The Order, the Bridge, and the Fire

At 4:45 in the afternoon on Thursday, July 28, 1932, Herbert Hoover signed a written instruction to his Secretary of War, Patrick J. Hurley. The paper directed federal troops to assist the District of Columbia police in clearing former soldiers from governmental property along Pennsylvania Avenue between the Capitol and the White House. The president’s directive contained a specific geographical limit. Troops were authorized to clear the partially demolished buildings near Third Street where former servicemen had taken shelter. They were not authorized to cross the Eleventh Street Bridge over the Anacostia River. They were not authorized to enter the main veterans’ encampment on the muddy flats across the water. They were not authorized to burn anything.

By 11:00 that night, every one of those limits had been violated. Cavalry under Major George S. Patton had charged civilians on Pennsylvania Avenue. Infantry under Major General Douglas MacArthur, the Army Chief of Staff who had personally taken field command in dress uniform with a row of decorations across his chest, had pushed across the bridge with bayonets fixed and tear gas canisters arcing through the summer air. The wooden shanties at Camp Marks on the Anacostia flats were burning. Two infants would die from the gas exposure in the days that followed. A bystander, William Hushka, lay dead from a District police bullet fired earlier in the day at Pennsylvania Avenue. Another veteran, Eric Carlson, lay mortally wounded. The newspapers the next morning would show photographs of mounted soldiers driving unarmed men in patched uniforms through clouds of gas, the Capitol dome visible in the background of several shots.

The order the president signed had been disobeyed in its specifics and in its spirit. The order Hoover would own for the rest of his life had not been the one he wrote. This article reconstructs how a written limit became an unlimited deployment, who decided what at each step, and why the gap between paper authorization and field execution mattered less to American voters in November than the photographs they had seen in late summer. The reconstruction draws on the actual written order to Hurley, on the contradictory accounts Douglas MacArthur and Dwight Eisenhower left in their respective memoirs, on Hoover’s own version in Memoirs Volume 3, on the War Department’s official report, and on a careful reading of the Washington Post and New York Times coverage from July 29 through August 1. It engages the scholarly disagreement between Donald Lisio, whose archival reconstruction in The President and Protest established the command-chain failure as a matter of documentary record, and Paul Dickson and Thomas Allen, whose narrative history The Bonus Army: An American Epic treats the White House as responsible regardless of what the field commander did or did not exceed.

The verdict here has two parts. First, the documentary evidence supports Lisio: MacArthur exceeded explicit written limits, did so knowingly, and falsified his account afterward to obscure the fact. Second, Lisio’s exoneration cannot rescue Hoover politically or historically because civilian control of the military is precisely the principle by which presidents own what generals do on their authority, particularly when he declines to publicly repudiate the excess afterward. The 1932 election was probably already lost on the economic numbers. July 28 made it ungettable. The day functions as a case study in what happens when an executive issues a constrained order to a subordinate who has incentives to expand its scope, and then refuses to discipline the expansion when it occurs.

Setup: Why Seventeen Thousand Veterans Came to Washington

The deferred pay that drew the Bonus Expeditionary Force to Washington in the spring of 1932 was not a Depression-era invention. It originated in the 1924 World War Adjusted Compensation Act, passed over the veto of Calvin Coolidge. The Act issued every World War I veteran a “service certificate” representing dollar-a-day pay for stateside service plus a quarter dollar per day for overseas service, with compound interest, payable in 1945. The twenty-one-year delay reflected actuarial calculations about veteran longevity and Treasury preferences for predictable obligations. In 1924, with the economy expanding and the veteran cohort still mostly under thirty, a 1945 payout date carried little urgency. By the spring of 1932, that calculation had changed entirely.

National unemployment reached 24 percent during the second quarter of 1932. The figure understated regional collapse. In Toledo, Ohio, four out of five industrial workers were unemployed. In Chicago, public school teachers had gone unpaid for eight months. In rural counties across the wheat belt, foreclosure auctions were drawing armed neighbors who bid pennies to return the farm to its owner. The certificates that promised the men payment in 1945 had, in this environment, a present-value problem. A man whose family was being evicted in 1932 could not eat the actuarial future. The men of the American Expeditionary Forces had been promised something concrete by the country they had fought for, and the promise was scheduled for redemption thirteen years after they had stopped being able to wait.

A Texas Democratic congressman named Wright Patman, a freshman from the Texarkana district who had himself served as a machine gunner in France, introduced legislation in 1929 to authorize early payment of the certificates. The Patman Bill went through several variants over three years. By the spring of 1932 it had taken its decisive form: full early payment in cash, financed by issuance of approximately $2.4 billion in new currency. The currency-issuance feature mattered because it gave the proposal a partisan coloration beyond simple veteran relief. To monetary conservatives, including most of the Republican congressional leadership and almost the entire banking establishment, financing payment by printing money was inflationary heresy that would damage the gold-standard credit structure the executive administration was attempting to preserve. To agrarian populists and Patman himself, the inflationary effect was a feature rather than a defect, because mild inflation would reduce the burden of fixed-dollar farm debt.

Walter W. Waters, an unemployed Oregon cannery superintendent who had served as a sergeant with the 146th Field Artillery in France, organized the first march from Portland in May 1932. Three hundred men set out riding the rails eastward. By the time they reached East Saint Louis, railroad police had pulled them from boxcars and Illinois Governor Louis Emmerson had ordered the state militia to escort them by truck to the Indiana line. Word of the march spread through the doughboys’ newspapers and through the informal networks of the unemployed. By early June approximately seventeen thousand men were converging on Washington from every state, many bringing wives and children. The Bonus Expeditionary Force, as Waters christened the contingent in conscious echo of the AEF designation under which most of them had served fourteen years earlier, established its main camp on the Anacostia flats, a marshy lowland across the river from the Navy Yard. Smaller groups occupied partially demolished federal buildings along Pennsylvania Avenue scheduled for replacement under the Federal Triangle redevelopment.

The camp at Anacostia was governed by Waters and a council of veteran representatives under rules that mirrored service discipline. Communists were ejected. Bathing was required. The American flag flew over the camp. Sunday services were held. The District police, under Superintendent Pelham Glassford, a former brigadier general who had himself served with the AEF in France and who knew Waters personally from veteran reunion circuits, treated the camp as a public-order responsibility rather than a national security threat. Glassford lent the BEF members cooking equipment, helped them obtain food donations, and maintained daily communication with Waters about the camp’s operations.

The Patman Bill Fails: June 15 Through June 17

The legislative crisis came in three days during the middle of June. On June 15, the House of Representatives passed the Patman Bill by a vote of 211 to 176. The margin was decisive enough to demonstrate House sentiment but well short of the two-thirds required to override the sitting executive veto. The marchers gathered outside the Capitol heard the result through portable radios. Approximately ten thousand attended an impromptu rally near the East Front. Waters spoke. Glassford spoke and was applauded. The mood was celebratory but contingent: the Senate vote two days later would decide everything.

On June 17, the Senate rejected the Patman Bill by a vote of 62 to 18. The margin was crushing in a way the House vote had not foreshadowed. The Senate vote demonstrated that the bill was not coming back in any form the president would be required to veto. The administration’s position had prevailed without the administration needing to expend public-opinion capital on a veto fight. The ex-servicemen outside the Capitol learned the result around 9:30 in the evening. Waters addressed the crowd by megaphone. The former soldiers sang “America” and dispersed in orderly fashion. The Senate vote, in retrospect, was the inflection point of the entire crisis. The electoral question had been answered. The legal question had been answered. What remained was the human question of what would happen to seventeen thousand men who had no work, no place to go, and now no reputational objective.

Most observers, including Glassford and most of the press, expected the encampment to dissolve over the following weeks. Some marchers did leave. Congress passed and the president signed a $100,000 appropriation for transportation home, structured as a loan against future bonus payment. Approximately five thousand men accepted the loan and departed. Approximately twelve thousand remained at Anacostia and along Pennsylvania Avenue through the second half of June and into the summer. The reasons for remaining varied. Some had no home to return to. Some hoped the lame-duck Congress in December might reconsider. Some had become invested in the camp itself as a community. A small number, perhaps a few hundred, were Communists or Communist sympathizers seeking to use the camp for propaganda purposes; Waters and his council continued to expel them when identified.

Why Hoover Would Not Meet Them

Hoover’s refusal to meet the protesters personally has become, in retrospective accounts, one of the iconic campaign-related miscalculations of the twentieth century. The refusal deserves careful examination because the standard treatment, which depicts him as aloof and tone-deaf, captures something true but misses the strategic logic that produced the choice.

The administration’s position was that no White House occupant should set the precedent of negotiating with a mass protest gathering. The reasoning was articulated by Theodore Joslin, his press secretary, in a July 1 memorandum: if the executive met with a group that had pressured Congress and lost, the precedent would invite every defeated lobby to assemble on the Mall and force a presidential audience. The constitutional channel for petitioning the government, Joslin argued, was through elected representatives, and Congress had decided. To negotiate with extraconstitutional pressure would compromise the separation of powers.

The argument was not absurd. It was, in fact, the standard executive position in the nineteenth century and through most of the twentieth. Lyndon Johnson did not personally meet with the 1963 March on Washington organizers as a group; he met with leadership separately afterward at the White House. Richard Nixon did not personally address the November 1969 Moratorium. The precedent that presidents should personally engage with mass demonstrations is a feature of the late twentieth and early twenty-first centuries, not a baseline that the executive violated.

