May 1998

Table of Contents

What's Happened in Floodplain Management Since the '93 Mississippi Flood?
-an invited comment

Civil Engineers Warn of Dam Failure Risk


Do Natural Disasters Lead to the Deliberate Ending of Human Life?

The Internet Pages

The National Disaster Reduction Initiative

Washington Update

Introducing EPC's NHEMATIS Project

EIIP Virtual Forum Expands Schedule

On The Line

Emergency Management Assistance Teams: Calling for the Cavalry During a Crisis

Delaware's Donation to Hazards Research

Conferences and Training

Contracts and Grants

Recent Publications

Electronic Stuff

Who We Are

What's Happened in Floodplain Management Since the '93 Mississippi Flood?

--an invited comment

Few people in the natural hazards business can forget the images of the Great Mississippi Flood of 1993. Week after week, TV screens were filled with scenes of millions of acres underwater and thousands of people forced from their homes by raging waters. The 1993 Mississippi flood was followed by another Mississippi flood in 1995, major floods in California in 1995 and again in 1997 and 1998, floods in Georgia and Alabama in 1994, in Texas in 1995, and in the eastern U.S. in 1996--each bringing destruction to both urban and rural areas. In the spring of 1997, national attention was focused on the Red River of the North and flooded office buildings burning in Grand Forks, North Dakota. Did we learn anything from the 1993 Mississippi Flood and the floods that have occurred since? Are we doing anything to reduce our floodplain vulnerability? What happened to the 1994 Sharing the Challenge report by the Interagency Task Force to Coordinate the Federal Flood Recovery Effort submitted to the White House (see the Observer, Vol. XVIII, No. 6, p. 10)?

Sharing the Challenge

As former members of that task force, we are frequently asked how many of the recommendations of the report have been carried out and what remains to be done. When asked about the report, rather than provide an enumeration of actions on report recommendations, we have preferred to examine the big picture and assess the overall progress in floodplain management. As readers may recall, Sharing the Challenge emphasized the need for responsibility and accountability for floodplain management to be shared among federal, state, and local governments and with the citizens of the nation. It also called for avoidance of unwise use of the floodplain, minimization of vulnerability when floodplains must be used, and mitigation of damage when floods do occur. All of those actions were to be focused on concurrent reduction of the flood vulnerability of the nation and the protection and enhancement of the natural resources and functions of floodplains.

Progress Made

In our view, in the years since the 1993 flood there has been positive action on many fronts. Education is improving. There is clearly greater national awareness of the hazards of flooding. Major flood specials have appeared on television and other news media. James Lee Witt, director of the Federal Emergency Management Agency, appears in television commercials reminding people of the flood risk and encouraging them to buy insurance.

Many state legislatures and executive agencies have examined their flood management policies and moved toward tighter controls. Federal and state governments have relocated over 12,000 families from Midwest floodplain locations (reaching a total of over 25,000 for the nation as a whole). State and federal agencies have acquired interest in over 250,000 acres of flood-prone land. Who, in 1992, would have thought that such relocations and land acquisition would have been possible?

In October 1994, Congress passed and the president signed a flood insurance reform act that addressed many of the issues raised by the 1993 flood. Flood insurance must now be purchased at least 30 days before the flood event (compared to five days prior to a flood in 1993). Pressure has increased on lenders to ensure that at-risk home buyers and owners purchase flood insurance. Insurance coverage is now available to help pay the costs of elevating or relocating substantially damaged buildings. Parallel changes for agriculture were embodied in the Crop Insurance Reform Act of 1994.

In its assessment of approaches to solving the flood problems of the upper Mississippi, a 1995 Corps of Engineers report validated the view that while structural flood control measures are an important part of an overall floodplain management program, they have limitations. The report noted that floodplains are best managed through a combination of structural and nonstructural measures that fully recognize the inherent risks of occupying flood-hazard areas.

In an effort to increase state and local sharing of the costs of flood-damage reduction efforts, in his 1995 budget submission to Congress, President Clinton proposed reductions in federal support of flood control construction activities. Unfortunately, Congress largely ignored his recommendations. The U.S. Environmental Protection Agency in-creased its support of basin management activities, and the 1996 Farm Bill included provisions to increase conservation activities in floodplains and riverine environments.

The federal government also has reformed a number of programs to remove subsidies and increase incentives to minimize vulnerabilities to flooding. Speaking for congressional committees examining the high costs of disaster recovery, Senator Ted Stevens and the late U.S. Representative Bill Emerson proposed that those at risk should take greater responsibility for avoiding hazards and sharing the costs of post-disaster recovery.

