A rural broadband grant is priced by its obligations, not its cheque

Few announcements please a regional broadband board more than a grant award. A state broadband office publishes a list, a press release names the counties, and a company that has argued for years that rural homes deserve better service finally has public money to build it. I have been in rooms where that list arrived, and I have felt the same satisfaction. I have also learnt to treat the award as the start of a valuation, not the end of one.

My view comes partly from where I started. At Quillan Valley Telephone Company in western Kansas, nearly half our revenue in the early 2000s came from support formulas and the charges other carriers paid us to complete calls. The money was real, but every rule change moved our results more than anything our managers did. Later, at Stavely Networks, I watched school circuits paid partly through the federal schools-and-libraries discount programme, where the paperwork was as much a part of the product as the fibre. Public money is never simply revenue. It is a contract with a government, and its terms deserve the same reading as a loan agreement.

The cheque is the smallest part of the award

A typical state grant for rural broadband pays a fixed sum towards the cost of reaching a defined set of locations. In return the company commits to several things, and each has a price.

It must fund a match, often a third or more of the budget in its application. It must finish construction by a deadline, usually measured in a few years, and may have to post a letter of credit or similar security that a bank will charge for and count against its credit lines. It must offer service at defined speeds, frequently including a low-cost plan for eligible households, and keep offering it for a stated number of years. It must report on progress and spending in a prescribed form, and it may have to repay some or all of the grant if it falls short. Above all, the grant is usually fixed while the cost is not. If construction runs over, the overrun belongs to the company.

None of these terms is unreasonable. Governments are right to ask for them, because some earlier subsidy programmes paid companies that later abandoned what they had promised. A federal reverse auction in 2020 produced many awards that winners later defaulted on, leaving the same homes waiting again. The terms exist to prevent a repeat. But they mean that a grant is an option the company writes to the government, and an option has a value to whoever holds it.

The federal programme that states now administer has added its own uncertainty, because its rules have been rewritten since the money was allocated. I do not think a company should design its plans around any particular version of those rules. It should design them around its own cost to build and its own revenue after building, and treat the grant as one source of funding among several.

Cost per location after the engineers have walked it

The weakest number in most grant applications is the construction cost. Applications are written quickly, against deadlines, using desktop designs and average costs. The real cost appears only after engineers have walked the routes, counted the poles, met the electric cooperative and found the creek crossings and rock that no map showed.

In 2023 Ridgeburn was awarded eleven project areas in two state grant rounds funded with federal recovery money, covering 40,400 rural locations. Our application had assumed an average construction cost a little under $7,000 a location. When our engineers walked the routes, eight areas held close to plan. Three did not. Across those three areas, covering 9,400 locations, the engineered cost came to $14,800 a location, because the homes were further apart than the mapping data showed and two cooperatives required pole replacements before attachment.

The grant for those three areas would have covered about $41 million. Our share of the engineered cost would have been about $98 million, for locations where we expected fewer than one household in three to buy service at prices the low-cost plan obligation would partly constrain. Even with optimistic take-up, the revenue would not have covered the network's upkeep, pole rents and storm repairs, let alone the capital.

Knowing when to hand an area back

I recommended that we accept the eight areas, covering 31,000 locations at an engineered cost of about $6,200 each, and hand back the other three before construction began. Our chief executive and our head of government affairs disagreed. They worried that returning awards would damage our standing with the state broadband office, that future scoring might count it against us, and that families in those areas would wait longer for service. All three concerns were legitimate, and the last one was the hardest to answer.

The board accepted my recommendation, and we told the state office within weeks of completing the engineering, explaining the cost and offering our design data to whichever provider might bid next. That timing mattered. Handing an area back before construction, with evidence, is different from abandoning it halfway. The state relisted the areas, and one has since been awarded to a wireless provider whose economics suit dispersed homes better than ours.

The eight areas we kept are being built. By December 2025, 14,000 of the 31,000 locations were complete, with take-up above 40 per cent where service had been available for a year. The grants cover about $118 million of an expected $192 million; our share, about $74 million, is funded within the fibre programme's delayed-draw facility.

For directors I would suggest five questions before any award is accepted. What is the engineered cost, not the application cost? Who bears an overrun? What does the low-cost plan obligation do to revenue per customer? What does the network cost to run once the grant is spent? And what is the company's plan, and its communication, if an area proves uneconomic? A board that asks those questions will accept fewer awards and complete more of them, which is the outcome a rural household actually needs.