Don't Discount the Investments in Internet Infrastructure that Content and Application Providers are Making

Benton Institute for Broadband & Society

Tuesday, October 18, 2022

Digital Beat

Don't Discount the Investments in Internet Infrastructure that Content and Application Providers are Making

David Abecassis

Should network usage fees be imposed on content and application providers to support internet infrastructure? New research from Analysys Mason shows such a mandate would be harmful to end users and the global internet ecosystem.

The internet is a network of networks, which must all be connected (directly or indirectly) to one another to enable traffic delivery from any source to any destination around the globe. Its evolution has been driven by a combination of competition, collaboration, and innovation by all the stakeholders in the value chain. These players include:

  • Internet service providers (ISPs), which provide residential and business end users the means to connect to the internet from their homes, offices, and mobile devices.
  • Tier 1’ global carriers, which invest and operate large-scale transmission networks that move content around the world and connect together the many networks that make up the internet.
  • A wide variety of other companies that provide technology, services, and content to end users and other stakeholders through internet access and are referred to as content and application providers (CAPs). This includes cloud providers which invest in and operate data centers, peering and caching infrastructure, and increasingly their own backbone networks around the world.

Some stakeholders, including large, vertically integrated ISPs, have argued that growing internet traffic creates a cost burden on ISPs, which they argue is unsustainable. A central part of the argument put forward by these stakeholders is the notion that CAPs are benefiting from the network without investing in network infrastructure. As such, they call for policymakers to mandate that CAPs pay ISPs network usage fees that would be based on the amount of traffic delivered to end users.

Analysys Mason demonstrates that:

  1. CAPs are investing significant amounts in internet infrastructure (above and beyond their investments in content and applications for end users), and these infrastructure investments increase over time, reaching nearly $900 billion in total over the period 2011–21.
  2. Network-related costs for ISPs have remained stable over time even while traffic volumes have grown significantly. Data traffic only drives a small share of ISP costs, which are further mitigated by the investments that CAPs make in internet infrastructure.
  3. The arguments put forth by proponents of network usage fees disregard ongoing trends in access network investment, and demonstrate an inadequate understanding of internet interconnection.
  4. If introduced, network usage fees would disrupt existing interconnection arrangements, as well as incentives for stakeholders in the ecosystem to continue investing to deliver a high quality of experience for end users.

Policymakers should consider the potential impact of network usage fees holistically when evaluating regulatory proposals that would mandate the introduction of such fees.

CAPs Invest over $120 Billion Annually in Internet Infrastructure

Over the period 2011–21, CAPs spent $883 billion on digital infrastructure including hosting, transport, and delivery networks, leading to positive impacts on end users, and broader economic benefits.

CAPs focus their internet infrastructure investments on three main clusters—hosting (i.e. data centers), transport (i.e. submarine and terrestrial cables), and delivery (i.e. peering and caching). This infrastructure spans tens of thousands of miles around the globe and is critical to deliver online content and services close to ISPs for the benefit of end users’ online experience.

CAPs are investing heavily in hosting, transport, and delivery networks. In 2018–21, CAPs increased their annual investment by 50% over the previous period (2014–17) and spent on average $120 billion each year on this infrastructure. As a result of these investments, CAPs have spent a total of $883 billion on infrastructure in these three main clusters from 2011 to 2021.

CAPs’ investment in internet infrastructure increases reliability and quality of experience for end users. More broadly, Analysys Mason highlights the many studies that have shown how these investments drive overall internet penetration and usage and, as a result, generate macroeconomic benefits through digitalization. These include increased GDP, job creation, and environmental benefits, as well as better societal outcomes (e.g. education, health, access to remote work) from the consumption of online services. Policymakers have also recognized the important role that the internet can play in unlocking these benefits.

Investments by CAPs in Transport and Delivery Networks have a Positive Impact on the Economics of ISPs

Network usage fees would effectively slow or reverse some of the advances in interconnection, peering, and caching that have evolved through voluntary, mutually beneficial arrangements that have aided ISPs and end users by lowering their costs and improving their service experience, respectively.

