On the evening of January 17, 1961, three days before he handed the presidency to John F. Kennedy, Dwight David Eisenhower sat at a desk in the Oval Office and read a farewell address into a bank of television cameras. He was seventy years old, a five-star General of the Army, the man who had commanded the largest amphibious invasion in human history and then governed the nation for eight years of Cold War. He had, by any measure, spent his entire adult life inside the institution he was about to warn the country against.
Near the end of the speech, in a passage that aides had worked over for more than a year through some twenty-one drafts, he said: "In the councils of government, we must guard against the acquisition of unwarranted influence, whether sought or unsought, by the military-industrial complex. The potential for the disastrous rise of misplaced power exists and will persist." The phrase entered the language permanently and has never left it.
What is less often remembered is that the speechwriting drafts — preserved in the Eisenhower Library and shaped by the speechwriter and political scientist Malcolm Moos and his assistant Ralph Williams — read, through much of their development, "military-industrial-congressional complex." The third word was struck before delivery. Eisenhower, by the surviving account, did not wish to leave office appearing to insult the legislative branch whose cooperation his successor would need. The edit is the warning in miniature: even the man cautioning the republic against the machine trimmed his caution to spare one of the machine's three arms. The branch that holds the purse had quietly excused itself from the indictment.
The temptation is to read the farewell address as the lament of a peacenik, and it is the opposite. Eisenhower was no dove. He had presided over the largest peacetime military buildup in American history, authorized the U-2 overflights, greenlit the coups in Iran and Guatemala, and built the nuclear arsenal into the tens of thousands of warheads. He believed in deterrence and in strength. Precisely because of that, his warning carried a weight no critic from outside the system could match. The man describing the danger had built much of the apparatus that posed it.
He was not saying the military was evil or the contractors corrupt. He was describing a structural danger — the emergence, for the first time in American history, of "an immense military establishment and a large arms industry" that had become permanent. "Until the latest of our world conflicts," he said, "the United States had no armament industry." American plowshares had been beaten into swords only in wartime and beaten back afterward. That cycle had ended.
His record backed the words. When the Gaither Report of 1957, written by a panel of defense-minded outsiders in the panic after Sputnik, urged a crash program of fallout shelters and a massive expansion of strategic forces, Eisenhower largely refused it, suspecting — correctly, as the U-2 imagery would confirm — that the Soviet threat was being inflated by men with an interest in inflating it. His "New Look" defense posture had deliberately leaned on nuclear deterrence precisely to hold down the cost of conventional forces, "more bang for the buck," resisting each service's perennial demand for more. He was not warning in 1961 about a danger he had indulged. He was warning about a pressure he had spent eight years pushing back against, and which he feared his successors would lack the standing or the will to resist.
The Cold War made mobilization permanent, and a permanent arms economy creates a permanent constituency for its own expansion — three and a half million people directly employed in the defense sector by 1961, the federal government spending more on military security annually than "the net income of all United States corporations" combined. The danger was not a coup. It was that the combined weight of the armed services, the firms that supplied them, and the legislators whose districts depended on the contracts would, in aggregate and without anyone intending it, bend national policy toward the permanent preparation for war. Eisenhower's prescription was not disarmament but vigilance: "an alert and knowledgeable citizenry." He was warning that the machine, left unwatched, would grow whether or not the country needed it to.
The farewell address was not the first time he had said it. Eight years earlier, in his April 16, 1953 "Chance for Peace" address before the American Society of Newspaper Editors, delivered weeks after Stalin's death, Eisenhower had laid out the moral arithmetic with a directness that has never been matched by a sitting president: "Every gun that is made, every warship launched, every rocket fired signifies, in the final sense, a theft from those who hunger and are not fed, those who are cold and are not clothed. This world in arms is not spending money alone. It is spending the sweat of its laborers, the genius of its scientists, the hopes of its children." A single heavy bomber, he calculated aloud, cost the equivalent of a modern brick school in more than thirty cities, or two electric power plants, or two fully equipped hospitals. The man who would warn of the complex in 1961 had already, in 1953, priced its product in schools, hospitals, and power plants not built. The opportunity cost was, to him, the whole point.
