ACKNOWLEDGMENTS

I wish to express my sincere appreciation to Henry H. Perritt, Jr., Deputy Director of the Office of Economic Stabilization, and Jonathan Brock, Director of the History and Archives Project for their support of this paper and espeially for assuring adequate resources to provide internal data from the Economic Stabilization Program, separately published in the data appendix.

I would also like to thank my summer staff for their competence and diligence in performing their jobs under the uncomfortable circumstances of temporary employment:

In addition, I would like to offer my special thanks to Bland Roberts and Dawn Thompson for managing the Schedule Credit during the last days of the Cost of Living Council, to Barry Goldberg for designing and managing the system update and edits of this same data, to Henry Resnikoff for his collaboration in the Phase IV data file redesign, and to Robert Mize and Ramon Morales for their unslacking efforts in editing the manual data base from the first days of Phase II through the Post-Phase IV cleanup. I also appreciate and wish to thank those who read and commented on the draft of this paper:

Robert Anthony

Jerry Dickens

Conrad J. Jacobs

William Covell
Charles Emley
David C. Lund

Robert F. Mize

Leslie Schwartz

Don Wortman

Edwin B. Parker
Gilmore S. Wheeler

Of course, all errors and evaluatory comments can only be attributed to the author.

Finally, I wish to thank all participants, public and private, in the data efforts of the Economic Stabilization Program for all those who believe that the resolution of opposing economic theories and policies is possible with the applications of sufficient amounts of quality data, which can be used to test and falsify inaccurate hypotheses.

PRICE DATA AND DATA SYSTEMS

Introduction

This paper on price data and data systems discusses the acquisition, retention, and use of price data during the Economic Stabilization Program. Second only to data availability and quality, the efficacy of the data systems was the largest constraint on the efficient use of the information resources of the Program. During most of the Program, there was a firm policy commitment to a computerized storage and retrieval data system. This commitment came, in part, from the desire to minimize the Program's personnel requirements by substituting more capital equipment that had been used or available during prior price control efforts.

This paper contains chapters concerned with the data and data systems in each of the major phases of the Economic Stabilization Program. It describes the hardware, the software, and the various interfaces between data input and report generation along with the extraordinary problems of installing complex data systems under severe time constraints.

In the general discussion of data uses, three distinct functions are particularly important: (1) data in a direct controls program used to decide on price increase requests and to monitor the compliance with the pricing regulations; (2) data collection used to support the analysis and development of policy alternatives by identifying upcoming problem areas; (3) the data used to support basic reasearch on inflation. The need for data under the first and second functions is (at this writing) no longer immediate, due to the demise of the controls and monitoring program. With respect to the basic research function, better data and more work are needed to understand the structure of prices and price formation in a dynamic economy.

In an effort to facilitate this research objective, some internal company price and cost data filed with the Price Commission and the Cost of Living Council during Phases II, III, and IV have been aggregated to a level that would not disclose proprietary information and are separately published as a statistical appendix to this volume. Although wage data systems are not the subject of this paper, Phase

II, III and IV wage data are also included in this appendix.1 A full explanation of all reproduced data and the assumptions involved in its preparation are also included in the appendix. In addition, efforts are being undertaken to place some of this non-proprietary data in computer-readable form. The availability of computer tapes can be ascertained by contacting the National Archives in early 1975 and inquiring about the Data Tapes of the Economic Stabilization Program.

Data Requirements and Data Availability

Two things became immediately apparent when considering the data requirements which arise from the implementation of a stabilization program. First, if the problem is to be regulatory with industry pricing actions monitored to check compliance with authorized increases, the data requirements may be burdensome to the stabilization authority as well as to those business entities which must comply with information-reporting regulations. Second, even if the program is not designed to be regulatory, there is a paucity of data to support government action attempting to affect prices through "jawboning," supply and productivity augmentation. The essential distinction, then, is whether or not data are needed for regulatory purposes. If the stabilization program includes a set of specific regulations governing price and wage increases, a different set of demands for data are made than those of a program primarily designed for monitoring inflation and inflationary pressures in the economy; a regulatory program, however, will also wish to consider these other data requirements as part of its on-going interest in the problem. THE USE OF DATA FOR CONTROLS AND COMPLIANCE

The majority of attention during the Economic Stabilization Program data collection effort was directed at the development of a data base for administering the pricing regulations of the Program. In such a case, the data requirements are obvious: the data must match the regulatory model of price formation. Since the regulations on price formation are not necessarily related to the economic pricing models followed by the firm, these data collection efforts have no inherent link to the economic and the natural functions of the market. For example, if the prices charged by a firm could be tied, in a regulatory scheme, to the amount of wage increases, the data requirements of the Economic Stabilization Program would then consist of information about the size and timing of wage increases.

