Public Networks, Private Costs: Why Government-Owned Broadband Fails Taxpayers

Download the PDF of “Public Networks, Private Costs: Why Government-Owned Broadband Fails Taxpayers,” here.

Introduction
GONs Are Solving the Wrong Problem
Why GONs Fail
- Bad Incentives:
Oversaturated Markets:
- Governments are Not Built to Provide Internet Access:
GONs Undermine the Competition They Claim to Create
The Florida Model
Policy Recommendations
Conclusion

Introduction

Broadband in America has never been faster, more available, or more affordable. As of December 2024, approximately 95 percent of serviceable broadband locations in the United States had at least one provider¹ offering speeds that satisfy the Federal Communications Commission’s broadband benchmark.² Between 2015 and 2025, inflation-adjusted broadband prices fell 63.4 percent even as average download speeds doubled.³ For the few locations that still lack access to reliable internet at modern speeds, the federal government has committed $42.5 billion through the federal Broadband Equity, Access, and Deployment (BEAD) program to close the remaining gaps in the coming years.⁴

Against that backdrop of robust connectivity and competition, local governments are still choosing to become internet service providers (ISPs). Government-owned networks (“GONs” or “municipal networks”) are broadband systems built, financed, and operated by municipal governments, with taxpayers and utility ratepayers bearing the financial risk. For communities that still lack sufficient broadband infrastructure, the urge to build a GON can be incredibly tempting. In sparsely populated or hard-to-reach areas, the economics of fiber deployment make private deployments cost-prohibitive, leaving residents with limited or outdated connectivity options. Even in places already served by private networks, city leaders frustrated with incumbent providers see a GON as a way to force incumbents to compete on price or service quality.

Today, approximately 400 GONs operate across the United States, with active networks in Florida, including in Ocala, Gainesville, and Fort Pierce.⁵ However, their long-term viability is far from certain. Independent financial analyses routinely find that GONs overestimate subscribers, underestimate costs, and take years longer than projected to reach profitability – if they become profitable at all.⁶ Cities have repeatedly sold networks to private companies at a fraction of their construction cost after failing to reach profitability or subsidies ran out.⁷ Several have shut down in the past few years alone.⁸ Nevertheless, new GON projects continue to launch and existing networks pursue expansion.

Before a city commits taxpayer dollars to a new fiber buildout, its leaders should understand why these networks fail so consistently, what those failures cost the communities that build them, and what alternatives already exist for closing the connectivity gap.

GONs Are Solving the Wrong Problem

The central premise of every GON is that communities need new networks to close the digital divide. However, the remaining divide is not caused by a lack of broadband deployment, but a lack of adoption. According to the National Telecommunications and Information Administration (NTIA), 71.1 percent of Americans remain offline primarily because they either don’t want internet access,⁹ or they cannot afford it.¹⁰ Building a new network does not give uninterested households a reason to subscribe, nor does it make service affordable for those who cannot pay.

During the COVID-19 pandemic, $14.2 billion was allocated to create the Affordable Connectivity Program (ACP). The ACP provided $30 per month to low-income households to purchase broadband service alongside a $100 discount for an internet-connected device.¹¹ With the program’s subsidy, 3.8 million new households connected to the internet and broadband adoption rates lifted by 6.9% in rural counties without building any new infrastructure.¹² However, when the program expired in June 2024, over 5 million households dropped their service because they could no longer afford it.¹³ The life and death of the ACP program shows that the primary barrier to connectivity is affordability, not availability.

GON proponents, stuck in deployment mode, argue that a municipal network can improve adoption by offering cheaper service than private providers.¹⁴ First, this claim is false, as GONs are, at best, found to have comparable prices to private ISPs.¹⁵ More often, the effective price paid by the consumer for municipal connectivity is actually higher when accounting for the taxes paid, higher electricity rates, and degraded municipal services elsewhere.¹⁶ Second, even if GONs lowered connectivity prices, a modest discount does nothing for households that cannot afford broadband at any price point.¹⁷ The broadband gap is overwhelmingly a demand-side problem, and GONs are a supply-side solution. Targeted subsidies that reduce the cost of existing service would be far more effective and less costly than building a new municipal network to deliver it.¹⁸

