In April 1927 the Mississippi River broke through its levees at Mounds Landing, Mississippi, and at many other points along the lower valley. The flooded area is commonly given as about 27,000 square miles, and hundreds of thousands of people were driven from their homes. The flood exposed a federal engineering doctrine that had relied on levees alone. The relief effort that followed produced documented forced labor and racial discrimination in the camps, and the disaster led directly to the Flood Control Act of 1928. This guide covers the doctrine, the breaches, the camps, the deliberate cut at Caernarvon and a death toll that remains uncertain.

1926 to 1927: months of rain and a river system out of its banks

The flood was built over many months. Heavy rain fell across large parts of the Mississippi basin from the late summer of 1926 into the spring of 1927. The tributaries rose with the main stem, among them the Arkansas, the White and the rivers draining the Ohio Valley, and the lower river stayed high for an unusually long time. The basin drains a large share of the continental United States, so a lower Mississippi flood is the sum of water that fell over many states and arrived at different times. When those arrivals overlap, as they did in 1927, the river below Cairo, Illinois has to carry more water than its channel can hold for weeks at a stretch.

The levees along the lower river were built to keep that water in the channel, and in the spring of 1927 they failed at many points in Missouri, Arkansas, Mississippi and Louisiana. Some breaks were natural crevasses, where the river cut through a weak section or overtopped it. At least one was cut on purpose, which I cover below. Once water was on the land, much of the floodplain was so flat that it had nowhere to drain quickly. In parts of the Mississippi and Arkansas Deltas it stood for weeks, and by most accounts it stood for months in some places.

Displacement figures vary between sources, which is common for a disaster of this size and era. Accounts usually put the number of people driven from their homes somewhere between about 600,000 and 700,000, and some retellings give higher numbers that I cannot trace to a count. The American Red Cross ran the main relief operation and published an official report on it afterward. That report is the place to start for what the Red Cross itself counted, keeping in mind that it describes people the Red Cross served, which is not the same thing as everyone who was displaced.

The two most useful book-length histories for anyone working on this event are Pete Daniel’s Deep’n As It Come, published in 1977, and John M. Barry’s Rising Tide, published in 1997. Much of what follows on the politics and the camps rests on their work and on the contemporary investigations they draw from. Where an account depends on one historian’s reading of the record, I say so.

Levees only: Humphreys and Abbot, Eads and the Mississippi River Commission

The engineering doctrine that failed in 1927 had roots in the 1850s. Congress funded surveys of the lower Mississippi after floods in that decade. The civil engineer Charles Ellet argued for a combination of measures that included reservoirs on the tributaries and outlets to carry floodwater away from the main channel. The Army survey led by Andrew A. Humphreys and Henry L. Abbot, whose report was published in 1861, reached a different conclusion. It rejected outlets and cutoffs as impractical or harmful and treated levees as the workable means of protection. Humphreys later became Chief of Engineers, and his report carried great weight in the Corps for decades.

James Buchanan Eads came to the river with a separate argument. He held that a confined current would scour and deepen its own channel. His jetties at South Pass, at the mouth of the river, were completed in the late 1870s and did deepen that passage for navigation, and his view lent support to the belief that confining the river would help it carry floods. Congress created the Mississippi River Commission in 1879 to coordinate work on the lower river. By the later nineteenth century, federal and commission policy had settled on what came to be called levees only, under which the river would be held between continuous levees and spillways and outlets were not part of the plan.

The funding arrangement shaped the doctrine as much as the engineering did. Levee districts and states built and paid for much of the levee system, and federal participation was long justified under the navigation power. A 1917 act is generally described as Congress’s first appropriation made expressly for flood control on the Mississippi, and it required local interests to share the cost. A doctrine that relied on one kind of structure, built partly by local bodies of very different means, left the system only as strong as its weakest reach.

