Partner content: Network traffic per se is not a reliable metric regarding cost; policy must make interconnect agreements transparent and prioritise digital inclusivity
The network fee (fair share) debate began in the European Union (EU) and spread to Brazil, India, and the US. It is notable for some eyebrow-raising claims about trends in internet traffic share and growth.
Pie charts suggesting that most internet traffic is generated by a small number of content and application providers (CAPs) such as Google, Meta, and Netflix make for great headlines, but many of these claims oversimplify the diverse structure of internet networks.
Claims that the largest CAPs take upwards of 56% of global internet traffic are disingenuous, projecting an image of the internet as a homogenous resource. The figures obscure the fact that public broadcasters, government services and start-ups would probably also face higher costs because of network fees being passed onto them. Currently, they depend on the content delivery network (CDN) infrastructure of big tech companies.
Finally, traffic per se is not the best metric to measure the burden on networks; rather, network operators’ cost and profit, and users’ experience of network speeds are much more appropriate.
More traffic doesn’t equal more costs
The metric of increased traffic, more than traffic levels, is vital to the network fee debate. Studies using proprietary and undisclosed network operators’ data find that increased traffic is driving huge costs for telcos.
Yet studies and rebuttals by CAPs have found the opposite, including one estimate by a telecoms consultancy which found that 80-90% of network operators’ costs are largely independent of traffic.
Further, the remaining costs that depend on traffic have become flat over time due to falling hardware costs and greater efficiencies: a BT presentation from 2018 reveals that costs per unit of traffic decreased by over 70% between 2012 and 2018. Modern network equipment can handle much more traffic while costs remain stable.
For all these reasons, broad claims correlating growth in internet traffic with investment in network infrastructure costs are misleading, and policymaking in this area must be based on thorough and independent analysis. Given the striking lack of publicly available data, it would be a mistake to move forward without transparent and independent studies.
Net neutrality
Without net neutrality, providers can control what consumers, who pay for their internet connections, see and do online. Any erosion of this principle will directly harm consumers and small businesses.
The empirical results of previous implementations of network fees are at odds with EU telecom carriers’ claims of the benefits of direct payments from content providers and underscore the negative impacts on network neutrality.
The peering and transit model of internet traffic has been hugely successful, unlike other proposals whereby ISPs pay to send traffic to each other’s customers – notably in South Korea.
In South Korea, the impacts of network fees on net neutrality have been significant, with some research suggesting it has caused increased prices and slowed speeds for high-traffic services. This selective impact on certain content constitutes a de facto violation of the principle of net neutrality.
Digital inclusion
The issue of digital inclusion only grows in urgency as more and more of the world’s social and economic activity goes online. However, network fees are clearly not the answer. If governments decide that CAPs are obliged to invest in bridging the digital divide, public taxes would be a simpler and safer investment model – and can be used to directly spur digital inclusion.
Network fee payments, untethered from the goals of digital inclusion, would represent a profit-subsidy to telcos which, for their part, have consistently failed to meet benchmarks of access for rural and underserved populations.
In the rural EU, for example, the World Economic forum found that in 2021 only 37% of households had access to high-speed internet. Broadly scoped payments to large telcos would further risk exacerbating the digital divide in the EU and globally because they do not guarantee network investments will serve less lucrative remote and low-income regions as well as more profitable urban environments.
Differing economics
Large telcos, which are the principal proponents of the new proposals, face vastly more favourable market dynamics than smaller, rural providers. For example, nearly 90% of respondents in a 2022 survey said that cost was a significant barrier to fibre deployment.
Despite this cost, one study found that broadband deployment in rural areas can have a significant benefit-to-cost ratio. Global telecoms regulators should therefore focus on the real underdogs – rural and marginalised communities.
If CAPs are making disproportionate profits, as network fee proponents seem to imply, it would make sense to apply policies that more explicitly connect any tax with more direct benefits to consumers. Policy-makers might consider windfall taxes similar to those recently implemented by the UK on the oil and gas sector.
Policymakers might also consider directing resources to networks that are designed to serve the public good, including municipal broadband, cooperatives and community-owned network infrastructure.
Open data standards
Given the complicated market dynamics, comprehensive analysis of network fees
must be based on data about interconnection agreements, which are often shrouded in
secrecy. Policymakers need to create transparency about the interconnection market to enable regular, independent analysis of data on internet traffic, network operators’ costs and users’ experience.
Mozilla is working with stakeholders including the World Bank, the International Telecommunication Union and the Internet Society to create open data standards for describing the telecoms sector. These standards could play an important foundational role in creating transparency concerning claims about telecoms infrastructure.
We strongly encourage the European Commission and government bodies to abandon current frameworks for network-fee contributions and instead focus on eliminating the digital divide. In its current framing, the outcome of these processes will benefit either large content providers or large telcos, depending on who wins the clash of titans.
Policymakers have an opportunity to benefit the public instead, particularly rural and poorer communities.
About the authors
Noam Kantor is Senior Policy and Government Relations Analyst and Steve Song is Policy Advisor at Mozilla