Global Consequences of Rushing IPv6 Deployment

IPv6 evangelists and Internet governance institutions have been urgently calling for immediate transition to IPv6 in light of the near global depletion of the IPv4 address free pool.

But companies are not rushing.  Instead, more and more are turning to the IPv4 market to purchase the IPv4 addresses they need to operate and grow their IP-based networks.   Why? Transitioning to IPv6 is expensive, and at the moment, there is little to be gained in return for the investment.

The Organisation for Economic Cooperation and Development (OECD) released a report, The Economics of Transition to Internet Protocol Version 6, this past November that sheds light on why global IPv6 deployment has been so slow. Applying the “probit model” of technology adoption, the report notes that large content providers, such as Google, Facebook and Yahoo, provide IPv6 content, and large ISPs, including Comcast, Verizon Wireless, DT and KDDI, have substantial IPv6 deployments. These early adopters have not influenced smaller providers to move toward IPv6 because the value proposition is either too costly or too remote, or both.

The report compares IPv6 migration to other examples of technology adoption and finds one key difference. For many technologies, the value of adoption increases with supply (e.g., CDs, VCRs and gaming systems). For IPv6, the value of adoption increases only after most of the IP-networks, content providers and end users transition. The OECD observes that achieving the critical network effect required to stimulate significant migration to IPv6 is more challenging than other technology adoption scenarios “because . . . there are few influential users to help facilitate transition to the new platform.”   Until there is wide-spread IPv6 adoption, switching to native IPv6 is not feasible, and dual stacking is a necessary intermediate step.   That intermediate step adds complexity, cost, and possible network security and performance problems.   In short, not only is the cost of migrating to IPv6 significant, but the value of migration will only be realized at some indeterminate time in the future after many others decide to migrate.

The OECD report exposes the powerful financial disincentives facing many network operators deciding whether and when to migrate to IPv6.  Adding artificial, regionally-specific IP address scarcity in an attempt to alter this decision calculus is bad policy. And it won’t work as long as there is remaining supply of IPv4 numbers in the private market. Companies will find ways to extend the life of IPv4 without regard to Internet governance policies.

Missing from this discussion, however, are the potentially devastating global implications of a hasty region-by-region v6 migration driven by artificial IPv4 address scarcity. Even if policy-based IPv4 address starvation causes network operators in North America, for example, to quickly migrate to IPv6, it will also shift and dramatically increase the burdens of migration to network operators in regions like Africa, with considerably more free supply. African network operators will be forced to make even tougher choices: spend valuable capital adopting IPv6 to stay connected to the rest of the world while free IPv4 numbers are still available, or continue to sweat their IPv4 infrastructure risking regional isolation on the legacy IPv4 Internet. If the transition from v4 to v6 in the most developed economies leaves behind -- on regionally isolated legacy v4 networks -- countries still struggling to bring developed-world communications services to their communities, the digital divide is likely to widen, with adverse global economic development consequences.

IPv6 is the long-term solution to the eventual exhaustion of IPv4. With considerable remaining IPv4 supplies available in the private market (estimates range from 800 million to 1 billion unused previously allocated IPv4 numbers), the life of IPv4 can be extended so network operators plan IPv6 migration strategies based on their specific technology and business objectives, not artificially imposed scarcity. The OECD recognizes the role the market can play in facilitating an effective IPv6 migration, and questions why the IPv4 transfer market has not more fully developed under the circumstances.   The report posits several possible reasons, including the lack of price transparency, the continuing availability of free numbers, and finally, “institutional opposition to paid transfers” even though “markets for IPv4 addresses can play a role in allocation addresses from low value uses to higher value uses.”

The free pool of IPv4 numbers is nearly depleted with the exception of Africa, IPv6 migration while moving forward is moving too slowly to address exhaustion in the near term, and there are no good alternatives. Institutional opposition to open and transparent IPv4 markets only results in creating an environment of scarcity where none need exist and comes at a heavy cost to the business community and ultimately the public at large. RIRs and other Internet governance institutions should, instead, adopt policies that increase market efficiency and transparency.

The IPv4 market, far from being an obstacle to IPv6 migration, offers network operators the flexibility they need to execute their individual transition plans in line with normal technology adoption patterns. It will also allow for global supplies of available IPv4 numbers (both free and available for sale) to re-balance -- reducing the likelihood that IPv6 migration will isolate countries in developing regions from the rest of the global digital economy.

x