Introduction

Regional Internet Registries, commonly known as RIRs, play a central role in the way the Internet is organised. They manage the registration and distribution of Internet number resources, including IP addresses and Autonomous System Numbers, across different geographic regions.

At first glance, this sounds like a purely technical function. Every network needs unique addresses. Those addresses must be registered somewhere. Someone must make sure that the same number resource is not assigned twice.

But the modern RIR system has grown far beyond a simple registry function. Today, RIRs operate as large institutions funded by membership fees paid by Internet service providers, hosting companies, enterprises, governments, universities, and other network operators. These costs do not remain isolated at the operator level. Over time, they flow through the connectivity supply chain and are ultimately reflected in the cost of Internet access.

This raises an important question: is the current scale of RIR spending justified by the actual function RIRs perform?

To answer that, we first need to understand how Regional Internet Registries handle IP allocation — and then examine whether the cost of maintaining this system remains proportionate, fair, and accountable.

What Are Regional Internet Registries?

A Regional Internet Registry is an organisation responsible for managing Internet number resources within a specific part of the world. These resources include:

  • IPv4 addresses
  • IPv6 addresses
  • Autonomous System Numbers
  • Registration records connected to those resources

There are five Regional Internet Registries:

  1. AFRINIC — Africa
  2. APNIC — Asia Pacific
  3. ARIN — North America
  4. LACNIC — Latin America and the Caribbean
  5. RIPE NCC — Europe, the Middle East, and parts of Central Asia

Together, these five organisations form the global RIR system. They do not “own” the Internet, but they control the registration layer for the number resources that networks need in order to operate.

This makes RIRs highly unusual institutions. They are private, non-profit, membership-based organisations, yet they perform a function that is essential to public connectivity. A network cannot simply choose from dozens of competing registries. In practice, each RIR holds a regional monopoly over the registration of Internet number resources in its service area.

That monopoly is the foundation of the current funding model.

How IP Allocation Works

IP allocation is the process through which address space is distributed to organisations that need it for network operations.

The general structure is hierarchical:

  1. The Internet Assigned Numbers Authority allocates large blocks of number resources to RIRs.
  2. RIRs allocate or assign resources to members and network operators in their regions.
  3. Operators use those resources to connect users, devices, servers, routers, and services to the Internet.
  4. Registration databases record who holds which resources and how they are routed or administered.

Historically, this process was designed around fairness, conservation, and technical coordination. IPv4 addresses were limited, so RIRs used needs-based policies to decide who could receive address space and how much they could receive.

An organisation requesting IP addresses typically had to show:

  • why it needed the resources;
  • how the resources would be used;
  • whether previous allocations had been efficiently used;
  • whether the request complied with regional policy.

This was the classic hostmaster function: reviewing requests, checking documentation, preventing hoarding, and ensuring that scarce IPv4 addresses were distributed responsibly.

In the 1990s and early 2000s, this role mattered greatly. The Internet was expanding quickly, IPv4 was finite, and global coordination was necessary to prevent chaos.

The Core Function: Maintaining a Registration Database

Despite the complexity of Internet governance discussions, the core function of an RIR is relatively simple: maintaining an authoritative registration database.

That database records which organisation holds which Internet number resources. It supports routing coordination, abuse contact lookup, network troubleshooting, and administrative clarity. The database must be accurate, secure, resilient, and publicly useful.

This is important work. But it is not the same as saying it must require large bureaucracies, expensive travel programmes, global conferences, and expanding administrative structures.

At its technical core, an IP registry is a database and a set of operational processes around that database. Modern database infrastructure is not unusually expensive to run. The cost of maintaining accurate records, secure systems, and customer support should be real — but it should also be proportionate.

The governance problem begins when a narrow technical mandate becomes the basis for a much broader institutional footprint.

For a deeper discussion of this imbalance, see: On the Cost Structure of Regional Internet Registries.

Why RIRs Became Large Institutions

The growth of RIR institutions is historical.

When IPv4 was still being actively distributed from available pools, RIRs had a clear scarcity-management role. They needed staff to evaluate requests, prevent waste, enforce community policies, and manage allocation records. Their work was tied directly to the fair distribution of a limited technical resource.

But the Internet changed.

IPv4 exhaustion reduced the traditional allocation role. IPv6 changed the scarcity logic. Transfer markets emerged. Many large IPv4 holdings became economic assets. The registry function remained necessary, but the original justification for intensive needs-based allocation weakened.