What Hoover did wrong was something more specific. He failed to authorize any senior administration figure to engage with Waters. He failed to develop a public message about why the administration opposed the service compensation that distinguished his position from indifference to veteran suffering. He failed to project public sympathy for the genuine economic distress that brought the doughboys to Washington while opposing the specific legislative mechanism they sought. The substantive case against the Patman Bill, that monetary financing of the deferred bonus would damage the credit structure needed for broader recovery, could have been made publicly with empathy. It was not. The administration retreated into procedural correctness. Veterans on the flats heard nothing from the executive branch but silence and through the District police a request that they go home.

The personal dimension matters too. The incumbent, by all surviving accounts, was a man of profound but constrained emotional range. His humanitarian work feeding Belgian children during the Great War and Russian peasants during the 1921 famine demonstrated genuine compassion for suffering. But the compassion expressed itself through systematic relief operations rather than through public displays of empathy. The Quaker temperament, the engineer’s preference for solving rather than emoting, the introvert’s discomfort with crowds, all combined to produce a chief executive who could feel deeply for the marchers while presenting to the public the appearance of glacial indifference. The gap between inner state and public projection was not unique to him, but in the summer of 1932 it became politically fatal in a way it might not have been at any other moment.

The Buildup to July 28: Tension on Pennsylvania Avenue

Through the first three weeks of late summer, the situation at the federal-buildings encampment along Pennsylvania Avenue and Third Street became progressively unstable. The buildings in question were not the Anacostia camp. They were a separate occupation by perhaps two thousand veterans of partially demolished structures on governmental property scheduled for replacement under the Federal Triangle project. The Treasury Department, which controlled the property, had been requesting that the ex-servicemen vacate since mid-June. Glassford had been negotiating gradual evacuation. The former soldiers on the national property were generally more radical than those at Anacostia. Communist organizers had penetrated the Pennsylvania Avenue contingent in ways that Waters’s discipline at Camp Marks had prevented.

On July 21, Treasury Secretary Ogden Mills sent Hoover a memorandum requesting national authorization to clear the buildings. Hoover declined, preferring that Glassford handle the eviction through the District police. The District commissioners were instructed to schedule the clearance for a weekday morning when the governmental workforce would be present and witnesses available. The eviction was originally scheduled for July 22, postponed to July 26, postponed again to July 28. The delays reflected Glassford’s preference for orderly negotiation and the administration’s reluctance to force a confrontation it would have to own.

On July 28 at 10:00 in the morning, Glassford and his officers began the eviction at Third Street and Pennsylvania Avenue. The first phase proceeded without major affair. Veterans were given time to gather possessions. Some left peacefully. Tensions rose around mid-morning as more men arrived from Camp Marks across the river to support the resistance. By early afternoon a crowd of perhaps two thousand had gathered on Pennsylvania Avenue. Bricks were thrown from a building under demolition. Glassford himself was struck on the shoulder by a brick.

At approximately 1:45 in the afternoon, two District police officers, Miles Zamcheck and George Shinault, were confronted on the second floor of a partially demolished building by a group of approximately a dozen the men. The encounter at the top of the stairs has been variously reported. The most credible reconstruction, by Dickson and Allen drawing on the District police inquiry, indicates that one officer was being beaten with a club when Shinault drew his revolver and fired. Two the protesters were hit. William Hushka, a Lithuanian-born former national armed forces private from Saint Louis, was killed instantly by a shot to the heart. Eric Carlson, a former the Regular forces private from Oakland, California, was hit and would die several days later from his wounds. Hushka, the District police would later determine, had been carrying a club but had not himself been the man beating the officer. The shootings ended the eviction immediately. Glassford pulled his officers back. The crowd dispersed slowly. By 2:30 in the afternoon the immediate confrontation was over, but the partisan situation had been transformed.

The Cabinet Meeting and Hoover’s Order

At 2:55 in the afternoon, District Commissioner Luther Reichelderfer telephoned Hoover at the White House. Reichelderfer reported that District police could no longer maintain order in the affected area and requested central troops to clear the governmental property. He convened an immediate meeting with Secretary of War Hurley, Attorney General William Mitchell, Treasury Secretary Mills, and several aides. The meeting lasted approximately ninety minutes.

The meeting’s deliberations have been reconstructed from the recollections of Joslin (his press secretary, in his 1934 book Hoover Off the Record), Hurley’s own later statements, Hoover’s Memoirs Volume 3, and contemporaneous notes by aide Lawrence Richey. The core question was whether to deploy federal troops at all. The case for deployment was that the District police had requested central-government assistance, that two men were dead, that the federal property had to be cleared of an occupation the legal owner sought to terminate, and that delay would permit Communist organizers to claim a victory. The case against deployment was that any troops intervention against civilians, regardless of how constrained, would be politically devastating during an election year.

Hoover ultimately decided to deploy. The decision was constrained by specific limits that he had Hurley reduce to writing. The written order, transmitted from Hurley to the general at approximately 2:55 p.m. and surviving in the War Department archives, reads in its critical portion: “You will have United States troops proceed immediately to the scene of disorder. Surround the affected area and clear it without delay. Turn over all prisoners to the civil authorities. In your orders insist that any women and children who may be in the affected area be accorded every consideration and kindness. Use all humanity consistent with the due execution of this order.”

The order’s geographic scope was specified by reference. “The affected area” meant the area where the disorder had occurred, which was the federal property along Pennsylvania Avenue between Third Street and Sixth Street. It did not mean the Anacostia camp, which was across the river and where no disorder had occurred. The order’s instruction to surround and clear “the affected area” was a limit, not a license. Multiple later reconstructions, including Lisio’s archival work, confirm that the president and Hurley both understood the order in this restricted sense, and that both expected the senior officer to execute it accordingly.

A second written communication followed approximately three hours later. As Douglas’s troops moved toward the Pennsylvania Avenue site and as the situation continued to develop, Hoover became concerned that the general might exceed the geographic limit. At approximately 6:00 in the evening, Hurley telephoned Colonel Clement Wright at the War Department with explicit instructions to find the general and order him not to cross the Anacostia bridge. Wright reportedly delivered the message to MacArthur in person at approximately 9:30 p.m., when the troops had pushed the doughboys back from Pennsylvania Avenue toward the bridge. A second messenger, Major General Perry Miles, delivered the same instruction at approximately 10:00 p.m. By that point his forces had begun crossing the bridge.

MacArthur Takes Command in Person

The decision that converted a constrained eviction into a military intervention against the main veterans’ encampment was the senior officer’s. The Chief of Staff was not required by regulation, by tradition, or by common sense to take personal field command of a domestic crowd-control mission. The deployment involved approximately six hundred soldiers from the Sixteenth Infantry, the Third Cavalry, and the Sixth Tank Company. The senior officer present, before Douglas’s intervention, would have been Brigadier General Perry Miles, who commanded the Sixteenth Infantry Brigade. Routine procedure called for the brigade commander to command the deployment. The top general is an administrative position concerned with the service’s overall doctrine, organization, and budget; he is not a field commander in the operational sense.

The senior officer appeared at the staging area near the Ellipse at approximately 4:30 in the afternoon in full dress uniform, mounted decorations across his chest, accompanied by his aide Major Dwight D. Eisenhower. Eisenhower would later record his immediate concern. In At Ease, the 1967 memoir Eisenhower published after his presidency, he wrote that he protested directly to MacArthur about the appropriateness of the top officer personally commanding the exercise and about wearing full dress uniform for a domestic action. Eisenhower’s specific recollection: “I told that dumb son-of-a-bitch he had no business going down there.” the general, Eisenhower recorded, dismissed the objection. He would command. The senior commander would wear the uniform he chose. The top general understood what he was doing.

his explanation, given to Hurley in person before the maneuver and later restated in Reminiscences, was that the situation was potentially revolutionary and required senior command attention. The “potentially revolutionary” framing was the senior officer’s, not the incumbent’s, not Hurley’s, and not supported by any intelligence assessment of the encampment’s actual composition. Glassford’s reports throughout that period had described the encampment as a politically defeated petition movement composed primarily of unemployed petitioners, with a Communist fringe that Waters had been actively suppressing. Douglas’s framing of the action as suppression of incipient revolution would shape his execution decisions through the rest of the day. It would also shape his memoir account for the next thirty-two years.

The deployment began at approximately 4:45 in the afternoon. The Third Cavalry, with Major Patton in field command of the cavalry element, advanced along Pennsylvania Avenue from the Capitol toward the Federal Triangle eviction site. Federal workers leaving offices for the day stopped on sidewalks to watch. The advance was supported by infantry with bayonets fixed and by approximately five small tanks. Tear gas was deployed as a primary crowd-control measure. The ex-servicemen on Pennsylvania Avenue, outnumbered and outgunned, fell back. Some retreated south toward the eastern bank of the Anacostia River. Some retreated west toward the Mall. By approximately 8:00 in the evening, the Pennsylvania Avenue area had been cleared.

This was the intervention Hoover had authorized. It was over by 8:00 p.m. What happened next was not the mission he had authorized.

The Bridge Crossing

At approximately 9:00 in the evening, the senior officer ordered his forces to advance across the Anacostia River by way of the Eleventh Street Bridge. The order was given in the presence of Major Eisenhower, who later recorded his own protest. Eisenhower’s recollection in At Ease emphasized that two separate messages from the White House, conveyed first by Colonel Wright and then by Brigadier General Miles, had explicitly forbidden crossing the bridge. the federal forces top officer, by Eisenhower’s account, claimed not to have received the messages, or claimed not to have received them clearly, or claimed that operational necessity overrode them.

