Disturbing History - The Merchants Of Death

Episode Date: September 4, 2026

On the evening of January 17, 1961, with just three days left in his presidency, Dwight D. Eisenhower delivered a 16-minute farewell address containing one warning that would outlive nearly everything else he said in public office: America must guard against the “unwarranted influence” of the military-industrial complex.This episode of Disturbing History examines what Eisenhower was really describing, because he was not predicting the future. He had spent eight years inside the system himself.Brian King-Sharp traces the rise of that system from the 1934 Nye Committee hearings, which investigated major American arms manufacturers, through the enormous industrial mobilization of World War II.The episode examines wartime cost-plus contracts, corporate executives embedded inside government, falsified inspections involving defective aircraft engines and substandard steel, and a 1944 proposal by General Electric president Charles E. Wilson to make the wartime partnership between industry and the military permanent.The Cold War did exactly that.From Operation Paperclip and the recruitment of roughly 1,600 German scientists, to secret human radiation experiments involving plutonium injections, radioactive iron given to pregnant women, and radioactive oatmeal fed to children at the Fernald State School, the episode follows what happened when national security programs operated behind layers of classification and limited oversight.The story then moves through Castle Bravo, the 15-megaton nuclear test that produced roughly 2.5 times its predicted yield, radioactive fallout across the Marshall Islands, nuclear testing in Nevada, and government efforts to minimize or conceal the consequences.From there, the episode opens up the machinery itself: defense contractors, congressional districts, the revolving door between government and industry, deliberately optimistic cost estimates, and the consequences faced by officials who challenged the numbers. Ernest Fitzgerald revealed the true cost overruns of the C-5A transport aircraft in 1968 and was fired, eventually becoming part of a Supreme Court case that helped establish modern presidential immunity doctrine.The investigation continues through the 1971 Lockheed bailout, international bribery scandals, Agent Orange, unexploded ordnance in Laos, human testing at Edgewood Arsenal, biological experiments over American cities, post-Cold War defense-industry consolidation, billions of dollars lost during the Iraq War, extreme contractor markups, and American-made weapons recovered after civilian deaths overseas. It ends with two numbers that reveal how large the system has become: the F-35 program, projected to cost more than $2 trillion over its lifetime, and a Department of Defense that has now failed eight consecutive financial audits.The final scene returns to Abilene, Kansas, and the surviving drafts of Eisenhower’s farewell address.According to his brother Milton, Eisenhower once considered describing a broader “military-industrial-congressional complex,” explicitly naming the political institution responsible for funding and overseeing the system.That version of the page has never been found.Email BrianJoin Our FREE NewsletterGet Brian's Books Leave Us A VoicemailVisit Our WebsiteHave a forgotten historical mystery, disturbing event, unsolved crime, or hidden conspiracy you think deserves investigation?Send your suggestions to [e-mail] History is a dark history podcast exploring unsolved mysteries, secret societies, historical conspiracies, lost civilizations, and the shadowy stories buried beneath the surface of the past.Follow the show and enable automatic downloads so you never miss a deep dive into history’s most unsettling secrets.Because sometimes the truth is darker than fiction.

Some stories were never meant to be told. Others were buried on purpose. This podcast digs them all up. Disturbing history peels back the layers of the past to uncover the strange, the sinister, and the stories that were never supposed to survive. From shadowy presidential secrets to government experiments that sound more like fiction than fact, this is history they hoped you'd forget. I'm Brian, investigator, author, and your guide through the dark corner.

of our collective memory. Each week I'll narrate some of the most chilling and little-known tales from history that will make you question everything you thought you knew. And here's the twist. Sometimes the history is disturbing to us. And sometimes, we have to disturb history itself,

just to get to the truth. If you like your facts with the side of fear, if you're not afraid to pull at threads, others leave alone. You're in the right place. History isn't just written by the victors. victors. Sometimes it's rewritten by the disturbed. On the evening of January 17th, 1961, a 70-year-old man sat down at a desk in the Oval Office in front of three television cameras

and read out loud from a speech he'd been working on, off and on, for about two years. He had three days left as president. He'd commanded the largest amphibious invasion in human history. He'd signed off on eight years of peacetime defense budgets larger than anything the country had ever seen, and he'd watched an entire industry grow up around the machinery he once commanded. The address ran about 16 minutes. One paragraph of it has outlived every other thing Dwight Eisenhower ever wrote or said. Almost nobody who quotes that paragraph has read the rest of the speech, and fewer still know what he'd been looking at for the previous 15 years when he wrote it. The farewell address wasn't written in a hurry. Eisenhower asked a speechwriter named Malcolm Moose to

start thinking about it back in 1959, and the drafts piled up over the next year and a half. The Eisenhower Presidential Library in Abilene, Kansas, holds a stack of them. Moose worked on it, a naval officer named Ralph Williams worked on it, the president's brother Milton worked on it, and Eisenhower himself went through the language line by line the weekend before he delivered it. What he said, in the part everybody knows, is that the United States had been forced to create a permanent armaments industry of vast proportions, that three and a half million people were directly employed in the defense establishment, that the annual military budget exceeded the net income of every corporation in the country combined, and that the American people had to guard

against the acquisition of unwarranted influence by what he called the military industrial complex. There's a story that's been repeated for 60 years, that the original phrasing was military industrial congressional complex, and that the word congressional complex. And that the word congressional got cut at the last minute, so the outgoing president wouldn't be seen taking a swing at the legislative branch on his way out the door. That claim comes mainly from Milton Eisenhower, who told the story years later. I went looking for the draft page that would show it, and the archivists at the Eisenhower Library have said for years that no surviving draft in their collection contains the word. That doesn't mean Milton was lying. It means we have a

recollection from a man who was in the room and no paper that backs it up. And those are two different categories of evidence. So that's where it has to rest for now. There's another part of that same speech people skip. Right after the famous paragraph, Eisenhower talked about scientific research becoming the province of federal contracts, about a solitary inventor being replaced by task forces of scientists working in laboratories funded by the government. And he said that public policy could itself become the captive of a scientific technological elite. He gave two warnings that night, and everybody remembers one of them. Most retellings also skip how the speech landed. When he gave it, the newspapers barely noticed. Kennedy's inauguration was three days out and the country

was looking forward. The phrase didn't really enter the bloodstream until later in the 60s, when a younger generation went hunting for something to hang the war in Vietnam on and found a Republican and five-star general who'd said it first. So the question I kept coming back to while putting this together is whether Eisenhower was predicting something or describing something that had already happened, something he'd participated in and knew a great deal more about than he was willing to say into a television camera on his way out of the building.

I think the answer is the second one. And I think the record makes that reasonably clear. Americans first had this argument in public in 1934, and it went badly. That spring, Fortune Magazine ran an unsigned piece called Arms and the Men. The same year, two writers named H.C. Engelbrecht and Frank Hanigan published a book called Merchants of Death. The thesis in both was that the arms industry had a commercial interest in war, that it sold to both sides, that it bribed officials, and that it worked quietly against disarmament agreements, because disarmament was bad for orders. The book was a bestseller and became a Book of the Month club selection.

The phrase, Merchants of Death went straight into the language. Congress responded the way Congress does. In April of 1934, the Senate created a special committee on investigation of the munitions industry, chaired by Gerald Nye, a Republican from North Dakota. The Nye committee sat for about two years, held 93 hearings, and put more than 200 witnesses under oath. Among them were the DuPont brothers, executives from Electric Boat, from Curtis Wright, from federal laboratories, and from the banking houses that had financed Allied purchasing before American entry into the First World War.

