FOIA Exemption 3 allows agencies to withhold documents in response to FOIA requests regarding “matters that are … specifically exempted from disclosure by statute.” 5 U.S.C. § 552(b)(3)(A). To be “specifically exempted” by statute, the statute must (i) prohibit disclosure “in such a manner as to leave no discretion on the issue,” or “(ii) establishe particular criteria for withholding or refer to particular types of matters to be withheld.” Id. Over the years, there has been considerable uncertainty as to which statutes qualified as Exemption 3 statutes.
The OPEN FOIA Act of 2009, Pub. L. No. 111-83, § 564, 123 Stat. 2142, 2184 (2009), was designed to ensure that Congress gave a clear answer to such questions and ensure that Congress deliberately weighed FOIA’s policy of transparency before carving out further exemptions shielding matters from disclosure. The OPEN FOIA Act added to Exemption 3 the requirement that any future statute Congress intended to operate as an Exemption 3 statute, must include a specific reference to Exemption 3 in its text. Thus, the requirement applied to all subsequent statutes, but not to previously-enacted ones. Simple enough. But what about a provision initially enacted before 2009 and then reenacted with modest changes thereafter. Does such a provision still qualify as an Exemption 3 provision if when the reenacted statute did not expressly reference Exemption 3?
The Tiahrt Rider, which in various iterations has limited the use of appropriated funds to disclose the contents of ATF’s Firearms Trace System database (“FTS database”), raises this question. The database houses every record law enforcement authorities have requested in seeking to track a weapon from manufacturer/importer, through the firearms distribution chain, to the first retail purchaser.
Since 2009, the Department of Justice had experienced great success in arguing that even though the Tiahrt Rider has been reenacted with modest amendments several times, the pre-2009 versions of the Rider remained operative and thus qualify as Exemption 3 statutes, even if the reenacted versions might not. District Judges in in the D.C., Ninth, and Sixth Circuits had accepted the argument.
The Department’s winning streak recently came to an end in Everytown for Gun Safety v. ATF, — F. Supp. 3d. —, No. 18 Civ. 2296, 2019 WL 3890220 (S.D.N.Y. Aug. 19, 2019)(available without charge here). District Judge Alison J. Nathan, of the Southern District of New York, relying on the canon of repeal by comprehensive revision, concluded that the post-2009 iterations of the Tiahrt Amendment had repealed the pre-2009 versions. The canon of repeal by comprehensive revision is an exception to the general presumption against replied appeal, i.e., the presumption that a later statute does not repeal earlier ones if the statutes can be reconciled.  When a “later act purports to cover the whole subject covered by [an] earlier act, embraces new provisions, and plainly shows that it was intended not only as a substitute for the earlier act but also to cover the whole subject involved and to prescribe the only rules with respect thereto,” courts presume that an earlier statutes is impliedly repealed. Because the post-2009 iterations did not reference Exemption 3 as required by the Open FOIA Act of 2009 Act, it was not an Exemption 3 statute. Thus the Tiahrt Rider no longer permits ATF to rebuff FOIA requests for documents from ATF’s FTS database.
Judge Nathan characterized the Government’s argument thusly: the pre-2009 Tiahrt Rider provisions “continue to apply because Congress intended their prohibition on public disclosure to be permanent,” as exemplified by the use of “language of futurity,” i.e., “language that applies the statute into the future.” But, Judge Nathan responded, Congress’s subsequent acts, including the currently operative 2012 Rider, cover the whole subject of the earlier acts, impliedly repealing the earlier acts, making them “no longer operative.”
In reaching that conclusion Judge Nathan compared versions of the riders to determine whether the later iterations covered the whole subject of the earlier ones, and found that they unquestionably did so. She noted that Congress’ repetition of language of futurity in every iteration of the Tiahrt Rider signified its understanding that each Rider would provide ongoing rules for the FTS database disclosure that did not necessitate examining prior iterations. Similarly, each Riders’ comprehensive treatment of identical subject matter clearly evidenced Congress’ intent that each “cover the entire field of Firearms Trace System database disclosure.”
Moreover, she noted, accepting the Government’s arguments would defeat Congress’ purpose in requiring an explicit reference to Exemption 3 in any future Exemption 3 statutes, namely “ensur[ing] an open and deliberative process in Congress” before adding Exemption 3 statutes. The Government’s approach would allow the Tiahrt Rider to continue to operate as an Exemption 3 statute even after its reenactment, without the inclusion of language necessary to alert members of Congress to the revised Tiahrt Rider’s implications for FOIA’s principle of transparency. Judge Nathan refused to follow the cases in the district courts of the D.C., Sixth, and Ninth Circuits because they either failed to examine the effect of the OPEN FOIA Act at all, or did not fully consider the implications of the canon of repeal by comprehensive revision.
The case raises two questions, one immediate and the other more hypothetical.
