On the Manufactured Narrative of IPv4 Scarcity

Is IPv4 scarce if its price barely registers in the cost of running a server?

Conceptual workshop with an empty tray in the foreground and people, shelves and a trolley carrying blue blocks behind it.
An empty allocation pool is only one part of the picture. Lu Heng argues that scarcity should be judged by whether the cost of obtaining addresses actually constrains network operators.

The claim that IPv4 scarcity is the defining constraint of the modern Internet is repeated so often that it has hardened into doctrine. But when examined through basic economics and operational reality, the claim begins to collapse.

Scarcity is not defined by administrative exhaustion of a registry pool.
Scarcity is defined by price pressure relative to utility.

If a resource is truly scarce, its cost must meaningfully constrain its use.

IPv4 does not meet that test.

Across the global market, leasing an IPv4 address typically costs about $0.25–$0.55 per month, with averages around $0.40 per address per month. 

That means a public IPv4 address costs roughly $5–6 per year.

Now compare that with the infrastructure that actually uses it.

Cloud servers range widely in cost depending on workload:
• Basic VPS: $5–$20 per month
• Typical production servers: $50–$200 per month
• Enterprise or large workloads: $500–$2,000+ per month 

A realistic operational midpoint for many services is therefore $100–$500 per month.

Against that cost structure, the IPv4 address required to run the machine costs well under 1% of the system.

Something that represents a rounding error in the infrastructure stack cannot credibly be described as the limiting resource of the Internet.

To understand how absurd the scarcity narrative becomes under economic inspection, consider a simple analogy.

If renting an apartment in the center of a major city cost 0.1% of a person’s salary, no economist would describe housing as scarce. Housing would be considered abundant relative to income.

IPv4 is in exactly that position.

The price signal does not support the claim of shortage.

Yet the narrative of scarcity persists.

Why?

Because it serves an institutional purpose.

By framing IPv4 as a scarce and exhausted resource, registry institutions can justify extraordinary administrative control over the registration layer. They present themselves as guardians of a fragile global resource rather than administrators of a ledger that increasingly supports a functioning market.

This narrative also conveniently obscures a deeper asymmetry.

Many registry contracts limit institutional liability to symbolic levels—sometimes around $100. That means institutions exercising administrative authority over infrastructure worth billions collectively assume almost no financial responsibility if discretionary actions cause harm.

The scarcity story helps sustain this arrangement.

If IPv4 is portrayed as fragile and finite, then centralized administrative control appears necessary, and legal accountability becomes secondary.

But once IPv4 addresses are recognized for what they have become—capital infrastructure assets traded and leased in global markets—that logic weakens.

Markets allocate scarce capital.
Administrators maintain ledgers.

When administrators exercise discretionary authority over capital assets while assuming almost no liability, the scarcity narrative becomes less an economic observation and more an institutional defense.

This framing has another consequence.

If IPv4 is not truly scarce in economic terms, then the central justification for pushing IPv6 weakens dramatically.

For decades, IPv6 has been promoted as the inevitable replacement because the Internet supposedly “ran out” of addresses. Yet the actual operational Internet continues to run primarily on IPv4 while markets supply addresses at trivial cost.

Operators are therefore required to maintain dual-stack infrastructure, running both IPv4 and IPv6 simultaneously. This creates real operational overhead:
• additional routing complexity
• additional firewall configuration
• additional troubleshooting
• additional engineering effort

In practice, the Internet has been paying a dual-stack tax for years.

If IPv4 scarcity is largely administrative rather than economic, then the justification for this tax becomes questionable.

The technical comparison is also less dramatic than often presented.

IPv4 uses a 32-bit address space. With technologies such as Network Address Translation (NAT), one public IPv4 address can serve thousands of internal devices, effectively multiplying usable addressing capacity beyond the raw 32-bit space.

Meanwhile IPv6 offers a much larger theoretical space, but real deployments commonly allocate networks in /48 or /56 blocks—far larger than most networks practically require. In everyday operational terms, the usable addressing capacity difference often becomes less consequential than the raw bit counts suggest.

The result is that IPv6’s primary advantage—address abundance—loses much of its practical force when IPv4 addresses remain cheap, tradable, and widely usable.

The scarcity narrative therefore begins to look less like a technical necessity and more like institutional propaganda.

If IPv4 scarcity disappears as a constraint, then the registry layer is no longer about rationing rare numbers.

It becomes what it always should have been:

A neutral ledger for infrastructure assets used by networks.

And once that reality becomes clear, the fundamental governance question emerges:

Should the Internet’s registration layer be governed by discretionary institutions with minimal liability?

Or by transparent, rule-bound systems aligned with the economic reality of the networks that depend on it?

When examined through price, engineering, and operational economics, the scarcity narrative becomes increasingly difficult to sustain.

Scarcity that does not constrain usage is not scarcity.

It is narrative.

And narratives eventually yield to mathematics.