1971]

PRIVACY AND THE COMPUTER

839

occasionally punitive or benevolent. Since Congress in recent years has investigated the operation of commercial agencies,1 we are most knowledgeable about them. These commercial agencies fall into two categories: the credit bureau and the investigatory reporting agency.

A. The Commercial Compilers

In a very rough way, the credit bureau is to the user of consumer credit what Dun & Bradstreet's reporting service is to a business organization seeking credit. The credit bureau is not a new institution in our society, but as consumer credit has burgeoned by more than 2,000% in the past quarter century2 so has the business of the credit bureau.

There are approximately 2,500 credit bureaus in the country, of which some 2,100 are members of one trade association, Associated Credit Bureaus, Inc. The files of bureaus affiliated with ACB include records on approximately 100 million persons, and those bureaus interchange their information. The largest credit bureau operation outside the ACB is the Credit Data Corporation, which has files on 27 million persons. While there is doubtless some overlap between the 100 million ACB files and the 27 million Credit Data Corporation files, the combined accumulation probably covers nearly all 131 million of us who are over eighteen years of age, particularly since most of the

1 There have been five separate Congressional hearings: Hearings on Commercial Credit Bureaus Before the Spec. Subcomm. on Invasion of Privacy of the House Comm. on Gov't Operations, 90th Cong., 2d Sess. (March 1968) [hereinafter cited as House Credit Bureau Hearings ; Hearings on Retail Credit Co. of Atlanta Before the Spec. Subcomm. on Ipwasion of Privacy of the House Comm. on Gov't Operations, 90th Cong., 2d Sess. (May 1968) [hereinafter cited as House Credit Bureau Hearings II]; Hearings on S. Res. 233 Before the Subcomm. on Antitrust and Monopoly of the Senate Judiciary Comm., 90th Cong., 2d Sess. (December 1968) [hereinafter cited as Senate Credit Bureau Hearings I]; Hearings on S. 823 Before the Subcomm. on Financial Institutions of the Senate Banking and Currency Comm., 91st Cong., 1st Sess. (May 1969) [hereinafter cited as Senate Credit Bureau Hearings II]; Hearings on H.R. 16340 Before the Subcomm. on Consumer Affairs of the House Comm. on Banking and Currency, 91st Cong., 2d Sess. (March-April 1970) [hereinafter cited as House Credit Bureau Hearings III.

2 From $5.7 billion at the end of 1945 to $123.9 billion at the end of November 1970. FED. RESERVE BULL. A-54 (Jan. 1971).

3 ACB is also operating under a 1933 antitrust consent decree that requires the interchange of data with the 4000 credit bureaus not affiliated with it. See United States v. National Retail Credit Ass'n, Eq. No. 10420 (E.D. Mont., Oct. 6, 1933) (abstracted in AM. ENTERPRISE INSTITUTE FOR PUB. POLICY Research, ANTITRUST CONSENT DECREES 1906-1966, at 371 (1968)); Senate Credit Bureau Hearings 1, supra note 1, at 23. In 1968 ACB entered into arrangements with International Telephone & Telegraph Corporation to provide a computer service to facilitate the interchange. Id. at 24, 33-34.

4 In November 1969 all of the stock of Credit Data Corp. was acquired by TRW Credit Information Services, Inc., which now operates the Credit reporting service as TRW Credit Data. House Credit Bureau Hearings III, supra note 1, at 156. This bureau is completely computerized and is adding files at the rate of one-half million per month. House Credit Bureau Hearings III, supra note 1, at 157, 165; Senate Credit Bureau Hearings 11, supra note 1, at 223, 227.

840

TEXAS LAW REVIEW

[Vol. 49:837 93 million of us who are married will be combined in some 46 million files with our spouses.