What came out of those hearings still reads like material the companies never expected to see daylight. Electric Boat, the submarine builder that eventually became part of General Dynamics, had a long-standing arrangement with the British firm Vickers, under which the two companies spent. territories and shared commissions on foreign submarine sales. The committee put correspondence into the record showing payments to agents in South America who were expected to move governments toward purchases. Curtis Wright's aviation salesmen were shown discussing the commercial upside of instability in the countries they sold to. Federal Laboratories is the one that stays with me.

That company sold tear gas and gas guns, and the committee produced correspondence showing sales pitched not to foreign armies, but to American industrial employers for use against striking workers, along with a salesman writing about the business opportunities that labor trouble created. The record includes a federal laboratories representative involved in shipping gas munitions during American labor disputes in the early 30s. That's a company selling chemical weapons into a domestic labor conflict and looking forward to more of it. The DuPont testimony produced the famous exchange about profits. And the DuPonts, to their credit, were relatively straightforward about the business they were in.

Pierre DuPont pointed out, accurately, that his company had made the powder the government asked for at the prices the government agreed to. Much of what the munitions makers had done was legal, ordinary commerce in a market where the product happened to be artillery shells, and that was part of the committee's problem. The Nye Committee did not end well. In January of 1936, Nye said on the Senate floor, lore that Woodrow Wilson had withheld from Congress his knowledge that the Lusitania had been carrying munitions. Senate Democrats came apart. Carter Glass of Virginia pounded the desk hard enough to bloody his hands, and Tom Connolly of Texas went after Nye for slandering a dead president. The committee's funding dried up, and it

wrapped up its work not long after, with no legislation to show for it and its chairman politically damaged. What it did leave behind was a mood. The neutrality acts of the mid-30s came out of that mood. So did a national reflex that treated the arms business as inherently disreputable. The reflex did not survive the next war. The same companies that got hauled in front of Nye's committee ended up as the backbone of American mobilization less than a decade later. Nobody apologized, and nobody had to because the country needed the powder.

The machine itself got built during the Second World War. Between 1940 and 1945, the federal government issued somewhere around $175 billion in prime supply contracts. About two-thirds of that went to the largest 100 corporations. Roughly a third went to the top 10. The War Production Board staffed itself with executives on loan from those same corporations, the so-called dollar-a-year men, who kept their company salaries while allocating national resources and awarding contracts

to industries they intended to return to. The dominant contract form was cost plus fixed fee, which does exactly what it sounds like. The government reimburses the contractor's costs and pays a fee on top. Under that arrangement, an inefficient contractor doesn't lose money. He just bills more. Everyone involved understood this at the time, and it was accepted as the price of speed, because speed was the whole point when the goal was 100,000 aircraft. The man who noticed the leakage was a first-term senator,

from Missouri named Harry Truman, who took a car trip through defense construction sites in 1941 and came back angry about what he'd seen. The Special Committee to investigate the National Defense Program, the Truman Committee, held hundreds of hearings over the next several years, and is generally credited with saving billions of dollars, though the exact figure has always been an estimate rather than an audit. Some of what the committee found went well beyond padded invoices. At the Wright Aeronautical Plan in Lockland, Ohio, run by Curtis Wright, investigators found that defective aircraft engines had been passed as airworthy. Inspection records had been falsified, inspectors who objected had been pressured, and

engines with known defects had gone out the door to be installed in aircraft, flown by American crews. The Truman Committee held hearings on it in 1943. Federal charges followed against plan inspectors, and there were convictions, though the appellant. pellet history is messy, and some of those convictions did not survive review. The core finding of the committee was blunt. The Army's own inspection apparatus at that plant had been captured by production pressure. The second case involved Carnegie, Illinois Steel, a United States Steel subsidiary, where test results on steel plate for ships and other war material were found to have been falsified at the Irvine works. Test specimens were manipulated so that plate that hadn't met

specification would pass. That plate went into shipholes. I've been in enough investigations to know how this happens. Nobody wakes up planning to send bad steel to a shipyard. What happens is a quota, a schedule, a boss who doesn't want to hear it, and a man at a bench who decides that this one time isn't going to matter. Then it happens twice, and then it's the way the shop runs. The mechanism is boring and the consequences are not. On the 19th of January, 1944, with the war still very much undecided, an executive named Charles E. Wilson gave a speech at the 25th anniversary dinner of the Army Ordinance Association. This is the Wilson who was president of General Electric and served as executive vice chairman of the war production board, not the Charles Wilson from

General Motors who came later, and the two of them get mixed up constantly. The speech was published that spring in the association's own journal under the title for the common defense, subtitle, a plea for a continuing program of industrial preparedness. Wilson's proposal was that the wartime partnership between industry and the armed services should not be dismantled when the shooting stopped. He wanted it made permanent. He suggested that every major company designate a liaison officer to work with the military, that the arrangement be insulated from ordinary political control, and he argued that the country could not afford to demobilize its industrial mobilization apparatus the way it had after 1918.

In January of 1944, the invasion of France hadn't happened yet, and a senior American industrialist was already lobbying to make wartime industrial mobilization a permanent peacetime institution. Wilson didn't use the phrase permanent war economy. That phrase came from his critics on the left the same year. What he described was the thing the phrase was coined to name. He got most of what he asked for, and the mechanism was the cold,

war. The National Security Act of 1947 created the Department of Defense, the Central Intelligence Agency, and the National Security Council. NSC-68, the 1950 policy paper drafted largely by Paul Nitzah, argued for a massive and sustained military buildup and assumed a threat environment with no obvious end date. Korea turned the argument into appropriations. The defense budget roughly tripled between 1950 and 1953, and never went back to where it had been. Before the Second World War, the United States built weapons during wars and stopped between them. After 1950, it stopped stopping. Every company, every laboratory, every congressional district, and every career inside that system now had a continuing interest in the continuation of the system.

That's an incentive structure rather than a conspiracy. Conspiracies require. require secrecy and coordination, and incentive structures just require people to act normally. An institution, that permanent, that well-funded, and that heavily classified is also an institution that can do things to people without ever being asked to explain itself. The record shows that it did. The scientists came first, and they came in on transport planes. In the summer of 1945, American teams were racing through occupied Germany, scooping up rocket engineers, chemists, aviation physicians, and metallurgists, partly to use them, and partly to keep the Soviets from using them. The program that formalized this was run by the Joint Intelligence Objectives Agency

and was eventually named Operation Paperclip. Stay tuned for more disturbing history. We'll be back after these messages. Something in the range of 1,600 German scientists and their families were brought to the United States over the following years. President Truman's policy on the program was explicit. It excluded anyone found to have been a member of the Nazi party for more than nominal reasons, an active supporter of Nazi militarism or a participant in its crimes. That policy was not followed. The way it wasn't followed is documented in the files themselves. Security reports that described individual scientists as ardent Nazis were rewritten. Dossiers were sanitized. Men whose files had described them as security threats were reissued with clean assessment.

so that visas could be granted. Linda Hunt's research in the 1980s, based on the agency's declassified records, laid out the rewriting in detail, and Annie Jacobson's later work went through it again, with more of the file material available. The name everybody knows is Vernher von Braun. He joined the Nazi party in 1937,

and the SS in 1940, and he held the rank of Sturm-Bahnfuerre, which is equivalent to a major. He said after the war that party membership had been a professional requirement and that his SS rank was honorary and pressed on him. Both of those claims have supporting context. Neither of them addresses the part that actually matters, which is where his rockets got built. The V2 production line at Middleverk, in a tunnel complex in the Hartz Mountains near Nordhausen,

was worked by concentration camp prisoners from the Middle Biodora camp. Prisoners cut the tunnels first, living and sleeping inside them without daylight or adequate sanitation. Roughly 60,000 prisoners passed through the Dora complex and at subcamps. The best estimates put the dead at around 20,000. Men died of tuberculosis, dysentery, starvation, and exhaustion, and men were hanged in the tunnels, sometimes from an overhead crane, in front of the assembled workforce. Roughly 3,000 people were killed by V2 rockets landing on London, Antwerp, and other targets.

