by Martin Doyle
Throughout the mid-nineteenth century, the Mississippi delta flooded repeatedly. The levee districts had been active in building and maintaining a growing line of levees, but their work was repeatedly destroyed by the spring floods on the river. Yet the region was a burgeoning agricultural area with growing importance in the national economy. The effect on the entire southern regional economy, along with the prevalence of New Orleans as a port of national significance, meant that the federal government felt compelled to do something about the chronic problem of flooding in the delta. Instead of doing any actual work itself on flood control, Congress chose to fund an enormous research project on flooding in the Mississippi valley as a way to nationalize the problem: the project would approach the problem in the context of the Mississippi valley, but with insights that would benefit the rest of the nation. That is, the federal government was going to fund basic scientific research.
Congress initially envisioned a single study of flooding on the Lower Mississippi that would clarify the problem and the solution. It ended up commissioning two separate studies, each proposing a course of action. The two engineers assigned to the studies were wildly different, as were the implications of their studies for the future of flood control and federalism.
Civilian engineer Charles Ellet Jr. was first out of the gate and first across the finish line. Congress had initially assumed it would fund work to be done by the Corps of Engineers, which typically was the scientific and engineering headquarters for the federal government. But the stature of civilian engineering had grown enormously by the mid-nineteenth century, as had civilian distrust of the military engineers. Civilian engineers were not just prominently involved in the levee districts of the rivers of the nation; they were also responsible for the vast canal works, locks and dams, and increasing number of railroads crisscrossing the nation. Ellet was chosen in response to the demand by civilian engineers to have one of their own do the work of evaluating the Mississippi River.
Ellet’s initial career was typical for civilian engineers of that time: He cut his engineering teeth at age 17 as an assistant engineer on the Susquehanna Canal and then worked on the earliest surveys of the Chesapeake and Ohio Canal. His early training in engineering was purely hands-on amid the construction of the internal works of the nation. But then Ellet set himself apart from civilian engineers by seeking more formal education. He moved to France, the heart of civil engineering education. After a meeting with Lafayette himself, Ellet gained entrance to the École des Ponts et Chaussées, the world’s oldest civil engineering school. In addition to his formal studies, Ellet traveled to see infrastructure that was far advanced compared to that in America of the 1830s: expansive canal networks, suspension bridges, and several massive dams and reservoirs in Switzerland.6
Ellet returned to the United States and developed into one of America’s foremost engineers with staggering breadth of expertise. He was the chief engineer for the infant James and Kanawha Canal Company and then moved on to propose and design the first suspension bridges in America across the Schuylkill River in Philadelphia and the Ohio River at Wheeling, Virginia. Between projects (canal building and bridge building), he wrote a treatise developing a theory for establishing toll rates and freight charges to finance internal improvements first for canals and eventually for railroads. During the Civil War, Ellet shifted his attention to naval architecture and warfare and then to military strategy overall. He was one of the earliest critics of the military tactics of General McClellan, who was later relieved of command by Abraham Lincoln.
In any one of these areas, Ellet would be viewed as an accomplished scientist and engineer. Taken together, his abilities won him wide regard from civilian engineers, who hailed him as the luminary of his generation—the Steve Jobs of nineteenth-century science and engineering, capable of innovating solutions to any problem. Upon his death in 1862, many engineering magazines eulogized Ellet for his technical achievements. Almost a century after his death, the prestigious Journal of Political Economy remembered Ellet for his ability to think both broadly and synthetically; he was “one of those rare men of affairs who scan their activities from a general position and are capable of marshaling a seemingly incoherent mass of physical and economic data into a series of significant relations.”7
The other engineer tasked by Congress to investigate flooding on the Mississippi River was Andrew Humphreys. If Ellet was an intellectual fox, Humphreys was the hedgehog—his life and career were bound inseparably and myopically to floods, levees, and the lower Mississippi River. Humphreys had graduated from West Point when it was still the single official nursery for the infant practice of American engineering. Humphreys graduated thirteenth in his class of thirty-one at West Point and then meandered through a series of unnoticeable posts intended for unnoticeable officers. Whereas Ellet had risen to prominence in civilian engineering circles, Humphreys stagnated in military ones.8
Humphreys got his break during a coastal survey, when he impressed his superior officer with his careful, meticulous work. After this relatively innocuous start, Humphreys was selected to lead the military engineers for the Mississippi River survey. He was assigned the same task as Ellet, but would represent a military engineer’s view of the problem.