Rather than shrinking back toward a lean registry model, RIRs expanded sideways. Over time, many built broader programmes around:

  • training;
  • policy forums;
  • conferences;
  • international travel;
  • community outreach;
  • communications;
  • governance events;
  • research;
  • administrative support;
  • institutional partnerships.

Some of these activities may provide value to parts of the Internet community. The question is not whether any non-registry activity is useful. The question is whether all of it should be funded through compulsory or near-compulsory fees attached to essential number resources.

When access to a required registry function becomes the funding base for a large institutional ecosystem, cost discipline becomes an ethical issue.

The Problem With Mandatory RIR Fees

RIRs are funded largely through membership and resource fees. For network operators that need direct access to IP addresses or ASNs, these fees are difficult to avoid.

This creates a de facto compulsory payment system. Operators pay the registry. ISPs pass costs into their broader operating model. End users eventually bear part of the cost through connectivity pricing.

In that sense, RIR fees function like a hidden global tax on Internet access.

This does not mean RIRs are governments. They are not. But the economic effect is similar: a universal infrastructure cost is imposed through a monopoly layer that users cannot meaningfully opt out of.

When a cost is compulsory, the ethical expectation should be minimisation. Essential infrastructure should be lean, predictable, and focused on its mandate. The registry should charge what is necessary to maintain accurate records, secure systems, accountable governance, and basic member services — not what is needed to sustain institutional expansion.

Yet monopoly systems often move in the opposite direction. When there is no real competitor and no realistic exit option, internal growth becomes easier to justify. Programmes multiply. Travel becomes normal. Meetings become annual rituals. Administrative structures become permanent.

The result is a mismatch between function and cost.

Who Pays the Price?

The burden of RIR costs does not fall equally.

A large multinational network may treat registry fees as a minor operational cost. A major cloud provider or global telecom company can absorb the expense with little difficulty.

But the situation is different for small operators, rural ISPs, schools, local networks, and connectivity providers in underserved regions. For them, every recurring infrastructure cost matters.

A child in a rural school in Southeast Asia, Africa, or Latin America may never attend an RIR meeting, never vote in a policy forum, and never benefit from international governance travel. Yet the cost of maintaining that system can still reach them indirectly through the price of connectivity.

This is the moral tension at the heart of the RIR funding model.

Critical Internet infrastructure should not extract maximum rent from the least powerful users on the planet. If the registry layer exists to support universal connectivity, its cost structure should reflect that mission. The poorest users should not be subsidising institutional activities that primarily benefit insiders, frequent participants, and professional governance communities.

The Gap Between Registry Work and Institutional Spending

Many members understand that RIRs provide IP addresses and maintain registration records. Fewer understand how much of an RIR budget may go toward non-core activities.

A registry budget can include costs related to:

  • staff salaries and benefits;
  • office administration;
  • legal services;
  • governance processes;
  • conferences and events;
  • travel;
  • training programmes;
  • communications;
  • community engagement;
  • external partnerships;
  • research and public policy work.

Some of these costs may be defensible. But members should be able to distinguish between the cost of operating the registry and the cost of maintaining the institution around the registry.

That distinction is essential.

If the actual registry database function is only a fraction of total spending, members deserve to know. If travel, events, communications, and institutional overhead consume a large share of resources, members should be able to question whether those activities are necessary, efficient, and fairly funded.

Transparency alone is not enough. Publishing a budget does not automatically create accountability. Accountability requires members to compare spending against mandate, challenge unnecessary expansion, and vote accordingly.

Thick Governance and Double Extraction

The modern RIR system also creates a deeper governance concern.

RIRs manage unique Internet number resources. That uniqueness gives their databases power. If a registry record determines who is recognised as the holder of a resource, then the registry function becomes more than clerical. It becomes economically significant.

This is especially true in the IPv4 era, where address space has market value. A registry’s policies, procedures, and recognition decisions can affect whether an operator can use, transfer, lease, secure, or monetise an IP resource.

This creates what critics describe as a form of double extraction. First, operators must pay fees to remain recognised within the registry system. Second, thick governance structures may restrict or shape the economic use of resources that operators already depend on for real networks.

For more on this argument, see: On Regional Internet Registries’ Thick Governance Turns Uniqueness into Double Extraction.

The more RIRs move beyond narrow registration and into broader governance, the more important it becomes to ask who benefits, who pays, and who carries the risk.

Are RIRs Still Necessary?

The answer is not simple.

The Internet still needs accurate number resource registration. It still needs uniqueness. It still needs coordination. If two networks claim the same address space, routing stability suffers. If registry data becomes unreliable, abuse reporting, network operations, and resource transfers become harder.

