Lessons from RDOF for BEAD

EDITOR'S NOTE: A previous version of this article stated that Nextlink defaulted on “a bit more than half” of its RDOF award areas before authorization. This claim was erroneous, based on the author’s miscalculation from FCC data. In fact, pre-authorization defaults affected only a tiny fraction of Nextlink’s RDOF award footprint. We apologize for this error, and the below article has been corrected. 

Washington, DC (April 30, 2025) - On April 22, Broadband Breakfast reported that the National Telecommunications and Information Administration (NTIA) told all states they would have an extra 90 days to submit their Broadband Equity, Access, and Deployment (BEAD) Final Proposals. New guidance is expected in mid-May, and the extension gives states including West Virginia, Kansas, Pennsylvania, and Washington, whose Final Proposals were soon coming due, a bit of time to adapt to new guidance.

As the NTIA deliberates on a BEAD course correction, it would be wise to review the experience of a recent major federal broadband program, the Rural Digital Opportunity Fund (RDOF), administered by the Federal Communications Commission (FCC), which embodies some of the features that BEAD has been criticized for lacking.

In particular, RDOF was technology neutral, in the sense that competition among projects did not explicitly consider, as part of the scoring system, what technology the bidders for funding proposed to use. RDOF was strikingly successful in buying deployment commitments from broadband providers at unusually low prices. However, many of those promises fell through, leaving gaps in RDOF’s coverage footprint. Also, while RDOF proceeded from program launch to preliminary selection of awardees much faster than BEAD has done, the process then slowed down, and there is still little, if any, documentation that locations have been connected.

BEAD was designed a.) to be more federalist than RDOF, and b.) to ensure universal broadband coverage, which for RDOF had been more of a stretch goal. BEAD uses a scoring rubric rather than the kind of sophisticated reverse-auction procedure that the FCC used. Requiring state broadband offices to run mini-RDOF auctions was never really compatible with the IIJA statute, the BEAD Notice of Funding Opportunity (NOFO), their own capabilities, or the mandate to achieve universal broadband coverage, so the NTIA prescribed a different approach, using a scoring rubric and no iterative competitive bidding.

Comparing RDOF with BEAD, a few main lessons emerge:

• Universal coverage was a stretch goal for RDOF but a mandate for BEAD.
• RDOF vetted organizations before selecting preliminary awardees, but reviewed projects only afterwards for auction winners, whereas BEAD is reviewing both organizations and projects first.
• In contrast with the explicit fiber preference in BEAD, RDOF’s implementation of technology neutrality favored performance metrics (outputs) rather than technologies (inputs), but that led to controversy and defaults when some winners seemed to have overpromised on what their technologies could do.
• RDOF structured its auction around rigidly predefined project areas, creating a process that lacked the project area flexibility and deconfliction procedures that many states have implemented for BEAD.
• Robust RDOF participation and aggressive bidding, notably by SpaceX (Starlink), drove prices down and caused RDOF to underspend its budget by nearly $7 billion.
• RDOF’s execution was initially faster than BEAD, going from launch to awardee selection in less than a year and a half, but then it ran into long delays getting the chosen projects authorized.
• About one-third of RDOF coverage plans fell through before auction-winning internet service providers (ISPs) even got authorized to receive funding, and defaults have continued.
• RDOF is on track to deliver a lot of high-quality broadband deployment at a proportionally low cost to taxpayers, but state BEAD subgrantee selection in frontrunner states like Louisiana and Delaware has been cost-effective, too.

Universal coverage as a stretch goal for RDOF but a mandate for BEAD

When RDOF was launched with a Notice of Proposals Rulemaking (NPRM) in August 2019, broadband — then defined as internet service at speeds of 25 Mbps download/3 Mbps upload (25/3) rather than 100/20 — was widely recognized as “a necessity, not a luxury.” State legislators with rural constituencies felt the urgency of the issue. But its national salience was limited.

The FCC’s longstanding mission of “universal service,” originally for landline telephony, had been updated in the Telecommunications Act of 1996 to cover ill-defined “advanced telecommunications and information services,” but the agency had an awkward job of adapting to rapid technological change in a framework of stable statutes and limited congressional attention.

Nonetheless, through the Connect America Fund program, launched in 2011, the FCC had pivoted from obsolescent landline technology to a focus on expanding access to the internet. As bandwidth demand escalated, definitions of “broadband” kept getting updated, from 4/1 (2010) to 25/3 (2015) to 100/20 (2024), with a 10/1 standard also being used for some programs.

