National Women's Law Center v. Office of Management and Budget

Opinion

UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

)

NATIONAL WOMEN’S LAW CENTER, ) et al., )

)

Plaintiffs, )

)

v. ) Civil Action No. 17-cv-2458 (TSC)

)

)

OFFICE OF MANAGEMENT AND ) BUDGET, et al., )

)

)

Defendants. )

)

MEMORANDUM OPINION

Pending before the court are Defendants’ Motion to Dismiss, ECF No. 11; Plaintiffs’

Motion for Summary Judgment, ECF No. 22; and Defendants’ Motion for Summary Judgment, ECF No. 27. Having reviewed the parties’ filings, the record, and the relevant case law, the court, for reasons set forth below, hereby DENIES Defendants’ Motion to Dismiss, GRANTS Plaintiffs’ Motion for Summary Judgment, DENIES Defendants’ Motion for Summary Judgment, and VACATES the Office of Management and Budget’s stay of the Equal Employment Opportunity Commission’s revised EEO-1 form and the September 15, 2017 Federal Register Notice (Stay the Effectiveness of the EEO-1 Pay Data Collection, 82 Fed. Reg. 43362) announcing the same. It is further ORDERED that the previous approval of the revised EEO-1 form shall be in effect.

I. BACKGROUND

A. The Paperwork Reduction Act The Paperwork Reduction Act of 1995, 44 U.S.C. § 3501 et seq. (“PRA”), was

established to “minimize the paperwork burden” that the federal government may require “for

2

individuals, small businesses, educational and nonprofit institutions, Federal contractors, State, local and tribal governments, and other persons resulting from the collection of information by or for the Federal Government.” 44 U.S.C. § 3501(1). The statute also strives to “improve the quality and use of Federal information to strengthen decisionmaking, accountability, and openness in Government and society.” Id. § 3501(4).

In striking the balance between minimizing the burden on the public and obtaining useful information for the government, Congress established a procedure in which federal agencies must obtain approval from the Office of Management and Budget (“OMB”) to collect certain types of information from the public. Under the PRA, an agency that proposes to collect information first conducts its own “evaluation of the need for the collection of information” and the burden collecting such information would create. Id. § 3506(c)(1)(A)(i). Frequently, the agency is also required to publish a “sixty-day notice” in the Federal Register to solicit comments on the agency’s proposal. Id. § 3506(c)(2)(A). After considering comments and making any revisions, the agency submits the proposed collection of information to OMB and publishes a second Federal Register notice. This notice announces the start of OMB’s review and begins a 30-day comment period. Id. § 3507(a)-(b). “In [this] notice, the agency must set forth (1) a title for the collection of information, (2) a summary of the collection of information, (3) a brief description of the need for the information and the proposed use of the information, (4) a description of the likely respondents and proposed frequency of response to the collection of information, and (5) an estimate of the burden that shall result from the collection of information.” United to Protect Democracy v. Presidential Advisory Comm'n on Election Integrity, 288 F. Supp. 3d 99, 102 (D.D.C. 2017) (citing 44 U.S.C. § 3507(a)(1)(D)(ii)(I)-(V)).

3

OMB may not act on the agency’s request until after the comment period has closed. 44 U.S.C. § 3507(b).

Upon completion of its review, OMB, through the Office of Information and Regulatory Affairs (“OIRA”), makes one of three determinations: it (1) approves the collection of information; (2) disapproves the collection of information; or (3) instructs the agency to make changes to the collection of information. Id. § 3507(c)(1), (e)(1). Before approving a proposed collection, OMB must “determine whether the collection of information by the agency is necessary for the proper performance of the functions of the agency, including whether the information shall have practical utility.” Id. § 3508. Once OMB grants approval, the agency may proceed with its collection, and OMB issues a control number that must be displayed on the collection-of-information form. Id. § 3507(a)(2), (3). An OMB approval is for three years, after which the agency must seek an extension from OMB. Id. §§ 3507(g), (h)(1).

At any point before the approval period expires, OMB “may decide on its own initiative, after consultation with the agency, to review the collection of information.” 5 C.F.R. § 1320.10(f). This review can be started only “when relevant circumstances have changed or the burden estimates provided by the agency at the time of initial submission were materially in error.” Id. OMB may also stay the prior approval of a collection of information not contained in a current rule, but only for “good cause.” Id. § 1320.10(g).

B. The EEO-1

Pursuant to Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e et seq.,

employers are required to “make and keep such records relevant to the determination[] of whether unlawful employment practices have been or are being committed, . . . preserve such records” and produce reports as mandated by EEOC. 42 U.S.C. § 2000e-8(c)(1)-(3). Since

4

1966, EEOC has required that employers with one hundred or more employees file with EEOC the “Employer Information Report EEO-1” (“EEO-1”). 29 C.F.R. § 1602.7. 1 The EEO-1 requires employers to report the number of individuals employed by job category, sex, race, and ethnicity. Agency Information Collection Activities: Revision of the Employer Information Report (EEO-1) and Comment Request, 81 Fed. Reg. 5113 (Feb. 1, 2016) (“Sixty-Day Notice”). EEOC makes aggregate EEO-1 information for major geographic areas and industry groups publicly available on an annual basis. Compl. ¶ 59.

C. Revision of EEO-1 – Component 2 In 2010, the EEOC “joined other federal agencies . . . to identify ways to improve

enforcement of federal laws prohibiting pay discrimination.” 81 Fed. Reg. at 5114. Subsequently, the EEOC “commissioned a study, and the NAS [National Academy of Sciences] convened a Panel on Measuring and Collecting Pay Information from U.S. Employers by Gender, Race, and National Origin.” Id. NAS issued a report which “recognized the potential value for enforcement of collecting pay data from employers by sex, race, and national origin through a survey such as the EEO-1, and emphasized the importance of a definitive plan for how the data would be used in coordination with other equal employment opportunity (EEO) enforcement agencies.” Id. NAS also “recommended that the EEOC conduct a pilot to inform the parameters for any pay data collection.” Id. (footnote omitted). Following NAS’s recommendation, “EEOC commissioned an independent Pilot Study to identify the most efficient means to collect pay data.” Id. The Pilot Study “made technical recommendations about several central components of a data collection, including: The unit of pay to be collected; the best

1 Certain federal contractors and subcontractors with more than fifty employees are also subject to this requirement, which is enforced by the U.S. Department of Labor Office of Federal Contract Compliance Programs (“OFCCP”). 41 C.F.R. § 60-1.7(a).

5

summary measures of central tendency and dispersion for rates of pay; appropriate statistical test(s) for analyzing pay data; and the most efficient and least costly methods for transmitting pay data from employers.” Id. It “also estimated employer burden-hour costs and the processing costs associated with the recommended method of collection.” Id.

In 2012, the EEOC held a two-day meeting with “employer representatives, statisticians, human resources information systems (HRIS) experts, and information technology specialists (work group).” Id. at 5114-15. This group “reviewed the current data collection procedures, provided feedback on future modernization of the EEO surveys, and engaged in brainstorming that led to ideas submitted individually by group participants on a number of topics, including collecting pay data as well as multiple-race category data on the EEO-1.” Id. at 5115. The report from this group “reflect[ed] feedback from participants that the burden of reporting pay data would be minimal for EEO-1 filers.” Id.

On February 1, 2016, following this interagency process, EEOC, in accordance with the PRA, published a Federal Register notice announcing its intention to seek a three-year approval from OMB of “a revised Employer Information Report (EEO-1) data collection.” 81 Fed. Reg. at 5113. The notice explained that the “revised data collection has two components. Component 1 collects the same data that is gathered by the currently approved EEO-1: Specifically, data about employees' ethnicity, race, and sex, by job category. Component 2 collects data on employees' W-2 earnings and hours worked, which EEO-1 filers already maintain in the ordinary course of business.” Id. EEOC proposed “collect[ing] aggregate W-2 data in 12 pay bands for the 10 EEO-1 job categories.” Id. at 5117. The notice provided a weblink to a sample data collection form. Id. at 5118 (“An illustration of the data to be collected by both Components 1

6

and 2 can be found at [link].”). 2 The notice also anticipated that employers would provide the information either through online filing or by uploading an electronic file. Id. at 5120. EEOC estimated that the new pay data collection would increase the reporting time per filer by 3.4 hours. Id. at 5119. On March 16, 2016, EEOC held a public hearing on its proposed pay data collection. Agency Information Collection Activities; Notice of Submission for OMB Review, Final Comment Request: Revision of the Employer Information Report (EEO-1), 81 Fed. Reg. 45479, 45480 (July 14, 2016).

