State v. Ross

358 F. Supp. 3d 965

• Court: District Court, N.D. California
• Decided: March 6, 2019
• Docket: Case No. 18-cv-01865-RS; 18-cv-02279-RS
• Status: Published

Opinion

RICHARD SEEBORG, United States District Judge *973TABLE OF CONTENTS

I. SUMMARY OF DECISION...973

II. BACKGROUND...976

III. STANDING...977

A. Legal Standard...977

B. Findings of Fact Related to Standing...977

C. Conclusions of Law Related to Standing...1002

1. Injury-in-Fact...1002

2. Traceability...1006

3. Redressability...1007

IV. APA CLAIM...1007

A. Legal Standard...1007

B. Scope of Review...1008

C. Findings of Fact Based Exclusively on the Administrative Record...1009

D. Findings of Fact Based on Extra-Record Evidence...1026

E. Conclusions of Law...1037

V. CONSTITUTIONAL CLAIMS...1046

A. Legal Standard...1046

B. Scope of Review...1047

C. Conclusions of Law...1048

VI. REMEDIES...1049

I. SUMMARY OF DECISION

The formal decision by Secretary of Commerce Wilbur L. Ross, Jr. on March 26, 2018 to add a citizenship question to the 2020 Decennial Census violated the Administrative Procedure Act ("APA") and the Enumeration Clause of the United States Constitution. Nearly a year before issuing that decision, on May 2, 2017, Secretary Ross sent an email to Deputy Chief of Staff Earl Comstock stating in part "I am mystified why nothing [has] been done in response to my months old request that we include the citizenship question. Why not?" What ensued was a cynical search to find some reason, any reason, or an agency request to justify that preordained result.

As to the APA, one need look no further than the Administrative Record¹ to conclude *974that the decision to include the citizenship question was arbitrary and capricious, represented an abuse of discretion, and was otherwise not in accordance with law. In response to Secretary Ross's demand, Comstock began to search for an agency that would be willing to request the inclusion of the citizenship question in the 2020 Census. When initially approached by Comstock about the citizenship question, the Department of Justice ("DOJ") opted not to request its inclusion in the census. Comstock then reached out to the Department of Homeland Security, which similarly declined to request the addition of the question. Only after Secretary Ross personally interceded with then Attorney General Jeff Sessions did the DOJ switch its position and request the inclusion of a citizenship question, ostensibly to assist in the enforcement of Section 2 of the Voting Rights Act ("VRA").

Despite unrefuted evidence produced by the professional staff of the Census Bureau that inclusion of a citizenship question would likely result in a significant differential decline in self-response rates within noncitizen and Latino communities and that the requested data could be obtained by other means, Secretary Ross insisted upon adding the citizenship question to the census. When Census Bureau staff offered to meet with DOJ staff to ascertain if other available data could be used to meet their VRA enforcement needs, DOJ took the unprecedented step of refusing to allow even such an inter-agency meeting to take place.

These facts and other evidence contained in the Administrative Record, along with all reasonable inferences to be drawn therefrom, demonstrate that Secretary Ross's reliance on VRA enforcement to justify inclusion of the citizenship question was mere pretext and the definition of an arbitrary and capricious governmental act. Moreover, Secretary Ross's conclusion that adding the citizenship question would enable the Census Burau to obtain more "complete and accurate data" in response to the DOJ's request is not only unsupported, it is directly contradicted by the scientific analysis contained in the Administrative Record. PTX-26 at 1, 7. While it is of course appropriate for an incoming cabinet member to advocate for different policy directions, to solicit support for such views from other agencies, and to disagree with his or her professional staff, this record reflects a profoundly different scenario: an effort to concoct a rationale bearing no plausible relation to the real reason, whatever that may be, underlying the decision.

Again confining review solely to the Administrative Record, it is evident that the inclusion of the citizenship question on the 2020 Census violated both Sections 6(c) and 141(f)(3) of the 1976 Census Act. Section 6(c) mandates that, to the maximum extent possible, the Secretary use administrative records as opposed to additional census questions to obtain secondary data, such as demographic information. Section 141(f) mandates certain timely reports to Congress regarding the subject and questions to be included on the census and limits the Secretary's ability subsequently to modify the contents of the census absent new circumstances that necessitate a change. Quite simply, Secretary Ross ignored these statutory requirements in issuing his March 26, 2018 decision.

While finding a violation of the APA logically flows from the Administrative Record in this action alone, the facts here satisfy the requisite standard warranting consideration of extra-record evidence. Such evidence includes the absence of any *975effort to test the impact of the addition of the citizenship question to the census, the deviation from the Census Bureau's usual process for adding new questions to the census, the troubling circumstances under which the DOJ's request letter was drafted and procured, and Sessions' order prohibiting DOJ staff from meeting with Census Bureau officials to discuss alternative sources of data that could meet DOJ's VRA enforcement needs. Going beyond the Administrative Record, in short, confirms that the decision to include a citizenship question runs afoul of the APA.

The analysis of the Enumeration Clause claim similarly involves evidence beyond the four corners of the Administrative Record. As a general proposition, the decision to include a specific question on the census is committed to the discretion of the Commerce Secretary and does not implicate the constitutional command that all persons in each state be counted every ten years. However, if the Secretary's decision to include a question affirmatively interferes with the actual enumeration and fulfills no reasonable governmental purpose, it may form the basis for a cognizable Enumeration Clause challenge.

The evidence admitted in the trial of these actions demonstrates that a significant differential undercount, particularly impacting noncitizen and Latino communities, will result from the inclusion of a citizenship question on the 2020 Census, compounded by macro-environmental factors arising out of the national immigration debate. Efforts to ameliorate these effects through Non-Response Follow-Up ("NRFU"), the evidence showed, would not remediate and could in fact exacerbate the differential undercount of noncitizens and Latino persons. While a citizenship question had been included in the decennial census in 1950 and before, the analysis now must turn on the impact of that question on the prospect of achieving the central constitutional purpose of an actual enumeration in 2020. Viewed through that lens, the inclusion of the question is contrary to the Constitution.

Plaintiffs in each of these actions satisfied their burden of demonstrating standing under Article III of the Constitution. The State of California demonstrated that it will suffer a loss of federal funding and face a substantial risk of losing political representation directly traceable to the inclusion of the citizenship question on the census. California established that the inclusion of this question will also require the expenditure of additional funds to attempt to mitigate the effects of the question and minimize the resulting undercount of California vis-à-vis other states. Similarly, the City of San Jose and the Black Alliance for Just Immigration ("BAJI") each established injury directly flowing from the addition of the citizenship question. In the case of San Jose, it showed the negative impact on federal funding it would receive for various programs dependent on census data and the additional resources that would be required to attempt to mitigate those effects. As to BAJI, the organization demonstrated that it will be obliged to commit additional time and resources to address the specific effects of the citizenship question on its constituents and to encourage them to participate despite the perceived risks.

In light of the statutory and constitutional violations outlined above, the issue becomes the appropriate remedy. With respect to the APA claim, consistent with and for the reasons stated in New York v. United States Dep't of Commerce , No. 18-cv-2921 (S.D.N.Y. Jan. 15, 2019) (the "New York matter"), vacatur of the Secretary's decision, remand to the agency, and an injunction against inclusion of the citizenship question on the 2020 census is warranted *976and will be ordered. As to the Enumeration Clause violation, an injunction is the proper relief. The Department of Commerce urges that any relief should be limited to the particular plaintiffs before the Court. While mindful of the concerns regarding individual district courts issuing orders of national scope, the limitation advanced by Defendants here is simply impractical in light of the nationwide nature of the questionnaire at issue. Accordingly, no such limitation will be included in the injunctive relief ordered by this Court.

II. BACKGROUND

This action, comprising two related cases, arises from the U.S. Census Bureau's decision to include a question regarding citizenship status on the 2020 Census questionnaire. Plaintiffs in Case No. 18-cv-1865 are the State of California, the County of Los Angeles, the City of Los Angeles, the City of Fremont, the City of Long Beach, the City of Oakland, the City of Stockton, and the Los Angeles Unified School District (collectively, "California Plaintiffs"). Plaintiffs in Case No. 18-cv-2279 are the City of San Jose and the Black Alliance for Just Immigration ("BAJI") (collectively, "San Jose Plaintiffs"). Defendants in both matters are Wilbur L. Ross, Jr., in his official capacity as Secretary of the U.S. Department of Commerce; the U.S. Department of Commerce; Stephen Dillingham, in his official capacity as Director of the United States Census Bureau; and the U.S. Census Bureau. During the relevant period Dr. Ron Jarmin served as Acting Director of the United States Census Bureau.

On March 26, 2018, Secretary Ross issued a memorandum (the "Decision Memo") directing the Census Bureau to add a question on citizenship status to the 2020 Census. PTX-1 at 1313-20. Plaintiffs contend the decision to include this question violated the Constitution and the APA. They specifically argue Secretary Ross's decision violated the Enumeration Clause, U.S. Const. art. I, § 2, cl. 3, and was "arbitrary, capricious, [and] otherwise not in accordance with law" under the APA, 5 U.S.C. § 706(2)(A).² The San Jose Plaintiffs also allege a violation of the Apportionment Clause of the Constitution. U.S. Const. amend. XIV, § 2.