However, “carrier of last resort” obligations that had been established to ensure universal access to landline telephone service as early as 1934 were never updated to include broadband. The FCC lacked clear authority to update the mission of the longstanding “Universal Service Fund” to include universal broadband service.

And there wasn’t enough money. The RDOF program’s initially planned budget of $20.4 billion, based on project USF revenues collected from incumbent telecommunications carriers, fell far short of the $80 billion that the FCC had estimated in its 2017 Broadband Deployment Report would be the cost of deploying broadband to all unserved areas in the United States. (An estimate by Deloitte put the number even higher.)

In this context, RDOF aimed at universal broadband service as a stretch goal, yet it was designed in a way that made it likely to fall short of that goal, while still achieving substantial progress in a cost-effective way.

All census blocks lacking 25/3 broadband were eligible for RDOF Phase I. Unserved locations in partially served census blocks — not yet visible in FCC data at the time of the RDOF reverse auction in 2020, but already expected to become visible through the new Broadband Data Collection (BDC) process that the FCC was mandated, and funded, to implement by the Broadband DATA Act of March 2020 — were slated for eligibility in a future Phase II of RDOF. “Reserve prices” based on a cost model sought to direct more funds to places where deployment costs were higher. Auction rules would dial down a “clock percentage” of the reserve price until they found a level where the national market cleared.

But gaps could occur if bidders were lacking for an area, or if they got priced out during the auction. More subtly, the way that RDOF vetted bidders and the way it grouped census blocks into larger predefined project areas could — and, as it turned out, did — lead to coverage gaps.

Such gaps might have been dealt with in RDOF Phase II. But foreseeably leaving coverage gaps wasn’t really a design flaw in RDOF, because it never had either sufficient funding or a statutory mandate to achieve universal broadband coverage. Making progress cost-effectively was success.

The political environment when the IIJA passed and created the BEAD program was very different.

In 2020, the COVID-19 pandemic greatly intensified society’s dependence on the internet, as people suddenly needed broadband in order to meet basic needs while complying with lockdown orders. And although lockdowns soon retreated, the impetus the pandemic had given to incorporate broadband access into the social contract lingered on.

Reflecting this, the IIJA explicitly established “universal broadband service” as the explicit overriding objective of the new BEAD program. State Five-Year Action Plans, drafted with NTIA oversight, were required to “assess the amount of time it would take to build out universal broadband service.”

Then, state Initial Proposals were required to submit plans that prioritized “unserved service projects” to ensure universal access to 25/3 broadband access, followed by “underserved service projects” that would ensure universal access to 100/20 broadband.

States had to implement BEAD challenge processes to ensure that locations did not miss out on universal broadband access through errors in the broadband coverage data. Later, the IIJA prescribed that states’ BEAD Final Proposals should detail specific projects sufficient to achieve universal broadband coverage. The FCC, while given a secondary role, was also tasked to report back on how to “improve … effectiveness in achieving the universal service goals for broadband.”

And now there was, at least arguably, enough money. Although the BEAD program’s $42.45 billion budget was less than earlier cost estimates for universal broadband deployment, private matching capital and new technology made it plausible that virtually all the remaining gaps could be closed.

That said, the BEAD NOFO included hints that a predefined list of “reliable” technologies — end-to-end fiber, cable, licensed fixed wireless, and (anomalously, obsolescent) DSL — might need to be supplemented with “alternative” technologies like unlicensed fixed wireless and low-Earth orbit (LEO) satellite.

Still, where RDOF’s priority was to stretch a dollar, BEAD’s priority was to eliminate coverage gaps. Some of the differences between BEAD and RDOF result from this difference in objectives.

RDOF’s two-stage pre-auction and post-auction vetting

In contrast with BEAD, RDOF had two application stages: a short pre-auction application, and a longer post-auction application. The selection of awardees occurred in between.

Many states, though not all, also have two application stages: a.) a prequalification (also called “pre-application,” “letter of intent,” etc.) phase, where organizations get qualified, and b.) a project application phase, where network designs and budgets to serve specific areas get submitted for review and scoring. But subgrantee selection occurs only after states have both applications in hand, and they can evaluate both the organization and the project.

By vetting applicants before the reverse auction, but reserving detailed review of projects until afterward, RDOF both increased competition for funding, helping it to get better deals, and economized the time that both applicants and FCC staff would have had to spend in preparing and reviewing detailed materials about uncompetitive projects.

