I’ll take that bet

Former CEO of ITXC and now high-priced consultant Tom Evslin says There Won’t Be Any Landlines in 2013.

I love bold predictions. It’s nice to see people put their neck out. In 1999 I proclaimed that light bulbs would have IP addresses by 2005. I think I lost that bet. And here’s why Tom is wrong on this one:

In a perfect world, there would already be no copper landlines. If telcos used normal math and logic, they would have pulled them years ago. The problem is, that’s not how telcos work. 2013 is just too soon. There are billions of landlines already operating today. Believe it or not, at the rate copper decays, many miles of NEW COPPER are laid into the ground every year. That sounds ridiculous to any thinking person, right? It would be way cheaper to replace that copper with fiber. But here’s why they don’t do it: supporting infrastructure costs. They have a bunch of guys in trucks already trained and loaded with spare parts and a massive support infrastructure already in place that knows how to maintain the copper. They don’t want to spend the money to set up all that infrastructure, train their staff, stock spare parts, testing, troubleshooting etc. for a new technology, even if it is cheaper to deploy and maintain. They will have to bite the bullet sometime, of course, but this will not be complete by 2013.

Telcos take a long time to do anything. The very fact that a project to replace or phase out the copper plant is not already on the books, to say nothing of already underway, means there is no way it could be completed by 2013.

People are also slow to change. Tom himself notes that “old technologies have a long tail. Mainframe computers are still around decades after people (including me) predicted their demise.” The copper plant will be no different. Even if 80% of the copper plant is gone by 2013, which I think is overly optimistic and far exceeds all analyst estimates, it will be many many decades before that last 20% goes away. Consider that here in 2007 there are still about as many people using leased rotary phones as there are people using Vonage.

Tom says people will switch to VoIP and that everyone will have “wireless hotspots in our homes by one means or another” by 2012. The question is, will those hotspots support VoIP, particularly third-party VoIP? Or the more basic question is, will there really be a compelling offer of some kind, whether VoIP, or mobile, or something else, that induces people to switch? The fact is, despite all the VoIP hype, penetration is still minuscule and it is not growing all that fast. Even the most aggressive estimates give VoIP less than 10 percent of all landlines in the U.S. today, and over half of these are second lines, or VoIP being used along with a traditional copper pair landline, not instead of. These aggressive estimates say about 1/3 of all lines will be VoIP by 2010 [1] — others say fixed line penetration will still be 74% in 2010 [2]. Either way, that leaves the clear majority of customers on good old copper lines, certainly enough to justify (in telco’s minds) maintaining the plant that supports those customers.

And then there’s the dirty little secret in the “copper plant goes away” logic. Even if everyone switches to VoIP, doing so requires broadband — and the most common broadband deployed today worldwide is, you guessed it, DSL over copper. So this means even if people switch to VoIP (or whatever), their broadband service would still have to be moved off the copper plant too, again, something that is just not going to happen universally by 2013.

So, in short, even if killing it is the sane thing to do, it won’t happen, and the copper plant will still exist in 2013 and well beyond. As ridiculkous as it sounds, it may even still be dominant in 2013.