be found in the broad coverages, which do not exclude the peril of flood, available for movable property, such as the various forms of marine, inland marine and automobile comprehensive coverage, and certain "floaters").

2. The companies believe that specific flood insurance covering fixed-location properties in areas subject to recurrent floods cannot feasibly be written because of the virtual certainty of loss, its catastrophic nature, and the reluctance or inability of the public to pay the premium charge required to make the insurance self-sustaining.

3. Any insurance program which does not cover areas subject to recurrent floods will not meet the public need.

4. The companies believe that it is impossible to tie in flood coverage with other coverage on fixed-location properties generally because, unlike other natural catastrophes which are unpredictable as to place of occurrence, floods can occur only where water flows or gathers and only those properties which are in the path of the flow or gathering have any need for it; competition would force the sale of coverage ex flood and the buyers would make the adverse selection.

5. There is no way in which the purchase of flood insurance can be made mandatory (even by Government compulsion) consistent with our American concept of free government and competitive selection.

6. The companies believe that the Government would encounter the same obstacles if it undertook a program of specific flood indemnity by means of insurance on a self-sustaining basis.

7. Any Government promise of indemnity on a non-self-sustaining basis is relief under the guise of insurance. In our opinion, a direct program of relief and rehabilitation would be more effective and more equitable, particularly in restoring essential services and providing food and shelter, which are the first forms of necessary relief in the case of a major flood disaster.

8. In our opinion, flood control and prevention (rather than insurance, indemnity, or relief) are of far greater importance to potential flood victims, especially when the many forms of irremediable losses are also taken into consideration, such as death, bodily injury, loss of employment, and loss of income. 9. In view of the magnitude of Government expenditures which are involved in the event of a major flood disaster, it would seem prudent for the Government to avoid fixed advance commitments in order to be in a position to use available funds most expeditiously and to the best advantage when the emergency arises. Mr. EDELSTEIN. Mr. Chairman, could we put into the record this Report on Floods and Flood Damage of 1955 in the Northeastern States?

Senator LEHMAN. Yes, without objection.

(The report above referred to follows:)

REPORT ON FLOODS AND FLOOD DAMAGE OF 1955 IN THE NORTHEASTERN STATES (Prepared for American Insurance Association, December 1955, by Parsons, Brinckerhoff, Hall & Macdonald, engineers, New York, N. Y.)

DEAR MR. HERD: In accordance with our proposal of October 17, 1955, we submit herewith our Report on Floods and Flood Damage of 1955 in the Northeastern States. This report is, in effect, a sequel to the Report on Floods and Flood Damage which we prepared for the Insurance Executives Association in April 1952. The purpose of the present report is to discuss the floods of August and October of this year; to obtain an approximate estimate of the damages caused by these floods in the States of Massachusetts, Connecticut, Rhode Island, New York, Pennsylvania, and New Jersey; and to consider these floods in connection with the general problem of the financial losses resulting therefrom. For your convenience in reviewing the report, its essential features are summarized below.

1955 flood damages. Various preliminary estimates of the amount of damages produced by the floods of August 1955 have been prepared by several governmental agencies, as well as by the American Red Cross. Analysis of these estimates indicates that the total amount of direct damage to physical property (both real estate and personal property) in the 6 States mentioned above was approximately $500 million. This figure may include a small amount of wind damage, but of a very minor nature.

The direct damages from the August 1955 storm may be subdivided as follows:

The distribution of the direct damages by watersheds is estimated as follows:

Percent

Housatonic River (Massachusetts and Connecticut)
Connecticut River (Massachusetts and Connecticut)
Thames River (Massachusetts and Connecticut).
Blackstone River (Massachusetts and Rhode Island)
Charles and Neponsit Rivers (Massachusetts).

38

14

12

6

6

Delaware and Hudson Rivers (New York, Pennsylvania, and New Jersey) –

24

Total___

100

Nearly one-half of the total direct damage in the August floods occurred in Connecticut.

Less than 2 months after the floods of August 1955, the Northeastern States were subjected to the effects of an extratropical cyclone. Fortunately, most of the sections that suffered severe damage in August were not badly hit by the October storm. However, properties adjacent to certain streams particularly along the Long Island shore of Connecticut, were damaged by high tides and water flooding, although they had not been seriously affected in the August storms.

