Wednesday, November 18, 2026

The Coming Washington Power Shift

 Democrats take the gavels in January with a target list built around Elon Musk. The contracts are real. The conflicts are real. So is the danger that oversight becomes a weapon aimed at the people who build things.

From Staff

Three weeks after the midterms, the shape of the next two years is clear enough. Democrats won the House. They have spent two years compiling lists of questions for Elon Musk, preferably answered under oath. Rep. Robert Garcia, the likely next chairman of the House Oversight Committee, said on CNN the Sunday after the election that there is "a lot of interest, not just from our committee, but from committees across the Congress to speak to Mr. Musk." Senate Democrats on the panels that write defense policy are signaling a harder posture toward how the Pentagon handles its SpaceX contracts.

The White House is not waiting for January. White House counsel Will Scharf sent Garcia a letter dated Nov. 9 warning that some of the foreshadowed probes could themselves be treated as criminal conduct, invoking the Supreme Court requirement that congressional investigations serve a legitimate legislative purpose rather than punish the investigated. A subsequent memo instructed agencies to route congressional requests through the White House Office of Legislative Affairs and the counsel's office before responding. President Trump, asked on a podcast about rumors of investigations into Musk, said: "It's talk. But if it is talk, we'll have to protect them." He also floated preemptive pardons for staff.

Musk answered on X by practically daring the government to move: he was "happy to cancel any that the government doesn't want. Just say the word!"

Every player in this fight is describing the same arrangement from a different seat. The largest government contractor in the country is run by the largest political donor in the country, and both parties now treat procurement as a lever to move the other.

The dollar figures are not in dispute. Their meaning is.

The Democracy Defenders Fund, a left leaning legal advocacy nonprofit, published a report in late October finding that Musk's companies have won at least $10 billion in federal contracts since he left his special government employee role in May 2025. That sits on top of roughly $6.3 billion awarded during his four months running the Department of Government Efficiency. The group says its tally is conservative because it excludes classified awards.

The largest items are two Golden Dome awards to SpaceX: about $4.2 billion to develop a space based missile tracking radar and about $2.3 billion to build a communications satellite constellation. NASA added $946 million in September for three more crew missions to the International Space Station. In July the Pentagon tripled a launch services contract ceiling to $17 billion and raised the planned number of launches from 60 to 170 over the next decade. In August the Pentagon redacted the amount it pays SpaceX to expand Starshield use aboard military aircraft.

Two structural facts sit underneath all of it. First, the military has no substitute. Nearly 94 percent of US orbital launches last year came from SpaceX. The company is running at roughly 90 percent this year. Second, the government is about to lose the rocket it depends on: SpaceX's planned retirement of Falcon 9 has reportedly left the Space Force scrambling for future launch capacity, which is why the Defense Department's Office of Strategic Capital signed a $150 million conditional loan commitment with Rocket Lab this month.

Clayton Swope of the Center for Strategic and International Studies described the long standing discomfort inside the Pentagon at putting "all eggs in the SpaceX basket." Dave Cavossa of the Commercial Space Federation countered that Starlink and Starshield are unique capabilities and that competitors such as Amazon Leo are a year or two from being operational. Both statements are true at once, which is the entire problem.

Most of the coming hearings will be about appearances. One thread is different and it is the one worth watching.

In October, Ukraine's military reported that Starshield, the military grade version of Starlink that SpaceX licenses to the US government, had been throttled. Ukrainian units reported link drops at speeds as low as 90 kilometers per hour, making strike drones that cruise at 160 kilometers per hour or more inoperable. CBS News reported that the decision came from Musk after multiple conversations with administration officials. Some officials inside the administration were surprised because neither the president nor the Pentagon had ordered limits. Musk said Ukraine violated the terms of service by using a civilian commercial service for offensive drone strikes and denied responsibility for the Starshield disruption. Ukraine says restrictions were imposed again in mid November. The Pentagon inspector general's office has said it is not investigating, despite a bipartisan request from senators.

That is an identifiable act by an identifiable party with identifiable consequences, which is precisely what oversight is for. The distinction that matters is between investigating conduct like that and running hearings because a donor backed the wrong candidate.

