The only thing that ties this post together is the Washington metro.  I was in DC the last few days and heard Michael Calabrese of the New America Foundation speak at a CITI conference.

Calabrese mentioned a study by Mark McHenry near Dupont Circle (that's the metro stop tie-in) that found that only about 20-40% of the spectrum under 3 GHz (that is, everything from AM radio to microwave devices) was in use at all.  There's a lot of warehousing going on by licensees.

Calabrese also said that there's a tremendous lag time for spectrum policy change.  With the Spectrum Policy Task Force report of 2002, there seemed to be an agreement that deregulation (unlicensing) of some spectrum was a good idea.  That report urged a hybrid approach, but we now seem to have less flexibility than we did before the report was written.  Yes, there are some exceptions, including in the "white spaces" proceeding, perhaps, but more on that later.

Calabrese noted that the FCC keeps taking a property/zoning approach -- "good fences make good neighbors."  What's scarce at this point is government permission for unlicensed uses, because the majority of licensed spectrum is underutilized.  Moreover, user demand and technical capabilities could allow reallocation of spectrum on a second-by-second, opportunistic, dynamic basis.  We could have cognitive radios certified to protect licensed services.  (Here are comments from Dell, Google, HP, Intel, Microsoft, and Philips pointing out that they've given the Commission a low-power prototype device that can sense when a particular channel is in use and that can politely, immediately, deftly vacate that channel.)

Calabrese called for the FCC to open up frequencies that haven't been assigned or aren't in use nationwide.  (There's a lot of this -- see Dupont Circle.)  He said that there are two objections that keep being made to the kind of dynamic frequency selection that he's talking about:  congestion and "letting the genie out of the bottle."  As to congestion, he noted that there's no evidence of "tragedy of the commons" caused by opportunistic spectrum use.  On the contrary -- wifi use has sparked enormous innovation.  We've got 300 million devices sharing what used to be thought of as "junk bands."  We've got 45 million home networks using wifi, and 60% of enterprises and 50% of higher-ed institutions are using it.  We've got rural WISPs using it, using clouds of connectivity to cover their customers.  At the same time, we're wasting a lot of licensed bandwidth that could be used much more efficiently.  If there turns out to be a congestion issue someday, we can use peak pricing to help.  We could be managing all of this much more usefully.

The "genie out of the bottle" complaint, Calabrese says, is the concern that once unlicensed devices move in to spectrum we won't be able to get rid of them.  But we could condition certification of smart radios on their ability to be upgraded remotely.  Technology can fix a lot of problems.

Calabrese was followed by Tom Hazlett, who has different views.  In a nutshell:  the most social value is created by markets that are facilitated by exclusive rights.  "Suboptimal valuation is the real tragedy."

The other DC metro station story I wanted to write about today was this one about Joshua Bell in the L'Enfant Plaza metro.  I hope you'll take a minute to stop and read the story.  It's prompted more than a thousand emails to its author, Gene Weingarten.  It's beautifully done.  I heard about it when I was playing viola quintets with people yesterday morning -- their view was that only about five out of a thousand people these days pay attention to classical music, so of course few people stopped to listen to Bell.  Take a look at the reader comments here.