Conference Agenda

This paper considers whether the geographic drivers of mid-band spectrum values have changed in the 5G era, in which capacity is increasingly important, and the wired infrastructure is being pushed deeper into the network. We examine changes in Relative Value Indices (RVIs), which have long played a central role in understanding geographic variation in spectrum values. Historically, models anchored in population and related demographic metrics have been used to explain a large share of cross-market variation in RVIs. These models reflect the intuition that spectrum demand is closely tied to the distribution of potential users across geographic markets.

We evaluate whether these traditional predictors continue to explain recent auction outcomes and whether complementary infrastructure provides additional explanatory power in the 5G era. Our empirical approach employs regression analysis to test the predictive performance of demographic and infrastructure variables, examine how well RVIs from earlier auctions predict subsequent auction values, and assess the predictability of spectrum at disaggregated geographies. We find that traditional geographic value predictors remain highly relevant, while complementary infrastructure is statistically significant in 5G auctions. Finally, we find evidence that relative value measures are more stable at larger geographies.