CBO, however, scored the entire incentive auction for $20B
In other words, House Rs predict that the lost revenue from the guard bands is 98.95% of what the actual auction experts predicted would be the total auction revenue.
That is one heck of a guard band. Perhaps we should auction the guard bands and keep the rest of the reclaimed spectrum for unlicensed.
Inflating the revenue "lost" relies on two things -- (a) an unrealistic inflated MHz/POP value that CBO explicitly rejected, and (b) a complete lack of understanding as to how wireless providers value spectrum at auction.
If you solve for (a), House Rs insist that you lose $7 billion in revenue. Which they also insist is enough to fund FirstNet. It's nice to see that their ignorance is consistent.
the problem with the $7B is that it makes the dumbass assumption that a MHz/pop has some absolute value. This is like assuming that gold should be the basis of all currency because gold -- unlike anything else in the universe -- has some absolute value. Oh wait, these ARE the guys who think we should use gold as currency.
The issue here is that without guard bands, a licensee must either internalize the guard band to protect itself from its neighbor, or rely on heavier and more expensive shielding. Worse, because your neighbor also has "flexibility," there is no way to know in advance the comparative cost (e.g., how much better is it to rely on a true guard band, an internalized guard band, or better shielding) As a result, bidders discount for the uncertainty. By contrast, an actual guardband that is sufficient in size to allow you to get away with cheap shielding and decent out-of-band-emissions (OOBE) is much more valuable, compensating for the lost value of the auctioned spectrum.
Teh stupid maketh my head to hurt.