^{Democracy & Technology Blog} $350-$700 billion for broadband

The FCC staff estimtates it would cost $350 billion to achieve universal access to the fastest broadband speeds, or $700 billion to build ubiquitous competing networks.
One way to raise this kind of money — without requiring a vote of Congress to raise taxes — is to impose a user fee on broadband subscribers.
Assuming you are a politician and you don’t want to be blamed for raising taxes or for increasing broadband rates in an election year, this plan would have to be paid for by Chinese and other foreign investors buying Treasury bonds which would increase our national debt.
But what if a massive investment in the fastest broadband speeds of today turns out to be an investment in the wrong technology?
Today fiber delivers the fastest broadband speeds.
But what if wireless becomes the dominant form of broadband? Wireless is cheaper than fiber to deploy, especially in rural areas. The Economist reports,

Within India, even the most remote areas are now judged to be on the verge of commercial viability, judging by the results of two auctions held in 2007. In each case bidders had to say how much government subsidy they would require to expand into rural areas, with the contract going to the lowest bidder.
In the first auction, for the right to build shared towers in 8,000 rural locations, the average subsidy requested was 35%, much less than expected. In the second auction, for the right to offer mobile services, many operators submitted zero bids or even negative ones–in effect offering to pay for the right to set up in rural areas. “The subsidies required are not as big as everyone thought, because the companies believe there’s a business case in being present in rural areas first,” says Mr Bajaj. In part this reflects the cut-throat competition in the Indian market. But it also shows that mandated tower-sharing can make the economics far more attractive for operators in rural areas, which could be a valuable lesson for other countries. A second round of rural expansion, with another 12,000 shared towers, has been announced.

I suspect the Economist has got it wrong about mandated tower sharing. The “newspaper” notes that voluntary, market-led tower sharing is already common in India, while it is mandated in China.
If producers can make money wholesaling, they will do it. If they will lose money — as a result of regulated prices or terms — they won’t. It’s only if they won’t make money that it is necessary to mandate it. But if they won’t make money, it doesn’t make economic sense– even if an Act of Congress requires it.
Why should government mandate something that doesn’t make economic sense?
If government does embrace a particular technology, it could be embarrassed if that technology becomes obsolete earlier than expected. France invested in the Minitel; China in the TD-SCDMA wireless technology.
Both gambles became jokes when overtaken by technology before they generated a sufficient return on significant political investment.