After the Affordable Connectivity Program, home Internet access on tribal lands ticks up

The COVID-19 pandemic exposed a critical lack of digital equity in the United States, particularly in Indian Country.

As previous Center for Indian Country Development (CICD) research has shown, before the pandemic households on federally recognized reservations were less likely to have home Internet and faced substantially higher prices for the most basic home Internet plans compared to households living outside of reservations.

COVID-19 exacerbated that digital divide.

In response, in December 2021 the U.S. government launched the Affordable Connectivity Program (ACP). Before sunsetting in June 2024, the ACP subsidized the monthly cost of high-speed Internet for low-income families. The program offered a $30 monthly subsidy for eligible households located outside of tribal lands and a $75 monthly subsidy for eligible households located on tribal lands.

Our analysis finds higher ACP enrollment among eligible households on tribal lands than on non-tribal lands. Analyses like this are part of CICD’s ongoing efforts to help policymakers better understand the conditions necessary for economic prosperity in Indian Country, such as access to broadband.

ACP participation rate on tribal lands exceeds the U.S. average

As shown in Figure 1, the ACP participation rate on tribal lands is much higher than the U.S. average. (See the Appendix section below for our definition of “tribal lands.”) From January 2022, which was the program’s first month, to December 2023, which was the last month before the Federal Communications Commission (FCC) announced that the ACP would not be renewed, the participation rate on tribal lands was consistently 3.5 to 5 percentage points higher than the U.S. average. By December 2023, the participation rate on tribal lands was 44 percent using a conservative definition of tribal lands and 42 percent using a looser definition, compared to the overall U.S. participation rate of 38 percent of eligible households.

The larger subsidy for households on tribal lands logically explains the higher participation rates. But how much of a factor was it? To test the effect of the larger subsidy, we used regression methods to estimate whether we still see higher participation rates on tribal lands after accounting for a host of factors that may also explain tribal land participation.

We find that, after controlling for a host of additional determinants, such as age distribution and poverty rate, households on tribal lands remain roughly 3 percentage points more likely to participate in the ACP than households located on non-tribal lands.¹ Thus, our results suggest that the higher subsidy rate for households living on tribal lands was likely a driver of the higher participation rates on tribal lands.

Unfortunately, the FCC identified instances of non-tribal Internet service providers (ISPs) fraudulently claiming ACP funding. For example, according to the FCC, one non-tribal ISP fraudulently claimed that many of its subscribers lived on tribal lands to pocket the higher tribal-lands reimbursement subsidy.

While we cannot identify which subscriptions are incorrectly assigned to tribal lands, we believe the identified fraud does not impact our analysis of differences in ACP participation rates.

First, if many subscribers were assigned to tribal lands because of fraudulent ISP activity, we would expect to see large swings in the tribal-lands ACP participation rate. As shown in Figure 1, we do not see those wide swings in ACP participation.

Second, according to the FCC, one of the ISPs erroneously assigned roughly half of the 7,700 subscribers in one particular month (February 2023) to tribal lands. If we drop 3,850 subscribers from the tribal-lands subscriptions each month, our average ACP participation rate on tribal lands would only decrease by roughly one-half of one percentage point.

To better understand ACP’s impact on tribal communities, we contacted Frances Goli, the tribal broadband manager of the Shoshone-Bannock Tribes. According to Goli, the ACP and the Lifeline Program, another FCC program, have made Internet service obtainable for many community members.

“The Affordable Connectivity Program and Lifeline Program have proven to be transformative within our tribal community, significantly enhancing Internet access and affordability,” she said. “The continuation of these programs is not just important but vital. In our efforts to serve the Fort Hall tribal lands, we’ve encountered numerous challenges in assisting our Shoshone-Bannock tribal and community members. However, by collaborating with tribal departments and focusing on participants enrolled in critical programs like Tribal Temporary Assistance for Needy Families, the Food Distribution Program on Indian Reservations, USDA Community Eligibility Provision Participation, Medicaid, and Supplemental Nutrition Assistance Program, we’ve identified individuals who automatically qualify for the Affordable Connectivity Program and Lifeline Program.”

Goli added that many tribal community members initially lacked awareness of Internet subsidies. “Still,” she said, “through this collaborative approach and government support, assistance has been provided to our tribal community members with Internet service that was previously financially out of reach.”

Digital divide still exists but is shrinking

The bad news is that full access to the Internet and broadband remains a concern in Native communities. But there’s good news. This digital divide is shrinking.

