Four rivers meet inside Dayton, Ohio. The Great Miami, the Stillwater, the Mad, and Wolf Creek all converge within three-quarters of a mile of the central business district. On a quiet afternoon you could stand downtown and barely register them. In late March 1913, every one of those rivers came for the city at once.
Between March 23 and 25, rain fell across the Great Miami River watershed in quantities that sources still can't agree on. The Midwestern Regional Climate Center recorded roughly 7.6 inches at Dayton proper. MCD histories cite nine to eleven inches across the wider valley. The disagreement is almost beside the point. All that water funneled toward the same three-quarter-mile radius, and the Great Miami crested at 29 feet. The previous record, set in 1866, was 21.3. The levees protecting the business district had been designed for 23.
They broke between eight and nine in the morning on March 25.
What Happened at Street Level
Brigadier General George H. Wood of the Ohio National Guard was in Dayton when the river came. His military report reads like the writing of a man who kept taking notes because there was nothing else he could do.
By sunrise, North Dayton and Riverdale were already under water. Breaks along the Mad River levee had let the flood into eastern and southern Dayton. Wood watched from the Lehman Street levee as Riverdale north of the Main Street bridge disappeared. About fifty people stood on the levee with him. Then water from Riverdale poured over the driveway west of the bridge and cut off the levee itself.
The current across Main Street grew strong enough that boats couldn't reach the bridge. Someone stretched a rope diagonally from near the Bellevue Apartments to a telegraph pole so rescuers could work a safer line. By 10:30 a.m. the current through Riverdale was running so fast that rescue work had to stop entirely.
Wood tried to push south on Main Street with a detachment. Private Coble was torn from the line by the current. After 11 a.m. the soldiers took shelter in houses. They would not move freely again for days.
By 2 a.m. on Wednesday, March 26, the water had reached its peak. Four separate fires were visible from upper floors. Gas mains had ruptured. One major explosion was reported near Fifth and Wilkinson. A fire broke out at St. Clair and Third streets after dark and spread westward, burning through nearly two blocks of business buildings while floodwater kept fire crews from reaching the blaze. A survivor's letter described water twelve to fourteen feet deep in the street while three paint stores burned above it.
On First Street, the current ran "like a mill race." Downtown Dayton sat under an average of ten feet of water at the crest. In northern Dayton, water rose above second floors. South of town, the river spread more than a mile wide. Telegraph and telephone lines were down. Wood did not reach a working telephone line until Thursday, when a Central Union Telephone Company wire to Columbus was found at the Ludlow Street office and he could finally report to Governor James Cox.
Think about what that means for the people in those houses. Days without communication. No way to tell anyone outside the city what was happening, no way to know if help was coming.
The death toll remains contested. The Midwestern Regional Climate Center lists approximately 123 fatalities in Dayton proper. The National Weather Service estimates 98 to 123. Across the Miami Valley, current MCD history says more than 360 people died, while an earlier MCD-authored booklet placed the figure at over 400. The cleanup required removing 1,420 dead horses and 2,000 other dead animals, clearing 133,600 wagon-loads of debris, and cleaning and disinfecting 13,991 houses and cellars. Nearly fourteen thousand homes had to be scrubbed out before families could return to them.
What the City Could Do, and What It Couldn't
The timing is what stings. Dayton had already been trying. The city had started a flood-control program around 1910, had issued bonds and let contracts, had a contractor ready to begin work. The flood swept away the contractor's equipment. The Ohio History Journal later noted that the proposed protection would have been inadequate anyway. Three years of planning, destroyed in a morning by a problem the plan was never scaled to solve.
And yet, while the water was still draining from the streets, the city showed what its people could mobilize. The National Cash Register factory sat on high ground and became a relief center. NCR workers built nearly 300 flat-bottomed rescue boats, baked 1,000 loaves of bread a day, produced hundreds of gallons of soup. Governor Cox officially named NCR president John H. Patterson head of the Dayton Citizens Relief Committee on March 27. Individual rescuers carried hundreds of people to safety. Wilber J. Schneider was later credited with rescuing 300 to 400 people. John Sloufman, 386.
Less than two months after the flood, on May 2, 1913, Dayton's business community and civic leaders formed the Flood Prevention Committee. Colonel Edward A. Deeds, an NCR executive, chaired it. Adam Schantz urged a fund of at least $2 million at a Dayton Club meeting. A fundraising campaign launched on May 26 ultimately drew contributions from about 23,000 citizens, raising more than $2 million for a comprehensive flood-protection program. This was not government money. This was a city taxing itself voluntarily because the legal machinery to do it officially did not yet exist.
