Big Tech Pays for Power—Why Not Broadband?
Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI signed the “Ratepayer Protection Pledge” at the White House (see video of ceremony) which acknowledges a simple principle: the largest users of infrastructure should help cover the costs they impose. Although the pledge is non-binding, it reflects the long-standing principle embedded across infrastructure: heavy users help finance the networks on which they depend. Airlines pay airport fees, trucks pay road tolls, merchants pay payment-processing charges, and large industrial customers pay higher electricity tariffs.
Yet broadband is the striking exception. The largest generators of internet traffic contribute little or nothing to the networks that carry their services or to the upgrades those networks require. The typical U.S. broadband household generates roughly $2,600 per year in revenue for the largest internet platforms—yet those same companies contribute nothing to last-mile infrastructure or universal service programs.
Some argue that network usage fees or Universal Service Fund (USF) contributions would amount to a “tax on the internet.” However, the core issue under discussion is not taxation but cost recovery for the delivery of large volumes of network traffic—essentially a shipping or usage fee common across infrastructure systems. These costs can be recovered through a variety of mechanisms chosen by market participants. It is also worth noting that the U.S. Supreme Court has confirmed that the USF contribution is regulatory fee administered by an federal agency (like 911, FDIC, AATF, FINRA etc), not a tax levied by Congress.
This note explores why the pledge represents an important manifestation of Big Tech agreeing to pay its own way, protecting consumers from free riding, and recovering cost for revenue for infrastructure investment. It describes how the pledge should be applied to broadband.
The White House’s Ratepayer Protection Pledge
“Ratepayer protection” has long been a foundational doctrine in U.S. public utility law. Emerging in the late nineteenth and early twentieth centuries, it reflects regulators’ duty to ensure that rates charged to captive customers remain “just and reasonable.” The doctrine prevents utilities from shifting imprudent costs, excessive returns, or cross-subsidized investments onto households and small businesses.
In electricity markets, large industrial users—including hyperscale data centers—regularly enter long-term power purchase agreements, reserve dedicated generation capacity, co-finance interconnection upgrades, and pay demand charges tied to peak electricity load. These mechanisms align costs with usage.
As hyperscale AI infrastructure expands, states have increasingly moved to ensure that grid upgrade costs are not socialized onto residential ratepayers. The emerging “Ratepayer Protection Pledge” reflects a growing political and economic consensus: households should not subsidize the infrastructure demands of the largest private users.
Importantly the pledge recognizes that consumers should not foot the bill for the benefit of private companies. Instead, the data center boom should be leveraged to address affordability and benefit all households and businesses. It calls on hyperscalers and AI companies “to build, bring, or buy all of the energy needed for building and operating data centers, paying the full cost of their energy and infrastructure, no matter what.” The signers agree to protect consumers from price hikes due to data center energy and infrastructure requirements and to pursue long term policy to lower electricity costs for consumers. This includes but is not limited to energy infrastructure investment, rate structures which protect consumers, investment in local job creation and workforce development ensuring that Americans benefit from “every step” of the “oncoming technological boom”, and resilience investment to make backup resources available to prevent blackouts and power shortages.
When Infrastructure Users Pay Their Way—Except in Broadband
Infrastructure systems are generally designed so that those who place greater demands on the system contribute proportionally to its maintenance and expansion. Whether through demand charges in electricity markets, landing fees in aviation, or transaction-based pricing in payment networks, cost recovery is typically aligned with usage intensity. Heavy users pay more because they require more capacity and often trigger the need for additional investment. This is the case around the world for airports, highways, electric grid operators, data centers, railways, and water systems. And yet broadband, the foundation of the Internet economy, is the striking exception—it is the only infrastructure in which the largest users contribute little or nothing for the network and or its upgrades. The following graphic illustrates how the largest users contribute financially to infrastructure.
Across infrastructure sectors, the pay your own way principle is widely accepted. Fees typically scale with volume, weight, capacity, or the number of transactions. Large users frequently trigger—or directly co-finance—system expansion when additional infrastructure is required. Usage-based charges are treated as a normal mechanism for allocating costs, and large-scale “free riding” is not an accepted operating norm. The expectation is simple: the biggest users pay their share.
