What Does “Fair Share” Actually Mean? Putting Math Behind Europe’s Network Cost Debate

Europe has an infrastructure problem. Telecom operators say they need billions to build 5G and fiber networks, but a handful of streaming and cloud companies generate most of the traffic without paying for the pipes. Meanwhile, content providers point out they already invest heavily in data centers and content delivery and warn that any “fair share” payments would violate net neutrality and ultimately hurt consumers. In fact, EU net neutrality was not meant to prevent welfare enhancing partnerships (indeed the law defines commercial categories of specialized services and 5G slicing to facilitate content provider innovation), but content providers have abused the rules, claiming that any legitimate payment is a violation of the rules. Parties like Meta have used lawfare to gum up interpretation in court, a process which takes years to resolve, even when ISPs are proven justified to charge for use of their networks. Just look at DT v. Meta in which Meta stopped paying for its agreed interconnection contract and claimed that net neutrality entitled it to continue “free” specialized service.

Both sides claim fairness is on their side. But what does “fair” actually mean?

The Missing Piece

In a working paper recently presented at major telecommunications conferences, we applied a tool economists have used for decades to solve exactly these kinds of problems: the Shapley value. Originally developed for cooperative games, it’s been used to allocate costs fairly in everything from airport runways to irrigation canals to climate change mitigation.

The insight is simple. When multiple parties jointly create value – or in this case, jointly create costs – we can determine each party’s fair share by looking at their contribution across every possible scenario. Not their bargaining power. Not their political influence. Their actual contribution to the outcome.

The 50-50 Finding

Here’s what the math shows: content providers and internet service providers should each pay half of traffic-related infrastructure costs. Consumers should pay nothing directly (besides their subscription fees of course).

This isn’t a political compromise. It’s what falls out when you apply rigorous fairness principles to the actual market structure. Both ISPs and content providers are indispensable to generating traffic – neither can create it alone. Consumers respond to what’s offered to them but don’t drive the underlying cost structure.

When there are multiple firms on one side of the market, they split their half further. So, in a market with four ISPs competing, each would carry half of their total traffic cost, while the content side collectively covers the other half.

Why It Matters

Europe faces a genuine infrastructure shortfall. Current market conditions make it difficult for ISPs to recover costs through retail prices.

The result: systematic underinvestment in the networks everyone depends on.

But if regulators were to implement cost-sharing based on this principle, my research shows several benefits:

• The total of internet access prices and content subscription fees would decrease
• Traffic would moderate to sustainable levels (content providers would have incentive to optimize, not just maximize, quality)
• Infrastructure investment would improve
• Consumer welfare would increase in most realistic scenarios

The key is getting the allocation right. Ad hoc negotiations favor whoever has more market power. Political pressure produces arbitrary splits that may not reflect economic reality. The Shapley approach provides a principled benchmark.

Beyond Europe

While this framework addresses European policy debates, it may be even more valuable in developing countries building high-capacity networks from scratch. Establishing sustainable funding models at the infrastructure planning stage – rather than retrofitting them later – could help ensure these massive investments don’t require perpetual subsidies or fall into disrepair. Countries in Sub-Saharan Africa, South Asia, and Latin America have an opportunity to implement efficient cost-sharing from day one.

South Korea has already wrestled with these questions through the Netflix-SK Broadband dispute. Other countries debate similar issues around universal service funding and infrastructure subsidies. The Shapley benchmark provides a rigorous measure: does any proposed cost split reflect actual contributions to value creation, or merely bargaining leverage?

The Path Forward

Europe’s “fair share” debate has been stuck because both sides appeal to fairness without defining it. Operators claim free-riding. Platforms warn about threats to the open internet. Consumers worry about higher prices. Many experts say that fair share regulation efforts have failed in the EU. The reason may be that stakeholders avoided talking about what fairness actually means, letting differences in bargaining power dominate a normative discussion. Efforts to rigorously define fairness in this context might resurrect the debate and move us closer to socially desirable outcomes.

The math cuts through the rhetoric. A 50-50 split between ISPs (collectively) and content providers isn’t a compromise – it’s what fairness looks like when both parties are essential to creating the costs in question.

Whether regulators mandate cost-sharing or not, this framework clarifies the conversation. It shows that properly designed allocation can close infrastructure gaps without harming consumers. As policymakers worldwide search for sustainable funding models, they now might be one step closer to a rigorous answer to what “fair share” actually means.

Request the Working Paper “What To Mean by “Fair” in Europe’s Fair Share Debate?”