In this post, we’ll discuss the history of the BEAD program, the technical/service comparison of satellite internet and fiber broadband, the consumer affordability comparison of satellite and fiber, and BEAD program reform proposals. Read our corresponding one-pager for a summary.
I. The State of Play
BEAD’s future lies at a crossroads. After years of federal and state coordination on staffing broadband offices, mapping coverage, managing challenge processes, and evaluating subgrantees, the Broadband Equity, Access, and Deployment (BEAD) funds are finally ready to be deployed to millions of Americans. However, the Commerce Department announced plans this month to revise the rules of this $42 billion high-speed internet initiative in a way that could further delay implementation and shift funds away from fiber and towards alternative technologies. In particular, low-Earth orbit satellite services (LEOs) like Elon Musk’s Starlink – which owns most of the satellites in the U.S. – stand to win big.
This move comes as no surprise to most telecommunications policy observers. Republicans have stressed the need for greater technology neutrality in BEAD implementation across many hearings and letters. The question was never whether alternative technologies like fixed wireless and LEO would get a bigger share of the pie; but rather, how much bigger that slice would be. While the exact figure remains uncertain, the Wall Street Journal reports that Starlink could receive funds to the tune of $10-20 billion – roughly 25-50% of the total BEAD funds, far exceeding the $4 billion it was potentially eligible for under the guidance issued by the Biden Administration in January.
However, people across the political spectrum have taken issue with the proposed changes. Former BEAD Director at the National Telecommunications and Information Administration (NTIA) Evan Feinman, who has consistently defended the program’s fiber-forward nature, cautioned that the administration “should fix BEAD by removing the requirements that have nothing to do with building infrastructure, NOT change it to benefit technology that delivers lower speeds at higher costs to the households paying the bill.”
Republicans across state and federal lines have also called for a measured approach. Sen. Shelley Capito (R-WV) issued a statement asserting that while she was “all for improving the program,” she “[does] not want to see West Virginia wait longer than is necessary or have to redo their proposals and application.”
This response makes a lot of sense when you consider where we’re at in the implementation process. Four states – Louisiana, Nevada, West Virginia, and Delaware – have already completed the selection process for their projects (with most of them prioritizing fiber buildout). An additional 6 states have already closed their application windows and are selecting among potential subgrantees. For many states, pausing and re-working of the BEAD program is akin to a football team with a 10-point lead calling a timeout at the 5-yard line with seconds left in the game – it halts momentum and introduces unnecessary risk.
People recognize these pauses are not costless – pauses cost providers, consumers, and the broader economy. With inflation projected to remain higher than expected and economic growth slowing, delayed BEAD implementation means higher costs for providers and reduced purchasing power for every BEAD dollar. Internet service providers that have budgeted around certain funding timelines face sunk costs and uncertainty, with smaller providers suffering the most. More critically, American families continue to bear the costs of prolonged lack of access to critical digital resources like work, telemedicine, education, and communication tools.
But beyond the short-term economic impacts – would significantly redistributing funds towards Starlink yield better long-term outcomes for consumers? Of even greater concern than the delays is the choice of technologies receiving this once-in-a-generation investment. Weeks prior to Commerce Secretary Howard Lutnick’s announcement of a review of the BEAD program, Sen. Dan Sullivan (R-AK) expressed his concerns with the technological distribution of investments, stating that “Our constituents appreciate Starlink in many circumstances, including when there are service disruptions associated with fiber, but the connection is not consistent and reliable yet and cannot cover the state with the current coverage. Will you assure me that Commerce will not rely on Starlink in Alaska as a solution to all our problems?”
So to steal Sec. Lutnick’s framing – would these modifications result in more American households getting the benefit of the bargain? Answering that requires a deep dive into the tradeoffs. To start, Congress required states to develop plans that the NTIA approves, the states implement, and the NTIA supervises. So, a change that requires the states to totally rework their plans will inevitably result in major delays. By contrast, changes that allow states to ignore certain obligations as part of their implementation allow states to decide how to proceed, and whether or not to slow things down to take advantage of the new flexibility granted by NTIA.
When considering whether to require states to rework their plans to shift from fiber and fixed wireless to LEOs, it’s helpful to consider why the NTIA prioritized fiber buildout to begin with. This reassessment should not only consider the technical merits and drawbacks of each technology, but also their long-term impact on the communities involved.
