Ten Things About ACP that Ted Cruz Cares About: #7 ACP and BEAD

Submitted on June 5, 2024

Benton Institute for Broadband & Society

Wednesday, June 5, 2024

Digital Beat

Ten Things About ACP that Ted Cruz Cares About

#7 ACP and BEAD

We're sharing ten questions about the Affordable Connectivity Program (ACP) that Senate Commerce Committee Ranking Member Ted Cruz (R-TX) asked New Street Research Policy Advisor and Brookings Nonresident Senior Fellow Blair Levin testified after a hearing entitled The Future of Broadband Affordability.

7. Your written testimony states: “But of course, if ACP goes away, those savings will also go away and to put a fine point on it, it is a mathematical certainty that there will be communities in, for example, Texas that instead of getting fiber will end up with fixed wireless or even satellite.”

a. Please provide specific numbers and equations that support the “mathematical certainty” portion of your statement.

The key points to Levin's answer are briefly summarized immediately below, with a more complete answer provided thereafter.

The Broadband, Equity, Access, and Deployment (BEAD) Program does not fund the cost of deployment broadband to an unserved or underserved area; rather it funds the difference between the cost of deployment and what a provider would be willing to invest to serve that area.
ACP increases the revenue for any such area, those closing the gap described above and decreasing the amount of BEAD funding necessary for that specific area, making the funds available for other areas.
BCG found that ACP reduces the amount needed to incentivize ISP investment by about 25% per location or $500.
A conservative projection indicates that the number of locations that will miss out on the opportunity to obtain fiber due to the end of the ACP program will be at least 47,705, though the actual number is likely to be larger.

As I noted in my written testimony, the consulting group BCG found that ACP reduces the subsidy needed to incentivize providers to build in rural areas by 25% per household, writing “the existence of ACP, which subsidizes subscriber service fees up to $360 per year, reduces the per-household subsidy required to incentive ISP investment by $500, generating benefit for the government and increasing the market attractiveness for new entrants and incumbent providers."

This makes sense from a capital markets perspective. After all, BEAD is not designed to fund 100% of the cost of any deployment. It will fund the difference between the cost of the deployment and what the private market would have paid to deploy a network to serve a designated geographic area. The greater the revenue, and hence likely profit, that the area would provide, the greater the private market’s willingness to invest, reducing the amount of BEAD funding necessary for the specific project.¹

So, let’s apply this to the situation in Texas, using numbers provided in a January 2024 report by Cartesian, a consulting group, and ACA, a trade association representing small cable providers.

Texas, under the Infrastructure Bill’s BEAD formula, will receive $3.3 billion, with an obligation to use the funds to connect all unserved locations and, if there are additional funds, to connect all underserved locations.

Cartesian finds that Texas has 364,700 unserved locations and 199,100 underserved locations. The current number is higher, but the FCC’s RDOF and EACAM programs are currently set to serve 322,434 unserved and underserved Texas locations.² Thus, that leaves approximately 563,800 locations as the responsibility of BEAD funds.

Cartesian estimates that the ISPs bidding for BEAD funds will provide $1.6 billion for the projects, bringing the total to be utilized for the capital expenditures of deployment to $4.9 billion. That would average out to about $8,675 per location, though in reality there is a wide variation in the actual cost to serve different locations. Cartesian estimates that the average private sector contribution would be $2,903.³

However, it costs more than $8,675 to connect the average location to a fiber connection. Cartesian estimates that it will cost, on average, $12,399 to connect an unserved location to fiber and an average of $10,468 to connect an underserved location to fiber. Therefore the BEAD contribution necessary for unserved locations would need to be $9,562⁴ and the BEAD contribution for underserved locations would need to be $7,631.⁵

Again, according to BCG, reduces the per-household subsidy required to incentive ISP investment by $500.⁶ Assuming the Cartesian study assumed a continuation of ACP,⁷ the BEAD contribution must be increased by $500, so the amount necessary for unserved locations would be $10,062,⁸and the amount for underserved locations would be $8,131.⁹

This then leaves us with four relevant numbers:

The average BEAD subsidy with ACP needed for a fiber connection to an unserved location at $9,562.
The average BEAD subsidy without ACP needed for a fiber connection to an unserved location at $10,062.
The average BEAD subsidy with ACP needed for a fiber connection to an underserved location at $7,631.
The average BEAD subsidy without ACP needed for a fiber connection to an underserved location at $8,131.

There is an interesting question as to whether Texas will prioritize connecting the unserved or the underserved to fiber. The state will likely try to optimize for both, with some percentage of both being forced to rely on fixed wireless or satellite. While that optimization will affect the math, it will not affect the conclusion that thousands of locations that would have been connected to fiber if ACP were to continue would have to be connected to an alternative.

