From Centralization to Market Concentration
Published in Concentration.
In 2021, a fourth focus area was published on Internet Society Pulse. We called it “Internet Centralization”. Today we’ve changed the name to “Market Concentration” and we wanted to explain why.
The concept of centralization can be seen as the degree to which an activity within a given set of relationships is dependent on a small set of actors or functions. In a fully centralized system, the activity is dependent on one actor or function, acting as a single point of control. In contrast, in a decentralized system, any dependencies or points of control for the activity are more evenly distributed. So we can think of centralization as a concept that describes the degree to which dependencies and control are concentrated among a few actors/functions or distributed among many.
When we started the work on the Centralization focus area we wanted to develop data collection and provide insights about how control, dependencies, and power in the Internet are evolving over time. Efforts to understand these trends are not new at the Internet Society, and go back to a report we published in 2019, “Consolidation in the Internet Economy“, as well as a collaboration with Chatham House in 2020 to inspire academic research on the topic.
What we learnt from this past work, and from similar efforts by others in the Internet community, is that the topic is both multifaceted and complex. And while concerns about the implications of centralization for the Internet persist, the answers are rarely clear cut. So we decided that one of the contributions we could make was to find a metric that can help to monitor the trends of power and control over time. And so the new focus area was born.
A Potential for Confusion
After receiving feedback from the community, we realized that the name we had selected—”Internet Centralization”—creates the potential for confusion. Internet Centralization is a very broad topic, and can be best understood as an umbrella term for a multifaceted set of questions about power, dependencies, and control in the Internet.
One way to approach this is to measure the market concentration held by some very large, centrally important international companies, and to infer the degree of legal influence that could be projected by their home countries. Which is what we did.
But Internet Centralization can also be measured by highlighting the local and regional dependencies on specific parts of centralized Internet infrastructure within a country, or by calling attention to the protective effects of adding new cables, new Internet Exchange Points (IXPs), and new Internet Service Providers (ISPs) within a given country to eliminate traffic chokepoints that can be used as leverage to control and shut down the Internet. In other words, there are many lenses through which the broader questions about Internet Centralization can be approached, and those mentioned above are just a few.
Looking to the Future
We are excited to continue exploring these insights on the Pulse platform. On the newly renamed “Market Concentration” focus area, our measurements and analysis show the potential for control by large international service providers. The Internet Resilience focus area continues to provide the complementary perspective on Internet centralization within a given country. By assessing the development of regional Internet Infrastructure, as well as levels of local diversity and competition, the metrics we use to calculate a country’s resilience help measure local and regional progress toward secure, high performance, unrestricted end-to-end delivery of Internet services.
We’re also learning that these two focus areas have the potential to interact in interesting and unexpected ways. For example, the arrival of international service providers in a market can provide competition for incumbents, make participation at IXPs more attractive, add additional options for international traffic paths, and increase the options available for local content hosting. This may provide a local protective effect, improving Internet performance, reliability, and decentralization, despite contributing to an increase in the global market share of a given multinational provider. The way to explore these complexities is to measure and assess, and to let the data tell the story. Our goal is to expand, deepen, and progress discussions on Internet centralization, not constrain them.
In renaming this focus area, we hope to attract those that are interested in market concentration but may not find the information they need on Pulse since they haven’t thought about it as an issue of “centralization”, while at the same time expanding our own contributions to the discussions under the broader umbrella term of “Internet centralization”.
We believe that this new naming will help do just that.