The Department of Justice, supported by an amicus curiae brief filed by SEARCH Group, Inc., has petitioned for rehearing. The government argues, inter alia, that the panel opinion misinterpreted the term "public interest," as used in the balancing of interests in Exemptions 6 and 7(C) of the Freedom of Information Act ("FOIA"). 5 U.S.C. Sec. 552(b)(6), (7)(C) (1982). The charge is that the panel incorrectly deferred to state or local government determinations that arrest or conviction records should be publicly available. Although we deny the petition for rehearing and reaffirm our prior judgment, we think the government has a valid point and therefore modify our rationale.

Our opinion rejected the district court's reasoning--that there was little or no public interest in Charles Medico's rap sheet because the records sought related to minor crimes that occurred long ago. Reporters Comm. for Freedom of the Press v. United States Dep't of Justice, 816 F.2d 730, 740-42 (D.C.Cir.1987). We concluded that there was no principled statutory basis to support that determination, and we said that the district court should have deferred to state or local determinations that publication of arrest or conviction records were in the public interest. Id. at 740-41. It is argued in the petition for rehearing, however, that such an approach could prove confusing and indeed unworkable since the district court may well encounter conflicting policies on disclosure of arrest records at the state and local level. SEARCH Group, an association of governors' appointees responsible for the operation of the agencies in their states that collect and maintain criminal history records, has brought to this court's attention the fact that many states have policies or laws that forbid the release of their own compiled law enforcement information, which includes rap sheets. Based in part on the amicus' presentation, we now agree that our prior position on this point should be abandoned. We must thus confront two questions that we previously thought appropriate to avoid: How do we determine, as a matter of law, the public interest in disclosure of the information that appellants seek? Does FOIA require the judiciary to make an individual determination of the general public interest in information sought in every case in which a section 6 or 7(C) Exemption is asserted? Id. at 740-42. In answering these questions, we find no standards or guidelines drawn from the statute to inform our determination. Prior cases of this circuit have purported to appraise and value the public interest in specific information sought,1 but in no case has this court ever articulated standards or a rationale for that process.

The government argues that the statute has a "core" purpose, i.e., "to ensure an informed citizenry, vital to the functioning of a democratic society, needed to check against corruption and to hold the governors accountable to the governed." NLRB v. Robbins Tire & Rubber Co., 437 U.S. 214, 242, 98 S. Ct. 2311, 2327, 57 L. Ed. 2d 159 (1978). And although the statute reaches beyond the core, the government argues, the courts should treat disclosure that serves core policies more favorably than those disclosures that do not. Over a decade ago, Judge Leventhal, speaking for this court, expressed doubt that the public interest considerations of the Act could be so limited. Ditlow v. Shultz, 517 F.2d 166, 172 & n. 24 (D.C.Cir.1975). It seems to us that he was quite correct. A core purpose does not, in our view, confer judicial power to predict whether particular information the government holds will, upon disclosure, aid an "informed citizenry" as to democratic political choices. Indeed, the government is utterly incapable of explaining to us why the information sought here does not serve the Act's "core" policy. The government and the panel concurring opinion argue that any convictions involved here are too old to require disclosure, Reporters Comm., 816 F.2d at 745, but from what principle does that observation flow? Surely it cannot be seriously argued that as a matter of law an "informed citizenry" should have available to it arrest records two years old but not five or ten. The subjects of appellants' requests are alleged to have had dealings with government officials; it is surely up to the citizenry, once informed, to determine the relevance of the age of the arrests or convictions.

Even if we could fashion a methodology--and we firmly believe we cannot--to grade the public interest in government-held information in the abstract, we must keep in mind that we are unable to foresee or monitor how the information will eventually be used. Ditlow, 517 F.2d at 171 & n. 18. As we have held, information disclosed to anyone must be disclosed to everyone. Durns v. Bureau of Prisons, 804 F.2d 701, 706 (D.C.Cir.1986); Ditlow, 517 F.2d at 171 & n. 18. Therefore, a particular requester's purpose in seeking information, or his proposed use, must be wholly irrelevant to a determination of the public interest since we cannot predict how other persons, including those to whom the requester might give the information, would use it.2