The reason any of this matters economically is that cheap, frequent reuse is the rare policy relevant technical fact.

An expendable rocket throws away its most expensive hardware after every flight. A stage that costs $100 million and flies once allocates $100 million of manufacturing cost to that flight. Fly it ten times and the allocation falls to $10 million. Fly it a hundred times and it falls to $1 million. Those are illustrative figures, not company estimates. They exclude development, infrastructure, propellant, labor, repairs, financing and losses. The real number is decided by inspection and refurbishment costs, which is exactly where the fight is.

Falcon 9 shows the mechanism working. ARK Invest's analysis, tracking turnaround times, estimated that first stage refurbishment fell from roughly $13 million to about $1 million over five years, with a reused Falcon 9 costing on the order of $800 per kilogram to low Earth orbit against roughly $2,700 for a new one. NASA's own Harry Jones, in a paper on the recent large reduction in launch cost, was more skeptical. Falcon 9 advertised $62 million for 22,800 kilograms to low Earth orbit, about $2,720 per kilogram. He noted that reuse had not yet delivered lower cost, that the Shuttle was extremely expensive largely because of refurbishment. That Falcon 9 reuse might reduce costs by a factor of two. Both readings are defensible because they count different things.

Starship is earlier in that progression. Flight 15 this month ended with the Super Heavy booster caught by the launch tower, a spectacular recovery. Recovery is step one. Step two is establishing the vehicle's condition afterward. Ars Technica's reporting on the previous flight described white streaks across the heat shield, all originating at tile boundaries, along with cracked tiles and broken edges, evidence that hot gas was finding paths between tiles. Dan Rasky, a former NASA engineer who helped invent the PICA material used on Crew Dragon, called the current thermal protection system a dead end for any mission requiring full and rapid reuse. Full means both stages. Rapid means days or weeks rather than months. SpaceX has not yet demonstrated rapid reflight of either vehicle.

Lower costs do not automatically mean a bigger launch business measured in dollars.

A hundred launches at $100 million each is $10 billion in revenue. Five hundred launches at $10 million each is $5 billion. Five times the missions, half the money. Whether that happens depends on how strongly demand responds to price, which is the central uncertainty in launch economics and on whether savings reach customers at all. Operating cost and price are different numbers. A dominant provider facing weak competition passes through only what competition forces it to pass through.

The market is already showing the shape of things. Worldwide orbital launches hit 329 in 2025, the most of any year. As of mid November, 2026 stood at 269 attempts with 261 reaching orbit, slightly behind the same point last year. SpaceX has flown 127 of them, about 47 percent of missions but roughly 78 percent of the mass delivered to orbit. Its share of launches slipped from 56 percent to about 48 percent as Chinese and other providers picked up the slack.

Much of the value from cheap access will show up outside the launch business entirely, in satellites, research and ventures that only pencil out when transportation stops being the dominant line item. Cheaper launches do not eliminate spacecraft manufacturing, operations, financing, or the need for paying customers.

Strip out the noise and the Democratic plan is a process, not a verdict. What has been described publicly is intended scrutiny of Musk's DOGE tenure. The relationships between his companies and the agencies that regulate and buy from them. What the administration has described is a plan to contest subpoenas and assert legal protections over certain communications. Neither of those is a finding of wrongdoing. Both are announced intentions.

The likely sequence runs like this:

Record preservation requests and document demands go out first, usually before the committee gavels are even warm.

Requests for voluntary testimony follow, with a handful of public hearings chosen for maximum attention.

Subpoenas arrive where cooperation fails.

Negotiations or litigation over disputed material consume the middle of the year.

Legislative recommendations or referrals, if any, come last, and depend entirely on what the documents show.

Congress has broad investigative authority, but it is tethered to legitimate legislative purposes and bounded by constitutional limits. Enforcement disputes run through negotiation, contempt procedures, and the courts. A congressional investigation is not a criminal conviction. A hearing is not a prosecution. The administration's threats against investigators over allegedly unlawful demands deserve exactly the same evidentiary scrutiny as the allegations coming from the committee chairmen. A claim is a claim no matter which letterhead it arrives on.