However, not all the post-auction applications were successful. The resulting “defaults” left coverage gaps, because RDOF didn’t have a way to re-open the bidding and circle back to other applicants whose projects might have been viable. This is one of the ways that RDOF pursued cost-effectiveness at the expense of universal coverage. 

The vetting of RDOF applicants both ex ante and ex post is also a reminder of the limits of technological neutrality. While it didn’t score projects based on their technology, the FCC didn’t want to fund failed projects, so it performed a technological gatekeeping role that sometimes involved the exercise of judgment.

RDOF was technology neutral, but designed to favor better network performance

RDOF’s technology neutrality did not preclude it from favoring better performance. Clearly, some broadband networks perform better than others, offering features such as higher bandwidth, and it’s not in the public interest to be blind to that fact. But rather than scoring project based on their technology — the “inputs” to network performance — RDOF scored them based on the key metrics of bandwidth (the “width” of the data pipe, measured in megabits per second) and latency (the “length” of the data pipe, measured in milliseconds)— the “outputs” of network performance.

Specifically, RDOF required bidders to declare whether they would offer:

• Gigabit low latency. Projects promising ultra-fast speeds of 1 Gbps download, 500 Mbps upload, and latency under 100 milliseconds (100 ms) got the highest preference, with a “weight,” meaning a penalty, of zero.
• Above baseline. Projects offering low latency but “above baseline” speeds of 100 Mbps download, 20 Mbps upload (100/20), rather than gigabit, got a penalty of 20%. That means that, to compete with an otherwise similar gigabit project for a given area, an “above baseline” bidder would have to accept subsidies that were less by 20% of the RDOF “reserve price.”
• Baseline. Projects offering low latency with “baseline” speeds of 25/3 got a 35% penalty.
• Minimum. Projects could bid even with speeds of just 10/1 but faced a penalty of 50%.
• High latency. Projects that would not offer latency of 100 ms or less could bid as high-latency projects with an additional penalty of 40% over and above whatever penalty was assigned because of their bandwidth. For example, above-baseline high-latency projects got a total penalty of 60%, baseline high-latency would get a total penalty of 75%, etc.

These “tiers” effectively favored some technologies over others. For example, most ISPs can easily meet the 100 ms latency standard, but geostationary orbit (GEO) satellite technology inherently cannot, because of the distance that data has to travel to the satellite and back for every interaction with the internet. The GEO satellite companies HughesNet and Viasat participated in the reverse auction, but because they could only offer high-latency service, they proved uncompetitive.

Other cases are less clear-cut. For example, after RDOF made gigabit awards to wireless ISPs, industry veteran Doug Dawson wrote about the “gigabit wireless controversy” in 2021:

"One of the big controversies in the RDOF auction was that the FCC allowed three of the top 10 grant winners to bid using gigabit wireless technology. This was Starry (Connect Everyone), Resound Networks, and Nextlink (AMG Technology). By bidding in the gigabit tier, these technologies were given the same technology and dollar weighting as somebody bidding to build fiber-to-the-premise. There was a big outcry from fiber providers that claim that these bidders gained an unfair advantage because the wireless technology will be unable to deliver gigabit speeds in rural areas."

Dawson proceeds to explore the technological feasibility of gigabit wireless, ultimately judging that wireless tech wasn’t up to meeting the gigabit standard even though it could provide “good” broadband. Of the three companies Dawson mentioned, Starry later “defaulted” on — did not get authorized for — all of its RDOF award areas. Nextlink and Resound Networks’ RDOF projects, minus a few census blocks, are on track at the time of writing, but whether they will ever stand up a capability for providing gigabit service to all residents of the RDOF territories is unknown.

FCC maps currently include a substantial footprint of gigabit coverage claims by wireless ISPs, as shown in Figure 1:

Figure 1: Gigabit coverage by fixed wireless in June 2024 (Source: FCC)

While this is encouraging for the viability of gigabit wireless, the FCC National Broadband Map’s gigabit speed standard, 1,000/100 Mbps, is somewhat slower on the upload side than the RDOF standard. Also, FCC maps use a “maximum advertised speed” methodology, which gives ISPs room to report speeds that are somewhat aspirational, and which their infrastructure couldn’t deliver to all residents at the same time. (Advertisements typically promise speeds “up to” specified bandwidths.)

If Resound Networks builds out to its RDOF footprint at speeds mildly below the 1 Gbps download/500 Mbps upload bandwidths required by RDOF’s “gigabit” tier, residents will probably be satisfied. But the company will still, arguably, have gotten an unfair advantage by overpromising in the auction. The pursuit of technology neutrality tends to give rise to such awkward edge cases.