The process by which the decennial census is taken has changed significantly over the years, as have the questions asked in the census instrument. From 1790 to 1960, the Bureau collected data directly from households through in-person interviews. Undisputed Fact ("UF") 75.³ Moreover, from 1820 to 1950, with the exception of 1840, respondents were asked a question concerning citizenship or birthplace. UF 67. The Census Bureau subsequently transitioned to a mailed questionnaire, which involved sending a "short form" questionnaire to most residences, and a "long form" questionnaire with significantly more questions to the remaining households. UF 77-78.

The long form questionnaires used in 1970, 1980, 1990, and 2000 included a question about citizenship status, whereas the short form questionnaires did not. UF 80. After the 2000 Decennial Census, the functions performed by the long form questionnaire were replaced by the American Community *977Survey ("ACS"). UF 83. The ACS is a yearly survey of approximately 2% of households-about 3.5 million-across the United States. UF 85. A question concerning citizenship status currently appears among more than 50 questions on the ACS questionnaire. UF 86.

In keeping with recent practice, the 2020 Census will be "short form only." UF 102. The ACS will continue to be distributed as usual and will continue to include a citizenship question. UF 103. Per Secretary Ross's Decision Memo, the 2020 Census will also include a citizenship question. The text of this newly added question will read, "Is this person a citizen of the United States?," with the answer options "Yes, born in the United States"; "Yes, born in Puerto Rico, Guam, the U.S. Virgin Islands, or Northern Marianas"; "Yes, born abroad of U.S. citizen parent or parents"; "Yes, U.S. citizen by naturalization - Print year of naturalization"; and "No, not a U.S. citizen." UF 104. As in past years, the 2020 Census questionnaire will also pose questions regarding sex, Hispanic origin, race, and relationship status. UF 106.⁴

III. STANDING

A. Legal Standard

In order to establish standing under Article III of the Constitution, a "plaintiff must have (1) suffered an injury in fact, (2) that is fairly traceable to the challenged conduct of the defendant, and (3) that is likely to be redressed by a favorable judicial decision." Spokeo, Inc. v. Robins , --- U.S. ----, 136 S.Ct. 1540, 1547, 194 L.Ed.2d 635 (2016) (citing Lujan v. Defs. of Wildlife , 504 U.S. 555, 560-61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992) ). As the party invoking federal jurisdiction, the plaintiff bears the burden of establishing all three requirements by a preponderance of the evidence. Lujan , 504 U.S. at 561, 112 S.Ct. 2130.

B. Findings of Fact Related to Standing ⁵

1. Inclusion of the Citizenship Question on the 2020 Census Will Cause a Differential Decline in Self-Response Rates

1. Undisputed evidence in this case shows that adding a citizenship question to the 2020 Census will cause a differential decline in self-response rates for noncitizen and Hispanic households.

2. Defense expert Dr. John Abowd, Chief Scientist and Associate Director for Research and Methodology at the Census Bureau, testified credibly that the Census Bureau has produced quantitative evidence that adding a citizenship question to the 2020 Census will lower self-response rates. Tr. 797:20-25 (Abowd). He specifically endorsed the Census Bureau's finding that the citizenship question will lead to a lower self-response rate in both noncitizen and Hispanic households, New York Tr. 881:19-882:1 (Abowd); id. at 918:3-919:1 (Abowd), and that this lower response rates will harm the quality of census data, id. at 882:2-5 (Abowd).⁶ The Plaintiff experts *978similarly endorsed these conclusions. See Part III.B.1.e, infra .

a. December 22 Census Bureau Memo

3. The analysis that underpins Dr. Abowd's testimony is set forth in three memoranda issued by the Census Bureau: the December 22 Memo, PTX-148; the January 19 Memo, PTX-22; and the Brown, et al. Memo, PTX-160. New York Tr. 896:7-15 (Abowd); Census Bureau 30(b)(6) Dep. Vol. II 353:2-6, 353:19-21.

4. The December 22, 2017 Memo was authored by senior professional staff at the Census Bureau (nicknamed the "SWAT Team"). They found that, based on a comparison of self-response rates to the 2010 Census and the 2010 ACS (which included a citizenship question), noncitizen households were 5.1 percent less likely than all-citizen households to respond to a survey with a citizenship question. PTX-103 at 6-7; PTX-148 at 6-7. This finding is "consistent with citizenship questions being more sensitive for households with noncitizens." PTX-103 at 7; PTX-148 at 7.

b. Dr. Abowd's January 19 Memo

5. Dr. Abowd's January 19, 2018 Memo conveyed the 5.1 percent differential self-response estimate to Secretary Ross. PTX-22 at 4.

6. This finding was the result of just one of the "[t]hree distinct analyses" in the January 19 Memo that "support the conclusion of an adverse impact on self-response and, as a result, on the accuracy and quality of the 2020 Census."Id. The other two analyses focused on indicators that suggest that Hispanic households are disproportionately less likely to respond to a survey with a citizenship question.

7. The first of these anlyses focused on the item nonresponse rates-the rate at which respondents do not answer a particular survey question. New York Tr. 905:10-24 (Abowd). The Census Bureau found that item nonresponse rates for the citizenship question on the ACS were more than twice as high for Hispanics as for non-Hispanic whites from 2013 through 2016, and that the nonresponse rate for Hispanics increased by 2.5 percent relative to non-Hispanic whites over that span. PTX-22 at 4; see also New York Tr. 906:12-908:6 (Abowd); Tr. 156:4-157:19 (O'Muircheartaigh).

8. The second analysis considered breakoff rates-the rate at which respondents stop completing a survey when presented with a particular question. New York Tr. 913:13-24 (Abowd). It found that the breakoff rate for the citizenship question on the 2016 ACS was more than eight times higher for Hispanics than for non-Hispanic whites. PTX-22 at 5; see also New York Tr. 914:5-8 (Abowd); Tr. 158:4-21 (O'Muircheartaigh). Similarly, the breakoff rate for three related questions on immigration status (citizenship, place of birth, and year of entry) on the 2016 ACS was more than three times higher for Hispanics than for non-Hispanic whites. PTX-22 at 5; see also New York Tr. 915:9-13 (Abowd).

9. Based on the Census Bureau's analysis of item nonresponse rates and breakoff rates, Dr. Abowd testified credibly that a citizenship question would be sensitive for Hispanics, and that the sensitivity of the question is increasing for Hispanics (but not for non-Hispanic whites). New York Tr. 917:4-918:2 (Abowd).

c. Brown, et al. Memo

10. The Brown, et al. Memo builds upon and updates the analysis in Dr. Abowd's January 19 Memo. New York Tr. 896:7-12 (Abowd). This memo represents the Census Bureau's best analysis of the consequences of adding a citizenship question to the 2020 Census. Census Bureau 30(b)(6)

*979Dep. Vol. II 355:15-356:15; New York Tr. 897:4-15 (Abowd).

11. The Brown, et al. Memo summarized its findings as follows:

This paper's examination of several Census Bureau surveys with and without citizenship questions suggests that households that may contain noncitizens are more sensitive to the inclusion of citizenship in the questionnaire than all-citizen households. The implication is that adding a citizenship question to the 2020 Census would lead to lower self-response rates in households potentially containing noncitizens, resulting in more nonresponse follow-up (NRFU) fieldwork, more proxy responses, and a lower-quality population count.

PTX-160 at 54.

12. The memo presented data showing that citizenship-related questions are more sensitive for Hispanics and that, because Hispanics have higher rates of nonresponse for citizenship than for sex or age, they could be disproportionately impacted by adding a citizenship question to the 2020 Census questionnaire. PTX-160 at 7-10.

13. The Census Bureau also updated the estimated 5.1 percent differential decline in the self-response rate of noncitizen households to 5.8 percent. PTX-160 at 39; Census Bureau 30(b)(6) Dep. Vol. II 372:2-12; New York Tr. 897:16-20 (Abowd); Tr. 161:13-21 (O'Muircheartaigh). This revised estimate is the result of a natural experiment that compared response rates on the 2016 ACS, which included a citizenship question, to response rates on the 2010 Census, which did not incorporate a citizenship question, and then compared the change in response rates between all-citizen households and all other households (i.e., households that contain or may contain one or more noncitizens). PTX-160 at 33-34; Census Bureau 30(b)(6) Dep. Vol. II 373:9-15, 374:10-16; New York Tr. 898:2-899:6 (Abowd); Tr. 161:22-164:17 (O'Muircheartaigh).

14. The 5.8 percent figure represents the Census Bureau's best conservative estimate of the differential effect of the citizenship question on noncitizen household self-response. New York Tr. 894:17-895:2, 897:9-12 (Abowd).