On July 14, 2016, EEOC published a second Federal Register notice (“Thirty-Day Notice”) seeking a three-year approval from OMB of a revised EEO-1 data collection. 81 Fed. Reg. 45479. This notice explained that EEOC had to revise the EEO-1 for the enforcement of equal pay laws. Id. at 45481-83. EEOC intended to maintain its earlier “proposal to collect W-2 income and hours-worked data in the twelve pay bands . . . for each of the 10 EEO-1 job categories.” Id. at 45489. The filings of EEO-1 reports would be done “either by digital upload or by data entry onto a password-protected, partially pre-populated digital EEO-1.” Id. at 45493. For those employers who filed through data uploads, EEOC would post online the new data file specifications “for Components 1 and 2 of the modified EEO-1 as soon as OMB approve[d] the information collection.” Id. at 45487. EEOC further explained that “[t]he EEO-1 data file specifications will be for data uploads (submitting EEO-1 data in one digital file), but they also will describe the formatting of data for direct data entry onto the firm's secure EEO-1 account with the Joint Reporting Committee.” Id. “For reference,” EEOC provided a link to a website with the then-current EEO-1 data file specifications. Id. EEOC estimated that the addition of Component 2 would increase the filing cost for each EEO-1 filed by $416.58. Id. at 45493-94.

2 This webpage is no longer active. Decl. of Benjamin Link ¶ 3, ECF No. 22-2.

7

The collection of pay data would begin with the 2017 reporting cycle, and EEO-1 respondents would be required to submit their reports by March 31, 2018. Id. at 45484. On September 28, 2016, EEOC provided its Final Supporting Statement for the EEO-1 report to OMB for review. Link Decl. ¶ 12, Ex. I. OMB approved the proposed collection on September 29, 2016 and issued an OMB control number for the revised EEO-1. Id. ¶ 11, Ex. H.

EEOC subsequently released an instruction booklet for the March 2018 EEO-1 survey, along with information about the revised EEO-1, including data file specifications for employers who planned to file through data upload. Id. ¶¶ 5, 6. The data file specifications included a sample EEO-1 form for pay data collection and a “data file layout” form. Id. ¶ 5. The data file layout form was a spreadsheet with instructions for formatting pay data submissions to EEOC. Id. ¶ 7, Ex. D.

D. OMB’s Decision to Review and Stay Component 2 Just under a year after OMB approved the data collection, on August 29, 2017, Neomi

Rao, the OIRA Administrator, sent a memorandum to Victoria Lipnic, the Acting Chair of the EEOC, stating that OMB had decided to initiate a review and stay of EEOC’s new collection of pay data under Component 2. In support of this decision, the memorandum stated in pertinent part:

The PRA authorizes the Director of OMB to determine the length of approvals of collections of information and to determine whether collections of information initially meet and continue to meet the standards of the PRA. In this context, under 5 CFR 1320.10(f) and (g), OMB may review an approved collection of information if OMB determines that the relevant circumstances related to the collection have changed and/or that the burden estimates provided by EEOC at the time of initial submission were materially in error. OMB has determined that each of these conditions for review has been met. For example, since approving the revised EEO-1 form on September 29, 2016, OMB understands that EEOC has released data file specifications for employers to use in submitting EEO-1 data. These specifications were not contained in the Federal Register notices as part of the public comment process nor were they outlined in the supporting

8

statement for the collection of information. As a result, the public did not receive an opportunity to provide comment on the method of data submission to EEOC. In addition, EEOC’s burden estimates did not account for the use of these particular data file specifications, which may have changed the initial burden estimate.

OMB has also decided to stay immediately the effectiveness of the revised aspects of the EEO-1 form for good cause, as we believe that continued collection of this information is contrary to the standards of the PRA. Among other things, OMB is concerned that some aspects of the revised collection of information lack practical utility, are unnecessarily burdensome, and do not adequately address privacy and confidentiality issues.

Memorandum from Neomi Rao, Adm’r, OIRA, to Victoria Lipnic, Acting Chair, EEOC (Aug.

29, 2017); JA020, ECF No. 44 (“Rao Memorandum”).

The memorandum further directed EEOC to publish a notice in the Federal Register “announcing the immediate stay of effectiveness of the” pay data collection but “confirming that businesses may use the previously approved EEO-1 form in order to comply with their report obligations for FY 2017.” Id. EEOC published this notice on September 15, 2017. Stay the Effectiveness of the EEO-1 Pay Data Collection, 82 Fed. Reg. 43362 (Sept. 15, 2017).

This stay remains in effect nearly a year and a half later.

E. Plaintiffs’ Lawsuit

Plaintiff National Women’s Law Center (“NWLC”) is “a 46-year-old nonpartisan,

nonprofit organization that advocates for the rights of women and girls at school, at work, at home, and in their communities.” Johnson Decl. ¶ 3. “[C]losing the gender wage gap, and in particular the race and gender wage gaps experienced by women of color,” is “[o]ne of NWLC’s primary and longstanding priorities.” Id. ¶ 4. As part of its mission, NWLC strives to “educate employers, the public, and policymakers about race and gender wage gaps,” id. ¶ 5, and “has published numerous analyses and reports about workplace pay disparities.” Id. Many of these reports “cite data on pay inequality across a number of factors and reflect time-consuming analysis of Bureau of Labor Statistics and Census data undertaken by NWLC staff.” Id.

9

Plaintiff Labor Council for Latin American Advancement (“LCLAA”) is “a national 501(c)(3) representing the interests of approximately 2 million Latino/a trade unionists throughout the United States and Puerto Rico, as well as other non-unionized Latino workers,” and “has 52 chapters around the country.” Sanchez Decl. ¶ 3. LCLAA’s “mission is to assist workers to advance their rights in their workplace and convince employers to take steps to improve working conditions, both through advocacy and through training and counseling workers and union members.” Id. ¶ 4. Recently, “[c]losing the pay gap has been an increasing focus of LCLAA’s work.” Id. ¶ 5. For example, in 2012 “LCLAA created the Trabajadoras Initiative which specifically seeks to protect and advance the interests of Latina workers on issues that impact them, including seeking to eradicate the persistent pay gap.” Id. ¶ 6. LCLAA’s activities include educating its members, chapter presidents, and the public about the pay gap. Id. ¶¶ 7, 8. Starting in 2016, LCLAA has “host[ed] an annual National Latina Equal Pay Summit, which provides information, education and solutions to closing the pay gap.” Id. ¶ 8. Moreover, as part of its education efforts, “LCLAA periodically issues reports . . . about the challenges encountered by Latinos and Latinas in the workforce. The reports discuss data compiled and published by the Government on income and employment—including EEO-I data. LCLAA uses these reports in advocating for policy change and enforcement of equal pay laws, and in educating its members on ways to negotiate with employers and encourage them to follow practices that reduce workforce discrimination.” Id. ¶ 9.

On November 15, 2017, two months after the Rao Memorandum and the subsequent notice in the Federal Register staying the effectiveness of the EEO-1 pay data collection, Plaintiffs sued, naming as defendants: OMB; John Michael Mulvaney, Director of OMB; Neomi Rao, Administrator of OIRA; EEOC; and Victoria A. Lipnic, Acting Chair of EEOC. Plaintiffs

10

ask this court to: 1) declare that OMB Defendants violated the PRA and Administrative Procedure Act (“APA”) and exceeded their statutory authority in reviewing and staying the collection of pay data as part of the EEO-1; 2) declare that the stay announced in the Rao Memorandum and the September 15, 2017 Federal Register notice was a nullity, and that the revised EEO-1 remains in effect; 3) vacate the stay and reinstate the revised EEO-1 reporting requirements; 4) order EEOC Defendants to publish a Federal Register notice announcing this reinstatement or take equivalent action necessary to immediately reinstate the pay data collection; 5) award Plaintiffs their costs, reasonable attorneys’ fees, and other disbursements incurred in this action; and 6) grant such other relief as the court may deem just and proper. Compl. at 34-35.

NWLC claims, among other injuries, that if OMB had not stayed the pay data collection NWLC “would have been able to make its reports and advocacy more robust with additional data and analysis.” Johnson Decl. ¶ 6. Possession of the aggregate EEO-1 pay data would allow NWLC to “focus its resources, analysis, and advocacy on the jobs, industries, and regions where intervention is most urgent.” Id. ¶ 7. LCLAA similarly claims injury from OMB’s decision to stay the pay data collection. With the information that otherwise would have been collected, LCLAA would “have presented statistics on pay equity within industries and across the nation, based on this data. This information would have materially improved LCLAA’s and its members’ ability to negotiate with and educate employers and to fulfill LCLAA’s mission of improving the condition of Latinos and Latinas in the workforce.” Sanchez Decl. ¶ 11.

Defendants have moved to dismiss pursuant to Federal Rule of Civil Procedure 12(b)(1), on the grounds that Plaintiffs lack standing, and pursuant to Federal Rule of Civil Procedure

11

12(b)(6), because Plaintiffs have not challenged a final agency action. Both sides have moved for summary judgment.