CAP investments to bring content closer to ISPs and end users generate benefits for end users in terms of better quality of experience, but also benefit ISPs in terms of cost avoidance or cost savings. For example, CAPs invest in large infrastructure projects like submarine cables, thus reducing the need for ISPs to invest in these systems. CAPs also use their global scale to deliver traffic broadly in internet exchange points (IXPs) and other peering locations across the world, reducing the need for ISPs to purchase transit or connect internationally to CAP ‘home bases.’ CAPs also invest in on-net caches that are embedded inside ISP networks, which reduces the backbone and backhaul capacity that ISPs require to deliver content to end users.

CAP investments contribute to ISP savings in two areas: CAP investments in embedded caching in ISP networks (at core/metro/aggregation nodes), and long-distance transport and peering locations (both public and private), which contribute to the widespread availability of ‘on-shore’ peering in ISP home markets. This enables ISPs to reduce capacity-related costs by between $5.0 billion and $6.4 billion each year, globally.

The central argument for network usage fees relies on two premises that Analysys Mason questions: that CAPs are responsible for large and growing traffic volumes, and that large growth in traffic drives much higher network costs.

CAPs deliver traffic to ISPs when end users demand such content, and as demand for online services grows so does the demand for faster and generally more expensive broadband services that ISPs sell. A small number of large CAPs and content delivery networks (CDNs) deliver a large share of traffic demanded by end users, in part because they are very successful with end users, and in part because of the cost and quality benefits for all CAPs, large and small, to use their services due to their widely distributed CDNs that bring traffic either close to or directly into ISPs’ networks. 

Network costs are expected to continue to remain relatively stable in the future while traffic volumes grow, as fixed networks move toward fiber-based architectures, and as mobile technologies evolve to enable operators to add network capacity more efficiently, further demonstrating the unreasonableness of any permanent transfer of mandated payments from CAPs to ISPs.

Policymakers should Consider Regulatory Objectives Holistically and Scrutinize Arguments in Favor of Network Usage Fees

Proposals largely call for fees to be transferred from CAPs to ISPs on the basis of traffic for internet interconnection, one argument being that this mirrors voice termination rates in the telephony market. These mechanisms have worked for voice services as it is easy to identify the party that originated the call.

Network usage fees would lead to regulatory and competition issues that policymakers already understand well: they have rejected network usage fees for the internet in the past, and have worked to mitigate similar issues in telephony markets for the last 20 years.

For internet traffic, however, it is usually difficult to identify the originator of a stream of traffic, not least because CAPs send traffic in response to an end-user request. There also would be the challenge of deciding what the rate should be, where it is imposed, which entities are charged, how to reconcile these charges with non-discrimination and net-neutrality policies, and how to limit ISPs’ ability to exercise their termination monopoly. These challenges could result in excessive rates, leading to further regulation of quality of service, in addition to higher costs for end users. Some of these concerns have been raised in the past, for instance, when European regulators rejected similar proposals to regulate interconnection that emerged a decade ago.

Proponents of network usage fees suggest that ISPs would invest more in connectivity and accelerate broadband deployment if they were able to charge CAPs for traffic. However, these arguments appear to disregard the large ongoing commitments made by ISPs themselves and by policymakers and other investors to roll out full-fiber networks throughout Europe, achieve ‘Internet for All’ in the US, and via other initiatives that are already underway for deploying broadband networks around the globe to unserved and underserved areas. Moreover, current proposals have not elaborated on mechanisms for ensuring ISPs use such fees on network investments that help to improve connectivity and end-user experience.

In this context, it seems unlikely that network usage fees would result in ISPs investing any more in networks. Instead, already large and vertically integrated ISPs would likely enjoy higher profits and shareholder returns at the expense of end users, who would face higher prices and a lower quality of experience.

