seemed advisable to consider, by comparison, the years 1948, 1951, and 1956.

You will see that the revenue-traffic units for those 3 years were very nearly identical. They fluctuated only about 2 percent, around 710 billion traffic units. In 1948, with that traffic level, there were 1,326,000 railway employees. In 1951, there were 1,275,000, and in 1956, 1,042,000.

If you will look at the figures for the years 1949 and 1954, you will see there, traffic was low and again very nearly the same for those 2 years, with traffic almost the same just slightly up-the number of employees of the railways dropped from 1,192,000 to 1,064,000.

These figures have to be considered in the light of the fact that the average workweek for the railway workers for 80 percent of the railways has declined over that period from 48 to 40 hours.

I should say the basic workweek shows that even though the number of hours per employee has dropped by 8 per week, the number of employees has still declined over 20 percent, with traffic at almost exactly the same level, that is, 1948.

The output per employee, man-hour, then, has increased from 1948 by 40 percent. That is to say, there is 40 percent more railway traffic produced per man-hour in 1956 than there was in 1948.

While this total employment was declining, there had been fluctuations in the number of persons drawing unemployment compensation. The figures for that are shown in table A-2.

The table there shows the total qualified employees; the total number of beneficiaries. You will see again, taking those 3 years together, which I draw your attention to, that in the benefits years 1947-48, there were 2,270,000 qualified employees which had dropped to 1,481,000 in the latest benefit year. And, the number of beneficiaries has also declined, but the percent of beneficiaries to qualified employees is slightly higher in the latter year than it was in either of the 2 earlier

years.

The unemployment of these workers has lengthened in time. There is more unemployment per unemployed man today than there was in the earlier years, and you will see on page A-4 that the average days of unemployment increasing over these particular years have been accompanied by increases in the proportion of benefits or of accounts exhausted by the beneficiaries.

The average days of unemployment in 1947-48 were 74. They were the same in 1950 and 1951, but they had risen to 90 in 1955 and 1956.

The accounts exhausted in 1955-56 were somewhat lower than in 1947-48, but the number of beneficiaries have climbed by almost one-third.

So that the accounts exhausted per 1,000 beneficiaries had risen very sharply over that period.

Reduced employment on the railways then has been accompanied by a great increase in output per employee of per man-hour.

The drop in railway employment between 1948 and 1956 is thus due to technological factors and may be properly called technological unemployment. This is not the total, of course, of railway unemployment. There are still very severe seasonal problems, and while the general subject of this statement is the technological unemployment,

I should like to call attention to some of the characteristics of this seasonal unemployment.

It is a fact that the irregularity of railway employment has risen. with the general decline in overall unemployment. The Railroad Retirement Board published figures showing the amount of unemployment on a monthly basis. It is instructive, I think, to look at the total for the skilled group of the shop crafts for the years 1953 and

1955.

In 1953, 83 percent of the shop craftsmen worked in each of the 12 months. The others worked something less than 12 months of the calendar year.

In 1955, according to figures which have just become available only 77.5 percent of these shopmen worked in each of the 12 months of the year.

Now, that figure for skilled workers is paralleled very closely by the data for their helpers-57.5 percent of the helpers worked in each of the 12 months in 1953 and only 52.5 percent worked in each of the 12 months in 1955.

So as I say, I want to discuss the problem of technological unemployment primarily in this statement. I think it is necessary to keep in mind that these railway employees also suffer from seasonal unemployment which seems to be definitely on the increase.

The technological unemployment to which I have referred is distinguished from this seasonal and irregular unemployment arising from changes in the volume of business or in the personnel policies of the carriers.

Technological change reduces the number of employees required in any industry or any given volume of production.

The employee who is furloughed because of seasonal fluctuations can expect to be returned to his job after a brief period of unemployment. The employee displaced for technological changes can be returned only if the total volume of business increases substantially or if the resignations, retirements or deaths of other employees create enough vacancies to counterbalance the technological reductions.

