3. H. R. 4408 would apply to a third major type of situation, that where the work involved requires comparable skills. I can best explain how the bill would apply in such cases by describing the experience of women workers employed in an aircraft instrument plant surveyed by the Women's Bureau in 1946. The plant in question made precision instruments.

Before the war, workers were roughly classified as unskilled, semiskilled, and skilled workers. All women workers were considered to be unskilled and their employment was limited chiefly to inspection and packing. The war opened up new jobs for women in the skilled and semiskilled categories including machine operators, high-grade inspectors, testers, and assemblers.

Women continued to be considered unskilled workers and their hourly rates remained in the lower rate brackets, with only a 20 cent range between the highest and lowest woman's job. Sometime during the war, a worker-management committee was appointed to draw up plans for a new wage structure in the plant.

The committee worked with industrial engineers in analyzing, describing, and rating several hundred jobs on the same basis for men and women. The committee proceeded to set up 15 labor grades into which all jobs were fitted. The result was that many of the jobs that were previously classed as unskilled because they were filled by women were moved from the bottom to the near top or midpoint of the rate structure. Jobs on which women were employed were found in all but the two highest grades.

Women who performed work requiring comparable skills to men's jobs received equal pay after the job rating plan was installed. Appropriate rate adjustments were made and the spread in the hourly rates paid to individual women widened from 20 cents per hour to 62 cents per hour.

The situations I have described are among the principal types in which the investigation and hearing procedure of H. R. 4408 would be utilized to remove wage discriminations against women found to be doing comparable work with men. No Government department or board other than a temporary war agency has yet had the job of factfinding over the whole equal-pay field, consequently no quantitative information exists on the extent of such practices. However, the Women's Bureau continually brings to light individual instances or types of cases in the course of its wage investigations and instances of wage discriminations such as I have cited could be multiplied indefinitely.

H. R. 4408 also seeks to reduce discriminatory practices on an industry-wide basis, through establishment of industry committees authorized to make voluntary recommendations for the industry they represent.

Admittedly, such industry committees would be doing pioneering work. The nine State equal-pay laws do not expressly provide for them, but in at least one State, New York, such a committee which the Administrator set up independently did yeoman service in the early period of enforcement by recommending criteria for determining comparability of jobs. Under many other types of law, use of tripartite industry committees is a procedure of proven worth.

The fact was repeatedly demonstrated in my own experience as administrator of the New York minimum wage law. The New York

law provided for recommendations by tripartite industry committees, or "wage boards," of standards for minimum wages for women and minors based on both the fair value of the worker's service and the least amount adequate to support such workers in health. The employer and industry members brought to such wage boards an expert knowledge of industry which enabled the early New York wage orders to give proper consideration to such industry practices as split shifts, purchase of uniforms, laundry expense, involuntary lost time, and other conditions that directly or indirectly affected a determination of fair minimum wages.

The divergent character of a tripartite board-the differences in experience and approach brought respectively by labor, management, and public members-preserved a desirable balance in committee deliberations and tended to keep the recommendations on a practical workable basis.

Use of the industry committee procedure in the equal pay field offers potentialities of great value and usefulness. It affords a means of securing the voluntary cooperation of both management and labor to establish, as general standards, the practices already in operation among a substantial part of their ranks.

Each industry has first-hand knowledge of unfair practices peculiar to it that may happen repeatedly and lead to wage discrimination against women. Industry committees would put such knowledge at the disposal of the Government on a voluntary cooperative basis. That employer cooperation might be expected is indicated by a "test area" survey recently made by the New York Labor Department, of employers covered by the New York State equal pay law.

Approximately three-fourths of the employers who knew of the law readily expressed approval of it. Among the group which acknowledged not knowing of the law before, a majority approved of both the theory and the law against discrimination in pay rates on the basis of sex alone.

At this early period of legislative experience in equal pay, the scope of industry committee work can only be outlined broadly. Its nature is naturally determined in accordance with the purpose of the law under which it is operating. Such a committee would attack the problem on a broad scale through recommendations for industry-wide policies and standards.

One set of industrial standards with a direct bearing on equal pay that industry committees would probably correct relate to standards for setting up job classifications and personnel ratings. Job evaluation based on job content, that is, on a clear statement of what is required of any worker in order to do a given job—and eliminating extraneous factors such as sex obviously facilitates elimination of wage discriminations against women.

A recent study by the National Industrial Conference Board-studies in personnel policy, No. 86, November 1947-shows that employers are increasingly adopting the practice of describing in general terms applicable to all workers on a given job-what performance on that job is to be expected of them.

Performance up to the standard becomes an impartially applicable basis for the payment of the rate set for the job. The 1947 report, based on replies from 3,498 companies, employing approximately six and half million persons, states:

The obvious need for a more accurate and systematic basis of determining wage levels has led to the development of job evaluation.

Job evaluation plans have increased remarkably since 1939. In the 1939 survey, only 13.3 percent of the 2,700 cooperators were using job evaluation. In the present survey, this figure has jumped to approximately 57 percent. A significant point in connection with this swift expansion is the fact that job evaluation plans are fairly evenly distributed among small, medium, and large companies, and among companies in all of the industrial classifications covered in the current survey as well.

The report goes on to state that the method of determining rates, however, is a matter of individual company policy. Hit-or-miss methods of establishing basic wages and salaries are being abandoned primarily because they usually defeat attempts to formulate longrange policies.