First, the immediate question — should reenactment serve to void a disclosure limitation’s status as an Exemption 3 provisions? The OPEN FOIA Act was clearly crafted to ensure deliberation of the implications of withholding statutes. FOIA exemptions themselves clearly alert Congress to the fact that they are carving out exceptions to FOIA’s general policy of disclosure. And Congress periodically revisits the FOIA statute, including the scope of its exemptions. Moreover, FOIA Amendments would typically be considered by the Committee with general jurisdiction over FOIA. On the other hand, some provisions limiting disclosure will be included in legislation within the jurisdiction of other committees, and may be lost in the substantive provisions of a broad legislative proposal that has little to do with information disclosure. Thus, requiring an explicit reference to FOIA Exemption 3 makes sense, as does strict judicial enforcement of the requirement.
On the other hand, lawyers tend to re-use language that works, and change only as little as need be. Thus, sponsors of legislation, committees considering legislative proposals, or members participating in floor debate may simply not realize that previously-sufficient language must be supplemented in view of the OPEN FOIA statute’s “new” requirements. The requirement can be a trap for the unwary.
But, while congressional intent with regard to some pre-2009 statutes’ status as Exemption 3 statutes is clear, Congress’ intent with regard to others may be decidedly less so. Thus there should be no presumption that Congress approved of court decisions declaring particular pre-2009 statutes to be Exemption 3 statutes. Congressional inaction is a hazardous basis on which to ground conclusions about congressional approval of executive branch or judicial decisions. See, Pension Benefit Guaranty Corporation v. LTV Corp., 496 U.S. 633, 650 (1990). Thus there is no reason to conclude that the sometimes isolated precedents regarding some statutes’ status as Exemption 3 statutes would be upheld by the Supreme Court or were consonant with the views of pre-2009 Congresses.
On the whole, I would be inclined to come down on the side District Judge Nathan came down on.
A second, more hypothetical question is whether appropriations riders should be viewed as Exemption 3 statutes, even if the rider references FOIA. The case for recognizing appropriations riders as Exemption 3 “statutes” is weaker than the case for according such status to substantive legislation. Appropriations riders are not supposed to make changes in substantive law. Indeed, they are designed to cover discretionary decisions. But, at base, disclosing information not covered by FOIA’s other 8 exceptions is not a discretionary action, but a mandatory one. That’s the whole point of changing from a standard that provided agencies with discretion to one that required agency disclosure. It is not clear that appropriations bills governing agency’s exercise of discretion should supersede mandatory obligations. However, the use of a rider to overturn a discretionary FOIA release of information could be covered by an appropriations rider.
Moreover, because appropriations bills must go through, there has always been a concern about allowing substantive legislation to be subject to revision through the appropriations process.
There is an irony here, of course, namely that the OPEN FOIA Act itself was a provision passed as a part of an appropriations bill, namely the Department of Homeland Security Appropriations Act.
But what would the remedy for the improper use of the appropriations process to create an Exemption 3 statute. Would it be requiring the agency to respond to requests in direct contravention of the Tiahrt Rider? Would the remedy be a non-remedy, namely allowing the agency to do as Congress directs, namely refuse to respond to FOIA requests? Or does a remedy require precluding ATF from adding to the FTS database because the appropriation prohibits the agency from doing what it is statutorily obligated to do if its continues adding documents to the database, namely providing information from the database to FOIA requesters? (Indeed, the logic of this remedy would suggest that the database be shut down, not merely than new entries not be added.) But this last remedy creates an even greater substantive impact than requiring the agency to respond to FOIA requests despite a congressional prohibition – in effect by persuading Congress to enact a disclosure provision without the language required by the OPEN FOIA Act, one could shut down a government program.
Don’t expect Congress to provide the needed reference to Exemption 3, and thus resolve the status of the Tiahrt Rider, anytime soon; the Rider is quite controversial. I suspect Judge Nathan’s decision will be appealed. And depending on how the Second Circuit rules, the U.S. Supreme Court may be induced to take the case.
In recent years, we have witnessed an alarming number of FOIA (b)(3) exemptions being offered in legislation—often in very ambiguous terms. . . . The bedrock principles of open Government lead me to believe that (b)(3) statutory exemptions should be clear and unambiguous, and vigorously debated before they are enacted into law. Too often, legislative exemptions to FOIA are buried within a few lines of very complex and lengthy bills, and these new exemptions are never debated openly before becoming law. The consequence of this troubling practice is the erosion of the public’s right to know, and the shirking of Congress’ duty to fully consider these exemptions.
 The provision added a subsection 5 U.S.C. § 552(b)(3)(B) which provided that section would apply to statutes “enacted after the date of enactment of the OPEN FOIA Act of 2009,” only if they “specifically cite to this paragraph,” i.e., 5 U.S.C. § 552(b)(3).
