Solutions to BEAD Funding Allocations – Approaches to Maximize Funds

By: Mark Kennet, Senior Consultant on behalf of CostQuest Associates


In this paper, we describe the BEAD funding mechanism and its assignment to States and propose an economically theoretically sound mechanism for utilizing that funding to maximize the benefit to State residents in unserved and underserved areas.


The Broadband Equity, Access, and Deployment (BEAD) program has been established by the US government to expand high-speed internet access for all citizens in the United States, District of Columbia, Puerto Rico, US Virgin Islands, Guam, American Samoa, and the Northern Mariana Islands. Funded by the 2021 Infrastructure Law, BEAD is a federal grant program that aims to provide broadband connectivity to Americans by funding partnerships among participants to build infrastructure where it is needed and increase adoption of high-speed internet. BEAD prioritizes unserved locations that have no internet access or that only have access under 25/3 Mbps and underserved locations only have access under 100/20 Mbps. The amount of money allocated to states to achieve this goal is about $42 billion. This funding is distributed according to the number of unserved and underserved locations present in each state.

The FCC, through development of the National Broadband Map, determines served, unserved, and underserved locations. The question examined in this paper is how states should best allocate their BEAD funding toward solving the broadband connectivity problem. To that end, we consider some alternatives and assess their economic efficiency.

Alternative approaches

The simplest approach to distribute BEAD funding would be for the State to hold an auction and allow competitive bidding for the minimum subsidy required to serve each unserved or underserved location. The economic logic of such an approach is clear; bidders would, after studying the locations to be served, bid the least amount of money they would require providing a determined level of service. There is extensive economics literature on auctions in general and on minimum bid auctions in particular; some well-known books on the topic are these:

  • Putting Auction Theory to Work by Paul Milgrom1. This book is the standard reference on auctions and the first source of authoritative information about multi-unit auctions. It develops the main concepts of auction theory from scratch in a self-contained and theoretically rigorous manner. It explores auctions and competitive bidding as games of incomplete information through detailed examinations of themes central to auction theory.
  • Auction Theory by Vijay Krishna2. This book improves upon his 2002 bestseller with a new chapter on package and position auctions as well as end-of-chapter questions and chapter notes. It provides clear and concise proofs of all results on bidding strategies, efficiency, and revenue maximization. It also discusses auction-related subjects, such as private value auctions, the Revenue Equivalence Principle, auctions with interdependent values, the Revenue Ranking (Linkage) Principle, mechanism design with interdependent values, bidding rings, multiple object auctions, equilibrium, and efficiency with private values, and nonidentical objects.
  • Auctions: Theory and Practice by Paul Klemperer3. This book provides a non-technical introduction to auction theory and emphasizes its practical application. It covers both traditional topics such as the Revenue Equivalence Theorem, the Vickrey Auction, and the theory of optimal auctions; and modern topics such as combinatorial auctions, spectrum auctions, and package auctions. It also includes exercises, suggestions for further reading, and a glossary of key terms.
  • A Course in Game Theory by Martin J. Osborne and Ariel Rubinstein4. This book contains a concise introduction to the main ideas and techniques of auction theory. It covers both single-object and multi-object auctions, with an emphasis on the connections between different models and the role of assumptions. It also includes exercises with hints or partial solutions.

Essentially, all these books and their associated research articles point to the fact that auctions extract information about the bidder’s willingness to perform the service and the amount of money the bidder will accept. The major conditions required for such mechanisms to work are the presence of multiple bidders, prevention of collusion between those bidders, and financial/technical capacity to perform the work required. Typical auction rules address these, and the track record of auctions used around the world to foment infrastructure deployments in unserved and underserved areas is substantial (See, for example, Wolfstetter (2022).5

The design of such minimum-subsidy auctions can be complex, since more traditional auctions are simply selling a single commodity to the highest bidder, while these service auctions are attempting to extract the best price (least cost) subsidy for providing a basket of services. The complexity of these sorts of auctions is discussed in the next section.

One alternative approach to a minimum-subsidy auction might be for States to utilize the location maps available and apply the cost estimates calculated for each site; and then simply let contracts to perform the service at some percentage of that defined price.

