by Martin Doyle
The original city was built on a topographic anomaly—a small natural levee on the north bank of a sweeping meander of the Mississippi River. This slight rise has stayed surprisingly dry through floods of the past, largely because the river had so much area to flood elsewhere: the river could send much of its flow west through the Atchafalaya basin over two hundred miles to the northwest, release more of the flow through Lake Pontchartrain just north of the city, and pass a smaller amount downriver past the small hill of downtown New Orleans. These natural outlets to the Atchafalaya and Lake Pontchartrain, however, were closed off during the Corps’ levees-only era in the late nineteenth and early twentieth centuries. Then the Mississippi was allowed to flow only through its channel past New Orleans. The city thus became fully dependent on the levees for passing floods. The other significant geographic turn was suburban sprawl. Starting after World War I, the population of New Orleans began spreading outward from downtown, off the high terrain of the natural levee toward lower-lying areas near the lakes, particularly Lake Pontchartrain to the north.
Levee breaks and flooding in New Orleans during Hurricane Katrina.
As the city and region have grown, two trends have been at work to ensconce the city in its own flood control infrastructure. First, a combination of factors—some natural and some human-induced—are causing subsidence; the city is literally sinking deeper into the swamp. The original downtown remains on a slight rise above sea level, as in the early eighteenth century. But an increasingly large portion of the expanding city outside downtown continues to sink below sea level, particularly the area known as the Lower Ninth Ward to the east of downtown and those areas to the north along Lake Pontchartrain. With large swaths of the city below sea level, rainwater and groundwater seep relentlessly into the marshy soil of the city—hence the need for aboveground tombs rather than belowground graves in the city’s cemeteries.
In addition to being at sea level and the delta of a massive river, New Orleans is located in one of the wettest places on Earth, and direct rainfall often causes flooding. Average annual rainfall exceeds 50 inches per year, and extreme events can be downright shocking: in 1927, before the Mississippi River flood crest reached New Orleans, the city received 18 inches of rainfall in 48 hours. To stay afloat, the entire city has to be dried through a series of drains and canals that collect rainwater or seeping groundwater. Enormous hydraulic pumps work to pump this ever-seeping water up over levees and into canals that run from the edge of downtown and drain north to Lake Pontchartrain. Without those pumps, local groundwater and rainwater have nowhere to go. They would accumulate and effectively turn the below-sea-level city into a massive puddle.
At the turn of the twentieth century, a substantial wetland fringe existed between the city and the shores of Lake Pontchartrain to the north, but the city began creeping into this bottomland. A local levee district built a seawall offshore in the lake and then created dry land behind the wall by pumping sand from the lake bed up and over the wall. As the wetlands along the canals drained, more dry land was available to develop the growing city. The city thus expanded northward toward the lake into former wetlands. Canals draining water from downtown toward the lake were lined with single houses; rather than passing from downtown through wetlands toward the lake, the canals passed from downtown through newly settled suburbs where land was chronically subsiding. Low ridges of dirt initially built to keep drainage flow in the canals instead became critical flood control infrastructure for these new portions of the city as it sprawled into the former wetlands and lake bed. As more homes were built and demand for land increased, it was more and more difficult to expand the footprint for levees; building levees higher meant expanding their base foundation, which consumed land that could otherwise be used for development. Instead, sheet pile walls and eventually concrete walls—floodwalls—were built on top of existing small levees to increase their height and protect the ever-sinking city.
With the levees and floodwalls growing higher and the land subsiding lower, the city became a crater surrounded by walls that kept out the rising river or the rising tides. The only dry place was in the middle, thanks to the ditches and canals draining the city. By 2005 there were 172 miles of drainage canals, and thousands of miles of storm sewer lines, along with vast pumps constantly shoving water up and over the encircling walls and into canals to flow toward the lake. New Orleans resident and historian Douglas Brinkley described the city as “no more stable than a delicate saucer floating in a bowl of water.”23
As cobbled together as this infrastructure was, the entities responsible for it were even more so. Each unit of government, from federal to local, had mixed-up and multilayered roles and responsibilities. Metropolitan New Orleans had a half dozen levee districts, each one a large, well-financed organization whose essential role was keeping its particular area of the city dry. These districts, like Nimrod’s Mississippi Levee Board hundreds of miles upstream, had broad powers to collect taxes, sell bonds to finance flood-control projects, and administer a wide range of activities. In fact, these activities could and did extend beyond simply flood control to include land development. By 2005, one of the larger districts—the Orleans Levee District—had an annual operating budget of $40 million and over three hundred employees. It was in charge of maintaining just over 100 miles of hurricane levees with 127 floodgates as well as 28 miles of Mississippi River levees with 76 floodgates.
The Corps of Engineers was also an integral part of the flood control system in New Orleans. The Corps had been building levees and strengthening existing levees along the lower Mississippi River since 1928. The agency also began construction on its vast series of bypasses and floodways to divert flows out of the Mississippi River. But by the mid-twentieth century, the Corps began to realize that floods in New Orleans were as likely to come from the lakes due to a storm surge as to come from the river due to a flood; the flood of 1849 was the last time the Mississippi River levee was breached within New Orleans city limits.
