone, whose 2013 QRA benchmark will be $796.02 after today’s order. Under this proposal, Bridgewater’s 2013 QRA benchmark would be last year’s benchmark ($814.20) times 1.1666 (5,834/5,001) to reflect the loss of 833 lines from one year to the next, for a total of $949.82. Because Bridgewater’s uncapped costs per line were $896.79, it will be capped after today’s order but would not have been under this proposal. 30 To illustrate, again consider Bridgewater Telephone. Because the recalculated benchmark for 2013 ($796.02) was lower than the benchmark for 2012 ($814.20), under this proposal the 2013 benchmark would be the average of the two ($805.11). 31 Sixth Order on Reconsideration, para. 15.