Will broadband pricing support 1Gbps services?

TitleWill broadband pricing support 1Gbps services?
Publication TypeConference Paper
Year of Publication2013
AuthorsA. Domingo, and W. Lehr
Conference Name24th European Regional International Telecommunication Society Conference
Date Published10/2013
Conference LocationFlorence
Keywordsbroadband services, competition, FTTH, NGA networks

Policymakers in many of the leading economies have identified the need for universal access to broadband services as a critical component of their Digital Agendas.[1] Over time, the definition of "broadband" and what ought to constitute appropriate goals in terms of service quality -- most often summarily articulated in terms of the peak speed offered downstream -- has evolved from hundreds of Kbps to tens of Mbps. While there is still no general agreement on how to define broadband nor what should constitute an appropriate speed goal, most policymakers and analysts have targeted services with speeds of 1Gbps or less to be implemented by 2020.

In this paper, we review data on retail service offerings in the United States and from around the world and tentatively conclude that prospects for 1Gbps services becoming more widely available are not implausible even at current pricing levels. The analysis and our conclusion in this paper should be regarded as preliminary and partial, but ought to provide some comfort to those who might fear that significant improvements in broadband service quality cannot be achieved without dramatic increases in broadband service pricing. Indeed, the pricing data and recent service offerings by Google in the United States and by others abroad suggest that order of magnitude improvements in broadband speeds are achievable with pricing close to today's range.

In reaching our conclusions, we rely on earlier work modeling fiber deployment costs for Europe modified to account for conditions in the United States. The focus of this earlier work was on estimating the aggregate investment costs associated with meeting Europe-wide goals of universal deployment of fiber-to-the-home. The model emphasized the significant savings to be realized if new fiber deployments are able to share the significant component of costs associated with conduit and/or outside structures.[2] Translating a techno-economic European FTTH model to the United States and using an aggregate investment model based on a stylized representation of the technical architecture provides, at best, only a noisy order-of-magnitude estimate of the likely total costs of deploying Gbps-capable infrastructure, but we believe this is sufficient for the goals of this paper, and also helpful for policymakers when framing digital agendas.[3] Absent some such attempt to quantify aggregate costs, there is no a priori reason not to anticipate a steep increase in revenue requirements from end-users or in government universal service subsidies if the goal is to achieve order-of-magnitude enhancements in last-mile infrastructure capacity.[4] Simply pointing to "Moore's Law" improvements in technology when much of last-mile investment includes civil-engineering and construction expenditures is of dubious value.

We compare our estimates of the aggregate investment requirements with recent tariff data for broadband services,[5] and make back-of-the-envelope estimates of the contribution that those retail prices imply for other elements of the broadband service in order to derive an estimate of the contribution available for investment in next-generation last-mile infrastructure. We also examine in some detail recent offerings of fast broadband from Google in the United States and by Free-Iliad and Adamo in Europe to explore how their business models, from a new entrant's point of view, exploit sharing opportunities and novel strategies to compete in today's markets. Our analysis of the potential revenue contribution available for investment under current broadband pricing and of the strategies being used by new players offering fast broadband services support our conclusion that it is feasible to extend much higher-speed offerings (up to 1Gbps) in the United States without requiring order-of-magnitude increases in operator revenue requirements (or public subsidization) to justify the requisite aggregate investment.

We stress that our analysis is agnostic about what may actually happen as a consequence of regulatory policies, competitive interactions, or market strategies. Addressing these factors and verifying the analysis provided herein will require further research. Clearly, there are many unknowns, including whether households generally need or will want 1Gbps services, whether mobile broadband from cellular providers or other wireless offerings will compete so as to make the transition to "FTTH"[6] unnecessary to realize order of magnitude performance improvements, or whether the goal will be to sustain near-universal facilities-based competition among wireline competitors.[7] These and other factors, as well as refined empirical analysis of broadband infrastructure economics has the potential to significantly alter our results. Nevertheless, we are encouraged that our analysis implies capacity can continue to scale without requiring commensurate scaling in end-user payments which we suspect is a dubious prospect in any case. Furthermore, we find our analysis useful in highlighting the potential benefits of infrastructure sharing, while recognizing that significant further research is needed to quantify the costs and benefits of alternative national broadband strategies and the NGAN deployments currently underway.


[1]Both the USA and EU have Digital Broadband Agendas: Digital Agenda for Europe by the European Commission, see EC (2010), and The National Broadband Plan for USA, see the FCC (2010.b). For a comparison of digital agendas from a number of countries, see OECD (2011).  Most of those articulate a national commitment and goals to provide universal access, although none are as ambitious as the commitment by the Australian Government to provide a national broadband network to all citizens.

[2]The main conclusion of that paper was that it could decrease the total investment on FTTH for Europe if sharing infrastructure was envisaged by approximately a 40%, see Domingo and Oliver (2011).

[3]The original model also does not attempt to address variations in regulatory and other local conditions across European markets, which we recognize are significant factors that a more complete analysis ought to include (e.g., variations in terrain that impact costs, cultural attitudes that may impact demand, or other factors).

[4]Throughout this paper, we will occasionally use "capacity" and "speed" as synonyms. We recognize that these are distinct concepts and equating faster data rates with higher quality infrastructure or services or with broadband capacity is only approximate.

[5]See the FCC (2012) where a Fixed Broadband Price Catalogue was provided.

[6]We put "FTTH" in quotes because what we are talking about is dense fiber deployments which may terminate close to the home, but use some other technology for the last couple of tens or hundreds of feet as the FCC has done in their analysis, see FCC (2010.a), which includes many different types of FTTx, where in addition to x being the home (FTTH), the node (FTTN), the cabinet (FTTC), etcetera.

[7]That is, the analysis undertaken here addresses the challenge of achieving (near) universal service rather than (near) universal duopoly service by wired competitors, which is a much more expensive proposition.