15. The Brown, et al. Memo emphasized that the 5.8 percent estimate is "conservative." PTX-160 at 39; New York Tr. 900:21-25 (Abowd); Tr. 164:21-24 (O'Muircheartaigh). The Bureau acknowledged that this figure may underestimate the impact of the citizenship question on census self-response rates for two reasons: (1) the question will be more prominent on the 2020 Census questionnaire, which has just ten other questions, than it was on the ACS questionnaire, which has 75 questions, PTX-160 at 39; New York Tr. 901:22-902:10 (Abowd); Tr. 164:25-165:14 (O'Muircheartaigh), and (2) given "the level of concern about using citizenship data for enforcement purposes," the macro-environment at the time of the 2020 Census may be worse than it was when the ACS data were collected, PTX-160 at 39; see also New York Tr. 902:11-24 (Abowd); Tr. 165:15-21 (O'Muircheartaigh).

16. The 5.8 percent estimate is also conservative because of limitations in the design of the natural experiment. For example, the natural experiment assumed that individuals whose citizenship information was missing from administrative records were citizens, which had the effect of reducing the estimated difference between the response rates of all-citizen households and noncitizen households. Tr. 165:25-166:15 (O'Muircheartaigh).

17. The Brown, et al. Memo also confirmed the findings in Dr. Abowd's January 19 Memo showing that (1) Hispanics *980were more than twice as likely as non-Hispanic whites to skip the citizenship question on the ACS and that the differential in such item nonresponse rates increased between 2013 and 2016, and (2) the breakoff rate for the citizenship question on the 2016 ACS was more than eight times higher for Hispanics than for non-Hispanic whites. PTX-160 at 8-11. Based on this data, the Census Bureau concluded that Hispanics are more sensitive to survey questions about citizenship than they were a few years ago but that non-Hispanic whites are not. Census Bureau 30(b)(6) Dep. Vol. II 369:1-19. This suggests that nonresponse rates to the citizenship question on the 2020 Census will be higher for Hispanics than for non-Hispanic whites. New York Tr. 910:7-13, 914:9-12 (Abowd).

18. Recent Census Bureau data show that the differential breakoff is escalating. After the January 19 Memo and the Brown, et al. Memo were issued, the Census Bureau made the 2017 ACS breakoff data publicly available. New York Tr. 915:19-916:3 (Abowd). That data, which was reviewed by the SWAT team, showed that the breakoff rate for the citizenship question on the 2017 ACS is now twelve times higher for Hispanics than for non-Hispanic whites. New York Tr. 916:4-917:3 (Abowd).

d. CSM Memo and CBAMS Results

19. Recent Census Bureau qualitative research suggests that the citizenship question will cause an even greater differential decline in self-response rates than estimated by Brown, et al. The macro-environment, particularly the political environment around immigration, has the potential to amplify the negative effect of the citizenship question on self-response rates. New York Tr. 926:21-927:10 (Abowd).

20. This research includes the Census Bureau's Center for Survey Measurement ("CSM") focus group testing in 2017, which revealed increased concern among immigrants about the confidentiality of their survey responses, PTX-157 at 1, and the Census Barriers, Attitudes, and Motivators Study (CBAMS) conducted in 2018, which revealed concerns among Spanish language respondents about the citizenship question, PTX-153 at 21-22.

i. CSM Findings

21. CSM researchers summarized the respondent confidentiality concerns they observed in a September 20, 2017 memo for the Associate Directorate for Research and Methodology at the Census Bureau, PTX-157, and in presentations of their findings to the American Association of Public Opinion Research (AAPOR), PTX-158, and to the National Advisory Committee on Racial, Ethnic, and Other Populations, PTX-326.

22. During the pretesting studies conducted in 2017, CSM researchers "noticed a recent increase in respondents spontaneously expressing concerns about confidentiality" and "reported that respondents' fears, particularly among immigrant respondents, have increased markedly this year." PTX-157 at 1.

23. For example, CSM researchers observed Spanish-speaking respondents who were "uncomfortable 'registering' other household members," who "left three or four roomers off the roster" and "mentioned being worried because of their '[immigration] status,' " and who stated that "the Latino community will not sign up because they will think that Census will pass their information on and people can come looking for them." Id. at 2.

24. CSM researchers observed that "this level of deliberate falsification of the household roster, and spontaneous mention of concerns regarding negative attitudes toward immigrants, is largely unprecedented in the usability interviews *981that CSM has been conducting since 2014 in preparation for the 2020 Census." Id. at 3. CSM researchers worried that the concerns expressed by immigrant respondents might be "even more pronounced" during the 2020 Census, because respondents are generally more willing to participate in pretesting surveys "given that they are being paid a cash incentive for their participation and [are] being interviewed by a researcher with whom they have established rapport." Id.

25. During focus group testing, respondents similarly expressed "fear of deportation[ ] [and] concern about how the data are used[ ] and which agencies can see it," specifically asking whether the Department of Homeland Security ("DHS") or Immigration and Customs Enforcement ("ICE") would have access to their data. PTX-326 at 9.

ii. CBAMS Findings

26. The CBAMS is a survey of 50,000 households in a series of 42 focus groups designed to inform the integrated partnership and communications program for the 2020 Census about the macro-environment. Census Bureau 30(b)(6) Dep. Vol. II 437:17-438:6; New York Tr. 927:22-928:6 (Abowd). The Census Bureau finds CBAMS research sufficiently reliable to provide actionable information for the integrated partnership and communications program. Census Bureau 30(b)(6) Dep. Vol. II 438:7-11.

27. After Secretary Ross announced that the 2020 Census would include a citizenship question, Census Bureau researchers began asking for feedback about the question from 30 of the 42 focus groups, including Spanish-language groups. PTX-161 at 6; see also New York Tr. 930:16-19 (Abowd).

28. The CBAMS found that in the Spanish-language (U.S. Mainland) focus groups, the citizenship question was a "determining factor for participation." PTX-153 at 22. Although most participants said that they were not afraid to answer the citizenship question because they are citizens or legal residents, they knew many others who would not participate in the 2020 Census "out of fear." Id. While all participants wanted to participate in the 2020 Census, "fear of deportation outweighs any benefit." Id.

29. The Census Bureau views the results of the Spanish-language focus groups with respect to the citizenship question as "extremely problematic." Census Bureau 30(b)(6) Dep. Vol. II 450:16-451:1; New York Tr. 934:8-12 (Abowd). Other immigrant and non-white groups raised similar concerns. New York Tr. 930:9-24, 938:22-939:17, 940:4-941:14 (Abowd).

30. Census Bureau researchers ultimately concluded that "[t]he citizenship question may be a major barrier" to participation in the 2020 Census because respondents, including citizens and legal residents, believed that the census's purpose "is to find undocumented immigrants" and because "[t]he political discourse is targeting their ethnic group." PTX-465 at 43.

31. The CBAMS results suggest that the citizenship question is sensitive in the current macro-environment and is a "major concern" for the Census Bureau's efforts to encourage participation in the 2020 Census within Hispanic communities. New York Tr. 944:7-24 (Abowd). Moreover, the increased sensitivity to the citizenship question that was observed in the 2018 CBAMS results was likely not captured in Brown, et al.'s 5.8 percent estimate, which was based on 2016 data. Id. at 944:25-945:4 (Abowd).

e. The Plaintiff Experts' Testimony

32. The Plaintiff experts' testimony further supports the conclusion that the citizenship question will cause a greater differential *982decline in self-response rates than estimated by Brown, et al.

33. Dr. Colm O'Muircheartaigh, professor in the Harris School of Public Policy and senior fellow at the National Opinion Research Center (NORC) at the University of Chicago, testified that he agrees with the Census Bureau research discussed above. Tr. 33:4-17; 145:15-166:15 (O'Muircheartaigh). Dr. O'Muircheartaigh also cited additional factors that will exacerbate the effects of the differential decline in self-response rates caused by the citizenship question on the ultimate enumeration. Id. at 166:16-174:20 (O'Muircheartaigh).

34. First, missing units in the Census Bureau's Master Address File (MAF) contain a disproportionate number of immigrant and noncitizen households. Tr. 166:21-25 (O'Muircheartaigh). The MAF is the "first building block" of census data collection. Id. at 122:4-6. The MAF is constantly updated throughout the census-taking process. Tr. 803:23-805.7 (Abowd). In general, the census is unlikely to count persons whose households do not appear on the MAF. Id. at 46:1-6 (O'Muircheartaigh). Dr. O'Muircheartaigh testified that social science research, including recent research on Mexican immigrants, has observed that the Census Bureau has particular difficulty identifying household addresses for immigrants and noncitizens. Id. at 122:7-123:13, 124:7-17 (O'Muircheartaigh). To the extent that immigrant and noncitizen households are not identified by the Census Bureau and included in the MAF, and the residents of these households choose not to come forward to be counted because of the citizenship question, such households and their residents will not be included in the 2020 Census despite the Census Bureau's NRFU efforts. Id. at 166:21-167:14 (O'Muircheartaigh).

