Commentary

Yep, the Broadband Stimulus Funded Overbuilds

Cynical readers won’t be surprised, but it turns out funds from the American Recovery and Reinvestment Act (ARRA), commonly known as the $787-billion stimulus enacted in 2009, financed public broadband buildouts in areas already served by commercial service providers.

Under the original plan, the $7.2 billion of the stimulus allocated for broadband was supposed to be directed to areas with no high-speed Internet. It turns out that the Rural Utilities Service, one of the agencies charged with funds distribution, dispensed $231.7 million to three areas where broadband penetration averaged 85 percent, according to a study by Jeffrey Eisenach and Kevin Caves of Navigant Economics of Washington, D.C. Eisenach is also an adjunct professor at George Mason University School of Law. The study was funded by the National Cable Television Association and can be downloaded here.

The report examines three large RUS stimulus awards:
* $101.2 million in western Kansas
* $66.4 million for Lake and St. Louis counties in northeastern Minnesota
* $64.1 million to cover a portion of Gallatin County in southwest Montana.

From the report:

Reports by the Department of Agriculture’s Inspector General (AIG) and the Government Accountability Office (GAO) have shown that RUS’ prior broadband subsidy programs have not been cost effective, in part because they have provided duplicative service to areas that were already served by existing providers, potentially creating “an uneven playing field for providers already operating without government subsidies.”

Among the key findings are the following:

* Definitions used by the RUS to determine where grants should be awarded permitted subsidies to areas which were already served by multiple wireline and wireless providers. Of the three projects analyzed, more than 85 percent of households were already passed by existing broadband providers, and in one project area, more than 98 percent of households were already passed by at least one provider.

* Based on the cost of the direct grants and subsidizing the loans, the study estimates that the cost per incremental home passed will be $30,104 if existing coverage by mobile broadband providers is ignored, and $349,234 if mobile broadband coverage is taken into account.

* The RUS approach of funding duplicative coverage is directly at odds with the National Broadband Plan recommendations and would massively increase the cost of extending broadband to all unserved homes. The FCC’s Omnibus Broadband Initiative estimated that the cost of extending broadband to every unserved household in the U.S. is approximately $23.5 billion, so long as duplicative service is not funded. But funding duplicative service (as RUS has done under BIP) increases the cost of a nationwide buildout by $63.7 billion, to $87.2 billion.

“While it may be too early for a comprehensive assessment of the ARRA’s broadband programs, it is not too early to conclude that, at least in some cases, millions of dollars in grants and loans have been made in areas where a significant majority of households already have broadband coverage, and the costs per incremental home passed are therefore far higher than existing evidence suggests should be necessary,” the study says.

The report goes on to layout some cost benchmarks and identify ways broadband build-out, even if government-funded, could be accomplished at far less cost. One problem with the stimulus—noted by across party lines—was that the timetable called for all funds to be distributed before an adequate map of broadband-deficient areas could be mapped.

Also, at the time of the stimulus, there were prevailing policy winds in favor of subsidizing competition. The report examines this approach and concludes it does not generate sufficient benefits.

Finally, one the projects Navigant examined appears to be a bailout of a municipal broadband-type system.

For Lake County Fiber Network, a taxpayer-funded fiber-to-the-home project in Lake and St. Louis county, Minn., the RUS funds were both timely and ironic. Eisenach and Caves cite news reports that LCFN had to shift a $56 million RUS debt to local taxpayers. They also point to articles that in which LCFN disclosed that it had lowered its projections of average revenue per household to between $100 and $125 per month, below the minimum $133 per month it said it needs to break even.

The municipal broadband debacles of the past decade have been well-chronicled, here and by others. Under the guidelines, the RUS was supposed to examine an operation’s long-term viability before releasing funds. That rule, along with a few others, it seems was ignored.