The failure mode should be stated plainly. Confining a river between levees raises flood stages within the channel, because water that once spread across the floodplain now has to pass between the embankments. Higher stages call for higher levees. When a levee built to that standard fails, the water behind it leaves through the breach with great force and floods land whose occupants have been told they are protected. The doctrine had no planned relief valve, so in 1927 the river found its own outlets at places nobody had chosen.

Mounds Landing, 21 April 1927: the crevasse that flooded the Delta

The single most destructive break came on 21 April 1927 at Mounds Landing in Bolivar County, Mississippi, north of Greenville. In the days before the failure, crews had been working the levee with sandbags against rising water and sand boils. Many of the laborers on that stretch were Black men drawn from the surrounding plantations and from Greenville. Barry’s account, drawn from contemporary records and later interviews, describes men held at work on the levee by armed guards. That is his reading of the evidence, and readers who need the detail should go to his book and the sources he cites.

When the levee gave way, the breach widened rapidly. Accounts commonly describe the crevasse as eventually reaching roughly three-quarters of a mile wide, and some describe it as nearly a mile. Water poured into the Yazoo-Mississippi Delta, the flat and heavily farmed region between the Mississippi and the Yazoo rivers, and spread across a large area within days. Many houses were covered to the roof. The number of people killed at the breach itself is not known. Accounts of workers swept away when it opened circulate widely, but I have not seen any reliable count of them, and I will not give one.

Greenville, the largest town in that part of the Delta, flooded soon after. Water stood in its streets for weeks. The levee along the river was the highest ground for miles, and the people who had lost their homes went there because nowhere else was dry. That decision, made by thousands of people in the first hours, defined the relief problem in Washington County for the rest of the spring.

Mounds Landing was one crevasse among many that spring, but its location and size made it the defining break of the flood in Mississippi. It also became the starting point for most of the later accounts of how Black residents of the Delta were treated during the relief effort.

The Greenville levee: stranded refugees and the boats that left without them

By most accounts more than ten thousand people were stranded on the Greenville levee in the days after the crevasse, most of them Black. Barry gives a figure of about 13,000, and I give it as his number. Some white residents were taken out of the area, while many Black refugees stayed on the levee top in tents and makeshift shelter, with the river on one side and floodwater on the other. Sanitation and food supply were difficult, and the camp stayed there for weeks.

The most consequential decision in Greenville is described in detail by Barry. The local relief committee was chaired by William Alexander Percy, son of the planter and former US senator LeRoy Percy. By Barry’s account, Will Percy arranged steamboats to evacuate the refugees to higher ground. Planters, including his father, opposed sending the Black labor force away from the Delta, because they feared the workers would not come back. The boats left without most of the Black refugees. Readers should take this as Barry’s reconstruction of private decisions from correspondence and interviews, and the full account is in Rising Tide.

The people kept on the levee were also put to work. Barry and the contemporary investigators describe Black men in Greenville being required to unload relief supplies and work on the levee under the direction of authorities, sometimes without pay, with arrest and force used against those who would not. Similar practices were reported elsewhere in the flooded region. The point for a modern reader is the decision structure. People with a direct economic interest in keeping refugees in place held the authority over whether those refugees would be moved, and the relief committee’s choices reflected that interest.

Red Cross camps, National Guard passes and the labor that was compelled

President Calvin Coolidge put Secretary of Commerce Herbert Hoover in charge of a special committee to coordinate the federal response, and Hoover worked closely with the American Red Cross, which ran the camps and distributed food, clothing, seed and feed. The camps were segregated. In many places National Guard troops guarded the camps for Black refugees, and accounts describe refugees who could not leave without a pass. Some were released only to the planter they had worked for, or once that planter agreed to take them back. Relief issued to sharecroppers was in some places routed through their landlords. Investigators reported cases in which planters charged tenants for Red Cross supplies on their plantation accounts, adding to debts that already bound them to the land.