35. Second, respondents, especially those that live in households containing noncitizens, may omit certain household members on the census questionnaire because of fears generated by the citizenship question. Id. at 167:15-20 (O'Muircheartaigh). In particular, the 2017 CSM research observed that Spanish-speaking respondents were reluctant to provide a complete roster of household members. Id. at 147:18-148:16 (O'Muircheartaigh) (citing PTX-157). Dr. O'Muircheartaigh testified that such rostering omissions are a particularly problematic form of nonresponse because "[t]he quality of the census is fundamentally dependent on complete rostering of individuals within households," and "the census protocol has no mechanism for remediating such a response." Id. at 147:10-16, 148:8-149:9 (O'Muircheartaigh); Census Bureau 30(b)(6) Dep. Vol. II 397:19-399:2, 459:21-460:7.

36. Dr. O'Muircheartaigh credibly testified to the following conclusions relating to the impact of the citizenship question on self-response: (1) current survey methodology research, primarily by the Census Bureau, has observed that Latinos and immigrants hold considerable fears about participating in the 2020 Census, (2) the citizenship question will increase the Census Bureau's misidentification of households as unoccupied, particularly among Latinos and households with noncitizens, (3) the citizenship question will depress self-response rates, particularly for Latinos and households with noncitizens, and the Census Bureau's conservative estimate is that the self-response rate for households containing a noncitizen will be 5.8 percent lower than for all-citizen households, and (4) factors such as rostering errors will exacerbate the difference in the effective self-response rates of noncitizens versus citizens. Tr. 175:1-19 (O'Muircheartaigh).

*98337. Dr. Matthew Barreto, a professor of political science and Chicano studies at the University of California, Los Angeles, Tr. 366:13-17 (Barreto), similarly testified that adding a citizenship question to the 2020 Census will reduce self-response rates, particularly among immigrants and Latinos, Id. at 374:7-15 (Barreto). Dr. Barreto's findings were based on a comprehensive literature review of research publications and reports, including those produced by the Census Bureau, related to response rates (as well as NRFU and imputation); an original survey he fielded in which he asked people about whether they intend to participate in the 2020 Census; and his expertise and years of experience implementing surveys in Latino and immigrant communities. Id. at 375:18-376:4, 379:19-380:7 (Barreto) (citing PTX-499).

38. Dr. Barreto identified three interrelated factors that affect survey participation: (1) trust, (2) sensitive questions, and (3) the macro-environment in which the survey is administered. Tr. 380:19-381:7, 383:13-16 (Barreto). Applying the literature on these factors to the citizenship question, Dr. Barreto concluded that the citizenship question will cause a significant decline in self-response rates on the 2020 Census because it is a sensitive question that will exacerbate trust issues in the current macro-environment, particularly for immigrants and immigrant-adjacent communities. Id. at 386:21-25, 411:5-14 (Barreto). Dr. Barreto defined "immigrant-adjacent communities" as communities with mixed-status households, where one family member is a U.S. citizen and another family member is not, and communities in which residents would interact with immigrants daily at work, school, or in other similar environments. Id. at 387:1-14 (Barreto).

39. A consistent finding in the social science research is that "if a potential respondent does not trust the survey taker to keep their information confidential and not put them at risk, then the survey respondent won't participate in the survey at all." Tr. 381:17-23 (Barreto). With regard to census participation specifically, Dr. Barreto observed that the Census Bureau, particularly in Manuel de la Puente's ethnographic studies of the 1990 and 2000 Censuses, found that "immigrant and undocumented populations in particular [ ] don't trust the federal government to fully protect or keep in confidence their information." Id. at 385:3-19, 390:12-395:7 (Barreto) (citing PTX-308 and PTX-309), see also id. at 388:11-389:8 (Barreto) (citing PTX-339). To break down the barriers he observed in his studies, Dr. de la Puente recommended that the Census Bureau work with community groups to assure them that the Census Bureau isn't seeking information about respondents' citizenship status. Id. at 393:25-394:15 (Barreto).

40. Moving on to the second factor, Dr. Barreto testified that "a sensitive question is one that asks a respondent for some very personal information that they may be uncomfortable revealing." Tr. 383:2-6 (Barreto). Social science research suggests that survey takers should "reduce unnecessary sensitive questions because they do create considerable trust issues with respondents." Id. at 383:10-12 (Barreto). Whether a question is sensitive varies in different environments and contexts and across subpopulations. Id. at 383:13-20, 384:19-385:2 (Barreto). Dr. Barreto observed that the citizenship question is likely to be most sensitive to "those who are closer to the immigrant experience or closer to [ ] immigrant communities," particularly "in the Latino community where there have been concerns over immigration-related issues over the past few years." Id. at 387:15-23 (Barreto).

*98441. The third factor, macro-environment, is "the context in which any survey is being implemented," including "the social and political environment, the atmosphere that is present when the survey is being administered." Id. at 395:11-19 (Barreto). A respondent "may be more willing to participate if the context or the environment seems very agreeable and welcoming, and they may be far less likely to participate if the environment seems threatening or concerning." Id. at 395:20-25 (Barreto). Dr. Barreto observed that social science research has found that "[i]mmigrants and mixed-status households are likely to avoid government contact when they suspect it is unsafe to participate." Id. at 397:19-398:2 (Barreto). This observation holds true for a census with a citizenship question, because the question will be asked in a macro-environment that is perceived by many immigrants to be "threatening or negative." Id. at 396:3-13 (Barreto).

42. To evaluate participation in the 2020 Census, Dr. Barreto conducted a large national survey that inquired about people's attitudes and behaviors. Id. at 411:15-23 (Barreto). Within the scientific community, survey research is considered reliable and has predictive value. Id. at 414:2-7 (Barreto).

43. Dr. Barreto conducted his survey on a sample of 6,309 respondents from across the United States, including oversamples of Latinos nationwide and residents of the State of California, the City of San Jose, and two border counties in Texas. Id. at 424:8-19 (Barreto). Respondents were randomly chosen, and weighting was applied to balance out the demographic characteristics of the sample. Id. at 415:19-418:12, 434:2-435:19 (Barreto). In addition, the survey response rate-28.1 percent-was within the American Association of Public Opinion Research (AAPOR) response rate guidelines (at least 20 to 30 percent) for telephone surveys. Id. at 425:2-23 (Barreto).

44. Dr. Barreto set forth the results of his survey in a number of tables. PTX-499A, PTX-863 through PTX-890. He estimates that, because of the citizenship question, census response rates are likely to decline between 6.3 and 8.0 percent nationally and between 10.5 and 14.1 percent in the State of California. PTX-870; PTX-871; see also Tr. 457:17-458:3 (Barreto) (explaining PTX 870); id. at 461:9-20 (Barreto) (explaining PTX-871). The nonresponse rate attributed to the citizenship question in California is statistically higher than the nationwide average. Tr. 463:8-464:15 (Barreto) (explaining PTX-873).⁷

45. Based on the Census Bureau's most current data, the average Latino household is larger than the average non-Latino household. Id. at 1036:12-1037:6 (Abowd). By factoring in the difference in average household size between Latino households and other households, Dr. Barreto estimated that Latinos would constitute approximately 35 percent (over 10 million) of the total number of persons (approximately 28 million) that would not self-respond to the 2020 Census because of the citizenship question, far surpassing the rate of Latinos in the national population (18 percent). Id. at 478:1-481:6 (Barreto) (explaining PTX-880 and PTX-881). This evidence further supports the conclusion that Latinos will be disproportionately affected by the *985citizenship question. Id. at 480:9-14 (Barreto).

46. Although Dr. Barreto's study provides credible evidence that the inclusion of the citizenship question on the 2020 Census is likely to cause a decline in self-response rates among certain demographic groups relative to the rest of the population, some aspects of the survey design and methodology limit the weight the Court affords to this evidence. In particular, Dr. Barreto asks in Question 2 whether respondents would participate in the 2020 census if the federal government were to include a citizenship question on the questionnaire. Tr. 576:4-13.⁸ By contrast, Question 1 specifically referred to the Census Bureau as the agency responsible for the census and asked respondents, without mentioning the citizenship question, whether they would participate in the 2020 census. Id. at 440:13-441:2.⁹ It is plausible that respondents who are more distrustful of the federal government writ large than they are of the Census Bureau in particular may have responded negatively to Question 2 in part because of this difference in terminology. While this hardly represents a glaring flaw, it does diminish somewhat the weight to be afforded to the drop in willingness to respond between Question 1 and Question 2.

47. Defendants' remaining arguments that the decline in self-response rates between Question 1 and Question 2 should not be credited are unpersuasive. In particular, the fact that a randomized controlled trial ("RCT") may produce more accurate results than a survey does not automatically render Dr. Barreto's survey unreliable. See Tr. 874:10-19.

