Rudolph Hecht was president of Hibernia Bank, chairman of the New Orleans Association of Commerce, and a member of the kind of overlapping civic boards that, in a city like New Orleans in 1927, meant he could walk into any room that mattered without knocking. James P. Butler Jr. was a Boston Club attorney who handled the negotiations with the communities that were about to be destroyed. These two men, along with a rotating cast of bankers, cotton factors, and levee board members who shared the same lunch tables and the same financial exposure, organized the campaign to dynamite the Caernarvon levee thirteen miles below New Orleans on the east bank of the Mississippi River.
On April 29, 1927, Army Corps engineers set off thirty-nine tons of dynamite. Over the following ten days they widened the breach to roughly 3,200 feet. Through that gap poured something like a fifth of the Mississippi's volume, drowning St. Bernard and Plaquemines Parishes, where thousands of people made their living trapping muskrat and fishing in a commercially integrated marsh economy that was anything but marginal.
The people who lived in those parishes did not decide to blow the levee. The people who decided to blow the levee did not live in those parishes. The authorization chain ran upward through Louisiana Governor Oramel H. Simpson, Secretary of War Dwight F. Davis, and the Orleans Levee Board, a state body governed by New Orleans appointees. Herbert Hoover, then Secretary of Commerce and already the most visible federal official managing the flood response, was kept informed and did not object. The chain ran upward through every relevant authority. It did not run outward to the people downstream.
What They Knew and When They Knew It
The argument for the breach was straightforward and, on its surface, reasonable. The Mississippi was at record levels. Upstream levees had been breaking for weeks across the delta. New Orleans, the commercial capital of the lower valley, appeared to be in genuine danger. If the city flooded, the economic consequences would ripple through the national economy. The solution proposed by Hecht and the Association of Commerce was to relieve pressure on the city's levees by opening a controlled breach downstream, sacrificing rural parishes to save the metropolis.
John M. Barry, in Rising Tide, forces an uncomfortable question into the open: whether the breach was necessary at all. Barry's reconstruction of Corps engineering data from late April 1927 argues that the river's crest at New Orleans had likely already peaked, or was peaking, by the time the dynamiting was completed. Corps engineers had information suggesting the city's levees would hold without intervention.
Barry's conclusion is an interpretive reconstruction from available data, not a simple factual finding. The Corps and the Mississippi River Commission maintained at the time that the breach was essential. The contemporary evidence is ambiguous enough that reasonable engineers could disagree. But the ambiguity matters here, because the people who bore the entire cost of that uncertainty had no meaningful role in resolving it. Opponents of the breach argued the levees would hold. They argued that the fur-trapping industry in St. Bernard and Plaquemines would be devastated. Barry documents that these arguments were heard and overridden. The parishes were rural, politically subordinate to New Orleans, and their residents had no effective representation in the rooms where Hecht and Butler made their case.
The Promise
Before the breach, something had to be done about the people downstream. You can't blow up a levee and drown a community's livelihood without at least saying you'll make it right.
The compensation commitment came primarily from the New Orleans business community, channeled through Hecht, Butler, and the Association of Commerce. St. Bernard and Plaquemines Parish officials initially resisted the breach. Their acquiescence was purchased with assurances of full compensation for all losses.
How these promises were made tells you more than what they said. They were public statements and direct negotiations between New Orleans civic leaders and parish officials. They were not written legal guarantees. They were not legislative appropriations. They were not backed by any governmental authority's formal commitment to pay. The entity making the promises was a collection of private citizens who had no mechanism to compel payment from any public treasury and no legal obligation to pay from their own pockets.
The pattern is old and it is not complicated. Management tells the workers there will be severance, retraining, placement assistance. The workers ask if it's in writing. Management says they don't need it in writing because they're men of their word. The workers accept this because the alternative is accepting that they have no power at all, which is the one thing people will go to considerable lengths to avoid knowing about themselves.
Whether the parish officials believed the promises fully or simply recognized they had no power to stop the dynamite is a question the historical record doesn't cleanly answer. The direct testimony of St. Bernard and Plaquemines residents is thin in the archive. The people most affected tend to leave the least complete records.
The Broken Promise
The compensation actually delivered was, by any honest accounting, an insult. Barry documents a figure widely reproduced in secondary accounts: residents received roughly $150,000 against losses that contemporary assessments placed in the millions. The claims process was administered by a committee controlled by representatives of the same New Orleans business community that had made the promises. Claims were systematically underpaid or rejected.
The marsh ecosystem damage extended well beyond the period of inundation, compounding the economic destruction. The people who had been promised they'd be made whole were handed a fraction of what they'd lost and told the matter was settled.
Where sources disagree on exact figures, look at what they agree on. Barry's numbers, the Mississippi Encyclopedia's account, and the 64 Parishes entry present the compensation figures with minor variations. None of them dispute the ratio. The compensation was pennies on the dollar, delivered late, administered by the people who caused the damage, with no independent oversight and no legal recourse for the recipients.
Hoover's Parallel Promise
Herbert Hoover understood the 1927 flood as a political instrument of enormous power. President Coolidge was notably reluctant to engage with the disaster personally, and Hoover filled the vacuum, treating the relief operation as both a logistics problem and a presidential campaign. He was running, and the flood was his stage.
His relationship to the Caernarvon decision was characteristically indirect. Hoover was not in the Corps chain of command. His interest was in managing consequences, and the consequence that worried him most was hundreds of miles upriver from Caernarvon.
The flooded Mississippi Delta contained large Black populations, mostly sharecroppers and tenant farmers. When levees broke throughout the delta, Black residents were disproportionately trapped. Barry documents that planters and local sheriffs in some flooded areas prevented Black refugees from evacuating, fearing loss of agricultural labor. Black refugees were confined to levee tops and refugee camps to preserve them as a workforce, with aid a secondary concern at best. The Red Cross, operating through local white committees, often accommodated rather than challenged this system.
