peacetime employment offers a relatively favorable prospect of absorbing the surplus. The demobilization rates for these States-Alabama, Colorado, Delaware, Illinois, Louisiana, Maine, Massachusetts, Missouri, New Hampshire, New York, Oklahoma, Oregon, Tennessee, Texas, Virginia, West Virginia, and Wisconsin-range between 24.7 and 34.3 percent.

It is significant that even in the 20 remaining States (including the District of Columbia) with ratios under 24.7 percent, the demobilization problem will in no case be less than 19 percent of pre-war employment. These States comprise a large part of the South and Southwest, as well as of the northern plain and northwest mountain areas. In these regions the development of war industry has been light, and the chief problem will be the placing of returned servicemen.

Three general comments are relevant. First, in no State is the problem negligible. A sudden expansion in the labor supply of only one-fifth, relative to normal labor requirements, is enough to cause serious disturbance unless foreseen and prepared for. Second, the gravity of the problem for each State will depend chiefly on the expansion of the industrial base during the war, since the demobilized servicemen will constitute a fairly constant proportion of the population and of the pre-war employment in each State. Third, even where the problem is not acute over the State as a whole, there may be violent local dislocations for which preparation will be necessary.

Number of Demobilized Persons, by States

In the accompanying table are given the estimates upon the basis of which the map was constructed. The national total of 8,500,000 men to be demobilized from the armed services is derived by subtracting the assumed 2,500,000 post-war armed force from the published estimates of a peak armed force of 11,000,000. The State totals were derived by dividing the 8,500,000 national total according to 1940 State population ratios, adjusted for sex and age distribution. It was assumed that servicemen will return to the States from which they were inducted.

The national total of slightly over 6,000,000 demobilized industrial workers was derived by estimating for each of 14 manufacturing industries (1) a rate of contraction of munitions production, and (2) the most rapid possible rate of expansion of civilian production, as explained above; and totaling the differences between December 1944 employment and lowest subsequent employment for each industry. The number of demobilized industrial workers in each State was derived by computing ratios of State to total employment in December 1944 for separate manufacturing industries, from forecasts of the Bureau of Labor Statistics and the Bureau of Employment Security, or from census data; applying these ratios to the indicated total demobilization for each industry; and totaling the results for each State. This assumes that demobilized industrial workers will, in the first instance, constitute a problem for the States where they held war jobs, either by seeking new jobs, drawing unemployment compensation or relief, requiring retraining services, or in other ways.

If a count of war jobs is made on the basis, not of war industry, but of war output not required in the post-war period-ships and planes in excess of peacetime needs, and ordnance items such as guns, ammunition, and tanks a smaller national total (4,600,000) of

industrial demobilization is obtained. This, however, would understate the industrial demobilization problem for the Nation as a whole and for all but 11 States, since in most cases the number of workers permanently displaced by the contraction of specialized munitions production will be less than the number temporarily disemployed by the reconversion of war industry in the broader sense.

It should be noted that the figures on industrial demobilization minimize the actual reemployment problem in the sense that they do not include job shifts within a given industry which produce no net contraction of employment. It should also be observed that all demobilization figures are totals for the reconversion period as a whole and do not measure unemployment at any particular time.

Military and Industrial Demobilization, Related to Pre-War Employment, by States

* Differences between United States totals and sums of State components are due to rounding of figures.

EARLY in 1943 the U. S. Maritime Commission and the U. S. Navy launched a safety and industrial health program in shipyards. Included in this program were all yards building boats for these two agencies, accounting for practically all the shipyards in the country, many of them Federally owned. Minimum-requirement rules were adopted after a conference in December 1942 in which several hundred safety men and physicians from nearly all of the yards participated and assisted in the drafting of these rules.

One of the requirements is that each yard must report monthly to the U. S. Bureau of Labor Statistics the data necessary for the computation of injury frequency rates. A monthly summary report was developed and adopted for this purpose. The tabulations to be based on these reports were to be used to indicate (1) the general injury experience in the shipbuilding industry, and (2) the yards with bad accident records. The safety consultants of the Maritime Commission and the Navy could then concentrate on yards in the second group, in order to reduce or eliminate the injury-producing hazards. Another requirement was that the monthly summary be accompanied by individual reports of disabling injuries which occurred during the month. The primary purpose of this report (the supervisor's report of accident or occupational disease) was to reveal the unsafe conditions and unsafe practices which caused these accidents. and to indicate what the management had done to prevent recurrences. The Bureau's analysis of these reports is intended to show the types of conditions or unsafe practices which prevail in individual yards, thus revealing to the Federal safety man the accident factors with which he should concern himself.

