In economics, someone who hoards resources to drive up the price is rather whimsically referred to as a troll. I was reminded of that recently when one of my colleagues suggested that perhaps the US government's interest in IPv6 is part of a plot to drive up the market price of IPv4 addresses.

See, the feds are the single largest owners of IPv4 address space -- and also the biggest proponent of moving to IPv6. My colleague's semifacetious theory is that by artificially tightening the market, the government can increase the value of the resources it owns. "Why wouldn't IPv4 space be sold off like unused spectrum or surplus computers?" he concludes.

He's got a point, even if calling the feds address trolls seems, well, a bit much. Lately, I've been getting a lot of requests from enterprises seeking help in smoothing their transitions to IPv6. Yet when I ask them why they want to move to v6, the answer boils down to, "We don't want to. We have to." There's a growing sense (accurate or not) that we're running out of v4 address spaces, and migration to v6 is a painful necessity. But that raises more questions than it answers, starting with the reality (or otherwise) of address exhaustion.

The reality is that thanks to rfc1918, which permits enterprises to operate their own private class A addresses, no enterprise will run out of private address spaces (so long as it's willing to implement network address translation, or NAT). Because the vast majority of enterprises I work with implement NAT as a security measure anyway, this is pretty much a nonissue.

That raises the question of exactly how necessary migration to v6 truly is. If you're a carrier in one of the countries that got shortchanged during the initial address allocations, you don't have a choice. Ditto carriers that serve the Feds (such as Verizon, which recently announced v6 capability). Everybody else does have a choice, however -- and there's the rub.

The only problem IPv6 indisputably fixes is address exhaustion -- which is a problem for carriers, not enterprises. All the other so-called benefits of v6 (integrated security and QoS) come at a cost. As I've discussed previously, v6 consumes considerably more bandwidth than v4 (because of its larger address space), which is an issue over low-speed links, such as wireless. And it introduces problems with multihoming, VoIP and a range of other applications. Essentially, it's a "new and improved" model that offers no clear value proposition to anyone who's already in possession of the tried-and-true version.

What happens in a free market when a manufacturer tries to promote an inferior product? Easy: The market value of the older version skyrockets. Remember New Coke? If eBay had been around back then, Coke Classic would have been selling at US$5 a bottle.

Which brings me back to my colleague's contention: There's no better way to drive up the market value of a resource than to create artificial scarcity. Hmmm. Calling the Feds "address trolls" isn't sounding quite so far-fetched after all.