Technical Note Consumer Expenditure Study, 1950: Field Methods and Purposes1 COLLECTION of consumer expenditure information in 91 cities throughout the country was recently started by the Bureau of Labor Statistics and will continue for about 3 months. More than 1,000 interviewers began knocking at doors of over 17,000 city families and single consumers who bave been selected, according to the most scientific sampling techniques, to represent all consumers in urban America. Questions asked include a description of family composition and living arrangements and the maximum possible detail on financial experience for the year 1950. Consumers are asked to recount their expenditures in 1950 for food, clothing, housing, transportation, recreation, etc. The Bureau wants to know the detail of these expenditures in prices and quantities for many of the thousands of things that make up the American way of living. And, in order to evaluate and analyze this information properly, data are also gathered on the family income, savings, and credit used during that year. Need for the Survey This extensive survey of consumers' expenditures is an essential step in the over-all revision of the Bureau's Consumers' Price Index 1 which measures changes in prices of goods and services commonly bought by moderate-income families in large cities. The index is in essence a ratio between the current cost of a specified "market basket" of goods and services and the average cost of the same "market basket" in the base period, 1935-39. The "market basket" currently priced for the indexes is based on records of purchases of moderate-income city families for a 1-year period in 1934-36. In the past 15 years, consumer incomes, prices, the kinds of things available to consumers, and the consumption habits of American families have changed greatly. Expenditure surveys made by the Bureau in seven cities (including Richmond which the prevent survey omits)2 between 1947 and 1950 provide accumulated evidence of such changes. Compared with the mid1930's families use less butter and more margarine; and less flour and bread and more ready-to-eat cereals, cakes and pies, more ice cream, soft drinks, fresh fruit, and canned fruit juices. Frozen foods have become commonplace; ice and ice boxes have given way to mechanical refrigeration; television sets are owned by almost every third family in cities which have well-established transmitting facilities. City families spend a greater proportion of their income for automobile transportation, medical and personal care, recreation and other less urgent needs than formerly; a somewhat smaller part of the total goes for food, housing, and clothing. Changes of this kind make it necessary to check the "market basket" of goods and services now used in the Consumers' Price Index and are the immediate reason for this survey. On the basis of the survey results, the content of the "basket" will be revised. The uses of the survey data, however, go beyond the adjustment of index weights. Through analysis of the survey data, businessmen may learn where they are likely to find their best customers; welfare workers may obtain information in planning family budgets and determining relief allowances; manufacturers may find a clue to the amount and kinds of unsatisfied wants the kinds of goods and services consumers demand. Summaries of the survey results will be of help in estimating the levels of industrial operation affecting labor requirements, the replacement requirements of durable goods, the need for new housing, the effect of soldier bonuses on the economy, and in dealing with many other important problems. Confidential Nature of Data Consumer expenditure information is collected on a purely voluntary basis and is treated as strictly confidential, in line with general Bureau policy. No one but sworn agents sees the report for an individual family, and the identity of the family is never disclosed. The Bureau itself does not want to know the family name or the first names of any of the members-the name of the family interviewed does not appear on any schedule form. No figure obtained from any family is made available to other Government agencies for taxation or regulatory purposes. The results are published in the form of averages for large groups of families. For its analytical work, the Bureau calculates averages for families of the same type, same size, same income level, and living in the same kind of community. Study of these figures yields information on the differences in living patterns of American families, and ultimately will reflect changes that have occurred since the last comprehensive Bureau survey in the 1934-36 period. Content of Schedules The schedule forms and collection methods for the 1951 survey were develped out of long experience with this type of study. In the spring of 1950, the Bureau tested its schedule and the training and field collection aspects of the survey in Memphis, Tenn. The results in Memphis were used in determining the final design of this survey. The schedule in use has 26 major sections. General information on the family composition, living arrangements and facilities, as well as on home owners' expenses for repairs and improvements, receipts from roomers and boarders, etc., is requested in the first eight sections. The following 10 sections are devoted to detailed information on expenditures for such groups as fuel, light, and refrigeration; miscellaneous household expenses; housefurnishings and equipment; food; clothing; medical and personal care; recreation, reading, and education; and travel and transportation. Questions on income, taxes, savings, and debts cover