A third approach might be for a State to directly fund the construction of infrastructure using internal resources.

Finally, a fourth approach to distributing BEAD funding is to create Public-Private Partnerships (PPPs) to address the gaps, wherein subsidies are paid to organizations formed for the purpose of expanding broadband coverage. In the United States, the Tennessee Valley Authority is an example of this, though it is not strictly speaking a PPP – it is a private sector entity that was first established by the federal government that collects and receives no tax funding. Some rural cooperatives may be considered a sort of PPP in that they receive tax benefits though they are operated outside of the public sector.

Each of these approaches has costs and benefits that must be weighed by the authorities. For example, the direct use of maps with contracts may result in an unwillingness by providers to service some areas, since very high-cost areas may be unprofitable; while the use of auctions may result in a State’s allocation of BEAD funding to be insufficient to provide 100% coverage, since the willingness of firms to provide access may depend on a price higher than that modeled. The latter problem can be addressed as has been in RDOF, by specifying in advance a maximum budget. Symmetrically, States could overpay using the direct contracting approach, since model estimates are only estimates; and conceivably some projects may in reality cost less than the estimates.

Auction Complexity

As discussed earlier, minimum price auctions with multiple attribute services are more complex than simple one-price, one-good auctions. Such an auction must, either explicitly or implicitly, define a value for each attribute or product in the bidding package to be able to compare bids. For example, a State may be interested in obtaining broadband services to a group of remote, high-cost locations; bidders may find that the cost for providing that service is not uniformly distributed and may choose to bid a lower price but serve only those locations it thinks it can manage at that price; another firm may bid on all the sites but have a higher cost. Even more complications arise if the technology for providing the services is defined; more typically, States refrain from defining the specific technology but specify quality attributes for the services.

An example of the sorts of complexity that can arise is the process set up for bids in the Rural Digital Opportunity Fund (RDOF). Run by the FCC, the specification has combinations of upload/download speed and latency. Each combination gets a determined number of points to be able to compare bids offering different sorts of service. The point schema gives a preference to fiber connections but allows technological flexibility for those areas where fiber connections would be prohibitively expensive.

The FCC’s schema is as follows:

BEAD Funding - FCC internet speed performance tiers.

The performance tiers are factored into the process using the following correspondence:

BEAD Funding - The FCC's internet latency performance tiers.

Thus, a bidder can, in principle, offer a low latency option if the offer scores enough points to make the bid competitive with other firms.

Other nations doing rural broadband initiatives have done analogous things. Peru, for example, had an auction that specified one upload/download speed of 10GB, but added hotspots and tablets to be distributed in the service area as a dimension of the bid. Bids had to specify the number of areas to be served, the number of hotspots to be included, and the number of tablets to be distributed in each area.

The Role of the FCC National Broadband Map

The Federal Communications Commission (FCC) has released a set of highly detailed maps that show the locations of all served, unserved, and underserved Broadband Serviceable Locations (BSL) in the United States. Using this information, the National Telecommunications Infrastructure Administration (NTIA) allocated BEAD funding6.

Each State decides, with NTIA approval, how to utilize its BEAD funds to address the broadband shortfalls. As discussed above, one possibility is that a State simply offers to let a contract for each location, or collection of locations, at the specified cost level, thus applying the maps directly.

The likely preferred approach is that non-served locations are packaged into clusters, and providers bid for each cluster using the techniques mentioned earlier. The clustering process involves using mathematical cluster analysis to group unserved points geographically, into service areas. That clustering process can be done using simple geographical connection rules like roads or rights-of-way, thus avoiding some problems experienced in the past with early models such as some of those proposed to determine universal service funding before the adoption by the FCC of its Hybrid Cost Proxy Model in 1997, like having network infrastructure that seems to cross bodies of water or across mountains.7

If a State chooses to utilize the auction approach, the FCC maps and cost estimates may serve as a point of reference for starting bids or for determining the feasibility of bid offers.

States will also find the FCC maps highly useful in any of the other approaches outlined here.