One other important feature of technological displacement in especially the railroad industry is that it often results from drastic changes at particular points or in particular processes, rather than being at an even general figure, such as comes from seasonal changes. The technological change may displace almost all of the workers in one occupation or at one point, while only slightly affecting those in other occupations or at other points.

In seasonal changes, in the railway industry, the seniority rules and practices result in concentration unemployment on the newer employees; but the drastic reductions brought about by technological change reach high up into the service and age levels of the particular groups most drastically affected.

Technological change of this kind is particularly serious in the railway industry. A very large proportion of railway workers have developed skills that are almost valueless in any other industry; displacement from the railways may require a highly skilled worker to start at the bottom in some other occupation or industry. Add to this the fact that a large proportion of these railway men are employed at railway points where virtually no other industry exists, and the problem becomes doubly acute. Skilled railway workers of long service and

no longer young may be forced to leave towns where they have bought homes and sunk deep roots, to seek employment at minimum wages in new locations.

This is much less true in other industries where typical industries are located in or near industrial centers, where other industries or other establishments perhaps in the same industry may be expanding as one contracts; where there is a variety of industries and the skills of the factory worker are easily adaptable to other factories in the same area. The expansion is characteristic of American industry. So that one industry either shrinks or through technological changes diminishes its work staff, others are expanding in the same area, and are able to use those skilled workers.

Neither of those is true typically of the railway industry.

Employees in the upper brackets of both age and service have been especially hard hit by the technological changes in the railway indus

try in recent years. If you will look at table A-5, you will see a meas

ure of the extent to which these circumstances have been especially hard on older employees. The table on that page shows the duration of unemployment by age groups.

I have chosen the 3 years I referred to earlier as having virtually equal traiffic levels so that there would be no complication from the standpoint of volume of traffic.

You will see, for example, under the heading of "Days of unemployment" 1947-48, that employees under 20 averaged 67 days of unemployment. That is to say those who were under 20 years of age. And if you will run down that column you will see the age 40-44, who were unemployed an average of 75 days, and all of the age group above that were more than 75 days of unemployment for those who were unemployed.

At the bottom of that column you will see the total figure of 75, the average for the whole group, and you will notice those upper-age brackets all had more unemployment than the average.

The second column, 1950-51-the same fact appears. The average is 73 days. Every group of 45 years or older had more than 73 days of unemployment. Those were the ones who were unemployed.

In 1955-56 two very significant changes. You will see the average unemployed, of the men unemployed has increased from 73 days to 90 days.

But if you will look at the individual group you will see that that increase of 15 days was greater than for many of the younger employees. The older employees again had more unemployment than the average, and the change from 1947-48 to 1955-56 is greater among the older employees than among the younger employees.

That means they are not only unemployed, left out of work longer, but that the difference between them and the younger employees is widening in the latter years over the earlier years; that these workers are not just a few, but a very substantial part of the railroad industry, is shown by the figures on page A-6. You will see there that of the employed in the railroad industry in the last benefit year, the cumulative percentage distribution on the right side of the table shows that 38.9 percent of those who drew unemployment benefits in the last year were in the age brackets of 45 years and above.

So that we have a very large number of employees unemployed a longer period of time; longer than the younger employees, and their

unemployment period increasing more rapidly than the unemployment period of the younger employees.

The high average days of unemployment for these older employees results, of course, from a blend of longer and shorter periods. Even among senior workers, many were out of work for shorter periods, having been furloughed for various temporary causes. But a large and increasing percentage of these older workers were permanently displaced. One measure of that displacement is in the number who were out of work long enough to use up all of the unemployment benefits to which they were entitled.

On page A-7, you will see a comparison of this matter of benefit exhaustions, in terms of age distributions.

Here again the age distribution is shown on the left side. The number of benefit exhaustions in each of the 3 years, the three benefit years are shown, and then the percentage of exhaustions to beneficiaries.