the last eight sections of the schedule. Information on income is recorded, according to its source. Savings data are intended to show the change over the calendar year and not the total amount saved by the family or the single consumer who supplies the figures. A supplementary questionnaire, covering food, household supplies, tobacco, and drugs and personal care items purchased in a 7-day period, has been prepared to provide more detailed information on expenditures of families who buy food and prepare it at home. This schedule is divided into three parts: (1) food items, (2) household supplies and tobacco, and (3) personal care and drug items. Sample Selection The families and single consumers to be included in this survey were carefully selected to be representative of all consumers in the urban areas of the United States. This is possible, of course, under modern sampling methods which provide a way of finding out about consumers in general by interviewing a comparatively small number of them. In drawing the sample for the present survey, great care was taken to insure proper geographic coverage of cities, and within cities, a good cross section of the population. All cities with populations 1,000,000 and above were taken into the sample. Cities below 1,000,000 were ranked by size and classified according to their climate, population density, and the general income level of the community. Very small cities were further classified by a measure of their distance from important market centers. Within each classification group, cities were selected at random.3 The list of cities selected, which is arranged by States, follows: Alabama-Birmingham, Demopolis. California-Los Angeles, San Francisco-Oakland, San District of Columbia-*Washington. Georgia-Atlanta. Illinois-Chicago, Bloomington, Anna. Kentucky Louisville, Middlesboro. surveys were not available. Addresses in the lists were arranged by their location in the city, by race of occupant, by tenure of occupant (owner or renter), by rent paid (for tenants), and by the number of persons in the dwelling and the income class of the occupants, to the extent that such information was available. In drawing the sample units, the distribution of these characteristics (race, tenure, rent, etc.) in the total population of the city was duplicated in the sample. The number of addresses selected varies from 65 in small places to over 600 in New York City. A subsample of 1,000 families was selected in 47 sampling points for a panel study of individual family patterns of expenditures over a period of time. Data thus obtained will be dynamic and will indicate how families react in their buying to changes in income, prices, or other economic or noneconomic factors. Moreover, the panel will serve to check the validity of the index weights over time; it therefore plays a major part in the maintenance of the index. Conduct of Survey The field work started in January is to be centrally directed from Washington through the Bureau's five regional offices. Work in each city is the direct responsibility of a survey supervisor hired and trained specifically for the purpose. These supervisors were recruited from all over the country by the Washington office and by the regional directors of the Bureau's regional offices. Seventy-five supervisors were selected and given 5 weeks of training during which they were taught the Bureau's interviewing and editing methods. In the training program the most recently developed techniques, combining lectures and discussions with auditive and visual aids, were utilized. Each supervisor was assigned between 65 and 250 interviews. In large cities like New York or Chicago, when the number of interviews exceeds 250, two or more supervisors are being employed. On the other hand two or three cities' surveys, in which the number of interviews is small, have been combined and assigned to one supervisor. The functions of the supervisor cover (1) recruiting and training the interviewers, (2) directing the collection work, and (3) checking and editing the schedules, before transmitting them to Washington. Interviewers are recruited in each city surveyed, and are trained for 6 days before their interview work begins. They are assigned a list of addresses, where they will obtain the required information from the housewife or other persons able to answer their questions. The use of elaborate and sound sampling techniques, careful selection and training of supervisors and interviewers, and constant control and reviewing in the process of collection are essential to this enterprise. Its success, however, depends in great part on the cooperation given by the public. 1 Prepared in the Bureau's Division of Prices and Cost of Living. * See a previous technical note, Revision of the Consumers' Price Index, in the Monthly Labor Review for July 1950 (p. 129); also reprinted as Serial No. R. 2003. The technique known as the Latin Square, which has been followed for the selection, is to be described in detail in a technical note in a forthcoming issue of the Monthly Labor Review. ♦ It is important not to consider this as the list of cities to be included in the revised Consumers' Price Index. The exclusion of a city from the above list does not preclude its inclusion in the revised Consumers' Price Indexand vice versa. The city characteristics used in the selection of this list are those which are related to and have some effect on consumer expenditure patterns. The index cities are used as pricing points to obtain information on the movements of prices over time; they will be selected according to the result of experimental pricing studies currently in progress. These studies are to be described in future issues of this journal. Recent Decisions of Interest to Labor' Wages and Hours 2 Coverage Employees of Plant Cafeteria.