Hoover recognized that Black voters were a significant Republican constituency in northern cities, and that reports of forced labor and abuse in southern refugee camps were a political problem. His solution was to engage Robert Russa Moton, principal of Tuskegee Institute and the most prominent Black public figure in America. The arrangement, as Barry reconstructs it from correspondence: Moton would inspect conditions in Black refugee camps and publicly vouch for the adequacy of relief. In return, Hoover reportedly promised federal support for a major land reform initiative for Black farmers in the delta.
Moton conducted his inspection. He issued a public statement that was favorable, though privately qualified. He gave Hoover the political cover he needed.
After the 1928 election, which Hoover won in a landslide, nothing happened on the land reform promise. The Black press, particularly the Chicago Defender and the Pittsburgh Courier, had been monitoring the gap between Hoover's promises and reality, and Barry documents their coverage of conditions that the official narrative obscured. The exposure contributed to what historians including Nancy Weiss, in Farewell to the Party of Lincoln, have documented as one of the early fractures in Black Republican loyalty, a loyalty that had held since Reconstruction and would collapse entirely under FDR.
The promise to Moton and the promise to St. Bernard Parish were structurally identical. Both were made by people who needed something from the recipients. Both were made without any formal mechanism to compel fulfillment. Both were broken. Moton's broken promise had electoral consequences that the historical record preserved. The trappers' broken promise had consequences only for the trappers.
The Act
The Flood Control Act of 1928 was, as Barry documents, the largest peacetime federal construction undertaking in American history to that point. It authorized the Army Corps of Engineers to design and build a comprehensive flood control system for the lower Mississippi: levees, floodways, channel improvements, and spillways. The Bonnet Carré Spillway, completed in 1931, could divert Mississippi water into Lake Pontchartrain when river levels threatened New Orleans. The spillway replaced the ad hoc Caernarvon solution with engineered infrastructure. No more dynamite. Same logic, better plumbing.
The Act also buried the "levees-only" doctrine that the Mississippi River Commission had enforced since 1879. That doctrine held that levees alone, by constricting the channel, would scour the riverbed deeper and increase carrying capacity. The theory had been championed by Army engineers and embedded in Commission policy for nearly fifty years against the objections of rival engineers who argued the river's energy had to go somewhere. The 1927 flood settled the argument the way floods settle most arguments: conclusively, and too late for the people already underwater.
The Act designated flood control as a federal responsibility and established the Corps as the dominant institution in American flood management, creating the institutional template for every major federal disaster infrastructure program that followed.
It also contained a provision that ought to be famous.
The Immunity Clause
Codified at 33 U.S.C. § 702c, the liability provision of the 1928 Flood Control Act states in substance:
"No liability shall attach to the United States for any damage from or by floods or flood waters at any place, including damage caused by flood control works constructed under the Act."
Read that again. The federal government authorized itself to build levees, spillways, and floodways. It authorized itself to direct flood waters deliberately, channeling them away from some communities and toward others. And it immunized itself, in advance, from any legal claim by the communities that got flooded.
The Caernarvon precedent became a capability to be protected from legal challenge.
The Act included designated sacrificial floodways. The Birds Point–New Madrid Floodway on the Missouri side was engineered so that the Corps could deliberately breach the levee to flood Missouri farmland and protect Cairo, Illinois. In 2011, the Corps did exactly that. Same decision, same asymmetry, same immunity. Eighty-four years later.
What the Act Did Not Do
It contained no compensation for the people of St. Bernard and Plaquemines Parishes. The communities deliberately destroyed in 1927 received nothing from the legislation their destruction had made politically possible. The Congressional Record of the 70th Congress contains no evidence that compensation for Caernarvon's victims was seriously debated during the Act's passage. Their absence from the legislative record is consistent with their absence from the rooms where the decision was made.
No equity framework appeared anywhere in the legislation. Nothing addressed how the benefits and burdens of flood protection would be distributed among communities. Engineering decisions would be made on engineering grounds, which in practice meant economic grounds. Protect the highest-value assets. The logic that had sacrificed rural trappers to save New Orleans bankers was given concrete and steel.
The local levee board system went unreformed. The Orleans Levee Board, the same body whose appointees had facilitated the Caernarvon decision, retained authority over local levee governance. The membership dynamics that produced the decision continued undisturbed.
The Flood Control Act of 1928 is remembered as the moment the federal government accepted responsibility for protecting Americans from floods. The same Act accepted no responsibility for the Americans it would flood in the process, and wrote that acceptance into permanent law.
- The NFIP's inherited logic: The National Flood Insurance Program, now $20 billion in debt and expiring September 30, 2026, descends directly from the institutional template the 1928 Act established, including the unresolved question of who bears cost when federal flood management creates winners and losers.
- Buyout programs and Caernarvon's echo: A peer-reviewed analysis of more than 40,000 voluntary flood buyouts found that bought-out properties concentrate in areas of greater social vulnerability, raising the same distributional question the Caernarvon decision posed a century ago.
- Managed retreat's human arithmetic: A 140-year survey of American managed retreat cases documents the recurring pattern in which communities relocated by federal programs lose cohesion even when individuals are compensated, a distinction the 1928 Act's framework was never designed to recognize.
- Galveston's compounding bet: A 2024 Washington Post investigation found new condominiums being built on Galveston Island despite projections that subsidence and sea-level rise could produce 1.9 meters of relative rise by 2100, another barrier island where protection infrastructure has encouraged development faster than risk has receded.