Ten Things About ACP that Ted Cruz Cares About: #9 Broadband Adoption Research

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.

9. You cite a study based on an input-output model by Sprintson and Oughton.

a. Are you aware that the authors relied on an Input-Output model that is unrelated to ACP, and that uses no empirical evidence from ACP?

Yes.  As the authors make clear, they are using models “to explore the potential macroeconomic impacts of broadband spending to Gross Domestic Product (GDP) and supply chain linkages.”  It is common for economic papers to use existing data to create models that estimate the likely outcome of various policies.  Some prove to be more accurate than others, but we should acknowledge that in developing public policies, much as the private sector does in evaluating large capital allocations, we don’t have the luxury of waiting for perfect data to evaluate the precise outcome.  Hindsight can be 20/20 but it is also true that waiting can prove counterproductive.

b. Are you aware that the authors relied on an Input-Output model that shows that any increased spending on the information sector would increase GDP disproportionately?

This was brought to my attention at the hearing.

c. In evaluating ACP, should the Committee rely on studies, such as Sprintson and Oughton, that offer no empirical evidence of the effects of ACP?

I think that is a fair question and I appreciated some of the comments and questions at the hearing.  I will say for myself that the questions made me more reluctant to cite the study in the future. 

But if I can add two other points, the problems with the Sprintson/Oughton study should not be used to disregard the many other studies I cited, including the Horrigan study, the telehealth studies, the education studies, the job training and placement studies, and others.  Many of them do not have specific data on ACP but that does not mean they should be disregarded in terms of projecting where the puck is going in terms of the impact of connectivity on the cost and outcomes of various healthcare, educational, job training and job placement services, among other services now online.

Second, the question raises a good point about the need to dig deeper into the studies all sides in good faith bring to the debate.

In that regard, I would note that one of the witnesses produced a study (the EPIC study) that concluded that ACP led to inflation in broadband services. In addressing your question of which studies Congress should rely on, I think the EPIC study demonstrates the very problem you raise.   While I am not an economist, my over two decades of working with institutional investors in communications causes me to immediately see several material problems with the study.

First, it is empirically wrong.  As a filing to the Committee from a conservative think tank, R Street, noted “According to recent studies, the most popular broadband speed tier plan prices dropped by 18.1 percent during ACP implementation."¹ The letter further noted that “recently-published research found that not only do Internet Service Providers (ISPs) pass cost savings on to their customers, but they also are ‘not inflating prices to appropriate government subsidies, and the ACP is successfully reducing the cost of internet plans for eligible households.’"²

The Bureau of Labor statistics offers data consistent with the R Street observation. Its Consumer Price Index for “internet services and electronic information providers in U.S. city average, all urban customers, comparing January 2021 to January 2024,” found that prices had dropped 7.8%, which compares quite favorably to another network business, electricity, for which prices rose 9%.³ The index for all prices in those three years show an increase of 17.9%.⁴

Further, even the FCC Urban Rate Survey on which the EPIC study relies demonstrates exactly the opposite of what the EPIC model would predict.  That survey concludes that prices for lower speed offers rose 5.4% between 2020 and 2023, well below the general rate of inflation in those years. So, the price differences EPIC uses to make its claims took place in speed and price tiers not likely to be relevant for ACP-eligible households.

The key error made by the EPIC study errors is that it relies on pricing data that doesn't reflect the experiences of ACP subscribers. The FCC's Urban Rate Survey tells us the prices ISPs offer in the marketplace. It tells us nothing about what we want to know about ACP subscribers: what they pay for the services they subscribe to using the ACP benefit.  In addition, the Urban Rate Survey data doesn’t include mobile, but a significant portion of ACP enrollment is mobile (as discussed earlier). So EPIC is comparing the price of fixed to trends in a mixed fixed/mobile program. This raises significant questions about whether the resulting correlations are meaningless. In other words, the data that is the foundation of the EPIC study does not have a sufficiently clear link to the pricing relevant for ACP subscribers.

Further, any analysis of price and ACP uptake needs to happen at "apples-to-apples" levels of geographic granularity. The EPIC study does not do this. It aggregates ACP enrollment to the state level (when zip code level ACP enrollment data is available) and compares it to statewide ACP enrollment and statewide pricing data. That's like saying ACP enrollment rates in the Bronx somehow have something to do with prices of service offerings in Westchester County NY.  As I believe any person working for ISPs on pricing strategy would tell you, they don’t. 

Second, the EPIC study is theoretically wrong.  Anyone who understands communications networks would incorporate into their analysis that communications networks are a shared resource for which the greater the number of customers using the network, the lower the per user cost structure for the provider.  That is, the networks involve high capital expenditures but low incremental costs, so the greater the number of customers, the lower the per user cost.  While we can argue about how many new customers ACP created or how many intermittent customers were kept on by virtue of the ACP, there should be no argument that the per user cost to the ISP would go down, not up, by virtue of the ACP.  We can also argue about the competitive intensity of the broadband services market, but we should all agree that if there is a competitive dynamic and the cost structure per user goes down, the price should also go down.  That understanding, however, is absent in the analysis of the EPIC study.