The most credible reconstruction, drawing on Lisio’s archival work, supports the conclusion that the general received the instructions and chose to ignore them. The Wright message had been delivered in person. The Miles message had been delivered in person. The content of both was unambiguous. his later claim, in Reminiscences and in interviews, that he had not received clear orders not to cross is contradicted by Lisio’s documentary findings, by Eisenhower’s contemporaneous account, and by Miles’s own later statements to service historians.

What the general received instead was an opportunity. The former soldiers falling back from Pennsylvania Avenue were not regrouping at Camp Marks for any organized purpose; they were going home, to the only home they had, which was the encampment on the flats. The Chief of Staff framed the retreat as a continuing threat that had to be pursued to its source. The framing was operationally and politically false. The marchers were beaten. They were not regrouping. The Anacostia camp was a settlement, not a base of operations. Crossing the bridge to attack a settlement was a separate exercise against a separate target on a separate piece of property under separate civil jurisdiction.

The troops crossed at approximately 9:30 in the evening. The crossing was opposed by the small contingent of the men still capable of resistance, but the resistance was symbolic rather than effective. By 10:30 p.m. the troops had reached Camp Marks. By 11:00 p.m. tear gas was being deployed against shanties in which families were sleeping. By 11:30 p.m. the shanties were on fire. The source of the fires has been disputed for ninety years. The most likely explanation, supported by War Department and District police investigations, is that the protesters set their own shanties alight as a final gesture of resistance, knowing the camp was lost. A smaller portion of the fires almost certainly came from tear gas canisters igniting flammable shelter materials. A few sources have alleged that soldiers deliberately set fires, but the evidence for direct military arson is weak. The electoral reality was that the fires were rising from a camp under military attack, which the photographs the next morning made impossible to spin.

The casualties from the night were lower than first reports suggested but real. Two infants, Bernard Myers (three months old) and an unnamed child reported by the Washington Post, died from the gas exposure in the days that followed. Approximately fifty-five marchers and a smaller number of bystanders received injuries serious enough to require hospital treatment. Several hundred received minor injuries from tear gas, bayonet wounds, and contusions. The two men killed by District police earlier in the day, Hushka and Carlson, brought the total directly attributable to the maneuver to two adults and two infants, with hundreds injured.

The partisan casualty was different in kind. The morning newspapers of July 29 showed photographs of mounted cavalry charging unarmed civilians on the streets of Washington. They showed columns of smoke rising from the Anacostia flats. They showed the petitioners in patched AEF uniforms being escorted away from their burning shelters with their families. They showed Patton on horseback with sabre drawn. The photographs were not staged, they were not retouched, and they could not be denied. American voters in every region of the country saw them by July 30. The administration’s public-opinion position, already weakened by three years of economic catastrophe, collapsed.

MacArthur’s Press Conference

At approximately 11:00 p.m. on July 28, with the camp still burning behind him, the field commander held an impromptu press conference at the Munitions Building. The conference, conducted with Hurley standing alongside, was attended by approximately a dozen reporters. Douglas’s statements that night, preserved in transcripts and in immediate published accounts, framed the action as the suppression of an incipient revolution.

The general’s key claims that night: the ex-servicemen had been infiltrated by Communist organizers seeking to overthrow the government; the situation had reached a point at which the District police could no longer maintain order; the federal government had acted with restraint to prevent insurrection; the country had been spared a revolutionary moment by decisive military action.

Each of these claims was false or misleading. The Communist infiltration was real but had been small, marginal, and resisted by Waters’s council. The District police had requested national assistance only for clearing the federal property, not for general crowd suppression and not for crossing the bridge. The federal government had not authorized the deployment that actually occurred. The country had not been on the brink of revolution.

his framing served two functions. It justified his own decisions, including the decisions that had exceeded his orders. It also positioned the intervention as a defense of constitutional order rather than as the suppression of a defeated petition movement. The framing required the former soldiers to be revolutionaries; if they were merely unemployed petitioners who had lost their legislative fight, the military mission was without justification. The framing required the incumbent and Hurley to back the framing publicly, because the alternative was to publicly admit that the Chief of Staff had exceeded his orders during a domestic operation that killed civilians.

Hoover and Hurley backed the framing. This was the second decision of the day that destroyed him politically. The first was authorizing the operation at all. The second was failing to publicly repudiate the senior officer’s excess once it was apparent.

Why Hoover Did Not Repudiate MacArthur

The president’s failure to disavow MacArthur’s bridge crossing is, after the underlying authorization, the most consequential single feature of the entire events. The failure has been explained in various ways by various scholars. Each explanation has merit. None is fully satisfying.

The military-civilian loyalty explanation. the incumbent may have believed that public repudiation of the top general during an ongoing domestic crisis would damage the military’s institutional position and weaken civilian control rather than affirm it. This explanation, drawn from his later self-defense in Memoirs Volume 3, has the virtue of consistency with his general administrative philosophy. He preferred to handle subordinate failures privately rather than publicly. He believed that an executive who undermined his own appointees in public lost the moral authority needed to lead them. This was the management style that had served him at Commerce and during his humanitarian work; it was less well suited to domestic electoral crisis.

The reputational resignation explanation. By the morning after, with the photographs in every newspaper, the incumbent may have concluded that the campaign-related damage was already done and could not be reversed by repudiating the general. Disavowing the operation would have admitted that the governmental response had been excessive without altering the public perception that the national response had been excessive. This view, suggested by Joslin in private and visible in his increasingly fatalistic correspondence through August and September, treats the failure to repudiate as recognition that the election was lost.

The substantive defense explanation. He may have actually believed, or persuaded himself to believe, MacArthur’s framing of the operation. The Communist infiltration of the Pennsylvania Avenue contingent was real even if exaggerated. The 1932 election cycle was full of genuine concerns about partisan extremism in Europe and at home. An executive confronting Depression-era unrest may have wanted to believe his military top officer that the operation had prevented something worse. This explanation, less flattering to the president, treats the failure to repudiate as a failure of skepticism.

The personal-loyalty explanation. MacArthur had been Hoover’s appointment to Army Chief of Staff in November 1930. The president had backed the general against considerable Senate skepticism. Disavowing the general in July 1932 would have been a personal repudiation of an appointment Hoover had defended. The Quaker temperament, which dislikes public conflict, would have weighed against the disavowal independently of its electoral costs.

Whatever the mix of motives, the consequence was the same. The president owned the operation that the general had executed. He owned the decisions he had not made because he refused to publicly distinguish them from the decisions he had made. The partisan price was unlimited. The November result was a 57 percent loss to Franklin Roosevelt, the worst loss for an incumbent since Taft in 1912.

The Election: November 8, 1932

The 1932 election was not lost on July 28. The Depression had been making it unwinnable for the incumbent since the 1930 midterms, when Republican losses had returned a Democratic House and a near-tied Senate. The agricultural collapse, the bank failures, the unemployment numbers, all of these had been crushing Republican prospects since 1930. The 1932 election was probably lost on the economic numbers alone, with or without the veterans’ movement.

The question is the margin. He received 39.7 percent of the popular vote to Roosevelt’s 57.4 percent. The Electoral College margin was 472 to 59. Polling does not exist for 1932 in the modern sense, but historians have constructed approximate state-by-state shifts using economic-only models. The economic-only models predict Hoover losing in 1932 by approximately twelve to fifteen percentage points. The actual margin was nearly eighteen points. The gap between the economic-only prediction and the actual margin is the rough measure of how much July 28 cost Hoover beyond what the economy alone would have produced.

Three to five points of margin loss does not change the outcome of an election that was unwinnable by twelve. But it does change the character of the loss. A twelve-point loss would have looked like a normal landslide of the FDR coalition gathering itself. A near-eighteen-point loss looked like national repudiation. A near-eighteen-point loss took Republican Senate seats and Republican House seats and Republican state legislative seats that an economic-only loss would have preserved. A near-eighteen-point loss positioned FDR to claim a mandate broader than economic recovery, and that mandate would license the New Deal’s scope across regulatory, fiscal, and social policy domains. The Bonus force crisis, by adding public-opinion revulsion to economic distress, helped produce the electoral conditions for the next decade’s policy revolution.

The Command-Chain Diagram

The findable artifact for this article is a command-chain diagram showing the gap between the president’s written authorization and MacArthur’s actual execution. Each box represents an action; each arrow represents an order or a deviation. The diagram makes visible at each step what was authorized and what was done.

The first level is the incumbent’s 2:55 p.m. written order to Hurley, authorizing federal troop deployment to assist District police in clearing the federal property where disorder had occurred. The geographic scope was Pennsylvania Avenue between Third and Sixth Streets, including the partially demolished federal buildings being occupied by the marchers. The instruction to use “all humanity consistent with the due execution of this order” set a standard of restraint.

The second level is Hurley’s transmission of the order to Douglas MacArthur, with verbal elaboration. Hurley’s recollection in later statements indicated that he had personally emphasized to the field commander that the Anacostia camp was not authorized as a target and that the bridge was not to be crossed. MacArthur’s response, by Hurley’s account, was acceptance of the limit.