48. Ultimately, respondents' increased reluctance to participate in the census between Question 1 and Question 2 of Dr. Barreto's survey provides credible evidence that the addition of the citizenship question is likely to result in a significant decline in self-response rates in California and within the Latino population relative to the public at large.

2. NRFU Will Not Remediate the Differential Decline in Self-Response Rates

49. In keeping with recent practice, the Census Bureau will implement a series of NRFU operations to attempt to count the significant number of persons who do not self-respond to the 2020 Census, UF 39-47, including the millions who will not self-respond because of the citizenship question, PTX-22 at 6; PTX-160 at 42; New York Tr. 894:1-16 (Abowd). All available evidence indicates that at every NRFU stage, including the imputation phase, the Census Bureau will be differentially less effective at counting noncitizens and Latinos-the very subpopulations most likely not to respond to the 2020 Census because *986of the citizenship question. Tr. 175:20-218:6 (O'Muircheartaigh).

a. Hard-to-Count Populations

50. The Census Bureau has always struggled to count hard-to-count subpopulations, including noncitizens and Latinos, even when the census count for the national population has been fairly accurate. Tr. 57:17-60:8 (O'Muircheartaigh). For example, as measured in the Census Bureau's post-enumeration surveys, Hispanics have been differentially undercounted compared to non-Hispanic whites in each of the last three censuses. Id. at 55:2-15, 56:11-57:5 (O'Muircheartaigh); UF 61-62. In the 2010 Census, Hispanics were undercounted by 1.54 percent and non-Hispanic whites were overcounted by .84 percent, resulting in a net differential undercount of Hispanics of 2.38 percent. Tr. 56:11-24; PTX-211 at 18. In the 2000 Census, Hispanics were undercounted by .71 percent and non-Hispanic whites were overcounted by 1.13 percent, resulting in a net differential undercount of Hispanics of 1.84 percent. PTX-211 at 18. In the 1990 Census, Hispanics were undercounted by 4.99 percent and non-Hispanic whites were undercounted by .68 percent, resulting in a net differential undercount of Hispanics of 4.31 percent. Id.

51. Hard-to-count subgroups include low-income persons, persons who do not live in traditional housing, persons who do not speak English fluently or have limited English proficiency, persons who have distrust in the government, racial and ethnic minorities, renters, undocumented immigrants or recent immigrants, and young children. Tr. 1021:19-1023:2 (Abowd); UF 59-60. Census Bureau research shows that there is "substantial overlap" between these hard-to count subgroups and those households most likely not to respond to the 2020 Census because of the citizenship question. Tr. 1023:3-7 (Abowd).

52. The Census Bureau has identified four primary obstacles to counting hard-to-count subpopulations: that they are hard to locate, hard to contact, hard to persuade, and hard to interview. Id. at 1023:8-24 (Abowd). For some hard-to-count subgroups, more than one of these obstacles applies. Id. at 1024:7-13 (Abowd). Census Bureau research acknowledges that these obstacles apply to those households most likely not to respond to the 2020 Census because of the citizenship question. Id. at 1023:25-1024:6 (Abowd).

b. The Census Bureau's Partnership and Communications Program

53. The Census Bureau has developed a range of strategies to address the net differential undercount of "hard-to-count" populations-including targeted marketing and outreach efforts, partnerships with community organizations, deployment of field staff to follow up with individuals who do not respond, and retention of staff with foreign language skills. UF 64.

54. In the 2000 and 2010 Censuses, the Census Bureau designed and implemented public advertising campaigns to reach hard-to-count immigrant communities, including using paid media in over a dozen different languages to improve responsiveness, and partnered with local businesses, faith-based groups, community organizations, elected officials, and ethnic organizations to reach these communities and improve the accuracy of the count. UF 65-66.

55. Defendants believe that a similar integrated partnership and communications campaign, in tandem with the Census Bureau's NRFU efforts, may mitigate the decline in self-response rates in the 2020 Census. Tr. 798:6-12, 799:21-800:14 (Abowd). Yet there is no evidence in the Administrative Record that Defendants' planned integrated partnership and communications campaign for the 2020 Census will significantly mitigate such a differential *987decline in self-response rates. Dr. Abowd agreed that it is "highly unlikely" that the integrated partnership and communications campaign can eliminate the negative effects of adding a citizenship question. Id. at 980:3-11 (Abowd).

56. The Census Bureau also acknowledges that the "trusted partners" that it relies on to convey the importance of participating in the census will have additional challenges communicating that message if the 2020 Census includes the citizenship question. Census Bureau 30(b)(6) Dep. Vol. II 451:21-452:4, 453:2-17; New York Tr. 937:16-23 (Abowd). The CBAMS focus groups of Spanish-speaking respondents found that, "while there were suggestions of trusted voices, there does not seem to be a single trusted voice that could mitigate [respondents'] distrust of the government to uphold the promise of confidentiality." PTX-153 at 22. Dr. O'Muircheartaigh persuasively testified that this observation shows the citizenship question will "reduce[ ] the potential impact of the positive input of constituency, community, and association leaders" as these trusted voices attempt to convince their constituents to participate in the 2020 Census. Tr. 153:1-154:9 (O'Muircheartaigh).

57. Census Bureau research has noted one messaging strategy that is reassuring to Spanish-speaking respondents is to convey that "[n]one of the questions in this survey will ask about immigration status" and that "[b]y law, [the respondent's] answers cannot be shared with Immigrations and Customs Enforcement." PTX-158 at 16. Dr. Barreto similarly observed that, consistent with the findings in Dr. de la Puente's ethnographic studies, the most effective way- indeed, perhaps the "only way"-to address confidentiality concerns related to the citizenship question is "to assure respondents that no citizenship information is being gathered" in the 2020 Census. Tr. 500:17-501:5 (Barreto). Neither the Census Bureau nor trusted partners can offer such assurances because the citizenship question will be on the 2020 Census, unless the Census Bureau is instructed to remove it. Tr. 1052:8-12 (Abowd).

58. Moreover, despite the barriers to participation in the 2020 Census associated with the citizenship question, the Census Bureau has not significantly increased its spending on 2020 census outreach relative to that expended in 2010. Tr. 1024:18-1025:9 (Abowd).

c. The Census Bureau's NRFU Operations

59. The Census Bureau's NRFU workload includes all households that do not initially self-respond to the census. Tr. 851:16-852:2 (Abowd). In the 2010 Census, over 27 percent of the persons enumerated were in the NRFU workload. PTX-211 at 32-33 (subtracting from the U.S. total population (300,703,000) those persons not in any NRFU universe (219,207,000) and dividing by the total population). The NRFU workload for the 2020 Census is expected to rise to between 34.5 and 44.5 percent of the total population. PTX-1 at 172.

60. The Census Bureau's best conservative estimate is that adding a citizenship question to the 2020 Census will increase the NRFU workload by 2.09 million households and 6.5 million persons. PTX-160 at 42.

61. Based on his survey data, Dr. Barreto estimated that adding a citizenship question to the 2020 Census will increase the NRFU workload by at least 28 million persons, and that Latinos will be disproportionately represented in that workload. PTX-880; Tr. 480:5-14 (Barreto).

62. The Bureau's NRFU operations are designed to obtain an accurate count-and thus, to prevent an undercount-at the *988national level. Tr. 918:11-16 (Abowd). In recent censuses, however, the Bureau's NRFU operations have been less effective at counting some subpopulations than others. Tr. 178:7-23 (O'Muircheartaigh).

63. Dr. Abowd testified that he is unaware of any "credible quantitative evidence" that adding a citizenship question will increase the net differential undercount of any subpopulation, after accounting for NRFU operations. Tr. 918:21-24 (Abowd). Dr. Abowd admitted, however, that it is "highly unlikely" that the Census Bureau's NRFU operations will eliminate a differential undercount in the 2020 Census. Id. at 980:12-981:2.

64. The Census Bureau's NRFU operations for the 2020 Census include in-person follow-up enumeration, proxy enumeration, administrative record enumeration, and imputation by other methods. UF 39-46; Tr. 176:13-177:20 (O'Muircheartaigh). The Census Bureau's NRFU operations for the 2010 Census included these same processes, with the exception of administrative record enumeration, which was used only on an experimental basis in 2010. Census Bureau 30(b)(6) Dep. Vol. II 400:19-401:21.

65. The weight of the evidence ultimately shows that these NRFU efforts are unlikely to mitigate significantly the differential decline in self-response caused by the citizenship question and may in fact exacerbate the problem. See Tr. 217:4-218:5 (O'Muircheartaigh).

i. In-Person Follow-Up Enumeration

66. The Census Bureau has repeatedly acknowledged that "[t]hose refusing to self-respond due to the citizenship question are particularly likely to refuse to respond in NRFU as well." PTX-25 at 4; see also PTX-160 at 41, 42 n.59 ("Households deciding not to self-respond because of the citizenship question are likely to refuse to cooperate with enumerators coming to their door....").