Third, it is logically wrong.  Think of it this way.  There are two kinds of broadband customers; those relying on ACP and those who don’t. 

As to the first group, whether they previously purchased broadband or not, ACP meant that as a practical matter, they were paying less for broadband.  So as to that group, ACP is clearly deflationary.

As to the second group, they would be unaffected by ACP.  That is, they are not eligible to receive it.  But as noted above, as the ACP lowers the cost structure for the ISPs, if anything, it should reduce prices.  As further noted above, it is illogical to think that ACP enrollment rates in the Bronx somehow have something to do with prices of service offerings in Westchester County NY. 

Further, to believe that ACP would cause ISPs to increase prices for those who don’t receive ACP requires the belief that the ISPs have been charging non-ACP recipients less than what would be the profit-maximizing price.  I do not mean this as any criticism of the ISPs, who are, and should be, profit maximizing enterprises.  But having spent most of the last two plus decades advising institutional investors who invest in ISPs, I believe that my clients would be, at a minimum, highly skeptical of any study dependent on an assumption that the ISPs charged less than what they could to maximize profits. 

In addition, there are also a number of technical problems with the study.  For example, the study’s timeframe doesn’t hold up, as it looks at price changes from 2021 to 2022 (It uses the 2023 FCC Urban Rate Survey which is 2022 data).  That is probably not the right timeframe for an ACP analysis.⁵

Further, it ignores the predecessor program, the EBB.⁶ In addition, the study does not control for quality changes in broadband (e.g., higher speeds). Also, the Urban Rate Survey data on which the study relies doesn’t consider promotions or bundles; it’s basically the rack rate for broadband.As anyone who knows the industry understands, a significant number of subscribers pay an amount less than the rack rate.

So, your question about how to weigh various studies is appropriate.  Again, I appreciate being directed to areas where one study I cited might have a deficiency.  I would hope, however, that in fairness and in a good faith effort to develop policy that serves the public interest, even when we lack perfect information, that all sides consider the validity of studies they have cited.

More in this Series

• Ten Things About ACP that Ted Cruz Cares About—And Ten Answers that Could Help Reshape How We Think About the Program
• #2 The Economic Benefit of ACP to the Health Care System
• #3 Net Cost Savings to Government
• #4 ACP and GDP
• #5 ACP vs Private Low-Income Plans 
• #6 ACP and Telemedicine
• #7 ACP and BEAD
• #8 ACP and Education
• #9 Broadband Adoption Research
• #10 What Companies Care About the Affordable Connectivity Program?

Notes

1. The letter cited Arthur Menko, 2023 Broadband Pricing Index Broadband Prices Continue to Decline, USTelecom https://ustelecom.org/wp-content/uploads/2023/10/USTelecom-2023-BPI-Repo... (last visited Apr. 30, 2024).
2. Here, the letter cited Schieberl, River and Ahmadi, Nikki, Measuring the Success of the Affordable Connectivity Program (July 31, 2023) https://ssrn.com/abstract=458690 or https://dx.doi.org/10.2139/ssrn.4528690
3. This does not include an adjustment for increases in speed, which would, if added, show a decrease of 80%.  But for purposes of this analysis, we are using the standard BLS CPI data.
4. The data set provides the data for 22-23 (an increase of 6.4%) and 21-22 (an increase of 7.5%).
5. The EPIC study's methodology notes are too sparse for me to be certain but based on the limited citations the study does provide, it appears the study is comparing ACP enrollment rates from 2022 to 2023 to FCC Urban Rate Survey data for 2021 to 2022 (i.e., the 2022 and 2023 Urban Rate Survey datasets). As a result, its entire analysis may attribute 2022 trends in broadband prices to 2023 trends in ACP enrollment, which would be, to put it mildly, problematic. Again, while I am not an economist, I am fairly certain that future activities cannot be seen as causing actions in the past.  Except, of course, in the Terminator movie series.
6. In this regard, there is also an interesting historical error.  While the study notes the ACP predecessor, the Emergency Broadband Benefit (EBB), it does not incorporate the subsidy, 67% larger than the ACP subsidy, into its model.  If the study’s model were correct, the EBB should have caused a greater level of inflation than the smaller ACP subsidy.  Further, and curiously, the paper says the benefits were “similar.” I have never read a study in which an economist refers to a benefit 67% greater than another as “similar.”  Again, while I am not an economist, I don’t believe the numbers, in this context, are similar.

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.

© Benton Institute for Broadband & Society 2024. Redistribution of this email publication - both internally and externally - is encouraged if it includes this copyright statement.

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:

Broadband Delivers Opportunities and Strengthens Communities