Page images
PDF
EPUB

priced bimonthly) to the maximum monthly pricing possible. The basis for this proposal is that increasing the number of monthly previous price estimates beyond that being used (which is a significant improvement over the old procedure) will reduce or eliminate the average lag in the price index. The new procedure introduced in the revision program reduced the lag from almost three-fourths of a month to slightly more than one-third of a month. However, while data lags are always an important problem in making month-to-month comparisons, the issue appears to be one of price volatility and the weight associated with the lagged items with volatile price movements. If there is little or no month-to-month price movement, monthly price collection at any increment in program cost is too expensive. If there are infrequent but sharp monthly price changes the problem is how to identify those changes without instituting monthly pricing. To this end further study of the pricing is necessary. In such a study consideration should be given to arranging for some form of monitoring through publications and/or exception reporting (i.e., intermediate price change reports) in each important situation where prices would otherwise be collected on a bimonthly or less frequent basis. There may be severe problems in instituting and maintaining such a program within the operational constraints of the CPI. Nevertheless, if operationally and technically feasible, such an approach could produce up-to-date accurate quotations on items usually priced less often than monthly and at a considerably lower cost to the Federal Government than more regular monthly pricing.

Consumer Expenditure Data

The revision of the CPI market basket weights and the updating of pricing outlets has traditionally been accomplished through massive periodic (10-15 years apart) consumer expenditure and point of purchase surveys. This procedure, given the rapid pace of innovation, accompanying changes in consumer preferences, and the availability of modern statistical procedures, should be discarded in favor of a continuing data collection program that would both: (1) allow for the maintenance of a statistically reliable sample of pricing outlets (e.g., stores), and (2) the continuing analysis of the relative importance and changes of different classes of goods and services to consumers. Such information which is of importance in its own right also would permit more timely revision of the CPI market basket and weights when much revision is necessary. A continuing consumers'

'Change in the market basket and in weights should be introduced probably not more frequently than about every 5 years.

expenditure survey would also form the framework for an improved family budget program.

The BLS has been instituting a continuing pointof-purchase survey for which funding was provided in FY 77. The President's budget for FY 78 provided for a Continuing Consumers Expenditure Survey (CCES) program. However, the program was delayed for a year until the CPI revision was complete so is being considered for the FY 79 budget. In planning the CCES, the Bureau of Labor Statistics should take into account the need for analyses of the income and expenditure patterns of major population components (e.g., the elderly, the poor population residing in urban areas and in rural areas, and so on).. This design work should be coordinated with the Department of Health, Education, and Welfare which is now planning a different but somewhat related study program, and the Bureau of Economic Analysis for estimates of personal consumption expenditures as an important part of the national income accounts.

Family Budgets

The family budget program provides data for three normative levels of living for a hypothetical family. It is used by many to form the basis for the evaluation of the adequacy of family income and in measuring place-to-place differences in living costs. In addition, the family budgets as published by BLS are used for many other purposes including wage negotiation. Nevertheless, the program as presently structured presents many conceptual and technical difficulties which give rise to serious questions as to the adequacy and suitability of the data for the purposes for which it is used. These deficiencies are thoroughly discussed in "Appendix 4: Bureau of Labor Statistics Family Budgets Program" to the HEW report The Measure of Poverty (April 1976) and to some extent in the BLS Handbook of Methods (BLS Bulletin 1910) on pages 82-86. A somewhat different viewpoint is to be found in the BLS Bulletin series 1570-1 to 6. In addition, a diametrically opposing view of the family budget program, arguing that the present formulation is correct and simply requires updating, has also been expressed by some users of the data series.

While the legislative requirement suggests that the program cannot be discontinued without statutory authority, the program does require major modification. Alternative and widely different conceptual approaches will be explored by a subcommittee of the Statistical Policy Coordination Committee, working in cooperation with a technical committee formed by BLS. These include, but are not limited, to:

1. Restating the normative values which form the basis of much of the budget into current terms and then pricing and updating the cost of these values on the basis of a specification pricing program conducted in conjunction with the CPI. This approach would be comparable to the basic program formulation initiated in 1946 and updated in 1959 and in 1966. It suffers from all of the deficiencies noted in Appendix 4 to The Measure of Poverty, primarily that there are no objective criteria to guide BLS in determining what items must be purchased by a family in order to enjoy a stated living standard which in turn is also subjectively determined, except that the data base would be current and would be maintained on a current basis.