The third level is MacArthur’s tactical execution of the authorized operation. Cavalry under Patton advanced along Pennsylvania Avenue. Infantry pushed the men south and east. Tear gas was deployed. By approximately 8:00 p.m. the federal property had been cleared. This portion of the operation, while harsh in execution, fell within the president’s authorization.

The fourth level is Hurley’s 6:00 p.m. transmission, via Colonel Wright, of an explicit instruction not to cross the bridge. The instruction was delivered in person by Wright to Douglas he at approximately 9:30 p.m.

The fifth level is the second transmission, via Brigadier General Miles, of the same instruction. The instruction was delivered in person by Miles to the general at approximately 10:00 p.m.

The sixth level is MacArthur’s decision, made between approximately 9:30 and 10:00 p.m., to cross the bridge despite both transmissions. The decision was MacArthur’s personally. Eisenhower’s protest is documented. The Wright transmission is documented. The Miles transmission is documented. The decision to cross was deliberate and informed.

The seventh level is the operation against Camp Marks. Tear gas deployment against family shanties. The burning of the camp, whether by veterans, by tear-gas-canister ignition, or by some military arson element. The expulsion of families across the flats and into the surrounding woods. This portion of the operation fell entirely outside the president’s authorization.

The eighth level is MacArthur’s press conference at approximately 11:00 p.m. on July 28. The framing of the operation as suppression of incipient revolution. The claim of Communist infiltration as the operative threat. The implicit assertion that the entire scope of the operation, including the bridge crossing, had been authorized at the highest level.

The ninth level is the absence, over the following weeks, of any public repudiation by the incumbent of the steps that exceeded his order. The silence allowed MacArthur’s framing to stand as the administration’s framing. The silence, more than the original authorization, defined his reputational ownership of the night.

The diagram, when traced step by step, reveals the structure of the failure. The president authorized a constrained operation. The top officer received clear instructions about the limits. The federal forces leader exceeded those limits in real time. The president, when the exceedance became apparent, did not publicly correct it. Each step is a documented decision by a specific actor. Each step has surviving evidence. The reconstruction is not a matter of interpretation but a matter of who decided what and when.

The Complication: Could Hoover Have Repudiated MacArthur?

The strongest counterargument to the analysis above is that he was politically unable to repudiate the Chief of Staff even if he had wanted to. The argument has several variants and each deserves examination.

Variant one: the institutional cost. Public disavowal of the Commander of Staff during the operation’s immediate aftermath would have created an open conflict between the president and the senior military officer. The federal military’s institutional standing, already shaken by the Depression-era budget cuts and reduced visibility, would have suffered. The civil-military relationship, which depends on mutual trust between the executive and senior officers, would have been damaged. the president, a former Belgian relief administrator who understood institutional fragility, may have valued the Regular forces’s institutional position more than his own campaign-related survival in the immediate aftermath.

This variant has merit but underestimates the alternatives. the incumbent could have requested MacArthur’s resignation privately, accepted the resignation, and announced a transition without ever publicly characterizing the general’s actions as exceeding his orders. Eisenhower’s later observation in At Ease, that the president should have insisted the general write a report acknowledging the order’s geographic limits, suggests an internal mechanism that would have permitted private correction without public conflict. The institutional-cost argument explains why the president did not publicly humiliate the Chief of Staff. It does not explain why the incumbent did not privately remove him.

Variant two: the partisan-resignation interpretation. By the next morning, with the photographs ubiquitous, the political damage was done. Repudiating the general would have admitted excessive force without changing the public’s view that excessive force had been used. The two interpretations the public could draw, that the president had ordered excessive force or that the president had failed to control his subordinate, were both politically devastating. The first might be slightly worse, but only slightly. Either way the election was lost.

This variant has merit if one accepts that Hoover believed the election was lost by July 29. Some of his correspondence from August through October suggests something like this fatalism. But other correspondence suggests the incumbent continued to believe through October that some recovery was possible if economic conditions stabilized. The fatalism reading may be hindsight; in real time Hoover was still campaigning.

Variant three: the substantive endorsement reading. the president may have actually believed MacArthur’s framing of the operation. The Communist infiltration was real even if exaggerated. The 1932 atmosphere was anxious enough about extremism that the general’s “potentially revolutionary” framing could have seemed plausible. the incumbent, who lacked direct military experience, may have deferred to MacArthur’s judgment about the operational situation. This deference, while politically costly, would have been consistent with a civilian executive’s general respect for senior military judgment.

The variant explains the failure to repudiate but does not exonerate the president. Civilian control of the military is not consistent with reflexive deference to military framing of domestic incidents. The general’s framing was empirically incorrect and contradicted by the District police’s own assessments. A president who deferred to it without independent investigation failed in his constitutional role.

Variant four: the personal-loyalty reading. the field commander was the president’s appointment. Disavowal would have been personal repudiation of a recent decision. The Quaker temperament, the preference for handling difficulties privately, the institutional conservatism, all weighed against public disavowal.

This variant captures something true about the incumbent’s character but illustrates exactly why his presidency failed. The presidency requires the capacity to override personal-loyalty instincts when public-interest considerations demand it. Hoover’s failure to override the instinct in July 1932 is part of a broader pattern of executive constraint that defined his administration through the entire Depression.

The complication, fully developed, does not rescue the president. It clarifies the structure of his failure. He authorized a constrained operation. The constraint was exceeded by his subordinate. He failed to publicly distinguish the authorization from the execution. He failed to privately remove the subordinate. Each failure was understandable given his character. Each failure was politically and historically consequential.

The Verdict

The reconstruction supports four specific conclusions. Each is defensible on the documentary record. Each carries a citation to the surviving evidence.

First, MacArthur exceeded the president’s written orders in crossing the Anacostia bridge and operating against Camp Marks. The order was specific in its geographic scope. The two follow-up communications from the White House were specific in their bridge-crossing prohibition. MacArthur’s later claim that he had not received clear instructions is contradicted by Lisio’s archival work, by Eisenhower’s contemporaneous account, and by the statements of the messengers who delivered the prohibitions. The exceedance was deliberate and documented.

Second, the president’s failure to publicly repudiate MacArthur’s exceedance made him politically responsible for the operation he had not authorized. The civilian-control-of-military principle works in both directions: the president owns what the military does on his authority, including the parts that exceed that authority, unless he publicly disavows them. the incumbent’s silence converted MacArthur’s freelance operation into the administration’s operation. The conversion was not legally required; it was politically and historically the consequence of his choice not to disavow.

Third, the political effect of July 28 was not the determining cause of the 1932 election outcome but was the determining cause of the margin. The election was almost certainly going to be lost on economic numbers. The election was lost by 17.7 percentage points rather than by 12 percentage points because of the photographs and the failure to respond to them. The five-point delta translated into broader Republican losses, a larger FDR mandate, and political conditions favorable to the New Deal’s eventual scope.

Fourth, the episode illustrates a general structural feature of executive command. Presidents who authorize constrained operations by subordinates with strong personalities and political ambitions of their own should expect the constraints to be tested. The expectation does not require the president to refuse all constrained operations, but it does require the president to be prepared to publicly repudiate exceedances when they occur. The cost of advance preparation is borne by the executive’s relationship with the subordinate. The cost of failing to prepare is borne by the executive’s political standing. the president chose the second cost without realizing he was choosing it.

The verdict is not that the incumbent ordered the burning of the BEF camp. He did not. The verdict is that Hoover, having authorized an operation whose scope was exceeded, declined to defend the limit publicly, and thereby made the exceedance his own. The 1932 election was over because the incumbent did not defend the line he had personally drawn.

Legacy: How July 28 Shaped What Came After

The BEF episode shaped American political development through several specific channels over the following decades. Tracing the channels distinguishes between the affair’s direct effects and the broader pattern it illustrates.

The direct legacy includes the eventual payment of the cash payment. The veterans who marched in 1932 did receive their compensation, though not until 1936. The Adjusted Compensation Payment Act of 1936 was passed over Roosevelt’s veto, with substantial Republican support, and disbursed approximately $1.7 billion to roughly 3.5 million veterans. The payment came four years after the Bonus Expeditionary Force had been driven from Anacostia. Most of the men who had marched received their pay checks in the mail at addresses across the country. The political logic of the 1936 payment owed much to the memory of 1932; Congress was determined that there would be no second march on Washington over the same issue.

The direct legacy also includes the GI Bill of Rights, formally the Servicemen’s Readjustment Act of 1944. The legislation, supported by President Roosevelt and championed by the American Legion, was explicitly framed by its proponents as the alternative to a postwar repeat of the Bonus force crisis. The G.I. Bill’s structure, with education, mortgage, and unemployment benefits available at the time of demobilization rather than deferred to a future actuarial date, reflected the lesson Congress had learned from 1932. Veterans should be compensated when their need was greatest, not when an actuarial calculation suggested. The G.I. Bill became one of the most consequential pieces of mid-century domestic legislation, with direct effects on the growth of suburbs, the expansion of higher education, and the formation of the postwar middle class. Its origin lay partly in the smoke rising from the Anacostia flats on the night of July 28, 1932.

The direct legacy includes the rehabilitation of Walter Waters and the marginal figures who had organized the Bonus Expeditionary Force. Waters returned to Oregon, took occasional menial work, served in the troops during the Second World War, and died in obscurity in 1959. Glassford was forced out of his District police position in October 1932, partly for his sympathetic handling of the veterans; he later served as a labor mediator and as a brigadier general in the Army Air Corps during the Second World War. The veterans themselves dispersed back to their states and gradually disappeared into the broader stream of Depression-era unemployment, though the Bonus March featured prominently in their own memoirs and family stories for the rest of their lives.