5

The content and the reliability of information in credit bureau files is dictated largely by the three principal sources from which the credit bureaus draw. A principal source of information is the bureau's own subscribers, the merchants, banks and finance companies who buy most of the credit reports. They supply information on their own customers with regard to employment, approximate income and credit performance. The value of this source is subject to three significant limitations. First, since the credit bureau files do not reveal the subject's net worth, or even his state of solvency, but only whether his accounts with the bureau's subscribers are delinquent, the extension of credit in reliance on a credit bureau report proceeds on the illogical assumption that the ability to meet current credit payments betokens the capacity to handle one additional debt. Second, credit bureau files cannot even approximate the amount of a subject's debts, since many creditors are not subscribers. Third, the delinquency of an account frequently results not from financial irresponsibility but from a bona fide dispute over the amount owing or the quality of merchandise sold.

As a second source of information, the more enterprising bureaus scrutinize official records for arrests, lawsuits, judgments, bankruptcies, mortgages, tax liens, marriages, divorces, births and deaths. The substantial possibility of mistaken identity, combined with the frequent failure of official records to include final dispositions, renders this source highly unreliable. As a third source of information, most credit bureaus also maintain a news-clipping service, although this occasionally serves as a substitute for checking official records. Obviously, this source poses even greater danger of error than the official record check with regard to mistaken identity or the failure to note the ultimate disposition of arrests, lawsuits, judgments and liens.

During Congressional hearings on the subject, much attention was focused on the denial of credit due to erroneous adverse information in credit bureau files. Unstressed but equally serious is the erroneous extension of credit. Just as mistaken identity leads to the insertion of data in an inappropriate file, so also its leads to the omission of relevant information from the appropriate file. And few, if any, files will record all debts. Inaccurate files thus mean that credit is both erroneously denied and granted. In their own defense, credit bureaus

5 U.S. BUREAU OF the Census, Statistical Abstract of the United States 25, 32 (1970).

1971]

PRIVACY AND THE COMPUTER

841

stress that they only collate information and that they never claim to make affirmative investigations of their subjects.

It is no coincidence that, as consumer credit expanded by more than 2,000% in the past twenty-five years, so did consumer bankruptcies-from 8,500 in 1946 to 178,000 in 1970." If a credit extender were to compare the report he received from the credit bureau with the debts scheduled by a subject in his bankruptcy proceeding, he might conclude that the credit report was not worth the 35 to 75 cents paid for it. Upon entries of such fragile reliability is your “credit rating" built. And when the credit bureaus also engage in debt collection-and many of them do, finding their ability to affect the credit rating an effective collection tool— the reliability of the entries is even further threatened by a built-in conflict of interest.

Credit bureau files are thus inadequate to satisfy some who contemplate commercial relationships. In consequence, prospective employers and insurers often turn for information to the investigatory reporting agency.10 Congressional committees heard from representatives of the largest of such agencies in this country-Retail Credit Company of Atlanta, with 1,225 offices, 7,000 inspectors, and files on 48 million persons." Inspectors from Retail Credit not only check

6 On occasion, however, bureaus join with local merchants in sponsoring the Welcome Wagon lady who reports back to the merchants on the apparent worldly needs of the newcomers she visits and to the credit bureau on their apparent credit worthiness-and on where the newcomer came from, so that his file can be obtained from a credit bureau at his former location. Senate Credit Bureau Hearings II, supra note 1, at 119, 167-172; House Credit Bureau Hearings III, supra note 1, at 33.

7 AD. OFFICE OF UNITED STATES COURTS, TABLES OF BANKRUPCY STATISTICS, TABLE F-3 (1946); AD. OFFICE OF United States Courts, Ann. Rep. of the Director, V-4 (1970).

8 This is the price the subscriber pays to learn what reposes in the compiler's file at the moment he makes inquiry. If he wants the file brought up to date by calls to other subscribers he must pay an additional fee. Save for routine checking of public records and newspaper clippings, the files are not updated until a subscriber requests, and pays for, that action.

In hearings held in Washington, D.C. in March, 1968, a New York Congressman asked for a demonstration of Credit Data Corporation's high-speed computerized retrieval of his New York City credit file. Within the time consumed by six pages of printed hearing record, the report came back on one bank loan as of June, 1967, and nothing else. The Congressman's response: "A very inefficient system, thank God." House Credit Bureau Hearings I, supra note 1, at 72-79.

9 Senate Credit Bureau Hearings I, supra note 1, at 5, 23, 28-30, 40-43, 45, 53, 78; Senate Credit Bureau Hearings II, supra note 1, at 180.