The 1996 Water Resources Development Act extended Corps authorities for nonstructural work under Public Law 84-99 and established requirements for floodplain management as part of Corps projects. The work of the Scientific Assessment and Strategy Team (SAST) initiated during the 1994 study has continued with the publication of several supplements to its initial report. It has also served as a model for other post-flood analyses and has clearly demonstrated the importance of integrating floodplain management into river basin planning.

Lessons from Recent Events

The California floods of 1995 sparked major hearings in that state's legislature over the approaches being taken to reduce the impact of frequent floods. Disastrous floods in 1997 on the Sacramento and San Joaquin Rivers and their tributaries reinvigorated the legislature, and additional hearings were held to develop legislation that would provide a balance between development and natural resource preservation in an environment of comprehensive flood management. In response to a charge from the state governor, a task force of state officials reported to him that a more holistic approach was required to deal with the continuing threat of floods and recommended a comprehensive water management planning effort for the region. Proposing both structural and nonstructural approaches, the task force report included recommendations that would require all those living behind levees providing less than 200-year protection to buy flood insurance and would increase the level of protection for urban areas above the federal flood insurance program standard of 100-year occurrences. These proposals would not have seen the light of day five years ago.

Following the 1997 flood on the Red River of the North, the International Joint Commission, a bi-national organization created by the U.S.-Canada Boundary Waters Treaty of 1909, established a task force to examine and report on the causes and effects of that flood and to identify measures that could be used to reduce future damage. The task force provided an interim report in December 1997 and indicated that a significant risk of flooding remains; that steps must be taken to improve, clarify, and coordinate flood policies throughout the basin and enforce a balanced approach to floodplain management; that major improvements must be made in the use of technology to improve forecasting, measurement, and modeling; and that environmental concerns need to be addressed.

The Next Steps

In December 1997, the Western Governors' Association, representing 18 western states, adopted an Action Plan for Reducing Flood Risk in the West (see the Observer, Vol. XXII, No. 3, p. 9). This plan generally endorsed the recommendations of Sharing the Challenge and specifically called on state and local jurisdictions to refrain from putting people and property at risk by avoiding development in the floodplain; move those at risk from the floodplain, when appropriate; share the risk among all levels of government and among flood-affected individuals; and treat the floodplain as part of a physical and biological system within the larger context of its watershed.

But floods will continue to occur and, unless more is done, so will damage. What still needs to be accomplished? We need a floodplain management act or an executive order that clearly lays out the floodplain management responsibilities at federal, state, and local levels. In the Mississippi River Basin the need for a comprehensive plan still remains, and until it is developed, the region will still be dealing with a loose amalgam of federal, state, and local flood damage reduction efforts. (At a March 1998 federal-state Mississippi River Summit, attendees committed to the development of this comprehensive plan and began formation of a task force to do so.) Complaints following the Red River flooding still point out that many people who could have purchased flood insurance and did not end up as well off as those that did. We continue to support payment for repetitive losses. Work has started on examining the ground rules for flood control projects to eliminate any structural bias, but the work is slow. We need to continue to exploit technology and all it offers us. Coordination of floodplain activities at the national level remains spotty.

In the aggregate, we believe we have made a great deal of progress. Still, we are all impatient and want action on all fronts now. However, when one considers that more and more people and organizations are endorsing a balanced structural/nonstructural approach and even avoidance of the unnecessary use of the floodplain, one must recognize clear progress. We all need to keep moving forward.

Gerry Galloway, National Defense University
Shannon Cunniff, Department of the Interior
John Kelmelis, U.S. Geological Survey
Mike Robinson, Federal Emergency Management Agency
Harry Shoudy, U.S. Army Corps of Engineers

Civil Engineers Warn of Dam Failure Risk

Many of America's dams have exceeded their intended lifespan, are in critical need of repair, and pose a serious safety risk, according to the American Society of Civil Engineers (ASCE). On March 6, 1998, the ASCE released its 1998 Report Card for America's Infrastructure (free via the Internet), which examines policy issues related to civil engineering structures in the U.S., including roads and bridges, mass transit, schools, dams, and waste management systems. The report notes that an alarming number of dams across the country are showing signs of age and lack proper maintenance. Downstream development is increasing. Most older dams were built without adequate spillways to release water in heavy rains, which causes water to run over the top. Inadequate spillway capacities are the most common deficiency and a major cause of dam failures. Dam safety officials estimate that thousands of dams are at risk of failing or are disasters waiting to happen.