The indirect legacy is more complex. The events shaped how subsequent presidents thought about deploying military force in domestic crowd-control settings. Roosevelt’s handling of the 1933 second Bonus March, with the offer of work through the Civilian Conservation Corps and a personal visit by Eleanor Roosevelt to the smaller 1933 encampment, was a deliberate contrast to Hoover’s handling. Truman’s reluctance to use federal troops during the 1946 railway strike, and his eventual cautious use during the 1948 air force strike, reflected awareness of the march on Washington precedent. Eisenhower, who had been MacArthur’s aide at Anacostia, used federal troops at Little Rock in 1957 with extraordinary care about scope, command, and public framing; his memoir explicitly drew the lesson from 1932 that domestic military deployment required precise authorization and clear limits.

The military side of the legacy is its own story. The Bonus Army crisis damaged MacArthur’s reputation in some quarters but burnished it in others. The conservative press generally praised his decisive handling of what it described as a revolutionary threat. The liberal press generally condemned his excessive force. The split set a pattern that would continue through MacArthur’s career. His Pacific command during the Second World War, his administration of postwar Japan, his strategy in Korea, and finally his firing by Truman in April 1951, all reflected the same characterological pattern visible at Anacostia: a senior military officer with strong political views, willing to expand his operational scope beyond civilian authorization, willing to frame his decisions publicly in terms that put his civilian superiors in difficult positions. The line from the Anacostia bridge to the firing in April 1951 is direct and instructive. Truman, who had been a Missouri politician with no special love for generals, understood MacArthur’s pattern from the president-era precedent. When the same pattern reappeared in Korea, Truman did what the incumbent had not done. He repudiated the general publicly, removed him from command, and absorbed the political cost. The principle of civilian control was reasserted at the moment of its sharpest test, and Truman’s success in 1951 owed something to Hoover’s failure in 1932.

The economic legacy connects to the broader pattern of presidents losing reelection during severe downturns. The pattern is empirically robust across two centuries: Van Buren in 1840, Cleveland in 1888, the president in 1932, Carter in 1980, Bush Sr. in 1992. In each case, economic distress provided the fundamental cause and a single dramatic episode provided the iconic image. For Van Buren, the bank failures of the Panic of 1837. For Cleveland, the labor unrest preceding the 1893 panic. For the incumbent, the veterans’ movement. For Carter, the failed Desert One rescue mission in April 1980. For Bush Sr., the 1990 budget summit reversal on taxes. The incidents did not cause the losses; the economic conditions caused the losses. But the incidents gave the losses their shape, their narrative, and their margin.

The administrative legacy includes the eventual professionalization of crowd control and the development of less-lethal tactics. The 1932 operation deployed cavalry with sabers, infantry with bayonets, tanks, and tear gas. By the early 1960s, federal crowd-control doctrine had evolved to emphasize containment over dispersal, negotiation over force, and graduated response over immediate escalation. The evolution drew on multiple sources including British colonial experience, urban riot control developed during the 1960s civil rights crises, and the lessons Eisenhower and his generation had drawn from witnessing or participating in operations like the one at Anacostia. The professionalization was incomplete and uneven; it did not prevent the 1970 Kent State shootings or the 1992 Los Angeles riots’ military deployment. But the trajectory from 1932 to the present runs through specific operational and doctrinal reforms, each of which can be traced back through the chain of lessons learned to the night of July 28.

The broader administrative pattern connects to the contrast with FDR’s Hundred Days response the following spring. Roosevelt’s first hundred days, beginning March 4, 1933, deployed the executive branch with a speed, scope, and public messaging discipline that his administration had never managed. The contrast between Hoover’s careful, restrained, procedural approach and Roosevelt’s improvisational, expansive, communicative approach defined a generation of expectations about presidential leadership during crisis. The Hundred Days was the policy answer to the question the administration had failed to answer: what would the federal government do for Americans in trouble. The Bonus Army episode, in its way, had been the dramatized statement of the president’s negative answer.

The historiographical legacy connects to the long failure of attempts to rehabilitate the incumbent. The 1932 incident is a recurring obstacle in every Hoover rehabilitation attempt. The rehabilitations emphasize the president’s pre-presidential humanitarian record, his post-presidential service under Truman and Eisenhower, his policy innovations such as the Reconstruction Finance Corporation, and his fundamental decency as a person. Each rehabilitation attempt is forced to address the Bonus Army incident, and each addresses it incompletely. The incident’s combination of executive failure and military excess does not fit the rehabilitation narrative. Burner, Wilson, Smith, Clements, and Whyte have each attempted to contextualize the incident in ways that preserve their broader sympathetic portrait. The contextualizations have not moved the needle in scholarly ranking surveys. the incumbent remained roughly thirtieth in the 2021 C-SPAN survey, having moved little since the 1962 Schlesinger survey first established the pattern of Depression-era opprobrium. The incident is not the only reason for the persistent low ranking. But it is a recurring presence in the explanations.

The Disagreement Among Historians

The scholarly literature on the Bonus Army incident is shaped by the foundational disagreement between Donald Lisio’s archival reconstruction and the broader narrative tradition. Understanding the disagreement clarifies what is in dispute and what is settled.

Donald J. Lisio’s The President and Protest: Hoover, MacArthur, and the Bonus Riot, originally published in 1974 with an updated edition in 1994, is the indispensable primary work on the command-chain question. Lisio worked through the Hoover Presidential Library archives, the War Department records, the diaries and correspondence of the principal actors, and the testimony from the Senate and District investigations that followed the incident. His conclusion: the Chief of Staff exceeded the president’s written orders, did so knowingly, and falsified his account in Reminiscences. The Lisio thesis is the documentary high-water mark of Hoover sympathy. If the question is “did the president order the burning of the camp,” the documentary answer is unequivocally no.

Paul Dickson and Thomas Allen’s The Bonus Army: An American Epic, published in 2004, is the most thorough narrative account of the entire incident from May 1932 through the eventual 1936 promised amount payment. Dickson and Allen accept Lisio’s documentary findings about the command chain but argue that the political and historical responsibility attaches to the incumbent regardless. Their position: a president who authorizes a domestic military operation owns the operation’s execution, particularly when he declines to publicly distinguish his order from the executed result. The Dickson-Allen argument is the political-responsibility argument. It does not contradict Lisio on the documents. It contests the inference Lisio draws from them.

Gary Dean Best’s The Politics of American Individualism: Herbert Hoover in Transition, 1918 to 1921, is the most defensive of the major presidential treatments. Best emphasizes the incumbent’s good intentions, his constrained options given the structural Depression dynamics, and his administrative competence within the limits of the political moment. Best’s treatment of the Bonus Army incident emphasizes MacArthur’s responsibility and minimizes the political-ownership question. His framing is closest to a full rehabilitation of Hoover on this incident.

Kendrick A. Clements’s The Life of Herbert the Republican incumbent: Imperfect Visionary, the most recent of the major the chief executive biographies, takes an intermediate position. Clements accepts Lisio’s documentary findings, accepts Dickson and Allen’s political-responsibility framing, and adds attention to the strategic-communication dimension. His view: the president’s failure was less in authorizing the operation than in failing to control its narrative aftermath. The Clements framing has become influential in recent the incumbent scholarship.

Geoffrey Perret’s Old Soldiers Never Die: The Life of Douglas MacArthur is the major MacArthur biography most sympathetic to its subject. Perret accepts that MacArthur exceeded his orders but argues that the exceedance was strategically justified given the operational situation as the field commander perceived it. Perret’s framing requires accepting MacArthur’s framing of the situation as potentially revolutionary, which most other historians reject. The Perret treatment is the contrarian limit case; it preserves MacArthur’s heroic narrative at the cost of accepting the empirical premises MacArthur himself constructed after the fact.

William E. Leuchtenburg, in his various surveys of the White House occupant and Roosevelt presidencies, offers the standard liberal-internationalist interpretation. Leuchtenburg treats the incident as a defining moment of the president failure, attaches political responsibility to the incumbent, accepts Lisio’s command-chain findings, and emphasizes the contrast with Roosevelt’s later handling of the 1933 second march. The Leuchtenburg framing is closest to the mainstream consensus that the incident defines Hoover’s political failure without exonerating Hoover’s role in the underlying authorization.

Where does the disagreement leave the careful reader? The documentary record supports Lisio on the command chain. The political record supports Dickson and Allen on responsibility. The combined picture is the one this article has constructed: the chief executive authorized a constrained operation, the general exceeded the constraints, and the president bore the political and historical consequences of failing to publicly distinguish his authorization from the execution. The disagreement among historians is less about facts than about how to weight the documentary and the political dimensions of executive accountability.

The Primary Sources, Examined

The reconstruction above draws on six specific primary-source bases. Each deserves direct engagement because each carries a particular evidentiary weight and a particular interpretive challenge.

the incumbent’s July 28 written order to Hurley survives in the War Department archives. The order’s text, transcribed in Lisio’s appendix and reproduced in the relevant volumes of the Hoover Presidential Library collections, contains the geographic limit and the humanity instruction. The order is the foundational document for the entire reconstruction. Its existence and its specific language are the basis for the claim that the senior officer exceeded his authorization. Any reading of the incident that ignores the written order’s specifics is reading the incident incorrectly.