10 Two of the subcommittees also heard testimony from representatives of one agency specializing in reports on prospective mortgagors to banks and institutional lenders. Their testimony was to the effect that they confined their investigation to public records and newspaper clippings, did not engage in "questioning neighbors," and always gave the subject an opportunity to explain adverse items of public record or news report-unless the client asked them not to contact the subject or unless the client asked for a quick, telephone report on what the files contained. Senate Credit Bureau Hearings II, supra note 1, at 237-254; House Credit Bureau Hearings III, supra note 1, at 211-230.

11 Senate Credit Bureau Hearings II, supra note 1, at 175; House Credit Bureau Hearings III, supra note 1, at 473-474.

842

TEXAS LAW REVIEW

[Vol. 49:837 public records and clip newspapers, but also interview friends, neighbors, former neighbors, acquaintances, employers, former employers, business associates-anyone who may know something, or have an opinion, about the subject.12 For life insurance companies Retail Credit inspectors inquire about the subject's drinking habits and domestic difficulties, seek out criticism of his character and morals, and report whether his living conditions are crowded or dirty.13 For automobile insurers Retail Credit investigates the subject's neighborhood business reputation, morals, and "antagonistic-antisocial conduct."'14 For employers, Retail Credit's inspectors also report whether the subject has any "known connection with a 'peace movement' or any other organization of a subversive type" and whether he is reputed to be "neurotic or psychotic."'15

13

In response to Congressional concern about the reliability of agency reports, spokesmen for Retail Credit proffered two assurances. First, its inspectors are carefully trained persons of “unusual inspection ability." This assurance lost some of its force when inquiry revealed that these highly qualified, well-trained inspectors commanded a starting salary of $475 to $500 per month, that they prepared anywhere from 2 to 16 reports per day, that half had no more than a high school education, and that another 30% were college dropouts.17 Second, Retail Credit reported that adverse information not coming from public records either is confirmed from a second source or is reported as unconfirmed.18 Whatever comfort otherwise might be drawn from this assurance is somewhat qualified by evidence to the contrary, 19

12 House Credit Bureau Hearings II, supra note 1, at 21; Senate Credit Bureau Hearings I, supra note 1, at 108.

13 Senate Credit Bureau Hearings II, supra note 1, at 196-203.

14 The auto insurers are convinced that there is some correlation between all of these factors (except antagonistic-antisocial conduct) and accident frequency, and that both immorality and antagonistic-antisocial conduct would impair the subject's effectiveness as a witness in the event of litigation. Senate Credit Bureau Hearings II, supra note 1, at 207212, 305-316; House Credit Bureau Hearings III, supra note 1, at 514-516. The latter consideration, of course, should dictate an inquiry into hare-lips, unsightly scars and birthmarks, and the use of deodorants.

15 Senate Credit Bureau Hearings I, supra note 1, at 109-110; Senate Credit Bureau Hearings 11, supra note 1, at 176-177.

10 House Credit Bureau Hearings II, supra note 1, at 19-20; see also Senate Credit Bureau Hearings II, supra note 1, at 175-176.

17 Senate Credit Bureau Hearings II, supra note 1, at 176, 192; House Credit Bureau Hearings III, supra note 1, at 473, 503.

18 House Credit Bureau Hearings II, supra note 1, at 21; House Credit Bureau Hearings III, supra note 1 at 474.

19 One well-trained, highly-qualified inspector, who heard from two sources that a subject had served in prison, reported the prison record as an unqualified fact although the inspector found no confirmation in court or prison records. Wagner v. Retail Credit Co., 338 F.2d 598 (7th Cir. 1964).

1971]

PRIVACY AND THE COMPUTER

843

Access to the files of these commercial compilers is reportedly restricted to "reputable" business organizations that possess a "legitimate" business interest.20 However, Congressional and private inquiry has undercut this assertion and ably demonstrated the easy availability of the files, not only to private parties21 but also to government agencies.22

20 House Credit Bureau Hearings I, supra note 1, at 63-64, 90; House Credit Bureau Hearings II, supra note 1, at 19; Senate Credit Bureau Hearings 1, supra note 1, at 62.