The ASCE also points out that, in the past 10 years, more than 200 dam failures have occurred. Approximately 9,200 regulated dams nationwide are categorized as high-hazard, that is, their failure will likely cause significant loss of life and property. Thirty-five percent of these dams have not been inspected since 1990 or earlier. In addition, the owners or managers of a majority of these dams do not have emergency action plans in place for timely downstream warning and evacuation.

The ASCE estimates the average cost of repairing one unsafe dam as approximately $500,000, meaning that it would cost $1 billion to rehabilitate all unsafe dams nationwide. They recommend improving the ability of states to regulate dam safety, establishing state revolving funds for dam rehabilitation and repair, and requiring states to contribute 20% in matching funds.

The complete text of the report and related references are available from the ASCE Web site: http://www.asce.org/govnpub/start.html.

Do Natural Disasters Lead to the Deliberate Ending of Human Life?

On February 5, 1998, the New England Journal of Medicine (Vol. 338, No. 6, pp. 373-378) presented the paper "Suicide After Natural Disasters," by Etienne G. Krug, Marcie-Jo Kresnow, John P. Peddicord, Linda L. Dahlberg, Kenneth E. Powell, Alex E. Crosby, and Joseph L. Annest. The authors conducted an epidemiological study to determine whether natural disasters affect suicide rates and conclude that suicide rates increased in the four years after floods by 13.8%, in the two years after hurricanes by 31.0%, and in the first year after earthquakes by 62.9%. The authors computed rates for the entire United States and found that they were stable. However, the increases in suicide rates following disasters were found for both sexes and for all age groups. Interestingly, suicides rates did not change significantly following tornadoes or severe storms. The authors conclude that this increase in suicide rates following severe earthquakes, floods, and hurricanes confirms the need for mental health support after severe disasters.

Below, David Alexander, professor of Geography, University of Massachusetts at Amherst, comments on the research. His remarks are followed by a response from the study's authors.

Point: Does a Link Really Exist?

Considerable interest has been generated by a recent article in the New England Journal of Medicine that suggests there is a causal relationship between natural disasters and suicide rates. The latter are supposed to increase as a result of post-traumatic stress disorder and depression in the wake of catastrophe. I would urge caution in accepting these findings, for they may be hard to verify in both statistical and causal terms. The same is true of the impact of natural disasters on murder rates.

My interest in the field was first stimulated by an attempted murder that I witnessed three days after a major earthquake disaster. It occurred late one night in the eerie glow of a sodium street lamp. One man, who was evidently profoundly disturbed by the earthquake, had got his hands around the neck of another and was attempting to throttle him. His victim lost consciousness and was within seconds of losing his life when other survivors realized what was happening, and 20 or 30 of them jumped onto the attacker and immobilized him. The reader may wonder what I did at this time. I regret that I stood rooted to the spot by fear and amazement. Unfortunately, academic researchers are seldom men [or women] of action! The episode, and the disaster that preceded it, left me profoundly depressed, but instead of taking my own life I decided to conduct some research into the subject.

I looked at the rates of both attempted and successful murders and suicides for a selection of countries and regions that had been affected by natural disasters during the periods covered by the data. After weeks of poring over statistical bulletins I gave up, as I could find no demonstrable link. In the first place, both murder and suicide are normally highly seasonal phenomena. Typically, in western, economically developed countries suicides peak in spring, with a secondary peak in late summer and a trough at Christmas. These fluctuations are slightly more pronounced for the 20- to 30- year-old age-group and for males than females. Evidently, in springtime, a young man's fancy turns, not so much to love, as to doing away with himself! One can speculate for hours on why such trends exist, but the tables of data are mute about the causes. Though murders and suicides after disaster tend to be highly publicized events, they are usually too few to make a dent in the statistics and the additional numbers are easily subsumed into the gigantic fluctuations that would have occurred regardless of natural disaster. Thus, 43,000 people were injured and 300,000 left homeless by the Kobe earthquake of January 1995, but Japanese researchers reported only eight confirmed cases of suicide among the survivors.

Causality is a thornier problem than statistical validity. We may hypothesize that disasters provoke drastic psychological reactions that can lead to the deliberate ending of human life. But, equally, after natural catastrophe there may be more solidarity, social participation, policing, and mental health counseling, all of which mitigate against the drastic solution. In any case, who is to say that the suicide victim or murderer would not have done the deed in the absence of disaster?

As with the victims of heart attack (acute myocardial infarction), there is an apparent causal link with disaster, but the figures are generally too small in relation to the usual trends for one to be very definite about it. Moreover, how long after disaster can a suicide or murder take place in order to be linked with it? On the one hand, traumatic events can lead to the taking of life many years later (for example, Primo Levi committed suicide more than 40 years after he was released from Auschwitz, and apparently as a delayed reaction to his incarceration there). But on the other hand, the longer the time that elapses, the more likely it is that factors unconnected with the disaster will intervene to sway the balance.