1 For a fuller discussion of wage data systems under the Economic Stabilization Program, see the Wage Case Processing paper in this compendium or Steven Alter, "Cost of Living Council-A, B,&C," an unpublished case study of on-line computer systems at the Cost of Living Council, Sloan School of Management, M.I.T., 1974.

The specific definition of data elements, the frequency of collection and the breadth of collection is defined by the regulations. For example, if prenotification of price increases is required by regulation, data are collected before each price movement. The extent of prenotification is determined by the requirements which specify which firms must prenotify and thus, determines the firms that supply data. While there may be some difficult technical issues to be resolved, such as designing forms and refining concepts for measurements, the needs of the regulations will determine whether or not all data must be collected de novo to meet stabilization program requirements. A number of agencies already collect data from a large population of firms. This includes income and cost data going to the Internal Revenue Service, profits and other operating data going to the Securities and Exchange Commission and the Federal Trade Commission. Clearly, if the path of tailoring regulations to the data is chosen, the stabilization program must accept the concepts, definitions, and reporting formats imposed by these other agencies. However, the range of permissible pricing rules for the regulations is severely restricted by the availability of appropriate data to enforce those rules. Therefore, it is not surprising that any regulatory scheme for stabilization would rely upon data collected by itself for its own specific purposes.

THE USE OF DATA FOR ANALYSIS AND
RESEARCH

The implementation of direct controls can be viewed as an attempt to break a cost-push inflationary cycle or to alter expectations about inflation that have become built into price and wage adjustments. There need be little link to the behavior of firms in the market place. However, even in a regulatory setting, there are questions which relate to expected behavior of firms within the market place. These relate to predicting future trouble spots in the economy, to ascertaining the economic effects of proposed regulations or policy changes, and to developing policies intended to alter the behavior of firms in the market place. These latter interests and activities represent both short run and long run functions of the stabilization agency.

The data requirements for these activities are much more nebulous than for the regulatory function because the significant underlying economic behavior is not well understood. Economists do not have good working models of price formation and investment behavior that predict well across various industries in the economy.

These latter data needs are those that the Cost of Living Council proposed be contained under a follow-on inflation "monitoring" agency. Even though this follow-on agency was not created by Congress at the end of Phase IV, it is still useful to discuss the types of activities it usefully might have considered-particularly with respect

to data availability and data requirements. Since this activity would be much broader in scope than the regulatory activity, it is more likely that other government data sources would be useful.

A useful starting place is the Economic Stabilization Program as it had developed in Phase IV. Price increases were allowed whenever production costs increased; the amount of increase allowed was equal to the dollar amount of a cost increase. Large firms prenotified price increases on the basis of increased costs. The Cost of Living Council attempted to utilize this data to determine what particular price increase approvals for specific product lines would mean in terms of observed prices in the economy.

In order to answer this question, it is necessary first to know how much of an authorized price increase was actually put into effect by the firm. The existence of increased costs and authorization to increase prices did not provide information on what proportion of the increase would actually be put into place. Clearly, prices are not entirely cost driven; historically there have been significant changes in aggregate profits in the economy over business cycles, and there is ample evidence to show that individual firm's profits also have considerable variance. Therefore, there must be factors other than cost factors at work that affect the prices of products. Traditional economic theory would suggest that the prices put into effect would partially reflect the competitiveness of the particular industry, the shape of the demand curve for the product including the availability of substitutes, etc., the capacity utilization in the industry at the point of operation, and the international trade situation in that particular product.

Economists have done little work in developing consistent models of pricing behavior for U.S. industries. This has resulted from a lack of detailed conceptual models and, more importantly, the general unavailability of appropriate data. There are a series of standard models of firm profit maximizing behavior, but these are simplistic and difficult to apply. Therefore, a number of suggested models of behavior, such as constant mark-ups, administered pricing, and bilateral monopolies, have appeared. But, the lack of data, especially at a disaggregated level, has precluded all but the crudest tests of such models.

Lack of both data and a conceptual framework restricted the Cost of Living Council from making much progress in the analysis of the structure of prices. During the Economic Stabilization Program, it was difficult even to link the authorized price increases to movements in aggregate price statistics. There were four problems in making such a linkage:

First, there was no way of telling what portion of an authorized