2. Basing the series on actual expenditure patterns at various income levels developed through the most recent Consumers' Expenditure Survey (CES) and updating the results through data collected in the proposed continuing CES. This approach would eliminate all normative judgments. It would suffer, as does the current approach, in not appropriately providing for place-to-place comparisons of the cost of maintaining the "same" purchasing patterns. This is a necessary concomitant of developing budgets based on actual expenditures since the expenditure patterns in different areas reflect regional differences in clothing and housing requirements and in food and recreation among other expenditure elements. Also, different purchasing patterns may result either because of preference or because some items freely available in one place may be either unavailable or available only at a high price in other areas. Thus, budgets for different places based on actual expenditure patterns in each place may not reflect the relative difference in cost of an identical or comparable living standard in the various places.

The Family Budget Subcommittee and the Statistical Policy Coordination Committee should provide guidance to BLS as to the appropriate changes that should be made in the family budget program so as to achieve the purpose(s) for which the program is used.

International Price Index Program

It is expected that data from this program will be used, among other purposes, to deflate the value of all exports and imports in the gross national product accounts as soon as sufficient information for that

purpose is available. This data will provide for more accurate deflation than is possible using the unit value indexes previously employed for this purpose. Resources currently available to the program will provide for the coverage of about 50% of the value of all imports and 65% of the value of all U.S. exports of products. The international price index program should be expanded to cover 100% of the value of imports and exports. The completion of this program, conducted by BLS on the basis of a sample of transactions from export declarations and import records held by the U.S. Customs Service, could permit the discontinuance of the unit value indexes compiled by the Census Bureau. However, only a few countries have a pricing system for exports and imports other than unit values. Accordingly, the need for unit value indexes may continue after the international price program has been completed.

Producer Price Program

The producer price program has three major elements: the Producer Price Index, formerly called the Wholesale Price Indexe (WPI), the stage of processing indexes, and the industry sector price indexes (ISPI). These series suffer from several conceptual, statistical, and operational inadequacies:

1. The PPI data are used as precursors of change in the CPI. This is without regard to the fact that the data are not designed to be used for this purpose. Some price indexes necessary to better understanding of the price mechanism are missing entirely: e.g., input materials indexes, final demand indexes, and industrybased stage of processing indexes;

2. The PPI is based on a hodgepodge of prices to different sectors-some of which are final demand sectors, and some of which are not. Further, the aggregation structure by total shipments in the PPI's leads to multiple counting of price changes;

3. Many major industry sectors are either unpriced or have poor coverage: ISPI's are available for only about one-fourth of the four-digit manufacturing and mining industries; PPI's are available for less than half of the total value of shipments in agriculture, manufacturing and mining; and almost nothing is priced in the service or construction sectors;

4. Prices collected are heterogeneous in naturesome are spot market prices, some contract prices, others are based on prices of delivered

orders, list prices, nominal or trade journal prices;

5. Sampling is judgmental and many items are priced on the basis of very few quotes. Probability sampling (with expanded samples) would provide measures of reliability. Although not all bias can be removed, it is likely that probability sampling will reduce it. Because of these limitations, gaps exist in our knowledge about both specific price movements and the total structure of price relationships. In addition, the current data may be misleading, but the degree of misrepresentation, if any, cannot be assessed.

To correct these deficiencies the producer price program should be improved using the recommendations in the Ruggles' report for the Council on Wage and Price Stability. To this end, the Bureau of Labor Statistics has launched a comprehensive program to revise the producer price program and its components and provide accurate measures of price change for the major sectors of the American economy. To achieve this objective, the revised program will provide a complete system of price measures consisting of the following major elements:

1. Output price indexes for all major industrial sectors;

2. Detailed commodity indexes and service indexes covering both primary and secondary production;

3. Stage-of-processing indexes which accurately reflect processing flows in the economy;

4. Input materials price indexes for all major industry sectors; and

5. Final demand price indexes of major industry

sectors.

The indexes will be based on a statistical sample of establishments within four-digit SIC's and within each establishment on a probability selection of commodities or services with each sample being stratified by size of firm and sales volume. The data obtained after initial selection should be specification priced and should permit the completion of the types of indexes noted earlier. (This type of approach seems to be consonant with the recommendations contained in the 1977 study, "Review and Evaluation of the Wholesale Price Index" conducted by Professor Richard Ruggles under the auspices of the National Bureau of Economic Research for the Council on Wage and Price Stability. The BLS approach to the revision of this program is comparable to virtually all of the other suggestions in the Ruggles report.)