This program is a clear-cut instance of the use of accident statistics for the guidance of safety men. Each yard is required to have its own safety organization. The Federal men act primarily as consultants, visiting yards which appear to require their help. This service is facilitated by monthly tabulations in which attention is called to yards with high frequency rates, particularly if these rates remain high for successive months. Monthly tabulations are also sent to each yard, with each yard identified by a code number. As every yard has been advised of its number, it can readily compare its experience with that of other yards, although unable to identify them. The attempt to direct safety work in an entire industry, employing well over a million workers, through the comprehensive use of accident statistics has rarely been tried before. Although the use of statistics may be of limited effectiveness in small establishments, it has decided advantages in a large establishment and still more when applied to an entire industry. Under the present reporting program, the safety engineers of the Maritime Commission and the U.S. Navy can easily direct their efforts toward the establishment of proper safety organizations in shipyards with bad accident records. They can assist these new organizations, or existing older organizations, in planning for proper remedial measures to eliminate accidentproducing working conditions and to teach workers and supervisors 1 Prepared in the Bureau's Division of Industrial Injuries, by Max D. Kossoris.

safe ways of working. That many yards need such assistance is amply illustrated by the consistently high frequency rates in a fairly high proportion of shipyards. Moreover, the individual injury reports requiring an analysis of each accident, serve a double purpose: (1) They require the safety engineer to make an analysis of the factors which produced each accident, and (2) they indicate whether the safety engineer understands his job or whether he needs assistance. On the basis of the 3-month experience of the reporting group, it appears that disabling industrial injuries in privately operated shipyards involve a time loss aggregating 1,368,000 working days annually, in addition to that involved in the fatal cases. The economic effects

of over 14 million man-days a year lost through work injuries at a time when ships are so urgently needed hardly require emphasis. In terms of a 300-day year, this time loss is sufficient to offset the fulltime efforts of nearly 5,000 workers.

Accident Experience

In table 1 are summarized the injury-frequency rates for the first quarter of 1943. Rates are shown separately for each region. TABLE 1.—Industrial Injury Frequency Rates in Shipyards, First Quarter of 1943

1 The frequency rate is the average number of disabling work injuries per million employee-hours worked.

? The number of reporting yards varied from 130 to 134.

3 The number of reporting yards varied from 37 to 43. The number of reporting yards varied from 90 to 93. Includes 9 yards.

Includes Pearl Harbor.

For all reporting privately operated yards there were 32 disabling injuries for every million employee-hours worked during the first 3 months of 1943. There was practically no difference between the experience of the total group building cargo vessels for the U. S. Maritime Commission and that of the total group building various types of warships for the U. S. Navy. There were, however, decided differences between these two groups regionally.

While there was a relatively narrow range between the regions in the U. S. Maritime Commission group-from a low of 26.5 in the Great Lakes region to a high of 37.0 for the Gulf Coast-the range within the U. S. Navy group was from 16.6 in the Great Lakes region to 60.4 in the South Atlantic region. The comparison, however, was not so one-sided as these figures indicate. In the Gulf Coast and Great Lakes regions the rates of the yards building for the U. S. Navy were considerably lower than those building for the U. S.

Maritime Commission, and on the Pacific Coast the difference between the two types of yards was small. In the Atlantic Coast regions, however, especially in the South Atlantic, the experience of the yards building for the Navy was decidedly the worse; the rate of 60.4 in the South Atlantic yards controlled by the Navy was nearly twice that of 33.6 for the U. S. Maritime Commission yards.

In sharp contrast to the rates for both of these groups were those of the 9 shipyards owned and operated by the U. S. Navy. The general average rate for the quarter was 16.7, or just about half that of each of the other two groups. The rate for the 3 yards on the Pacific Coast was at an even lower level, being only about a quarter of those of the other two groups.

A good share of the reason for this marked difference lies in the fact that the naval shipyards have for years had a systematic accidentprevention program. Many of the privately operated yards, on the other hand, have instituted safety programs only recently and occasionally the newly appointed safety men have had very little experience in scientific accident prevention. Many of the yards expanded tremendously in a short period of time, without adequate attention to accident prevention, whether in the form of adequate instruction and supervision or of safe working conditions.

The differences pointed out between the regions shrink into insignificance when examined in the light of the experiences of individual yards. In a considerable proportion of these, the frequency rates were as good as or lower than those for the group as a whole, industry-wide or regionally. In others, however, the rates were consistently high. In the U. S. Maritime Commission group, a yard on the Gulf Coast averaged 128 disabling injuries per million hours worked during the quarter; the monthly rates for January, February, and March were 123, 144, and 120. A yard on the Pacific Coast averaged 94 for the quarter, and 97, 42, and 126, respectively, for the 3 months. A Great Lakes yard averaged 108, with monthly rates of 160, 107, and 73.

Among the yards with Navy Department contracts, a North Atlantic yard averaged 137 for the quarter, and 154, 101, and 157 for the 3 months. A South Atlantic yard averaged 116, and had monthly frequency rates of 126, 123, and 94. A Gulf Coast yard ran rates of 175, 156, and 156 for the 3 months, averaging 162. The worst yard of all, on the Pacific Coast, had rates of 548, 257, and 315, ending the quarter with an average of 364 disabling injuries per million employee-hours.

Frequency rates which remain persistently high for 3 successive months are indicative of deplorable working conditions and unsafe practices. If the experience of the reporting group is fairly representative of the industry as a whole, the actual time lost because of disabling industrial injuries in privately operated yards may be estimated as about 114,000 days per month, or 1,368,000 days per year. This estimate makes no allowance for deaths and the economic effects of permanent impairments, which would increase the total time loss considerably. It represents a minimum.

Accident Factors

ACCIDENT TYPE AND NATURE OF INJURY

Over one-third of the more than 8,000 disabling injuries reported by about 130 yards during the first 3 months of 1943 involved workers