Broadband access networks are unusual in this respect. Retail broadband pricing remains largely flat-rate, and interconnection between access networks and major platforms is often settlement-free—meaning no payment changes hands. As a result, the largest marginal generators of internet traffic frequently pay nothing at the network interface, even though they account for a substantial share of peak downstream traffic.
The contrast with other infrastructure sectors is striking. Consumers are seeing electricity prices rise in part due to the massive grid demand of hyperscale data centers. Yet broadband prices have generally fallen even as network traffic has surged. One reason is that broadband operates in a competitive retail market. Providers often absorb the costs associated with hyperscaler-driven traffic growth rather than passing them on to consumers. But this dynamic comes with consequences: reduced margins for network operators and lower incentives for investment.
The impact is particularly acute in rural areas, where margins are thin and capital for infrastructure expansion is scarce. To address the resulting gap, policymakers often layer on subsidies, appropriations, and universal service programs—social interventions intended to correct a distortion in the cost structure. If the largest traffic generators contributed proportionally to the infrastructure their services rely on, the need for such interventions would be far smaller.
The current arrangement reflects historical assumptions from the early commercial internet, when traffic flows were relatively symmetrical between networks. Those conditions no longer describe today’s broadband ecosystem. Traffic has become highly concentrated, with a small number of platforms accounting for a large share of peak network load. Yet the cost allocation model still reflects an earlier era, assuming symmetry where none exists.
Competitive Broadband Prices Defend Consumers in an Inflationary Economy
The Ratepayer Protection Pledge is a response to rising electricity bills faced by households as energy demand from large-scale computing infrastructure grows. Broadband offers an instructive contrast. Even as internet traffic has surged, broadband prices have remained flat or declined in many markets. In this sense, broadband has quietly protected household purchasing power. While the prices of many goods and services have increased, affordable connectivity allows people to work, learn, receive healthcare, and access services from home—activities that are often significantly more costly when performed offline.
Stable broadband pricing has therefore played an important, if underappreciated, role in maintaining the affordability of modern digital life. For example, the price of a typical US wireless plan is down 8 percent down from 2023. This translates to roughly $6 per month in savings for a single line — about $70 to $80 per year — and for a family of four, approximately $20 to $25 per month, or $250 to $300 annually.
Over the past two decades, broadband has undergone a remarkable transformation driven by technological innovation and competitive market dynamics. Twenty years ago, a typical broadband connection delivered roughly 3 Mbps downstream. Today, speeds are often hundreds of times faster, with some connections reaching levels unimaginable in the early 2000s. Consumers are not just paying less—they are receiving dramatically more in both speed and usage. In fact, today’s broadband connections are roughly 500 times faster than in 2005. At that time, mobile data usage was barely measurable; today, U.S. wireless networks deliver roughly 130 trillion megabytes of data annually according to CTIA, representing exponential growth. Measured per megabyte, prices have fallen sharply: data costs declined 62 percent between 2020 and 2024, while unlimited wireless plans are about 40 percent cheaper than in 2010.
Broadband investment, a staggering $90 billion annually in the USA, accelerates this dynamic: DOCSIS 4.0 increases cable capacity, 5G and fixed wireless access extend high-speed connectivity through the air, low-earth-orbit satellite expands coverage, and fiber deployments continue nationwide.
Three-quarters of Americans today can choose among at least three broadband technologies — cable, fiber, fixed wireless, or satellite — with national networks, regional carriers, and resellers competing aggressively for customers. are doing more with less — squeezing greater capacity out of existing infrastructure while racing to deploy next-generation networks. This competitive pressure has disciplined prices and encouraged rational consolidation to achieve economies of scale necessary for capital-intensive upgrades.