That response was extraordinary. Twenty-three thousand people reaching into their own pockets while their basements were still damp. And none of it was enough, because the problem lived at the wrong scale. The rivers that converged at Dayton gathered water from hundreds of square miles of watershed. Protecting Dayton meant controlling water that fell on farms and towns that had never flooded, in counties whose residents had no particular reason to care about Dayton's levees. The city could rescue its own people, feed them, raise millions in weeks. What it could not do was govern a valley.
Following the Water Upstream
Deeds hired Arthur E. Morgan of the Morgan Engineering Company of Memphis. Morgan arrived in Dayton on May 5, 1913. He was 34 years old, a college dropout who had learned drainage work from his civil-engineer father. By 26 he had written Minnesota's Water Control Code. He had served as Supervising Drainage Engineer for the U.S. Department of Agriculture from 1907 to 1910 before founding his own firm, specializing in floodplain drainage and reclamation.
Dayton hired him for Dayton. But Morgan's method was to investigate every possible approach and survey flood conditions across the entire valley before committing to a plan. His engineers worked the watershed systematically, tracing tributaries upstream, measuring where water gathered and how fast it moved toward the confluence that had wrecked the city. Within six months, the investigation had reached a conclusion that changed the scope of everything: no single city in the Miami Valley could financially secure complete protection on its own. A coordinated valley-control system was feasible. A city-scale system was not.
Morgan and the committee studied and discarded alternatives. Diverting the Mad River into the Little Miami. Building bypasses around cities. Relying on channel improvement alone. Many small retarding basins. Combining flood prevention with hydroelectric power. Each was found impractical, too expensive, or insufficient. The plan that survived called for an integrated system of large storage basins with earthen dams upstream of the cities, combined with levees and channel improvements within them. A watershed-scale answer.
The trouble was that Ohio had no law capable of building it.
Inventing the Legal Container
Ohio's existing flood-control statutes were locally limited. No mechanism existed for uniting multiple interests across county lines, adopting an official plan, distributing costs, or enforcing project requirements. The Flood Prevention Committee needed a kind of law that didn't exist yet.
Veteran Dayton attorney John A. McMahon, working with Oren Britt Brown, drafted the Ohio Conservancy Act. The bill was introduced in the Ohio House on January 19, 1914. Governor Cox signed it on February 7, though some near-contemporary sources date passage to February 18 or final enactment to March 17. The discrepancy likely reflects the gap between signing, legislative passage, and effective date, but the documentary record doesn't resolve it cleanly.
What the Act created was genuinely new. It authorized the formation of conservancy districts through property-owner petitions to common pleas courts. A Conservancy Court, consisting of one common pleas judge from each county with land in the proposed district, would oversee the process. The court would appoint three directors. The directors' chief engineer would prepare the plan. Objections would be heard formally. Three appraisers would assess benefits and damages to every affected property. The district could exercise eminent domain, levy assessments tied to benefits, and issue bonds.
The Act was not without legal precedent in every respect. The U.S. Supreme Court, when it upheld the statute in Orr v. Allen in December 1918, cited Houck v. Little River Drainage District, a Missouri case involving a special-assessment drainage district. Multi-county assessment districts were not unknown. What was unprecedented, according to the ASCE, was the result: the first regionally coordinated flood-control system in the United States that employed retention reservoirs for controlled release of floodwaters. The legal innovation and the engineering innovation were inseparable. The Act existed because the engineering plan required governance at a scale Ohio law had never contemplated.
The Ohio Supreme Court upheld the Act in Miami County v. Dayton in 1915, ruling it a valid exercise of state police power. Even so, the Conservancy Court organized the Miami Conservancy District on June 28, 1915, by a vote of just 5-4. An earlier April 18 vote of 5-4 had overruled objections but failed to establish the district because six votes were required. The margin tells you something. One judge, either way, and the whole thing collapses. The most consequential flood governance experiment in American history survived by a single vote.
Who Paid for the Watershed
Morgan was appointed chief engineer on July 7, 1915. The Official Plan was adopted by the board on May 10, 1916, and formally approved by the Conservancy Court on November 24. Between the board's adoption and the court's approval lay seven weeks of public hearings, more than 3,000 typed pages of testimony, and five consecutive days of cross-examination of Morgan by opposing attorneys.
