Conference Agenda

We study the economic effects of broadband deployment in a setting that is central to communications and internet policy: a geographically remote Tribal reservation with thin markets, high fixed costs of service provision, and very low baseline connectivity. Our case is the Colville Reservation in Washington State, where the tribal government launched its own broadband network beginning in 2013 after private carriers did not extend comparable service across the reservation boundary. The rollout sequence was shaped by engineering feasibility, terrain, line-of-sight constraints, and overlapping tribal, state, and federal permitting requirements rather than by short-run local income trends. Public anchor institutions, including government offices, schools, and community centers, were connected first, with the initial network delivering a discrete jump from legacy T1 connections to symmetric 100 Mbps service.

This setting is useful for internet policy because it isolates a case in which market provision was limited, public and tribal institutions played a first-order role in deployment, and the main policy question is not marginal speed upgrades in an already connected market, but whether initial modern broadband access changes economic participation in a high-friction environment. We evaluate that question using two complementary designs. First, we construct a reservation-level synthetic control to estimate the effect of broadband onset in 2013 on income and labor-market outcomes for American Indian residents living on reservation land. Second, we exploit staggered adoption across ZIP code areas within the reservation, where treatment begins in the first year a public anchor institution receives operational broadband, to estimate within-reservation effects on annual administrative outcomes from IRS Statistics of Income and County Business Patterns.

We find large and persistent gains. Relative to a synthetic counterfactual, American Indian per capita income rises by about $4,000 in 2016 dollars, labor-force participation increases by 7.4 percentage points, and the employed share increases by 2.8 percentage points, with gains persisting through 2022. Within the reservation, treated ZIP code areas experience higher adjusted gross income per return, higher wages and salary per return, and increased paid employment at establishments. The ZIP-level estimates imply roughly a 6 to 7 percent increase in adjusted gross income per return, a 9 to 10 percent increase in wages per return, and a 12 to 13 percent increase in paid employment at establishments. Pre-trend tests are supportive of the identifying assumptions, and the paper reports placebo-based inference and robustness checks addressing alternative donor composition, treatment timing, and later-year disclosure suppression in one treated ZIP.

The pattern of results is consistent with a transaction-cost mechanism that is directly relevant for communications policy. Broadband appears to lower search, coordination, and administrative frictions in a place where institutional complexity and remoteness make those frictions especially costly. Household participation margins respond quickly, while establishment employment adjusts more gradually. The broader implication is that broadband policy may generate especially large returns where baseline connectivity is minimal and where local institutional capacity, here tribal capacity, determines whether deployment occurs at all. For communications and internet policy, the Colville case points to the importance of institutionally tailored deployment strategies, anchor-based rollout, and governance structures that can overcome the market and permitting barriers that have historically limited connectivity in Tribal areas.