MacArthur’s account in Reminiscences, published in 1964, contains the alternative version of events. The Chief of Staff describes receiving authorization to clear the entire area, including Anacostia, and describes the operation as the suppression of an organized revolutionary threat. The Reminiscences account is contradicted by the documentary record but has shaped popular memory of the incident because Reminiscences is the more accessible source and because the general was a more famous figure than the White House occupant by 1964. The Reminiscences account is what most general readers, if they have read anything, have read on the incident. The documentary correction has not penetrated the popular memory.

Dwight Eisenhower’s At Ease, published in 1967, provides the indispensable corroborating account from the inside of MacArthur’s staff. Eisenhower’s recollection of his own protests, of the messages received from the White House, and of MacArthur’s responses, all corroborate the Lisio documentary findings. Eisenhower’s account carries weight because he was personally present, because his subsequent career gives him no incentive to falsify the record favorably to the president, and because his political relationship with the Chief of Staff (consistently fraught) gives him incentives that point in different directions on different points. The At Ease recollection is the contemporaneous corroboration of the documentary chain.

Memoirs Volume 3, published in 1952, contains Hoover’s own retrospective account. The Memoirs treatment is defensive and partial. It emphasizes the order’s geographic limit, acknowledges that the operation exceeded the limit, but does not explicitly characterize the exceedance as a violation of orders by the Chief of Staff. The hedging in the Memoirs reflects the Republican incumbent’s continuing reluctance, twenty years after the incident, to publicly criticize the Army Chief. The treatment is a primary source for the president’s framing but not a fully reliable source on the operational facts.

The War Department’s official report, completed in August 1932 and released in modified form for the congressional investigations of late 1932, provides the institutional military account. The report is detailed on operational facts and circumspect on chain-of-command questions. It accepts MacArthur’s framing of the situation and provides the institutional basis for the framing’s continuing influence in military-history circles. The report is more reliable on what happened operationally than on why it happened or whether it should have happened.

The contemporary press coverage in the Washington Post and the New York Times from July 29 through August 1, 1932, provides the immediate public record. The coverage was photographic, vivid, and politically devastating to the administration. The photographs that appeared on front pages around the country shaped national opinion within seventy-two hours. The press coverage is the source for the political-effect analysis above and is the basis for understanding why the incident, regardless of its documentary truth about command chains, was politically determinative.

Each source carries its particular bias and its particular evidentiary weight. The combination, carefully read, yields the reconstruction this article has built.

The Two-Image Comparison: 1932 and 1933

A side-by-side reading of two specific moments clarifies what changed between the president and Roosevelt and why the change mattered.

The first moment is the night of July 28, 1932. MacArthur’s troops are advancing across the Anacostia bridge. Tear gas canisters are arcing into the air. The shanties on the flats are beginning to burn. the incumbent is in the White House, in conference with Hurley and Mitchell, receiving updates by telephone. The decision to cross the bridge has been made by the field commander. The two messages forbidding the crossing have failed to stop it. the White House occupant, learning of the crossing, does not order the troops back. He does not order their commander relieved. He instructs that the situation be monitored. The night proceeds. The camp burns. The veterans are scattered. The administration makes no public statement disavowing the operation. The photographs are taken. The papers print them.

The second moment is May 1933. A smaller second Bonus March, perhaps three thousand men, has arrived in Washington. The Roosevelt administration responds within forty-eight hours. The men are housed at Fort Hunt, an military facility in Virginia, with food and medical care. They are visited there by Eleanor Roosevelt, who walks among the tents, talks with the veterans, sings songs from the World War with them, and tells reporters afterward that “the president sent the military. Roosevelt sent his wife.” The men are offered work through the Civilian Conservation Corps. Most accept. The encampment dissolves peacefully. There are no photographs of burning shelters. There are photographs of Eleanor Roosevelt with veterans.

The two images are the same incident structurally and the opposite politically. Veterans seeking economic relief at the federal government in conditions of mass unemployment. The structural setup is identical. The political response is opposite. The 1932 response was procedural correctness backed by force, with no public empathy and no human face. The 1933 response was substantive accommodation with visible empathy and a familiar face. The difference shaped the public memory of the two administrations.

The comparison is not entirely fair to the incumbent. The 1933 response was easier because the second march was smaller, because the New Deal had given the administration tools to offer that the administration had not possessed, and because the precedent of 1932 made the political stakes obvious to anyone considering a repeat of Hoover’s approach. The conditions Roosevelt faced were structurally different. But the political artistry of Roosevelt’s response was a real factor independent of conditions. Sending Eleanor was a choice. Offering work through the CCC was a choice. Treating veterans as fellow citizens facing hard times rather than as a public-order problem was a choice. The incumbent, given the same situation in 1932, made different choices, and the choices mattered.

The Forgotten Veterans

The reconstruction above has focused on the president and the general, on orders and exceedances, on political consequences and historiographical disagreements. The men whose camp burned on the night of July 28 deserve their own attention. They were not interchangeable units in a political drama. They were specific individuals with specific histories and specific stories. A reconstruction that loses them in the institutional analysis fails to honor what they came to Washington seeking.

William Hushka, the veteran killed by District police on Pennsylvania Avenue at approximately 1:50 p.m. on July 28, was a Lithuanian immigrant who had enlisted in the American federal military in November 1917 and served stateside before being discharged in March 1919. He had worked as a butcher in Saint Louis after the war, lost his job in 1929, and joined the march from Saint Louis in June 1932. He was thirty-five years old. He left a widow and a daughter. He was buried at Arlington National Cemetery on August 2, 1932, with full military honors. The Arlington burial, which the administration could have prevented but did not, was a small acknowledgment of what could not be fully repaired.

Eric Carlson, the veteran killed by police bullets at approximately the same time, was a Swedish immigrant who had served with the AEF in France and been wounded at the Argonne. He was forty years old. He died several days after the shooting from his injuries. He, too, was buried at Arlington.

Bernard Myers, the three-month-old infant who died from tear gas exposure in the days after the operation, was the son of John and Bertha Myers, veterans of the AEF and Red Cross respectively. The Myers family had been living at Camp Marks since June. Their son’s death, attributable directly to the gas used during the bridge crossing and camp burning, was the kind of casualty that no military operation against civilians can avoid producing. The death was not in any operational sense necessary. It was a consequence of the decision to cross the bridge.

The thousands of veterans who survived the night dispersed into the broader Depression. Their stories appeared in personal memoirs, in family oral histories, and in the records collected by the Federal Writers’ Project during the 1930s. The patterns in these stories are remarkably consistent. The veterans had come to Washington in hope. They had been disappointed by the Senate vote. They had remained because they had nowhere else to go. They had been driven from their camp by their own the Regular forces. They had returned to states and counties that had no work, no money, and no answer for them. Most lived to see the 1936 payment of the promised payment, to send sons to the Second World War, to receive Social Security checks under the system FDR had built, to die in the 1960s and 1970s in nursing homes and at family bedsides. Their political legacy was the G.I. Bill and the New Deal coalition. Their personal stories were what their grandchildren told at family gatherings and what historians eventually collected.

The night of July 28, 1932, was a specific event in the specific lives of specific individuals. The institutional analysis is necessary, but it does not exhaust the meaning of what happened. The men who came to Washington seeking what their country owed them, who were driven from their encampment by the country’s own soldiers, who watched their meager shelters burn in the summer dusk, were not abstractions. They were Americans being failed by the executive officer of their own government. The failure was institutional and political. It was also personal and irrevocable.

What Hoover Could Have Done Differently

The standard counterfactual exercise in presidential decision reconstructions identifies what the president could have done differently within the constraints he actually faced. The exercise is not about retrospective wisdom; it is about identifying the decisions that were available at each moment.

In May, when the first marchers began arriving, the White House occupant could have authorized a White House meeting with Waters and a small delegation. The precedent against negotiating with mass protest movements would have been preserved by limiting the meeting to formal delegation status. The meeting itself, by demonstrating presidential engagement with veteran concerns, would have changed the political coloration of the entire crisis.

In June, after the Senate rejection of the Patman Bill, the president could have announced a transportation-and-relief package for the veterans that exceeded the eventual $100,000 loan-for-fare appropriation. A larger package framed as veteran relief rather than as compensation payment would have provided a face-saving exit for the men and a face-saving political message for the administration.

In early July, when the Pennsylvania Avenue federal buildings were becoming a flashpoint, the incumbent could have allocated alternative federal property for the veterans’ use during a graduated transition. Empty CCC-style barracks at Fort Meade or Fort Belvoir could have housed the veterans temporarily while the Federal Triangle redevelopment proceeded. The displacement would have been managed rather than forced.

On July 21, when Mills requested federal authorization to clear the federal buildings, the president could have continued to refuse and instructed Glassford to negotiate further. The buildings were not so important that they required clearance on a specific day. Delay was costless.

On July 28 at 2:55 p.m., when Reichelderfer requested federal troops, the president could have declined. The District police had a serious incident on their hands, but the incident did not require federal military intervention. Glassford’s resources, supplemented by additional District personnel, could have managed the situation through the rest of the day without federal troops.

On July 28 at 9:00 p.m., when reports began arriving that the general might cross the bridge, the president could have personally telephoned the senior officer at his Munitions Building command post. A direct presidential order, with the president’s voice on the line, would have been impossible for the Army Chief of Staff to claim he had not received clearly. The fact that the prohibition was transmitted through written orders and through intermediaries gave the general the deniability he would later use.