Implementing Network Usage Fees could Disrupt Existing Interconnection Arrangements and Investment Dynamics, and Reverse Gains made in Connectivity

Network usage fees would effectively slow or reverse some of the advances in interconnection, peering, and caching that have evolved through voluntary, mutually beneficial arrangements that have aided ISPs and end users by lowering their costs and improving their service experience.

The introduction of network usage fees would disrupt existing interconnection arrangements. This is likely to affect incentives for both CAPs and ISPs to continue making investments that deliver ongoing improvements in quality of experience for end users. Network usage fees would raise costs for all CAPs, not just larger ones, resulting in barriers to entry and expansion for online content and service providers. Reduced incentives for CAPs to continue investing in infrastructure and processes that optimize traffic delivery will result in higher costs for ISPs as well, constraining resources for organic investment in ISP networks. Moreover, fees proportional to traffic paid directly to ISPs would favor larger ISPs, which may distort competition in the ISP market. As a result of these effects, end users are likely to face higher prices, reduced quality, and less choice in the ISP market, while also receiving a lower quality of experience for online services.

South Korea is currently the only country where the regulator has mandated payments from domestic CAPs and ISPs. The added costs imposed by network usage fees have led to higher transit costs, diverging from other countries in the region. As a result, Korean CAPs have found it challenging to host content domestically due to higher costs and have either moved overseas or have become less competitive. Likewise, service quality is affected as the overall average latency experienced by users in South Korea is the highest among Organization for Economic Co-operation and Development countries. Importantly, the introduction of network usage fees elsewhere could disincentivize CAPs or CDNs from deploying caches domestically in those other countries as well, leading to similar negative effects as those seen in South Korea.

Demand for online services and demand for broadband access are inherently linked. The impact of introducing network usage fees, and the resulting impact on end users, could be long-lasting and harmful for both markets. Lower consumption of online services by individuals and businesses could also result in further negative effects in terms of slower digitalization and economic growth.


Tech’s Investment in the Modern InternetBased on current proposals, network usage fees are unlikely to be beneficial to end users. These proposals are supported by arguments that mischaracterize the relationship between traffic delivery and cost, and that appear to be based on an inadequate understanding of internet interconnection. If implemented, network usage fees would result in a fundamental shift away from the voluntary collaboration that has sustained the rapid growth of the internet thus far, and negatively affect a wide range of stakeholders. Policymakers and regulators should scrutinize any proposal on network usage fees and take a holistic perspective on the potential harmful impact of those fees on the internet ecosystem.

For more see Analysys Mason's The Impact of Tech Companies' Network Investment on the Economic of Broadband ISPs

David Abecassis is a Partner in Analysys Mason’s London office. Since joining the firm in 2004, he has advised clients on cutting-edge topics including public policy, economic regulation, corporate strategy, due diligence and corporate finance. He has worked in over 30 countries across the world – previously based in our offices in Singapore and Milan, he continues to serve clients worldwide. 

The Impact of Tech Companies' Network Investment on the Economic of Broadband ISPs was commissioned by INCOMPAS and prepared independently by Analysys Mason.

The Benton Institute for Broadband & Society is a non-profit organization dedicated to ensuring that all people in the U.S. have access to competitive, High-Performance Broadband regardless of where they live or who they are. We believe communication policy - rooted in the values of access, equity, and diversity - has the power to deliver new opportunities and strengthen communities.

© Benton Institute for Broadband & Society 2022. Redistribution of this email publication - both internally and externally - is encouraged if it includes this copyright statement.

For subscribe/unsubscribe info, please email headlinesATbentonDOTorg

Kevin Taglang

Kevin Taglang
Executive Editor, Communications-related Headlines
Benton Institute
for Broadband & Society
1041 Ridge Rd, Unit 214
Wilmette, IL 60091
headlines AT benton DOT org

Share this edition:

Benton Institute for Broadband & Society Benton Institute for Broadband & Society Benton Institute for Broadband & Society

Benton Institute for Broadband & Society

Broadband Delivers Opportunities and Strengthens Communities