hundred employed Negro men were engaged in March 1940, claimed only 23 of each hundred in April 1947. Simultaneously, the proportion of all Negro men workers who had employment in the crafts rose by 25 percent; in the semiskilled occupations the increase was 50 percent, representing the entry of thousands of Negro workers into industrial plants. The ratio of clerk and salesman positions increased by over 27 percent, and that of proprietors, managers, and officials, 40 percent.

Among Negro women, the shifts between kinds of work were even more striking. In 1940, out of each 100 employed, 75 were in service occupations; in 1947, such work engaged only 49. During this period, the proportion of Negro women workers who were employed in retail stores and in various occupations in the clerical field quadrupled. The Urban League reported that more than 500 Negro telephone operators were in the employ of the area telephone company alone. In the semiskilled occupations, particularly of laundries and manufacturing establishments, Negro women opertives increased from 16 percent to 31 percent of all employed Negro women.

The Negro population of the city was estimated by the Urban League at 700,000. For the New York section of the New York-Northeast New Jersey metropolitan district, the Census Bureau estimated a total nonwhite population of 819,450, compared with 532,950 in 1940. The large increase during the period was attributed to migration of Negroes to New York in search of work.

The accompanying table shows the trend from service to more specialized occupations during the period covered.

Percent distribution of New York City Negro workers in specified occupational groups, March 1940 and April 1947

Labor-Management Disputes

in December 1948

WORK STOPPAGES due to labor-management disputes in December 1948 followed the usual seasonal trend, with a substantial decline noted in new stoppages, number of workers involved, and amount of time lost.

At the end of 1948 preliminary estimates of the Bureau of Labor Statistics show slight reductions in the year's strike activity as compared with 1947. The estimates indicate about 3,300 stoppages, which involved approximately 2,000,000 workers, and about 34,000,000 man-days of idleness as compared with 3,693 stoppages in 1947 involving 2,170,000 workers and 34,600,000 mandays of idleness.

No large stoppages began in December. An emergency board made recommendations for settlement of a threatened strike of the Nation's nonoperating railroad employees; a threatened stoppage of the Southern Pacific Railroad was postponed until January 15, 1949; the West Coast shipping strike of 28,000 workers which began September 2 was terminated on December 4, 1948; but a strike of 4,500 workers at the Kennecott Copper Co., Bingham, Utah, which began on October 25, was still in effect at the end of December. The strike of 1,600 printers against 6 Chicago newspapers which began November 25, 1947, continued through December 1948 although negotiations indicated a possible early settlement.

Shorter Workweek: Nonoperating Railroad Employees

On October 22, 1948, President Truman appointed an Emergency Board under provisions of the Railway Labor Act, to investigate the unresolved dispute between the Nation's railroads and their approximately 1,000,000 nonoperating employees. The employees were represented by 16 unions and had demanded a shorter workweek and increased wages. A strike vote was in process when the Board was appointed.

The Emergency Board made its report to the President on December 17 recommending that

1 The board consisted of William M. Leiserson, chairman, David L. Cole and George A. Cook.

the workweek be reduced from 48 to 40 hours with the same weekly pay for 40 hours as wa being paid for 48 hours and that present wage rates be increased by 7 cents an hour. It was recommended that the shorter workweek becom effective September 1, 1949, and that the 7-cen wage increase be made retroactive to October 1 1948. The Board also recommended that the parties negotiate directly on specified rules change with a view to reaching an agreement before September 1, 1949, suggested effective date of the shorter workweek.

Both parties agreed to further negotiations early in January using the Board's recommenda tions as the basis of their discussions.

Southern Pacific Railroad Dispute

A strike of 3,500 workers against the Souther Pacific Railroad, set for December 15, 1948, by the Brotherhood of Locomotive Firemen and Enginemen, to settle 296 issues which prio negotiations had failed to reconcile, was post poned until January 15, 1949. This delay was to enable a mediator of the National Mediation Board under the Federal Railway Labor Act to arrange further labor-management conferences and provide uninterrupted holiday travel. Fail ure of the parties to agree within the period may lead to the appointment of an emergency board The dispute involved operations in seven western States-Arizona, California, Nevada, New Mex ico, Oregon, Texas, and Utah.

Southwestern Bell Telephone Co. Controversy

Early in December it was reported that 81 per cent of the 50,000 members in Southwestern Divi sion No. 20, Communications Workers of Americ (Ind.), voted a strike against the Southwester Bell Telephone Co. affecting the company's man ual service in Texas, Oklahoma, Kansas, Arkansas Missouri, and a small portion of Illinois, as wel as the maintenance, repair, and installation service The union demanded a new contract with a wag increase, equal to that granted by other Bell Sys tem companies. Representatives of the Federa Mediation and Conciliation Service were instru mental in having the parties resume negotiation on December 10 and, as a result, the compan

offered increases of from $2 to $6 a week. The union countered with a demand for a flat increase of $6 and announced it had filed charges with the NLRB against the company for failure to bargain in good faith. Involved also is the dispute over the company's demand for $50,000 a year to cover the expense of handling deductions for union dues. A settlement, reached December 26, provided for a wage increase of 101⁄2 cents per hour or an average increase of $4.20 per week.

Musicians Union: New 5-Year Contract

After a ban of almost a year the general manufacture of phonograph recordings was resumed the middle of December, when James C. Petrillo, president of the American Federation of Musicians, signed a new 5-year contract with major record producers. Signing of the contract had been delayed until opinions as to its legality under the Labor Management Relations Act had been obtained from the Secretary of Labor and the Attorney General of the United States.