To address the threat of hurricane storm surges, in the 1960s, Congress authorized the Corps to construct protection along Lake Pontchartrain. This work was to include barriers along the lake itself to defend against storm surges, but it would also protect the primary canals going from the lake into the city. Under normal conditions these canals work as drains to take seeping water away from the city to the lake. During a hurricane, they potentially reverse and serve as a funneled pathway for a storm surge to flow past the lakefront barrier and deep into the city. To address this potential, the Corps proposed using floodgates at the mouth of each canal where it enters the lake. The gates would close during a hurricane and keep the potential surge from entering the canals and funneling into the city. Floodgates at the mouths of canals would be, according to the Corps, effective and the least costly option.
The levee districts and the Sewerage and Water Board, which managed the drainage system, also had a say in the decision because they were required by Congress to pay for some of the project costs. The local boards and districts didn’t like the floodgate approach proposed by the Corps. If the gates were closed, the accumulating rain and groundwater that normally would be pumped into the canals would have nowhere to flow. Keeping the floodgates closed for any period of time would fill the canals with water that then backed up into the city, causing more water backup in drainage ditches throughout the city and returning the city to its natural puddle-like condition. Local districts argued that instead of floodgates, the canals should be left open at the lake—leaving them susceptible to storm surges—but levees and floodwalls along the canals should be flood-proofed against these surges by being made higher and stronger. Higher floodwalls would allow a storm surge to enter the open-ended canals without overtopping. The districts would benefit further because the increased height of levees and floodwalls would help with their chronic interior drainage problems. Floodgates addressed only the issue of flooding due to storm surges; higher floodwalls would help with both f
lood control and interior drainage.
At that point, federalism raised its complex head: the Corps generally considered interior drainage to be a local issue and therefore not fundable by the federal government. So the local Sewerage and Water Board maintained internal drainage—all the pumps and ditches and canals—while the levee districts maintained the levees and floodwalls along the canals. As long as the levees along the canals were tied only to internal drainage, the levee districts had to pay all the costs of construction and maintenance. But by tying the floodwalls along canals to hurricane protection, the levee districts could potentially get the federal government to pay for a large portion of improvements to local interior drainage infrastructure. With the Corps equivocal on whether floodgates or floodwalls were chosen, so long as it was the least cost to the Corps, the levee districts lobbied and convinced Congress to authorize this alternative approach.24
From the perspectives of engineering and risk avoidance, these approaches were fundamentally different. The Corps would have focused its attention and resources on ensuring the integrity and functionality of three critical floodgates—effectively hydrological strongholds. The floodwall approach, in contrast, would require a strong infrastructure for the entire length of the levees and floodwalls along the canal system; the soundness of this flood control system would rely on the strength of miles and miles of levees and floodwalls. The adage about a chain being only as strong as its weakest link is appropriate in thinking about using a levee and floodwall system to protect a sub-sea-level city; a single weak point undermines the entire system.
Federalism relies on the local government doing its part—and on the central government not doing more than is needed. When infrastructure is involved, this means the federal government—the Corps—does some of the work but then hands off significant roles and responsibilities to local groups. One such case is Peter Nimrod’s maintenance of the levees at Greenville. In New Orleans, the handoff was complex and problematic. First were the half dozen levee districts, each with its own board of directors, spending priorities, and habits. Second was the involvement of various other government units, such as the separate sewerage board in charge of pumps in the canals. Adding to these layers, the Louisiana Department of Transportation had the role of approving activities that might compromise the levees, as well as directing training for levee district board members and inspectors. What’s more, the local government units meant to focus solely on flood control and the levees had a surprisingly undisciplined attention span. The Orleans Levee District, for instance, used some of its revenue to build parks, marinas, an airport, and a dock for casino boats. In 1995 the district published a brochure that read, “We protect against hurricanes, floods, and boredom.”25
Another problem was that risky future development was an essential element of the agreement to build levees along the lakes and canals. To justify the entire project when authorized and appropriated, the Corps and levee districts had factored projected new development into the cost-benefit analysis; fully 79 percent of the project’s financial benefits were to come from future development rather than existing development. Jefferson Parish added 47,000 housing units, and Orleans Parish added 29,000 units in what had previously been swampland but was rapidly turning into drained, leveed, but subsiding suburban neighborhoods.26
After completing the construction work for its part of the project, the Corps had handed off most of the levees to the local districts, which were then expected to take up the responsibilities of operation and maintenance. But some levee districts refused to take over those responsibilities because they were unclear when they were to do so or because they were willing to take over only minimal roles of maintenance, such as grass and weed control. In general, the local districts claimed that “major” problems remained the responsibility of the Corps, while “minor” problems were the responsibility of the districts—a casual distinction at best. The conflicting or at least inconsistent roles of the different agencies could be seen in the elevation of the levees themselves; their heights varied depending on who was in charge. At one pumping station where jurisdictions had (opaquely) changed hands, a concrete wall was connected to an earthen levee that had a completely different elevation. The elevation of levees maintained by Peter Nimrod’s Mississippi Levee Board is ensured with laser-like precision, but New Orleans had at least five different elevations of different parts of the levee system protecting the canals, and these variations likely corresponded to whatever agency was responsible for that section of levee.27
On August 29, 2005, the weaknesses in the New Orleans flood control system were laid bare as Katrina worked its way across the region. Gaps and inconsistencies in the floodwalls and levees created weak points through which water first came into the city. The storm made landfall just east of New Orleans at 6:00 a.m. At 6:30 the floodwall along the 17th Street Canal breached, sending water into Jefferson Parish just west of downtown. By 9:00 a.m. the Lower Ninth Ward had over 6 feet of water; later that morning a storm surge sent water into St. Bernard Parish, where water levels eventually reached 10 to 15 feet and covered 95 percent of the parish. The Industrial Canal floodwalls breached after one more hour, sending water into the Lower Ninth Ward through a gap large enough for an entire barge to be sucked through it and deposited in the neighborhood. When the London Avenue Canal levees breached later that afternoon, water poured into broad areas throughout New Orleans. It kept pouring through breaches in levees and floodwalls for two days, until noon on Wednesday, covering large swaths of the city with water. Only the old part of the city, sitting on the natural rise along the river, remained dry.