Why GONs Fail

GONs collapse for structural reasons that persist regardless of the region, the political party in charge, or the technology deployed. A private broadband provider operates inside a chain of market tests, where the cost of capital screens out unviable buildouts, the prospect of profit determines which markets are worth entering, and the threat of insolvency forces exit when a network cannot pay for itself. Government networks face none of those tests in their full force. State and federal grants underwrite the buildout regardless of viability, causing cities to routinely enter markets that cannot support another competitor, and the political cost of admitting failure keeps networks operating long after their projections have collapsed. The cost of each distorted decision falls on local taxpayers and ratepayers.

Bad Incentives:

To assist with the high upfront capital investment required to build a new network, municipalities almost always leverage taxpayer funds.¹⁹ Billions of dollars made available by the United States Department of Agriculture’s ReConnect program and, most recently, the BEAD program artificially lower the initial cost of GON development, especially since the money rarely needs to be repaid.²⁰

Online: https://www.usda.gov/reconnect.

Covering a large portion of the upfront costs by the federal government shields municipalities from the financial consequences if the project proves to be unviable, leading them to build when they otherwise shouldn’t. These costs are a useful check on whether a network should be built at all. When municipalities don’t have to face those upfront costs, they are more vulnerable to being enticed into a hyper-competitive, dynamic market without the requisite market discipline, operational expertise, or ongoing capital reserves required for long-term sustainability. For example, in 2010, the North Florida Broadband Authority (NFBA) was awarded a $30.1 million federal grant to build an open-access middle-mile fiber network, covering 76 percent of the project’s $39.3 million cost.²¹ By 2013, the physical infrastructure was completed, but the NFBA was facing administrative and operating expenses exceeding $250,000 per month while only generating $11,000 in monthly revenue.²² When the project was sold, the private operator concluded that the network possessed no conceivable path to profitability.²³

Cities considering a GON often point to Chattanooga, Tennessee as proof that municipal networks can work. In doing so, they point to a subsidy stack that no other municipality can replicate. The Electric Power Board of Chattanooga (EPB) launched its $390 million municipal network with a $111.5 million grant from the Department of Energy,²⁴ a $50 million loan from EPB’s electric division, and $229 million in tax-exempt local revenue bonds. The massive cash injections that enabled Chattanooga to finance its network buildout were an anomaly. The DOE grant was a onetime stimulus disbursement in response to the Great Recession and historically low interest rates allowed EPB to secure favorable terms on its $229 million initial bond offering and refinance its debt. Moreover, only 15 percent of American electricity customers are served by a government-owned utility, leaving most cities with no municipal electric provider to draw from.²⁵

Once external funding covers the network’s upfront costs, the entry decision turns on whether a grant is available rather than on whether the network can sustain itself. While state and federal grants can subsidize the cost of building a new network, they do not account for the ongoing costs of operating it. When these networks fail to achieve their projected market penetration rates or adequately forecast their perpetual operational expenditures, the shortfall is absorbed by the local taxpayer, not the state or federal government that subsidized their entry.

Oversaturated Markets:

Government-owned networks were built to connect geographically isolated or sparsely populated areas that private providers could not afford to reach. However, the modern municipal broadband movement has largely abandoned these unserved frontiers, choosing instead to launch highly speculative, capital-intensive infrastructure projects in densely populated markets that already boast robust private competition. Since there are only so many customers in a given market, overbuilding disrupts the local broadband ecosystem and erodes the long-term viability of all networks operating in the area.