II. How Fiber Became the Backbone of BEAD
To determine whether it was wise to prioritize connecting users via fiber – despite its higher cost and longer deployment time – over alternative technologies like LEOs, we should examine two key factors: 1) the objectives of the BEAD program and 2) the technological capabilities of both options.
BEAD was born of the COVID-19 pandemic, which shifted nearly every facet of daily life online and starkly exposed the existing digital divide. Recognizing that this divide could not be bridged overnight, Congress established targeted programs like the Emergency Broadband Benefit and Emergency Connectivity Fund as quick supports – but it knew Americans’ connectivity needs were far broader, and would persist well beyond the pandemic. As the Infrastructure Investment and Jobs Act (IIJA) states, “Access to affordable, reliable, high-speed broadband is essential to full participation in modern life in the United States.” With that animating principle in mind, Congress created the Broadband Equity, Access, and Deployment (BEAD) Program to build quality broadband infrastructure for Americans in unserved and underserved areas.
Congress was clear on what it wanted out of the networks that BEAD is funding. It required networks to offer at least 100 mbps download / 20 mbps upload speeds, no more than 100 ms of latency, and uninterrupted service (with no more than 48 hours of downtime over the course of a 365-day period). However, it left some discretion to the NTIA as to specifics, such as what is to be considered a “priority broadband project.” Specifically, Congress stated that a priority broadband project must deliver “broadband service that meets speed, latency, reliability, consistency in quality of service, and related criteria as the Assistant Secretary [of Commerce for Communications and Information] shall determine,” and ensures “that the network built by the project can easily scale speeds over time to — (I) meet the evolving connectivity needs of households and businesses; and (II) support the deployment of 5G, successor wireless technologies, and other advanced services.”
Since the IIJA does not explicitly prioritize technologies, the NTIA – the nation’s lead adviser on telecommunications issues – was left to discern which technologies are best suited to meet the IIJA’s stipulations. The NTIA sought public input, and as then-NTIA Administrator Alan Davidson weighed industry and public perspectives, as well as the Congressional imperatives, Administrator Davidson began devising the guidelines for the BEAD Notice of Funding Opportunity (NOFO). At this point in early 2022, he understood two critical things:
- Fiber broadband is considered the gold standard of broadband infrastructure, with the fastest speeds, lowest latency, and overall most reliable and resilient service of any broadband technology, and;
- Designing a fiber-exclusive broadband infrastructure program would be both fiscally and technically impossible – the cost to build out to certain high-cost locations is simply too steep, and the geographic obstacles too difficult to surmount.
Administrator Davidson acknowledged the latter issue in 2022, noting that “…we fully expect that the program will be implemented with mixes of technology. Some states will have more fiber. We expect there’ll be fixed wireless, we expect there will be low-Earth orbit satellites, and different states will make those choices.”
That said, it makes no sense to pretend that different technologies share the same capabilities or challenges. The NTIA was statutorily compelled to assess technologies for both their current and future potential, and consequently chose to adopt a strategy that emphasized fiber deployment as the primary solution, while utilizing alternative technologies like fixed wireless and LEO satellites in areas where fiber installation is economically or physically impractical. As the 2022 BEAD NOFO explained, fiber is best suited to ensure networks can scale speeds over time and support the deployment of emerging wireless technologies, “which rely extensively on fiber for essential backhaul.” Critically, fiber service also remains significantly more affordable for consumers to subscribe to than LEO service.
However, the NOFO also explicitly articulated funding opportunities for “alternative technologies” like Starlink. The original 2022 BEAD NOFO stated that while alternative technologies were considered lower priority relative to fiber networks for service, reliability, and scale reasons, they were eligible for funding when no Reliable Broadband Service (RBS) projects were feasible below the Extremely High Cost Per Location Threshold (EHCPLT) that each state was directed to establish based on its unique circumstances.
In the final days of the Biden administration, the NTIA responded to the evolving technological landscape, as well as public and industry input, with an additional missive – its 2025 Alternative Technology guidance. The NTIA’s updated guidance clarified the conditions under which alternative technologies can be employed, simplified the process for states to determine their applicability, and reduced barriers for LEO satellite providers (including by allowing states to reimburse these providers based on subscriber milestones or number of locations served). Critically, this guidance did not diminish or remove the emphasis on fiber deployment, but allowed for greater flexibility in determining where alternative technologies are appropriate.