Figure 2 illustrates what is happening to the gap in Internet outcomes for households located in Native Geographies.² The share of households without Internet in the United States has declined from 17.6 percent in 2017 to 9 percent in 2022, while the share of households without Internet in Native Geographies has declined even more: from 26.7 percent in 2017 to 15.3 percent in 2022.

Figure 2 also shows that the relative share of households in Native Geographies that use only cellular plans to access home Internet or satellite Internet has also shrunk.

Despite the increase in the share of households with an Internet connection, connectivity in Native Geographies continues to fall shy of the national average. For ACP participation on tribal lands, we estimate that the highest participation rate in any given month, while higher than elsewhere, was still less than 50 percent.

Our prior research showed that the cost of basic home Internet is higher on tribal lands than on neighboring non-tribal lands. The ACP directly addressed this issue by providing an incentive to make this critical service more affordable for households on tribal lands. According to our new analysis and research by others, providing such Internet subsidies may reduce digital inequities. These findings suggest that by taking into account the unique needs of Indian Country, the ACP may be instructive for governments, broadband providers, and stakeholders engaged in ongoing efforts to close the tribal digital divide.

Appendix: How we define ACP participation rates for tribal lands

To determine the ACP participation rate, we divide the number of households with an ACP subscription by the estimated number of households eligible to participate in the program. ACP subscription data come from the Universal Service Administrative Company (USAC), an independent nonprofit organization designated by the FCC to administer phone and Internet programs like the ACP. These data were released to the public every month from January 2022 to February 2024 and published by ZIP Code.³

Households can qualify through many criteria, including having a household income of 200 percent or less of the federal poverty level, being eligible for the Supplemental Nutrition Assistance Program (SNAP) or Medicaid, or participating in qualifying tribal programs. Because some of these criteria are hard to observe in public data, the true number of eligible households is unknown to researchers.

As a result, we estimate ACP eligibility using a method developed by Hernan Galperin. We first use the 2018–2022 American Community Survey to find the number of households in a ZIP Code Tabulation Area or ZCTA (a unit of geography equivalent to a ZIP Code) that qualify for the ACP because their household income is at or below 200 percent of the federal poverty level. Since households can also be eligible for the ACP if a household member participates in federal welfare programs, we compute state-level participation rates in qualifying programs, such as Medicaid, SNAP, Supplemental Security Income, and other sources of public assistance income taken from the American Community Survey microdata. The final estimate of the ACP eligibility for each ZCTA adjusts the number of income-eligible households to account for differences in state-level participation in these qualifying programs.

Since the total number of ACP subscriptions per month is published by ZIP Code, we first crosswalk ZIP Codes to ZCTAs, which allows us to map ZIP Codes to geographic areas that they serve and then incorporate the USAC shapefile of tribal lands to determine the share of each ZCTA’s area that overlaps with tribal lands.

We conservatively characterize a household as being located on tribal lands if a ZCTA is entirely within tribal lands. Since this assignment rule omits many households on smaller tribal lands, we alternatively characterize households as being located on tribal lands if 50 percent or more of a ZCTA’s area overlaps with tribal lands. These two assignment rules together capture the impact of the ACP on tribal lands.

Using our broad definition of a tribal land ZCTA, we find that approximately 374,816 households in these ZCTAs enrolled in the program. This number is greater than the 329,459 that the USAC has confirmed are on tribal lands because it includes people who live near a reservation but not on tribal lands. USAC’s number, however, may exclude those living on tribal lands who had trouble proving they lived on tribal lands in the enrollment process. USAC’s numbers may also include those living near tribal lands who have been erroneously marked as on tribal lands.

Endnotes

1 For our analysis, we regressed the share of households with ACP subscriptions in December 2023 on the share of households that were eligible, an indicator equal to one if 50 percent or more of their Zip Code Tabulation Areas overlapped with Native Geographies, and additional factors used by Horrigan et al. (2023).

2 The Census Bureau collects data on “Native Geographies,” defined as all American Indian, Alaska Native, and Native Hawaiian Areas. Thus, when we compare Internet-related outcomes over time in this section of the article, we are technically referring to households living in Native Geographies.

3 We narrow our analysis to the period between January 2022 and December 2023. Announcements about the potential discontinuation of ACP funding in January 2024 and February 2024 may have contributed to a drop in enrollments.