Something on the order of 20,000 died making them. More people died building that weapon than died from it. Fawn Brown visited Middlewerk, and he acknowledged that later in life, describing the conditions there as repellent. Testimony from surviving prisoners and from Middleverk personnel places him in the tunnels on multiple occasions. There is also documentary evidence cited in the post-war record of his involvement in requesting skilled prisoner labor from camps. What there isn't is evidence that he personally ordered a killing or rancers. ran the camp. The office of special investigations examined him and did not bring a case before his

death in 1977. The honest statement is that he was an officer in the SS who ran the technical development of a weapon he knew was being manufactured by slave labor in a tunnel, where thousands were dying and that he continued the work. The man the government did move against was Arthur Rudolph. Rudolph was the operations director at Middlewerk, meaning production, meaning the labor. He came to the United States under paperclip, worked on the Pershing missile, and became the project director for the Saturn 5,

the rocket that carried Apollo to the moon. NASA gave him its distinguished service medal. In 1982, the Office of Special Investigations opened an inquiry into him. In 1984, he signed an agreement to leave the United States and renounce his American citizenship rather than face denaturalization proceedings. He moved to Hamburg and died there in 1996.

West German prosecutors looked at him and closed their file without charges. He always maintained he'd done nothing but managed production. A number of his former colleagues at NASA defended him for years afterward and argued the government had treated him unfairly. The rockets get the attention, but paperclip also brought over the Luftwaffe's aviation doctors. Hubertus Strughold ran aviation medicine research for the Luftwaffe and came to the United States to work for the Air Force,

where he was later called the Father of Space Medicine. An aeromedical library at Brooks Air Force Base was named for him. Colleagues and subordinates in his professional circle conducted the low-pressure and freezing experiments at Dachau, in which prisoners were killed. Whether Strughold himself directed or approved any of those experiments has never been established to a legal standard, and he denied it until his death.

What is established is that he attended a 1942 conference where such research was discussed. His name was removed from that library in 1995, and a space medicine award name for him was retired in 2013. And then there's Walter Schreiber, a former Surgeon General of the German Army, who ended up working at the Air Force School of Aviation Medicine in Texas, until the columnist Drew Pearson published material connecting him to human experimentation in 1952. The Air Force arranged for him to go to Argentina. What I keep circling back to is the calendar. The Nuremberg doctor's trial ran from December of 1946 to August of 1947.

Seven defendants were hanged. The tribunal produced the Nuremberg Code, the founding document of modern research ethics, which starts with the principle that voluntary consent of the human subject is absolutely essential. While that court was sitting in Germany, American agencies were actively importing German scientists in the same, same fields, editing their files to get them past immigration and putting them to work. And at the same time, on American soil, the United States government was conducting experiments

on its own citizens that would have violated every line of the code it had just written. On April 10, 1945, at the Army Hospital on the Clinton Engineer Works Reservation in Oak Ridge, Tennessee, a cement mixer named Eb Cade was injected with 4.7 micrograms of plutonium. He'd been in a car wreck on the 24th of March and had fractures in his arm and leg. In the files he's referred to as HP12, the H and the P stand for human product. Kade did not consent, and there's no evidence he was told. His fractures were left unset for about 20 days, until after the injection had been given and the excretion sampling was underway.

When his bones were finally set on the 15th of April, 15 of his teeth were extracted and a bone biopsy was taken so his tissue could be analyzed for plutonium uptake. He discharged himself from the hospital, left Tennessee, and died of heart failure on April 13, 1953, in Greensboro, North Carolina. He was black. He was a laborer, and he was hospitalized in a segregated facility in wartime, Tennessee. The record does not establish with certainty who ordered the injection. The physicians most likely to, have known contradicted one another afterward. Cade was the first of 18 people injected with plutonium between 1945 and 1947, at Oak Ridge, at the University of Rochester, at the University of Chicago, and at the University of California in San Francisco. Eleven of them were at Rochester, three at Chicago,

three in San Francisco, and one at Oak Ridge. The subjects included a 58-year-old house painter named Albert Stevens, coded CAL-1, who was misdiagnosed with terminal stomach cancer and injected in May of 1945 on the assumption he'd be dead within months. The diagnosis was wrong, and what he actually had was a benign ulcer that had been surgically removed. He lived another 20 years and carried the largest known accumulated radiation dose of any human being. There was a four-year-old Australian boy, coded CAL-2, who'd been flown to California for treatment of bone cancer and was injected while hospitalized there.

And there was Elmer Allen, coded C-A-L-3, a railroad porter whose leg was amputated three days after the injection, so the bone could be studied. None of these people were told what they'd been given. Some of them were tracked for decades afterward by government researchers who visited them under other pretenses to collect samples. Elmer Allen's family later described how researchers would arrive, periodically and take him for tests. The purpose of all of it was dosymmetry. The Manhattan

Project needed to know how much plutonium a worker's body would retain and excrete, because it had thousands of employees handling the material and no human data. The scientific question was legitimate. The method was to inject people who had no idea it was happening. The plutonium injections were not an isolated program. At Vanderbilt University between 1945, in 1949. Roughly 829 pregnant women attending a prenatal clinic were given radioactive iron and what they were told was a vitamin drink. The purpose was to study iron absorption in pregnancy.

A follow-up study, decades later, found several cancers among the children of those women at a higher rate than in the comparison group, and litigation over the study was settled in the 1990s. At the Walter E. Fernaud State School in Massachusetts, an institution for boys labeled feeble-minded. Researchers from MIT ran nutrition studies in which boys were fed oatmeal containing radioactive calcium and iron tracers. The oatmeal was supplied by Quaker oats, which funded part of the research. The boys were recruited into what was called the Science Club, and members got extra milk, trips to baseball games and parties. Consent letters went to parents describing

the nutrition study and mentioning special privileges. The letters did not mention radioisotaph, hopes. The doses were low and the panel that later reviewed the case concluded that no boy was likely to have suffered physical harm. A lawsuit brought decades later, settled for around $1.85 million. None of this came out because the government volunteered it. In 1993, a reporter named Eileen Wellsom at the Albuquerque Tribune published a series that identified five of the 18 plutonium subjects by name after years of digging through documents and tracking down families. She won the Pulitzer Prize for it. Energy Secretary Hazel O'Leary responded by opening the department's records,

and President Clinton created the Advisory Committee on Human Radiation Experiments, which spent more than a year going through hundreds of thousands of pages. The committee's report delivered in 1995, documented thousands of human radiation experiments conducted or funded by the federal government between 1944 and 1974. Some were minimal risk, some were not. It found a consistent pattern in which officials knew that public disclosure would be a problem and made decisions with secrecy in mind rather than consent. On October 3, 1995, Clinton apologized on behalf of the government to the subjects and their families.