67. Although in-person follow-up enumeration is typically more effective than mail solicitation, "in this case for this population, the level of threat embodied by a federal agent arriving at your residence to collect the information is far greater than the threat that might be implied by a piece of paper [ ] that arrives at your residence." Tr. 190:2-10 (O'Muircheartaigh).

68. Given these conditions, the enumeration errors that will result "may not be avoidable simply by spending more money on fieldwork. Once a household decides not to cooperate, it may not be possible to obtain an accurate enumeration no matter how many times an enumerator knocks on their door." PTX-160 at 43 n.60; see also Tr. 190:20-191:21 (O'Muircheartaigh).

69. Recent data from ACS in-person follow-up enumeration efforts, specifically the Computer-Assisted Personal Interviewing (CAPI) operation, underscores the challenges that enumerators will face in the 2020 Census if, like the ACS, the census includes a citizenship question. Census Bureau 30(b)(6) Dep. Vol. I 124:19-133:17; Tr. 178:24-185:19 (O'Muircheartaigh) (describing PTX-138). The data, which was collected between 2010 through 2016, is consistent with the notion that questions on citizenship have become more sensitive since 2010. Census Bureau 30(b)(6) Dep. Vol. I 131:4-11.

70. The CAPI data exhibit the following trends: (1) in-person follow-up enumeration has been less effective over time in all census tracts, (2) in-person follow-up enumeration has been differentially less effective in census tracts with a higher proportion of households containing a noncitizen, and (3) the differential between census tracts with a higher proportion of households containing a noncitizen and census tracts with a lower proportion of households containing a noncitizen has *989grown over time. Census Bureau 30(b)(6) Dep. Vol. I 129:22-130:4, 131:4-18, 133:8-17, Tr. 180:17-181:3 (O'Muircheartaigh).

71. The most recent CAPI data-from 2016-for the half of the population with a higher proportion of households containing a noncitizen indicate that in-person follow-up enumeration was 86.63 percent successful. Tr. 183:21-185:9 (O'Muircheartaigh). This rate "is an approximate representation of how...such households might behave in the context of the census." Id. at 185:10-19 (O'Muircheartaigh). Indeed, the success rate was lower (and conversely, the non-interview rate was higher) for in-person follow-up enumeration in the 2016 End-to-End Test and the 2018 End-to-End Test. Id. at 186:19-187:14 (describing PTX-482 at 26).

72. None of the testing that has been used to plan NRFU staffing levels, the number of field offices, enumerator training, NRFU protocols, or census questionnaire assistance has accounted for a citizenship question on the 2020 Census. Census Bureau 30(b)(6) Dep. Vol. I 198:2-10, 200:9-201:10. Although the Census Bureau's NRFU operations were used in the 2018 End-to End Test, Tr. 819:15-820:9 (Abowd), it did not include a citizenship question, Census Bureau 30(b)(6) Dep. Vol. I 225:13-16; Tr. 820:14-15 (Abowd).

73. The Census Bureau considers the NRFU operations to have been a success in the 2018 End-to-End Test. Tr. 820:19-23 (Abowd). But a U.S. Government Accountability ("GAO") report on NRFU implementation in connection with that test "raises some serious concerns." Tr. 98:3-8 (O'Muircheartaigh) (describing PTX-482).

74. That the Census Bureau did not determine the procedures for late-NRFU data collection until after it started work, for example, "seriously undermines the potential of the activity to be successful." Id. at 98:9-99:7. (O'Muircheartaigh) (describing PTX-482 at 11). This finding, in combination with similar findings that the field workforce was unprepared for certain enumeration challenges, id. at 99:8-100:15 (O'Muircheartaigh), and lacked adequate training, id. at 186:9-18 (O'Muircheartaigh) (describing PTX-482), led Dr. O'Muircheartaigh to conclude that the report was "a little disturbing." Id. at 101:9-12 (O'Muircheartaigh). These findings "cast[ ] doubt on...any projections that the Census Bureau has about how successfully it will operate in 2020, compared, for example, to 2010." Id. at 101:16-102:4 (O'Muircheartaigh).

ii. Proxy Enumeration

75. Locating a proxy respondent-a neighbor, landlord, postal worker, or other knowledgeable person who will provide information about another household-is generally not easy. Tr. 195:2-10 (O'Muircheartaigh). The Census Bureau expects that, just as with in-person follow-up enumeration, in census tracts with a higher proportion of households containing a noncitizen, the proxy enumeration rate will be lower than in other tracts. Census Bureau 30(b)(6) Dep. Vol. II 386:2-15; Tr. 196:25-197:6 (O'Muircheartaigh).

76. In other words, the challenge of finding willing proxy respondents will be greater in neighborhoods with households that are "fearful of the Administration and fearful of Census." Tr. 195:13-25 (O'Muircheartaigh). Potential proxy respondents will be "less likely to want to cooperate" if they are concerned about reporting undocumented immigrants. Id. at 521:15-522:2 (Barreto). Given that "reference persons are much less likely to answer the citizenship question for nonrelatives in the household than for themselves...they may be even less likely to answer it for neighbors." PTX-160 at 43; Census Bureau *99030(b)(6) Dep. Vol. II 386:16-387:10; Tr. 523:3-17 (Barreto).

77. Even if located and willing to provide a response, proxy respondents generally provide lower quality enumeration data than self-responses. Census Bureau 30(b)(6) Vol. II 382:17-21; PTX-22 at 6; Tr. 931:14-24, 951:11-14 (Abowd). For example, in the 2010 Census, 97.3 percent of self-responses resulted in a correct enumeration, but the correct enumeration rate for proxy responses was just 70.2 percent. PTX-160 at 42 (citing PTX-211 at 33); Tr. 197:14-198:5 (O'Muircheartaigh).

78. Proxy responses are particularly inaccurate for persons in tenuous residential arrangements-a subpopulation that is disproportionately made up of Latinos and immigrants. Tr. 198:6-200:4 (O'Muircheartaigh). Because of the nature of these living arrangements-which include, for example, converted garages-proxy respondents "may not actually know how many people live there." Id. at 522:3-8 (Barreto).

79. Census Bureau research has also found that "proxies supply poor quality individual demographic and socioeconomic information about the person on behalf of whom they are responding." PTX-160 at 41-42; Tr. 200:23-201:8 (O'Muircheartaigh); id. at 937:6-19 (Abowd). Dr. Abowd conceded that the increased use of proxy responses "does impact data quality," including the quality of characteristic data. Id. at 887:13-24 (Abowd).

iii. Administrative Record Enumeration

80. Census Bureau research has observed that the quality of administrative records varies depending on the subpopulation. Tr. 204:18-205:3 (O'Muircheartiagh) (describing PTX-288). More specifically, the Bureau is less likely to be able to use administrative records to enumerate hard-to-count subpopulations, including noncitizens and Hispanics. Jarmin Dep. 285:1-286:20; Tr. 948:7-949:12 (Abowd), 205:4-12 (O'Muircheartaigh). Undocumented immigrants are particularly unlikely to be found in administrative records and will be harder to enumerate using such records. Census Bureau 30(b)(6) Dep. Vol. II 391:4-19; Tr. 205:13-17 (O'Muircheartaigh). Accordingly, the Census Bureau does not expect administrative record enumeration to be as successful with noncitizens as with citizens. Census Bureau 30(b)(6) Dep. Vol. II 391:21-392:4.

81. Similarly, the Census Bureau will be unable to link Hispanics to administrative records at as high a rate as it can link non-Hispanic whites. Census Bureau 30(b)(6) Dep. Vol. II 389:12-390:5.

82. Given the inability of the Census Bureau to use administrative records to count the very subpopulations most likely not respond to the 2020 Census because of the citizenship question, administrative record enumeration will not remediate the differential decline in self-response rates and may indeed exacerbate any differential undercount of noncitizens and Latinos. Tr. 206:4-19 (O'Muircheartaigh).

iv. Imputation

83. If the Census Bureau is unable to enumerate a household through other NRFU operations, it will impute, or model, the number of persons in the household and their characteristics. Tr. 942:17-20 (Abowd). In the decennial census, the Bureau uses "count imputation" to impute the size of the household, and "whole-person imputation" to impute both the size of the household and the characteristics of the people in the household. Id. at 892:10-15 (Abowd); PTX-22 at 5.

84. The Census Bureau concedes that whole-person imputations "are not very accurate." Census Bureau 30(b)(6) Dep. Vol. I 253:7-15.

*99185. The Census Bureau anticipates that there will be 1.477 million more whole-person imputations in the 2020 Census because of the citizenship question. PTX-160 at 42-43.

86. The Census Bureau has not finalized the algorithms it will use for count imputation in the 2020 Census. Tr. 892:16-19 (Abowd). The accuracy of the Census Bureau's imputation model "is unknown at this time." PTX-160 at 44. The Census Bureau has recognized that any attempt to use imputation to count nonresponding persons "will be challenging due to the fact that nonresponse is highly correlated with citizenship." Id.