It was characteristic of the August storms that they caused maximum damages on the smaller streams, accompanied with serious loss of life in certain localities. The path of the storm was generally transverse to the main axis of the Delaware, Hudson, and Connecticut Rivers, so that the degree of flooding on these larger rivers was relatively small as compared with maximum floods in the past. Meteorological conditions.-The storms that caused the severe floods in August 1955 were tropical hurricanes that were diverted from their previous normal paths by unusual meteorological conditions. Such storms have been considerably more frequent on the North Atlantic seaboard since 1938 than in the preceding hundred years. Whether this situation is a result of a permanent change in climatic conditions or is only in the nature of a temporary weather cycle is a moot question among meteorologists, and probably will not be definitely answered for many years.

Past experience and theoretical studies indicate that, while the floods of August 1955 were unprecedented in magnitude in the region affected by the storm, it is entirely possible that even greater floods may occur in almost any portion of the area under consideration.

Tropical hurricanes approaching coastal regions often cause serious flooding along the shore because of exceptionally high tides. Under certain conditions such tides may exceed 16 feet at any location between southern New Jersey and Cape Cod.

Comparison with previous floods.-The hurricane storms of August 1955 caused unprecedented depths of precipitation over certain areas, resulting in floods of extraordinary magnitude on many of the smaller rivers, in some cases as great as eight times the maximum previous flood of record.

Climatic changes.-The unprecedented nature of the precipitation and stream discharge during the floods of August 1955 has been interpreted by some people

69096-56-pt. 2- -12

as an indication of a change of climatic conditions in the Northern Hemisphere. Although climatic cycles have occurred on the earth in past ages, they are known to be very slow, extending over periods of thousands of years. Recent apparent changes in climate affecting the paths of hurricanes along the North Atlantic seaboard cover too short a time period to justify the assumption of any long-term major climatic changes in this region. Nevertheless, the possibility of such cyclic variations in climate should not be neglected in estimating flood probabilities, although it would involve serious difficulties in estimating mean annual flood losses in any locality by statistical methods.

Effect on estimates of mean annual flood losses.-A preliminary investigation of the effect of the 1955 floods on estimates of mean annual flood losses indicates that it would be of a minor nature because of the apparent low probability of occurrence of such unprecedented floods. However, if future experience should prove that there has been a definite change in climatic conditions resulting in more frequent hurricane storms in this area, the damage losses of the 1955 floods would indicate an appreciable increase in mean annual flood losses. It is probable that further light on the matters discussed in this report will become available in the near future, as a result of extensive studies now under way by several governmental and private agencies. We will be pleased to analyze such material and expand our report accordingly, if you so desire. Very truly yours,

PARSONS, BRINCKERHOFF, HALL & MACDONALD.
M. N. QUADE.

INTRODUCTION

In April 1952, the firm of Parsons, Brinckerhoff, Hall & Macdonald submitted a report on floods and flood damage to the Insurance Executives Association, the purpose of which was to determine the technical engineering problems that would have to be solved if a practical and effective program of flood insurance in the United States were to be established. The insurance business had never been able to devise a method of providing specific flood-insurance coverage on a basis in conformity with sound insurance principles. Following the disastrous floods of 1951 in Missouri and Kansas, insurance underwriters began a reexamination of their traditional position respecting flood insurance. The report of April 1952 was prepared for use in this connection. An abstract of the 1952 report was published by the American Society of Civil Engineers in August 1954, as Proceedings Separate No. 483, Flood Insurance, by H. Alden Foster.

General interest in the possibilities of flood insurance has been greatly intensified since the disastrous floods in the Northeastern States in August 1955, resulting from hurricanes Connie and Diane. In anticipation of public discussion of this matter, the current study was initiated on October 4, 1955. It was agreed that the investigation would cover the floods and flood damage that occurred in Massachusetts, Connecticut, Rhode Island, New York, Pennsylvania, and New Jersey on August 17-20, 1955, and during the floods of October 1955. The work would consist principally of:

(a) Collecting data on 1955 flood damages to obtain realistic estimates of physical damage classified to the extent possible, in accordance with the general character and use of the property from the viewpoint of an insurance interest; and

(b) Securing information on hydrology and meteorology of the recent floods and the interpretation of such data in general terms with respect to flood damage

Field work by P. B. H. and M. was well under way prior to the severe weekend storms of October 14-17, which naturally resulted in considerable changes in the overall problem of flood damages in the investigated area. The investigation was then extended to include an estimate of damages during the October flood as compared with those experienced in August, and the general effect of such repetitive floods.

