Back in 2018, in District Court Allows IRS to Use Glomar Defense In FOIA Suit Seeking Whistleblower Info I discussed one of the many installments of the litigation between Thomas and Beth Montgomery and the IRS. The case involved the Mr and Mrs Montgomery’s desire to see documents that shed light on how IRS had gotten wind of their partnerships that threw off mostly phony losses.
Late last month, the DC Circuit Court of Appeals affirmed the district court.
read more...The Montgomery family’s partnership transactions attracted the attention of the IRS, leading to a Fifth Circuit opinion finding that most of the partnerships were shams but that one had a legitimate business purpose. The finding that one partnership was legitimate led the Montgomerys to initiate separate refund suits that the IRS ultimately settled, leading to an almost $500,000 refund.
The Montgomerys filed a FOIA claim because they wanted to unearth how the IRS came to scrutinize the transactions, with a particular interest in finding out if there was an informant.
In asserting a Glomar defense to the Montgomery request for documents pertaining to whistleblower forms, the IRS refused to either confirm or deny the existence of the records requested.
As I discussed in my original post, the Glomar defense originates from the Cold War and the CIA wanting to keep secret its efforts to uncover a sunken Soviet submarine:
In certain circumstances, […] an agency may refuse to confirm or deny that it has relevant records. This “Glomar response” derives from a ship, the Hughes Glomar Explorer, about which the CIA refused to confirm or deny the existence of records. See Phillippi v. CIA, 546 F.2d 1009, 1011 (D.C. Cir. 1976). Such responses are appropriate only when “`confirming or denying the existence of records would’ itself reveal protected information.” Bartko v. DOJ, 62 F. Supp. 3d 134, 141 (D.D.C. 2014) (quoting Nation Magazine v. U.S. Customs Serv., 71 F.3d 885, 893 (D.C. Cir. 1995)).
When the government raises a Glomar defense, it still needs to tether the defense to one of the nine statutory FOIA exemptions. In Montgomery, the IRS asserted Exemption 7(D).
Exemption 7(D) excludes from disclosure “records or information compiled for law enforcement purposes, but only to the extent that the production of such law enforcement records or information could reasonably be expected to disclose the identity of a confidential source … [who] furnished information on a confidential basis.” 5 U.S.C § 552(b)(7)(D).
As the DC Circuit described, the district court was “persuaded by the IRS’s explanation that it asserted a Glomar Response because a confirmation of the existence or absence of whistleblower documents in a particular case may lead a savvy requester to the very whistleblower himself. “
In sustaining the district court, it rejected a variety of estoppel type arguments that were premised on prior government admissions that no informant existed. That admission was not sufficient to estop the government from asserting a Glomar defense because it was not quite precisely what was at stake in the FOIA litigation:
The issue of the existence of a confidential informant…is not the same as the issue confronting us now; namely, whether the IRS possesses any documents pertaining to a confidential informant… Indeed, the documents requested by the Montgomerys…include such items as award applications and reportable transaction forms.. As the IRS and the district court correctly point out, the IRS does not pay awards for every form submitted to it.
The key point that the DC Circuit made was that “documents pertaining to a potential whistleblower can exist regardless of whether a whistleblower himself exists” any prior admission about the absence of an informant should not serve as a basis to assert collateral or judicial estoppel or otherwise serve to waive its right to assert a Glomar defense.
As to the merits of the underlying Glomar claim, the Montgomerys argued that the 7(D) exemption was not properly at play, claiming that the exemption only runs to the identity of the whistleblower and any information that the whistleblower provided, rather than their mere existence.
The DC Circuit squarely rejected that argument:
We disagree. Exemption 7(D) permits the IRS to withhold information that “could reasonably be expected to disclose the identity of a confidential source.”
Noting that the whistleblower regs require the IRS to use its best efforts to protect a whistleblower’s identity, the IRS’s policy to neither confirm nor deny the existence “makes sense”:
If the IRS only asserts Glomar when whistleblower records exist, and gives a negative answer when no records exist, savvy requesters would both (1) recognize that a Glomar Response indicates the positive existence of whistleblower documents; and (2) may well be able to deduce the identity of a potential whistleblower himself, the very information the IRS is required to protect. This is especially true when the pool of potential whistleblowers is very small, leading a revenge-seeking requester to narrow down the informant with relative ease. Far from violating FOIA’s statutory scheme, the IRS’s Glomar Response to FOIA requests for whistleblower documents aligns with the purpose of Exemption 7(D) and the duties of the IRS to protect whistleblower identities.