Best practices internationally utilizing map data

The World Bank (WB) and Interamerican Development Bank (IDB), among other international organizations, have worked with a number of client states to help subsidize broadband connectivity. Two examples are the World Bank’s participation in Mexico’s Red Compartida and the IDB’s participation in Brazil’s plan to connect unserved and underserved areas.

In the case of Mexico, a World Bank team worked with the Government of Mexico (GOM) to determine the feasibility of the Red Compartida, a national broadband network that was intended to extend the reach of broadband services to the largely unserved rural and poor areas of the country. As of 2014, broadband penetration was as low as 44.6 connections per hundred people; and mobile penetration was only 87—in a region where poorer countries, like Peru, had already attained 100 or more mobile lines per hundred people8. By 2023, mobile teledensity had risen to 989 and broadband to 5410, suggesting an ameliorative effect resulting from the Red Compartida effort. In this case, the plan was to involve government and WB funding to deliver the national shared network, with private corporations connecting to provide end-user services in underserved areas.

Brazil’s project is still (as of this writing) in the formative stages. Brazil has used extensive geographical mapping data together with ‘crowdsourced’ data from Ookla showing where data transmission at various speeds is located within various parts of the country, including the lightly populated Amazon River region. IDB researchers have identified the approximate cost of providing broadband services to those locations and calculated the likely benefit to Gross Domestic Product based on studies in other countries, producing a schedule of costs and benefits to providing broadband services for all locations.11

How States can utilize map data to maximize the impact of BEAD funding

Clearly, there are a variety of methods that can be employed to obtain the maximum value of BEAD funding to States’ residents. If the objective is to maximize the economic well-being of the State, that suggests utilizing broadband deployment cost estimates to determine where the least costly unserved/underserved locations exist and applying resources to reach them first. This approach can be augmented with information regarding the willingness-to-pay for the types of customers found there if demand studies are available and applicable. That suggests that an auction-based approach is likely to be most successful.

An approach that emphasizes the equity of broadband access may sometimes imply either a different sort of auction or perhaps the PPP approach. The economic rationale of an equity strategy—a strategy built around the concept that all network clients are equally important—is that since connection is a network good12 it is worth extending the network to all possible and feasible users. One equity-based strategy is to combine ‘expensive’ unserved areas with ‘less expensive’ unserved areas to create service areas that have a feasible average cost of service, and then auction rights to operators. This appears to be the dominant approach thus far in State BEAD projects, and it relies heavily on cost-estimate data to create appropriate service areas.

The PPP approach may also be useful in some cases. In this approach, cost-estimate data may be used to identify the areas that need attention; and then the State would collaborate with a local provider to deliver the access. This approach may be used in combination with an auction approach.


  1. Milgram, Paul. Putting Auction Theory to Work. Cambridge University Press, 2004. ↩︎
  2. Krishna, Vijay. Auction Theory. Academic Press, 2009. ↩︎
  3. Klemperer, Paul. Auctions Theory and Practice. Princeton University Press, 2004. ↩︎
  4. Osborne, Martin J. and Rubenstein, Ariel. A Course in Game Theory. MIT Press, 1994. ↩︎
  5. WolfStetter, Elmar G., 2022. “Universal high-speed broadband provision: A simple auction approach,” Information Economics and Policy, Elsevier, vol. 60(C). ↩︎
  6. NTIA Announces Final Guidance for States to Develop Their BEAD Challenge Process | National Telecommunications and Information Administration. ↩︎
  7. Reference to ACAM documentation. ↩︎
  8. World Bank Group, “Providing Technical Inputs to the National Shared Wholesale Network ‘Red Compartida'”, Final Report May 2015, ↩︎
  9. Mexico Teledensity: Mobile, 1960 – 2023 | CEIC Data. ↩︎
  10. Fixed broadband penetration in Mexico 2022 | Statista. ↩︎
  11. Luis Guillermo Alarón López, Mauricio Ayala Rao, and Eduardo Marques da Costa Jacome’s, C2DB: Crowdsourcing to Identify Digital Gaps and to Estimate the Cost of Bridging those Gaps. Interamerican Development Bank, 2022. ↩︎
  12. Network goods are goods or services whose value does not only accrue to the network member, but to other existing members as new members are included. ↩︎

Continue Reading

Back to Top