The Source
Page 9
Gerry Galloway is blue-blood Corps. His father was a West Point grad who then became the commandant of engineers at West Point and retired as a major general after commanding the vast Army base at Fort Belvoir outside of Washington, D.C. Following in his father’s footsteps, Gerry graduated from West Point and joined the Corps. He rose quickly through its ranks while picking up a master’s degree from Princeton in engineering, another from Penn State in public administration, and a Ph.D. in geography from Chapel Hill. His early and middle career in the Corps included one of the most prestigious appointments; he became commander for the Mississippi Valley District, managing all federally constructed mainstem levees through the delta.
But then Galloway pivoted and took a different trajectory. Rather than continuing in the Corps hierarchy and potentially becoming chief of engineers like Andrew Humphreys a century earlier, Galloway left the Corps—in a way—by joining the faculty at West Point. His career there was no less illustrious; he quickly became the dean and chief academic officer at West Point and was promoted to brigadier general. Through his role at the Corps as a practicing engineer and then his role in training future generations of engineers at West Point, Galloway became known among river managers in the United States as a flood control engineer; when his name is brought up in Iowa or Mississippi, levee district managers—including Nimrod in Greenville—inevitably know who he is and refer to him simply as “the General.” But Galloway was also outside the normal Corps in the early 1990s; he had the bona fides and blue blood of a career Corps man, but he had also spent a portion of his time at the margins of the typical Corps—part academic, part Army engineer.
In 1993, perhaps because of this insider/outsider status, President Clinton tapped Galloway to analyze the aftermath of a massive flood in the Midwest. That year had been the wettest summer in 100 years for much of the Midwest. More than 1,800 miles of rivers and tributaries were inundated by record high flows, including over 500 miles of the upper Mississippi, 400 miles of the Missouri, and almost 200 miles of the Iowa and Kansas. Every river and every tributary was swollen with water and stayed that way, so that flooding became almost normal. The Missouri River at St. Louis first went above flood stage in early April and would not fully go back down until mid-October. Over 26 million sandbags were used in fighting floods in 1993, but to little avail: as rivers went over their banks, they eroded the roads, along with millions of tons of topsoil as the flooding broke levees and raged across the steep, narrow Missouri River floodplain. In some places the river scoured massive divots in the floodplain, creating a landscape that resembled an area bombed during a war. In other places the flooding river blanketed farms with several feet of sand, effectively sterilizing the fields. The 1993 flood far exceeded the 1927 Mississippi flood in spatial extent and economic impacts: it inundated over 20 million acres as compared to the almost 13 million acres flooded in 1927.12
At the time of the flood, Gerry Galloway was at West Point. Twenty-three years after the flood, Galloway—now retired from the Corps and West Point but still working more than full-time for the University of Maryland—recalls the 1993 flood as “the first CNN flood.” It was also the first natural disaster of the Clinton administration, which tapped Galloway to wrestle with floods in the United States. The general went on leave from West Point and was given a team of experts to work with him in office space in a brownstone right on Lafayette Square, facing the White House. But Galloway’s task was not to oversee the rebuilding of infrastructure, the operational aftermath, or disaster relief. Rather, he was brought in to untangle the causes and consequences of the flooding and evaluate the government—federal, state, and local—responses to floods. Like Humphreys and Ellet over a century earlier, Galloway was asked to take a step back and rethink how floods were being dealt with.
When recalling where he started to tackle the problem, Galloway went to the immediate aftermath of the 1936 Flood Control Act: “After thirty-six, nobody asked whether we should continue developing and building in the wrong areas.” This was the central sentiment in the Galloway team’s approach to the 1993 flood, and for the remainder of Galloway’s career: Before talking about flood control, should we be living and working in floodplains in the first place?
In the late twentieth century, flood control infrastructure had created an enormous sense of hydrologic security. But it also unintentionally created a perverse incentive that drew people into areas previously considered too high risk for developing. Levees initially constructed to protect farms and a few rural houses were now protecting neighborhoods, industrial parks, and portions of cities. Flood control infrastructure began to be economically justified based not on protecting what actually existed, but on what could exist if floods were eliminated. When evaluating a potential flood control project, a cost-benefit analysis may show the project was not justified if used only to protect farms. But if flood control led to new industry and housing developments replacing the farms, then the cost-benefit ratio for justifying projects could be flipped in a favorable direction to support the building of new or expanded infrastructure.