But, it might be asked, if it is impossible to judge the public interest in requested information by its proposed use or its inherent value to an informed citizenry, how then can it be judged at all? We do not think it can. We do not believe that the phrase "public interest," as used in the balancing in Exemptions 6 and 7(C) of the Act, means anything more or less than the general disclosure policies of the statute.3 Since Congress gave us no standards against which to judge the public interest in disclosure, we do not believe Congress intended the federal judiciary--when applying only Exemptions 6 and 7(C) of the Act--to construct its own hierarchy of the public interest in disclosure of particular information. As we said in our panel opinion, Reporters Comm., 816 F.2d at 741, such an unbounded delegation would raise serious constitutional problems. Cf. Illinois v. United States, 460 U.S. 1001, 1004-06, 103 S. Ct. 1240, 1242-43, 75 L. Ed. 2d 472 (1983) (Rehnquist, J., dissenting); Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 646, 100 S. Ct. 2844, 2866, 65 L. Ed. 2d 1010 (1980) (plurality opinion). It is true that the Supreme Court (and this court) has held that the federal courts in applying Exemptions 6 and 7(C) must weigh the public interest against the privacy interest and determine whether disclosure would be an unwarranted invasion of the privacy interest. Department of Air Force v. Rose, 425 U.S. 352, 372, 96 S. Ct. 1592, 1604, 48 L. Ed. 2d 11 (1976) (Exemption 6); FBI v. Abramson, 456 U.S. 615, 631, 102 S. Ct. 2054, 2064, 72 L. Ed. 2d 376 (1982) (Exemption 7(C)). That of course means that the federal courts must, as we have done in this case, consider whether there is a cognizable privacy interest in the information sought, and then appraise the impairment to that interest that would result from disclosure. For instance, had it been shown to us that disclosure of Medico's "rap sheet" would cause him particular harm, even though his privacy interest is slight, we might well have reached a different conclusion. But, that we must balance the prospective damage to the privacy interest against the public's interest does not necessarily mean, and as we conclude, could not mean, that the public interest depends on our own appraisal of the public's need to know particular information.

The petition for rehearing sharpens another argument the government previously presented: The Justice Department compilation of criminal histories, by putting the public record information in different form, somehow changes the nature of the information sought. Our dissenting colleague now agrees with this argument. Dissent at 1128. But it seems to us this argument is undermined by the government's contention, with which the dissent also agrees, dissent at 1129-30, that if the Department refers a requester to the original source it provides the requester with essentially the basic information sought--at least the fact that criminal history information exists. The requester can then go to the original source and, if the information is publicly available, easily obtain it.4 That, of course, is true, but if such referral is tantamount to disclosure, it is abundantly clear that the form of the government's compilation of criminal histories is immaterial to the issue of whether disclosure of the facts compiled is required. Cf. Abramson, 456 U.S. at 624, 102 S.Ct. at 2060 (holding that information originally gathered for law enforcement purposes by the FBI did not lose its status under Exemption 7 because it was placed in a different compilation for a political purpose by the White House).

The government and the dissent also claim that our analysis of Charles Medico's privacy interest implies a delegation to a state or local body for a policy determination. Dissent at 1128. Not so. We direct the district court only to make a factual determination in these kinds of cases: Has a legitimate privacy interest of the subject in his rap sheets faded because they appear on the public record? Reporters Comm., 816 F.2d at 740. That a particular state may treat its own compilations of criminal histories drawn from public record information on file in cities and counties as confidential is irrelevant to that determination. The district court's task is limited to determining, as a matter of fact, not law, whether, by reason of the actual practices of the jurisdiction that is the original source, the subject's privacy interest has faded. It may well be, as both the government and dissent suggest, that state statutes that limit disclosure of compilations of criminal histories are eminently sensible and should guide federal policy. But as we said in our panel opinion:

Congress could of course decide the public ought not have access to federal compilations of public criminal records. Congress has considered such a possibility in the past, and might do so again. See, e.g., Dissemination of Criminal Justice Information: Hearings on H.R. 188, H.R. 9783, H.R. 12574, and H.R. 12575 Before the Subcomm. on Civil Rights and Constitutional Rights of the House Comm. on the Judiciary, 93d Cong., 2d Sess. (1973).

Reporters Comm., 816 F.2d at 741 n. 15. The point, of course, is that Congress, unlike a number of states, has chosen not to pass a statute that would establish the federal policy that appellees now urge us to announce without a legislative mandate. We decline their suggestion because we do not believe we have the authority to accept it.

Our conclusion in this case therefore does not change. We reverse and remand to the district court to determine whether the Department of Justice holds criminal record information on Charles Medico that, in accordance with our opinion, must be disclosed.