Serious property damage and loss of life was caused in the Northeastern States, from Massachusetts to Pennsylvania, by floods resulting from two tropical hurricanes that passed over this region in August 1955. The first storm, christened Connie by the United States Weather Bureau, was of moderate intensity, and passed over North Carolina, The Chesapeake Bay region of Virginia, Maryland, western Pennsylvania, Lake Erie, and finally into Canada, on August 11 to 13. The second storm, identified as Diane, caused much

heavier precipitation, passing over North Carolina, Virginia, Pennsylvania, New Jersey, Long Island, and Cape Cod and then over the Atlantic Ocean, on August 17 to 20, as shown on figure 1.

This study is an analysis of available reports on the damage caused by these two hurricanes. Diane is usually referred to as the one which caused the damage, but as it followed on the heels of Connie the damage actually was the result of both. To differentiate between the damage caused by these storms would be impracticable.

Estimates of damage were available in all the Northeastern States which suffered from the storms; Massachusetts, Rhode Island, Connecticut, New York, New Jersey, and Pennsylvania. Four organizations, mobilized immediately by the President and the governors of the States, prepared such estimates for each of the States:

(1) United States Corps of Engineers.-Cost of cleaning and restoring river channels, property damage to all types of property, and loss of production and employment, with a view toward constructing facilities, under Federal Government sponsorship, to minimize danger in the future.

(2) Federal and State civil defense agencies.-Property damage to all types of property, including farms and agricultural crops.

(3) American Red Cross.-Care of human beings and repair of damage to their dwellings to provide needed assistance for relief and rehabilitation.

(4) State governments.—Under the leadership of flood-disaster committees, damage to all types of properties and care of human beings in order to establish need for relief and rehabilitation and to institute measures for reducing damage in the future.

Except for the Corps of Engineers, these organizations did not make their own appraisals of all types of damage but accepted the figures developed by other groups particularly qualified to estimate damages to certain types of structures. For example, utility commissions of the States were relied on for public-utility data, highway departments for figures on roads and highway bridges, and State and local governments for estimates on public buildings and schools as well as damage to private residences.

The work of evaluating the damage was commenced immediately after the floods and the earliest estimates were generally the highest, resulting partly from the shock of the disasters and the inability to obtain a clear view of the damage because so much was entirely covered by water. Subsequently, detailed evaluations were made at the State and local levels. These estimates revised the original data, generally resulting in a lowering of the figures, and were used in the present study.

The several estimates are consistent as to the classification of the categories of property, but the results vary because of differing approaches to the question of the value to assign to the properties-whether depreciated value, replacement in kind, or replacement on the basis of modern standards.

The United States Corps of Engineers estimates are probably the most standardized and are the ones which can be used for a comparison with other disasters of a like nature. The corps has made such estimates for many years for use in obtaining congressional appropriations to institute corrective measures. In their final form, which will not be available for some months from now, they will be presented in detail-by classes of property, and covering all the types of damage. At present, however, these evaluations have been prepared only as overall figures by watersheds and do not segregate losses by classes of property and by types of damage. They represent the overall economic effects of the floods, including loss of production and wages. Present estimates are not final and are subject to revisions as refinements are made in preparing the final figures.

The Federal Civil Defense Administration (FCDA), in cooperation with the several State civil-defense organizations, has also prepared estimates for each of the States. These do not include the values of loss of production and wages, nor do they include the cost of restoring and improving river channels. In their preparation, the values determined by State agencies, especially for certain categories such as roads and highway bridges, public utilities and public facilities, were used directly.

FCDA has adopted the following categories in its breakdowns of estimates: (1) Industrial.-Manufacturing establishments of all types, metals, textiles, finished products, etc.

(2) Commercial.-Service establishments, wholesale and retail distribution, automotive dealers and other occupants of business buildings.