The complex is not a conspiracy and does not require one. It is a set of incentives that all point in the same direction. The first is the revolving door. A general officer manages a weapons program, retires, and is hired by the contractor whose program he oversaw; a contractor executive is appointed to a senior Pentagon acquisition post and writes the requirements his former firm is positioned to fill. A 2018 study by the Project On Government Oversight tallied hundreds of senior Defense officials and military officers passing through the door into the top contractors; Boeing, Lockheed Martin, Raytheon, General Dynamics, and Northrop Grumman together sustain a standing population of former flag officers and acquisition officials. The personnel on both sides of the table are, across a career, the same people.
The second mechanism is the contract itself. Major weapons systems are routinely procured on a "cost-plus" basis — the government reimburses the contractor's costs and adds a guaranteed fee on top. The structure inverts the ordinary discipline of commerce: overruns do not punish the seller, they enlarge the fee. The F-35 Joint Strike Fighter, the most expensive weapons program in history, is projected to cost more than 1.7 trillion dollars across its life cycle, years behind schedule and over budget at nearly every milestone — and yet politically unkillable, because Lockheed Martin deliberately distributed its subcontracts across forty-five states and most congressional districts.
The pathology is old and well documented from the inside. In 1968 a Pentagon cost analyst named A. Ernest Fitzgerald told a Senate subcommittee that Lockheed's C-5A transport was running two billion dollars over its estimate; for the candor he was stripped of his duties and ultimately fired, in a retaliation that Richard Nixon later admitted on tape ordering — "get rid of that son of a bitch." Fitzgerald spent years in the courts winning his job back, and became the patron saint of a small, recurring genre of insider disclosure that the system absorbs and outlasts. The 1980s produced its iconic artifacts: the Pentagon paying, by its own auditors' accounting, four hundred and thirty-five dollars for a hammer and over six hundred dollars for a toilet-seat cover, the totems of a procurement culture in which price discipline had simply ceased to operate. And the revolving door occasionally turns criminal: in 2004 Darleen Druyun, the Air Force's top acquisition official, went to federal prison after admitting she had favored Boeing on a tanker-lease deal worth tens of billions while negotiating a job with the company for herself and her family. These are not the system breaking down. They are the system's incentives expressed without their usual discretion.
That distribution is the third mechanism made visible: every job is a vote. A representative who moves to cut a system cuts payrolls at home, and so the systems do not get cut. The Government Accountability Office's annual assessment of major weapons programs has for years found cumulative cost growth across the portfolio measured in the hundreds of billions of dollars, with schedule slips routine and performance shortfalls common — and yet the programs almost never die, because the political cost of cancellation is borne by legislators while the financial cost of continuation is borne by a taxpayer who does not see the line item. The fourth and deepest mechanism is the threat premium. The entire apparatus is funded in proportion to the danger the nation perceives, which means every institution within it has a standing incentive to perceive danger generously — to inflate the adversary's capabilities, to discover missile gaps and bomber gaps that later prove illusory, to find in every regional conflict an existential stake.
The supposed "missile gap" that dominated the 1960 campaign was, the Eisenhower administration already knew from U-2 photography, fictional; the Soviets had a handful of operational ICBMs, not the hundreds claimed. But the gap had already done its work on the budget by the time the photographs settled the question. Manufactured or merely exaggerated, the threat is the raw material the complex converts into appropriations. This is why threat-inflation is not incidental to the system but central to it — the incentive runs in only one direction, and the institutions that benefit from danger are the same institutions trusted to measure it.
The estimating itself was institutionalized. The RAND Corporation, founded in 1948 as an Air Force–funded research arm before spinning off as a nominally independent think tank, became the prototype of an ecosystem of analysts, contractors, and foundations whose business was to quantify the Soviet threat — and whose largest patrons had a direct interest in the answer coming back large. The "bomber gap" of the mid-1950s and the "missile gap" of 1960 both originated in estimates congenial to the services that wanted the buildup, and both were later shown to be inversions of reality: it was the United States, in each case, that held the decisive lead. The threat is not always cynically invented. More often it is sincerely overestimated by people whose careers, institutions, and worldview all incline them to see the worst case, in a system that punishes the analyst who guesses too low far more harshly than the one who guesses too high.