The procedure can be carried out by BLS on the basis of the establishment universe material currently in its possession. This revision, which is now being phased in, is expected to continue for many years. When the revision is completed the data are expected to be significantly more complete, accurate, and useful than the present data. Nevertheless, the availability of sales volume data, as would be available if the Standard Statistical Establishment List information were available to BLS, would enhance the sample procedure by allowing for direct stratification of establishments on the basis of volume instead of employment (as in the BLS universe material) as a proxy for volume (see related chapter on standards development). It is believed that this approach would produce statistically reliable data in considerably more detail which can be aggregated in more meaningful ways than is now the case. However, the extent to which the availability of the Standard Statistical Establishment List would enhance the BLS effort is, in fact, not known, and furthermore there appears to be little question that the BLS approach will produce statistically reliable data in considerably more detail and aggregated in more meaningful ways than is now the case.

While the comprehensive revision being implemented will provide for a vast improvement in the quality of price data now available, there are some difficult data problems which would still remain. For example, price indexes do not now include "build-to-order" commodities such as commercial aircraft, ships, and some computers. The very nature of a build-to order commodity introduces severe technical questions of measurement and quality adjustment. Efforts toward developing a conceptual framework for the pricing of such commodities should continue.

Tuesday Spot Market Price Index

This index which shows changes in prices for 23 commodities is not based on direct transactions data, but rather on reports of prices as reported through trade and government sources. Moreover, spot prices reflect only the price that a seller in a spot market is willing to accept at that moment, and does not reflect actual transactions. Spot prices, moreover, do not generally reflect total transactions prices in the regular course of commercial events. The items priced, however, are considered to be highly price sensitive. Presently the Bureau of Economic Analysis uses data from this Index in developing preliminary estimates for the current month in several Business

'See separate chapter.

Conditions Digest series. However, the producer price program when revised will provide reliable monthly data for all or many of the commodities included in the Tuesday Spot Market Index. When those data are

available the long-term value of this measure to the Federal Government and its relative importance in terms of all other statistical program priorities should be reassessed.

Chapter 15. PRODUCTION AND DISTRIBUTION STATISTICS

Introduction

The production and distribution of goods and services is the most basic economic fact of life, ever since man first bartered one product for another. Although the measurement of production did not start until much later (probably with a flour miller making marks on the wall), statistics of production and distribution have been kept for centuries.

Early statistics were probably limited to a count of products made and sold and the number of persons employed. These two statistics are still the main measures of production and distribution today. The complexity of today's marketplace has meant that each and every different product cannot be counted separately, and dollar values are now used more and more to measure the production of a group of similar products.

Production, for purposes of this section, means those goods which are produced or those services which are performed by labor and capital resources and managerial skills and enter directly or indirectly (through intracompany sales or consumption in the same plant) into commercial trade. Distribution means the offering of goods or services at a specific time and place, in a particular quantity or amount by companies with wholesale or retail trade or transportation activities.

Many aspects of production and distribution are of special importance or of current interest. They are covered separately in the chapters on agricultural commodities, construction, transportation, energy, health, and education.

The areas that are the subject of this chapter are mining and manufacturing, wholesale and retail trade (including international trade in these commodities), communication and public utilities, finance, insurance, real estate, services including repair, and public administration.

The agencies collecting production and distribution data are the Bureau of the Census, Bureau of Mines, Interstate Commerce Commission, U.S. International Trade Commission, Social Security Administration, Department of Trans

portation, Federal Trade Commission, Corps of Engineers, Department of Energy, Federal Reserve Board, Small Business Administration, U.S. Forest Service, and many others.

Users of Production and Distribution
Statistics

Production and distribution statistics are used by almost every organization-public, private, national, international, state, or local-concerned with the economy of the United States as well as many organizations concerned with the social aspects of the nation. They are used in projections and analytic studies of capital, energy, labor, and civil defense requirements; defense mobilization; pollution control; international trade agreements; taxation; small business definition and evaluation; macroeconomic projections; production bottlenecks; antitrust actions; and input-output, industrialoccupational, and import-export models. In most cases the information is used directly, but in many others the use is through the Gross National Product (GNP), Industrial Production Index, or the Wholesale Price Index. In some series, these statistics are not used directly, but are needed to complete the measures of the economy at a higher level of aggregation.

Basic Source Program

Collectively, the quinquennial economic census is the largest Federal program of production and distribution statistics. In addition to this census, various Census Bureau monthly, quarterly, annual, or occasional programs fill in the data needs where more frequent data are required. Most of these intermediate series are benchmarked to annual and/or quinquennial series.

The quinquennial censuses provide the basic statistical profile of the Nation's businesses. Included in these censuses are the censuses of agriculture, mineral industries, construction, manufactures, wholesale trade, retail trade, services and governments; the census of transportation collects

« PreviousContinue »