Broadband providers need to garner a sizable market share to recoup the cost of building and maintaining the network. In more urban areas, a firm can become profitable with a take rate of 20 to 30 percent, but may require a take rate around 80 percent for more rural areas since the fixed cost of infrastructure is relatively similar, but there are fewer customers to serve.²⁶ A new provider will enter a market if they can do so profitably. As Beard et al. explain, “If market conditions only permit two firms to operate profitably, then three firms cannot do so, and no amount of wishful thinking will change that fact.”²⁷

Still, municipal networks, insulated from costs by grants and subsidies, are led to believe that adding a new competitor will be universally beneficial to the marketplace. A cottage industry of private consultants, shielded from the financial risk of network failures, are a major reason municipalities are convinced to build new networks when they otherwise shouldn’t. Since these third-party contractors are hired to recommend building a network and can then be hired again to build it, they are structurally biased toward recommending infrastructure deployment.²⁸ It is therefore unsurprising that Magellan Advisors, a prominent consulting firm, discloses that 75 to 90 percent of its studies recommend that the city proceed.²⁹

Consultants routinely overestimate future subscription rates by relying on push polls to justify entry into an oversaturated broadband market. In one case, the consultant firm Uptown Services reported that while some feasibility metrics were based on research, others they described as “Creative Writing 101.”³⁰ After Lakeland, Florida’s CFO concluded that Magellan’s $220 million feasibility study was ‘highly speculative,’ the city scrapped the project.³¹ Unfortunately, most cities do not catch the problem in time. An analysis of recent GON projects by the Information Technology and Innovation Foundation (ITIF) found that 18 of the 20 projects were launched in markets that were already being served by comparable private options.³² A University of Pennsylvania study corroborated ITIF’s findings, noting that all 15 municipal fiber projects it examined were overbuilds of already-served communities.³³ Accordingly, a GON will fail to attract enough customers and become unsustainable, requiring new cash injections that are borne by local taxpayers.

Governments are Not Built to Provide Internet Access:

Local governments routinely provide water and electricity to their communities, which is often used to justify their operational capacity to provide internet service as well. For instance, the IT director of Williston, Florida, argued “Williston already manages electric, water, gas, and sewer — so why not internet as well?”³⁴ Like broadband, these services require a large upfront investment repaid over time, aided in part by the low marginal cost of serving each additional customer. However, unlike water and electricity, the broadband market is incredibly competitive and the technology evolves quickly, requiring frequent upgrades.

Successfully operating in this environment requires sustained capital investment, competitive pricing to attract and retain subscribers, and sufficient scale to service the debt that financed the original buildout.³⁵ These pressures apply equally to all internet providers, public or private. But, since governments lack the adequate signals, expertise, and incentives that private ISPs possess, an unfortunate pattern of failure persists.³⁶

When a private ISP overbuilds a market, shareholders absorb the loss. When a GON overbuilds, the costs are shifted to taxpayers and ratepayers through new bonds, general fund transfers, or higher utility bills. In fact, by their very nature, government networks are not allowed to fail, so there is no counterbalancing mechanism to avoid sunk cost investment into a failing project.³⁷ When the GON in Burlington, Vermont could not cover its costs, the mayor used the city’s pooled cash management system to divert $16.9 million from other city departments, (water, schools, general operations) without council authorization. The network was eventually sold in 2019 for $30.8 million, but the city recovered only approximately $6.4 million of the diverted taxpayer funds.³⁸

A municipal network that cannot cover its costs will always face pressure to draw on other city funds, because the political cost of admitting failure exceeds the financial cost of quietly subsidizing the loss. Similar failures have occurred over the years in Florida, including collapsing GONs in Orlando,³⁹ Quincy,⁴⁰ and Dunnellon.⁴¹

GONs Undermine the Competition They Claim to Create

GON proponents argue that municipalities should create new competitors to encourage incumbent providers to improve their service quality or lower their prices. However, the same structural advantages that allow GONs to defy financial failure predictably undermine those goals. Because GONs enter a marketplace with extensive subsidies, they reduce private investment in the surrounding market, shift consumer costs onto electricity bills and taxes, and can paradoxically reduce competition overall.⁴²

Private broadband providers invest in markets where they can recover the fixed cost of building a network and earn a return on capital. Entry by an additional firm changes that investment strategy, especially if the competing firm is heavily subsidized. GONs can absorb losses by borrowing at tax exempt rates or cross-subsidizing from other government revenues for as long as the political will to do so persists. Private providers do not have that luxury. Faced with an entrant that can sustain below-cost pricing indefinitely, private firms rationally respond by reducing investment, delaying upgrades, or exiting the market entirely.⁴³

