The US government has spent about $100 billion on rural telecommunications in the last 20 years. (That figure doesn’t include the billions of dollars in private investment and state subsidies.) It doesn’t feel like it in many rural areas.
The lion’s share of rural telecom subsidies come from the FCC’s “high-cost” fund, which is part of the Universal Service Fund. The high-cost fund currently disburses about $4.5 billion per year to rural carriers and large carriers serving rural areas.
Excess in the high-cost program
Bill drafters in Congress and the CBO, after the passage of the 1996 Telecom Act creating the Fund, expected the USF program subsidies to decrease over time. That hasn’t happened. The high-cost fund has increased from $800 million in 1997 to $4.5 billion today.
The GAO and independent scholars find evidence of waste in the rural fund, which traditionally funded rural telephone (voice) service. For instance, former FCC chief economist Prof. Tom Hazlett and Scott Wallsten estimate that “each additional household is added to voice networks at an annual USF cost of about $25,000.” There are at least seven high-cost programs and each has its own complex nomenclature and disbursement mechanisms.
These programs violate many best practices for public finance. Shelanski and Hausman point out, for instance, that a huge distortion for decades has been US regulators’ choice to tax (demand-elastic) long-distance phone services to fund the (demand-inelastic) local phone services. The rural fund disbursement mechanisms also tempt providers to overinvest in goldplated services or, alternatively, inflate operational costs. Wallsten found that about 59 cent for every dollar of rural subsidy goes to carriers’ overhead.
To that end, the high-cost program appears to be supporting fewer households despite the program’s increasing costs. I found in Montana, for instance, that from 1999 to 2009 subsidies to carriers rose 40 percent even while the number of subsidized rural lines fell 30 percent. The FCC’s administrative costs for the four USF programs also seem high. According to the FCC’s most recent report, administrative costs are about $172 million annually, which is more than what 45 states received in high-cost funds in 2016.
A proposal: give consumers tech vouchers
A much more transparent and, I suspect, more effective way of satisfying Congress’ requirement that rural customers have “reasonably comparable” rates to urban customers’s rates for telecom services is to give “tech vouchers.” Vouchers are used in housing, heating, and food purchases in the US, and the UK is using them for rural broadband.
My colleague Trace Mitchell and I are using Census and FCC data to calculate about how much rural households could receive if the program were voucher-ized. Assuming all high-cost funds disbursed to states in 2016 were converted into broadband vouchers, these are our estimates.
If vouchers were distributed equally among rural households today, every rural household in the US (about 20% of US households) would receive about $15 per month to spend on the broadband provider and service of their choice. Low-income rural households could tack on the $9.25 USF Lifeline subsidy and any state subsidies they’re eligible for.
Perfect equality probably isn’t the best way to subsidize rural broadband. The cost of rural service is driven primarily by the housing density, and providing telecom to a rural household in the American West and Great Plains is typically more expensive than providing telecom to a rural household in the denser Northeast, and this is borne out in the FCC’s current high-cost disbursements. For instance, Vermont and Idaho have about the same number of rural households but rural carriers in Idaho receive about 2x as much as rural carriers in Vermont.
However, some disparities are hard to explain. For example, despite South Carolina’s flatter geography than and similar rural population as North Carolina, North Carolina carriers receive, on a per-household basis, only about 40% what South Carolina carriers receive. Alabama and Mississippi have similar geographies and rural populations but Alabama carriers receive only about 20% of what Mississippi carriers receive.
A tiered system of telecom vouchers smooths the disparities, empowers consumers, and simplifies the program. We’ve sorted the states into six tiers based on how much the state received on a per-household basis in 2016. This ranking puts large, Western states in the top tier and denser, Northeastern states in the bottom tier.
In our plan, every rural household in five hardest-to-serve Tier 1 states (Alaska, Kansas, Montana, North Dakota, and South Dakota) would receive a $45 monthly discount on the Internet service of their choice, whether DSL, cable, fixed wireless, LTE, or satellite. As they do in the UK, eligible rural households would enter a coupon code when they receive their telecom services bill and the carrier would reduce the price of service accordingly.
Similarly, every rural household in:
Tier 2 states (ten states) would receive a $30 monthly discount.
Tier 3 states (ten states) would receive a $19 monthly discount.
Tier 4 states (ten states) would receive a $13 monthly discount.
Tier 5 states (ten states) would receive a $6 monthly discount.
Tier 6 states (five states) would receive a $3 monthly discount.
$3 per month per rural household doesn’t sound like much but, for each of these states (Connecticut, Delaware, Massachusetts, New Jersey, Rhode Island), this is more than the state currently receives in rural funds. In Connecticut, for instance, the current high-cost funding amounts to about 25 cents per rural household per month.
Under this (tentative) scheme, the US government would actually save $25 million per year from the current disbursements. And these are conservative numbers since they assume 100% participation from every rural household in the US. It’s hard to know what participation would look like but consider Lifeline, which is essentially a phone and broadband voucher program for low-income households. At $9.25 per month, 28% of those eligible for Lifeline participate. This is just a starting point and needs more analysis (see link below for spreadsheet), but it seems conceivable that the FCC could increase the rural voucher amounts above, expect 50% participation, and still save the program money.
Conclusion
As Jerry Hausman and Howard Shelanski have said, “It is well established that targeted subsidies paid from general income tax revenues are often the most efficient way to fund specific activities.” Current law doesn’t allow allow for tech vouchers from general income taxes, but the FCC could allow states to convert their current high-cost funds into tech vouchers for rural households. Vouchers would be more tech-neutral, less costly to administer, and, I suspect, more effective and popular.
Excel spreadsheet of tech vouchers by state (Dropbox): link.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.