-A Federal district court held that the employees of a cafeteria company operating on the premises of a steel corporation were not within the coverage of the minimum wage and overtime compensation provisions of the Fair Labor Standards Act of 1938. Since the steel company employees were not compelled to patronize the cafeteria and could, and many did, go to several other nearby restaurants, the cafeteria employees were held not to be "necessary" to the production of goods for commerce within the meaning of the act prior to its amendment in 1949. The steel company's plant was located in a town of 12,000 near Los Angeles. There were three restaurants across the street from the main plant gate during 1945 and 1946, the period for which wages were alleged to be due. The court stated that, for an employee operation to be "necessary" to the production of goods for commerce, the operation need not be indispensable-in that the goods could not have been produced without such operation. In this case, however, the court held that the furnishing of food to the steel plant's employees was too remote from the production of goods by such employees to be considered "necessary" thereto within the meaning of the FLSA. Cases holding plant cafeteria employees to be integrated with production were distinguished because they involved plants in which employees had no ready access to other eating places. The cafeteria-canteen was operated under a "leasehold" contract with the steel company. Under this contract the cafeteria was permitted to use, at no cost, a building erected by the company. In return, the cafeteria operator agreed to serve meals only to the plant employees and their visitors, and to pay expenses of operating the cafeteria. Cafeteria profits were limited by the agreement to 10 percent of gross income. The cafeteria was available to employees on any of the three 8-hour shifts maintained by the plant. Other than that, the agreement did not prescribe the hours worked by cafeteria employees. The district court held that the agreement between the steel company and the cafeteria operator was in fact a lease and not a license or a joint adventure, despite the provision limiting profits. It was pointed out that, at least, no formal agreement existed by which the steel company was bound to reimburse the cafeteria for operating losses. employees of the steel company, since they were found not to be subject to the direction of the company except indirectly in determining the period when plant employees were to be served. Portal Act-"Good Faith" Defense. The Chairman of the National Railway Labor Panel ruled that employees working at a "modification" center of an airline company were subject to the Railway Labor Act and exempt from the Fair Labor Standards Act. The Court of Appeals for the Eighth Circuit held that this was an "administrative ruling" of an "administrative agency" within the meaning of the good faith defense provisions of the Portal-to-Portal Act. Section 9 of that act, which applies to the period prior to May 14, 1947, relieves an employer from liability for violations of the FLSA when such violations are in good faith in conformity with and in reliance upon any administrative ruling of any agency of the United States. During World War II the company, besides operating an airline, "modified" army planes produced elsewhere. Such planes were later used in combat. The company sought rulings from certain Government agencies as to whether modification-center employees were covered by the Railway Labor Act and, therefore, exempt from the FLSA under its section 13 (a) (4). Regarding these rulings as inconclusive, the company segregated modification activities from airline activities so as to take the former out of the exemption and allow payment of overtime compensation to employees. However, the Chairman of the National Railway Labor Panel, disapproved the payment of such overtime on the ground that the employees were covered by the Railway Labor Act. Accordingly, the company paid no overtime for hours worked over 40 a week. Modification-center employees subsequently sued the company for overtime compensation. A Federal district court permitted recovery by the employees. It held that the employees were engaged in the production of goods for commerce. It also held the modification activities too remote from carrier activities for the employees to be exempt from FLSA because of coverage by the Railway Labor Act. In these respects the decision was upheld by the court of appeals, in view of a recent United States Supreme Court decision that employees working for a cost-plus-fixed-fee contractor in a governmentowned munitions plant were covered by the FLSA. the court of appeals reversed the district court decision to the effect that the good faith defense of the Portal Act was not applicable. But The district court had held section 9 inapplicable on the ground that the ruling by the Chairman of the National Railway Labor Panel was not the ruling of an agency, and was unauthorized. Executive Order 9299 required carriers to file notice of proposed changes in wage rates with the Chairman of the Railway Labor Panel. If he had reason to believe the proposed change did not conform with the standards prescribed in Executive Order 9250 (wage stabilization order) and such change was not modified to conform to the standards, the chairman was to designate three members of the Panel as an Emergency Board to investigate the change. The district court held that the |