II. MOTION TO DISMISS

A. Legal Standard

A motion pursuant to Federal Rule of Civil Procedure 12(b)(1) “presents a threshold

challenge to the court’s jurisdiction.” Haase v. Sessions, 835 F.2d 902, 906 (D.C. Cir. 1987). “[T]he core component of standing is an essential and unchanging part of the case-or- controversy requirement of Article III.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992). The plaintiff bears the burden of establishing the elements of standing, id. at 561, and each element “must be supported in the same way as any other matter on which the plaintiff bears the burden of proof, i.e., with the manner and degree of evidence required at the successive stages of the litigation.” Arpaio v. Obama, 797 F.3d 11, 19 (D.C. Cir. 2015) (quoting Lujan, 504 U.S. at 561). The plaintiff must “show a ‘substantial probability’ that it has been injured, that the defendant caused its injury, and that the court could redress that injury.” Sierra Club v. EPA, 292 F.3d 895, 899 (D.C. Cir. 2002) (citation omitted). With respect to a facial 12(b)(1) motion to dismiss, the court must “accept the well-pleaded factual allegations as true and draw all reasonable inferences from those allegations in the plaintiff's favor.” Arpaio, 797 F.3d at 19. At the summary judgment stage, the plaintiff “must support each element of its claim to standing by affidavit or other evidence.” Scenic Am., Inc. v. U.S. Dep't of Transp., 836 F.3d 42, 48 n.2 (D.C. Cir. 2016) (quotation marks and citation omitted).

“To survive a motion to dismiss” under Federal Rule of Civil Procedure 12(b)(6), “a complaint must contain sufficient factual matter, accepted as true, to state a claim to relief that is plausible on its face.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (quotation marks and citation

12

omitted). However, a court “is not bound to accept as true a legal conclusion couched as a factual allegation.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555 (2007) (quotation marks and citation omitted).

B. Informational and Organizational Standing Because Plaintiffs are two organizations seeking to sue on their own behalf, i.e., seeking

organizational standing, “like an individual plaintiff” they must show “[1] actual or threatened injury in fact [2] that is fairly traceable to the alleged illegal action and [3] likely to be redressed by a favorable court decision.” People for the Ethical Treatment of Animals v. USDA (“PETA”), 797 F.3d 1087, 1093 (D.C. Cir. 2015) (quotation marks and citation omitted).

1. Informational Standing Although both plaintiffs base their claim of injury on the loss of information that

Component 2 would have provided, both nevertheless lack so-called “informational standing.” The D.C. Circuit has been clear that to demonstrate informational standing, a plaintiff must show that “(1) it has been deprived of information that, on its interpretation, a statute requires the government or a third party to disclose to it, and (2) it suffers, by being denied access to that information, the type of harm Congress sought to prevent by requiring disclosure.” Friends of Animals v. Jewell, 828 F.3d 989, 992 (D.C. Cir. 2016) (citation omitted). It is undisputed that Plaintiffs here do not have a statutory right to the information they allege forms their injury in fact. Therefore, any claim to informational standing “fails at the first part of the inquiry, the sine qua non of informational injury.” Id.

2. Organizational Standing Whether Plaintiffs’ loss of information can also serve as an injury in fact as a component

of organizational standing, however, presents a more complicated question. While the D.C.

13

Circuit has provided some conflicting guidance, this court concludes that under binding precedent, in particular PETA and Action Alliance of Senior Citizens of Greater Philadelphia v. Heckler, 789 F.2d 931 (D.C. Cir. 1986), Plaintiffs have demonstrated organizational standing.

The Supreme Court has said that “the injury-in-fact requirement requires a plaintiff to allege an injury that is both concrete and particularized.” Spokeo, Inc. v. Robins, 136 S. Ct. 1540, 1545 (2016) (quotation marks and citation omitted) (emphasis in original). “A particularized injury is personal, individual, distinct, and differentiated—not generalized or undifferentiated. An actual or imminent injury is certainly impending and immediate—not remote, speculative, conjectural, or hypothetical.” Food & Water Watch, Inc. v. Vilsack, 808 F.3d 905, 914 (D.C. Cir. 2015) (quotation marks and citation omitted). Plaintiffs must show that they have “suffered a concrete and demonstrable injury to [their] activities.” Id. at 919 (quotation marks and citation omitted).

In order to determine whether an organization has suffered an injury in fact, courts conduct a “two-part inquiry— we ask, first, whether the agency's action or omission to act injured the [organization's] interest and, second, whether the organization used its resources to counteract that harm.” Id. (quotation marks and citation omitted) (alteration in original).

In Action Alliance, plaintiffs were “four organizations that endeavor[ed], through informational, counseling, referral, and other services, to improve the lives of elderly citizens.” 789 F.2d at 935. They sued the Department of Health and Human Services (“HHS”), alleging that HHS’s regulation “significantly restrict[ed]” the flow of “information regarding services available to the elderly” that, if possessed by plaintiffs, “would enhance [their] capacity . . . to refer members to appropriate services and to counsel members when unlawful age discrimination may have figured in a benefit denial.” Id. at 937. The D.C. Circuit found that the plaintiffs

14

established standing, because the regulations kept plaintiffs from “access to information and avenues of redress they wish[ed] to use in their routine information-dispensing, counseling, and referral activities. Unlike the mere ‘interest in a problem’ or ideological injury in Sierra Club [v. Morton, 405 U.S. 727, 735, 739 (1972)],” plaintiffs had “alleged inhibition of their daily operations, an injury both concrete and specific to the work in which they [were] engaged.” Action Alliance, 789 F.2d at 937-38 (quotation marks omitted) (footnote omitted).

In PETA, plaintiff, an animal rights organization, sued the USDA and requested that the court order USDA to “extend enforcement of the AWA [Animal Welfare Act] to birds covered by the AWA, by enforcing the general AWA standards that presently exist.” 797 F.3d at 1091 (quotation marks omitted) (footnote omitted). PETA claimed that USDA’s failure to investigate allegations of bird mistreatment denied the public reports of those alleged instances, and that it used the information in the reports to educate its members and the public. Id. at 1095-96. The district court found that PETA had standing because USDA’s decision not to apply the AWA to birds “precluded PETA from preventing cruelty to and inhumane treatment of these animals through its normal process of submitting USDA complaints and it deprived PETA of key information that it relies on to educate the public.” Id. at 1094 (quotation marks and citation omitted). The D.C. Circuit affirmed, noting that “[t]he key issue is whether PETA has suffered a concrete and demonstrable injury to [its] activities, mindful that, under our precedent, a mere setback to [PETA's] abstract social interests is not sufficient.” Id. at 1093 (second alteration in original) (quotation marks and citations omitted). The Court explained that in determining “whether an organization's injury is concrete and demonstrable,” a court asks “first, whether the agency's action or omission to act injured the [organization's] interest and, second, whether the organization used its resources to counteract that harm.” Id. at 1094 (alteration in original)

15

(quotation marks and citations omitted). Applying these standards, the Court found that PETA’s alleged injuries were “materially indistinguishable from those alleged by the organizations in Action Alliance.” Id. at 1094. It held that the “USDA's allegedly unlawful failure to apply the AWA's general animal welfare regulations to birds has perceptibly impaired PETA's ability to both bring AWA violations to the attention of the agency charged with preventing avian cruelty and continue to educate the public.” Id. at 1095 (quotation marks and brackets omitted). Because PETA expended resources to counters its injuries, it established organization standing. Id. Significantly, plaintiffs did not allege a statutory right to the information. See id. at 1103 (Millett, J., concurring dubitante) (“Even as alleged by PETA, there is no suggestion that anything in the Animal Welfare Act or any regulation gives PETA any legal right to such information or reports.”) (emphasis in original).

Under Action Alliance and PETA, Plaintiffs have demonstrated an injury in fact. First, they have shown that the loss of pay data caused a “concrete and demonstrable injury” that is “more than simply a setback to [their] abstract social interests.” Havens Realty Corp. v. Coleman, 455 U.S. 363, 379 (1982). As part of its mission to eradicate the gender wage gap, NWLC seeks to educate the public by publishing analyses and reports about workplace pay disparities. Johnson Decl. ¶ 4. When NWLC became aware that EEOC was planning to publish the pay data based on the revised EEO-1, it determined that this data could have revealed new information about the race and gender gaps. Id. ¶ 6. This information would have made NWLC’s reports and advocacy more robust because they would have additional data and analysis. Id. The aggregate EEO-1 pay data would have enabled NWLC to focus its efforts where intervention was most urgent. Id. ¶ 7. With respect to NWLC’s efforts to enforce and redress workplace discrimination, NWLC’s pro bono representation would have benefited from

16

the EEO-1 pay data collection. Id. ¶ 8. With the aggregate pay data, NWLC would more easily have been able to evaluate clients’ claims. Id. ¶ 9. When representing their clients before the EEOC and in courts, the NWLC “bear[s] the costs of amassing evidence about internal pay inequities and comparators’ pay practices to obtain redress.” Id. ¶ 10. This evidence, the NWLC represents, may help persuade EEOC to act or prove claims to a court or jury. Id. Finally, the aggregate employee pay data would reduce NWLC’s costs when representing clients “in the EEOC charge process and beyond.” Id. ¶ 11.