The opposition was not irrational, and the reasons matter for understanding everything that followed.
The original district included portions of nine counties: Montgomery, Shelby, Miami, Clarke, Greene, Warren, Preble, Butler, and Hamilton. Levees would be constructed in nine cities, from Piqua in the north to Hamilton in the south. Five dams would be built upstream. Four railroad lines would have to be relocated.
And in the Huffman basin, the entire town of Osborn would have to be purchased and moved to make room for a reservoir. Not a neighborhood. Not a block. A whole town, erased from its location so that a dam could hold water that would protect cities downstream. The district found it necessary, the Ohio History Journal reported, in the careful language institutions use when they are describing what they did to someone else's home.
About 60,000 pieces of property owned by nearly 40,000 owners were appraised. Roughly 30,000 acres were purchased and resold with flood easements attached. Construction would be financed by nearly $34 million in bonds, secured by appraised benefits and retired through assessments against benefited property.
The costs fell unevenly, and they fell hardest on the people farthest from the decision. Farmers living in or near proposed reservoir basins were told by local attorneys that their lands would be confiscated and they would also be taxed for the project. The Ohio History Journal account notes this as a source of "stubborn opposition." The Act required that damages be paid where damage was sustained, and no property could simply be confiscated. But the distinction between a flood easement and confiscation is clearer in a statute than it is standing on your own land, watching surveyors mark the boundaries of a reservoir you never asked for. Lands outside the cities were only partially protected because channel improvements were built only within city limits, and the extra cost of fully protecting rural lands would have exceeded the assessed benefits. The people who gave up the most land received the least protection.
Earlier prejudice against dams, partly rooted in previous government reports, had to be overcome through public education campaigns using newspapers, illustrated talks, maps, diagrams, and a working model of a retarding reservoir. Consulting engineers, including the respected H.M. Chittenden, testified that the plan had been worked out with extraordinary care and thoroughness. The first $15 million bond installment was sold in December 1917, after consultation with U.S. Treasury Secretary William McAdoo because wartime financing conditions were unfavorable. Construction ran from 1918 to 1922.
What the Flood Forced Into Existence
The system that emerged protected cities across five counties with five dams, 43 miles of levees, and channel improvements through nine communities. In the ASCE's careful language, it was the first regionally coordinated flood-control system in the United States. The concrete mattered, but the coordination mattered more. What Dayton's flood forced into existence was a legal and political framework that made a nine-county watershed into a single governed unit, with costs distributed by assessed benefit, damages paid where sustained, and a court overseeing the whole arrangement.
Five dams, 43 miles of levees, nine communities protected, ~60,000 properties appraised across nine counties, ~$34 million in bonds, built 1918–1922.
It was built against real friction. Farmers absorbed costs for cities downstream. Counties submitted to an authority imposed across their boundaries. A town was relocated for a reservoir it would never need for itself. The 5-4 vote that created the district reflected genuine disagreement about whether this kind of governance was legitimate, and the U.S. Supreme Court had to weigh in before the question was settled.
None of that friction was illegitimate. The people who objected were, in many cases, the people who would pay the most and benefit the least. What the flood had revealed, and what the long institutional chain from committee to legislation to court challenge to construction confirmed, was that the rivers did not care about the boundaries humans had drawn. Protecting a city required governing a valley. And the question the Act couldn't fully answer was what you owe the people upstream who make that protection possible.
- MCD's funding gap today: The Miami Conservancy District says its Dam Safety Initiative expired in 2022 and that it currently has no long-term funding mechanism for the infrastructure that has protected the valley for over a century.
- Buyouts as permanent retreat: After the 1997 Grand Forks flood, the city ran a voluntary buyout program that purchased roughly 850 properties, about 10 percent of its housing stock, converting neighborhoods into open space that changed the city's map permanently.
- National buyout patterns: Mach et al. found that FEMA-funded voluntary buyouts from 1989 to 2017 totaled 43,633 flood-prone property acquisitions across 49 states, with a median time from flood to buyout completion exceeding five years.
- When code meets enforcement: After Hurricane Andrew destroyed or damaged roughly 157,000 homes in South Miami-Dade, a University of Florida review found Florida's building code was administered differently by more than 400 local jurisdictions, revealing that the problem was enforcement capacity as much as code text.