87. As in previous censuses, the Census Bureau expects to use a "hot-deck" imputation model that imputes missing households based on nearby households that the Census Bureau has counted and believes are similar in size, location, and other characteristics. Tr. 892:10-893:11 (Abowd), id at 208:19-209:12 (O'Muircheartaigh).

88. Because hot-deck imputation fills in missing data based on data that the Census Bureau has already collected, it is not neutral; by definition, it over-represents the household characteristics of the known population, in which those most likely not to respond to the citizenship question-in particular, noncitizens and Latinos-are underrepresented. Id. at 210:14-211:6, 211:24-212:6 (O'Muircheartaigh). Dr. Abowd confirmed that hard-to-count subpopulations will be imputed at a greater rate than the rest of the population. Id. at 981:8-13 (Abowd).

89. The Census Bureau's imputation model also fails to account for the larger household size, on average, of Hispanic households compared to other households. Tr. 528:1-24 (Barreto), 1036:25-1037:6 (Abowd). The Census Bureau's imputation model is built on the assumption that household size is "ignorable" missing data-that it is not correlated with nonresponse. Id. at 525:10-13 (Barreto). Given that those persons most likely not to respond to the 2020 Census because of the citizenship question, however, tend to come from larger households, household size is, in fact, "non-ignorable" data. Id. at 528:1-11 (Barreto); see also Tr. 1036:12-1037:6 (Abowd) (noting that, based on the Census Bureau's most current data, the average Latino household is larger than the average non-Latino household). The result is bias in the Census Bureau's imputation model. Id. at 985:10-14 (Abowd), 525:20-25 (Barreto).

90. Based on his survey data, Dr. Barreto presented quantitative evidence that the Census Bureau's imputation model will systematically undercount nonresponding households. Id. at 529:14-530:6 (Barreto). The data reveal that, on a national level, households that will not respond to a census with the citizenship question are larger, on average, than households that will respond, and that in California, the gap between these groups expands. Id. at 529:21-25, 530:7-9 (Barreto) (describing PTX-888 and PTX-889).

91. Dr. Barreto also constructed an imputation model based on the Census Bureau's 2010 imputation model, as described in PTX-344, the Bureau's J-12 memorandum. Like the Census Bureau's imputation model, Dr. Barreto's version predicted the household size of nonresponding households based on their 20 nearest neighbors, with controls for such factors as housing type, geographic proximity, and household demographics. Tr. 535:3-14 (Barreto). Because Dr. Barreto's survey data contained the household size of each nonresponding household, he was able to compare the imputation model's predicted household size to the actual size of these households. Id. at 535:15-19 (Barreto).

*99292. Dr. Barreto's imputation analysis suggests that the Census Bureau's imputation model is likely to under-impute the household size of Latinos that do not respond to the 2020 Census because of the citizenship question at a rate of three-quarters of a person per household on average, as compared to similarly-situated Latino households that will respond to the census. Id. at 538:22-539:6 (Barreto) (describing PTX-468).

93. This evidence, suggests that imputation will not remediate the differential decline in self-response rates. Id. at 540:24-541:8 (Barreto), 212:9-17 (O'Muircheartaigh). In sum, the relative ineffectiveness of the Census Bureau's NRFU operations with respect to individuals who are unlikely to self-respond as a result of the citizenship question inevitably leads to the conclusion that the NRFU process is unlikely significantly to mitigate the disproportionate effect of the citizenship question on Latino and noncitizen households.

3. Inclusion of the Citizenship Question on the 2020 Census Will Result in a Differential Undercount of Noncitizens and Latinos

94. The weight of the evidence, both qualitative and quantitative, strongly suggests that the citizenship question will cause a net differential undercount of noncitizens and Latinos relative to all-citizen households.

95. The Census Bureau concedes, based on its own natural experiment, that the citizenship question will cause the self-response rate of noncitizen households to decline at least 5.8 percent. Part III.B.1.c, supra . The Census Bureau has also produced considerable qualitative research suggesting that the citizenship question will cause an even larger differential decline in the self-response rate of noncitizen households, and that these negative effects of the citizenship question will extend to other subpopulations, such as Hispanics. Part III.B.1.d, supra .

96. Dr. Barreto produced quantitative evidence that is consistent with the Census Bureau's research. Dr. Barreto's survey results further suggest that the citizenship question will cause a decline in the self-response rate in California that will be greater than the decline in the nation as a whole. Part III.B.1.e, supra .

97. In all recent censuses, the Census Bureau has differentially undercounted hard-to count subpopulations, most notably Hispanics, even after implementing all NRFU operations. Part III.B.2.a, supra .

98. The persons most likely not to self-respond to the citizenship question are also some of the most unlikely to be counted at every NRFU stage-in-person follow-up enumeration, proxy enumeration, administrative record enumeration, and imputation by other methods. Part III.B.2.c, supra.

99. In addition, the Census Bureau's NRFU operations are not designed to count persons that are missing from the MAF or are left off the roster by a family member or proxy respondent. Part III.B.1.e, supra . If such persons do not self-respond to the 2020 Census because of the citizenship question, they will not be counted. Moreover, noncitizens and immigrants, particularly Mexican immigrants, are more likely to live in housing that is missing from the MAF. Part III.E.1.b.

100. In sum, it is more likely than not that the citizenship question will cause a substantial net differential undercount of noncitizens and Latinos.

4. The Citizenship Question Will Harm Data Quality

101. Harm to the quality of census data is something the Census Bureau "tr[ies] to avoid." New York Tr. 953:18-20 (Abowd).

*993It is undisputed, however, that the citizenship question will damage the quality of characteristic data collected through the 2020 Census, separate and apart from the damage to the count. These characteristics include gender, age, race, and ethnicity. Tr. 1001:17-24 (Abowd). The damage to data quality will also cause some people to be counted in the wrong place, including in the wrong area of a municipality, or even in the wrong state. Id. at 1003:5-16 (Abowd).

102. In the January 19 Memo, the Census Bureau concluded that adding a citizenship question to the 2020 Census will have an adverse impact on the quality of the data collected by the census. PTX-22 at 4. Because the citizenship question will lower self-response rates, the NRFU workload will increase, which will "degrade data quality because data obtained from NRFU have greater erroneous enumeration and whole-person imputation rates." Id. at 5. One reason that data quality will suffer is that data collected during NRFU are "much more likely to be collected from a proxy rather than a household member and, when they do come from a household member, that person has less accurate information than self-responders." Id. at 6.

103. In the March 1 Memo, the Census Bureau similarly concluded that a citizenship question will reduce data quality. PTX-25 at 4. The Brown, et al. Memo reached the same conclusion. PTX-160 at 54. Dr. Abowd's testimony confirms that adding the citizenship question will damage the quality of the data collected in the 2020 Census. New York Tr. 885:17-21 (Abowd). Dr. Abowd observed that data produced by lower self-response rates is less accurate than data produced by higher self-response rates. Id. at 881:19-882:5 (Abowd). Likewise, data produced by self-response is much more "reliable" than data produced by NRFU efforts. Id. at 953:2-14 (Abowd); Tr. 942:21-943:2 (Abowd). Therefore, by decreasing self-response rates and thus increasing reliance on NRFU efforts, the citizenship question will reduce the quality and accuracy of data produced during the 2020 Census. New York Tr. 881:19-882:5, 952:23-953:14 (Abowd); Tr. 934:16-935:1, 1001:17-24 (Abowd).

104. Dr. Abowd also acknowledged that inclusion of the citizenship question on the census would result in fewer persons being linked to administrative records, New York Tr. 969:2-23, 979:16-20, 981-17-19 (Abowd), which would reduce data quality, id. at 981:20-25. In contrast, using administrative records to provide DOJ with block-level CVAP data without adding the citizenship question to the census would not harm the quality of the census data. Id. at 958:5-18 (Abowd).

105. Dr. O'Muircheartaigh, Dr. Barreto, and Dr. Habermann confirmed that adding a citizenship question to the 2020 Census will harm the quality of the census data. Tr. 114:11-15, 217:21-22 (O'Muircheartaigh); id. at 492:16-21 (Barreto); PTX-821 at ¶¶ 47-54, 68. This is the consensus among scientists within and outside the Census Bureau. Tr. 114:11-15 (O'Muircheartaigh).

106. The increased degradation of data quality that results from adding a citizenship question to the 2020 Census cannot be mitigated. Id. at 935:3-5, 950:6-13, 1001:25-1002:8 (Abowd).

5. Inclusion of the Citizenship Question on the Census Will Result in a Loss of Federal Funding to Several Plaintiffs

a. The California Plaintiffs

i. The State of California

107. The citizenship question is more likely than not to cause the State of California to lose federal funding. This is because *994any measurable differential undercount of households containing noncitizens will cause California to lose funding for its state-share programs. Reamer Decl. ¶¶ 20, 74; Tr. at 676:1-2, 677:6-14 (Reamer).

108. Dr. Andrew Reamer, who is an expert in the relationship between census data and federal financial assistance, Reamer Decl. ¶¶ 1-8, testified that a significant portion of federal domestic financial assistance is distributed on the basis of statistics derived from the decennial census, id. ¶ 10.