Similarly, as part of its advocacy and educational work, Plaintiff LCLAA issues reports “about the challenges encountered by Latinos and Latinas in the workforce.” Sanchez Decl. ¶ 9. LCLAA would have used the EEO-1 aggregate pay data to present “statistics on pay equity within industries and across the nation, based on this data. This information would have materially improved LCLAA’s and its members’ ability to negotiate with and educate employers and to fulfill LCLAAs mission of improving the condition of Latinos and Latinas in the workforce.” Id. ¶ 11.

Plaintiffs’ declarations establish that NWLC and LCLAA would have used the EEOC pay data to advance their missions by using the data in their publications and as part of their public education and advocacy campaigns. The loss of the pay data constitutes the type of informational harm the D.C. Circuit has found to be “both concrete and specific to the work in which [these organizations] are engaged.” PETA, 797 F.3d at 1095 (quotation marks and citation omitted); see id. at 1094 (stating that the government “deprived PETA of key information that it relies on to educate the public”) (quotation marks and citation omitted); Action Alliance, 789 F.2d at 937-38 (noting that “the challenged regulations denied plaintiffs access to information

17

and avenues of redress they wish to use in their routine information-dispending, counseling, and referral activities”).

Second, Plaintiffs have shown that they have used their resources to “counteract [the]

harm” caused by the stay of the data collection. Food & Water Watch, Inc., 808 F.3d at 919 (citation omitted). After OMB stayed the data collection, NWLC “expended funding and staff time that would have been unnecessary absent the stay.” Johnson Decl. ¶ 12. These additional expenditures included “time to engage with employers to encourage voluntary compliance with equal pay laws and changes in compensation practices and policies” and “voluntary self-audits of pay practices and internal wage gaps, in order to compensate for the self-evaluation of internal wage gaps that the EEO-1 pay data collection would have required employers to undertake.” Id. ¶ 12. Moreover, “in order to meet the objective of improved enforcement of pay discrimination law and increased self-evaluation by employers of their own practices, NWLC has expended and will expend funding and staff time” on tasks that it is “compelled” to take in pursuit of “its mission of eradicating the pay gap.” Id. ¶¶ 13, 14. These tasks require NWLC staff to “create model state and/or local pay data collection legislation; create educational materials setting out the need for such pay data collection and the interests it serves; build coalition and grassroots support in targets states and/or localities for these measures; and advocate at the state and local level for adoption and implementation of such legislation.” Id. ¶ 13. These and similar endeavors have forced NWLC staff members to divert time from their daily operations. Id. ¶ 15.

LCLAA states that “[t]o replicate the same information LCLAA would obtain from the revised EEO-l, LCLAA would have to convince thousands of employers to voluntarily provide data and then employ statisticians to analyze the resulting raw data—efforts that would require enormous expense and still likely be unsuccessful, given LCLAA’s inability to require

18

employers to comply.” Sanchez Decl. ¶ 12 (emphasis in original). LCLAA further represents that “any attempt to compile such data would almost certainly underreport pay inequities, because employers with significant disparities would be particularly unlikely to provide data voluntarily.” Id. Because of the stay, LCLAA is considering whether to use a “very costly” survey of workers to obtain the information it otherwise would have had. Id. ¶ 13. The stay also “forced LCLAA to redeploy its limited staff resources away from their daily operations.” Id. ¶ 14. For example, LCLAA had planned to have staff work on educational materials related to the North American Free Trade Agreement and its impact for Latino/a workers—an initiative that “is a core and common part of LCLAA’s daily operations”—but it could not do so because it had to redeploy its resources to working on obtaining the information that it likely would have received but for OMB’s decision to stay the data collection. Id. ¶ 16.

The court must next determine whether Defendants caused Plaintiffs’ injuries and whether they can be redressed by a favorable court decision. To establish causation, Plaintiffs must show “a causal connection between the injury and the conduct complained of—the injury has to be fairly ... trace[able] to the challenged action of the defendant, and not ... th[e] result [of] the independent action of some third party not before the court.” Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992) (quotation marks and citation omitted) (alterations in original). “Redressability examines whether the relief sought, assuming that the court chooses to grant it, will likely alleviate the particularized injury alleged by the plaintiff.” West v. Lynch, 845 F.3d 1228, 1235 (D.C. Cir. 2017) (quoting Fla. Audubon Soc'y v. Bentsen, 94 F.3d 658, 663–64 (D.C. Cir. 1996) (en banc) (footnote omitted)). Standing is “substantially more difficult to establish,” because Plaintiffs are not the object of the government action. Lujan, 504 U.S. at 562 (quotation marks and citation omitted). “If the challenged conduct is at best an indirect or contributing

19

cause of the plaintiff’s injury—i.e., if the injury may or may not follow from the conduct, based on a chain of contingencies—the plaintiff faces an uphill climb in pleading and proving redressability.” West, 845 F.3d at 1235–36 (quotation marks and citation omitted). Thus, “it must be likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision.” Lujan, 504 U.S. at 561 (quotation marks and citations omitted).

The government does not challenge the proposition that, but for OMB’s stay of the data collection, EEOC would have collected the Component 2 pay data. Therefore, the key issue with respect to causation and redressability is whether Plaintiffs have shown that EEOC likely would have publicly published the aggregate pay data collected from Component 2.

EEOC has discretion whether to publicly report the Component 2 pay data. Therefore, this court pays particular attention to EEOC’s statements about its intention to publicly publish this information. In its Thirty-Day Notice, EEOC said: “Using aggregated EEO-1 data, Census data, and potentially other data sources, the EEOC expects to periodically publish reports on pay disparities by race, sex, industry, occupational groupings, and Metropolitan Statistical Area (MSA).” 81 Fed. Reg. at 45491 (emphasis added).

Based on EEOC’s statement that it “expects to periodically publish reports” based on the aggregated EEO-1 date—a declaration that EEOC has never repudiated—the court finds that EEOC likely would have published the information and made it available to the Plaintiffs. Moreover, EEOC explained how publishing the information would be useful: “Particularly after a few years of data collection, these reports will provide useful comparative data. For smaller employers and others that do not hire consultants to analyze their compensation structures, these reports will be especially informative in light of the business case for equal pay and the need to comply with state equal pay laws.” 81 Fed. Reg. at 45491. The court has no reason not to take

20

the government at its word. Cf. Wheaton Coll. v. Sebelius, 703 F.3d 551, 552 (D.C. Cir. 2012) (“We take the government at its word and will hold it to it.”). 3 Therefore, the court finds that Plaintiffs have demonstrated causation and redressability.

There is some uncertainty about whether an informational injury can exist in the absence of a statutory or regulatory right to the information. Judge Millett expressed this doubt in PETA, even as she recognized that “the majority opinion hew[ed] faithfully to precedential lines” in finding standing where the plaintiffs lacked any statutory or regulatory right to the information. PETA, 797 F.3d at 1099 (Millett, J., concurring dubitante). In Judge Millett’s view, finding injury in fact where plaintiffs seek information unconnected to the exercise of a right conferred by statute is “tantamount to recognizing a justiciable interest in the enforcement of the law.” Id. at 1106 (quotations marks and citation omitted); see also Food & Water Watch, Inc., 808 F.3d at 926 (D.C. Cir. 2015) (Millett J., concurring) (“Because the majority opinion properly applies our precedent to keep a bad jurisprudential situation from getting worse, I concur. But I continue to believe that our organizational standing doctrine should be revisited in an appropriate case.”).

In a difference concurrence, Judge Williams questioned whether informational injury and organizational injury should be viewed as separate categories: “[O]rganizational standing is merely the label assigned to the capacity in which the organization contends it has been harmed; it is not a separate type of injury.” Elec. Privacy Info. Ctr. v. Presidential Advisory Comm'n on Election Integrity, 878 F.3d 371, 381 (D.C. Cir. 2017) (Williams, J., concurring) (emphasis in original) (“EPIC”).

3 Plaintiffs note that EEOC’s expressed intention to publish reports is consistent with its “historical practice of publishing annual reports based on aggregate EE0-1 data, in addition to periodic industry specific reports.” Pls.’ Opp. to Defs.’ Mot. to Dismiss at 34, ECF No. 16 (citing EEOC, Job Patterns for Minorities and Women in Private Industry (EEO-1), https://www.eeoc.gov/eeoc/statistics/employment/jobpat-eeo1/index.cfm).

21

Other cases have indicated that the Court intends for informational and organizational standing to continue to be viewed as distinct categories. See Friends of Animals v. Jewell, 828 F.3d 989, 994 (D.C. Cir. 2016) (criticizing the Sixth Circuit for conflating informational and organizational standing). Some, however, point in a different direction. See EPIC, 878 F.3d at 377 (“We conclude that EPIC lacks standing to obtain such relief because it has suffered no cognizable informational or organizational injury. We analyze and reject those two asserted types of injury in turn without necessarily agreeing that they are in fact analytically separate here.”). Nonetheless, Action Alliance and PETA have not been overruled and remain binding on this court.