I respectfully dissent from denial of the petition for rehearing. After reflecting on the filings, I have come to the conclusion that the panel's opinion is wrong; worse still, it will likely lead to unfortunate consequences, misconstruing as it does FOIA Exemption 7(C) and, by analogy, Exemption 6 as well.

The panel opinion speaks for itself, and its rationale therefore need not be reviewed in detail. Suffice it to say that the panel majority, finding the balancing process ordained by Congress to be too "awkward," embarked on a quest for someone to defer to in making determinations under Exemption 7(C).1 Discerning in this case a determination by political bodies that the criminal-record information sought here was to be "public," the panel accordingly found any privacy interest eviscerated; at the same time, it found a significant public interest in disclosure by virtue of these same political determinations. As my concurring opinion in the case sought to point out, this approach reflected no de novo balancing at all;2 it was, rather, a single-factor test.

I nonetheless concurred in the judgment and indicated my general agreement with the majority's privacy-interest analysis, limiting my criticisms to its public-interest analysis. On the latter score, I wholeheartedly disagreed with my colleagues. See Reporters Committee, 816 F.2d at 743 (concurring opinion). Now, the filings on rehearing, including a helpful amicus brief surveying state legislation regarding the confidentiality of criminal histories,3 have led me to conclude that the majority's entire analysis in section II, 816 F.2d at 737-43, is flawed.

First. In its original presentation to the panel, the Government argued that while the underlying information was indeed "public" in the sense that it was on record at local courthouses or police stations, the particular items sought by appellants are of a different character by virtue of their cumulative, indexed, computerized nature. The panel opinion rejected this argument, see Reporters Committee, 816 F.2d at 739, concluding instead that the public availability at the "original source" was the only relevant fact. See id. at 743. After reviewing the rehearing petition, I am now of the view that our conclusion was wrong.

As I see it, computerized data banks of the sort involved here present issues considerably more difficult than, and certainly very different from, a case involving the source records themselves. This conclusion is buttressed by what I now know to be the host of state laws requiring that cumulative, indexed criminal history information be kept confidential, as well as by general Congressional indications of concern about the privacy implications of computerized data banks. See H.R.Rep. No. 1416, 93d Cong., 2d Sess. 3, 6-9 (1974), reprinted in Legislative History of the Privacy Act of 1974, Source Book on Privacy, 296, 299-302 (1974).

Second. With respect to the majority's public-interest analysis, I need not restate the objections already voiced in my concurrence. Suffice it to say that the majority fails to carry out its obligation to make a separate valuation of that interest under the precedents of this circuit and the Supreme Court.

The majority vigorously objects to an analysis of the public interest that depends upon the identity of the requester or the use to which the information will be put. Even granting the majority's position that calls into question a rigorous application of our court's analysis in Getman v. NLRB, 450 F.2d 670 (D.C.Cir.1971), this still does not leave the Congressionally mandated public-interest standard devoid of meaning. This is scarcely the place to speculate about the public-interest analysis in a post-Getman era, but rather than give Getman no burial, I think a few words should be said out of respect.

Getman's critical insight is that there is meaning in the public-interest standard; the way in which meaning is imparted to that term will depend on the information that is sought and the circumstances in each case. Where what is at stake is the disclosure of personal information about a particular subject, one helpful vantage point is the public interest in the subject of the information request. Although there may be no public interest in disclosure of the FBI rap sheet of one's otherwise inconspicuously anonymous next-door neighbor, there may be a significant public interest--one that overcomes the substantial privacy interest at stake--in the rap sheet of a public figure or an official holding high government office. For guidance in fleshing out that analysis, it seems sensible to me to draw upon the substantial body of defamation law dealing with "public personages." See e.g., Providence Journal Co. v. F.B.I., 460 F. Supp. 778, 790-91 (D.R.I.1978) (striking differing balances in an Exemption 7(C) case depending upon whether the information obtained through an illegal wiretap concerned the subject's dealings with public figures), rev'd, 602 F.2d 1010 (1st Cir.1979) (holding all information obtained in this manner exempt from disclosure), cert. denied, 444 U.S. 1071, 100 S. Ct. 1015, 62 L. Ed. 2d 752 (1980). But however fruitful that analytical mode may (or may not) eventually prove to be in the post-Getman era, the fundamental point is that the public interest in any particular case can vary beyond the "general disclosure policies of the statute," Majority op. supra at 1126, and is to be seriously weighed against the subject's privacy interest.4

Third. The Bureau and the District Court will have considerable difficulty in applying the majority's single-factor test in future cases. The rehearing filings highlight two problems.