This new logic had developed after the 1936 act and was prevalent throughout much of Galloway’s career in the Corps. It created a tautologous rationale that led to flood control infrastructure and floodplain development being codependent—working together to encourage and even require development in extremely hazardous areas. Not only did proposed flood control projects increase the likelihood of such development, they depended on it. The combination of levee districts, state governments, and federal engineers building mammoth reservoirs and precisely engineered levees created the appearance and assumption that flooding had been eliminated. Society came to believe that the floodplains could be made flood-proof—and so, even though intense investment in flood control reduced the number of floods nationwide, economic impact of floods increased annually as more and more people lived on floodplains. Because of this shift, more infrastructure went up along floodplains, thus drawing more people to live on floodplains. And when events like the 1993 flood took place, they were disasters of staggering scale despite—or perhaps because of—the growth of flood control infrastructure.13
The federalism question of the early and mid-twentieth century had been “What should be the role of the federal government in preventing floods?” The federalism question of the late twentieth century became “What should be the role of government during and after floods?” The United States lacked an overarching policy or specific program for responding to natural disasters, but it did provide some financial relief following specific, significant floods like the 1927 flood. In 1950, however, Congress authorized a relatively minor bill to fund the repair of roads and bridges that had flooded along the Red River in Minnesota. This small act was a big deal: its funds were solely for reconstructive work, and its funds were solely for local work. From that precedent the federal role in disaster relief ballooned like the federal role in infrastructure had before it. By 1980, disaster response and relief had been reorganized into a new administrative home—the Federal Emergency Management Agency (FEMA). Through FEMA and growing spending and responsibilities, the federal government took on the role of providing temporary housing, grants for repairing damaged state property, unemployment compensation, health services, and payments to communities to offset lost tax revenue.14
In addition to funding these purely local effects of flooding, the federal government also developed questionable practices for evaluating disasters as being eligible or ineligible for relief funding. A governor from an affected state would request a disaster declaration from the president; if the request was granted, the state gained access to federal funding. These disaster declaration requests were routed through FEMA, which was outside normal cabinet-level oversight, so that their review was responsive only to the White House. Along with this lack of systematic agency oversight, the Stafford Act in 1988 explicitly prohibited the use of “arithmetic formulae” (such as benefit-cost analysis) as a basis for di
saster declarations. Disaster relief became codified as a solely political decision, outside traditional economic evaluation. Because disaster relief allows federal politicians to helicopter into stricken areas with the promise of federal funds close behind, for the White House a disaster request from a governor was an attractive request to approve: the Clinton administration declared 709 disasters, or about 1.7 disasters per week; the Bush Jr. administration declared 1,037 disasters, or 2.5 per week; and the Obama administration approved 938, or 2.25 per week.15
Federalizing disaster relief in turn reduced the burden carried by local and state governments: federal disaster relief required only a 25 percent nonfederal cost share, thus ensuring that the federal government would bear most of the relief costs. Yet this effort at local and state responsibility was consistently undermined: the nonfederal share was reduced or waived altogether following many severe disasters, and many local and state governments avoided most disaster relief costs altogether. Few local governments were calling for disaster funding reform, and essentially no sustained calls for fiscal discipline came from federal leadership because no senator or congressional representative was certain that his or her constituents would escape flood damage during the next season. The sprawling federal flood control infrastructure program of the mid-twentieth century had been replaced by an equally massive federal disaster relief and recovery assistance program of the late twentieth century.16
Even the few efforts by the federal government to correct this series of systemic mistakes backfired. When floodplain property owners were left without insurance options because the private insurance industry had little appetite for the enormous risk created by river floods, the federal government stepped in with a new low-cost insurance program: the National Flood Insurance Program (NFIP). But NFIP was the only insurance option for properties in flood-prone areas, meaning that the federal government was insuring only high-risk properties. With such high-risk exposure, NFIP payouts were higher than the revenues taken in by the program, so the program had to borrow chronically from the federal treasury. The unintended effect of NFIP was to federally subsidize the costs of living on floodplains.17
This entire flood control and disaster relief system drew the scrutiny of Gerry Galloway as he and his team sequestered themselves to analyze the causes and consequences of the 1993 Missouri River disaster. The existing infrastructure—all the massive dams upstream and levees downstream—had reduced much of the damage that otherwise would have occurred. But Galloway’s analysis started to reveal that insurance and disaster relief programs were especially problematic. Only about one in five properties damaged by the 1993 flood were actually insured; and while the NFIP paid out over $293 million, this amounted to less than 7 percent of the $4.2 billion of direct federal disaster relief costs. Even the cost-share efforts that had been introduced to increase local responsibility were undermined. In response to the flood, a Housing and Urban Development (HUD) program provided $450 million in grants for flooded communities to use as they saw fit. Some communities turned around and used this HUD money as their “nonfederal share” of flood disaster recovery, thus routing federal funding from one agency to provide the nonfederal funding of another agency.18
Galloway and his team chronicled these types of examples in high detail across the flood-affected region. Then they summarized and analyzed them in a report that reads more like a treatise on flood policy than a government report. This report to the White House—eventually known as “The Galloway Report”—was one of the first systematic analyses and criticisms of flood control and floodplain management in the United States. It was a full frontal analysis, and maybe even assault, on assumptions and practices that had accrued over the decades since Humphrey’s report became the bible of the Corps. Although the Galloway Report blamed the flood damages squarely on Mother Nature, calling the Midwest flood a “significant hydrometeorological event,” it also blamed the impacts of the flood on the demise of federalism: “Through provision of disaster assistance, the government may in fact be reducing incentives for local governments and individuals to be more prudent in their actions.” The thrust of the report was for the federal government to exhibit restraint so that local governments would exhibit initiative.19
The report was also one of the first to describe federal flood disaster relief as a moral hazard. When one person decides how much risk to take on while another person bears the costs if things go badly, the result is a moral hazard. Insurance companies have long worried about moral hazards created by overgenerous reimbursements: if their insurance policies were too generous, then their clients might behave in systemically risky ways. Similarly, in finance, the federal bailout of banks that were deemed too big to fail during the 2008 Great Economic Recession created a moral hazard: banks could grant risky loans and enjoy high returns, yet the U.S. Treasury bore the costs of those risks. This system created a moral hazard of giving banks an incentive to engage in risky behavior.