Reasonable people may disagree on whether the additional guidance arrived too late, or even if the IIJA’s authors prioritized the right things, but the more pressing question is how to move forward. With BEAD now several years into its implementation, and given Starlink’s growth and technological advancements, should BEAD still preference fiber? Additionally, are there ways to make the program more tech neutral without forcing states’ hands? To answer this, we’ll evaluate Starlink from a technological, logistical, and affordability perspective, as well as some of the BEAD reform proposals.
III. How Starlink Stacks Up to Fiber
Service Standards
Low-Earth orbit satellites work by beaming internet down from satellites to a corresponding dish on Earth’s surface, unlike terrestrial internet, which typically relies on fiber-optic cable or transmission towers. This portability makes them a lifesaver for everything from emergency response efforts to people living in RV’s, and also for highly remote locales in the U.S. and around the globe – for instance, Starlink has been deployed in 19 out of 54 African countries.
Starlink has also grown its constellation of satellites rapidly since the company’s founding in 2019. Its total fleet now sits at roughly 7,000 functional satellites, with 1.4 million subscribers in the U.S. alone – about 1% of internet connections in the country. Additionally, with Starlink’s internet speeds ranging from 25 to 220 Mbps, relative to Hughesnet and Viasat’s 100 to 150 Mbps, Starlink’s service represents a dramatic improvement over other satellite broadband providers. These speeds and coverage capabilities stand to improve with further launches of satellites.
Congress’ decision to require networks to offer at least 100/20 Mbps speed and no more than 100 ms of latency reflected a recognition that the previous standard of 25/3 Mbps, which the Federal Communications Commission defined as “high-speed broadband” in 2015, no longer meets the demands of many bandwidth-intensive services such as telehealth, online education, teleconferencing, video streaming, artificial intelligence applications, and gaming.
When the BEAD NOFO was drafted in 2021 and 2022, Starlink’s ability to meet the 100 Mbps download and 20 Mbps upload speed benchmarks strongly varied by time, network congestion levels, and region. In the first quarter of 2021, Starlink’s median download speed in the U.S. was 65.72 Mbps, which increased to 90.55 Mbps by the first quarter of 2022. However, during the same period, median upload speeds declined from 16.29 Mbps to 9.33 Mbps, falling short of the 20 Mbps upload target. These days, Starlink’s speeds are hewing much closer to the 100/20 standard established by Congress in the IIJA; but they’re still highly variable and lagging behind other providers – as of February 2025, Ookla shows that Americans are getting on average 279.93 Mbps of download speed and over 40 Mbps of upload speed through fixed broadband connections. Meanwhile, although LEOs demonstrate much lower latency levels than traditional Geostationary Orbit (GEO) satellites, their latency remains higher on average than terrestrial fiber-optic connections, with throttled speeds common during “peak usage” hours.
Some of these issues will be ameliorated with further satellite deployments, but some of them are intrinsic to LEOs as a technology. One key physical issue is the distance signals must travel to and from space. Even though LEO satellites are closer to Earth than traditional geostationary satellites, they still face inherent latency due to the time it takes for signals to travel between the Earth and the satellite constellation. This delay is typically higher than the latency experienced with fiber-optic networks, which benefit from transiting a wired network and being able to operate in closer proximity to the consumer, resulting in lower propagation losses. While technically this should make for easier signal transmission in rural areas relative to urban ones – which sustain significant signal losses due to dense buildings and more overall architectural “clutter” – environmental features and weather conditions pose their own challenges.
Starlink has some important advantages and can work where technologies such as fiber and fixed wireless don’t. For example, loss of power to the wireless tower or fiber network operating center will knock out these connections, whereas satellite service will keep functioning (as long as you have a home generator). That’s what makes satellite services useful in disaster recovery. And with Starlink, you can take your receiver dish with you and set up elsewhere – very handy if you have to evacuate.
But these disaster-related advantages are offset by the many conditions where satellite service works poorly, or does not work at all. In particular, satellites don’t work in mountainous regions where the terrain blocks the receiver from “seeing” the satellites in space.