Predefined project areas in RDOF and BEAD

One of the trickiest challenges states had to wrestle with during BEAD planning was whether, and how, to predefine project areas. RDOF dealt with that, too. Its solution was predefined project areas consisting of all the unserved census blocks — not the broadband serviceable locations (BSLs), since at the time no national BSL-level broadband coverage maps existed for use in program planning — in each census block group (CBG).

There are a lot of downsides to that. CBGs have nothing inherently to do with telecommunications networks and are in no way optimized for low-cost broadband deployment. It might make zero economic sense for all the unserved locations in one CBG to get served by one ISP. The FCC will likely need to deal with requests to descope specific locations or census blocks from RDOF projects as an alternative to larger defaults.

Geospatial planning is an area where BEAD has probed some fruitful new directions. The BEAD NOFO allows states either to predefine project areas or let ISPs define their own. Despite this, in 2024 the NTIA strongly favored predefined project areas, but it eventually let a few states, such as South and North Dakota, get approved for subgrantee selection using ISP-defined project areas.

Most states predefine areas for bidding, varying greatly in size from census blocks to whole counties, sometimes with ISP input, sometimes varying by round, etc. Whatever the “atoms” of geospatial planning may be, they raise the challenge of what to do when projects overlap. The BEAD NOFO coined the concept of “deconfliction,” and states have developed a variety of implementations.

RDOF, however, lacked the concept of deconfliction. Each CBG was a separate auction, and the only recourse for bidders on contiguous areas that didn’t win enough of their footprint to make the project viable was to default, as many did, resulting in coverage gaps, even if other projects might viably have covered the defaulted areas. While predefined project areas helped give structure to the RDOF reverse auction, they were not a cost-saving feature. On the contrary, predefining project areas can get in the way of allocating expansion territories among ISPs in the most cost-effective way.

RDOF reserve prices

Some areas are more expensive to deploy broadband to than others. Reflecting this, the FCC defined “reserve prices” for RDOF, relying on data from its longtime partner CostQuest Associates. These reserve prices played a key role in the reverse auction. Instead of dialing down the price per CBG or per BSL that the FCC would pay for deployment commitments, it dialed down a “clock percentage,” which was then applied to the reserve price to determine the support that was offered for each CBG in each round of the auction to gigabit low-latency projects. Projects in other performance tiers would be offered a support level based on a reduced percentage, corresponding to their disadvantage in promised performance, as described above.

RDOF reserve prices varied greatly. Some CBGs had reserve prices of $26,700 per BSL in 10-year support, which was the cap. The 90th percentile price per BSL was well below that, at $9,830 in 10-year support. That was almost four times the median of $2,544, which in turn was almost six times the 10th percentile level of $450. For a few CBGs, the reserve price was less than $1 in 10-year support.

RDOF awards could not exceed the reserve prices, and as it turned out, the bidding pushed awards below 70% of the reserve price in all cases. However, awards were not proportional to the reserve prices either, because some CBGs were assigned as soon as the national budget cleared. In other cases, multiple bidders were still in play and drove the price much lower. The fact that some CBGs got no bids, while the price was bid as low as 1% of the reserve price in others, suggests that the reserve prices were imperfectly correlated with actual deployment costs.

The RDOF auction, by the numbers

A total of 317 bidders participated in the RDOF auction. Of those, 227 bid in just one state, and most of the rest bid in less than 10 states. But the three satellite bidders — SpaceX, HughesNet, and Viasat — bid in almost every state. Other large footprint RDOF bidders included CenturyLink (34 states), CommNet Wireless (30 states), and Charter (25 states), and two consortiums, one comprising Sparklight and Wisper (25 states) and the other composed of many rural electric coops (23 states).

A substantial minority of RDOF bidders limited themselves to one to three CBGs. The vast majority of bidders — 243 out of 317 — offered gigabit low-latency projects. The next most common tier was “above baseline” (100/20) low-latency projects, offered by 71 bidders.

Overwhelmingly, however, RDOF funds were awarded either for gigabit low-latency service ($8.3 billion) or to SpaceX ($885 million) for the 100/20 low-latency service that it proposed to offer by means of LEO satellites.

The RDOF reverse auction began on October 29, 2020, and went through a dozen innocuous rounds before a falling clock percentage cleared the national $16 billion threshold, and awardees began to be selected. Table 1 shows how more and more CBGs were taken off the table in the last few rounds as the reverse auction closed in on a national winners list.