For purposes of answering the question, let us assume that Texas will prioritize connecting all the unserved locations to fiber and use any remaining amount to connect as many underserved locations as possible to fiber. To connect all the unserved locations with fiber with ACP continuing would cost $3.487 billion.¹⁰ Subtracting that amount from the total available capital funds of $4.9 billion would leave $1.413 billion. Using that to connect unserved locations would enable Texas to connect 185,165 such locations, leaving only 13,935 underserved locations unconnected with fiber.¹¹

Now let’s compare that to the result if ACP goes away. To connect all the unserved locations to fiber would cost $3.669 billion,¹² leaving only $1.231 billion to connect underserved locations.¹³That sum would only be enough to connect 151,395 underserved locations.¹⁴

In short, ACP going away would result in at least 47,705 underserved locations not receiving fiber but being connected with another type of network, such as fixed wireless or satellite.¹⁵

The actual number of underserved locations that will end up with wireless rather than fiber is likely to be materially larger. Under the terms of the BEAD program, Texas must connect all unserved and underserved locations. So, if ACP were to continue, the 13,935 underserved locations would still have to be connected. The cost of those connections to either fixed wireless or satellite would then reduce further the number of locations who would obtain a fiber connection.

As Cartesian does not estimate the cost for a fixed wireless connection, I did not do that math here of incorporating those costs, which would be necessary to connect all the remaining underserved locations.¹⁶ But whatever the cost of connecting 13,935 underserved locations to fixed wireless (or satellite), the cost of connecting 47,705 underserved locations would be more than triple, meaning that even more underserved locations would be forced to settle for fixed wireless rather than fiber if ACP funding ends.

Further, using the 25% figure would have resulted in a significantly higher number of homes that Texas would be forced to connect with fixed wireless or satellite.

The bottom line is that while there can be some uncertainty about the number of homes that will lose out on fiber if ACP, there can be no uncertainty that tens of thousands of underserved Texas locations will be forced to connect with a wireless or satellite provider, instead of fiber, due to the end of the ACP program.

b. The Wall Street Journal, in an article titled “The $53,000 Connection: The High Cost of High-Speed Internet for Everyone,” reported that the government is financing fiber-line buildouts that are higher per location than the value of the home getting the hookup. Why would it be the case that expensive fiber buildout financed entirely by the government would need continued cash payments to run a low-maintenance fiber line?

The key points to Levin's answer are briefly summarized immediately below, with a more complete answer provided thereafter.

Congress should be interested in the question of whether certain subsidies per location are beyond an acceptable limit, something the National Broadband Plan and the recent NTIA reauthorization addressed.
In the most low-density areas of the country, there simply aren’t enough customers to generate enough revenue to pay for the ongoing operating expenses.
The shortfall in operating expenses is likely to only affect about 2% of the locations in the U.S.
That, however, has nothing to do with the ACP. ACP does not change the cost to deploy; rather it changes the incentives to deploy.

As to the logic of financing fiber deployments that are more expensive than the value of the house the deployment is connecting, I offer two thoughts. First, the 2010 National Broadband Plan recommended a limit on the per-house subsidy for deployment, a recommendation I strongly supported and which the FCC subsequently adopted. That was aggressively opposed by rural telephone companies, although in practice the rather generous limit adopted by the FCC only impacted a very small number of companies.

Moreover, as the Wall Street Journal article you cite documents, the FCC subsidy dollars per location served ($1,753) is considerably lower than other federal programs, such as the Department of Agriculture’s Reconnect program (which averages $9,014 per location.) While there are differences in the program that make a simple dollar-to-dollar comparison misleading, the differences suggest that the Senate should follow the wisdom of the recent action by the House in passing an NTIA reauthorization that included a mandate to develop a digital divide strategy. Within that mandate, the House legislation would require NTIA’s proposed strategy to support better management of Federal broadband programs to deliver on the goal of providing high-speed, affordable broadband internet access service to all individuals in the United States; and synchronize interagency coordination among covered agencies for Federal broadband programs. I support the House’s proposal and think it would be useful to address the problem you correctly raise. (I discuss the House’s proposal further in answering part c below.)

Whatever I personally think today about limits on per location support mechanisms, however, is irrelevant as the BEAD program is now the primary mechanism for connecting unserved and underserved locations. That program does not have an ex-ante limit to the amount of capex that may be provided to connect locations. The program, however, wisely incorporates the concept of the “Extremely High Cost Per Location Threshold.” As described by telecommunications expert Carol Mattey, that concept involves “a BEAD subsidy cost per location that determines when states may select a less expensive technology over fiber. Essentially, the threshold determines where awardees will be able to use fixed wireless and potentially other technologies. Setting the threshold lower reduces the opportunity for fiber applicants to automatically win over non-fiber applicants, which allows finite BEAD dollars to go further." That is, NTIA’s administration of the BEAD program is addressing the concern you appropriately raise.