Streaming Windfalls, Broadband Burdens
At the same time that broadband connectivity helps households save money across the economy, it has also enabled the rapid proliferation of streaming platforms that now generate enormous volumes of downstream traffic. Consumers increasingly must subscribe to multiple services to access the content they want, while the largest sports leagues and hyperscale platforms that drive peak traffic volumes typically pay nothing at the network interface, despite contributing significantly to network congestion.
Senator Mike Lee (R-UT) has has drawn attention to the changing economics of sports broadcasting, urging federal antitrust authorities to examine whether the Sports Broadcasting Act of 1961 still serves its original purpose in the streaming era. The law granted professional leagues a limited antitrust exemption to pool broadcast rights and ensure games remained widely available on local television. Today, however, those rights are increasingly packaged into multibillion-dollar deals with streaming platforms, fragmenting access for fans and weakening the local broadcasting ecosystem the law was meant to protect. The consequences ripple beyond media markets: escalating sports rights inflate subscription costs for consumers, erode the economics of local news tied to broadcast advertising, and drive massive volumes of streaming traffic across broadband networks—without any mechanism for those networks to recover the associated infrastructure costs.
Advertising technology adds another layer to the imbalance. Systems of ads and ad-tech—often delivering video and other content users did not actively request—account for roughly one-fifth or more of downstream internet traffic. Yet consumers must pay uniformly for all data delivered over broadband networks, rather than benefiting from pricing arrangements that might otherwise reduce their cost burden, as occurs in many other networked industries.
The result is difficult to reconcile with a consumer-focused broadband policy. The issue is not advertising itself, but that broadband providers have been systematically discouraged from adopting cost-recovery mechanisms that reflect the traffic burdens created by these services. The problem is likely to intensify as hyperscale AI data centers expand the automated generation and distribution of video, advertising, and promotional content—often initiated algorithmically without consumers actively searching for or requesting it.
To top it off, consumers also face a growing surcharge tied to legacy telephone services. The typical U.S. broadband household contributes roughly $9 per month in Universal Service Fund (USF) charges, which support programs designed to ensure affordable connectivity for rural communities, low-income households, schools, libraries, and healthcare providers. Yet the contribution base for USF remains tied largely to traditional telecommunications revenues rather than the modern digital ecosystem that now drives network demand. At the same time, that same household generates an estimated $2,600 per year in revenue for the largest internet platforms, whose services account for a substantial share of network traffic but which contribute nothing directly to universal service programs or to the last-mile infrastructure investments their traffic growth necessitates. The result is a structural mismatch: consumers shoulder the costs of maintaining and expanding broadband networks while the largest traffic generators benefit from those networks without contributing to the affordability mechanisms designed to sustain them. Broadband remains one of the most competitive and innovative sectors of the digital economy—but its cost allocation model remains uniquely misaligned.
A clear illustration of this imbalance can be seen in the Caribbean, a region of roughly two dozen small economies where broadband investment challenges are particularly visible. Strand Consult’s research describes the region as a microcosm of Big Tech’s “digital colonialism,” where global platforms extract significant revenue from local users while contributing little to the networks that make those services possible. Caribbean telecom operators pay taxes, regulatory fees, employ local workers, and reinvest roughly 18 percent of revenue into infrastructure, while the largest internet platforms generate billions in revenue from the region without comparable contributions to network investment. The result is a widening investment gap—estimated at $8–13 billion—that threatens the ability of these small economies to deploy next-generation networks and close the digital divide. The Caribbean case illustrates how the broadband cost-recovery imbalance is magnified in smaller and emerging markets where infrastructure costs cannot easily be spread across large populations.
The Policy Choice
Congress is already engaged in efforts to reform the Universal Service Fund (USF). The U.S. Supreme Court upheld the constitutionality of the program, confirming that USF contributions constitute a regulatory fee administered by an agency rather than a tax levied by Congress. This ruling provides important legal clarity and reinforces Congress’s authority to modernize the program’s funding structure.
Against this backdrop, policymakers face two primary pathways. The first option is to modernize the contribution base of the fund. A clear statutory framework could reduce the regressive burden currently placed on consumers while preserving the universal service system. Establishing a predictable contribution mechanism would align broadband’s funding model with longstanding infrastructure principles: the largest users contribute proportionally to the systems they rely upon.