The percentage in the three right-hand columns-if you will, I would like to run down that-you will see that those who exhausted the benefits in the younger age groups, in 1947-48 were in every instance smaller than the percentage of exhausted benefits in the upper age groups. That was true in 1950-51. But the situation of each of those years had been aggravated by developments in 1955 and 1956.

There is not only the greater proportion of older employees who completely exhausted their unemployment benefits but that percentage is rising, rising more rapidly than the percentage of younger men exhausting their benefits.

In making this analysis, as I arrived at this point it naturally occurred to me to see that proportion of the employees of the upperage brackets were also long-service men. It would not be impossible, theoretically, at least, that the men of the upper-age brackets really had shorter service, and for that reason more unemployment.

If you will look at page A-9, you will see a distribution of the railroad employees in the year 1954, which is the latest for which these figures were available according to their period of service in the industry.

In 1954, there were 1,682,000 employees in the railroad service. You will see these are distributed according to their age in the lefthand column. There were in age groups from 45 and up, 801,000 employees in that year. Of those, 671,000 had more than 10 years of service in the industry.

There were 772,000 over 40 years of age with more than 10 years of service in the industry.

I would invite your attention to the figures on the right-hand side of the page under the column "Employees with less than 10 years of service." You will see there the heading "New entrants in 1954." The railways hired that year 87,700 workers without previous railroad service. Of those you will see 60,000, 70 percent, were under 30 years of age. Another 9,400 were under 35 years of age. So that percent, of the total of new entrants were under 55

70,000, or 81 years of age.

I have other information on these new entrants but I wish to call that particular point to the attention of the committee members here to show that the older employees are also the long service employees

and, on the other hand, the new employees are very young employees, by comparison with the average for the industry.

Mr. DINGELL. Mr. Chairman, may I ask a question at this point? Mr. MACK. Mr. Dingell.

Mr. DINGELL. Will you tell us, sir, if you can, what change would have taken place in this table since 1954 when it was compiled?

Mr. OLIVER. Well, from other figures that are available it would seem to me that the number of employees in the older service brackets would have increased considerably; but the number of younger men would not have diminished.

Mr. DINGELL. In other words, those are relative figures.

Mr. OLIVER. Yes, relative figures. I do not believe that the situation overall would have changed very greatly.

Mr. DINGELL. What would be the situation with respect to those 2 new hirings?

Mr. OLIVER. Probably the same situation. The same proportion would be in the very young age groups.

I have some additional figures on that, Mr. Dingell, which I want to put before the committee in the next few minutes, and perhaps they will answer the questions you have in mind.

Mr. DINGELL. I just wanted to know whether the same situation still exists and if the table is reliable.

Mr. OLIVER. I think that is very accurate.

Mr. DINGELL. Thank you very much.

Mr. OLIVER. Of the railway workers separated, not employed, in 1953, those still living and not retired, 11,000 had more than 10 years of service and another 26,000 had more than 5 years of service. 20,000 of this 37,000 were more than 40 years old, and 16,000 were more than 45 years old.

Taken together, these unemployment figures establish a very grave problem of the displaced older railway worker in the industry for which the current unemployment compensation provisions is clearly inadequate.

The problem has been aggravated greatly by the employment policies of the railroads; part of the remedy lies within the control of the carriers themselves.

POSSIBILITY OF IMPROVED PLACEMENT

The net reduction in total railway employment, from year to year, is made up of a great many separations, due to force reductions, resignations, deaths, and retirements, which are counterbalanced in large part by new hirings.

If you will look at Table A-3 you will see that comparison in some detail. These figures were compiled by the Railroad Retirement Board. They are perhaps a little confusing-at any rate they were to me at first.

The column headed "Employees in service during year" covers every employee who drew pay in accordance with their definition here.

The "employees separated during year" means those who were eniployed in the calendar year shown but were not employed at all in the following year.