On July 29, when the photographs appeared, the president could have issued a statement acknowledging that the operation had exceeded its authorization. The statement would have been politically costly but would have established the documentary record at the moment when it would have shaped public memory. Two decades later, in the Memoirs, the incumbent hedged on the question. The hedging in 1932 would have been more useful than the hedging in 1952.

On July 30, the president could have requested MacArthur’s resignation privately. The transition would have been internal and could have been announced as a routine rotation. The personal cost to the president of dismissing his own appointment would have been less than the political cost of being identified permanently with MacArthur’s operation.

Each of these decisions was available. Each was declined. The accumulation of declined alternatives is the structure of the incumbent’s failure. The failure was not a single moment of bad judgment but a sequence of opportunities to alter course, each declined for reasons that made sense to the Quaker engineer at the moment but cumulatively converged on disaster.

Frequently Asked Questions

Q: Did Herbert Hoover order the burning of the Bonus Army camp at Anacostia?

No. The documentary record, established most thoroughly by Donald Lisio in The President and Protest, demonstrates that the chief executive’s written July 28, 1932 order to Secretary of War Patrick Hurley authorized clearing only the federal property along Pennsylvania Avenue. The order specifically did not authorize crossing the Anacostia bridge or operating against the main veterans’ encampment at Camp Marks. Two follow-up communications from the White House, one through Colonel Clement Wright and one through Brigadier General Perry Miles, explicitly forbade the bridge crossing. The decision to cross the bridge and operate against Camp Marks was made by the Chief of Staff in defiance of these orders. The fires that consumed the Anacostia shanties were ignited during the operation the Army Chief chose to conduct, not the operation the president had authorized.

Q: Why did MacArthur exceed Hoover’s orders?

MacArthur’s stated reasoning, given that night to the press and later in his 1964 memoir Reminiscences, was that the situation was potentially revolutionary and required complete suppression of the threat. The framing was empirically false. The Bonus Army had been a politically defeated petition movement since the Senate’s rejection of the Patman Bill on June 17. Waters’s council had been actively expelling the small Communist contingent. The District police, under Pelham Glassford, had been managing the encampment as a public-order matter, not a national-security threat. MacArthur’s framing of the situation as revolutionary reflected his own political views about leftist threats and his personal preference for decisive military action. The framing required the camp to be a base of insurrection rather than the shelter of unemployed veterans, which the framing then justified destroying.

Q: What was the Bonus Expeditionary Force?

The Bonus Expeditionary Force, or BEF, was the name adopted by approximately seventeen thousand World War I veterans who gathered in Washington in June 1932 to demand early payment of the service-compensation certificates promised by the 1924 World War Adjusted Compensation Act. The certificates had been scheduled for payment in 1945. The Depression-era unemployment crisis had made the deferred payment a hardship for veterans whose families faced immediate need. The march was organized loosely under Walter W. Waters, an unemployed Oregon veteran who had served as a sergeant with the 146th Field Artillery in France. The name “Bonus Expeditionary Force” was a conscious echo of the American Expeditionary Forces under which most of the participants had served fourteen years earlier.

Q: What was the Patman Bill?

The Patman Bill was the legislation introduced by Representative Wright Patman, a Texas Democrat and World War I veteran, to authorize immediate cash payment of the service pay certificates rather than deferred payment in 1945. The bill passed the House of Representatives on June 15, 1932, by a vote of 211 to 176, and was rejected by the Senate on June 17, 1932, by a vote of 62 to 18. The Senate’s decisive rejection ended the legislative possibility of immediate payment during the Hoover administration. The compensation would eventually be paid through the Adjusted Compensation Payment Act of 1936, passed over President Roosevelt’s veto with substantial bipartisan support.

Q: How many people died as a direct result of the Bonus Army incident?

Four deaths are directly attributable to the events of July 28 and the following days. William Hushka, a veteran from Saint Louis, was killed by District police bullets at approximately 1:50 p.m. on July 28 during the initial eviction at the federal buildings. Eric Carlson, a veteran from Oakland, was hit at the same incident and died several days later from his wounds. Bernard Myers, a three-month-old infant at Camp Marks, died from tear gas exposure during the bridge-crossing operation. A fourth infant, reported by the Washington Post but not definitively identified by name in the surviving record, also died from gas exposure. Approximately fifty-five additional veterans and bystanders required hospital treatment for serious injuries, and hundreds received minor injuries from tear gas, bayonet wounds, and contusions.

Q: Why didn’t Hoover meet with the Bonus Army leaders personally?

Hoover’s refusal to meet with Waters or other BEF leaders reflected his administration’s position that the president should not establish a precedent of negotiating directly with mass protest movements. The reasoning, articulated by press secretary Theodore Joslin, was that direct presidential engagement would invite every defeated legislative coalition to assemble in Washington and pressure the executive. The procedural argument was not unreasonable, but its political effect was disastrous. Hoover’s failure to provide any administration-level engagement, even by a senior subordinate, communicated indifference to the veterans’ situation and to the broader population watching the standoff. A meeting with a small formal delegation would have preserved the procedural concern while demonstrating presidential awareness of the veteran community’s distress.

Q: Did Dwight Eisenhower really protest MacArthur’s actions at the Bonus Army incident?

Yes. Eisenhower, who was serving as MacArthur’s aide at the Munitions Building in 1932, recorded his protests in his 1967 memoir At Ease. Eisenhower wrote that he had told MacArthur it was inappropriate for the Chief of Staff to personally command a domestic crowd-control operation, that the full dress uniform with mounted decorations was an inappropriate choice for the occasion, and that the bridge crossing exceeded the orders received from the White House. Eisenhower’s specific recollected language about the general, “I told that dumb son-of-a-bitch he had no business going down there,” was published in the memoir thirty-five years after the events. The recollection is consistent with the documentary record and with the broader patterns of Eisenhower’s relationship with the senior officer over the following two decades.

Q: Did the Bonus Army incident cost Hoover the 1932 election?

The election was almost certainly going to be lost on economic grounds alone. National unemployment exceeded 24 percent. The agricultural sector had collapsed. Bank failures had accelerated through 1931 and 1932. Economic-only models of the election predict Hoover losing by approximately twelve to fifteen percentage points. The actual margin was 17.7 points. The Bonus Army incident did not cause the loss; it expanded the margin by roughly three to five points. The margin expansion mattered for down-ballot Republican losses and for FDR’s mandate to undertake the New Deal. The incident shaped the character of the loss without changing the outcome.

Q: How did the Bonus Army incident affect the eventual passage of the G.I. Bill?

The G.I. Bill of Rights, formally the Servicemen’s Readjustment Act of 1944, was structured explicitly to prevent a postwar repeat of the Bonus Army crisis. The bill’s architects, including American Legion lobbyists and Franklin Roosevelt’s domestic policy advisors, designed the legislation to provide veterans with education, mortgage, and unemployment benefits at the time of demobilization rather than through deferred compensation. The memory of 1932 made the political case for immediate benefits self-evidently compelling. The G.I. Bill became one of the most consequential pieces of mid-century domestic legislation, with effects on suburban development, higher education expansion, and middle-class formation that continued for decades. Its origin lay partly in lessons learned from the Hoover-era failure.

Q: What was the relationship between Hoover and the Army Chief of Staff after the incident?

The relationship continued formally without public rupture. Hoover did not request MacArthur’s resignation, did not publicly criticize his handling of the operation, and did not in subsequent correspondence (during the remaining months of the Hoover administration) characterize the bridge crossing as a violation of orders. the general continued as Chief of Staff through 1935, when he was succeeded by Malin Craig under President Roosevelt. The personal relationship between Hoover and the Chief of Staff became distant after 1932, with limited correspondence and no surviving evidence of substantive collaboration during the long post-presidential decades that followed. Both men lived into the 1960s. Both wrote memoirs that addressed the Bonus Army incident with notably different emphases.

Q: Why did Franklin Roosevelt veto the 1936 deferred-bonus payment legislation?

Roosevelt vetoed the Adjusted Compensation Payment Act of 1936 on January 24, 1936, citing fiscal concerns about adding to the federal deficit during the early stages of recovery and concerns about the precedent of single-group benefits over broader social-insurance approaches. The veto was overridden by both houses of Congress within days, with substantial Republican support. The legislation paid approximately $1.7 billion to roughly 3.5 million veterans through bonds redeemable for cash. The political dynamics of the override reflected both the lingering memory of 1932 and the broader pre-Social Security debate about how the federal government should structure benefits for specific population groups versus universal programs.

Q: How does the Bonus Army incident compare to the 1951 Truman firing of the Army Chief?

The two incidents are structurally connected by MacArthur’s characterological pattern of expanding his operational scope beyond civilian authorization, framing his decisions publicly in terms that constrained his civilian superiors, and forcing his commanders-in-chief into politically difficult positions. In 1932, Hoover declined to publicly repudiate MacArthur’s exceedance, and the political cost attached to Hoover. In 1951, Truman did publicly repudiate the field commander, fired him from his Korean command, and absorbed the political cost directly. Truman’s success in reasserting civilian control of the military in 1951 owed something to the lesson available from Hoover’s failure in 1932. The same general was tested against the same civilian-control principle by two different presidents with different temperaments, with different results.

Q: Where can I find Hoover’s original July 28, 1932 order?