Federalism created the backdrop for flood control infrastructure that failed during Katrina, and federalism also drove part of the chaos of disaster response. Much went wrong before and after Katrina, but at the local level much also went right. In the two days before Hurricane Katrina, over a million people were evacuated from New Orleans—about 85 percent of the population—a significant achievement for a geographically isolated city. Following Katrina, search and rescue efforts were dominated by local and state organizations. The Louisiana Department of Wildlife and Fisheries, composed largely of personnel familiar with the affected area, used more than two hundred flat-bottomed boats to scour the city and rescue people from the rising waters. The locally placed Coast Guard likewise used 4,000 people, close to 40 aircraft, and dozens of small boats to conduct an outpouring of relief works.28
But much also went terribly wrong, particularly at the federal level. Response to Katrina was heavily influenced by the shadow of the terrorist attack of September 11, 2001. Before 9/11, disaster relief and FEMA reported directly to the White House, forming a direct chain of command from president to actions on the ground. Following 9/11, however, the administrative home of disaster relief was shifted to the Department of Homeland Security. This decision reflected the assumption that the most ominous disasters of the future likely would be terrorist attacks. Thus an apparently simple bureaucratic move by the federal government made natural disaster response the stepchild of homeland security.29
Federal disaster response was mortally slow, in part because of this reorganization. Federal officials were slow to appreciate the magnitude of the catastrophe. Over 24 hours after 20,000 people had fled to the Morial Convention Center, where they were trapped without food or water, Secretary of DHS Michael Chertoff stated on National Public Radio that “Actually, I have not heard a report of thousands of people in the convention center who don’t have food and water.” Even after federal responders recognized the scale and severity of the situation, FEMA’s response was governed by the new 400-page National Response Plan, a manual prepared in the wake of 9/11, which became a chokepoint for federal decision making in the rapidly evolving events following Katrina.30
The use of National Guard troops was an obscure but critical battle line drawn by federalism amid the fetid waters of Katrina. In the late twentieth century the National Guard was typically used
for natural disaster recovery, but its involvement was always in response to explicit requests by governors of affected states. This practice represents an important distinction between the National Guard and the regular Army. The National Guard is intended to be the military presence within U.S. boundaries and under the command of the affected state’s governor. Federal troops—the regular Army—cannot be used within the United States unless the state government is unwilling or unable to suppress violence that is in opposition to the U.S. Constitution.
A president can also federalize National Guard troops, effectively putting a state militia under federal command; this would be much the same as sending in federal troops to regulate state behavior. While seemingly outdated, this practice is a militaristic version of federalism: a state check on federal power. In 1957, against the governor’s will, Eisenhower used the Army’s 101st Airborne Division and also federalized the Arkansas National Guard to enforce school desegregation in Little Rock. It was one of the very few times that such actions have been taken. Sending federal troops to enforce law and order against the wishes of a state governor, or federalizing a state’s National Guard against the governor’s wishes, raises the ire of federalists perhaps more than any other potential action a president could take—especially in a southern state.31
Against the backdrop of a flooded New Orleans, Louisiana Governor Kathleen Blanco requested federal troops to assist in disaster recovery so that Louisiana National Guard troops could focus on law enforcement in the city; this way, federal troops would be doing humanitarian work only while National Guard troops would be conducting actions of a more militaristic nature. The states affected by Katrina did not have full use of their own National Guard troops because in late summer of 2005, when Katrina struck New Orleans, about 40 percent of the National Guard units from Louisiana and Mississippi were in Iraq for the war on terrorism. The U.S. Northern Command, a unit of the federal military created after September 11, 2001, to protect the continental United States, moved to the Gulf Coast to support FEMA. But stories of widespread looting and murders coming out of New Orleans led the Bush administration to question whether local law enforcement and the National Guard, under local command, were really up to the task. Yet federalizing National Guard troops or sending in federal troops would require the governor’s permission unless an actual insurrection was in progress. Despite the continuous violence and chaos, looting is not the same thing as an insurrection.32