Not only did the two men come from different backgrounds, but their approaches were wildly different. Humphreys came at the problem militaristically and empirically; he organized three teams to study the science of hydraulics and sediment movement in the Lower Mississippi River. He collected and compiled data, working the problem like an army officer leading a platoon of data-collecting soldiers. In contrast, Ellet treated the task as if he were an academic on sabbatical; he moved to New Orleans to study the problem, bringing his wife and family instead of a team of surveyors. He relied on already available information about the Mississippi, as well as some of his own previous work on navigation from the Ohio River. Mostly, Ellet thought and wrote.9
In the summer of 1851, while Ellet was wrapping up his report, Humphreys had what can only be described as a mental breakdown. The physical and mental exertion of the survey had overwhelmed the army officer, and Humphreys ceased all work on the project. He first moved to Philadelphia to recuperate and then arranged to be sent to Europe to continue his recovery and examine the river works there. It was not until 1857 that Humphreys restarted the survey work of the Mississippi. By then Ellet’s report had sat in congressional hands for over five years. On his return Humphreys took on an assistant, Lieutenant Henry Abbot, whom he sent back to the river to finish the remaining surveys and measurements. Finally, as the storm clouds of the Civil War gathered, Humphreys and his assistant Abbott wrote their report—a 500-page treatise on river hydraulics and the Mississippi River that would be filed to Congress just after the Civil War began.
There are many differences between the reports, but at the most fundamental level, Ellet saw the potential of the horizon while Humphreys and Abbot saw the constraints of the immediate. The Humphreys and Abbot report began, continued, and ended with sheer density, starting with the rather intimidating partial title Report on the Physics and Hydraulics of the Mississippi River. From there on, the military engineers laid out a masterful dissertation on the way rivers work. Their report is thick with data, tables, calculations, and footnotes. The survey had, in the words of Abbot years later, taken the “pulse of the great river,” painfully measuring the intricacies of its operation upstream and downstream through all depths and all available floods. In almost every way, it represented the determinism and reductionism that would be the hallmarks of early twentieth-century engineering. Indeed, the science in the report was, and is, regarded as pathbreaking. Over the decades following, Humphreys’s report established him worldwide as an authority on the science of hydraulics, and along with his conquests in the Civil War would earn him the coveted position of chief of engineers for the U.S. Army Corps of Engineers.10
Humphreys concluded his analysis with what became the drumbeat of his r
emaining career: levees were the only option. After a detailed evaluation of data and formulas, Humphreys laid out his full attack on all other opinions, condescendingly brushing aside non-levee approaches—such as upstream storage reservoirs or bypasses—as “unrealistic.” Whenever possible, Humphreys used his data to undermine any other argument, particularly those of Ellet. He would often italicize the words direct measurement to drive home his superior basis for authority on all matters; he had measured, others had speculated. And after hundreds of pages of his initial assault on other options for flood control, he finally arrived at his recommendation that “An organized levee system must be depended upon for protection against floods in the Mississippi valley.” The reader is left with a sense of relief; the pummeling is over, and we have a solution. He then goes on to italicize his empirical, rather than intuitive, method: “It has been demonstrated that no advantage can be derived either from diverting tributaries or constructing reservoirs, and that the plans of [meander] cutoffs, and of new or enlarged outlets to the gulf are too costly and too dangerous to be attempted.”11
But the essential point was not just that Humphreys had solidified the science of levees, but that he also doubled down on the existing policy of flood control: federalism. Included with his tabulations of air temperatures and elevations of historic flood stages are thirteen pages covering the history of levee district governance—a policy document within a scientific study. He pointed specifically to the levee districts of Mississippi as a “judicious State system of operations” that other states should emulate. The end result of the Humphreys-Abbott analysis was a recommendation to sustain the existing mode of federalism: the federal government should provide the science; the states should oversee systems of levee districts.