The original written order is preserved at the Herbert Hoover Presidential Library in West Branch, Iowa, in the papers relating to the Bonus Army incident. The text has been published in full in Donald Lisio’s The President and Protest as an appendix, and in various edited collections of Hoover-era documents. The order’s language, particularly the geographic limit on the affected area and the instruction regarding humanity in execution, is the foundation for the documentary case that Douglas he exceeded his orders. Reading the original text alongside MacArthur’s account in Reminiscences is the single most illuminating exercise for understanding the command-chain dispute.

Q: Was the camp at Anacostia actually a Communist organization?

No. The Bonus Expeditionary Force at Camp Marks was governed by Walter W. Waters and a council of veteran representatives under explicitly anti-Communist rules. Communists who attempted to recruit at the camp were ejected. The American flag flew over the camp. Sunday religious services were held. The District police superintendent, Pelham Glassford, characterized the camp throughout June and late summer as a politically defeated petition movement composed of unemployed veterans, not a national-security threat. The Pennsylvania Avenue federal-buildings contingent included a small Communist element that Waters had less ability to control because the federal buildings were a separate occupation, but this contingent was a small minority of perhaps a few hundred among the thousands present in Washington. MacArthur’s framing of the operation as suppression of Communist revolution required ignoring the actual composition of the Bonus Army.

Q: What happened to Walter Waters after the camp was destroyed?

Walter W. Waters returned to Oregon after the camp’s destruction and worked occasional menial jobs through the Depression. He briefly attempted to organize a successor movement called the Khaki Shirts, which dissolved quickly. He served in the troops during the Second World War. He worked in obscurity for the remainder of his life and died in 1959 at age sixty-one. The Khaki Shirts attempt and his other post-1932 organizational efforts reflected the same impulse that had produced the BEF but failed to achieve comparable scale or attention. Waters’s brief moment of historical prominence in the summer of 1932 was the high point of his public life, and his eventual return to obscurity paralleled the broader dispersal of the men he had led to Washington.

Q: Why is the Bonus Army incident less well known today than it was in 1932?

The incident’s prominence has declined in popular memory partly because subsequent crises (the New Deal’s transformation of the federal government, the Second World War, the postwar consensus) overshadowed Depression-era events that did not lead to permanent institutional change. The G.I. Bill, which arguably grew from lessons learned from the incident, is well remembered, but its connection to the 1932 origins is less commonly traced. The incident also lacks a single iconic image or single iconic name that could have anchored popular memory. The fires at Anacostia, the cavalry on Pennsylvania Avenue, and the names of the dead veterans never achieved the cultural penetration of, for example, the assassination of Kennedy or the Watergate scandal. The incident remains important in scholarly Hoover studies and in military-history treatments of the general, but it has faded in general historical consciousness.

Q: How did the Bonus Army incident shape later civil-military relations in the United States?

The incident contributed to the development of more precise doctrines around the deployment of federal military force in domestic crowd-control settings. Subsequent presidents, including Eisenhower at Little Rock in 1957, used federal troops with significantly greater attention to scope, command structure, and public framing than Hoover had demonstrated. Eisenhower in particular had personally observed the command-chain failure of July 28 and applied the lesson in his own decisions. The Posse Comitatus Act of 1878 already constrained federal military domestic deployment, but the Bonus Army incident illustrated that legal limits were insufficient without operational discipline and clear civilian authorization. The professionalization of crowd-control doctrine over the following decades drew partly from the 1932 lessons, even where the connection was implicit rather than explicit.

Q: What was Eleanor Roosevelt’s role in handling the 1933 second Bonus March?

Eleanor Roosevelt personally visited the second Bonus March encampment at Fort Hunt, Virginia, in May 1933. She walked among the tents, talked with veterans, sang World War-era songs with them, and helped administration officials arrange for the men to be offered work through the Civilian Conservation Corps. Her visit was a deliberate counterpoint to Hoover’s distance from the 1932 march. Veterans participating in the 1933 march later said that “Hoover sent the military; Roosevelt sent his wife.” The contrast became a defining element of the early Roosevelt administration’s political messaging about how the federal government should engage with citizens in distress. Eleanor’s role illustrated a different theory of presidential representation, with the First Lady serving as a personal face of administration concern for circumstances the formal presidency could address only at a procedural distance.

Q: Did the Reconstruction Finance Corporation or other Hoover-era programs help the Bonus Army veterans?

The RFC and other Hoover-era programs were designed for institutional lending rather than direct individual relief, and they did not provide assistance to the Bonus Army veterans during the 1932 crisis. The RFC’s loan structure required collateral and institutional borrowers, which excluded unemployed individuals. The Emergency Relief and Construction Act of July 1932, signed by Hoover only days before the Bonus Army crisis erupted, provided some federal relief funding to states but not directly to veterans. The mismatch between the administration’s program design and the actual needs of the unemployed was one of the defining features of the Hoover policy approach. The administration’s tools were institutional; the population’s needs were personal. The Bonus Army incident was, among other things, the visible collision between these mismatched levels.

Q: How did the Bonus Army incident affect Hoover’s reputation in his post-presidential decades?

Hoover lived until 1964, more than three decades after the incident. He served the Truman administration through the Hoover Commissions of the late 1940s, and he served the Eisenhower administration through similar commissions in the 1950s. His humanitarian reputation slowly recovered, but the Bonus Army incident remained a fixed point in any reckoning of his presidency. Rehabilitation attempts by historians including George Nash, Joan Hoff, and others addressed the incident with the documentary care it deserves but could not change the political-responsibility verdict. The incident is, by general scholarly consensus, the single defining political failure of Hoover’s presidency, and no reappraisal effort has succeeded in removing it from that position. Hoover’s broader legacy as a Quaker humanitarian and administrative innovator survives the incident in scholarly treatments, but the presidency-specific evaluation remains marked by the night of July 28, 1932.

Q: What role did the press play in shaping public response to the events of July 28?

The press response was rapid, photographic, and politically devastating to the administration. The Washington Post and the New York Times both led front pages on July 29 with images of mounted cavalry, fixed bayonets, and burning shelters. Wire service photographs reached newspapers across the country within twelve hours of the operation’s conclusion. Editorial response was mixed along partisan lines: conservative editorial pages generally supported MacArthur’s framing of the operation as suppression of a revolutionary threat, while liberal and centrist pages condemned the use of regular military troops against unemployed veterans. The Hearst chain, which had been broadly supportive of the administration, broke decisively against Hoover during the week following July 28. Walter Lippmann’s syndicated column, widely read among policy-attentive readers, framed the operation as a fundamental failure of executive judgment. The press coverage shaped national opinion within seventy-two hours of the operation. By the time the administration could have responded with a coherent counter-framing, the photographic record had already established the dominant narrative. The lesson, learned by every subsequent president, was that domestic military operations against civilians create photographic evidence that political messaging cannot easily counter.

Q: How did veterans’ organizations respond to the July 28 operation?

The American Legion, the largest veterans’ organization, was internally divided. The national leadership had opposed early bonus payment during the legislative debate, partly out of concern that the compensation would damage the broader credit conditions the country needed for recovery. After July 28, the Legion’s national leadership was placed in a difficult position by the operation against fellow veterans. The Legion’s national commander, Henry L. Stevens Jr., issued a measured statement on July 30 expressing regret at the violence without explicitly criticizing the administration. State-level Legion posts, particularly in regions where the Bonus March had drawn significant local participation, were considerably more critical. The Veterans of Foreign Wars, generally more sympathetic to early compensation payment, condemned the operation directly. The Disabled American Veterans, whose constituency had been heavily represented in the encampment, issued the sharpest criticism. The fracture within the veterans’ community between national-leadership procedural caution and local-membership solidarity continued through the 1936 bonus payment fight and shaped veterans’ political behavior into the 1940s.

Q: Was there a congressional investigation of the July 28 events?

Yes. A House subcommittee held hearings during the lame-duck session in early 1933, taking testimony from District police officials, Army officers including some who had participated in the operation, and veterans who had been present at Camp Marks. The Senate held parallel but more limited hearings. The investigations produced detailed factual records but limited political conclusions because the change of administration in March 1933 shifted political attention to New Deal legislation. The records of these hearings remain valuable primary sources for the operational details, particularly for the District police perspective and for the testimony of junior Army officers whose accounts contradicted MacArthur’s framing. The investigations did not produce formal sanctions against any officer involved. the senior officer continued as Chief of Staff until 1935. The chain-of-command questions that the investigations might have addressed institutionally were left to scholarly reconstruction in the decades that followed.

Q: What lessons did Dwight Eisenhower take from his experience as MacArthur’s aide that day?

Eisenhower drew several specific lessons that shaped his later approach to civil-military matters. First, he concluded that the Chief of Staff should not personally command a domestic operation; that role belonged properly to the brigade commander or theater commander whose normal responsibilities included the affected area. Second, he concluded that domestic military deployment required unambiguous written authorization with explicit geographic and tactical limits, and that the senior officer in command had a duty to confirm those limits rather than expand them. Third, he concluded that dress uniform with mounted decorations was an inappropriate choice for domestic crowd-control operations and that operational rather than ceremonial appearance better suited such contexts. Each of these lessons informed Eisenhower’s own decisions twenty-five years later at Little Rock, when his careful attention to authorization, scope, and command structure produced an operation that achieved its constitutional purpose without political damage. Eisenhower’s At Ease memoir explicitly draws the connection between the 1932 lessons and the 1957 execution, making the chain of institutional learning visible across the generation that bridged the two events.