In the first place, the Government points out that the FBI will often have no way of knowing whether a particular item in a particular criminal history compilation is, in fact, "publicly available at the original source," Reporters Committee, 816 F.2d at 743, without a specific inquiry to the source at the time of the FOIA request. To require the Bureau to undertake such inquiries is to impose a substantial (if indeed not crippling) administrative burden. The majority's response to this unhappy, practical result of our holding is that "the Department might satisfy its obligations under FOIA ... by merely referring appellants to the law enforcement agency that provided the information to the Department." Id. (footnote omitted). In its rehearing petition, the Government suggests that this "option" is illusory; to refer a requester in the manner suggested by the majority would only confirm that criminal history information exists, which may be the most pertinent (if not the only) "fact" that a requester is seeking. Government's Rehearing Petition at 10 n. 9.

In the second place, the rehearing filings make it abundantly clear that an entirely different level of "political determinations," Reporters Committee, 816 F.2d at 741, exists. That is, most state legislatures have expressly provided that cumulative, indexed criminal history information of the type sought here is to be held confidential. See, e.g., Cal.Penal Code Sec. 11105 (West Supp.1986); Conn.Gen.Stat.Ann. Secs. 54-142i(c), 54-142n (West 1985); Hawaii Rev.Stat. Sec. 846-9 (1985); Tenn.Code Ann. Secs. 40-15-106(b)-(c) (1982). The exact approach and precise degree of confidentiality, not surprisingly, varies significantly from State to State.5 But the common theme is nonetheless one of confidentiality.

Moreover, the majority provides no guidance to the Bureau or the District Court as to the source of law for determining whether the information is "publicly available at the original source." Often the state legislature requires both that the underlying criminal history source records be publicly available and that cumulative, indexed criminal history information be kept confidential. Compare Ohio Rev.Code Ann. Sec. 1901.31(E) (Anderson Supp.1986) (court proceedings shall be public records) with Ohio Rev.Code Ann. Sec. 109.57(E), (D) (Anderson 1984 & Supp.1986) (criminal history information to be held confidential). In the latter situation--which would appear to be the rule rather than the exception--I fear that the District Court will be left on remand with no criteria for determining whether the privacy interest of the subject in his rap sheet has "faded." Majority op. supra at 1127. Instead the District Court will be left with only the majority's original direction to look to availability at the "original source," as well as the majority's initial rejection of the argument that cumulative, indexed, computerized information was somehow different.

Thus it is that I must confess that I was wrong the first time around. The rehearing filings have convinced me that the panel opinion's entire Exemption 7(C) analysis is in error. Its transmogrification of de novo balancing into a deference-driven, single-factor test does violence to what had heretofore been settled law. In the process, it raises the unsettling prospect of future FOIA requests tailored precisely to fall within the dictates of Reporters Committee. We are now informed that many federal agencies collect items of information on individuals that are ostensibly matters of public record. For example, Veterans Administration and Social Security records include birth certificates, marriage licenses, and divorce decrees (which may recite findings of fault); the Department of Housing and Urban Development maintains data on millions of home mortgages that are presumably "public records" at county clerks' offices. Government's Rehearing Petition at 2. Under the majority's approach, in the absence of state confidentiality laws, there would appear to be a virtual per se rule requiring all such information to be released. The federal government is thereby transformed in one fell swoop into the clearinghouse for highly personal information, releasing records on any person, to any requester, for any purpose. This Congress did not intend.

Quite apart from the potentially daunting administrative burden this will impose, I harbor the gravest concerns that this new-fangled regime will have a pernicious effect on personal privacy interests in conflict with Congress' express will. Congress chose not to enact separate exemptions for each set of personal files kept by the federal government, but chose instead to delegate to the courts the task of making the ultimate determination with respect to disclosure on the basis of criteria that Congress set. Surely we should not eschew this task because we find it awkward. The rehearing filings have referred the court to previously unknown legal authorities, arguing strongly that the information sought here should be held confidential. We should abandon right now our unfortunate departure from traditional FOIA analysis; having repented, we should then conduct an old-fashioned Exemption 7(C) balancing.

In that process, we would be well advised to consider these state laws as one, quite powerful factor. In carrying on the traditional Exemption 7(C) analysis I would, on reflection, conclude that the privacy interest here outweighs the limited public interest in Charles Medico. On that basis, I would affirm the District Court's judgment. As my colleagues have concluded otherwise, I respectfully dissent.