That's 50 years between Eb Cade's injection and the apology. Cade had been dead for 42 of them. The weapons themselves were even less discriminating than the experiments. On March 1st, 1954, the United States detonated a thermonuclear device called Shrimp on Bikini Atoll in the Marshall Islands, under the test name Castle Bravo. The predicted yield was around six megatons. The actual yield was about 15. The error came from the physics. The designers had assumed that lithium 7 in the device's fuel would be essentially inert, and it turned out to contribute enormously to the reaction. It remains the largest device the United States has ever detonated, and it came in about two and a half times the predicted size. The record shows some warning, though.

Documents released later showed that designers at Los Alamos understood the yield might run as high as 15 megatons, and passed that warning along so the task force could protect its own personnel. The fallout went the wrong way, and the question of how much warning there was has been argued over ever since. The Atomic Energy Commission's public position was that the wind shifted unexpectedly after the shot. Weather information available in the hours before the detonation showed winds at altitude trending toward the inhabited atolls to the east, and the shot went forward. Within hours, a fine white ash was falling on Rongalap Atoll, about 100 miles away. Survivors have described children playing in it, because they'd never seen anything like it

and thought it was snow. People ate food it had settled on and drank water it had fallen into. Within a day, residents were vomiting, their skin was burning and their hair was coming out. The Navy evacuated Rongalap about two days after the detonation. Udurik was evacuated later, and the residents of Eiling Nye, who'd been on the atoll gathering food, were picked up as well. The exposed marshalese became the subjects of a study designated Project 4.1, run by the the Atomic Energy Commission's Brookhaven National Laboratory, described in its own documentation as a study of the response of human beings exposed to significant beta and gamma radiation due to fall out from high-yield weapons. Whether that study was conceived before the detonation

as an intended part of the test or created afterward as a medical response to an accident, has been argued about for decades. And Marshallese advocates and some researchers have pointed to the pre-test planning documents as evidence that it was a part of the plan. The AEC has always said it was a response to an accident. What isn't in dispute is what the commission did afterward.

In 1957, the AEC returned the people of Rangalop to their atoll and told them it was safe while its own scientists were documenting residual contamination. An AEC document from that period described the resettled atoll as by far the most contaminated place

in the world and noted the opportunity to gather data on human beings living in a contaminated environment. Thyroid nodules and thyroid cancers appeared in the exposed population, especially among people who'd been children in 1954. Women reported miscarriages and stillbirths at rates they had not seen before. In 1985, no longer willing to wait for the United States government, the people of Rangalop left. Greenpeace's ship, the Rainbow Warrior, moved about 300 people

and their belongings to Majado, over the course. of several trips. Stay tuned for more disturbing history. We'll be back after these messages. Two months later, French intelligence agents bombed that ship in Auckland Harbor over an unrelated matter, which is its own story for another day.

The Marshall Islands absorbed 67 American nuclear detonations between 1946 and 1958. On Runeut Island in any Wetak atoll, the United States eventually scraped up contaminated soil and debris and dumped roughly 85,000 cubic meters of it into an unlined crater left by an earlier test, then capped it with a concrete dome about 18 inches thick. Soil contaminated with plutonium was also shipped in from the Nevada test site and buried there.

The dome has cracked, and the Department of Energy has acknowledged the cracking while saying the material below poses no immediate threat, while conceding that the lagoon sediment around it is contaminated, and it wasn't only the marshalese. On the same day as Castle Bravo, a Japanese tuna boat named the Daigo Fukuir Yu Maru, the Lucky Dragon Five, was fishing outside the declared danger zone

when the ash fell on it. 23 crewmen were aboard, and all 23 of them got sick. The radio operator, Aikichi Kuboyama, died that September. The United States paid Japan a settlement without accepting liability, and Japanese consumers stopped buying tuna for months.

Then there's what the testing did at home. Between 1951 and 1962, the Atomic Energy Commission detonated around 100 atmospheric nuclear devices at the Nevada Proving Ground, northwest of Las Vegas. Fallout drifted east and northeast over Nevada, Utah, and Arizona. The people who lived under it are called downwinders. In the spring of 1953, after a test series called Upshot Not Hole, sheep ranchers in southern Utah started losing animals. ewes aborted their lambs. Sheep developed lesions around the mouth and nose and lost their wool, and thousands of animals died. The AEC investigated, and its investigators in the field reported findings consistent with radiation injury. The final AEC position was that malnutrition and cold

weather had killed the sheep, and that radiation had nothing to do with it. The ranchers sued and lost in 1956. In 1982, after documents surfaced showing that the government had pressured its own scientists to change conclusions and had withheld material from the court, District Judge A. Sherman Christensen, the same judge who'd heard the original case, found that the United States had committed fraud upon the court and reopened it. That finding was later reversed on appeal on procedural grounds and the ranchers never collected. The health consequences for the people are harder to pinned down individually and clearer in the aggregate. Leukemia rates in southern Utah counties rose in the period following the tests,

and a series of epidemiological studies have connected exposure to iodine 131 from fallout to thyroid cancer. A National Cancer Institute study released in 1997, estimated that fallout from Nevada testing exposed essentially the entire population of the continental United States to some level of iodine 131. largely through milk, and that it would eventually cause tens of thousands of thyroid cancer cases. That estimate has been debated in its particulars. But nobody argues the exposure didn't happen. Congress passed the Radiation Exposure Compensation Act in 1990, which set up a compensation fund for downwinders, uranium miners, and on-site test participants, with a formal apology written into the statute. It's paid out well over $2 billion. It also excluded whole categories of people

who lived downwind, drew county lines that put some communities in and their neighbors out, and required documentation that many claimants couldn't produce 40 years after the fact. Uranium miners in the Navajo Nation, who worked unventilated mines that the Public Health Service was studying for lung cancer while declining to warn the miners, had particular trouble meeting the paperwork requirements. The public health service study behind that last point began in 1950 and tracked the miners lung cancer rates over time. Researchers involved in it did not tell the miners that the radon exposure in the mines was likely to kill some of them, because informing them was considered likely to disrupt the study and to cause trouble with the mining companies and the

AEC. The miners themselves were the data. The shorthand for all of it is the iron and Triangle. The contractors want revenue. The armed services want programs, budgets, and postings, and the members of Congress want jobs in their districts. Each corner needs the other two, and no corner has any structural interest in a weapons program ending. The most effective single technique in that arrangement is geographic distribution of subcontracts. A large aircraft can be built mostly in two or three plants, or the work can be spread across hundreds of suppliers and dozens of states, which is more expensive and less efficient and makes the program nearly impossible to cancel because canceling it means a member voting against employment in his own district.

Every major prime contractor publishes maps showing exactly which congressional districts benefit from each program. Those maps are marketing material aimed at legislators. The second technique is the revolving door. Officers and civilian officials who manage procurement, retire and go to work for the companies they were buying from. Company executives take senior positions in the department. Both directions are legal, subject to cooling off periods that vary by position and are frequently narrower than people assume. A government accountability office report published in September of 2021, looked at 14 major weapons manufacturers, and counted more than 1,700 former senior officials, acquisition officials, generals and generals hired between 24,

and 2019. That's better than 300 a year. One company alone had over 300 of them on the payroll while holding billions of dollars in defense contracts that year. To be fair to the GAO, its own conclusions were measured. It found the rules had been strengthened and recommended further tightening rather than declaring the system corrupt. That's a structural relationship, not a scandal.