The fifth mechanism greases the other four: money in politics. The major contractors maintain among the largest corporate lobbying operations in Washington, collectively spending well over a hundred million dollars a year, and they direct campaign contributions with precision toward the members of the Armed Services and Defense Appropriations committees who write their budgets. The traffic is legal, disclosed, and routine. Its effect is to align the financial interest of individual legislators with the expansion of the very programs they are supposed to scrutinize. When Congress, unable to trust itself to shut surplus bases, created the Base Realignment and Closure (BRAC) commission in 1988 — an independent body whose recommendations the legislature could only accept or reject as a single package — it was a confession in institutional form: the people's representatives had to tie their own hands to do what the parochial interest in every district made otherwise impossible.
Political scientists gave the arrangement a name: the "iron triangle," the mutually reinforcing relationship between the armed services that want the weapons, the contractors that build them, and the congressional committees that fund them, each leg sustaining the other two. Gordon Adams documented its mechanics in The Politics of Defense Contracting: The Iron Triangle (1981), tracing how the same eight companies captured the bulk of procurement dollars decade after decade regardless of which were performing well. The most candid measure of where the incentives lead came from inside the industry itself. Norman Augustine, the longtime chief executive of Lockheed Martin, observed half-seriously in his collection Augustine's Laws that unit costs of military aircraft had risen so steadily for so long that, extrapolated forward, "in the year 2054, the entire defense budget will purchase just one aircraft," to be shared between the Air Force and Navy three and a half days each per week. The joke is an admission: the cost curve has no internal brake, because nothing in the structure rewards restraint.
Eisenhower named the thing in 1961, but the diagnosis was older and, in its earlier forms, far angrier. In 1935, Major General Smedley Darlington Butler — twice awarded the Medal of Honor, at the time of his death the most decorated Marine in U.S. history — published a slim, incendiary pamphlet titled War Is a Racket. "War is a racket," it opens. "It always has been. It is possibly the oldest, easily the most profitable, surely the most vicious."
Butler wrote from the inside with a bluntness no civilian could claim: "I spent thirty-three years and four months in active military service... And during that period I spent most of my time being a high-class muscle-man for Big Business, for Wall Street and for the Bankers. In short, I was a racketeer, a gangster for capitalism." He itemized it — Mexico for American oil interests, Haiti and Cuba for National City Bank, Nicaragua for the Brown Brothers banking house, the Dominican Republic for sugar, Honduras for the fruit companies. This was the same Butler who, in 1934, testified to Congress that a cabal of financiers had approached him to lead a coup against Franklin Roosevelt — the episode now called the The Business Plot. The man who blew the whistle on a plot to militarize American politics was also its sharpest analyst of why such a plot would be profitable.
Two decades later the sociologist C. Wright Mills supplied the theoretical architecture. The Power Elite (Oxford University Press, 1956) argued that American power had concentrated in three interlocking hierarchies — the corporate rich, the political directorate, and what Mills called the "warlords," the ascendant senior military — whose members were increasingly interchangeable, schooled alike, intermarried, and rotating among the commanding heights. The "military ascendancy," Mills wrote, had reached a point where the warlords sat at the center of national decision-making for the first time in American history, and a "permanent war economy" had become the normal condition of the country rather than an emergency to be ended.
Mills himself was nearly forgotten in the conformist 1950s and then resurrected by the movement that read The Power Elite as prophecy. The New Left of the 1960s — the Students for a Democratic Society whose 1962 Port Huron Statement borrowed his vocabulary — took up the warlords-and-corporate-rich framing and aimed it at the war in Vietnam, and the phrase "military-industrial complex," handed down from a Republican president, became a chant at antiwar demonstrations. That a Marine general's pamphlet, a radical sociologist's treatise, and a five-star president's farewell converged on the same charge is why the idea proved impossible to quarantine on either flank: it was the consensus suspicion of a country that had just built the largest war machine in history and was uneasy about what it had made.
Butler had company in his own decade. From 1934 to 1936 the Senate Special Committee Investigating the Munitions Industry, chaired by the North Dakota Republican Gerald Nye, took testimony from arms makers and bankers and concluded that the firms supplying weapons in the First World War had lobbied for conflict, bribed foreign officials, and profited enormously from the slaughter. The phrase "merchants of death," drawn from a 1934 bestseller by H. C. Engelbrecht and F. C. Hanighen, entered the language; the Nye Committee's work fed directly into the Neutrality Acts of the late 1930s. The American suspicion that arms and bankers manufacture the wars that enrich them is not a 1960s invention. It was mainstream enough in 1936 to drive federal legislation.
