Who Really Pays What for Internet Service? The Answer Is: Who Knows?

Since Congress established the Broadband Equity, Access, and Deployment (BEAD) program, the media, government officials, and the public want to know how much funding will be available, which technologies will be deployed, and how the program will be administered. The BEAD program, part of the 2021 Infrastructure Investment and Jobs Act in 2021, will provide more than $42 billion to help states expand the availability of affordable high-speed internet.

The Pew Charitable Trusts

Often lost in the shuffle, however, are important steps Congress took to address long-standing challenges to equitable broadband access to ensure that funds received by state broadband offices are used to advance the BEAD program’s goals. Among the steps are stricter requirements for state broadband officials to follow when distributing federal funds to internet service providers (ISPs) and when collecting data for the National Telecommunications and Information Administration, including measures for tackling low- and middle-income affordability concerns.

These reporting requirements underscore the need for state broadband offices to have accurate pricing data to help inform their implementation of BEAD and other federal programs and policies.  But no definitive source for what consumers pay for broadband service exists. This has led some researchers and advocates to rely on imprecise Federal Communications Commission (FCC) sources when assessing year-over-year changes to broadband pricing in each state.

This interview with Colby Humphrey, research officer with The Pew Charitable Trusts’ broadband access initiative, has been edited for length and clarity.

Q: Let’s start with the basics. How does the broadband field measure internet prices in the U.S.?

A: There are different ways to analyze broadband costs, but the primary measurement is the monthly price charged to consumers for their service.

Q: So researchers and policymakers are looking at prices that consumers pay for their internet service. But can’t those prices vary quite a bit from household to household and geographical area to geographical area?

A: Yes. Internet plans differ based on the download and upload speeds offered, and providers can also charge additional fees beyond the monthly cost of the service. So, national and even state-level averages can often misrepresent what households or communities are actually paying for internet services. To try and get more accurate, localized information, researchers are using a range of strategies, like surveying ISPs, collecting customers’ monthly internet bills, and even analyzing online marketing offers.

Q: What causes the disparities in what consumers pay for internet service?

A: Typically, the price depends on the consumer’s plan’s minimum download and upload speeds and the type of network providing the connection—for example, fiber, cable, or satellite. The FCC has put together a Household Broadband Guide to help consumers figure out which type of service (basic, medium, or advanced) they should choose, based on their bandwidth needs. But the guide hasn’t been updated since 2022 and doesn’t reflect the most current download and upload speeds that define “high-speed fixed broadband.”

Q: So that’s one problem.

A: Yes. And another problem is that while several entities are involved in compiling pricing data, including the FCC; the Bureau of Labor Statistics; USTelecom, the trade association of internet service providers; and BroadbandNow, an independent research organization that follows the industry, there’s no national standard for pricing. And Congress hasn’t mandated that the FCC report pricing data as part of its National Broadband Map, which, in the FCC’s own words, “displays location-level information about the mass-market internet services available across the United States,” based on data supplied by internet service providers.  

Q: But the FCC does compile some pricing data through something called the Urban Rate Survey. Can you say more about that?

A: The survey, URS for short, is an annual sample of internet service providers that operate in urban census tracts that focuses on the prices they report across different speed tiers and technology types. It’s designed for a specific purpose: to ensure that customers in rural communities served by providers that receive grants or subsidies from the FCC don’t, in general, pay more than their urban counterparts. So the FCC sets benchmark prices across a range of speed and plan types based on each year’s responses to the URS. Rural providers that receive an FCC subsidy cannot charge consumers more than those benchmark prices for comparable plans.

That means that the URS can’t be used to explain what consumers actually pay for internet service in a given year because it doesn’t collect prices from rural areas nor does it include nondiscounted and nonpromotional prices from the service providers.

Q: How does the URS work?

A: The FCC collects a representative sample of internet prices, speeds, and technology types from ISPs operating in urban areas throughout the country. Participation is voluntary for providers, but for every ISP that submits pricing, the FCC collects all the speeds offered and the minimum and maximum price for each speed. And from that data, the FCC calculates an average rate. Lastly, the FCC adjusts the responses according to a set of market variables to create reasonable benchmark prices for the entire country, based on the different tiers of plan speeds and technologies. The agency has a complex methodology process to control for the survey’s sample size that is publicly available, but those are the highlights.