These practices were documented at the time. Reports of abuse reached Hoover, and in response he appointed the Colored Advisory Commission, chaired by Robert Russa Moton, president of the Tuskegee Institute, to investigate conditions in the camps. The commission’s investigators reported forced labor, mistreatment by guards, restrictions on movement and relief withheld or diverted. Walter White of the NAACP investigated separately and published findings that were more openly critical. Historians including Daniel and Barry describe Hoover as accepting some of the commission’s recommendations while softening or holding back the harsher findings. Barry also describes a promise of land reform for Black tenant farmers, discussed between Hoover and Moton, that never came to anything after the 1928 presidential campaign. That account is Barry’s interpretation, and his book should be read for the evidence behind it.

Conditions were not identical across the camps. Some camps had better sanitation and medical care than their inmates had known before the flood, and Red Cross reporting emphasized vaccination and public health work. Those facts sit alongside the documented coercion rather than cancelling it. The abuses came largely from the way relief was administered through the existing local power structure, so a camp could be medically well run while its residents were held there against their will.

Caernarvon, 29 April 1927: the dynamited levee and the compensation that fell short

As the crest moved downriver, the business and banking leadership of New Orleans pressed to relieve pressure on the city by deliberately breaching the levee below it. The site chosen was Caernarvon, in St. Bernard Parish near the Plaquemines Parish line, downstream of the city. As Barry describes it, the cut went ahead with the approval of Louisiana’s governor and the agreement of federal officials. Dynamiting began on 29 April 1927. The first charges did less than expected, and it took repeated blasting over several days to open a substantial cut. The water flooded large parts of St. Bernard and Plaquemines parishes. The people living there included trappers, fishermen and the descendants of Canary Islands settlers known as Islenos, and the number displaced is usually given as around ten thousand.

The residents were promised compensation. New Orleans civic and financial leaders publicly pledged that the people flooded out would be made whole, and a process was set up to hear claims. Barry’s account, drawn from the claims records, is that the payments came to a small fraction of what was claimed and that many claimants received little or nothing. I am not giving dollar figures here, because the ones I have seen in retellings vary and I cannot reconcile them with confidence. Anyone who needs the numbers should go to Barry’s sources and to Louisiana archival material on the claims.

Whether the cut was needed is a separate question. Barry and others argue that crevasses upriver had already lowered the flood stage enough that the river at New Orleans would probably have stayed within its levees without the Caernarvon breach. That is a historian’s argument about a counterfactual, and it is widely accepted. It is not a finding that anyone at the time was in a position to make with certainty, and I present it as an argument.

One popular retelling needs to be labeled for what it is. A myth that circulated after Hurricane Katrina in 2005 held that the levees were blown deliberately to flood Black neighborhoods of New Orleans, and the 1927 cut was often cited as precedent. That claim about 2005 is false, and the floodwall failures of that year are covered in the site’s article on New Orleans. The 1927 cut was real, but it flooded the parishes downriver of the city and did not flood the Ninth Ward. The fact that residents were promised compensation and largely did not receive it helps explain why the distrust has lasted.

The death toll: 246 officially, and a count historians do not accept

The figure most often given as the official death toll is 246. I have seen it attributed to federal and Red Cross reporting of the time, and I am not certain which document first fixed it, so I give it as the commonly cited official count rather than tracing it to a specific report. Some sources give somewhat different official figures.

Historians do not accept 246 as the real number. Barry argues that the true toll was considerably higher, and the higher estimates I have seen in the secondary literature run from several hundred to more than a thousand. I cannot trace those higher figures to any systematic count. These sources disagree, and I am not going to average them or pick one. The honest statement is that the official count is about 246, that the true number was almost certainly higher, and that nobody has produced a reconstructed count good enough to replace it.

The reasons for the undercount are not hard to see. Most of the dead in the Delta were rural, poor and Black, and their deaths were less likely to be recorded by any official body. People drowned in isolated houses and on farms where the water stood for weeks, and some bodies were never recovered. Recordkeeping by county officials and relief agencies in 1927 was thin even in good conditions. The deaths at the Mounds Landing breach, which are widely reported in accounts but never counted, are one example of a category that the official figure is unlikely to have captured.

