So you are the IT regulator for a country and you are convinced that the shortage of IPv4 address space represents a threat to the development of the Internet in your country and you want to do something about it. Being that as regulator you don’t really run the countries IP networks, what can you really do? I’ve heard many regulators in over 30 countries grapple with this problem. The purpose of this article is to think through some ideas to guide action on using (or not) regulation to drive IPv6 adoption.

The January–February 2011 edition of the Harvard Business Review has an interesting article by Michael Porter & Mark Kramer titled “Government Regulation and Shared Value” on page 74. They gave some broad prescriptions on what key elements of effective regulation and I want to analyse those prescriptions in the context of IPv6.

First off, deploying IPv6 is the only way to sustain the explosive growth of the Internet and connect the more than 4 billion people who are yet to be connected. Deploying IPv6 will enable the Internet to largely return to a true end-to-end network which is how the Internet was designed to be. The end-to-end networking model is responsible for the huge innovations that have made the Internet more valuable than traditional telephony networks. The associated innovations such as smart grids, Internet of Things, machine to machine communications, VoIP, IPTV etc. have the potential to spur both technology-driven economic progress and as communication, education and quality of life is improved by these technologies, general societal progress will follow. While I am not a fan of trickle down economics, I have seen first hand the difference that having high quality Internet access can make in the lives of even those in rural areas. What I’d like to advocate for is that governments have a deliberate policy to really take advantage of Information Technology for economic development and that specific strategies be implemented and monitored to actualise those policies. Regulation is a vehicle of government policy and if those policies are well-crafted, thoughtfully implemented and ruthlessly reviewed and adjusted for effectiveness, the results can be truly amazing.

Secondly, IPv6 deployment is not an end-user problem and should not be cast primarily as such. End-users (ok apart from the nerds like us who set up multiple subnet networks in their homes) really don’t care about IPv4 or IPv6—they only care about the services that run on those protocols. Network operators should just build networks that work and shield the user from the underbelly of IPv6. Thus in communicating to end users, the IPv6 adoption problem must not be framed in terms of IP addresses but in terms of applications that matter to the end users.

According to Porter & Kramer, effective regulation encourages shared value and focuses on connections between societal and economic progress. Such regulation has the following characteristics:

1. Clear social and measurable goals.
2. Set performance standards but do not prescribe the methods for achieving them.
3. Define phase-in periods for meeting standards. These must reflect investment of new-product cycle in the industry.
4. Deploy universal measurement and performance reporting systems, combined with efficient and timely reporting of results.

Let us now examine each in the context of IPv6 adoption.

1. Setting clear, social and measurable goals

This should be a no brainer. I advocate breaking down the goal(s) down to objectives along a set of related metrics that are indicative of the objective and that will be tracked throughout the period of transition. Accepting that particularly in the social sense, while not everything that counts can be measured, “that which doesn’t get measured, doesn’t get improved.” So if you want to improve it, measure it objectively—period!

The over-aching goal will essentially boil down to something akin to “We shall have a X% of this country’s network infrastructure and services fully IPv6-enabled by date MM.YYYYY, leading with services used and run by, for and behalf of the government”. These goals could be within the framework of a national IPv6 Strategy or Policy. It will be progressively enforced by a set of incentives, including limiting access to government contracts and infrastructure projects only to organisations that demonstrate commitment to IPv6 deployment.

2. Set performance standards but don’t prescribe the methods for achieving them

Dictating what to do (the goal) and how to do it (the means) hasn’t achieved much (apart from troves of demotivated employees) in the workplace, why should it be expected to be successful in setting of interdependent but mostly economically independent entities? I have dealt with this issue in several countries during conversations with managers of IT regulatory agencies who want to impose a “Deploy IPv6” imperative on all the networks in the country. When I tell them that that might not be a good idea and could backfire, they are shocked considering I had usually just spent three days teaching their engineers how to plan and deploy IPv6. I often point to how resistant network operators are to being told how to run their operations (especially by some “clueless” folks in government—their typical choice of words in quotes). If you want to see this resistance live, you only have to attend one of your RIR meetings (or check its policy mailing list) and notice the number of times operational network folks will say “I don’t want to be told how to run my network” in response to some proposed policy prescription that tries to dictate how some aspect of networks should be run.

IPv6 deployment is a means (actually the only sustainable one) to some loftier goals/objectives and regulators must look to and focus on those goals rather than being fixated on means. I would rather have them institute a national broadband policy that defines the standard for broadband as “a connection that is capable of hosting Internet services” in addition to standards for access speed. Such a policy prescription could set end-to-end capability as a requirement for networks built and operated in the country. The network operators will ultimately work it out for themselves that IPv6 is the best method to use achieve that prescription and if they could achieve that by other means and still maintain compatibility with other networks at a cost that is acceptable to them, then fine. By focuses on standards, the regulator can frame the IPv6 deployment problem as an exciting opportunity that brings value rather than problem that just costs money to fix (which is what prescribing a method will mostly achieve).

Of course operators of government networks can go ahead and prescribe and mandate IPv6, but that is not a regulatory problem, just a government IT policy and strategy one.

3. Define phase-in periods for meeting standards

Unless you are a regulator for a brand new country where you have the luxury of having to oversee the development of infrastructure from scratch, it is just not possible to ‘turn off IPv4 and turn on IPv6’. Phase-in periods are going to be more effective if they are aligned with new-product and or equipment refresh cycle in the industry. This gives organisations time to deploy and integrate IPv6 into their service offerings and processes in a way consistent with the economics of their business. Sadly, most organisations don’t see a traditional business case for IPv6 and thus will only deploy it by incrementally building the capability during infrastructure refresh cycles. Those cycles again will differ from country to country and the regulators must study what applies in their own economy. Of course regulators can use the right incentives to ensure that where the cycles are unreasonably long, the country doesn’t get left behind by the IPv6 adoption train—like banning importation of networking equipment that is not IPv6 capable (this by the way has a direct social good element particularly in developing countries—stop these countries being a dumping ground for legacy technology).

4. Deploy universal measurement and performance reporting systems, combined with efficient and timely reporting of results.

This is the system that enables the regulator track performance on the metrics they set in step (1) above. The more granular the data and reports are on progress or lack thereof, the easier it is to adapt whatever regulatory strategies have been implemented. Such a monitoring system would be an open website that tracks key IPv6 deployment metrics across various sectors of the economy: government, telecommunications providers, universities and major content providers. This platform can be a great tool to use in driving adoption—social pressure, naming and shaming on laggards but also celebrating front runners—all in a implicit and non-confrontational way (people generally don’t argue with facts without tarnishing their intellectual credibility).

As always, these are my personal opinions and do not necessarily represent the views and positions of any of my past or present employer. I have other blog posts on techxcellence.net and perfexcellence.org and am on twitter as @perfexcellent.