And that's exactly what makes it hard to argue about. The third technique is the, the cost estimate. A program priced honestly over 30 years may never get started. A program priced to get started develops constituencies. And the true number arrives when cancellation would mean writing off everything spent so far. There's a term for this inside the building. Some people call it buying in. Others call it the front loading and political engineering cycle, which is the phrasing the analyst Franklin Spinney used when he laid the pattern out for Congress in the 1980s.

One man tried to say all of this out loud in 1968, and he got run over for it. A. Ernest Fitzgerald was a civilian cost analyst in the Air Force, a deputy for management systems, and he was good at his job in the specific way that gets people in trouble, which is that he actually read the numbers. On November 13th, 1968, Fitzgerald testified before the subcommittee on economy and government of the Joint Economic Committee,

chaired by Senator William Proxmire. He was asked directly about cost growth on the C-5A, Galaxy Transport, being built by Lockheed. He said the overrun could approach $2 billion, and he described technical problems in development. That answer was accurate. It was already known inside the Air Force, and it had not been given to Congress. What followed is documented in the litigation. Within two months, Air Force staff prepared a memorandum for the Secretary of the Air Force. Air Force, laying out three ways Fitzgerald might be removed, one of them a reduction in force.

His performance evaluations changed, and he was assigned to review a Bowling Alley construction project in Thailand. He was told that his civil service tenure had been the result of a computer error. In November of 1969, the Air Force announced a reorganization. It described as an economy measure. His position was eliminated, and he was separated effective the 5th of January, 1970. Around 60 members of Congress signed a letter protesting it. We know what the White House thought about it because it's on the tapes. In a recorded Oval Office conversation, President Nixon discussed Fitzgerald and said he wanted him out, using language that leaves no ambiguity about intent. In another exchange, his aides discussed whether Fitzgerald could be dealt with, and Nixon's response

has been transcribed and published repeatedly. He wanted the man gone. The Civil Service Commission's finding cuts against the clean version of the story. The examiner found that the reorganization, though presented as an economy measure, had actually been motivated by reasons personal to Fitzgerald, and recommended he be put back in his old job, or one of comparable authority. But the Commission explicitly declined to find that he'd been fired in retaliation for the testimony. So the official finding is narrower than the way this case usually gets told. What sits alongside it is the White House tape record

and Nixon's own later acknowledgement of responsibility, neither of which the commission had in front of it. Fitzgerald fought the removal for years. He got back on the payroll in 1973, but not into his old work, and it took further litigation over more than a decade before he held a comparable position again. He sued Nixon personally,

and in 1982 the Supreme Court held in Nixon v. Fitzgerald that a president has absolute immunity from civil damages for official acts. Fitzgerald's case is the reason that precedent exists. He stayed at the Air Force until 2006, and he kept doing exactly what got him fired the first time. He died in 2019 at 92. The C5A, for what it's worth, entered service and flew for decades and did useful work.

That's part of the point. The scandal was the $2 billion that the Congress funding the president, plane wasn't told about, and the punishment fell on the man who mentioned it. 1971 is when the arrangement became explicit. Lockheed, the largest defense contractor in the country, was going broke. The L1011 TriStar Commercial Airliner Program was in trouble. The engine supplier Rolls-Royce had gone into receivership, and the C-5A losses were real. So Lockheed went to Congress and asked the federal government to guarantee $250 million in bank loans.

After a fight, the House passed it on the 30th of July by 192 to 189. The Senate passed it on the 2nd of August, 49 to 48, with a Republican from Kentucky named Marlowe Cook, casting the deciding vote and telling reporters he'd never been enthusiastic about it. Nixon signed it on the 9th. What that vote settled reaches well past the loan. A private corporation was declared too important to be allowed to fail. on national security grounds, four years before anybody used the phrase too big to fail about a bank. Lockheed never actually defaulted.

The guarantee commitment was wound up in 1977, and the fees Lockheed and its banks paid for the privilege netted the treasury somewhere around $30 million, but the principal had been established. The largest defense contractors were now, functionally, public utilities that distributed profits privately. Four years later, the same company produced the biggest corruption scandal in the industry's history. In 1975, the Senate Subcommittee on Multinational Corporations, part of the Church Committee's broader work, started pulling on Lockheed's foreign sales practices. Lockheed's vice chairman, Carl Kochian, ended up testifying about payments to foreign officials and agents, and the company eventually acknowledged something on the order of $22 million in such payments across multiple countries.

In Japan, money moved through a right-wing fixer named Yoshio Kadama, a man with deep ties to organized crime who'd been held as a war crime suspect and released. Payments were made to secure orders for the L-1011 from all Nippon Airways. Prime Minister Kakwe Tanaka, by then out of office, was arrested in July of 1976 on charges of taking 500 million yen. He was convicted in 1983, appealed and died in 1993 with the appeal. appeal unresolved. In the Netherlands, Prince Bernhard, the husband of Queen Juliana, and a serving inspector general of the armed forces, was found by a government commission to have accepted payments

and to have damaged the state's interests. He was never prosecuted, in part because prosecuting him risked the monarchy. He resigned his military and business positions and stopped wearing his uniforms in public. In Italy and West Germany, investigations produced their own casualties and their own denials. West Germany is where the allegations sit next to a body count. In the late 1950s, the Federal Republic chose the Lockheed F-1 104 Starfighter as its principal combat aircraft, and eventually acquired 916 of them, the largest starfighter fleet in the world outside the United States. The aircraft had been designed as a high-altitude interceptor. The Germans configured it as a low level. All weather strike aircraft capable of carrying nuclear weapons in some of the

worst flying weather in Europe, which is not what it was built for. 292 of those aircraft were lost in accidents, and 116 pilots were killed. Eight of them, American exchange pilots. That's roughly a third of the fleet. German crews called it the Vitwinmacher, the Widowmaker. Most analysts who've gone through the accident files conclude the Starfighter was not simply a defective machine and the historiography has moved in the aircraft's favor over the years. The engine was vulnerable to bird strikes in the crowded lowlands. The wing loading was punishing at low altitude and early training was rushed. The comparable German loss rate on an earlier type, the F-84, was actually worse. Stay tuned for more disturbing history. We'll be back after these messages.

What remains true regardless is that a country bought a high-altitude interceptor, and flew it into the weeds for 30 years. And there were allegations examined in the 70s and never conclusively proven that money had changed hands to secure that sale. In 1975, Lockheed paid settlements to survivors of 32 Luftwaffe pilots killed in the type. The direct result of the Lockheed hearings

was the Foreign Corrupt Practices Act of 1977, which made it a federal crime for American companies to bribe foreign officials. That law is real, and it has teeth and it's been enforced. The arms trade has also continued to run on agents, commissions, and offsets ever since. And enforcement actions against defense firms for foreign bribery have kept arriving decade after decade.

Operation Ranch Hand was authorized in late 1961 and flown from January of 1962 until 1971. It sprayed roughly 19 to 20 million gallons of herbicide over South Vietnam and parts of Laos and Cambodia. About 11 million gallons of that was Agent Orange, a 50-50 mix of two herbicides known as 24D and 245T. The purpose was defoliation to strip the canopy that concealed movement and crop destruction to deny food. The problem was a manufacturing contaminant. The process for making 245T produces a dioxin called TCDD as a byproduct and how much of it ends up in the the finished product depends on how hot and how fast the reaction runs.