The 1928 Flood Control Act: the Jadwin Plan, floodways and federal responsibility

Congress responded with the Flood Control Act of 1928, which President Coolidge signed on 15 May 1928. The act adopted a plan prepared under Major General Edgar Jadwin, Chief of Engineers, and authorized the Mississippi River and Tributaries project, usually shortened to MR&T. The authorization usually cited in the histories is $325 million. More important than the dollar figure was the principle. The federal government accepted responsibility for flood control on the lower Mississippi and took on most of the construction cost, while local interests remained responsible for some obligations such as providing rights of way.

The Jadwin Plan abandoned levees only. It kept and strengthened the levees but added floodways and spillways designed to divert water out of the main channel at chosen points during a major flood. The Bonnet Carre Spillway above New Orleans and the Birds Point-New Madrid Floodway in Missouri were both first used in the 1937 flood. Later work added further structures, including the Morganza Spillway, built in the 1950s, which feeds the Atchafalaya Basin. The system concedes in advance that some land will be flooded deliberately in order to protect other land. In May 2011 the Corps operated the Birds Point-New Madrid Floodway, breaching its levee with explosives, and opened Morganza. That record is the place to look for how the modern version of a Caernarvon decision is made under a plan written in advance.

The act also contained a provision shielding the United States from liability for damage from floods or flood waters, and courts have applied that provision in later litigation. I am describing it rather than citing the section, and anyone who needs the current text should read the law as codified today. Whether owners of floodway land were adequately compensated through flowage easements has been a recurring dispute in the decades since. The general principle that flood control is a proper federal activity was extended beyond the Mississippi by the Flood Control Act of 1936, which is the usual starting point for histories of the national program.

What to do at your agency

• The emergency management coordinator reads the shelter annex of the existing emergency operations plan this month, confirms in writing that it states shelter residents may leave at any time without anyone’s permission, and adds that sentence if it is missing.
• The mass care lead checks the commodity distribution plan for any step that routes relief through an employer, landlord, property manager or other third party, and rewrites that step so that residents register and collect directly.
• The emergency management coordinator asks the levee district, drainage district or Corps district office for the emergency action plan and inundation map covering any levee in the jurisdiction, and records in the hazard mitigation plan which shelters, dispatch centers and hospitals sit behind it.
• The emergency management coordinator adds one item to the agenda of the next regular emergency management meeting asking whether any part of the jurisdiction lies in a designated floodway or spillway, and who would notify residents and on what timeline if it were operated.
• The volunteer coordinator writes one paragraph into the existing volunteer management annex stating that shelter residents are never required to work as a condition of shelter or relief, and that any work they choose to do is recorded as voluntary.
• The public information officer adds to the existing crisis communications plan a standing rule that any casualty figure released or repeated must name its source and date, and checks the last after-action report for any number that does not.

Takeaways

• The 1927 flood followed months of heavy rain across the Mississippi basin and broke levees at many points from Missouri to Louisiana, flooding an area commonly given as about 27,000 square miles.
• The levees-only doctrine, rooted in the Humphreys and Abbot report of 1861 and later Mississippi River Commission policy, relied on continuous levees and had no planned outlet for a flood larger than the channel could carry.
• The crevasse at Mounds Landing on 21 April 1927 flooded much of the Yazoo-Mississippi Delta, and the number of people killed at the breach itself has never been counted.
• By Barry’s account, thousands of mostly Black refugees were kept on the Greenville levee after planters opposed evacuating their labor force, and many were made to work.
• The Colored Advisory Commission and the NAACP’s Walter White documented forced labor, guarded camps, restricted movement and relief routed through planters.
• The Caernarvon levee was dynamited beginning 29 April 1927 to protect New Orleans, and Barry’s account of the claims records shows that the promised compensation came to a small fraction of what was claimed.
• The commonly cited official death toll is about 246, and historians including Barry argue the true number was considerably higher, with estimates running from several hundred to over a thousand.
• The Flood Control Act of 1928 adopted the Jadwin Plan with floodways and spillways and committed the federal government to flood control on the lower Mississippi.