Conclusion
For those with a keen interest in FOIA , I recommend reading the opinion, which includes a discussion of when it is appropriate for the government to overcome the presumption of confidentiality for informants, the nature of district court in camera review, and the adequacy of the IRS’ search for other records that the Montgomerys wanted. To that end and for an even deeper dive, in Saltzman and Book Chapter 2, Taxpayer Access to Information, we have an extensive FOIA discussion, covering materials that are not subject to IRS disclosure as well as myriad FOIA procedural issues.
As I discussed when blogging the 2018 opinion, this case is a significant victory for the IRS and paves the way for future Glomar responses in FOIA cases where someone is seeking information to determine if there was an informant that led to an IRS investigation.
About Leslie Book
The Glomar defense would be all the more difficult to overcome where the desired information involves the tax affairs of an unconnected party. Such was the case — sort of — with the IRS’s settlement agreement with the Church of Scientology as it pertained in Sklar v. Commissioner, 125 T.C. 281 (2005), aff’d 549 F.3d 1252 (9th Cir. 2008). As a practical matter, the Glomar issue was mooted when the existence of the agreement was stipulated by the IRS, following a leak of its text in the news media.
{Perhaps the Sklars would have remained under the radar if they had not, as noted in a separate Tax Court opinion [Sklar v. Commissioner, T.C. Memo. 2000-118], filed their tax return late, a violation the IRS excuses rarely for common, ordinary taxpayers, and all the less for taxpayers such as Michael Sklar, CPA, whose business is preparing tax returns for others.].
As logical as this seems, the law of unintended consequences will come into play. There often is no difference between knowing that a whistleblower exists and not knowing but suspecting that there may be.
Consider the possibilities.
1. If in the case at hand the information that there was a whistleblower COULD NOT enable a savvy taxpayer to figure out who it was, then the IRS’s refusal to confirm or deny that there was a whistleblower is no different from a confirmation that there was one or a confirmation that there was not one.
2. If in the case at hand the information that there was a whistleblower COULD enable a savvy taxpayer to figure out who it was, then suppose that the IRS refuses to confirm or deny that there was a whistleblower. In at least some subset of these cases, perhaps many or most, the same knowledge that the taxpayer has that would lead to the identity of the whistleblower if the IRS had confirmed that there was one also would lead the taxpayer to dismiss the likelihood that the audit (which may have been clearly targeted at one item) was a random occurrence or the IRS’s own idea and to suspect that there was a whistleblower, leading to the same actions as if the IRS had confirmed that there was one. Then, either —
a. The taxpayer will decide to accuse the actual whistleblower of being the whistleblower, or
b. The taxpayer will falsely accuse someone.
3. In some cases, if the IRS refuses to confirm or deny that there was a whistleblower, the taxpayer will not investigate further.
So what it boils down to is whether a particular case is, or in crafting a uniform policy cases in general are likely to be, in category 1 or 2a (it does not matter what the IRS does as long as it does no more that say whether there was or was not a whistleblower), in category 2b (a very bad result comes from not confirming or denying) or in category 3 (a positive result comes from not confirming or denying). The court’s observation that if the IRS asserted a Glomar defense only when there was in fact a whistleblower then the Glomar defense would be tantamount to a confirmation that there was a whistleblower can be justified only if categories 2a & 2b are very small relative to category 3, which may or may not be true in general and certainly may or may not be true in a particular instance.
Accordingly, in many cases, especially where the pool of possible whistleblowers is small, if the IRS knows that there is not a whistleblower it would behoove the IRS to say so in order to protect the possible whistleblowers from possible retaliation for something they didn’t do, even if that means that in other cases there might be a greater suspicion that the IRS’s refusal to say anything either way means that there was a whistleblower., and even if the taxpayer who knows that the pool is small takes a refusal to confirm or deny as confirmation that there was one.
Finally, poker players know from experience (and game theory can quantify and optimize) how to achieve the best results over a series of events through selective use of bluffing. But the limits of the legal process and tax policy, of course, dictate that the IRS could NOT adopt a policy that in a given case it might or might not say whether there is a whistleblower and its statement may or may not be truthful even if that is, in fact, the method that best protects whistleblowers’ identities.