A 2026 analysis of France’s national broadband subsidy program found that nearly half of the public funds disbursed went to areas where private operators would have built within three years anyway — displacing investment that would have happened without taxpayer support.⁴⁴ The mere threat of public entry is enough to reduce private investment in service upgrades.⁴⁵ In Colorado municipalities that chose to allow municipal networks, internet providers offer speeds that are 11 to 26 percent slower than municipalities that maintained restrictions, likely because they understood that they could not recover costs with a GON competitor.⁴⁶

Even if a GON delivers lower prices, the consumer benefit is nominal compared to the higher costs paid elsewhere.⁴⁷ The subsidies that allow below cost pricing on their broadband bill have to be paid back somewhere, typically through higher electricity bills, property taxes, or general fund transfers. For instance, an examination of four Tennessee cities found that GONs funded through electric utility debt raised residential and commercial electricity rates by approximately 5.4 percent — close to $12 per month for an average customer.⁴⁸

GONs subsidies let it undercut private competitors, discouraging them from investing in their own networks and shifting losses onto electric customers or tax payers who never signed up for the service. When the GON later fails, as most do, the community is left with fewer private providers than before and no net gain in competition.

The Florida Model

States have taken different approaches to mitigate the risks municipal networks pose to taxpayers. In 1995, Texas became the first state to enact restrictions on municipal networks,⁴⁹ shortly followed by Missouri in 1997.⁵⁰Over the next decade, around a dozen states enacted proposals ranging from banning municipal networks to guardrails on how GONs can and can’t operate. In 2005, Florida joined the movement by enacting the state’s own restrictions on municipal GON development as part of a comprehensive telecom reform package that year.

Prior to 2005, a series of GON failures in Orlando⁵¹ and Quincy⁵² Florida drove the legislature to create a framework to protect taxpayers going forward. Instead of a stringent ban on building new municipal networks, Florida Statute 350.81 imposes core requirements on any municipality that wants to provide communications services.⁵³ The statute requires any municipality considering broadband service to conduct two public hearings, produce a written business plan showing the network will be revenue positive within four years, use separate enterprise-fund accounting, refrain from subsidizing service with revenue from other utilities, pay the same ad valorem taxes as private providers, and obtain voter approval before issuing revenue bonds with maturities longer than 15 years.⁵⁴ If the network cannot cover its operational costs and financial obligations within four years, the municipality must hold a public hearing to consider whether to continue operations, sell the network, partner with a private entity, or close down the network entirely.⁵⁵

This framework, along with other deregulatory actions,⁵⁶ is a major reason that Florida has a robust competitive broadband market.⁵⁷ By holding municipal entrants to the same financial discipline that private firms face, the law has deterred speculative buildouts that would otherwise crowd out private investment. As of December 31, 2024, the FCC estimates that 97% of Florida locations have access to 100/20 Mbps fixed broadband, two points higher than the national average.⁵⁸ Independent third-party rankings place Florida ninth in the nation for broadband availability⁵⁹ and second in the nation for internet speeds.⁶⁰ These restrictions signal to private capital that the state would not allow municipal governments to undercut private providers, incentivizing investment.

The framework did, however, exempt municipalities that were already offering service as of April 1, 2005 from the procedural requirements, such as convening public hearings or equal taxation. Networks in Ocala, Gainesville, and Fort Pierce predate the 2005 law and remain exempt from the law’s full force. Each network has stayed in business under the carveout, but has recently sought expansion that threatens to put taxpayers at risk. For example, Ocala Fiber Network now serves just 10 percent of the city after 30 years of operation.⁶¹ Yet now, despite a private competitor’s plan to invest $100 million to build new connectivity⁶² and the area already boasting at least 13 internet providers,⁶³ the city is considering a $90 million expansion that taxpayers will be on the hook for.⁶⁴In 2022, Gainesville’s GON attempted a $9.2 million proposal to expand service to residential customers, though the city commission ultimately rejected the project.⁶⁵ In Fort Pierce, FPUAnet appears to be operating with consistent losses that are cross-subsidized by other government services⁶⁶ although 11 other broadband providers operate in the area.⁶⁷

Florida’s experience offers other states a demonstrable alternative to either unfettered municipal broadband or outright prohibition. The state allows cities to build networks, but only if they can meet the same financial test a private firm would face. Two decades later, Florida’s broadband market is among the most competitive in the country.