The government advances numerous arguments why Plaintiffs have not established organizational standing. All are unavailing. First, the government argues that the D.C. Circuit’s decision in EPIC means that if an organization lacks informational standing it must also lack organizational standing. But EPIC could not and does not purport to overrule Action Alliance and PETA. See General Comm. of Adjustment, GO–386 v. Burlington Northern & Santa Fe Ry. Co., 295 F.3d 1337, 1340 (D.C. Cir. 2002) (explaining that Circuit precedent “binds us, unless and until overturned by the court en banc or by Higher Authority”) (quotation marks and citation omitted). Moreover, as indicated above, EPIC did not hold that informational and organizational standing are the same thing. EPIC, 878 F.3d at 377. EPIC did not make a statutory right to withheld information a necessary condition for both informational and organizational standing. Moreover, its analysis in rejecting standing was more nuanced than the government suggests.

In EPIC, plaintiff was “a nonprofit organization whose stated mission [was] to focus public attention on emerging privacy and civil liberties issues.” Id. at 374 (quotation marks omitted). EPIC sought an injunction under the APA requiring the Presidential Advisory

22

Commission on Election Integrity to undertake privacy assessment review pursuant to the E- Government Act of 2002 before collecting voter data information. Id. EPIC based its claims of both information and organizational standing on the lack of access to these privacy assessments. Id. at 377. The D.C. Circuit found that EPIC lacked informational standing because the statute was designed to protect individuals’ privacy rights, and therefore EPIC as an organization had no cognizable interest in the privacy assessment reviews. Id. at 378 (“As we read it, the provision is intended to protect individuals—in the present context, voters—by requiring an agency to fully consider their privacy before collecting their personal information. EPIC is not a voter and is therefore not the type of plaintiff the Congress had in mind.”) (emphasis in original). The D.C. Circuit then found a lack of organizational injury because without a cognizable interest in the information, EPIC could not show that its efforts to obtain the information were not entirely discretionary budgetary choices. Id. at 378-79.

In contrast, this court has already found that both NWLC and LCLAA expended resources because of the government’s refusal to collect and disseminate the Component 2 information. And, unlike in EPIC, Plaintiffs have sought information that is designed to protect far more than individuals’ privacy rights.

The government further argues that PETA is inapplicable based on the Court’s decision in Food & Water Watch, Inc. Defs.’ Reply in Supp. of Mot. to Dismiss at 5, ECF No. 18. However, in that case, the D.C. Circuit described PETA as “instructive” in explaining why plaintiffs lacked standing. 808 F.3d at 920. The plaintiffs in Food & Water Watch, Inc. challenged a new USDA regulation, claiming that it would lead to an increase in foodborne illness from contaminated poultry—a claim that the D.C. Circuit found to be unsubstantiated. The organizational plaintiff also claimed standing because it would have to spend resources

23

educating the public about the fact that the new regulations did not sufficiently protect the public from unsafe poultry. Id. In PETA, the D.C. Circuit explained, because USDA did not apply animal welfare requirements to birds, it did not generate inspection reports. Id. This “denial of information perceptibly impaired” PETA’s ability to bring statutory violations to USDA’s attention and to continue to educate the public. Id. at 920-21 (quotation mark and citation omitted). In contrast, the organizational plaintiff in Food & Water Watch, Inc. could not “allege that the USDA's action restricts the flow of information that FWW uses to educate its members.” Id. at 921. Thus, Food & Water Watch, Inc. does not advance the government’s argument here, because as in PETA, Plaintiffs NWLC and LCLAA have shown that the stay of EEOC’s data collection deprives them of information that they use to educate their members and the public.

The government further contends that it could not have caused Plaintiffs to incur costs “because the EEOC has never before collected Component 2 pay data, let alone provided aggregate pay data to the public.” Defs.’ Mot. to Dismiss at 16, ECF No. 11. Therefore, the government argues, “OMB’s decision to initiate a review and stay of the same before covered employers ever submitted such data simply maintains the status quo.” Id. Here, Defendants misunderstand the law. “[T]he baseline for measuring the impact of a change or rescission of a final rule is the requirements of the rule itself, not the world as it would have been had the rule never been promulgated.” Air All. Hous. v. EPA, 906 F.3d 1049, 1068 (D.C. Cir. 2018). The government’s argument is also inconsistent with both PETA and Action Alliance, where the Court of Appeals found injury in fact even though plaintiffs claimed an entitlement to information to which they previously did not have access. See PETA, 797 F.3d at 1089 (“Although the Agency has taken steps to craft avian-specific animal welfare regulations, it has yet to complete its task after more than ten years and, during the intervening time, it has

24

allegedly not applied the Act's general animal welfare regulations to birds.”); Action Alliance, 789 F.2d at 937 (plaintiffs challenging a new Department of Health and Human Services regulation).

The government advances a new standing argument in its Motion for Summary Judgment, asserting that “Plaintiffs lack standing to challenge a decision rendered in the context of an informal adjudication to which they are not a party.” Defs.’ Opp. to Pls.’ Mot. for Summ. J. and Mot. for Summ. J. at 16, ECF No. 27-1 (citation omitted).

An adjudication is “retroactive, determining whether a party violated legal policy . . .

[while] rulemaking . . . is of only future effect.” Ameren Servs. Co. v. Fed. Energy Regulatory Comm'n, 880 F.3d 571, 584 (D.C. Cir. 2018); see also Catholic Health Initiatives Iowa Corp. v. Sebelius, 718 F.3d 914, 922 (D.C. Cir. 2013) (“[A]n adjudication must have retroactive effect, or else it would be considered a rulemaking.”) (emphasis in original). The Rao Memorandum relieves covered employers from their obligations to provide Component 2 pay data in the future, and EEOC’s notice in the Federal Register does the same. The government provides the court with no precedent holding that an OMB determination under the PRA is considered an adjudication, and this court will not set a new course. 4 In light of the foregoing, this court finds that Plaintiffs have met their burden to demonstrate organizational standing. 5

4 Moreover, even if the court were to determine that OMB’s stay is an adjudication, Plaintiffs would still have standing to challenge it, because they “point [] to a particular imminent application of the disputed agency policy, the firmness of which is not in dispute, on a fastarriving date certain.” Conference Grp., LLC v. FCC, 720 F.3d 957, 964 (D.C. Cir. 2013) (quotation marks and citation omitted) (alteration in original).

5 Given this finding, the court does not need to reach Plaintiffs’ additional standing claims.

25

C. Final Agency Action

The government moves to dismiss pursuant to Rule 12(b)(6), arguing that OMB’s

decision to reconsider and stay EEOC’s collection of pay data is not a final agency action and thus not subject to judicial review. The APA limits judicial review to “final agency action for which there is no other adequate remedy in a court.” 5 U.S.C. §704. “To be ‘final,’ agency action must ‘mark the consummation of the agency's decisionmaking process’ and ‘be one by which rights or obligations have been determined, or from which legal consequences will flow.’” Clean Air Council v. Pruitt, 862 F.3d 1, 6 (D.C. Cir. 2017) (quoting Bennett v. Spear, 520 U.S. 154, 177-78 (1997)). “Case law interpreting this standard is ‘hardly crisp,’ and it ‘lacks many self-implementing, bright-line rules, given the pragmatic and flexible nature of the inquiry as a whole.’” Friedman v. Fed. Aviation Admin., 841 F.3d 537, 541 (D.C. Cir. 2016) (quoting Rhea Lana, Inc. v. Dep’t of Labor, 824 F.3d 1023, 1027 (D.C. Cir. 2016)).

In Clean Air Council, “a group of environmental organizations, challenge[d] the Environmental Protection Agency's decision to stay implementation of portions of a final rule concerning methane and other greenhouse gas emissions.” 862 F.3d at 4. The government argued that the EPA’s decision to stay the rule was not reviewable because it was not a final agency action. The D.C. Circuit followed the familiar two-pronged Bennett analysis and held that the court had jurisdiction. Id. at 6-7. It found that the EPA’s stay satisfied the first prong – that the agency action “mark[s] the consummation of the agency’s decisionmaking process,” Bennett, 520 U.S. at 177-78 (quotation marks and citation omitted) – because it “suspended the rule’s compliance deadlines,” which, “[was] essentially an order delaying the rule’s effective date, and this court has held that such orders are tantamount to amending or revoking a rule.” Clean Air Council, 862 F.3d at 6. As the court explained, “By itself, EPA's decision to grant reconsideration, which merely begins a process that could culminate in no change to the rule,

26

fails this [finality] test. The imposition of the stay, however, is an entirely different matter.” Id. at 6.