In short, it is all part of the mystery of the human condition.

David Alexander, Professor of Geography, University of Massachusetts at Amherst

The author is currently on leave from the University of Massachusetts and can be contacted at Borgo Sarchiani 19, 50026 San Casciano Val di Pesa, Firenze, Italy; tel: +39 [0]55 822-9423; fax: (c/o Rossella Rossi-Alexander): +39 [0]55 587087; e-mail: alexande@dada.it.

Counterpoint: Suicide Rates Increase after Natural Disasters

We thank Dr. Alexander for his interest in our study of suicide after natural disasters.(1) We analyzed suicide rates in 377 U.S. counties affected by a single natural disaster between 1975 and 1993. Comparisons were made between pre- and post-disaster suicide rates for the disaster counties and for the entire United States. We found that suicide rates increased in the four years after floods by 13.8% (p<0.001), in the two years after hurricanes by 31.0% (p<0.001), and in the first year after earthquakes by 62.9% (p<0.001). Rates computed in a similar manner for the entire United States increased by less than 1.3%. We concluded from our study that suicide rates increase after severe earthquakes, floods, and hurricanes. Our findings highlight the importance of suicide as a public health outcome of natural disasters.

Dr. Alexander urges caution in accepting these findings and states that they are difficult to verify in both statistical and causal terms. While it is true that some of the previous studies on the psychological effects of disasters are methodologically weak and inconclusive,(2) our study utilized rigorous epidemiological and statistical methods. The population included in our study was large--19,453,931 people lived in the 377 counties at the time of the disasters. Baseline information was available, which allowed us to compare suicide rates before the disasters with rates after the disasters. We also compared suicide rates in the affected counties with those of a control group (the entire United States). Previous research has not been able to find a link between suicides and natural disasters, in part, because the focus has been on only one disaster (e.g., Hurricane Andrew) and the population affected by the disaster has been too small to make significant comparisons between pre- and post-disaster suicide rates.(3),(4) Research that has examined countries or regions, such as that conducted by Dr. Alexander, is also not likely to find a link between suicides and natural disasters. The entire population of a country is rarely affected by a disaster; therefore, the effects of the disaster on suicide rates are not likely to be detectable at the country or region level--a point demonstrated in our study (again, rates for the entire U.S. increased by less than 1.3%).

Dr. Alexander raises a number of other issues, including the seasonality of suicide rates; the possibility that factors other than the disaster may have precipitated the suicides; and the possibility that after a natural catastrophe there may be more solidarity, social participation, and mental health counseling. While it is true that suicide rates peak in the spring months, seasonality is not an issue in our study because we used annual, instead of monthly, suicide data. Trends over several years were also taken into account by comparing rates in disaster counties to rates in the United States as a whole.

We do not know the precise reasons why people commit suicide in the aftermath of a natural disaster. More research is clearly needed in this area. The patterns seen in our data do not support the notion that the suicide victims would have done the deed in the absence of the disaster. Finally, counter to what Dr. Alexander has hypothesized, previous research shows a decrease in the feeling of social support and integration after a natural disaster.(5)

Dr. Alexander concludes that, in short, [suicide] is all part of the mystery of the human condition. Here too we have to disagree. Suicide is an important public health problem. It is our duty as public health officials to understand suicidal behavior and to prevent it.

Etienne G. Krug, M.D., M.P.H.
Linda L. Dahlberg, Ph.D.
Centers for Disease Control and Prevention

1. Krug, E.G., Kresnow, M.J., Peddicord, J., Dahlberg, L.L., Powell, K.E., Crosby, A.E., and Annest J.L., Suicide After Natural Disasters, New England Journal of Medicine, 1998: 338: 373-8.

2. Bromet, E., and Dew, M.A.. Review of Psychiatric Epidemiologic Research on Disasters, Epidemiological Review 1995, 17: 113-9.

3. Imamura, K. Mental Health in Japan [letter]. Lancet 346 (8973): 509-510.

4. Lew, E.O., and Wetli C.V. Mortality from Hurricane Andrew, Journal of Forensic Science 41 (3): 449-52.

5. Kaniasty, K., Norris, F., and Murrell, S.A., Received and Perceived Social Support Following Natural Disaster, Journal of Applied Psychology 20: 85-114.

Next page

Return to the index of the Natural Hazards Observer

Return to the Natural Hazards Center's Home Page