Wartime production ran hot and fast. The company's new dioxin was dangerous. Occupational exposure at chemical plants had produced a severe disfiguring skin condition called Chloracne, going back to industrial accidents in the 1940s and 50s, including an explosion at a Monsanto plant in Nitro, West Virginia in 1949, and an incident at a BASF plant in Germany in Nicar. 1993. Dow ran toxicology work on dioxin in the early 60s. In March of 1965, Dow hosted a meeting with representatives from other manufacturers of 245T at its headquarters in Midland, Michigan,

where the toxicity of the dioxin contaminant was discussed along with concern about what would happen if the government found out. Records of that meeting surfaced in later litigation. Internal correspondence introduced in those cases in, includes Dow language describing dioxin as one of the most toxic materials the company had encountered. None of that stopped the production lines. Dow, Monsanto, Diamond Shamrock, Hercules, Uniroyal, Thompson Chemical and several others kept supplying the government, and the government kept spraying. The bill came due on the veterans first.

Men who'd walked through defoliated zones or handled the drums or flown the aircraft, started reporting soft tissue sarcomas, non-Hodgkin lymphoma, chloracin, and a cluster of other conditions, and reported birth defects in their children. The Veterans Administration resisted claims for years. Class action litigation against the manufacturers settled in 1984 for $180 million with no admission of liability, which worked out to small individual payments and which many veterans considered an insult. The Agent Orange Act of 1991 finally established presumptive service connection for a list of conditions, and that list has kept growing as the science has accumulated, most recently adding conditions decades after the war ended. The Vietnamese side got almost nothing.

Vietnamese plaintiffs sued the manufacturers in American federal court and had the case dismissed in 2005, a ruling upheld on appeal. on the reasoning that herbicides used as defolients weren't chemical weapons under international law at the time, and that the manufacturers were protected as government contractors. Vietnamese estimates of affected people run into the millions, and those numbers are contested because separating dioxin effects from everything else that happened to that country is genuinely difficult. What isn't contested is the soil. dioxin concentrated at former storage and loading sites,

particularly the air bases at Da Nang and Bienhoa, at levels far above any safety standard, and it stayed there for decades, leaching into fish ponds and sediment. The United States eventually funded remediation at Da Nang starting in 2012, and at Bienhoa starting in 2019, which is 50 years after the spring stopped and only after Vietnamese and American advocates

spent years pushing for it. The other Dow product from that war was Napalm. Specifically, Napalm B, a mix of polystyrene, benzene, and gasoline that burned longer and stuck better than earlier formulations. Dow made it at a plant in Torrance, California, under a government contract that was in revenue terms, almost nothing to accompany that size. When student protests against Dow recruiters spread across campuses in 1966 and 67,

and they were significant. The company's board considered dropping the contract and decided to keep it, taking the position that a company shouldn't refuse a lawful government order in wartime. That was a defensible position, and they made it publicly. It also meant that for a generation of Americans, the name of a chemical company became attached to photographs of burned children. And then there's the ordinance the war left behind. Between 1964 and 1973, the United States dropped over two,

million tons of ordinance on Laos in more than 500,000 bombing runs, in a campaign that was not acknowledged publicly while it was happening. A large fraction of that tonnage was cluster munitions, and each cluster bomb dispensed hundreds of submunitions. The tennis ball-sized bomblets Laotians call, bombies. Something on the order of 270 million of those bomblets came down. Failure rate estimates vary between about 10 and 30% depending on who's counting and under what conditions. And the working figure used by the clearance organizations is that roughly 80 million of them never went off. Those bomblets are still killing people. Since the bombing ended, the recorded casualty

count from unexploded ordinance in Laos runs to 20,000 or more killed and injured, and a large share of the victims have been children because a bombie is small and round and metal and looks like something worth picking up. Clearance operations have been running for over 30 years and have cleared a small percentage of the contaminated land. At the current pace, the work runs past the end of this century. Nobody was bribed in Laos and nobody was being studied. The bombing was the ordinary output of the system doing what it was built to do, on a schedule and at a volume set by production capacity in a country the American public was not told about. Back in Maryland at Edgewood Arsenal on the Chesapeake, the Army was running a different kind of program on its own soldiers.

Between 1955 and 1975, roughly 7,000 American service members volunteered for testing at Edgewood. They were told they'd be helping evaluate protective clothing and equipment, and many later said they understood it as a light duty assignment with a break from routine and a chance at extra leave. What they were actually being given, depending on the study, included nerve agents like sarin and vx, mustard agent, the incapacitating agent BZD LSD, PCP, and a long list of other compounds. The commonly cited figure is more than 250 different chemicals across the life of the program. BZ was the agent the program pursued hardest. It's an anti-colonergic delirient. At effective doses, it produces confusion, hallucinations that the subject cannot distinguish from reality,

and effects that last for days. Soldiers under BZ were filmed, and some of that footage still exists. Men wandered, spoke to people who weren't there, and had no reliable memory afterward. The consent forms these men signed did not name the agents. They couldn't, because the agents were classified. Volunteers were told they could withdraw,

which is a limited kind of freedom when the information required to make the choice, is being withheld by the people asking. Some were required to sign secrecy oaths and understood themselves to be legally barred from discussing what had happened to them, including with their own doctors, for decades afterward. The Army's own Inspector General reviewed the program in 1975. The National Academy of Sciences reviewed it across several volumes in the 1980s. Follow-up on the individual men was inconsistent at best.

Many were never contacted again. A class action lawsuit brought by the, Vietnam Veterans of America eventually forced the Department of Defense and the Department of Veterans Affairs to notify surviving participants of what they'd been exposed to and to provide medical care. And that litigation ran into the 2010s. Some of these men learned in their 70s what had been put into their bodies in their 20s. By the 1980s, the abuses had gotten more mundane, and the money had gotten much larger. The Reagan buildup pushed defense spending up sharply, and with the money came a run of procurement fraud that produced the most famous numbers in the history of government waste.

The $600 toilet seat and the $400 hammer both came out of this period, and both are real, and both are usually misunderstood. The toilet seat was actually a molded plastic and fiberglass cover assembly that fit over the toilet in a Navy P3C Orion anti-submarine aircraft, supplied by Lockheed at $640 a piece for an aircraft no longer in production. When the Pentagon leaned on Lockheed ahead of a congressional hearing, the company cut the price. The hammer was billed to the Navy at $435, and the invoice showed a hammer worth something like $15, carrying more than $400 in allocated engineering overhead. Because the contractor had spread overhead evenly across every line item on a bulk spare parts order. The hammer got the same overhead charge as an engine component. The real problem

is worse than the cartoon version. A gold-plated hammer is a joke, but an accounting system where nobody can say what anything actually costs, where overhead is smeared across line items in a way that hides both the padding and the genuine expense. It's a structural condition, and it's still the condition today. The genuine fraud in that era was prosecuted. Operation Illwind was a multi-year FBI and Naval Investigative Service investigation into bribery in Navy procurement. And on June 14, 1988, agents executed search warrants at dozens of locations simultaneously, hitting contractors, consultants, and Pentagon offices. It produced more than 60 convictions of individuals and companies.

