148

The foregoing description of current practice relating to credit information, like those of recent viewers-with-alarm, may seem to be an introduction to a call for legislative action of a fundamental kind. It is not. The information in standard credit bureau files (as distinguished from those resulting from "special" investigations is not overly sensitive. Only in relatively few cases are requests for credit reports made by persons other than those who have been asked to give the subjects credit, others similarly situated (such as insurers), or law enforcement officials. A law that required proof of need to know or required permission from the subject for each credit report would be expensive. Credit bureaus provide much of their information to subscribers over the telephone. Even if the crudest sort of control were provided-a call-back system to identify the caller, for example-the number of phone calls would be doubled. Even that control would not prevent a subscriber from passing on the information to someone else. A control system based on "instant consent" of the credit applicant would require elaborate schemes to identify the person who claims to be the subject of the file, such as a card that could be inserted in a telephone, or perhaps a voice-identification system with corresponding losses of privacy for the purposes of identification. Finally, special protections would have to be devised for the interstate information request. At present, inter-bureau exchanges of credit data are routinely given.149 A national enforcement system to give teeth to access controls is most unlikely. Does that imply that there will be a Reno for credit data? One local control over the export of credit data might take the form of an insistence that an out-of-state request for the data be accompanied by written permission of the subject of the file or an identification of the requesting bureau's customer and a demonstration of the customer's need to know. But at what cost are we willing to impose controls of that kinds? Beyond the very considerable money costs imposing effective limits on access to credit information, there are hidden costs of another type, which are hard to estimate. To the extent that Adam Smith's "invisible hand" continues to guide the economy, there is a vital interest in a friction-free flow of information that is relevant to individual economic decision-making. The data in credit bureau files should be readily available to those who find it useful in making such decisions. A poor credit risk who fails to repay his loan burdens the whole economy; if subsidies of that type are to be paid, they should be identified and evaluated as such and not hidden in the price of all consumers' purchases.

150

It is not argued that the idly curious should be encouraged to snoop. One institutional protection that deserves consideration is a requirement of notice to the subject of the file that a credit report has been made arguably excluding reports to the police.' That, too, involves additional expense to the consumer however the cost of the notice is initially absorbed. The notice might be given in the form of a quarterly list of all those who have requested and received reports. In whatever form it is given, a notice of this kind would certainly discourage inquiries for essentially social reasons, as where the parents of the girl check on the credit rating of the parents of the boy."

151

The credit bureaus do not, as a matter of course, send each subject a copy of his own credit report. Access is so easy, however, that anyone who wants to see his own file can do so at low cost. Perhaps that cost to the subject should be eliminated, so that he might be entitled to a copy of his file once in a specified period, such as a year. Perhaps also this free access to the subject should be limited to certain critical occasions, such as the denial of credit. However the subject gains access to his own file, he should have the right to add a statement of moderate length about any derogatory entry that he thinks creates an in

148 See PACKARD, op. cit. supra note 91, ch. 10, for a description of special investigations of an elaborate character by insurance investigators. Credit "specials" are less detailed. 149 BRENTON, op. cit. supra note 91, at 28, says that the ACBA bureaus "provide approximately 7.500,000 interbureau reports annually."

150 In Wisconsin, state income tax and gift tax returns were at one time open to public inspection. Anyone could inspect anyone else's return, but immediate notice was sent to the taxpayer of the name of each person who inspected his return and the reason for the inspection. WIS. LAWS 1951, ch. 714. In 1953, the law was amended to permit access only to the amount of income tax or gift tax paid, with the same provision for disclosure of the identity of the person to whom the information is given and his stated reason for inquiring. Nonresidents are not allowed to receive this information unless their states make similar information available to Wisconsin residents or corporations. Wis. Laws 1953, ch. 303, now in WIS. STAT. § 71.11 (44) (b), (bm) (1963).

151 Such instances are related by BRENTON, op. cit. supra note 91, at 36-37. The problem of interstate exchanges of credit information would still remain, and may be insoluble. One who wishes to know of another's credit standing without disclosing his inquiry can employ enough intermediates to insulate him from that disclosure.

accurate impression if left unexplained. The removal of claimed inaccuracies should not be required; however, once the subject has notified the credit bureau that he considers a derogatory entry to be false, the bureau may have difficulty in establishing its qualified privilege in a later defamation action based on the publication of the disputed entry.12 There are a great many cases which hold that "malice" which overcomes the asserted privilege may be established by proof of a reckless disregard of the likelihood that the derogatory information is false.1 The award of damages in cases of that kind may be a risk that credit bureaus are willing to run in order to serve their customers. Such damages, like other costs of doing business, will surely be paid by consumers in the end, but that is a more attractive result than leaving the loss where it has fallen, on the victim of the defamatory report.