Q: Fair enough. But the survey has its limitations, doesn’t it?

A: Yes. First of all, as the name indicates, the Urban Rate Survey doesn’t represent the rates that rural households pay, because only providers in urban areas are selected to participate. Furthermore, the Urban Rate Survey does not represent all providers or census tracts. For context, the U.S. has more than 84,000 census tracts and over 2,200 ISPs operating nationwide. The 2024 survey collected nearly 2,000 responses from more than 200 providers across 1,800 urban census tracts. And while that’s substantial, it’s still missing pricing data from many urban areas.

Another key limitation is that it doesn’t indicate which plans are more popular among consumers. For example, if a provider reports the availability of three plans: a 25/3 Mbps (megabits per second) plan with a 1 gigabyte data cap at $50 a month, an unlimited 25/3 Mbps at $75 a month, and an unlimited 100/100 Mbps plan at $100 a month, there’s no indication in the URS data of which plan consumers are selecting. So while the URS can show trends for urban areas in terms of available plans, it can’t tell us how broadband prices impact various household budgets.

What’s more, the URS doesn’t represent promotional pricing from ISPs or capture the bundled packages available to consumers throughout the country.

Q: Does that make the survey data useless?

A: No. The data itself is useful, and researchers and advocates have proposed additional pricing indices that could be created with it. But ultimately, the data is designed for a specific purpose and can’t be used to clarify broadband adoption patterns or prices throughout the country.

Q: But it sounds like you can’t use URS data to come up with state averages of broadband prices?

A: Not really. The fact that the sample, by design, excludes rural areas is important because rural residents on average, pay more for broadband then urban residents do. So, while the FCC shows the submission it gathered from each state, you can’t extrapolate those to come up with statewide prices. There’s a similar problem with national pricing data in the Bureau of Labor Statistic’s Consumer Price Index: It reflects city averages and urban consumers, so it shouldn’t be used to assess trends in state-level pricing.

And the surveyed providers and census tracts aren’t equally distributed for each state. For example, the 2024 URS collected just three different reports on pricing data from a single provider in Montana, but collected more than 350 reports from 10 providers in Michigan.

The survey’s scattershot approach, which is intentional, may also misrepresent the rates being paid in an individual urban area. For example, certain locations in the city of Wilmington, Delaware, have four providers offering cable, fiber, or fixed-wireless service—but only one of the four was selected to participate in the 2024 URS. So the data isn’t a precise indicator of what consumers might be paying throughout a given state, particularly at the local level.

Q: You mentioned earlier that additional pricing data could be used to come up with the cost of broadband throughout the country. What data would that be?

A: There are third-party data sources that could be used. In 2021, BroadbandNow created a list of propriety plans and pricing data from 2,000 providers, but the list hasn’t been updated since then. And a 2023 report from Ready.net (a private sector manufacturer of technology that powers broadband internet) relied on a combination of pricing data from the website Broadband Search; internet scraping for marketed plans; and government sources, including ISPs that participated in the federal government’s Affordability Connectivity Program, which subsidized broadband for low-income households before running out of funds this past spring, and the National Broadband Map. Also, the Cost of Connectivity report, which was published in 2020, detailed consumer broadband prices by collecting online pricing data. But none of the available reports use a standard pricing measure in the analysis.

There’s some hope on the horizon: As part of the Broadband Consumer Labels program, which the FCC launched earlier this year, ISPs have until October 10—less than three months from now—to make their base prices, speed, added fees, data allowances, and other plan information publicly available online and in stores using the program’s “label” format, which was designed to resemble the familiar nutrition labels for food. This approach should benefit not just consumers, but also researchers trying to evaluate the true cost of broadband in each state.

Q: Should policymakers care about this? And why?

A: Policymakers need precise data to make informed decisions about policy and the design and implementation of state broadband expansion and affordability programs. Any public policy analysis that’s based on faulty information or misuse of data is distorting the facts. And if that analysis leads policymakers to propose solutions, it could result in ineffective policy innovations.

We’re not advocating that policymakers ignore the URS; we’re only cautioning against its application in assessing pricing trends throughout the states. We’re also encouraging federal and state policymakers to prioritize pricing transparency from internet service providers applying for government grant programs. That transparency would allow researchers and public officials to evaluate the impact of broadband policy on a range of issues, including pricing and affordability.