Conference Agenda

Cable operators across the United States are at varying stages of planning and deploying DOCSIS 4.0 upgrades — encompassing both DOCSIS 4.0 Full Duplex (FDX) and Extended Spectrum (ESD) paths — that promise symmetric multi-gigabit speeds over existing hybrid fiber-coaxial plant. These investment decisions are unfolding against a backdrop of intensifying intermodal competition: fiber-to-the-premises (FTTP) overbuilders are expanding into cable incumbent territories, and fixed wireless access (FWA) providers are capturing share at the value end of the broadband market. This paper investigates the role of competitive pressure as a determinant of cable operators' upgrade timing and geographic prioritization, combining reduced-form empirical analysis with a calibrated structural model of investment under competition.

The research question asks: to what extent do the presence and intensity of fiber and FWA competition explain observed variation in DOCSIS 4.0 deployment decisions across local geographies (e.g., Census Block, or Census Block Groups)? A large theoretical literature on investment under competition offers competing predictions. For example, Arrow-type replacement effects suggest incumbents underinvest when they cannibalize existing rents,[1] while incentives to defend market position and fight new entry suggest existing incumbents could generate more investment. More recent "investment ladder" type models of competition generate nuanced predictions about how the type of competitive threat (e.g., symmetric fiber versus capacity-constrained FWA) and the quality gap it creates interact with incumbent cost structures to shape upgrade incentives. The paper situates DOCSIS 4.0 deployment within this theoretical landscape and tests which patterns dominate empirically.

The empirical methodology proceeds in two stages. The first stage estimates reduced-form models using the FCC Broadband Map data, relating observed DOCSIS 4.0 upgrade activity to local market characteristics, including measures of fiber overbuild presence and penetration, FWA availability and take-up, demographic and demand-side variables, and cable plant characteristics that affect upgrade cost (such as node density and homes passed per node). The second stage develops a structural model of the cable operator's upgrade decisions. The model represents the operator as choosing the timing and geographic sequence of upgrades to maximize the net present value of its broadband subscriber base, facing heterogeneous competitive threats across its footprint and incurring technology-specific upgrade costs. We explore the parameter space of the upgrade decisions in equilibrium. Comparing the model results to the reduced-form patterns can give us insight into the state of fixed broadband competition in the U.S. and allows us to make inferences about comparative statics such as, whether DOCSIS 4.0 deployment would slow with reduced fiber entry, or whether FWA alone offers enough competitive pressures to trigger DOCSIS 4.0 upgrades.

The paper contributes to the telecommunications economics and competition policy literatures. Its novelty lies in applying the theoretical framework of investment under competition to a network upgrade cycle that is actively underway, using granular data that captures the geographic heterogeneity of competitive conditions within operator footprints. This is directly relevant to contemporary communications policy: regulators and policymakers are evaluating the competitive sufficiency of cable broadband markets, the need for further facilities-based entry, and whether existing incentive structures are producing adequate investment in next-generation capacity. Understanding to what extent and how competition drives these upgrade decisions informs debates over broadband competition policy, permitting and rights-of-way for fiber overbuilders, and the role of FWA as a competitive discipline.

Our priors suggest that fiber overbuild presence is a strong positive predictor of DOCSIS 4.0 upgrade prioritization, consistent with a preemption or competitive-response motive dominating the replacement effect in markets where the quality gap is large. FWA competition is expected to exert a weaker but nonzero effect, consistent with its more limited bandwidth trajectory posing a less acute threat to cable's core subscriber base. The structural model is expected to confirm that competitive intensity is a first-order determinant of upgrade timing, with implications for the pace of next-generation broadband deployment nationwide.

[1] Arrow, Kenneth. 1962. “Economic Welfare and the Allocation of Resources for Invention.” In The Rate and Direction of Inventive Activity: Economic and Social Factors, pp. 609-625. Princeton, NJ: Princeton University Press.