IPv6 Should Be Invisible… To Most
By Brent Rowe
Research Economist, RTI International

IPv6 is coming, but at this point the transition is still moving rather slowly. Many potential users don't have the motivation to move, since presently there appear to be no easily-demonstrated quantifiable benefits to do so, and many of the touted benefits appear to be available without IPv6. Still, the efforts of the U.S. Department of Defense (DoD) and the Office of Management and Budget (OMB) are certainly moving things forward much more quickly than the demand from corporate users. Even though the benefits are still difficult to forecast for most users, the DoD has a vision into which IPv6 fits firmly. And OMB's requirement that all government agencies move to IPv6, while possibly not resulting in benefits for government agencies equal to those the DoD foresees in the short-term, should result in benefits to all Internet users (public and private) by accelerating its adoption.

A study that RTI International conducted to look at the economic impact of IPv6 adoption concluded that, although many users may never know that they are using IPv6, the cost to the organizations for which they work, to the ISPs that provide their Internet connectivity, and to the vendors who supply their Internet hardware and software, will be sizable. We estimated that ISPs will have to spend $136 million, hardware and software "infrastructure" vendors $1.4 billion, applications vendors $593 million, and users $23.3 billion, all spread out over approximately 30 years.

However, on a per-employee basis, the cost picture looks very different. As compared to users, ISPs will incur a much higher cost on a per-employee basis by spending approximately $19,000 per IT staff member on training, installation, and maintenance. Users will incur a cost per IT staff member of approximately $10,500. To be clear, we did not estimate any costs to be borne by home users, who we assumed would not adopt IPv6 until it is rolled into their products — operating systems, routers, etc. — through their self-determined routine hardware upgrade needs, which will likely lag significantly behind corporate, institutional, and government users. Since infrastructure and applications vendors' total costs were calculated as a percentage of R&D spending, per employee costs cannot be accurately approximated.

There is unfortunately still a void of specific examples of easily demonstrable benefits. A "killer app" has not been developed, and many of the touted applications of IPv6 (e.g., IPTV, faster Internet access on phones, etc.) seem to be already appearing without the use of IPv6. Still, some potential users whose services or products require a large supply of IP addresses do see benefits from IPv6 adoption. For example, several ISPs we spoke to were planning to move to IPv6 very quickly, so that they could put IP addresses in all of their phones. But, average users are not willing to incur any cost increases at this point because they still don't see the benefits and they certainly see the costs.

The benefits cited often appear incremental and, in many cases, are tied to other changes (costs) which are needed, such as network restructuring. If NATs are removed, our interviews and discussions with many experts indicate that this would have a cost-savings impact on network administration costs. Of course, to remove NATs, more IP addresses would be needed; hence, IPv6 would solve the problem if, for example, all NATs were removed in the U.S. But, many IT security administrators do not want to remove their NATs, and around and around we go. The benefits — in this example, reduced network administration costs — won't be realized until some costs are incurred (e.g., network restructuring). And when these costs are incurred, users must give up a network feature that many view as beneficial, namely NATs. This tradeoff does not appeal to many who are unable to see or easily quantify the benefits of IPv6.

Still, even those experts who typically have negative things to say about IPv6 agree that the world will be better with IPv6. They simply don't think there is a huge rush, and as such, they do not believe that companies should spend resources to switch sooner rather than later. In the next three to four years, a major infrastructure component, Microsoft Windows Vista, will provide the large majority of Internet users with IPv6 "on by default." With this installed base, users could begin to benefit from the network effects of IPv6 usage — said another way, as more users start sending (and receiving) IPv6 packets, all IPv6 users will see greater cost savings and application benefits (e.g., end-to-end security, etc.).

The actions of OMB and DOD are pushing vendors and ISPs to offer IPv6 as product features and/or service offerings more quickly than they would have otherwise, causing an acceleration effect which will likely shift the timing of adoption of others — corporate, institutional, and government organizations, as well as home users — by several years. But, until the benefits are more clearly demonstrated, fewer perceived potential problems exist, and the costs (even simply labor) become much smaller, it is likely that many users will continue to drag their feet.