The federal government can count broadband connections. It has a harder time counting whether people can actually afford to keep them. In 2024, the U.S. Census Bureau estimated that 123.9 million of the nation’s 132.7 million households had an internet subscription, while roughly 6.0 million households reported no internet access at home and another 2.8 million had access without a paid subscription. Those figures show how close the country is to treating connectivity as ordinary infrastructure — and how many households still sit outside that assumption. Census data makes the gap visible, but it does not explain why some households remain disconnected even when service exists nearby.
The first problem is that availability and access are not the same thing. A provider can technically serve a neighborhood while the monthly bill remains too high for some households, especially when internet service competes with rent, groceries, transportation and child care. That tension became harder to ignore after the federal Affordable Connectivity Program ended in June 2024. The FCC said more than 23 million households were enrolled when the program stopped accepting new participants, and the program had offered eligible households discounts of up to $30 a month, or up to $75 on qualifying Tribal lands. The FCC’s final ACP notice made clear that infrastructure alone could not resolve the affordability side of the digital divide.
That gap matters because the internet has moved from a convenience to the operating layer beneath ordinary life. Employers expect applicants to upload documents, monitor email and complete multi-step applications online. Schools route assignments, grades and parent communication through digital platforms. Health systems increasingly rely on patient portals. Government agencies move forms, benefit applications and appointment scheduling onto websites. A household without reliable home access is therefore not simply missing a communications service. It is spending more time and effort completing tasks that institutions increasingly design around uninterrupted connectivity.
The burden also changes by geography. Rural households can face fewer providers and weaker infrastructure. Urban households may live in buildings where service is technically available but prices remain unaffordable or where building-level wiring constrains provider choice. Lower-income households can be pushed toward mobile-only access, which may be adequate for messaging and basic browsing but far less useful for a long job application, school assignment or telehealth appointment. The Census Bureau’s 2024 data counted about 101.6 million households with cable, fiber or DSL service, meaning millions of subscribing households relied on other forms of connectivity such as cellular or satellite service. That is another reason a simple count of connected households can overstate the consistency of the experience.
Federal broadband policy has spent years emphasizing construction, including the Broadband Equity, Access, and Deployment program created to expand high-speed infrastructure in underserved areas. That investment remains essential, but the end of ACP demonstrated the limits of a strategy that treats the network as the final outcome. A fiber line can reach a block without making the service affordable. A household can subscribe without owning a suitable computer. A student can have internet access that becomes unreliable when several people use the connection at once. Each of those situations produces a different kind of access gap, even though a coverage map may classify the household as served.
For SSC’s broader Geography of Opportunity framework, broadband offers one of the clearest examples of how place and household resources interact. The question is no longer only whether infrastructure reaches a community. It is whether the people living there can convert that infrastructure into usable access to jobs, education, health care and public services. That requires policymakers to track affordability, reliability, device access and sustained subscription — not simply construction milestones. It also requires providers and public agencies to recognize that connectivity failures create costs elsewhere: missed deadlines, longer commutes to libraries or public Wi-Fi, interrupted schoolwork and fewer options for remote employment.
The country has already decided that broadband is necessary enough to spend billions extending it. The unfinished work is deciding what success looks like after the line reaches the neighborhood. The 6 million households the Census Bureau counted without home internet access are not a rounding error in an otherwise completed system. They are evidence that building the network and being able to use it remain separate achievements. If broadband is going to function as basic infrastructure, policy will eventually have to measure both.