Focused review of specific decisions, contracts can establish facts, protect taxpayers, and surface genuine conflicts. The cost is compliance expense and management distraction. Including for the engineers and program managers who have to answer the requests instead of doing the work.

Evidence based reforms can improve safeguards and fair competition. Badly designed requirements can also saddle a functioning industry with burdens that serve no safety purpose and simply entrench whoever can afford the lawyers.

Prolonged investigations without clear objectives produce little beyond reputational damage and uncertainty unless new evidence actually emerges. Broad restrictions on essential operations might address a serious proven problem, but the collateral lands on customers, employees, suppliers. And public services that had nothing to do with it.

These are scenarios, not predictions. Nobody can calculate the financial cost of the coming investigations or quantify the innovation that might be delayed. The existence of compliance costs does not by itself prove an investigation is improper. The largest danger is narrower and darker than either side admits. Officials treating damage to a productive enterprise as an objective rather than as a consequence to be avoided wherever possible.

Any investigation should be judged against six practical tests.

Specificity. Identify the conduct at issue rather than relying on political association.

Consistency. Apply comparable standards to political allies and political opponents. If the standard only exists for one party's donors, it is not a standard, it is a weapon.

Evidence. Distinguish suspicions, documented facts, established violations and say which one you have.

Proportionality. Match remedies to proven problems, not to headlines.

Continuity. Account for effects on essential services, innocent third parties, including the troops and astronauts who fly on the hardware.

Closure. Narrow or end inquiries when the evidence does not support the allegations. An investigation that cannot conclude is an instrument of punishment wearing procedure as a costume.

Political disagreement alone should not justify legal pressure. Technical achievement alone should not excuse misconduct.

There is an older argument underneath all of this, and it should discipline the hearings.

An economy cannot keep expanding its productive capacity while undermining the people and organizations that create it. The postwar United States broadly grew output and typical worker compensation together until the late 1970s, when the two diverged. The Economic Policy Institute points to declining collective bargaining as an important contributor without claiming it is the only one. Growth does not distribute its gains automatically.

China's household responsibility system in the late 1970s pushed production. Income decisions down to households and was associated with large gains in agricultural output, a reminder that rewarding producers expands the pie rather than merely slicing it. Germany's codetermination system put employee representatives on supervisory boards. The research record does not support the claim that worker participation necessarily damages productive enterprise. None of these is a transferable formula. All of them point the same direction. Incentives, participation, and accountability can coexist.

The strongest version of this argument is not that Elon Musk should be beyond scrutiny. It is that the United States should not casually endanger productive capability because the person running it has become politically inconvenient.

Rapid, economical reuse could cut the resources consumed in reaching orbit and unlock scientific and commercial activity that currently does not pencil out. Whether that promise is kept depends on repeatable engineering, sustainable operating costs, sufficient demand and actual competition. None of those four things is improved by two years of subpoenas.

The defensible forecast remains legal expense, distraction and political conflict rather than operational collapse. The risk worth naming is the precedent. If contracts are understood to follow political allegiance, the next administration rewrites the ledger. A generation of effort to build competitive private launch capacity becomes an appendage of whoever holds the gavel. Congress cannot investigate its way to a second supplier. It can only fund one and for a decade it has done the opposite while congratulating itself on the savings.

Investigate specific evidence. Correct proven abuses. Preserve the capacity to build. Accountability is supposed to protect the conditions for innovation, not replace it with political retaliation.

Sources:

Axios, "SpaceX congressional scrutiny looms," Thomas Novelly, Nov. 12, 2026. The $10 billion figure, the $2.3 billion and $4.2 billion Golden Dome awards, the $946 million NASA award, the 94 percent and 90 percent US launch shares, the Falcon 9 retirement bind, the $150 million Rocket Lab loan commitment, the Swope and Cavossa quotes, and Musk's "happy to cancel any" post. https://www.axios.com/2026/11/12/spacex-scrutiny-looms-congress


No comments:

Post a Comment

The Coming Washington Power Shift

 Democrats take the gavels in January with a target list built around Elon Musk. The contracts are real. The conflicts are real. So is the d...