Franklin Roosevelt knew Pearl Harbor was coming and let it happen

The case revisionists have built

Take the revisionist case at its strongest, because serious historians — not just internet skeptics — have argued versions of it for eighty years, starting with Charles Beard's 1948 President Roosevelt and the Coming of the War, 1941 and revived most fully by Robert Stinnett's 2000 book Day of Deceit.

Roosevelt wanted into the war, and said as much privately. By late 1941 he believed Nazi Germany was an existential threat to Britain and, eventually, the United States, but he faced a public and a Congress still committed to neutrality after the trauma of the First World War. A dramatic, unambiguous act of aggression — one that unified the country overnight — was, in cold strategic terms, exactly what his position needed. That motive is not in serious dispute; only what he did about it is.

A 1940 planning memo shows officials thinking in exactly these terms. Naval intelligence officer Arthur McCollum's October 1940 “eight action” memo proposed a set of pressure tactics against Japan and predicted that such moves would likely provoke an attack the United States could then answer. Stinnett argued this amounted to a designed provocation, and several of McCollum's eight proposed actions were, in fact, later carried out.

Washington had broken Japan's codes and had real, specific warning traffic. The MAGIC program was reading Japanese diplomatic cable traffic throughout 1941. Among the intercepts was the so-called “bomb plot” message of 24 September 1941, in which Tokyo instructed its Honolulu consulate to divide Pearl Harbor into a grid and report the exact berthing positions of US warships — a request no previous intercept had ever made. That is a specific, targeting-grade signal, translated in Washington by early October, roughly two months before the attack.

The Hawaii commanders were not given what Washington had, and were then blamed. Admiral Husband Kimmel and General Walter Short, who commanded naval and army forces at Pearl Harbor, were relieved of command within weeks of the attack and were, for decades, treated as the officers primarily responsible for the disaster — even though the MAGIC intercepts were tightly restricted in Washington and not routinely forwarded to them. If information that might have prompted a full alert was kept from the men who needed it, and they alone paid the price, that pattern looks to many like more than an accident.

What the intercepts actually said, and didn't

Set against this case is a mountain of documentary evidence, and it points toward failure, not design.

The single biggest fact is what MAGIC could not read. PURPLE was a diplomatic cipher, used by Japan's Foreign Ministry — which its own military deliberately kept in the dark about the Pearl Harbor plan, precisely to guard against a leak. The operational code that mattered, the Imperial Navy's JN-25, was barely broken at all: US Navy cryptanalysts' own contemporaneous logs show that in the twelve months before the attack, not a single JN-25 message had been fully read, and by December 1941 analysts could recover perhaps 10 to 15 percent of any given message. The attacking fleet sailed under strict radio silence. There was no decrypted order, from any code, that named Pearl Harbor as the target.

The “bomb plot” message was real, but it was one signal among hundreds asking similar questions. Japan's consulates were sending comparable ship-location requests about Manila, Panama, San Diego and other ports throughout 1941; in isolation, a request for more granular detail about Pearl Harbor did not obviously separate itself as uniquely alarming from routine naval intelligence-gathering, especially against a backdrop in which the US high command expected Japan's first blow to fall in the Philippines or Southeast Asia, not Hawaii. Colonel Rufus Bratton, who translated it, later said he considered it significant — but “significant” in a mountain of comparable traffic is precisely the problem, not the solution.

This is the intelligence-failure thesis that has dominated serious scholarship since 1962, when Roberta Wohlstetter's Pearl Harbor: Warning and Decision reframed the whole debate. Her core finding was that the failure was not a shortage of warning signals but an overload of noise: real signals of danger were mixed in with a much larger volume of ambiguous, contradictory or misleading information, and — with attention focused on the Philippines — analysts had no reliable way to isolate the one pattern that mattered from the many that did not. Gordon Prange's exhaustive 1981 At Dawn We Slept, the product of 37 years of research including interviews with Japanese planners, reached the same conclusion from the opposite direction: bureaucratic rivalry between the Army and Navy, a chronic underestimate of Japanese capability, and plain bad luck, not conspiracy.

The specific documentary claims in Stinnett's case do not hold up when checked against the record. Stinnett never explains how Roosevelt could have brought Army Chief of Staff George Marshall and Chief of Naval Operations Harold Stark into a secret conspiracy without any of them ever revealing it — a gap British historian John Keegan flagged directly, writing that the charge “defies logic.” A close reading of the McCollum memo's own text, laid out by military historian Conrad Crane, shows its recommendations aimed at deterring and containing Japan while buying time to prepare, not at engineering a surprise attack on the fleet. And the timeline itself is hard to square with deliberate foreknowledge: on the night of 6 December, Roosevelt personally drafted a direct appeal for peace to Japan's emperor — an odd final act for a man who supposedly knew the attack was hours away and wanted it to succeed.

Why the theory endures

Even readers persuaded that the intelligence-failure account is correct should understand why the foreknowledge theory has never fully died, because its pull is not irrational.

The first driver is hindsight bias. Once you know Pearl Harbor was the target, intercepts like the bomb-plot message look glaringly obvious — of course grid-mapping ship berths meant an attack was coming. But that clarity is only visible looking backward. At the time, analysts were reading that message alongside hundreds of others, with attention and resources tilted toward an expected blow in the Philippines, and no framework for knowing which single cable, out of thousands, was the one that mattered. It is a near-universal feature of surprise-attack post-mortems, seen again after Pearl Harbor's 1962 study became required reading before 11 September 2001: warnings are always easier to find after the disaster than before it.

The second is proportionality. More than 2,400 Americans died in two hours, the Pacific Fleet was crippled, and the country was hurled into a global war — a catastrophe that feels too large to have resulted from mere miscommunication between rival Army and Navy bureaucracies. A deliberate plot restores the emotional proportion between cause and effect in a way that a chain of ordinary human errors does not.

The third is that part of the surrounding story is genuinely true, which lends borrowed credibility to the rest. Washington really did possess codebreaking capability the Hawaii commanders lacked; it really did withhold some of what it had; Kimmel and Short really were relieved of command and blamed for years for a failure that was substantially not theirs alone — both men were posthumously exonerated by Congress of dereliction of duty in 1999. And Roosevelt undeniably wanted the United States in the war against Hitler. None of that proves foreknowledge of the specific attack — but each true thread makes the larger theory feel more plausible than a cold read of the documentary record supports.