By contrast, Ellet’s report could not have been more different in tone and vision: The Mississippi and Ohio Rivers: Containing Plans for the Protection of the Delta from Inundation and Investigations of the Practicality and Cost of Improving the Navigation of the Ohio and Other Rivers by Means of Reservoirs. Right from the start, Ellet’s report was different; its title linked the Mississippi to the Ohio and identified the solution—reservoirs. Whereas Humphreys produced a treatise on hydraulics and the Mississippi River, Ellet focused on watersheds and governance and analyzed the failures of federalism.12
The thrust of Ellet’s report was that the states of the Lower Mississippi valley, particularly Louisiana, were being negatively affected by the actions of many upstream states. Thus the entire nation had a responsibility to the downstream states. Ellet looked at the hydrology of floods and realized that planning levees around the peaks of historic floods, as Humphreys was doing, was simply naïve because past floods were poor predictors of deluges that were inevitable in the future. Ellet estimated that swamps and forested areas of river valleys upstream were themselves vast reservoirs for floods, because large floods could simply spread out onto the valley floor and be retained just as they might be in reservoirs. However, as the immense agricultural development in the Midwest continued, it would lead to the removal of forests and bottomlands upstream, which in turn would affect the Mississippi valley downstream.
For this reason, Ellet took the federal government itself to task. Until 1849 undeveloped swamplands, lying mostly along rivers and floodplains, were in the public domain. They were owned by the federal government, similar to U.S. Forest Service lands of the twenty-first century. To stimulate land development and settling of the interior, Congress passed the Swamp Land Acts of 1849, 1850, and 1860, which transferred ownership of nearly 65 million acres of swamplands in fifteen states to the governments of those states. This transfer of ownership had a caveat: funds from the sale of these lands by the states had to be used to build the levees and drainage required for land development. The federal government was not only encouraging settlers to move into flood-prone lands but also requiring deforestation of upstream swamps and floodplains of the upper Midwest and thus increasing the flood load on the downstream Lower Mississippi valley. Ellet saw the federal government as part of the problem and said that it had to be part of the solution as well.13
Because the downstream floods were so crippling, and because they would only get worse, Ellet’s solution demanded that anything and everything should be tried. He argued for opening river outlets on the Lower Mississippi River to Lake Borgne and enlarging outlets to Bayou Plaquemine and the Atchafalaya River. Ever mindful of the effect of upstream actions on downstream communities, he was against straightening the river not because he thought the result would be ineffective, but because he worried that doing so would exacerbate flooding downstream. Most importantly, Ellet advocated for reservoirs. He suggested establishing a system of upstream reservoirs in the tributaries of the Ohio and Mississippi. The reservoirs could store floods in the spring and then release the flows gradually during the summer months to minimize floods as well as supplement low flows for navigation. The reservoirs Ellet envisioned were nontrivial, estimated to be staggering in size and costs. But Ellet believed that reservoirs were essential. Levees had a role, but they should be used cautiously and as a last resort. Relying solely on levees, Ellet thought, gave “a delusive hope, and most dangerous to indulge because it encourages a false security.”14
In 1862, only months after Humphreys submitted his report—but years since Ellet had submitted his—Ellet was killed on the Mississippi River at the First Battle of Memphis. Humphreys, in contrast, rose to fame during the Civil War. He went on to serve as chief of engineers for the entire Corps, where he would lead the implementation of his own recommendations. The centerpiece of his recommendations, and therefore the centerpiece of Corps flood control policy, was a strict “levees only” doctrine for flood control. The Corps moved gradually into the realm of flood control, but only on its own terms: levees.
