Exploring the heterogeneous link between broadband investment and coverage expansion using unconditional quantile regressions

Abstract

One of the primary challenges of the telecommunications industry is to address the broadband connectivity divide. Policy makers, regulators, and network operators are in need to understand the amount of capital required to address this gap and generate solid evidence upon which policy options and regulatory remedies should be discussed. To date, most studies aimed at estimating the investment associated with broadband deployment have consisted in ad-hoc approaches based on cost modelling from average estimations, which lack rigor in terms of considering differences in technologies, topography, and population density. In this paper we propose a general approach, based on the use of Unconditional Quantile Regressions (UQR), that addresses the heterogeneities that arise from considering differentiated coverage targets and investment required across the different deployment phases. Our tool was found to successfully account for the increasing investment to fulfil coverage targets. Moreover, it can be applicable to a wide range of countries, as robustness checks conducted provided evidence of being suitable, to some extent, for countries with more geographic challenges (e.g., mountains, forests, etc.). On this basis, the estimated UQR coefficients were also used to simulate 4G, 5G and FTTH expansion for scenarios of accelerated deployments in Latin America, by trading off cost and target network quality requirements. We simulated some coverage goals above the current trends for 2030, targeting on average 98% of the population covered by 4G, 80% of the population covered by 5G, and two-thirds of households passed by FTTH. For the overall Latin America region, the extra capital needed above current investment trends accounts for $ 17,101 million over the period 2023–2029. Considering this challenge, potential regulatory remedies are considered.