Second, I don’t see what that issue of extremely high cost per location thresholds has to do with the ACP’s impact on the incentives to deploy. ACP does not change the cost to deploy; rather it changes the incentives to deploy. Reasonable minds can differ on whether there should be a limit on public subsidies to deploy where private market incentives are not sufficient and if so, what that limit, if any, should be. On the other hand, it would be axiomatic to anyone who understands capital markets that continuing the ACP would reduce the capital gap between what the private sector would be willing to pay to deploy to an unserved or underserved area and what the cost of the deployment would be, as discussed in the answer to 7a, immediately above. In other words, as the BCG study and the mathematical explanation detailed above demonstrates, whatever is the government subsidy necessary to deploy to an unserved or underserved location, the ACP reduces that amount.¹⁷

As to the specific question of whether the ACP subsidy would affect the ongoing shortfalls in the operating expenses of an ISP deploying to an unserved or underserved area, a lot depends on the specific deployment. But policy makers should keep in mind that in the most low-density areas of the country, there simply aren’t enough customers to generate enough revenue to pay for the ongoing operating expenses. Maintenance expenses in this context is something of a red herring; the network operator in the most remote areas must pay backhaul to someone else to get the traffic to the nearest internet exchange point. Those backhaul expenses are operating expenses, not capital expenses. Other operating expenses that are higher on a per customer basis include such things as customer care and repairs, as the economies of scale available in denser areas are lower in less dense areas.

As to data related to how many locations would be affected, a good source is the 2017 FCC study on the costs of deploying fiber to locations lacking reliable service using fiber or cable, which at the time were estimated to be 14% of U.S. homes and business locations. The study found that for the vast majority of these locations (12% out of the 14%) the initial subsidy should be sufficient, noting that “we do not expect these first 12% of locations will require material ongoing support once the network has been built, as subscriber revenues should be sufficient to pay for ongoing network costs.” The study continued, however, that for the last 2%—representing the costliest locations making up the difference between 98% and 100% deployment across the nation—“there is not enough addressable revenue to cover ongoing costs, so—in addition to the initial capex—an annual subsidy of about $2b would be required to keep the networks operating."

In short, while it is correct that the operating costs for fiber are lower than for copper, there still are likely to be low-density areas where operating subsidies for the rural ISPs are necessary to maintain economic viability.

c. Indeed, you are quoted in this same article as having said “The problem is money is not infinite.” Are you aware that the Biden administration has made it clear that it seeks to enroll 40% of Americans in ACP, which could cost American taxpayers $20 billion annually, just $5 billion less than the annual NASA budget?

I stand by my observation that money is not infinite.

I would also extend that observation to note that the statement is as true for the companies my equity research firm covers for institutional investors as it is for government.

As to the cost of the ACP program, I also stand by my comment at the hearing and earlier in this document that while I favor an extension, I would hope that the FCC provides data to Congress so that Congress, in considering the costs and benefits of the different Universal Service Fund programs and the trade-offs inherent in different program designs, can design a more efficient and sustainable program that moves us toward universal adoption.

As to the benefits and comparisons with NASA, I would just note that all of us can find programs and expenditures that we believe are not justified. For example, in 2023, the GAO found that the federal government lost an estimated $247 billion in payment errors during the most recently completed fiscal year (FY 2022). According to the report, “these errors include overpayments or payments that should not have been made—for example, payments to deceased individuals or those no longer eligible for government programs." That is an amount more than 10 times larger than the amount spent on NASA. I am sure everyone in the Congress could come up with their own list of problematic expenditures that exceed NASA’s budget.

But, with respect, that is not the right way to look at the issue. It is perfectly fair to ask whether our economy and society is achieving a benefit that is commensurate with the cost to the public. It is, however, confusing, rather than helpful, to compare one program to another without a reference to the benefits.

What would be helpful is for the Senate to follow the wisdom of the House, which, as noted above, recently passed an NTIA reauthorization that included a mandate to develop a digital divide strategy. As further noted above, within that mandate, the House language would instruct NTIA to support better management of Federal broadband programs to deliver on the goal of providing high-speed, affordable broadband internet access service to all individuals in the United States; and synchronize interagency coordination among covered agencies for Federal broadband programs.