As a long-time objective, equal pay for women is concerned not only with outright discrimination in individual instances, but with the broadening of women's opportunities for training and advancement. H. R. 4408, section 7 (b) makes this objective the continuing responsibility of industry committees. The problem is illustrated by the Women's Bureau study of Women in Shipyards (Bulletin 192-6, 1944), which reports that in the shipyards employing nearly 61,000 women, discrimination in upgrading women or in giving upgraded women the rate for the job was common practice. The study states that with the exception of the electric arc welding classification, case after case could be cited where women actually work as group leaders or as skilled craftsmen and only receive helper's ratings. Foremen, management, and unions resist extending an equal opportunity to women for supplementary training, upgrading, and supervisory work. Statistical data relating to need for equal pay:

In considering the need for legislation the committee will naturally want to know how extensive at present are inequities in rates between men and women for comparable work. I wish it were possible for me to be specific and all-inclusive on this point. It would be gratifying to give credit to employers, of whom we know there are considerable numbers, who have voluntarily adopted the practice of equal pay. It would be useful also to be able to say, for example, that so many hundreds of plants are paying women employees from 10 to 20 cents per hour less than men for a list of specific jobs.

I am not able to do that. Such a task has always been beyond us because of the size of the problem to be explored. That this practice of wage differentials is widespread, however, falls in the area of "common knowledge." The fact that the Federal Government issued specific directive orders during both World Wars to correct this situation in plants making war goods is conclusive evidence that such a situation existed, and will continue to exist if action is not taken to correct it.

We do, however, have some specific data which throw considerable light on this question. I should like to refer here to my testimony in 1945 relating to wage rates paid in the miscellaneous machinery industry, an industry in which considerable numbers of both men and women were then employed. The study was made by the Bureau of Labor Statistics for a pay-roll period in January 1945. That study indicated that

A distribution of straight time average hourly earnings of men and women in all plant occupations shows that while about 50 percent of the men earned $1 or more an hour, 85 percent of the women earned less than $1 an hour.

Now, these earnings figures include varying numbers of men and women doing different kinds of work, varying numbers paid by time and piece rates. They are not, therefore, an index of the extent of unequal pay on the same jobs. The distribution does serve, however, as a vivid example of the generally lower level of women's earnings even in an industry where large numbers of them are performing work similar to that of men.

If, on the basis of this same survey, we look only at the earnings of men and women classified in the same jobs, we find that in all but one of the 27 important classifications studied, in which both men and women were employed, men still had higher average straight-time hourly earnings, the differentials ranging, on a national basis, from 4 cents to 26 cents. In 11 occupations the differences are 10 cents or more. Incentive earnings account for some of the differences, but the influence of this and other factors, such as possible differences in seniority, cannot entirely explain the persistent one-directional differences in earnings.

If we narrow the comparison further and observe four important class C occupations in which both men and women were employed and were paid on a time basis only, differentials continue to be apparent, even though they are smaller than when incentive-paid workers are included. These differences range, on a national basis, from 1 to 7 cents, and in varying amounts in different geographical

areas.

The comparison from this study of the machinery industry that is most pertinent for equal pay is that of indexes of men's and women's earnings-on a time basis only-in five plant occupations in which both were employed. Earnings of both sexes are shown in relation to a common base, that of unskilled male workers-janitors and hand truckers. In relation to the base of 100 the men's earnings ranged from 105 to 130 percent, while the women's earnings in the same five occupations ranged from 94 to 125 percent, three of the five being from 4 to 6 points below unskilled male workers. These jobs, for which the average earnings of women were less were actually semiskilled jobs. In the job where men's and women's earnings were more nearly the same, there was a 4-percent difference; the other differences were 9, 18, 182, and about 20 percent.

More recent data for other industries indicate that unequal rates for jobs classified under the same job titles persist. I should like to point out comparative straight-time hourly rates, for example, from a BLS survey in the radio, radio equipment and phonograph industry, for a few occupations in which men and women worked in considerable numbers. These figures, as yet preliminary and unpublished, are for the United States as a whole; and relate to a January 1947 pay-roll period.

(The matter referred to is as follows:)

The differences range from 5 cents for class C inspectors to 22 cents for class B machine welders and 23 cents for working foremen. Four of the jobs listed show a 9- or 10-cent differential.

For the paint and varnish industry there are interesting figures for 12 cities for one occupation, labelers and packers, in which both men and women were employed. Data are available for two dates, July 1946, and August 1947. It is interesting to observe that while wages increased for both men and women in the year's period, the women are still earning noticeably less than the men. The differences range, for 1947, from 8 and 9 cents in Newark-Jersey City and Boston to 27 cents in St. Louis and Chicago.

(The following table was submitted:)

In the field of office work there is less extensive data.

However, a BLS study of 15 manufacturing industries in 1946 showed that among office occupations in those industries men general clerks averaged $1.03 an hour, compared to 80 cents for women. Men pay-roll clerks averaged $1.11 an hour, compared with 92 cents for women; and men hand bookkeepers, the highest-paid occupation, averaged $1.29, as compared with $1.12 for women.

I do not want to burden the committee with a great deal of statistical data in my direct testimony, but I attach to it as supplementary material three of the more recent occupational wage relationship surveys issued by the BLS, which will indicate the relationship between the straight-time average hourly earnings of men and women employed in the same job classifications in certain industries. The tables and charts in these studies for the wood furniture, textile cotton, and woolen and worsted textile industries are particularly pertinent illustrations of what we are trying to point out here that women's rates are, with minor exceptions, consistently lower than men's, even though they are doing generally comparable work.

Mr. McCONNELL. Miss Miller, you will see that the reporter gets the supplementary information?

Miss MILLER. Yes; I will give it to him right now. (The data referred to are as follows:)

NATION-WIDE WAGE RELATIONSHIPS

Men time workers occupied substantially higher positions in plant wage scales than women in similar jobs (chart 2). Wage differentials in favor of men ranged from 5 percent for boring-machine operators (operate only) to approximately 20 percent for drum and belt sanders. In general, the differences were greater among the higher-ranking occupations.