It was widely reported that Senator Daniel Inouye (D-HI) had some tough questions for the FCC last week saying that the FCC’s National Broadband Plan was a “vision that appears to be firmly rooted in the second tier of countries”, but there are a few problems with this.  For one thing, it’s based on widely cited but grossly exaggerated broadband speeds and prices of other nations.

Much of the bad data can be traced back to the Berkman Study which was widely discredited.  The Berkman study claimed that Japan’s broadband speeds in 2008 averaged 93 Mbps when actual real-world data from Akamai suggested that actual performance was in the 7 Mbps range.  The US numbers were less inflated to 10 Mbps when Akamai data showed it averaged 3.9 Mbps.  That means the actual differential between the US and Japan was around 1.79 but the Berkman study claimed it was a 9.3 fold difference which is nearly an order of magnitude difference.

The other problem with the Berkman study is its bogus pricing data which claimed that the top speed tiers of broadband service were only $32 when the actual verifiable pricing for high speed in Japan is over $76 per month.  It also conflates apartment service for single-unit family home pricing which is ridiculous since apartment broadband prices can be partially masked by the rent.  There are apartments in San Francisco that offer 100 Mbps service for around $30/month but it’s masked by the fact that a tiny apartment costs $2000/month in rent.  The actual “last mile” network infrastructure in an apartment is covered under facilities which in turn are covered by rent.  Single unit homes don’t mask their cost of broadband so they should not be compared to apartment subsidized broadband.  Furthermore, those cheap apartment broadband services are far more oversubscribed and they tend to fall far short of their advertised performance.

Another important point is that last-mile broadband speed is only part of the equation and server bandwidth costs are the other half.  Broadband speeds and prices have to be measured in conjunction with server bandwidth pricing and speeds to get an accurate gauge on Internet performance.  Server bandwidth in Japan costs many times more than in the US and that restricts the bandwidth their content and applications can run at.  The reality is that server bandwidth is the constraining factor because it can’t be shared like broadband bandwidth.  This is why applications have and always will lag broadband performance in the United States and elsewhere.

At a more fundamental level, most of the application innovation on the Internet has little to do with high bandwidth.  We just need to look at Twitter or Facebook or Google or Amazon or Ebay to see that innovation occurs on the low bandwidth end of the spectrum because that is what facilitates scalability.  It’s also an undeniable fact that the US leads the world in Internet innovation and it is where most of the leading dotcoms are born.

The other problem with the questions from the Senator is that the FCC doesn’t control the purse strings and they have to work within the budget they have.  The House and Senate barely approved $7.2 billion (not to mention all the poison pill strings attached that has scared away all the major operators) in total broadband stimulus funding which is barely enough scratch the surface.  Verizon alone spent many times that figure to cover a portion of their footprint.  For the FCC to build or spur the kind of fictitious broadband networks abroad that we keep hearing about, the FCC would almost need the power to print money.

We have to have realistic bars to measure against and the FCC seems to have put forth a realistic broadband plan.  We cannot afford to get into the business of competing against fantasy broadband rivalries.