The Director of the Philadelphia Orchestra Association, Samuel R. Rosenbaum, was named

Wage Demands of West Coast Sailors

In early December, after the settlement of the 95-day maritime strike on the West Coast, the AFL Sailors Union of the Pacific presented demands for wage increases ranging from $20 to $60 per month to restore the union's relative position in the maritime wage pattern, and threatened to strike if a settlement was not reached within a 30-day negotiating period. Shipping operators countered with an offer to increase wages by $16 to $20 per month, but this was promptly rejected by the union. By December 15, the union had filed formal notice of its intention to reopen and modify the contract, under terms of the Labor Management Relations Act.

Errata

Technical Notes

The Rent Index-Part 2: Methodology of Measurement1

PRIOR TO 1942, the Bureau of Labor Statistics obtained its rental data from records of real estate firms, banks, trust companies, "large" landlords, etc. Since the summer of 1942, rent information has been obtained directly from tenants living in representative samples of dwellings in each city. Currently, the Bureau collects the information once each year by personal visit to tenants occupying the sample dwellings. At 3-month intervals between personal visits, data are obtained from the same dwellings by mail questionnaire.

Dwelling Unit Survey Sampling

In June 1944, the Bureau began a program of comprehensive dwelling unit surveys to provide master listings of tenant and owner-occupied dwellings for the selection of the samples for regular recurring rent surveys, and for other surveys requiring listings of individual dwellings in the sample design.

Size of Sample. The total number of dwellings included in these comprehensive surveys was determined on the basis of the total number of dwellings in each city. A minimum sample for each city was determined by the formula 10 times the square root of the total dwelling units, in order to provide larger samples for the larger cities and also to decrease the proportionate coverage as the city increased in size. Since the characteristics of the housing supply vary more in some cities than in others, this minimum size sample was then adjusted upward by a factor

1 Prepared by Helen Humes and Bruno Sehiro of the Bureau's Division of Prices and Cost of Living.

Part 1-Concept and Measurement appeared in the December 1948 issue of the Monthly Labor Review.

y dividing the total dwelling units in the city by the desired sample size. For the 34 large cities the consumers' price index the 1940 Census Housing Bulletins of Block Statistics 2 were availble and provided the basis for selecting repreentative sample blocks with dwellings in 1940, nd for identifying areas within the city which ad no residential dwellings at that time, but in hich construction may have taken place later. In the sample design, separate treatment was iven to heavily populated blocks in developed reas, to blocks which contained little or no develpment in 1940, to suburban areas, and to public busing.

esign of Sample. "Semi- and un-developed" meas included all areas within a city which were gnificantly larger than blocks in geographic size, ad irregularly shaped, and thus likely to be sites new construction. These were identified on aps and investigated by the field staff. If they ntained dwellings on the survey date, sample eas, equivalent to blocks, were selected in the eld by applying the city sampling ratios. This sured adequate representation of newly develmed areas of the city. The separate classification such areas in the sample design permits periodic spection for new construction.

"Zero" blocks-i. e., all blocks of normal shape ad geographic size which contained no residential wellings in 1940-were listed by Census Tract. he sampling ratio for the city was applied to this t, and the selected "zero" blocks were likewise vestigated for possible new construction. The paration of "zero" blocks also permits periodic spection for new construction.

"Developed" blocks included the remaining ocks of the shape and geographic size normal for e city which contained dwellings in 1940. mese were stratified by (a) predominant race of cupants in each block, and (b) number of wellings within each block. Blocks which coned less than twice the average number of its per block for the city were classified as mall" blocks. The remaining blocks were ssified as "large."

Sirteenth Census of the U. S. 1940 Housing Supplement to the First Housing Bulletin of Block Statistics. This bulletin is available for hof 191 cities which in 1930 had a population of 50,000 or more.

The city sampling ratio was applied to the listings of small blocks in the Census Bulletins of Block Statistics. Since the listings are arranged by Census Tract, their use insured adequate representation of all geographic areas of the city. The classification by race before the application of the sampling ratio insured adequate racial representation. For these small sample blocks, detailed information on the tenure and description was recorded on a form (BLS 1674) for every dwelling located within the block boundaries.

For each large block, a separate card, indicating Census Tract and block number and total number of dwellings in the block, was prepared and arranged in ascending order of the total number of dwellings in the block.

In large blocks, dwellings are concentrated in relatively few structures and thus are likely to be very similar both in structural and price characteristics. There may, however, be wide differences both in price and dwelling unit description among these blocks. In order to obtain greater representation of large blocks and to reduce the coverage within these blocks, multiple samples of large blocks were selected and detailed information obtained on the listing form (BLS 1674) for only a portion of the dwellings ("in-block" sampling) in these blocks. The "in-block" sampling ratio was determined to give the same proportionate coverage of units in the large and the small block strata. Generally speaking, triple samples of large blocks were selected, and data were obtained for every third dwelling within these sample blocks. For example, in Scranton, where the unit sampling ratio was 1 in 15, every fifteenth small block (under 50 dwellings) was included in the sample, and data were obtained for all dwellings in these sample blocks. One-fifth of the large blocks (50 dwellings or more) or a number equivalent to three samples was selected, and data were obtained for every third dwelling unit within each of these sample blocks.

In cities where there was a wide variation in the number of units per block in the large stratum, the large block stratum was divided into several substrata requiring variations in the "in-block" sampling ratios. For example, in Manhattan borough of New York City, the large block stratum varied in size from 100 dwellings units

817250-49