What can country panels tell us about broadband price responsiveness? Evidence from the European Union and the Eastern Partnership, 2010–2024

Regulators often judge how strongly broadband adoption responds to price from country panels, because comparable household data are scarce. We examine what such a panel can reveal, using ITU entry-level price baskets for 27 European Union (EU) and six Eastern Partnership (EaP) countries over 2010–2024. A two-way fixed-effects regression of log subscriptions on log prices gives a sizeable negative coefficient whose year-specific values rise steadily towards zero and above, a pattern that suggests broadband has become a necessity. A simulation with no price effect reproduces this pattern when late adopters start with high prices, and the pattern disappears in the data once country-specific trends are removed or the model is estimated in differences. In the EU the association between entry-level prices and adoption is then close to zero throughout; the 90% confidence interval in first differences is [−0.04, 0.02]. In the EaP adoption grew fastest in the years in which affordability improved most, giving coefficients between −0.17 and −0.46. This association rests on movements common to the six countries: with EaP-specific year effects it vanishes, and with small-sample inference it is fragile. Regional data for 175 EU regions show no stronger association in less-developed regions. Country panels of this kind do not identify a price response, and they give no evidence of a break in 2020.