The Court also found that EPA’s stay affected the rights or obligations of the regulated parties. Id. at 7. (quotation marks omitted). The court explained:

Absent the stay, regulated entities would have had to complete their initial monitoring surveys by June 3 and repair any leaks within thirty days. Failure to comply with these requirements could have subjected oil and gas companies to civil penalties, citizens' suits, fines, and imprisonment. The stay—which EPA made retroactive to one day before the June 3 compliance deadline—eliminates that threat, and thus relieves regulated parties of liability they would otherwise face.

Id. (citations omitted) (emphasis in original).

Just as in Clean Air Council, OMB’s decision to reconsider and stay EEOC’s collection of Component 2 pay data “mark[ed] the consummation of the agency’s decisionmaking process.” Clean Air Council, 862 F.3d at 6 (quotation marks and citation omitted). OMB—like EPA—did not choose solely to reconsider or review the previously approved collection of information; rather, it chose to review and stay it, thereby revoking its earlier approval of the Component 2 information collection. OMB’s stay is not interlocutory simply because OMB will conduct future administrative proceedings to decide whether to approve EEOC’s Component 2 data collection. “[T]he applicable test is not whether there are further administrative proceedings available, but rather whether the impact of the order is sufficiently final to warrant review in the context of the particular case.” Id. at 7 (quoting Friedman v. FAA, 841 F.3d 537, 542 (D.C. Cir. 2016).

Moreover, OMB did not have to “review” and “stay” the data collection. The regulation’s terms establish that these are separate decisions. OMB may “review” a data collection “when relevant circumstances have changed or the burden estimates . . . were

27

materially in error.” 5 C.F.R. § 1320.10(f). OMB may “stay” a data collection only for “good cause.” Id. § 1320.10(g). OMB concedes that it made two decisions by acknowledging that it had “also decided to stay immediately the effectiveness of revised aspects of the EEO-1 form.” Rao Memorandum at 2 (emphasis added).

OMB’s stay also satisfies the second step of the Bennett test. Before the stay, covered employers were legally obligated to submit the Component 2 pay data to the EEOC by March 31, 2018, and were required to file the EEO-1, 29 C.F.R. § 1602.7; 41 C.F.R. § 60-1.7(a)(1), and false reporting subjected them to fines or imprisonment, 29 C.F.R. § 1602.8. Covered employers could be compelled by court order to file their EEO-1 if they failed or refused to do so. Id. § 1602.9. The stay eliminated the pay data from this mandatory requirement, therefore creating legal consequences for the regulated entities, no matter whether OMB ultimately decides to approve or disapprove the pay data. See Clean Air Council, 862 F.3d at 7 (“Failure to comply with these requirements could have subjected oil and gas companies to civil penalties, citizens' suits, fines, and imprisonment. The stay . . . eliminates that threat, and thus relieves regulated parties of liability they would otherwise face.”) (citations omitted).

The government nevertheless contends that because the stay does not directly affect the Plaintiffs’ rights or obligations it fails the second prong of the finality test because Plaintiffs do not have any legal right to the pay data EEOC would have collected. Defs.’ Mot. to Dismiss at 26-27, ECF No. 11. This argument fails because the correct test asks whether the stay has legal consequences, not whether it affects the plaintiffs’ legal obligations. In Clean Air Council, the D.C. Circuit found that an environmental organization had standing where the stay affected the obligations of the regulated industries – oil and gas – not the environmentalists. 862 F.3d at 4; see also Nat’l Ass’n of Home Builders v. U.S. Army Corps of Engineers, 417 F.3d 1272, 1281

28

(D.C. Cir. 2005) (“Because the Corps' [policies] mark the completion of the Corps' decision- making process and affect the [regulated entities’] day-to-day operations, they constitute final agency action regardless of the fact that the Corps' action might carry different (or no) consequences for a different challenger, such as an environmental group.”). The government relies on DRG Funding Corp. v. Sec’y of Hous. & Urban Dev., 76 F.3d 1212 (D.C. Cir. 1996), but in that case the agency’s actions did not have any legal consequence. Id. at 1215 (“[This] is not a case in which legal consequences will flow from the agency action taken thus far.”) (quotation marks and citation omitted).

Defendants argue that Clean Air Council is inapposite. They point out that Clean Air Council involved the stay of a rule promulgated pursuant to notice and comment, the PRA permits the review and stay of the collection of information not contained in a rule, and permits only review but not the stay of information contained in a rule, and OMB’s stay of Component 2 data is a stay of a collection of information not contained in a rule. While each of these points is correct, the Defendants cannot weave them together in a way that demonstrates why the stay in this case is not final.

First, OMB discusses the legislative history of the PRA with respect to the decision to permit stays of some collections of information but not others, see Defs.’ Reply in Supp. of Mot. to Dismiss at 14-15, but this discussion reveals only that Congress granted substantive authority to allow a stay in one place and withheld authority to grant a stay elsewhere. It does not speak to the finality of a stay. 6

6 Defendants suggest that the notice and comment underpinnings of the rule stayed in Clean Air Council were critical to the court’s finding that the stay amounted to final action. See Defs.’ Reply in Supp. of Mot. to Dismiss at 17. However, the language that Defendants quote does not

29

In any event, Defendants have failed to account for the fact that this case involves the interplay between two governmental entities, OMB and EEOC, and by staying EEOC’s collection of information, OMB has effectively stayed its own prior approval of the collection— an approval which was itself granted only after two rounds of notice and comment. See 81 Fed. Reg. 5113; 81 Fed. Reg. 45479. Moreover, Defendants do not argue that judicial review would be unavailable if OMB had initially disapproved the collection of information by EEOC. Cf. 44 U.S.C. § 3507 (d)(6) (expressly stating that certain actions are not subject to judicial review). If such disapproval would have been judicially reviewable, it does not make any sense that OMB could evade judicial review by first approving the collection and then staying it.

III. SUMMARY JUDGMENT

A. Legal Standard

When a plaintiff challenges an agency’s final action under the APA, summary judgment

“is the mechanism for deciding whether as a matter of law an agency action is supported by the administrative record and is otherwise consistent with the APA standard of review.” Louisiana v. Salazar, 170 F. Supp. 3d 75, 83 (D.D.C. 2016) (citing Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415 (1971)). A court must “hold unlawful and set aside agency action” that it finds to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A).

come from the portion of the opinion dealing with the finality of the stay; it comes from the portion of the opinion finding the stay arbitrary and capricious.

30

B. Analysis of Whether the Stay Violates OMB’s Regulations or is Arbitrary and Capricious

Under its own regulations, OMB may review a previously approved collection of

information only when “relevant circumstances have changed or the burden estimates provided by the agency at the time of initial submission were materially in error.” 5 C.F.R. § 1320.10(f). If either requirement is met, OMB may stay a prior approval of a collection of information not contained in a rule, but only for “good cause” shown and “after consultation with the agency.” Id. § 1320.10(g). OMB has not shown that a relevant circumstance has changed or that the burden estimate provided was materially in error. Moreover, it has not shown good cause.

In staying EEOC’s data collection, OMB stated that the “data file specifications for employers to use in submitting EEO-1 data . . . were not contained in the Federal Register notices as part of the public comment process nor were they outlined in the supporting statement for the collection of information.” Rao Memorandum at 1. Therefore, OMB claimed that 1) “the public did not receive an opportunity to provide comment on the method of data submission to EEOC”; and 2) “EEOC’s burden estimates did not account for the use of these particular data file specifications, which may have changed the initial burden estimate.” Id.

OMB’s assertion that the data file specifications were not contained in the Federal Register, thereby depriving the public of an opportunity to comment on them, is misdirected, inaccurate, and ultimately unpersuasive. The claim is misdirected, because the data file specifications consisted of a sample spreadsheet and formatting specifications which explained how to format a spreadsheet; they did not change the content of the information collected. In both the Sixty-Day and Thirty-Day notices, EEOC described in detail the information it proposed to collect. The government’s argument therefore focuses on a technicality that did not affect the employers submitting the data.

31

The government’s claim that employers lacked an opportunity to comment on the data file specifications is strained, because the Thirty-Day Notice explicitly stated that for employers who were going to file through data uploads, the EEOC would post online the new data file specifications “for Components 1 and 2 of the modified EEO-1 as soon as OMB approves the information collection.” 81 Fed. Reg. at 45487. Moreover, the Sixty-Day Notice directed employers to a website with sample data collection forms which provided “an illustration of the data to be collected by both Components 1 and 2 [of the revised EEO-1].” 81 Fed. Reg. at 5118. The sample spreadsheets in the data file specification match in all material respects the sample form provided in the Sixty-Day Notice. Compare id. with Link Decl. ¶ 8, Ex. E. Thus, employers not only knew that the data file specifications would be posted after OMB’s approval, but also knew what the specifications would look like, putting the employers in a good position to comment if they desired.