The second option is to allow market renegotiation to evolve organically. Private ordering could move interconnection toward traffic-metered or capacity-priced regimes that internalize network costs. Such an approach could restore economic symmetry.
South Korea—widely recognized as one of the world’s leading broadband economies in part due to its policy innovation—has long incorporated network usage fees into its broadband framework. For two decades, Korean policy has recognized that large traffic generators should participate in the cost of delivering their services over broadband networks. Major domestic platforms have long engaged in such arrangements, and negotiations with the largest global platforms have increasingly followed the same logic. These policies emerged as video streaming and platform services began generating large shares of network traffic and imposing growing infrastructure costs on operators. Big Tech companies, which now account for most Korean traffic, have mounted sustained opposition to these arrangements, even as they rely on the very broadband networks that enable their services and revenues. The Korean experience illustrates that market-based cost recovery between content providers and broadband networks can coexist with world-leading connectivity and a thriving digital content economy. Such arrangements have supported sustained broadband investment, allowing Korea to develop world-leading networks, maintain competitive prices, and emerge as a global powerhouse in digital culture exports.
The first pathway offers stability, transparency, and Congressional oversight. The second relies on private negotiation and may produce greater uncertainty, at least initially. Yet both approaches reflect the same underlying economic principle: infrastructure sustainability requires proportional contributions from the largest users. This principle is not new. It is already reshaping energy policy.
The Ratepayer Protection Precedent and Justice for Broadband Consumers
Broadband now faces a structural question similar to the one emerging in energy policy. As hyperscale computing expands, policymakers have increasingly recognized that electricity ratepayers should not be required to subsidize the infrastructure demands of large data centers. The same logic applies to broadband. If the principle of ratepayer protection guides decisions in electricity markets, it should also inform how policymakers think about the funding of broadband infrastructure.
Across infrastructure sectors, a simple principle governs cost allocation: heavy users pay more. Fees typically scale with volume, weight, capacity, or the number of transactions generated. Large users often trigger infrastructure upgrades and frequently participate in financing the additional capacity they require. In this framework, usage-based charges are treated as standard cost allocation mechanisms, not punitive measures.
Broadband interconnection remains the exception. If statutory reform does not correct this imbalance, economic pressures may eventually push the system toward mechanisms that internalize costs more directly—potentially through traffic-metered or capacity-priced interconnection regimes.
Broadband underpins America’s $5 trillion digital economy, yet it operates under a funding structure that differs markedly from other essential infrastructure sectors. In most systems—electric grids, highways, rail networks, and water systems—users whose activities place the greatest demands on infrastructure contribute proportionally to its maintenance and expansion. Broadband remains unusual in that the largest marginal generators of network traffic contribute little or nothing at the point where those costs are incurred.
This imbalance does not reflect the absence of competition or innovation in broadband markets. On the contrary, providers have delivered dramatic improvements in speed, capacity, and affordability while investing tens of billions of dollars annually in network upgrades. Yet the underlying cost allocation framework has not evolved alongside the modern digital ecosystem. As a result, infrastructure funding gaps are increasingly addressed through rulemaking workarounds rather than through mechanisms that align costs with usage.
Ensuring the long-term sustainability of broadband infrastructure therefore requires policy clarity. Congress has an opportunity to modernize the contribution framework in a way that preserves the universal service system while aligning broadband financing with the principles that govern other infrastructure networks. Doing so would not represent a departure from established policy norms; rather, it would bring broadband into alignment with the longstanding expectation that those who rely most heavily on shared infrastructure contribute proportionally to its upkeep and expansion.
As debates around ratepayer protection continue in the energy sector, broadband policy faces a parallel challenge. Digital infrastructure delivering the digital economy and exporting the AI stack requires a funding model that reflects the realities of modern network use and ensures that the costs of maintaining the system are shared proportionately among those who benefit from it.
Learn more from Strand Consult’s series of Universal Service Fund Reform and Broadband Cost Recovery.