Environmental disruptions, such as rain, storms, heavy foliage, or thick cloud cover can make connectivity inconsistent. The signals that connect users to Starlink satellites can be blocked or degraded by such conditions, leading to slower speeds or complete outages. While satellite networks continue to improve with better technology and more satellite coverage, these environmental challenges will remain a limitation because they are a matter of physics. In contrast, fiber-optic networks, which rely on physical cables, are unaffected by these environmental factors, offering more reliable service under a wider range of conditions.
Capacity and Scalability
In addition to service quality, the IIJA statute instructs states to prioritize grantmaking for networks that “can easily scale speeds over time to meet the evolving connectivity needs of households and businesses; and support the deployment of 5G, successor wireless technologies, and other advanced services.”
Can Starlink realistically significantly scale up its speeds and provision of broadband to rural and underserved communities within the four years of receiving grant awards mandated under the NTIA’s BEAD NOFO – or even reasonably soon? We can look to Starlink’s current capacity for some clues. In regions around Seattle, Portland, San Diego, and Austin, service is currently listed as “sold out.” Beyond the physical logistics of building and launching additional satellites, the system also frequently has congestion issues. Adding millions of users over a short time frame will worsen speeds and reliability, at least until Starlink can launch enough satellites and build enough ground stations to compensate for the additional traffic.
Furthermore, while satellite technology is rapidly evolving, it is still constrained by factors that fiber-optic networks do not face. Fiber infrastructure, once installed, remains operable for decades with remarkably low maintenance and operating expenses. In contrast, Starlink and other satellite services require ongoing replenishment of satellites – with the average satellite having a lifespan of 5-7 years – incurring significant operational expenditures. This reliance on frequent satellite launches and maintenance adds complexity and cost that fiber-optic networks do not share.
Lastly, the issue of space debris poses a significant challenge for the scalability of satellite networks like Starlink. As more satellites are launched, the risk of increasing debris in space – and potentially, collisions that compromise satellite functionality and safety – grows substantially. Space debris management and collision avoidance are already complex issues for the satellite industry, and navigating these challenges becomes more difficult if Starlink attempts to scale up quickly. It will be imperative for the FCC to develop stronger domestic and international coordination of launches, debris clearance, and general space safety protocols prior to a significant acceleration of satellite deployment.
Affordability and Community Economic Impact
Starlink deserves real credit for swiftly expanding its fleet and working to deliver better service to its subscribers, but Starlink’s chief growth impediment remains its high price point. As New Street Research’s Blair Levin noted in October 2024, Starlink is unlikely to significantly grow its market share in served areas, even with higher speeds, under its current pricing plans.
The IIJA tasked the NTIA with defining “high-cost areas” as part of the BEAD allocation formula. The NTIA decided to define “high cost” using a cost model that incorporates an area’s remoteness, population density, topography, and poverty levels, and then measures costs over the life of the network. Data indicates that rural areas in the United States tend to have higher poverty rates compared to urban regions. For instance, in 2022, the poverty rate in nonmetropolitan (rural) areas was 15.5%, while metropolitan (urban) areas had a poverty rate of 12.1% – a trend that has been consistent for over 50 years. Therefore, when we consider the average BEAD beneficiary in rural or Tribal America, it is imperative that we consider whether households will actually be able to afford the networks deployed to their communities.
Starlink continues to be about twice the price of fiber and three times the price of fixed wireless broadband, with upfront costs of $349+ for hardware and monthly service fees ranging from $120-500. Even its newly announced Residential Lite plan – which is only available to select states and is designed for households with lighter data usage needs who don’t mind being deprioritized during peak hours – is priced at $80 per month.
These costs will prove prohibitive for many low-income and rural users, though they can and should be mitigated through an affordability subsidy akin to the one offered under the Affordable Connectivity Program (ACP). Regardless of how BEAD’s funding distribution shakes out, aligning infrastructure investments with affordability initiatives will be critical to stimulating demand in BEAD’s newly connected markets. As Congress considers possible reforms to the distribution and contribution sides of the Universal Service Fund (USF), Public Knowledge encourages them to integrate the lessons we learned from the ACP, which provided discounts of up to $30 per month for broadband services to eligible consumers, and up to $75 per month for those on qualifying Tribal lands. An enhanced affordability support would ensure that we don’t build “networks to nowhere” in rural America, where the difference between five subscribers per mile and seven can make or break a business.