Butler supplied the witness; Mills supplied the map; Eisenhower, five years after Mills, supplied the phrase — and the unimpeachable authority of a man who had run the whole apparatus and could not be dismissed as a leftist or a crank. The line of argument runs from a Marine general through a Republican-led Senate investigation and a radical sociologist to a Republican president, which is part of what makes it durable. It was never confined to one ideological corner.
What Mills called the permanent war economy is now, for much of the country, simply the economy. After 1945 the United States never demobilized to a peacetime footing. The decisive turn came in 1950, when Paul Nitze's National Security Council paper NSC-68 argued for a permanent, tripled defense budget to contain the Soviet Union — a document that languished until the outbreak of the Korean War that June supplied the political moment to enact it. From then on the budget contracted somewhat between conflicts and resumed climbing through the missile buildup, Vietnam, the Reagan rearmament, and the wars after 2001. The economist Seymour Melman, in The Permanent War Economy (1974) and a string of works before and after, documented how military production had become a self-sustaining sector that consumed the nation's engineering talent and capital while producing nothing that re-entered the civilian economy as wealth — a vast, permanent, state-managed enterprise wearing the costume of free-market contracting.
The industry's concentration made the dependency structural. At a 1993 Pentagon dinner that defense executives came to call "the Last Supper," Deputy Secretary of Defense William Perry told the assembled contractors that the post-Cold War budget could no longer support all of them and that the government would subsidize mergers. Within a few years some fifty prime contractors had consolidated into five — Lockheed Martin, Boeing, Raytheon, Northrop Grumman, and General Dynamics — each too large and too geographically dispersed across congressional districts to be allowed to fail. The "free market" in defense became an oligopoly of five firms selling almost exclusively to one buyer, a buyer politically incapable of refusing to buy.
September 11 opened a second front for the complex in the intelligence and surveillance economy. The reporters Dana Priest and William Arkin, mapping what they called "Top Secret America" for the Washington Post in 2010, found that some 1,900 private companies were working on counterterrorism, homeland security, and intelligence at over 10,000 locations, that nearly a third of the people holding the highest clearances were contractors, and that no single official could say with confidence how much the whole enterprise cost or whether it made the country safer. The arrangement is the same iron triangle ported to a new threat: a permanent danger, a forest of firms paid to answer it, and an oversight apparatus that cannot see the whole. Edward Snowden, when he copied the NSA's surveillance archive in 2013, was not an NSA employee at all. He was a contractor for Booz Allen Hamilton, a consulting firm that earns the large majority of its revenue from the government it was, in effect, surveilling on behalf of.
Vietnam was its proving ground. The The Gulf of Tonkin Incident incident of August 1964 — an exchange whose second night, by the later admission of the National Security Agency's own historians, almost certainly never happened — was inflated into the Gulf of Tonkin Resolution that authorized a decade of escalation, hundreds of billions in spending, and revenue for every prime contractor on the order of battle. That the same Joint Chiefs had, two years earlier, formally drafted Operation Northwoods — a signed proposal to stage terror attacks against Americans to manufacture a pretext for war with Cuba — establishes that the institutional willingness to generate a casus belli was not a paranoid hypothesis but a documented appetite.
The manufactured atrocity recurs as a method because it works. In October 1990, a fifteen-year-old girl identified only as "Nayirah" testified before the Congressional Human Rights Caucus that she had watched Iraqi soldiers pull Kuwaiti babies from hospital incubators and leave them on the floor to die. The testimony was cited by President George H. W. Bush and by senators in the debate that authorized the Gulf War. It was later revealed that Nayirah was the daughter of the Kuwaiti ambassador to the United States, that her account was unverified and almost certainly false, and that her appearance had been arranged by the public-relations firm Hill & Knowlton, retained by a Kuwaiti government lobby. The incubator story was a product, manufactured to a specification, and it did exactly what threat-inflation is meant to do: it converted a contested foreign policy into a moral emergency, and the emergency into appropriations and ordnance.