The central figure was Melvin Paisley, a former Assistant Secretary of the Navy for Research, Engineering, and Systems, who took payments to steer contracts and who went to prison. Consultants were selling inside information about competitors' bids. Companies were buying it. Eunice pleaded guilty and paid a fine in the neighborhood of $190 million. In 1985, General Electric pleaded guilty to defrauding the government on a Minuteman missile contract by falsifying employee time cards, more than 100 counts of it. In the same period, Pratt and Whitney, Sperry, Boeing,

Hughes, Northrop, and others all faced charges, settlements, or suspensions for various forms of contract fraud. By the late 80s, a substantial share of the largest defense contractors in the country had been under criminal investigation. And then almost all of them kept their contracts, which is the part of that record I keep going back to. Debarment, meaning exclusion from federal contracting, is the only real deterrent in that world, and it is almost never applied to a major prime for long. because the government needs the aircraft, and there are only a handful of firms that can build them. The penalty becomes a fine. The fine becomes a cost of doing business, and the cost of doing business gets billed back on the next contract. The Cold War ended, and instead of shrinking the way people expected, the machine concentrated.

In July of 1993, Defense Secretary Les Aspen and his deputy William Perry hosted a dinner for defense industry executives at the Pentagon. Perry delivered the message, which was that the post-Cold War budget would not support the existing number of prime contractors and that the department expected consolidation. Accounts of exactly how bluntly it was put vary depending on which executive is telling it, but the industry has referred to that meeting ever since as the Last Supper. John Deutsch, then Undersecretary for Acquisition, was part of the discussion. What followed was one of the fastest consolidations in American and industrial history. Something like 50 major defense contractors became five. Lockheed merged with

Martin Marietta. Northrop bought Grumman. Boeing absorbed McDonnell Douglas and Rockwell's aerospace business. Raytheon took Hughes and Texas Instruments defense electronics. General Dynamics sold off pieces and bought others. The government paid for the mergers. Restructuring costs were treated as allowable charges against defense contracts on the theory that consolidation would save money in the long run. In one case, executives at Martin Marietta and Lockheed received large payouts tied to the merger that were partially reimbursed by the Pentagon, and after the resulting political noise, Congress restricted that specific practice. The broader policy of reimbursing restructuring costs continued. The savings did materialize at first, and then the structural consequence arrived,

which was that the buyer had reduced its own supplier base to a handful of firms in most categories and in some categories to exactly one. The department's own industrial base reports have documented the result over the last decade. Single source and sole source contracting has grown. Competition on major programs is often nominal. When there are two possible builders of a class of ship or aircraft, and both are needed to preserve the industrial base, Neither can be allowed to lose badly enough to exit.

Stay tuned for more disturbing history. We'll be back after these messages. Nobody negotiates hard with a supplier that has been declared in advance, too important to fail. No moral failing on anybody's part is required, because that's the arithmetic of a market with one buyer and five sellers, all of whom the buyer has committed to keeping alive. September 11, 2001 opened the money valve wider than at any point. since the Second World War, and it opened it in a form that had not existed at scale before, because this time, the contractors deployed to the war zone alongside the troops.

The Logistics Civil Augmentation Program, Log Cap, put a single company in charge of the support functions an Army used to do for itself. Housing, food service, laundry, fuel hauling, water treatment, and mail all went to the contractor. Log Cap 3 was awarded to Kellogg Brown and route, then a Halliburton subsidiary, and it eventually paid out in the tens of billions of dollars. Vice President Dick Cheney had been Halliburton's chief executive between his time as Secretary of Defense and his time in the White House, and he retained deferred compensation from the company while in office. He said correctly that he had no role in the contract awards. He'd also been Secretary of Defense when the Department commissioned a Halliburton subsidiary study in 1992 on how

to privatize military logistics, which is the study that helped produce log cap in the first place. At its peak, the number of contractors in Iraq and Afghanistan roughly equaled or exceeded the number of American troops. That had never happened before in an American war at that scale. The Commission on wartime contracting in Iraq and Afghanistan, a bipartisan body created by Congress, delivered its final report in 2011. It concluded that of the roughly two $206 billion spent on contracts and grants in the two theaters. Somewhere between $31 and $60 billion had been lost to waste and fraud. That range is enormous, and the commission was clear about why. Nobody could tell them precisely, because the records didn't support a tighter number.

18 electrocution deaths of American personnel in Iraq were reviewed by the Defense Department Inspector General, which concluded that nine of them were. caused by improper grounding or faulty equipment. One of the nine was Staff Sergeant Ryan Masseth, a 24-year-old Green Beret from outside Pittsburgh, who died on the 2nd of January 2008, in a shower at the Radwania Palace Complex in Baghdad when an improperly grounded water pump energized the plumbing. The Inspector General's finding was that multiple systems and organizations failed and that Masseth had been left exposed to unacceptable risk. KBR, which held the maintenance contract, had inspected that building months earlier and documented improper grounding.

And the company's position was that its contract did not cover fixing hazards it found, only repairs it was tasked to make, and that it was not responsible for his death. The Army's criminal investigation concluded there was insufficient evidence to prove or disprove criminal culpability by anyone, and no charges were brought. Civil litigation by the family ran for years. In 2007, a man named Ephraim Diveroli, barely into his 20s, running a company called AEY, incorporated out of an office in Miami Beach, held an army contract worth up to nearly $300 million to supply ammunition for Afghan security forces.

AEY delivered Chinese manufactured ammunition, some of it dating back to the 1960s, sourced from Albanian stockpiles, repackaged to hide its origin in violation of the contract, and of a ban on Chinese munitions. The company got the contract through the normal competitive process. Di Viroli pleaded guilty to fraud in 2009, and in September of 2007, Blackwater security contractors escorting a State Department convoy, opened fire in Nesore Square in Baghdad,

killing 17 Iraqi civilians and wounding around 20 more. An FBI investigation concluded that at least 14 of those killings were unjustified. The first prosecution collapsed on a technicality involving the use of compelled statements. A second effort produced convictions in 2014. One contractor was convicted of murder and three of manslaughter and weapons charges. In December of 2020, President Trump pardoned all four of them. The reconstruction money went out the door in a manner that has few parallels. In the period around the transfer of sovereignty in 2004,

the coalition provisional authority shipped roughly 20, $12 billion in shrink-wrapped bundles of American $100 bills into Baghdad aboard military cargo aircraft, money drawn largely from Iraqi oil revenues held in the Development Fund for Iraq. Congressional investigators later found that a large portion of it could not be accounted for. The Special Inspector General for Iraq Reconstruction spent years documenting projects that were paid for and never built, or built and never used. The catalog includes a children's hospital in Basra and a prison at Khan Bani Sad that consumed around $40 million and was never completed or occupied. The F-35 is the largest weapons program in history.