Moreover, OMB gave its approval knowing that the Thirty-Day Notice stated that the file data specifications would be posted “as soon as” OMB approved the information collection. Plainly, OMB and employers expected some file specifications, so OMB’s assertion that anyone was caught off guard by the mere posting of data file specifications rings hollow. Unable to point to any material variance between the sample form in the Sixty-Day Notice and the specifications ultimately posted, OMB’s burden was to show how the file data specifications altered the content of the information required from employers or significantly increased the burden on employers in producing that information. OMB did not carry that burden. Moreover, OMB did not explain why posting the data file specifications after OMB approved the Component 2 collection should jeopardize the collection, when OMB had previously allowed data file specifications to be submitted after it approved EEO-1 forms. See, e.g., Agency

32

Information Collection Activities: Notice of Submission for OMB Review; Final Comment Request, 70 Fed. Reg. 71294, 71301 (Nov. 28, 2005); Agency Information Collection Activities: Proposed Collection; Submission for OMB Review, 79 Fed. Reg. 72678 (Dec. 8, 2014).

OMB’s assertion that “EEOC’s burden estimates did not account for the use of these particular data file specifications, which may have changed the initial burden estimate,” Rao Memorandum at 1, is also unavailing. First, as noted, EEOC’s sample form in the Sixty-Day Notice and description in both the Thirty-Day and Sixty-Day notices of the pay data that would be collected were consistent with the data file specifications. Second, OMB did not actually find that the data file specifications would change the initial burden estimates; rather, it stated that data file specifications “may change the initial burden estimate.” Rao Memorandum at 1 (emphasis added). This equivocation is not surprising given that OMB’s assertion was unsupported by any analysis.

The only “changed circumstance” alleged by OMB is the release of the data file specifications. But this is far from a “changed circumstance” in any meaningful sense. As mentioned, in the Thirty-Day Notice, EEOC informed OMB and the public that it would post file specifications after OMB approved the collection of information. After OMB approved the collection of information, EEOC released the data file specifications. There was no changed circumstance. EEOC proceeded exactly as planned and as OMB had approved.

Moreover, OMB merely speculated that the initial burden estimate “may” be incorrect, while failing to support its speculation with any analysis. This omission is remarkable because OMB did not assert that EEOC’s previous analyses were based on formats meaningfully different from the file data specifications. See 81 Fed. Reg. at 5119-20, 81 Fed. Reg. at 45492- 95 (earlier burden analyses). OMB’s unsupported conjecture falls short of showing any prior

33

error, let alone a material one. “Though an agency's predictive judgments . . . are entitled to deference . . . deference to such ... judgment[s] must be based on some logic and evidence, not sheer speculation.” Sorenson Commc'ns Inc. v. FCC, 755 F.3d 702, 708 (D.C. Cir. 2014) (quotation marks and citations omitted) (alteration in original); see also Water Quality Ins. Syndicate v. United States, 225 F. Supp. 3d 41, 70 (D.D.C. 2016) (“speculation … is an inadequate replacement for the agency’s duty to undertake an examination of the relevant data and reasoned analysis”) (quoting Horsehead Res. Dev. Co. v. Browner, 16 F.3d 1246, 1269 (D.C. Cir. 1994)).

OMB also failed to demonstrate good cause for the stay. It did not explain in any substantive way why it believed that the revised EEO-1 was contrary to PRA standards. See Amerijet Int’l, Inc. v. Pistole, 753 F.3d 1343, 1350 (D.C. Cir. 2014) (“[A]n agency must explain ‘why it chose to do what it did.’ And to this end, conclusory statements will not do; an ‘agency’s statement must be one of reasoning.’”) (emphasis in original) (quoting Tourus Records v. Drug Enf’t Admin., 259 F.3d 731, 737 (D.C. Cir. 2001); Butte Cty., Cal. v. Hogen, 613 F.3d 190, 194 (D.C. Cir. 2010)); see also Bauer v. DeVos, 325 F. Supp. 3d 74, 105 (D.D.C. 2018) (“[T]he APA . . . requires that agencies ‘provide an explanation that will enable the court to evaluate the agency’s rationale at the time of decision.’”) (citation omitted). OMB’s stated reason when issuing the stay also conflicted with its prior findings that EEOC’s data collection had practical utility, 81 Fed. Reg. at 45481-83, was designed to minimize the burden on reporting employers, id. at 45492-95, and provided adequate privacy and confidentiality protections, id. at 45491-92. OMB failed to explain these inconsistencies. See Dist. Hosp. Partners, L.P. v. Burwell, 786 F.3d 46, 59 (D.C. Cir. 2015) (“We have often declined to affirm an agency decision if there are unexplained inconsistencies in the final rule.”); Bauer, 325 F. Supp. 3d at 109 (“[T]he

34

Department failed to acknowledge, much less to address, the inconsistency between its current view that those provisions stand on legally questionable footing, and its prior conclusion that they were legally sound. . . . [A]n unacknowledged and unexplained inconsistency is the hallmark of arbitrary and capricious decision-making.”). 7 An agency’s action is arbitrary and capricious if the agency “(1) has relied on factors which Congress has not intended it to consider, (2) entirely failed to consider an important aspect of the problem, (3) offered an explanation for its decision that runs counter to the evidence before the agency, or (4) offers an explanation that is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.” Am. Wild Horse Pres. Campaign v. Perdue, 873 F.3d 914, 923 (D.C. Cir. 2017) (quotation marks and citation omitted). While “[a]gencies are free to change their existing policies,” in doing so they must “provide a reasoned explanation for the change.” Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125 (2016) (citations omitted). OMB’s action in staying EEOC’s collection of Component 2 was arbitrary and capricious for the same reasons the agency violated its own regulation: OMB’s decision to stay the collection of information totally lacked the reasoned explanation that the APA requires. 8 In urging this court to uphold OMB’s decision to stay EEOC’s pay data collection, the government contends that “OMB’s interpretation of its ‘review and stay’ regulations is entitled to deference,” unless it is “plainly erroneous or inconsistent with the regulation,” because “neither the PRA nor OMB’s regulations define when ‘relevant circumstances have changed,’

7 OMB did not explain why a stay was necessary, when the percentage of employers likely to use data upload is small, perhaps as low as two percent, 81 Fed. Reg. at 5119 n.55, 5120 n.62, and no employer is required to use data upload rather than data entry. 8 Because the court finds that OMB’s stay of EEOC’s pay data collection violated the PRA and was arbitrary and capricious, it does not need to reach Plaintiffs other arguments.

35

what constitutes a ‘material error’ in the burden estimates initially provided, or when ‘good cause exists.’” Defs.’ Opp. to Pls. Mot. for Summ. J and Mot. for Summ. J. at 19-20. Plaintiffs counter that because “the Rao Memorandum is almost entirely conclusory, parroting regulatory language,” it lacks reasoning and is not entitled to deference. Pls.’ Reply in Supp. of Mot. for Summ. J. and Opp. to Defs.’ Mot. for Summ. J. at 6, ECF No. 32. Moreover, Plaintiffs argue that OMB has previously articulated a definition of a “good cause” standard and should be held to it. In the preamble to a PRA regulation, OMB stated that good cause “exists when continued collection of information would be significantly contrary to the standards of the Act.” Controlling Paperwork Burdens on the Public; Proposed Rulemaking, 47 Fed. Reg. 39515, 39522 (Sept. 8, 1982). As part of this good cause standard, OMB must “take into consideration any disruption such reconsideration would create in the agency’s program.” Id. The government responds that as a general matter, preambles to rules are not binding on agency actions. See Defs.’ Mot. for Summ. J. at 21 n.4.

The court need not decide whether the good cause standard provided in the preamble applies here. Even without holding OMB to those requirements, the court finds that OMB provided inadequate reasoning to support its decision to stay the data collection. Courts “do not defer to an agency's conclusory or unsupported suppositions.” Nat'l Shooting Sports Found., Inc. v. Jones, 716 F.3d 200, 214 (D.C. Cir. 2013) (quotation marks and citation omitted).