A few highlights here.

The GEO satellite companies, HughesNet and Viasat, stayed in until Round 13, but they couldn’t go beyond that because, with plans for “above baseline” 100/20 coverage at high latencies, their penalty of 60% meant that any price point below 61% would have meant no subsidies for them at all. Since no awards were made before Round 13, they had no effect on RDOF outcomes.

SpaceX, by contrast, had a big effect. With 100/20 as its promised speed, but with low latency thanks to the low orbit of its satellites, it faced only a 20% penalty. Like HughesNet and Viasat, it could afford to accept very low subsidies for a service that it was planning to deploy everywhere in any case. So, in Rounds 13, 14, 15, and 16, it competed in almost every CBG and was the biggest winner.

SpaceX ultimately won $885 million nationally (only to be disqualified later) but beyond that, it was also a big reason why the FCC underspent its $16 billion budget by nearly $7 billion. SpaceX reduced the RDOF Phase I spend both by outcompeting more expensive bidders, and by forcing bidders that ultimately outlasted SpaceX and won funding, to lower their prices first. Only in the last three rounds did the falling clock percentage, combined with the 20% penalty, push SpaceX to the sidelines.

The 0% clock percentage in Round 19 didn’t mean that the FCC was buying deployment commitments for free. Rather, Round 19 was the time for all the remaining bidders to make their best and final offers, after which the FCC would select the cheapest, since by then only gigabit low-latency tier bidders were still in play.

In many CBGs, the price was bid down all the way to at or near 1%. The most tenacious RDOF bidders were willing to accept obligations to deploy to over 725,000 locations for roughly $30 million in 10-year support, or just a few dollars per year in subsidies per location. That’s extreme, but RDOF’s national average of $176.80 per location in annual support is strikingly cheap compared to the capital expenditure budgets of broadband deployment or to the per location subsidy costs of other programs like USDA ReConnect.

At the national level, RDOF was impressively strategic in the way it leveraged competition among providers to buy deployment commitments from the broadband industry at a remarkably low price per location. Locally, however, results could be counterintuitive.

There were a lot of reasons, starting with different reserve prices, why similar areas with similar proposed fiber projects might get or lose them at very different subsidy rates. The process was opaque not only because of its inherent complexity, but also because the FCC had to be secretive and impose secrecy on bidders before and during the auction to prevent them from colluding.

In stark contrast, the BEAD program has been executed with a strong emphasis on local consultation throughout. Execution by state broadband offices brings the program closer to the people the program is serving. 

Speed of execution: RDOF and BEAD had different pain points

BEAD has been criticized for its slow execution. The IIJA passed in November 2021. Three and a half years later, no one has been connected to broadband, no funds have been passed on to ISPs to start construction, and most states have not completed their subgrantee selection. RDOF provides an interesting counterpoint.

On the one hand, RDOF moved more quickly from launch to awardee selection. It was three years from the passage of the IIJA to the first BEAD subgrantee selection announcements, and most states still have not picked their winners. From RDOF launch in August 2019 to the announcement of winners in December 2020 was less than 18 months. But RDOF took a long time to go from selecting awardees to authorizing them to receive money.

Resound Networks, mentioned above, was an RDOF winner in December 2020 but only got authorization two years later, in December 2022. Charter, another big winner, got authorized in mid-2022. To get authorized, RDOF winners had to submit a long-form application, get a letter of credit, and get authorized as eligible telecommunications carriers with requirements varying by state. That took time, and some RDOF winners never got authorized at all.

State broadband offices should be able to convert BEAD awards into funded projects more quickly that RDOF did, because by the time they submit Final Proposals, they will already have vetted not only applicant organizations but also specific projects. If so, that will reduce RDOF’s speed advantage.

But BEAD was slowed down somewhat by the program’s federalist design, and by the need to either establish state broadband offices, or for those offices to build capacity for a new state government function. Also, the more intentional pursuit of universal broadband access created special challenges in planning for deployment to the hardest-to-serve locations.

RDOF ‘defaults’ and the ‘winner’s curse’

If there is a single Achilles heel of the RDOF program, it is that so many of its promises fell through. Of just over $9.2 billion awarded in the auction, about $3.2 billion in funding for about 1.76 million locations never got authorized. The money wasn’t wasted. It stayed with the FCC. But it left big coverage gaps in what should have been RDOF’s footprint.