 A variant of this type of question arose in Dongkuk International, Inc. v. U.S. Department of Justice, No. 16-CV-01584, 2016 WL 4543995 (D.D.C. 2016). Mutual legal assistance treaties (MLATs), governing requests for assistance with criminal investigations, provide that the signatories will attempt keep a request for assistance confidential if the nation seeking assistance so requests. Should a court conclude that pre-2009 MLATs, which did not mention Exemption 3, qualify as Exemption 3 statutes, but not identical MLATs concluded after 2009, because they lack explicit reference to Exemption 3? It would seem likely the Congress would prefer consistent treatment of substantively-similar MLATs.
 The Rider is named after the proponent of the original version of the rider, Representative Tom Tiahrt of Kansas.
 See generally, William J. Krouse, Gun Control: Statutory Disclosure Limitations on ATF Firearms Trace Data and Multiple Handgun Sales Report (May 27, 2009). Such traces are generally performed on firearms recovered at a crime scene or seized from a person who possesses the weapon illegally.
 E.g., Ctr. for Investigative Reporting v. U.S. Dep’t of Justice, No. 17 Civ. 6557, 2018 WL 3368884, at *8 (N.D. Cal. July 10, 2018); Smith v. ATF, No. 13-13079, 2014 WL 3565634, at *5 n.2 (E.D. Mich. July 18, 2014); Higgins v. U.S. Dep’t of Justice, 919 F. Supp. 2d 131, 144-45 (D.D.C. 2013).
 WILLIAM D. POPKIN, A DICTIONARY OF STATUTORY INTERPRETATION 196-98 (2007)(doctrine against implied repeal).
 United States v. Lovely, 319 F.2d 673, 679-80 (4th Cir. 1963); see, SHAMBIE SINGER & NORMAN J. SINGER, SUTHERLAND STATUTES AND STATUTORY CONSTRUCTION §23.13 (accessible on Westlaw).
 Indeed, at least one major Exemption 3 “statute” is Federal Rule of Criminal Procedure 6(e), which protects grand jury records. Rule of Criminal Procedure are drafted and revised by advisory committees to the U.S. Supreme Court, which may be even less aware of the OPEN FOIA Act requirements.
 Granted, cases such as Pension Benefit Guaranty Corporation make this observation in the context of failed legislative proposals, and it is unclear whether there were any proposals to overturn any court decisions regarding whether particular statute’s qualified as Exemption 3 statutes.
 A perusal of Statutes Found to Qualify under Exemption 3 of the FOIA (published December 2016) suggests that the precedential authority for a number of statutes qualifying as Exemption 3 statutes consists of a few district court cases (some of which are unreported).
 Rivers v. Roadway Express, 511 U.S. 298 (1994)(only the U.S. Supreme Court can settle questions of federal law, and if the Supreme Court overturns established lower court precedent, that means the lower court precedent was never “the law”).
 The canon disfavoring implied repeal by appropriations rider is not absolute. Robertson v. Seattle Audobon Soc’y, 503 U.S. 429, 440 (1992)(“although repeals by implication are especially disfavored in the appropriations context, Congress nonetheless may amend substantive law in an appropriations statute, as long as it does so clearly”). Moreover, two scholars have presented powerful arguments that the canon rests on bad political science, namely that the “authorization” process is more deliberative, representative, and transparent than the appropriations process. Mathew D. McCubbins & Daniel B. Rodriguez, Canonical Construction And Statutory Revisionism: The Strange Case Of The Appropriations Canon, 14 J. CONTEMP. LEGAL ISSUES 669 (2005).
 See, H. Rep. No. 1497, 89th Cong, 2d Sess., 1-2 (May 9, 1966); S. Rep. No. 813, 89th Cong, 1st Sess. 38, 40-41 (Oct. 4, 1965). One of the clearest statements of the philosophy of the proposed FOIA came, ironically, from the Assistant Attorney General Norbert A. Schlei who testified in opposition to the proposed FOIA. Federal Public Records Law: Hearings on H.R. 5012, etc. before Subcomm. of the Gov’t Operations Comm., 89th Cong., 1st Sess. 14 (1965)(“the basic thrust of . . . the bill . . . is to eliminate any application of judgment to questions of disclosure and nondisclosure, and to substitute therefor, a simple self-executing legislative rule which would automatically determine the availability to any person of all records in the possession of all agencies”).
 Agencies have the discretion to disclose much information that is covered by exemptions. Chrysler Corp. v. Brown, 441 U.S. 281, 293 (1979). Granted, the FOIA Improvement Act of 2016, Pub. L. No. 114-185, 130 Stat. 538 (2016) codifies a judicially enforceable “foreseeable harm” standard to govern such “discretionary” disclosures.
 WILLIAM D. POPKIN, A DICTIONARY OF STATUTORY INTERPRETATION 196-98 (2007).