Lastly, although local economic impacts were not explicitly addressed in the IIJA’s statute or the BEAD NOFO, fiber also enjoys the benefit of fostering local job growth in the communities where it’s deployed, whereas the economic benefits of Starlink expansion are contained to Starlink. First, there is the actual construction and maintenance of the network itself. But more importantly, once fiber (or fixed wireless) is in a community, it’s available for everyone – whereas satellites can be remotely shut off, and only work as long as you’re paying for them. Even if fiber networks need to be maintained, they are durable infrastructure under local control.That means businesses, schools, hospitals, and other anchor institutions stand to benefit from it long-term. If you want to grow an area economically, you need a broadband network that reliably supports the broader community. Whether it’s running multiple credit card readers for retail, running complicated precision farming, or even just fixing cars and tractors, you need a stable broadband connection that not only meets your needs today, but can grow as your community grows. A satellite dish is better than nothing, but it doesn’t give you the reliability or capacity of fiber – and this problem only stands to get worse as we get rid of the existing phone network that these rural areas rely on today.
IV. BEAD Changes On the Table
There are two primary buckets of modifications on the table – ones that grant states flexibility, so they can either proceed according to their original plans or incorporate some changes, and those that would compel states to return to square one. Since the latter could cost us significant time and money, it’s critical to consider how reforms would interface with existing state plans.
For instance, Rep. Richard Hudson (R-NC)’s Streamlining Program Efficiency and Expanding Deployment (SPEED) for BEAD Act would do many things – like allowing LEOs and unlicensed fixed wireless to compete for BEAD awards by redefining the term “reliable broadband service” to include any technology that meets the performance criteria – that would still allow states to to choose which of the services they’d like to invest in. Even if the NTIA removes BEAD’s fiber prioritization, if early results are any indication, many states may still opt for a primarily fiber-oriented approach.
On the other hand, whether or not it was intended, other portions of SPEED for BEAD like amending the IIJA’s affordable broadband plan requirements to prohibit the NTIA and states from regulating, setting, or otherwise mandating the rates charged for broadband service would implicate many states’ BEAD grant rules which have already won approval – potentially adding months or over a year to the implementation process.
Elections have consequences, and nobody expects the Trump administration to maintain the exact same focus and priorities in program implementation – trimming off Democratic priorities like labor or climate provisions is what you’d expect. But the overall, consistent message from Republicans, like Louisiana Governor Jeff Landry, who sent a letter to Sec. Lutnick suggesting tech neutral modifications to the BEAD program, has been to streamline and reform – not restart – the program.
If Sec. Lutnick pursues a more heavy-handed approach by setting a low nationwide cap on the per-location cost of fiber deployment, it would compel states to use the lowest-cost technology in a given area – likely Starlink. While experts believe that the cap being discussed, about $10k-$30k, would shift funds from fiber to satellite in a more marginal way, a limit of $10,000 per passing or lower would materially shift funds from fiber providers to satellite. While ostensibly tech-neutral, this position would make most fiber networks cost-prohibitive and effectively shift BEAD funds from fiber to LEO and wireless – regardless of communities’ distinct economic and connectivity needs.
The Commerce Department should forgo such a blunt approach. If policymakers want to speed up deployment, there are gains to be had by further streamlining permitting; only removing requirements that aren’t essential to the program – such as labor and climate conditions; and ensuring states have more autonomy in distributing funds. But any modifications should accelerate our progress, not force states back to the drawing board.
V. In Conclusion
Low-Earth Orbit satellite service is a truly unique and valuable tool for rural and remote areas, even if its use case is currently limited. Furthermore, its continual technological advancements make it a promising addition to the constellation (hehe) of broadband options. However, we won’t get a chance at building broadband infrastructure at this scale again. As applications such as artificial intelligence, telehealth, precision agriculture, and online education become increasingly integral to daily life, the demand for robust and resilient broadband infrastructure will only increase. Fiber networks are best equipped to meet these evolving needs – but we must allow states continued autonomy to determine the best technology mix for their constituents. Reworking BEAD to favor Starlink risks squandering this opportunity to build robust, future-proof broadband infrastructure. Congress and the administration should stay the course, support states’ autonomy over the process, and ensure rural communities are set up to truly receive the benefit of the bargain.