Galloway suggested that the existing federal disaster relief program created a nationwide moral hazard for floodplains. By providing artificially low-cost insurance along with disaster relief to entire communities, the federal government was encouraging risky decisions by individuals and communities on floodplains. Indeed, the most damning components of the Galloway Report focused on federal insurance programs as the root cause of the moral hazard. While only 20 percent of the flooded properties had flood insurance, over a third of those policies were purchased within a few days of the property being flooded—because the federal insurance program had only a five-day waiting period for claims. Those living along a river simply waited and watched the weather and flood reports over the winter and spring to see if they might be flooded. Then, within a week of the flood peak reaching their part of the river, they purchased insurance. They later received full payments from the federal insurance coffers.20
In the same way that Ellet took the federal government to task for its role in mid-nineteenth-century flooding of the Mississippi delta, Galloway took the federal government to task for exhibiting little restraint following almost any natural disaster. During the three natural disasters that occurred in 1993—the Midwest flood, Hurricane Andrew, and the Northridge, California, earthquake—nonfederal cost-share requirements decreased while disaster costs increased. Consequently, insurance programs went further and further into debt. This weaning of local responsibility perpetuated the view that the federal government would always be a benevolent big brother with ever-deeper pockets—a financial safety net. The end result was to encourage further development on dangerous floodplains.
Galloway was not a lone voice. An increasing number of scientists, engineers, and planners had likewise begun arguing for a first step of federal restraint. Perhaps more surprising, some state and local communities and governments—those benefiting directly from federal disaster relief—began wanting federal involvement reduced rather than enlarged after experiencing the moral hazard effect. Following a series of floods along North Carolina’s rivers in the 1990s, an editorial in the Raleigh News & Observer chastised the federal government for its role in creating the moral hazard at taxpayer expense: “The allocation of hundreds of millions in taxpayer dollars has led the federal government to undermine what state officials have been trying to do for decades—discourage development in floodplains.”21
Over the next decade, the Galloway Report became famous among engineers and planners as the kind of thoughtful document that sheds light on systemic problems and systematic solutions. Yet, in the Corps of Engineers and even within the federal government, Galloway did not benefit directly from his analysis and critique. A key implication of his argument was that floodplains might be better suited as environmental corridors—natural ecosystems of swamps and wetlands—rather than as shopping malls and suburban sprawl. Environmental groups praised Galloway’s report and its insights, but land developers and levee districts began to question his dedication to flood control. In 1995, a year after the Gallowa
y Report was published, Galloway’s name was floated for the position of Assistant Secretary of the Army for Civil Works—the chief political appointee over the Corps, and an appropriate placement for someone widely regarded as a luminary of river policy. Yet traditional flood control interests were nervous about Galloway’s acquired green label, fearing that he would undermine the status quo on which they depended. When Galloway was passed over for the appointment, the Washington Post noted the irony that the hint of an environmental label was enough to sink the nomination of an active duty brigadier general at West Point.22
After being passed over for Assistant Secretary, Galloway moved on. He became dean at the National Defense University and then in 1998 formally retired from public service. For the next two decades, General Gerry Galloway spoke on the radio or television to give his opinions about why the costs of floods were increasing rather than decreasing. He called for federalism to be applied more stringently—reducing the federal role and increasing local responsibility for flood control and disaster response. He occasionally penned editorials to revisit the recommendations of his earlier report. And he pointed to places that were particularly vulnerable to catastrophic flooding—places like New Orleans.
Of all the disasters, Hurricane Katrina has the dubious distinction of revealing the chaos of federalism, flood control, and moral hazards on live national television. If the 1993 Midwestern flood was the first to be called a CNN flood—for media coverage of the aftermath of a flood—then Katrina was the first flood to be covered as it took place. But in many other ways, the flooding disaster of Katrina is simply a continuation of historic trends, although focused in a small place for a short, intense period of time.
Like other cities prone to flooding, New Orleans has a complicated hydrology that convolves multiple sources of water with flatness. The city proper is wedged between the Mississippi River to its south, Lake Pontchartrain to its north, and Lake Borgne to its east (the lakes actually are tidal inlets to the Gulf of Mexico). Floods thus come at the city from all sides: from the river upstream, as in the case of the 1927 flood, or from coastal surges up into the lakes, as happens during hurricanes.