153

Any insitutional suggestions in the area of credit reporting are short-term proposals. Here, as much as anywhere, "the rate of change of technology and society threatens to make footless fantasying of any speculations about the impact of selected factors." For a radical restructuring of the whole consumer credit system seems to have begun. Retailers as a class have already begun to withdraw from the credit business in favor of financial institutions. The credit card is gradually replacing the charge-a-plate and bids fair to replace the personal check.155 Tomorrow's consumer will have one cash-credit account with one financial institution. When he buys a suit, the salesman will insert the customer's cash-credit card in a slot in a telephone, dial the appropriate numbers to represent the amount of the sale, and wait for an approving signal from the bank's computer. The customer's bank account will be reduced-or his loan account charged. The retailer will have loaned no money; for him, every sale will be a cash sale. Thus he need not make a credit investigation.

In fact, no one need make a credit investigation relating to the granting of consumer credit, except for the bank. What kind of investigation will the bank make? When it accepts a customer for an account which authorizes loan-overdrafts ("Instant Money," in the promotional material of at least one bank 156). it will be entering into a relationship which may be a lifelong one. The initial investigation will probably be more thorough than the investigation now made of a typical credit buyer. But soon the bank will have built up its own body of experience with the customer. Its own records will provide his complete financial history, superior to what a credit bureau can now produce at comparable cost. When the customer begins to look like a "no-pay," the bank can tug at the string. If he goes elsewhere in search of credit, the new bank will ask the old one for his record. If the customer moves to another city, his bank-credit record will tag along. It may become necessary to police concerted denials of credit as other concerted refusals to deal are now policed.157 But access to the information will otherwise be more easily controlled. Without question, it will cost more than ten dollars to get a bank-credit file in that era of centralized credit and centralized data processing. At the same time, inaccuracies in the files will amost certainly be reduced. It will be to the bank's advantage to provide its customers with cashcredit statements, to assist in keeping him within his means. There will be little need for evaluative reporting; disputes over the quality of merchandise bought with borrowed money will be none of the bank's concern, and a credit buyer will not be called a "slow pay" because he tried to get his money back from the retailer.

152 In 1965, a bill was introduced in the California legislature to "require" credit bureaus to correct any false report. The sanction for failing, after notice, to disseminate the corrected report to recipients of the false one was to be loss of the qualified privilege (the common law privilege written into statute by CAL. CIV. CODE § 47(3), A.B. 920, 1965 Regular (General) Sess.). The bill was not enacted into law, but much of its substance could be achieved by judicial decision within the existing law of defamation. Cf. Stationers Corp. v. Dun & Bradstreet. Inc., 62 Cal. 2d 412, 42 Cal. Rptr. 449. 398 P.2d 785 (1965) (using the expressing "probable cause" to describe the non-"malicious" state of mind). See also Annot., Libel and slander: report of mercantile agency as privileged, 30 A.L.R.2d 776 (1953).

153 Even negligence is enough to overcome the privilege in some jurisdictions. See Hallen, Character of Belief Necessary for the Conditional Privilege in Defamation, 25 ILL. L. REV. 865 (1931). It should be added that there are some decisions to the effect that it is not libelous to impute a low credit standing to one who is not a merchant. See Annot., Imputing credit unworthiness to nontrader. 99 A.L.R.2d 700 (1965).

154 Michael, supra note 84, at 270.

155 See Check Writing Seen Being Reduced Sharply "In Foreseeable Future," Wall St. Journal, Feb. 10, 1966. p. 4, col. 3.

156 See BLACK, op. cit. supra note 136, at 118-22.

157 E.g., Ruddy Brook Clothes, Inc. v. British & Foreign Marine Ins. Co., 195 F.2d 86 (7th Cir. 1952) (injunction under Clayton Act against blacklisting by fire insurance companies).

CONCLUSION

One need not believe, as I do on my gloomier days, that we are marching steadily to the Anthill in order to share the view that our main concern about "the files" should be directed to their accuracy, not to their accessibility. Any society, whatever its degree of freedom, has a great deal to gain from unimpeded access to information. Any society risks serious losses in the quality of low-level decision-making if its leaders decide before the event that certain classes of information must be kept secret because they will not be relevant to future decisions. In the society of the near future, much more personal information will be relevant to predicting individual performance. Concurrently, much more will be thought to rest on predicting accurately. Whatever we do today to restrict access to personal data, our successors will probably discard. The same cannot be said of efforts to make the files accurate. If there is one sure, positive contribution we can make to those who follow us, whatever social forms they choose, it is to make all our reservoirs of information into truer reflections of the world we see.