In further support of its claim that OMB’s decision to stay EEOC’s collection of Component 2 information is a reasonable interpretation of the regulations and supported by the record, the government relies on four letters OMB received in early 2017, after the posting of the file specifications. Defs.’ Opp. to Pls.’ Mot. for Summ. J. and Mot. for Summ. J. at 22-23. These letters, according to the government, expressed concern about the lack of adequate notice

36

and comment and buttressed OMB’s finding that “Component 2 could impose a burden on regulated entities.” Id. at 23. However, the Rao Memorandum does not indicate that OMB relied on these comments or even took them into consideration. Moreover, one of the four letters (from the Chamber of Commerce) does not discuss the data file specifications. See Letter re Request for Review; EEOC’s Revision of the Employer Information Report from Randel K. Johnson, Senior Vice President, Labor, Immigration & Employee Benefits and James Plunkett, Director, Labor Law Policy, U.S. Chamber of Commerce, to John M. Mulvaney, Dir., Office of Management and Budget (Feb. 27, 2017); JA046. The other letters do not provide any analysis or conclude that the data formatting specifications increase the burden on EEO-1 filers. An agency’s speculation cannot suffice for APA review purposes. Sorenson Commc'ns Inc., 755 F.3d at 708. Similarly, an agency cannot simply rely on the speculation of commenters. Instead, the agency must conduct a critical examination of comments on which it relies. See Nat’l Ass’n of Regulatory Util. Comm’rs v. FCC, 737 F.2d 1095, 1125 (D.C. Cir. 1984) (agency may not rely on comments “uncritically” and must apply its “expert evaluation”); Am. Great Lakes Ports Ass'n v. Zukunft, 296 F. Supp. 3d 27, 39 (D.D.C. 2017) (“Thus, an agency may, for example, rely on comments submitted during the notice and comment period as justification for [a] rule, so long as the submissions are examined critically.”) (quotation marks and citations omitted) (alteration in original). Finally, the government omits mention of the comments OMB received in support of the file data specifications. For instance, NWLC, with 82 organization signatories, submitted a comment on April 12, 2017, around the same time that OMB received the four letters it cites. NWLC explained both why the data file specifications did not increase any burden and did not constitute a change in circumstances. See Letter re Opposition to Reopening Review of the Equal Employment Opportunity Commission's Employment Information (EEO-1)

37

Report (OMB Control Number 3046-0007), from NWLC, et al., to John M. Mulvaney, Dir., Office of Management and Budget (Apr. 12, 2017); JA115. Without explaining why, an agency cannot rely on some comments while ignoring comments advocating a different position. AARP v. U.S. Equal Emp’t Opportunity Comm'n, 267 F. Supp. 3d 14, 32 (D.D.C. 2017) (“An agency must explain why it chose to rely on certain comments rather than others.”) (emphasis in original).

The government further contends that the Rao Memorandum “simply summarizes the four months of meetings, emails, and conference calls between OMB and EEOC during which each issue underlying the Administrator’s decision to review and stay was discussed.” Defs.’ Opp. to Pls.’ Mot. for Summ. J. and Mot. for Summ. J. at 30. Nothing in the record supports this claim. Further, the government’s assertion does not mean that the discussions involved an analysis of evidence supporting a conclusion of changed circumstances or materially erroneous burden estimates. In fact, the only evidence in the record of documented agency analysis rejects the government’s current contention about the data file specifications. According to an April 14, 2017 EEOC memorandum:

The EEOC’s published filed specifications are not a new requirement but rather simply a familiar tool to make it easier for employers to submit EEO-1 data to the EEOC. The EEOC provided this tool to employers for use with the 2016 EEO-1, as evident on the EEOC website where the file specifications for the 2016 EEO-1 were published … Like those later published for the EEO-1 pay data collection, the 2016 specifications were in comma-separated values format (“csv”), a format that enables employers to convert tabular data (like that in the EEO-1) for importation and exportation to a database.

Employers that used data upload technology for their 2016 EEO-1 reports used these csv specifications.

Publication of the file specifications after OMB’s approval of the pay data collection is not a “significant” change that warrants OMB’s reconsideration. As noted above, EEOC indicated in the 30-day notice its intent to post updated specifications, and stated that it would provide support to employers and HRIS vendors as they transitioned to the new reporting requirements. EEAC had notice of these data specifications, was probably

38

familiar with the 2016 version, and had ample opportunity to raise questions or concerns before OMB’s approval of the revised EEO-1, but did not do so.

Memorandum from Peggy Mastroianni to Victoria Lipnic, Acting Chair, EEOC (Apr. 14, 2017);

JA126.

The government now claims that Component 2 required “significant formatting changes.”

Defs. Mem at 23. But the Rao Memorandum did not reach that conclusion. Administrator Rao’s letter said only that the data file specifications “may have changed the initial burden estimate.” Rao Memorandum at 1. Moreover, even if OMB had relied on a finding of “significant formatting changes” at the time of the stay, its conclusion would have lacked analytic and evidentiary support. This court cannot “fill in critical gaps in an agency’s reasoning. We can only look to the [agency’s] stated rationale. We cannot sustain its action on some other basis the [agency] did not mention.” Point Park Univ. v. NLRB, 457 F.3d 42, 50 (D.C. Cir. 2006); see also NAACP v. Trump, 298 F. Supp. 3d 209, 237 (D.D.C. 2018) (“[B]ecause ‘a reviewing court ... must judge the propriety of [agency] action solely by the grounds invoked by the agency,’ post hoc explanations that the agency did not articulate when it acted are insufficient.” (quoting SEC v. Chenery Corp., 332 U.S. 194, 196, (1947)) (alteration in original).

The government’s position rests on hyper-technical formatting changes that have no real consequences for employers. While there may be instances when formatting changes could be burdensome, that is not the case here. Similarly, it is conceivable that the government could impose formatting requirements that made the presentation of collected information far more difficult for employers. But in this case, OMB and the regulated entities knew what the formatting would look like before OMB gave its approval to the Component 2 collection, and the government has failed to demonstrate any likelihood that the data file specifications meaningfully increase the burden on employers. Indeed, neither the Rao Memorandum nor the

39

administrative record as a whole demonstrates that before issuing its stay, OMB ever analyzed the data file specifications in an effort to determine whether they meaningfully changed the burden of collecting Component 2 information.

IV. REMEDY Because the court has determined that OMB’s stay of EEOC’s pay data collection was

illegal, it must fashion an appropriate remedy. The APA provides that a court shall “hold unlawful and set aside agency action, findings, and conclusions found to be . . . arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). The D.C. Circuit has stated that “vacatur is the normal remedy” for an APA violation. Allina Health Servs. v. Sebelius, 746 F.3d 1102, 1110 (D.C. Cir. 2014). The APA “itself contemplates vacatur as the usual remedy when an agency fails to provide a reasoned explanation for its regulations.” AARP v. U.S. Equal Emp’t Opportunity Comm'n, 292 F. Supp. 3d 238, 242 (D.D.C. 2017). The presumption of vacatur, “however, is not absolute, and a remand without vacatur may be ‘appropriate [if] “there is at least a serious possibility that the [agency] will be able to substantiate its decision” given an opportunity to do so, and when vacating would be “disruptive.”’” Bauer v. DeVos, 332 F. Supp. 3d 181, 184 (D.D.C. 2018) (quoting Radio-Television News Dirs. Ass'n v. FCC, 184 F.3d 872, 888 (D.C. Cir. 1999)) (alterations in original). “Courts in this Circuit . . . have long recognized that ‘when equity demands, an unlawfully promulgated regulation can be left in place while the agency provides the proper procedural remedy.’” Shands Jacksonville Med. Ctr. v. Burwell, 139 F. Supp. 3d 240, 267 (D.D.C. 2015) (quoting Fertilizer Inst. v. EPA, 935 F.2d 1303, 1312 (D.C. Cir. 1991) (citation omitted) (footnote omitted)). A court’s decision to remand without vacatur “depends on the ‘seriousness of the order’s deficiencies’ and the likely ‘disruptive consequences’ of vacatur.”

40

Allina Health Servs., 746 F.3d at 1110 (quoting Allied–Signal, Inc. v. U.S. Nuclear Regulatory Comm'n, 988 F.2d 146, 150-51 (D.C. Cir. 1993)).

OMB’s deficiencies were substantial, and the court finds it unlikely that the government could justify its decision on remand, despite its assertion that “OMB could easily cure the defects in its memorandum by further explanation of its reasoning.” Defs.’ Opp. to Pls.’ Mot. for Summ. J. and Mot. for Summ. J. at 33. The government’s deficiency is not that it failed to explain OMB’s “reasoning,” but that OMB’s reasoning lacked support in the record.

The government also argues that “vacatur could upset the current expectation of filers, who may not be aware of this litigation nor its potential impact on their obligation to complete their Component 2 submissions.” Defs.’ Reply in Supp. of Mot. to Dismiss or in Alt. Mot. for Summ. J. at 19 (emphasis added), ECF No. 36. This speculation is unsupported by the record. Moreover, the revised pay data collection had been in place for almost a year by the time it was stayed. The government’s argument that a stay is only temporary further undercuts its argument that vacatur would be disruptive, because affected entities were on notice that the stay could be withdrawn at any time. The court therefore finds that vacatur will not have potentially disruptive consequences.

Accordingly, this court finds that both Allied-Signal factors favor vacatur as the appropriate remedy.

V. CONCLUSION For the foregoing reasons, the court hereby DENIES Defendants’ Motion to Dismiss, ECF No. 11; GRANTS Plaintiffs’ Motion for Summary Judgment, ECF No. 22; DENIES Defendants’ Motion for Summary Judgment, ECF No. 27; and VACATES OMB’s stay of the

41

EEOC’s revised EEO-1 form and the September 15, 2017 Federal Register Notice (Stay the Effectiveness of the EEO-1 Pay Data Collection, 82 Fed. Reg. 43362) announcing the same.

It is further ORDERED that the previous approval of the revised EEO-1 form shall be in effect. Date: March 4, 2019

Tanya S. Chutkan

TANYA S. CHUTKAN

United States District Judge