The “defaults” continued after authorization, though so far there are fewer. As of January 14, 2025, about $113 million in planned RDOF support has been defaulted, removing about 144,000 more locations from RDOF’s footprint.

RDOF grantees are not required to report on construction until three years after authorization, so the lack of documentation of RDOF-funded broadband expansion is not, so far, a sign of trouble. But soon, some RDOF winners will need to start supplying evidence that they have, in fact, deployed to 40% of their awarded locations. If they can’t, more defaults will follow.

While the defaults partly reflect the way RDOF put the cart of awardee selection before the horse of detailed review of project applications, it also reflected a troublesome phenomenon that is well known to auction theorists, under the moniker of “the winner’s curse.” Auctions tend to select bidders who get it wrong, bidding over-aggressively because they overestimate the value of what they’re bidding for. Auction winners who realize their mistake will then look for exits. RDOF experienced a lot of that.

Some RDOF defaults also show the limits of technology neutrality. LTD Broadband, a primarily wireless ISP, initially won $1.32 billion in RDOF awards but never got authorized because the FCC decided, ex post, that it was unconvinced of LTD’s ability to deliver the promised gigabit low-latency service across multiple states. LTD sued.

SpaceX didn’t promise gigabit, but it won the vast majority of the above-baseline (100/20) funding, only to be denied authorization. Again, the FCC decided, ex post, that it wasn’t willing to entrust the broadband future of over 600,000 locations to SpaceX’s then rather novel and unproven technology.

While these are reasonable judgments, it’s awkward for them to come after the auction. Yet as we have seen, allowing marginal providers to bid in the auction was key to why RDOF was so cost-effective. By first approving and then rejecting these companies, the FCC heightened competition and stretched scarce deployment dollars, albeit at the cost of leaving gaps. But that’s not a trade-off that BEAD, with its strong mandate for universal service, could emulate.

The emergence of the BEAD program, which wasn’t even a twinkle in Congress’s eye when RDOF ran its auction but was funded and moving ahead by the time RDOF made its last authorization decisions, altered the terms of the FCC’s authorization decisions. Had it authorized SpaceX’s funding, for example, the FCC would almost certainly have harmed the impacted areas by excluding them from BEAD eligibility. By the same token, in early 2024, there was pressure for an “RDOF amnesty” so that areas where RDOF projects had been authorized but were at high risk of failure could exit in time for BEAD to pick the areas up.

Conclusion

There is a saying that “the grass is always greener on the other side.” In 2021, when RDOF’s technology-neutral approach had led to the gigabit wireless controversy and a spate of defaults, an explicit preference for the one technology that everyone knew could really deliver future-proof gigabit service — end-to-end fiber — seemed to have merit. And so Congress, in the IIJA statute, directed the NTIA to favor “priority broadband projects” that could “meet the evolving connectivity needs of households and businesses” and “support the deployment of 5G, successor wireless technologies, and other advanced services,” criteria that the NTIA determined, in the BEAD NOFO, could only be satisfied by end-to-end fiber projects.

But fiber preference, too, has its counterarguments and its discontents. Similarly, the turn from RDOF’s opaque and technocratic approach to a strong emphasis on federalism and local consultation has contributed to BEAD’s much-criticized slower rollout.

At this stage, neither program can be assessed in terms of the principal goal, which is to expand broadband access. But a suggestive comparison of RDOF auction outcomes nationally with BEAD subgrantee selection outcomes in a frontrunner state, Louisiana, is shown in Figure 3. The similarity is striking. Louisiana BEAD is awarding a bit more per BSL for its projects than RDOF did. But that was to be expected, what with price inflation for broadband network components and BEAD’s need to push out to higher cost locations as it reaches for universal coverage.

But RDOF didn’t get dramatically lower prices than Louisiana BEAD, despite its expertly structured multiround auction, light initial vetting, technology neutrality, and lack of the special pricing and workforce burdens that come with BEAD. Offsetting these factors, Louisiana’s use of small project areas and sophisticated deconfliction may have been in practice, as it is in theory, more conducive to the design and selection of cost-effective projects.

Figure 2: Comparing price per BSL for RDOF and Louisiana BEAD

There are a lot of reasons why it’s not quite a like-to-like comparison. For example, Louisiana’s flat dirt is favorable terrain for broadband deployment compared with many other states’ sloping and/or rocky ground. Also, Louisiana’s broadband office has been consistently outstanding in its BEAD implementation.

So, while there are lessons to be learned from RDOF, it will be at least as helpful to encourage and enable state broadband offices implementing BEAD to learn from and emulate each other.