FEDERAL DATA CENTERS-PRESENT AND PROPOSED

(By Julian Nixon, Council of Social Science Data Archives)

(This paper has been prepared by the Council of Social Science Data Archives, 605 West 115th Street, New York, New York 10025. It is circulated at present among persons interested in archives and in government data resources; eventually it will be published. We would be grateful if any readers pointed out errors or suggested additional information. Please address the office of the Council.) This paper covers a variety of subjects in an attempt to indicate some of the most recent efforts by the Federal Government to establish new data banks. Included are two accounts of the establishment of relatively new data systems in the Bureau of Labor Statistics and the Federal Bureau of Investigation; and most lengthily, a history to date of efforts to establish a National Data Center.

THE MANPOWER INFORMATION SYSTEM

The Bureau of Labor Statistics of the Department of Labor in Washington under Rudolf C. Mendelssohn, Chief, Data Systems Research and Development Staff, has developed within the last few years a system for the storage, retrieval and analysis of the nation's major collection of employment and unemployment data. This is the Manpower Information System.

It is a production oriented, batch processing, system employing IBM 1460-7074 machines with magnetic tape reels serving as the data storage medium. It will include about 25,000 published monthly time series extending mainly in time to 1947 but in a few cases to 1909. These data are augmented each month by the current findings of the Federal Government's major statistical surveys on employment, unemployment, hours of work, earnings, and labor turnover.

The data comes in the main from two major programs. The first of these is a series of monthly interviews by the Bureau of the Census of a sample of 35,000 households, mainly providing data on occupation, race, marital status, age, employment status, level of skill and other demographic data. Some 2,000 monthly statistical series originate here. The second is monthly historical data for about 350 industries on hours of work, hourly and weekly earnings, number employed, and labor turnover. The figures are extended for about 4,000 series each month. In addition under the technical guidance of BLS, cooperating government agencies in each state working with the Bureau of Employment Security, have been providing comparable labor information for the state and for over 130 metropolitan areas throughout the country, ranging in size from New York City to Casper, Wyoming. Seventeen thousand statewide and area series are thus extended each month.

In addition to the statistical summaries from these sources which constitute the main time series material in the system, there are well over 12 million reports submitted monthly by 170,000 business and industrial establishments since 1957, which are on tape and can be entered. Arranged by establishment, month and industry, they are susceptible to many forms of analysis in relation to the business cycle and other economic and social factors.

A great deal of information on the nonagricultural labor force comes from the annual records produced by the Social Security Administration. These records give age, race, sex, industry, geographic area, and earnings of each worker covered. Since 1957 the Bureau of Labor Statistics has obtained each year from the Social Security Administration material from a one-percent sample of these records. The selection is done on the basis of certain combinations of digits in the individual social security number, and since the same combinations are picked each year it is possible to follow the activities of the individuals in the sample from year to year.

The Manpower Information System was made operational during the first half of 1965 and is open to the use of qualified researchers. Search statements for the system are fairly simple and manuals for its use are available from the Bureau of Labor Statistics in Washington.

THE NATIONAL CRIME INFORMATION CENTER

In September, 1965, the Federal Bureau of Investigation in Washington embarked on the development of a national electronic information system to be known as the National Crime Information Center. It is visualized ultimately as an information network encompassing the entire United States which will make available to each law enforcement agency in a matter of seconds, the facilities of an information file national in scope.

The National system which will be tied together and coordinated by the F.B.I. will contain not only material from F.B.I. files but be built up by linking local metropolitan systems to statewide systems and thence to a national system. Prototypes of such local systems now exist in many areas, for example, in New York City and St. Louis, New York State and California.

The F.B.I. has contracted with the Institute for Telecommunications and Aeronomy, Environmental Science Services Administration, Department of Commerce, to survey all existing telecommunications networks throughout the United States and to recommend a network that would best support a nationwide computerized system. The study was due for completion by July 1, 1966 and was to cover the following points: desirable computer characteristics, existing information systems, existing and planned telecommunication networks, data transmission requirements, and the development of telecommunications operations to fulfill National Crime Information Center requirements.

The information to be stored at each level will depend on actual need, with local systems having a data base much broader than either the statewide or national system. Initially the National Center will include information on stolen automobiles, certain classes of stolen property, and some wanted persons.

