Digital Infrastructure in Thin Markets: Evidence from a Tribal-Led Broadband Rollout

About the paper

We estimate the economic returns to broadband on the Colville Reservation in Washington State, a thin market with near-zero baseline connectivity where a tribal-led rollout began in 2013. We combine a reservation-level synthetic control with a within-reservation staggered-adoption design, using IRS Statistics of Income and County Business Patterns. Because several reservation-wide developments coincided with the rollout, we read the synthetic control as a descriptive upper bound on the total post-2013 divergence, about 33 percent for American Indian per-capita income. Identification rests on the within-reservation design, which differences out developments common across the reservation. There, adjusted gross income rises by about four percent and paid employment at establishments by about nineteen percent, with the income response falling on the intensive margin. The gap between the large reservation-level divergence and the smaller within-reservation effect is the expected signature of bundling set against differencing. The evidence aligns with a transaction-cost mechanism in which broadband lowers search, coordination, and compliance frictions in a high-friction governance environment. A back-of-the-envelope welfare calculation implies benefit--cost ratios well above one, which we read as an upper bound.

Coverage

Presentations

• Southern Economic Association Meetings November 2025
• Markets and Society Conference October 2025