Policy Recommendations

Florida’s framework reduces the financial risk to taxpayers, but state-level guardrails alone cannot resolve the conditions that draw municipalities into broadband construction in the first place. Before considering government competitors, policymakers should first make private entry and competition more robust. That means removing the local barriers that drive deployment costs and letting federal broadband programs finish closing the remaining gaps. Where municipal networks are still built, they should face the same financial and regulatory standards as their private competitors. Municipalities should be required to exhaust the alternatives below before committing taxpayer dollars to building their own network.

Remove Deployment Barriers

The most effective way to expand private broadband deployment is to fix the barriers of deployment that drive higher costs. A broadband provider that has the capital and the customers can still wait months for local approval to access poles, dig trenches, or cross rights-of-way. Permitting alone accounts for roughly 10 percent of total fiber project costs, and a majority of providers surveyed by the Fiber Broadband Association expected deployment delays in 2026.⁶⁸

States should set and enforce clear timelines for permitting decisions. Several states, including Florida, have adopted “shot clock” statutes that require local governments to act on broadband permit applications within a defined window — typically 10 to 60 days — with the application automatically approved if the deadline passes without action. Instead of a provider waiting indefinitely for a decision, the burden falls on the local government to respond promptly or lose the ability to block the project.

Let Federal Investment Finish the Job

The largest broadband infrastructure investment in American history is directing $42.5 billion toward unserved and underserved areas through competitive grants. As those deployments come online over the next several years, the remaining locations without broadband access will shrink dramatically.

A municipal network that breaks ground today will take five to seven years to complete construction. The market it enters will look nothing like the market that initially justified its expansion. With federally funded and privately deployed alternatives already underway, taxpayers should not bear the risk of new municipal networks before the BEAD program is finalized.

Level the Playing Field

Where government-owned networks do operate, they should face the same financial reporting, tax obligations, and regulatory compliance as their private competitors. Under current law, private ISPs that receive federal broadband grants owe corporate income tax on those funds — effectively losing 21 cents of every grant dollar. Municipal recipients are tax-exempt and keep the full amount. The Broadband Grant Tax Treatment Act, introduced with bipartisan support in the U.S. Senate, would eliminate this asymmetry by excluding qualified broadband grants from taxable income.⁶⁹ States should also maintain guardrails like Florida’s business plan requirements, separate accounting, profitability tests, and cross-subsidization prohibitions.

Require Ongoing Re-authorization for Grandfathered Networks

If a state already has guardrails on GON development, they should extend those guardrails to existing networks that were previously grandfathered in. To ensure continued transparency and long-term solvency, networks that predate state restrictions should be required to file a public business plan and profitability review on a fixed cycle, measured against the same revenue-positive standard that new networks face. Networks that cannot meet re-authorization requirements should face the same options as unviable new networks: seek community approval to continue, partner with an existing private network, or cease operations.

Conclusion

Subsidize entry, remove the cost of failure, and insulate decision-makers from the consequences, and the result is networks built in markets that cannot support them, run by governments with no mechanism or incentive to stop. The remaining connectivity gap in America is real, but it is a problem of affordability and adoption, not deployment. GONs are an expensive solution to a problem that no longer exists.

Lawmakers in Tallahassee recognized this many years ago, and their efforts should serve as a model for other states seeking to protect taxpayers from municipal networks that are, in the long term, unsustainable. Rather than outright prohibiting GONs, force them to survive on the same terms as private providers. Two decades later, Florida has one of the most competitive broadband markets in the country, with private investment filling gaps that a municipal competitor would have crowded out.

Other states should follow that model. Remove the barriers that keep private capital from reaching underserved areas. Let federal broadband programs finish the job they were designed to do. Where municipal networks already exist, hold them to the same standards every private ISP must meet. And close the grandfathering loopholes that allow networks operating outside those standards to expand at taxpayer expense. In doing so, legislatures across the country will be able to provide better broadband policy and help the very people GON advocates claim to be helping.