This is an admirable and important goal. As the GAO found in 2023, “Federal broadband efforts are fragmented and overlapping, with more than 133 funding programs administered by 15 agencies. Among these programs, 25 have broadband as their main purpose, and 13 of those programs overlap because they can each be used for the purpose of broadband deployment. Having numerous broadband programs can be helpful to address a multifaceted issue like broadband access, but this fragmentation and overlap can lead to the risk of duplicative support. However, determining whether program overlap results in duplicative support can be challenging."¹⁸ For purposes of evaluating the ACP, we should consider how to better coordinate and administer all federal (and to extent practical, state and local) efforts to improve broadband access, adoption, and utilization, rather than comparing ACP to unrelated programs.

As NTIA hopefully conducts its analysis and as Congress responds, I would again urge Congress to adopt the wisdom of skating to where the puck is going. Getting the unconnected or the intermittently sustainably has significant value to our economy and society. Whatever value it has today is likely to be increased as Artificial Intelligence and other technology trends accelerate the movement of the delivery of goods and services to online sources. I am in favor of continually studying and evaluating both the costs and the benefits. Simple comparisons to other programs, without any reference to the costs and benefits does not advance a thoughtful public policy debate.

More in this Series

Notes

There may be areas where the revenue prospects are so low that there are no bidders willing to put in any of their own capital and BEAD does have to fund 100% of the capital costs. But those are likely to be relatively rare and are not important to the answer to this question.
This number was provided to me from Mike Conlow, whose Substack on BEAD has been a valuable source of information for those involved in the BEAD process.
I am not sure how Cartesian arrived at that number. When I ran the numbers, I assumed the $1.6 billion in the private sector contribution, if distributed equally among the 563,800 unserved and underserved locations, would be $2,837. Cartesian likely weighed the unserved and underserved in a manner that is slightly more accurate so I will use their number for the average private sector contribution.
That number is derived by subtracting the average private sector contribution ($2,903) from the average cost of the connecting an unserved location ($12,399.)
That number is derived by subtracting the average private sector contribution ($2,903) from the average cost of the connecting an undeserved location ($10,468.)
I could have also used the 25% factor but after consulting with the authors of the study, I decided to use the more conservative number.
The Cartesian study does not discuss its assumption on ACP, but I contacted the co-authors at ACA who told me that while the study didn’t directly address ACP in the cost modelling, one could argue that it assumes that ACP – or a successor program – remains in place.
This number represents the sum of $9,562 plus $500.
This number represents the sum of $7,631 plus $500.
That number is derived by multiplying the 364,700 unserved locations by $9,562.
That number is derived by dividing the $1.413 billion by the average $7,631 for each underserved location.
That amount is derived by multiplying the 364,000 unserved locations by $10,062.
That number is derived by subtracting the amount necessary to connect all the unserved homes ($3.671 billion) from the total capital Cartesian projects will be available ($4.9 billion.)
That number is derived by dividing the available $1.228 billion by the average cost of connecting an underserved location to fiber of $8,136.
That number is derived by subtracting the number of underserved homes that can be reached by fiber under the scenario without ACP from the total number of underserved homes that could be connected to fiber if ACP were to continue (184,782 minus 33,779.)
The math is complicated as comparing fixed wired deployments to fixed wireless deployments do not yield an apples-to-apples cost comparison. Further, while fixed wireless is cheaper to deploy, it’s operating costs are more expensive.
There is an interesting question, as government seeks to accomplish both deployment and adoption, as to what would be the right ratio of spending on each goal. I have not seen any studies that evaluate that question but if the government ever decides to again make new investments in deployment, I hope some evaluation of the right ratio is undertaken.
I might add that a report by the National Urban League, with which I assisted, made a similar point about the need to coordinate federal broadband programs. https://nul.org/program/lewis-latimer-plan

Blair Levin is the Policy Advisor to New Street Research and a nonresident senior fellow at Brookings Metro. Prior to joining New Street, Blair served as Chief of Staff to FCC Chairman Reed Hundt (1993-1997), directed the writing of the United States National Broadband Plan (2009-2010), and was a policy analyst for the equity research teams at Legg Mason and Stif Nicolaus. Levin is a graduate of Yale College and Yale Law School.

The Benton Institute for Broadband & Society is a non-profit organization dedicated to ensuring that all people in the U.S. have access to competitive, High-Performance Broadband regardless of where they live or who they are. We believe communication policy - rooted in the values of access, equity, and diversity - has the power to deliver new opportunities and strengthen communities.

For subscribe/unsubscribe info, please email headlinesATbentonDOTorg

Kevin Taglang
Executive Editor, Communications-related Headlines
Benton Institute
for Broadband & Society
1041 Ridge Rd, Unit 214
Wilmette, IL 60091
847-220-4531
headlines AT benton DOT org

Share this edition:

Benton Institute for Broadband & Society

Broadband Delivers Opportunities and Strengthens Communities

By Blair Levin.

benton's blog

Benton

Search