The Government Accountability Office estimated in 2024 that the total lifetime cost of the program, including sustainment through the 2080s, would exceed $2 trillion. The program was built on a concept called concurrency, meaning production began while testing was still underway, on the theory that this would save time, and it didn't. Aircraft delivered early required extensive retrofitting. The program was years late and hundreds of billions over its original estimates, and the aircraft is now a real and capable machine flown by many countries, which is precisely why nobody was ever going to cancel it. Work on it is spread across 45 states and hundreds of suppliers. Congress finally required the Department of Defense to submit to a full financial audit, with the first one conducted for fiscal year 2018. It failed, and it has failed every year since. In December of 2025, the department announced the results of the audit for fiscal year 2025,

and it failed for the eighth consecutive time. Auditors identified 26 material weaknesses in its financials. controls. The department reported roughly $4.65 trillion in assets and about $4.7 trillion in liabilities, spread across all 50 states and more than 40 countries. Among the specific findings was that the department had failed to properly account for assets in the F-35 program's global spares pool. The Department of Defense is the only one of the government's 24 major agencies that has never passed an audit. Its stated goal is to achieve a clean opinion by 2028, and its own chief financial officer has acknowledged that meeting that

date will require a significant acceleration of current efforts. A material weakness is not an accusation of theft. It's a finding that the controls are inadequate to detect an error or a theft if one occurred. That's the honest way to put it, and it's also the point. On a budget that has now crossed the trillion dollar mark for a single year, the institution cannot demonstrate to an independent auditor where its property is. In February of 2019, the Defense Department Inspector General released an audit of parts bought from Transdime Group, a company that had spent years acquiring small manufacturers of aircraft components. The auditors looked at 47 spare parts sold across 113 contracts. On 46 of the 47, the company had earned what the department

calls excess profit, with margins running from 17 percent. to more than 4,400 percent. On those 46 parts, which cost the government $26.2 million, the auditors calculated $16.1 million in excess profit. The mechanism was control of information. Contracting officers asked TransDOM for cost data 16 times. The company was not required by law to hand it over, and it declined 15 of those 16 times.

The one time it did provide certified cost data was the one contract that couldn't be awarded without it, and that was the single part in the whole sample where the profit came in under 15%. Transdime's chief executive went in front of the House Oversight Committee in May of 2019 and made an argument worth understanding. The prices were legal. The government had agreed to them. No law had been broken, and his obligation ran to shareholders. He also acknowledged that employees had withheld cost data from contracting officers. Nine days after the hearing, the company refunded the full $16.1 million. Then the committee asked the Inspector General to look at the business model as a whole. That report came out in December

of 2021 and found roughly $20.8 million in additional excess profit on 105 out of 106 part types sold between 2017 and mid-2019. The Inspector General recommended the department seek a refund. and members of Congress called for one. I couldn't establish that the second amount was ever repaid. That's the machine functioning exactly as designed. There's one buyer. The buyer needs the part. The part fits only one aircraft.

And the company that owns the tooling has no obligation to be reasonable. The refunds are a public relations expense. The United States is by a wide margin the largest arms exporter on earth, accounting for something in the range of 40% of global exports by volume, in recent years, according to the Stockholm International Peace Research Institute, and that share has been rising rather than falling. Foreign military sales are administered by the Defense Department itself, which negotiates on behalf of American manufacturers,

which means the government is simultaneously the regulator of the arms trade and its broker. On the 9th of August, 2018, in the town of Dahayan in Saada province in northern Yemen, a bus carrying schoolboys back from a field trip stopped, at a crowded market. An airstrike by the Saudi led coalition hit it just after 820 in the morning. Health authorities in Sayata put the toll at 51 killed, 40 of them children, all under 15 and most under 10, with 79 wounded. Open source investigators who reconstructed the strike put the figures somewhat higher. The Red Cross said its supported hospital received the bodies of 29 children from that one attack. Local journalists at the scene photographed a fragment of the

weapon with its stamped markings visible. And CNN, working with munitions experts, established that the markings identified a 500-pound laser-guided bomb of a type manufactured by Lockheed Martin and sold to Saudi Arabia through a State Department approved transfer. The coalition's first statement called the strike a legitimate military operation carried out in accordance with international humanitarian law and said it was a response to a ballistic missile launched at the Saudi city of Jizan the previous day. Weeks later, after its own incident assessment team reviewed it, the coalition acknowledged the strike had been unjustified and said those responsible would be held accountable. A United Nations panel of experts examined the strike among many others,

and found a pattern of coalition airstrikes on civilian targets in that conflict. American munition sales, to the coalition continued. There was a serious fight about it in Congress, and in 2019, both chambers voted to block certain sales, using emergency authority the administration had invoked to bypass the normal review, and the vetoes held. The Government Accountability Office later reported that the state and defense departments had not adequately tracked whether American supplied weapons were being used in compliance

with the agreements they'd been sold under. There's no way to establish what any individual engineer at that company knew or intended. What the record does establish is that a serial number on a bomb fragment in a Yemeni market trace back through a legal export license to an American production line and that the sale continued after the fragment was photographed. I don't conclude that every defense contractor is a criminal enterprise, because that's lazy, and it isn't true. The people who built the C-5 and the F-35 and the ships and the radars are engineers doing hard work, and some of what this system has produced has kept people alive. including a lot of American service members. What I do conclude is harder to argue with. When war production becomes permanent, it creates a permanent constituency.

And that constituency behaves the way constituencies behave. It defends its budget, hires the people who oversee it, distributes its work to make itself politically indestructible, and classifies what embarrasses it. And when it needs data, it can't otherwise obtain. The record shows that it has repeatedly taken that data, data from people who were not asked, who were not told, and who were selected precisely because they were unlikely to be able to object.

The people it took from were a cement worker in a segregated hospital. Boys at a state school for the feeble-minded, soldiers who signed a form that couldn't legally tell them what it was for, and Islanders 100 miles down wind, of a device whose yield the designers got wrong by a factor of two and a half. Every one of those decisions was made by someone who could have decided otherwise, and And in most cases, the decision was reviewed by lawyers and approved by supervisors and documented in writing, which is how we know about it now. That's the part I keep coming back to as a former investigator. Almost none of this was done in the dark by unknown hands.

Almost all of it is in the files, written down at the time by people who assumed the files would stay closed. Eisenhower knew a fair amount of this. He'd approved the Nevada test series. He was president during the Marshall Island shots and during the resettlement of Rangalap. He presided over the buildup he was warning about. He was the man who'd signed the paperwork, and I think the speech reads differently with that in view. Eight years before the farewell address, in April of 1953, a few weeks after Stalin died, he'd given a speech to the American Society of Newspaper editors that made the same argument in plainer terms.

He said that every gun made and every warship launched signified a theft from those who were hungry and not fed and cold and not clothed. He listed the trades, a modern heavy bomber for 30 schools or two power plants or two fine hospitals. Then he went home and submitted the budget and he kept submitting it for seven more years. The machine he described has outlived every person in this episode. Ernest Fitzgerald went back into that building and kept auditing programs until 2006, and the department he audited has now failed eight consecutive audits of its own, on a budget that has crossed a trillion dollars a year. The system absorbed the man who confirmed one true number,

and it absorbed the president who warned about it, and it is larger today than it was when either of them was alive, which leaves the drafts. They're still in Abilene, in the Eisenhower Presidential Library, the stack that Moos and Williams and Milton and the president himself worked over for a year and a half. Milton Eisenhower said for the rest of his life

that the famous phrase originally had a third word in it, that his brother had planned to warn the country about the military industrial congressional complex, and that the word congressional came out before airtime so a departing president wouldn't spend his last three days in office at war with the legislature. The archivists who keep those drafts have looked. The word is not in any surviving page.

So either Milton misremembered a conversation from a weekend in January of 1961. Or a draft existed that named the third corner of the triangle, the one corner that votes on its own oversight. And that page didn't survive. Every other decision in this story got written down. The injection at Oak Ridge is in a file. The rewritten paperclip dossiers are in files. The Midland meeting, the sheep findings, the memo laying out how to remove Fitzgerald,

all of it is on paper because the people who did those things assumed the paper would stay closed. The one page that would tell us whether Eisenhower tried to name all three partners and decided not to is the page nobody has ever found.