The highspeed computers which are the heart of the national system are located at the F.B.I. in Washington. The Center is expected to be in operation with several local systems tied into it by January, 1967. Until local systems are developed sufficiently to be tied into the network, state centers may utilize the national system through remote terminals. Security of information in the file will be assured by a number of means both in the equipment and its programs. Communications switching procedures will eventually permit each state and/or metropolitan area to exchange with one another. Other Federal investigative agencies and F.B.I. field offices will also be in direct contact with the national system.

The development of standard codes and formats has been undertaken with the assistance of an advisory committee of the International Association of Chiefs of Police. This work was begun in March, 1966 and initial recommendations were to have been submitted by May, 1966.

With the implementation of this system, the F.B.I. believes that "instant information" will become a reality and the officer on the street and the investigator will get "real time" information on current situations.

THE RUGGLES REPORT

In 1960 the American Economic Association, recognizing the fact that economics was coming to depend more and more on large systematic collections of microdata, and being aware of the large bodies of empirical data available within the Federal government was concerned with the problem of its preservation and its availability to research workers. It approached the Social Science Research Council about setting up a Committee on the Preservation and Use of Economic

Data. This the SSRC did, drawing on social scientists from universities and the Federal Government.

During the three years from 1962 through 1964, the Committee, under the chairmanship of Richard Ruggles of Yale University, undertook to study on an agency by agency basis the problem of providing access to specific bodies of information. Meetings were held with a considerable number of independent agencies in the Departments of Commerce, Labor, Treasury, Agriculture, Interior, and Health, Education, and Welfare. In addition there was close contact with the Bureau of the Budget and the National Archives. At the end of this period it was conIcluded that some more general solution was needed and that this would require that the Federal Government develop a systematic policy for ensuring the preser vation of important data and mechanisms whereby data could be made available to universities and research institutions.

In its survey of twenty Federal agencies the Ruggles Committee, aided by Rudolf Mendelssohn of the Bureau of Labor Statistics, found over 600 major bodies of data stored on 100 million punch cards and 30,000 computer tapes. The decentralized nature of the Federal statistical system makes it extremely difficult to know what data exists on various topics and how to obtain access to it. Different agencies have entirely different policies with respect to access and outsiders must know exactly whom to contact for each kind of information. Most agencys other than the Census Bureau collect and process as a part of their primary administrative or regulatory mission and find it inconvenient and costly to respond to requests for information which might disrupt or delay their own work. In addition to the problems of location and access another difficulty within the Federal statistical system was uncovered. Federal agencies often fail to develop clean edited tapes and to provide supporting information on the data contained in the tapes.

For these and other reasons, the Committee on the Preservation and Use of Economic Data in April, 1965 recommended that "a Federal Data Center be established by the Federal Government to preserve and make available to both Federal agencies and non-governmental users basic statistical data originating in all Federal agencies."

The Center would be given authority to obtain computer tapes produced by other Federal agencies. It would have the duty of following statistical projects underway in the Federal Government and seeing that budget and other provisions were made which would result in a cleaned, edited, and well supported tape. The proposed Center would require substantial computer capability if it were to furnish basic information, and on a reimbursable basis make production runs and furnish aggregated tapes to scholars so as to eliminate problems of disclosure. It should be staffed with computer analysts who are subject specialists and can understand the nature of the data with which they work.

The recommended Center would provide servicing facilities so that Federal agencies and individuals could obtain specific information directly and it would publish descriptions of the data available. In this sense the Federal Data Center would serve in somewhat the same role as the Library of Congress in providing a systematic and comprehensive coverage of the material in its area of competence and the same role as the National Archives does for basic records and documents, in the area of machine readable data.

Finally, it was recommended that since the Bureau of the Budget has been given the responsibility for developing programs and issuing regulations and orders for the improved gathering, compiling, analyzing, publishing, and disseminating of statistical information for any purpose by the various agencies in the Executive Branch of the Federal Government, it should immediately take steps to establish a Federal Data Center. There was no recommendation as to the agency under which such a Center should ultimately operate.

In addition to its specific recommendations concerning the Center, the Ruggles Committee also urged "that the Office of Statistical Standards of the Bureau of the Budget place increased emphasis on the systematic preservation in usable form of important data prepared by those agencies engaged in statistical programs."

THE DUNN REPORT

The Ruggles Committee made its report in April, 1965 to the Social Science Research Council which presented it to the Director of the Bureau of the Budget. In May, 1965 Edgar S. Dunn, Jr. of Resources for the Future, Inc., a consultant to the Bureau of the Budget, was asked to examine the proposal and to study