The pattern recurred at the largest scale in 2003. The invasion of Iraq was sold on weapons of mass destruction that did not exist — see The Iraq WMDs & PNAC — and its largest material beneficiaries were the contractors: Halliburton and its subsidiary KBR booked tens of billions in no-bid reconstruction and logistics work, while Lockheed, Raytheon, and Boeing supplied the munitions and platforms. Dick Cheney, who as vice president was the war's most forceful internal advocate, had been Halliburton's chief executive in the years immediately before taking office — the revolving door turning at the very summit of the decision to invade.
Afghanistan supplied the longer record of the same dynamic. The Special Inspector General for Afghanistan Reconstruction (SIGAR) and the trove of interviews published as the "Afghanistan Papers" by the Washington Post in 2019 documented two decades in which officials privately conceded the war was unwinnable and the reconstruction money largely wasted while publicly insisting on progress. The Costs of War Project at Brown University estimates the post-9/11 wars in Iraq and Afghanistan ran past 8 trillion dollars. The Stockholm International Peace Research Institute (SIPRI) records U.S. military spending climbing past 800 billion dollars annually in the 2020s — more than the next nine nations combined, China included. Defense-sector equities, meanwhile, outperformed the broad market across the war years; the firms were structurally indifferent to whether the wars were won, only to whether they continued.
And the institution cannot say where the money goes. Mandated by law since the Chief Financial Officers Act of 1990 to produce an auditable financial statement, the Department of Defense did not attempt its first full agency-wide audit until fiscal year 2018. It failed. It has failed every consecutive year since — seven straight as of the November 2024 audit — unable to fully account for assets its own Inspector General values in the trillions. The largest single line in the discretionary budget is, by its own admission, unauditable. A system that cannot be counted cannot be checked, and a system that cannot be checked is precisely the "misplaced power" against which the farewell address warned.
Eisenhower had already supplied the moral accounting in 1953, and the permanent war economy is best understood as the steady running of that ledger in reverse. Every dollar that flows into a weapons program that overruns by design, into a war that the people prosecuting it privately know is lost, into bases that exist because no legislator can close them, is — in the 1953 formula — a school not built, a hospital not equipped, a power plant not financed. The genius of the nation's scientists and the sweat of its laborers are routed, decade after decade, into hardware much of which is never used and some of which the government cannot even locate on its own books. This is the cost the deterrence argument has to be weighed against: not merely the wars fought, but the civilian future continuously deferred to fund a readiness that has its own appetite and no natural ceiling.
The case against the complex is powerful, but the case for the apparatus it describes is not weak, and an honest map must draw it at full strength. The first argument is deterrence. The permanent military establishment Eisenhower built and warned about presided over the longest peace among great powers in modern history; no nuclear weapon has been used in anger since 1945, and the Soviet Union collapsed without the third world war that every prior multipolar rivalry had eventually produced. Peace through strength is not a slogan invented by contractors — it is the explicit logic under which deterrence is held to have worked, and its proponents argue that the absence of catastrophe is precisely the product you cannot find on a budget line.
The second argument is spillover. The same defense research apparatus that critics indict produced an extraordinary share of the modern world's civilian technology. ARPA, the Pentagon's research arm, funded the packet-switched network — ARPANET — that became the internet. The Global Positioning System was a U.S. Air Force constellation built to guide weapons; it now underlies civilian navigation, agriculture, and the timing of financial markets. Jet aviation, radar, nuclear power, satellite communications, and a long list of materials and computing advances came out of defense money that no private market would have risked at that scale. The economist Mariana Mazzucato, no apologist for militarism, argues in The Entrepreneurial State that the most consequential technologies of the era were funded by exactly this kind of mission-driven state spending — the iPhone in your hand is a basket of inventions the Pentagon paid to invent.
Silicon Valley itself was a defense project before it was anything else. The integrated circuit, invented at Texas Instruments and Fairchild Semiconductor in 1958–59, found its first volume buyers not in any consumer market but in the guidance computer of the Minuteman II nuclear missile and the Apollo Guidance Computer — government programs willing to pay almost any unit price for miniaturized reliability, and in paying it, driving the chip down the cost curve that eventually put one in every household. The transistor radio, the jet airliner, microwave technology, modern aviation electronics: each descends from a military requirement that a civilian market would not, on its own, have funded into existence. The honest version of the indictment has to concede this. The machine that kills also invents, and some of what it invents outlives every war it was built for.