So the issue is not whether the registry function matters. It does.

The issue is whether the current RIR institutional model is the only way — or the best way — to perform that function.

A leaner model could focus on:

  • maintaining accurate registration records;
  • securing registry databases;
  • supporting routing stability;
  • providing basic member services;
  • publishing transparent financial reports;
  • limiting non-core spending;
  • lowering barriers for underserved operators;
  • separating essential registry costs from optional community programmes.

This would not eliminate governance. It would make governance more disciplined.

What Members Should Demand

RIR members should not treat registry fees as unavoidable background noise. They should ask direct questions:

  1. What percentage of the budget is spent on core registry operations?
  2. What percentage is spent on conferences, travel, outreach, and communications?
  3. Are fees structured fairly for small operators and underserved regions?
  4. Are non-core programmes funded voluntarily or through compulsory member fees?
  5. Are staffing levels proportionate to the registry’s actual workload?
  6. Are reserves, surpluses, and fee increases clearly justified?
  7. Do members have meaningful power to reject unnecessary spending?

These questions are not anti-RIR. They are pro-accountability.

The Internet community should not confuse institutional loyalty with responsible stewardship. A registry that performs an essential function should welcome scrutiny. The more critical the infrastructure, the stronger the obligation to remain lean, transparent, and accountable.

The Future of IP Allocation

The future of IP allocation will not look like the past.

IPv4 scarcity has already changed the economics of number resources. IPv6 adoption continues, but IPv4 remains deeply embedded in global networks. Transfer markets, leasing models, routing security, legal disputes, and regional policy differences all place pressure on the old RIR framework.

At the same time, the legitimacy of the RIR system depends on trust. If members believe fees are excessive, governance is too thick, or institutions are expanding beyond mandate, that trust weakens.

For a broader explanation of RIRs and their role, see: What You Need to Know About Regional Internet Registries.

For a deeper critique of the model’s structural weaknesses, see: Why the Current Regional Internet Registries Model Is Structurally Fragile.

The next phase of Internet number governance should be built around a simple principle: essential infrastructure must be financially restrained.

Conclusion

Regional Internet Registries handle IP allocation by maintaining the systems, policies, and records that allow Internet number resources to remain unique and usable. That function is important. Without reliable registration, the Internet would become less stable, less secure, and harder to operate.

But importance does not justify unlimited institutional growth.

The core RIR function is narrow: maintain accurate registration of Internet number resources. Around that function, large organisations have developed with expanding budgets, staff, meetings, travel, and governance programmes. Because these organisations hold regional monopolies over essential number resource registration, their costs are ultimately borne by the wider Internet community.

That burden falls hardest on those least able to pay.

If RIRs want to maintain legitimacy, they must prove that their spending is proportionate to their mandate. Members must demand clearer budget separation between core registry operations and optional institutional activity. Fees should be minimised wherever possible, especially for operators serving underserved communities.

The Internet is for everyone. Its essential infrastructure should not be financed through hidden extraction from the most remote and least powerful users.

At minimum, RIR governance must confront one reality: critical Internet infrastructure should be lean, accountable, and focused on service — not institutional self-preservation.

FAQs

What is a Regional Internet Registry?

A Regional Internet Registry, or RIR, is an organisation that manages and registers Internet number resources such as IP addresses and Autonomous System Numbers within a specific geographic region.

How many Regional Internet Registries are there?

There are five RIRs: AFRINIC, APNIC, ARIN, LACNIC, and RIPE NCC. Each serves a different region of the world.

How do RIRs allocate IP addresses?

RIRs allocate IP addresses based on regional policies. Historically, organisations had to demonstrate operational need before receiving IPv4 address space. RIRs also maintain registration records for allocated and assigned resources.

Why do organisations pay RIR fees?

Organisations pay RIR fees to become members, receive number resources, maintain registrations, and access registry services. These fees fund the RIR’s operations and activities.

Why is the RIR cost model controversial?

The RIR cost model is controversial because RIRs hold de facto regional monopolies over essential number resource registration. Critics argue that compulsory fees should fund only lean registry operations, not excessive institutional overhead.

Are RIRs still needed?

The registry function is still needed because the Internet requires unique and accurate number resource registration. The open question is whether today’s high-cost institutional model is the most efficient and accountable way to provide that function.

What should RIR members do?

RIR members should review budgets, question non-core spending, participate in governance, vote in elections, and demand that registry fees remain proportionate to essential registry services.