21 Credit bureau spokesmen have admitted the possibility of an employee of a credit bureau subscriber obtaining a report for purposes unrelated to his employer's business. House Credit Bureau Hearings 1, supra note 1, at 111; Senate Credit Bureau Hearings II, supra note 1, at 228. Retail Credit's spokesmen admitted that it sometimes gave out reports as a "favor" when, for example, an executive of a subscriber asked for a report on a man being considered as a new minister for his church. House Credit Bureau Hearings II, supra note 1, at 19, 38; Senate Credit Bureau Hearings 1, supra note 1, at 101.

Moreover, the compilers had been under interrogation by Congressional committees for over a year when CBS News conducted an experiment. Using a fictitious company name, it sent out twenty letters to credit bureaus requesting reports on named individuals. It received ten reports and offers of two more if it would sign a subscriber's contract. On a second effort, the fictitious company sent out twenty-eight letters. This time it did not state that it was considering granting credit-it simply asked for a full credit report, without stating why it wanted it. And this time it asked only about individuals who had been complaining to Congressional committees about the credit bureaus. This time it received only seven of the requested reports-plus one more when it accepted the offer of one credit bureau to sign a subscriber's contract. Senate Credit Bureau Hearings II, supra note 1, at 378-381; House Credit Bureau Hearings III, supra note 1, at 59-61.

22 This includes not only such governmental credit-granting agencies as the Federal Housing Administration and the Veterans Administration, which buy such reports just as do private subscribers. It includes also such law enforcement agencies as the FBI and the Internal Revenue Service. Members of ACB and the Retail Credit Company make their files available to the law-enforcers "as a public service." House Credit Bureau Hearings 1, supra note 1, at 134-138; House Credit Bureau Hearings II, supra note 1, at 23, 39; Senate Credit Bureau Hearings II, supra note 1, at 149, 161; House Credit Bureau Hearings III, supra note 1, at 603, 605. The Credit Data Corporation took a different view, and refused to turn over its reports to the IRS. House Credit Bureau Hearings 1, supra note 1, at 90-91; Senate Credit Bureau Hearings I, supra note 1, at 62-63. It was then met with a statutory summons calling for all credit information relative to named taxpayers. The IRS is authorized, for the purposes of "determining the liability of any person for any internal revenue tax" to summon the taxpayer "or any other person the [IRS] may deem proper," to produce "such books, papers, records, or other data.. as may be relevant or material to such inquiry...." INT. REV. CODE OF 1954, § 7602. When Credit Data refused to obey the summons, IRS got a judicial order of enforcement pursuant to the statute, requiring it to comply on payment by the IRS of 75 cents per report, the fee which Credit Data charged its regular subscribers. On appeal, Credit Data won a great victory. The decision below was affirmed in all respects save that the case was remanded to determine the "fair value" which IRS must pay for the reports the rate paid by subscribers was not to be taken as conclusive because subscribers supply "valuable credit information" to Credit Data. United States v. Davey, 426 F.2d 842, 844 (2d Cir. 1970).

The result was foreshadowed by previous decisions of the Supreme Court. In a long line of cases that court has sustained judicial enforcement of an administrative agency's statutory subpoenas against fourth amendment attack where the subpoena sought testimony about the affairs or the records of the person subpoenaed, if the subpoena was sufficiently specific to satisfy the fourth amendment, if the administrative inquiry was authorized by Congress, and if the evidence sought was relevant to the inquiry-the court's application of the last two requirements when its enforcement order was sought being held to satisfy the fourth amendment's requirement of probable cause. United States v. Morton Salt Co., 338 U.S. 632 (1950); Oklahoma Press Publishing Co. v. Walling, 327 U.S. 186 (1946). Earlier decisions are discussed in 1 K. DAVIS, ADMINISTRATIVE LAW TREATISE § 3.12 (1958) and Comment, Agency Access to Credit Bureau Files: Federal Invasion of Privacy? 12 B.C. IND. & COM. L. REV. 110, 113-118 (1970). More than forty-five years ago the Supreme Court also