Over the ensuing decades, levees grew in size and number, expanding well beyond the Lower Mississippi valley. By the start of World War I, fifty-two levee districts had been created on the Upper Mississippi, protecting several hundred thousand floodplain acres, and over 330 miles of levee were built along the Illinois River. Several hundred more miles of levees were built to protect the agricultural lands of the Sacramento Valley in California.15
Ellet had lost the flood control battle to Humphreys for ideological reasons as much as scientific ones. Humphreys’s support for state-driven, levee-centric flood control was a logical extension of late-nineteenth-century federalism; Ellet’s proposed solutions—requiring systematic planning on a national scale and substantial financing by the federal government—would have necessitated reorganization of federalism itself. If there was going to be any approach to flood control other than building and maintaining levees under state government oversight, it would require some kind of event or events that undermined the nation’s perception of levees and federalism. A series of mega-floods provided one such event, and the Great Depression provided the other.
CHAPTER 4
Flood Control
When the spring rains began in 1927, farmers along the Mississippi were in their typically precarious position of estimating the odds of losing that year’s crop. On the one hand, flood control had improved: there were over 250 million cubic yards of earth forming the levees of the Lower Mississippi—enough to bury 240 square miles under a foot of earth. Levee district engineers were widely recognized scientists who traveled internationally and wrote articles for prestigious scientific journals. Levees were also more effective: from the close of the Civil War to 1927, the chances that a flood would inundate farms along the river had been cut roughly in half.1
Yet booms in population, industry, and agriculture had made regions along major rivers all the more vulnerable to the impacts of levee failure, and the levees—along with floodwalls in urban areas—were far from perfect. The 1897 Mississippi River flood inundated 87 percent of the area protected by the Mississippi Levee Board (2,270 out of 2,593 square miles). And after considerable rebuilding, the levees fail
ed again in 1912, inundating over 52 percent of the district’s protected area. Urban areas bore no less impact. The flooding of Pittsburgh’s central business district in 1907 put 100,000 workers out of work amid growing labor tensions. The subsequent 1913 flood on the Ohio River was 15 feet above any previously recorded heights, caused over $300 million in damages, and killed more than seven hundred people—more than had died in the Great Chicago Fire of 1871. The growing population in California was also experiencing devastating floods as the Sacramento River continued to surprise everyone who worked on it. Local engineers at the turn of the twentieth century estimated that the peak flooding the Sacramento could ever experience would be 300,000 cubic feet per second (cfs); in both 1907 and 1909, the Sacramento neared 600,000 cfs.2
But no flood had greater effect on the psyche of a region, and the nation, than the 1927 Mississippi River flood. Almost any description or statistic of the 1927 flood seems over the top, but no single account fully conveys the severity and scope of the disaster. One hundred breaks in the mainline Mississippi levees inundated over 26,000 square miles across seven states—a region larger than the total area of West Virginia. Over 700,000 people lost their homes and as many as 300,000 were rescued from houses, rooftops, levee crowns, and even trees. Tens of thousands of livestock were killed, and millions of acres of the nation’s most productive farmlands were destroyed.3
Beyond the direct impacts, farmers in the Midwest were dependent on the Mississippi to export their products through New Orleans. With the river and region crippled by the flood, all the agricultural exports were bottlenecked in the Upper Midwest. For weeks, trains were unable to cross the Mississippi south of St. Louis because more than three thousand miles of track were underwater. Total losses were diffused across the nation and hard to estimate, but they reached as much as $1 billion at a time when the federal budget was typically less than $3 billion. As agricultural commodities spoiled and wasted in Missouri and Kansas because of flood waters in Louisiana and Mississippi, flood control began to appear more and more a federal rather than local issue.4