There is a sharper rebuttal to the spillover argument, and an honest map must give it too: the fact that defense funding produced the internet does not prove defense was the efficient way to produce it. Melman's retort was that a nation can fund civilian research, infrastructure, and basic science directly, without routing the money through weapons it must then find wars to use — that the spillovers are real but are a defense of public R&D, not of procurement and combat. The internet did not require Vietnam. That said, the counter's defenders answer that mission-driven, deep-pocketed, patient state spending of the kind only a security imperative reliably musters is exactly what produced the breakthroughs, and that peacetime civilian budgets have rarely shown the same tolerance for expensive, decades-long failure on the road to a GPS or a jet engine. The argument does not resolve cleanly, which is the mark of a real one.
The third argument is the most important methodologically: the complex is a real interest group, but it is not an omnipotent one. It loses. Major weapons systems are cancelled — the Army's Comanche helicopter, the Navy's A-12 Avenger, the Crusader artillery program — over the furious objection of contractors and the legislators who carry their water. The Cold War buildup was followed by a genuine "peace dividend" in the 1990s that cut the force and the budget substantially. If the complex truly controlled policy, neither the cancellations nor the dividend could happen.
The disciplined version of the thesis, made by the historian Andrew Bacevich in The New American Militarism (Oxford University Press, 2005) — himself a West Point graduate and retired Army colonel who lost a son in Iraq — is not that a cabal runs the country but that Americans, across both parties and the broad public, have come to see military power as the truest expression of national greatness, and have built institutions that faithfully serve that preference. There is also the matter of the alliances. The same forward-deployed force that critics call imperial overreach is, to much of the democratic world, the security guarantee under which it chose not to build its own. Extended deterrence — the American nuclear umbrella over NATO, Japan, and South Korea — is widely credited with suppressing the nuclear proliferation and regional arms races that a vacuum of American power would invite. Whatever one thinks of the cost, the public good is real and is purchased in carriers and bases; allies free-ride on it precisely because they trust it exists. The complex's defenders argue that this stabilizing function, like deterrence itself, is invisible in the indictment because its product is a war that did not happen.
On this reading the complex is not a hidden hand overriding the public will; it is the public will, given budget and steel. The contractors do not have to manufacture the appetite for war. They only have to be standing where it lands.
Eisenhower's most important sentence is not actually the one with the famous phrase in it. It is the one that follows: "We must never let the weight of this combination endanger our liberties or democratic processes. We should take nothing for granted." The verb is guard, not abolish. He did not believe the military establishment could or should be dismantled — he had built much of it and considered it necessary in a dangerous world.
He believed it had to be watched, continuously, by a citizenry alert enough to tell the difference between the security the nation genuinely required and the security the apparatus was structurally inclined to oversell. Sixty-five years later, the instruments of that vigilance — the audits, the oversight committees, the press — strain against an apparatus that spends more than ever, cannot balance its own books, and has discovered in the open-ended emergencies of the present century a permanence Eisenhower could only have feared.
The complex is also where the abstraction of the The Deep State becomes concrete and countable. The permanent national-security bureaucracy that survives every election does not float free of material interest; it is anchored in contracts, payrolls, and the revolving door, and it is funded through a budget so large and so opaque that no elected body fully tracks it. When critics describe a government that pursues the same foreign and military policy regardless of who wins, the contractor base and the appropriations machinery are a large part of what they are describing — the economic gravity that bends policy in a consistent direction no single administration sets or can reverse. Eisenhower's "combination" and Lofgren's "deep state" are, at this point, two names for overlapping halves of one structure.
The struck word in the draft was congressional, and its absence remains the most honest thing about the speech: the branch that was supposed to hold the purse is the branch that distributes the contracts. The fusion of the corporation with state violence that the contractor embodies — the form examined under Corporate Personhood & The Corporation — and the global military primacy that underwrites the dollar order described under The Petrodollar System are the same machine seen from two angles. Whether that machine is a danger to the republic or the republic's own considered preference is the question Eisenhower left open, and the question the unbalanced ledger keeps reopening every November the Pentagon fails its audit again.