109. At least 320 federal domestic assistance programs used census-derived data to distribute about $ 900 billion in FY2016. Id. ¶ 10. Of these, there are 24 large federal financial assistance programs with geographic allocation formulas that rely in whole or part on census-derived data. Id. ¶¶ 10-11, Ex. D (PTX-245); Tr. 668:12-669:9 (Reamer); see also UF 52-56. Eighteen of these 24 programs are "state-share" programs, in that they rely in whole or in part on state share of a U.S. population total. Reamer Decl. ¶¶ 11, 17, Ex. D PTX-245).¹⁰

110. As previously discussed, the citizenship question will cause a differential undercount of persons living in households containing noncitizens. This, in turn, will lead to a differential undercount of the population of states that have a disproportionate number of such persons, like California. See id. ¶ 17. This will impact a number of federal domestic financial assistance programs with census-tied geographic allocation formulas. Id. ¶¶ 16-18, 74.

111. Specifically, a differential undercount in the decennial census among persons who live in households containing noncitizens will lead to measurable fiscal losses across numerous federal programs for states with population percentages of households containing noncitizens that are above the national average, including California. Id. ¶¶ 17-18.

112. Dr. Reamer performed calculations using two alternative projections of the potential undercount of households containing at least one noncitizen resulting from the addition of a citizenship question, which undercount scenarios were applied to projections of the 2020 population by state. Id. ¶ 14. These projections were prepared by Plaintiff expert witness Dr. Bernard Fraga, id. ¶¶ 15, 35, and are discussed in greater detail in Parts III.B.6 and III.C.1.b, infra .

113. The two scenarios involve: (1) an undercount of 5.8 percent of households containing at least one noncitizen, and (2) using the same starting point but assuming 86.63 percent of these households are ultimately counted successfully through NRFU efforts. Id. ¶ 36.

114. Dr. Reamer calculated the specific financial impact of these projections on three of the 18 state-share programs-Title I grants to local educational agencies, the Supplemental Nutrition Program for Women, Infants, and Children grants, and *995Social Services Block Grants-to illustrate certain losses that would occur in the event of a differential undercount. Id. ¶¶ 17, 20, 33, 37-40, 43-48, 52-53, 57-63; Tr. 667:8-19 (Reamer).

115. Under either undercount scenario, California, among other states, would lose funding annually under all three programs. Reamer Decl. ¶¶ 49-50, 54-55, 64-65, and accompanying charts.

116. Dr. Reamer's conclusion that a differential undercount will result in lost funding extends to the other 15 state-share programs he identified, meaning that California, among other states, will lose population share and thus funding under these programs if the citizenship question causes an undercount of individuals living in households containing noncitizens. Id. ¶ 34; Tr. 678:18-679:2 (Reamer).

117. Dr. Reamer also opined that the magnitude of the impact varies depending on the extent of the undercount. Id. ¶¶ 15-18. A change in the amount of the differential undercount would impact only the magnitude of the loss to a state-share program, not the existence of a loss. Id. ¶ 20.

118. Similarly, a change in the funding level or allocation formula would impact only the magnitude of the loss, not the existence of a loss, so long as the allocation formula retains a degree of state-share-based calculation. Id. ¶ 19; Tr. at 669:24-670:11, 675:19-22 (Reamer).

119. Therefore, because the inclusion of the citizenship question on the 2020 Census is likely to result in a significant differential undercount of households containing noncitizens, the State of California is more likely than not to lose funding for its state-share programs. Reamer Decl. ¶¶ 20, 74; Tr. at 676:1-2, 677:6-14 (Reamer).

ii. LAUSD

120. The citizenship question will also cause the Los Angeles Unified School District ("LAUSD") to lose federal funding. The funding for certain federal assistance programs is distributed among localities within the state according to formulas prescribed by law. Tr. at 677:23-678:10 (Reamer). For example, Title I grants are ultimately distributed to local educational agencies, and grants authorized by the Workforce Innovation and Opportunity Act (WIOA) are distributed to Local Workforce Development ("LWD") areas. Reamer Decl. ¶¶ 66, 67; see also id. ¶¶ 45, 71 (WIC and Community Development Block Grants); Tr. 677:15-22, 678:11-13 (Reamer) (Community Development Block Grants and WIOA grants).

121. Where there is a differential undercount of noncitizens within the locality that is a funding recipient of a state-share program relative to the national population, those localities will experience a loss of federal funding. Tr. 677:23-678:10 (Reamer).

122. For example, LAUSD, which has a higher-than average share of households containing noncitizens than the state and national population, would incur a further decrease in Title I funding when the funding received by California is distributed among the local educational agencies within the state. Reamer Decl. ¶ 66 n.2; Escudero Decl. ¶¶ 16, 27; Ryback Decl. ¶ 33.

b. San Jose Plaintiffs

123. Among the 18 "state share" programs discussed in paragraph 109, supra , are grants authorized under WIOA, including the Youth Activities Program, 29 U.S.C. § 3163, the Adult Activities program, 29 U.S.C. § 3173(b)(2)(A), and the Dislocated Workers Program, 29 U.S.C. § 3173(b)(2)(B). Reamer Decl. ¶ 67; Tr. at 677:19-22; 678:11-13. Also among the 18 "state share" programs are grants distributed via the Community Development *996Block Grant ("CDBG") Entitlement Program. Reamer Decl. ¶ 71; Tr. at 677:15-18.

124. Under WIOA, San Jose operates a workforce development program called "work2future" that serves a LWD area composed of the cities of San Jose, Campbell, Morgan Hill, Los Altos Hills, Gilroy, Los Gatos, Saratoga, and Monte Sereno, along with the unincorporated areas of Santa Clara County. Melchor Aff. ¶ 2.

125. The cities that comprise work2future's LWD have a combined population of 1,243,043 residents, of whom at least 197,663, or 16.00%, are noncitizens. Judicially Noticed Facts ¶ 14.

126. Of the 321,004,407 residents of the United States, 22,337,765, or 6.96%, are noncitizens. Judicially Noticed Facts ¶ 5.¹¹

127. Of the 38,982,847 residents of California, 5,250,604, or 13.47%, are noncitizens. Judicially Noticed Facts ¶ 6.

128. Among the funding programs that use Bureau data are programs administered by the Department of Labor under WIOA, which use Bureau data as part of the allocation formulas set forth in 29 U.S.C. §§ 3162(C) and § 3172(C). UF 56. WIOA provides funding to work2future under a two-part formula: first funding is delivered to a state (the "State Allotment") and the State of California distributes the State Allotment among the LWD's (the "Sub-State Allotment"). Melchor Aff. ¶ 5.

129. Based on his calculations regarding the three example programs, and the fact that WIOA is one of the 18 "state share" programs he identified, Dr. Reamer concluded with a high degree of certainty that California's State Allotment under WIOA will be lower under each of the scenarios set forth by Dr. Fraga because California's percentage of noncitizens is higher than the national average. Reamer Decl. ¶ 68.

130. Monique Melchor, Director of work2future, Workforce Development Board, Office of Economic Development for the City of San Jose, is tasked with ensuring that the program operates in compliance with federal law and regulations and to ensure that it is properly funded. As part of her duties, she regularly uses the WIOA formula for calculating the Sub-State Allotment for work2future's LWD to ensure it was properly delivered. Melchor Aff. ¶¶ 5, 8. Melchor uses publicly available Bureau data to make these calculations, which are based on the LWD's relative share of the total unemployed, the relative share of the excess unemployed, and the local area's share of disadvantaged adults or youth. Id. ¶ 10.

131. Because the LWD, mainly including San Jose, has a higher percentage of noncitizens than California as a whole, a differential undercount of noncitizens, according to Dr. Barreto, will "be particularly severe in San Jose and other plaintiffs' jurisdictions." Trial Tr. 375:6-7; Judicially Noticed Facts ¶ 14.

132. If the population of the LWD (including its disadvantaged adults and disadvantaged youth) is undercounted relative to the State of California, then the LWD's share of the Sub-State Allocation will decrease. Melchor Decl. ¶¶ 12-13.

133. Therefore, if there is a differential undercount of noncitizens in the 2020 Census, California will receive a lower State Share of WIOA funding, and the work2future LWD, which includes San Jose, will receive a smaller proportion of the State Share in its Sub-State Allocation, resulting *997in a double funding loss for the City of San Jose.

134. The CDBG program, administered by the U.S. Department of Housing and Urban Development ("HUD"), provides funding to eligible "entitlement communities" including the City of San Jose. Reamer Decl. ¶ 71; Judicially Noticed Facts ¶ 16. One of the programs that uses Bureau data is the Home Investment Partnership Program ("HOME"), run by HUD, which uses Bureau data as part of its allocation formula under 42 U.S.C. § 12747(b). UF 54.

135. The statutory formula for HOME grants is required to reflect "each jurisdiction's share of total need among eligible jurisdiction[s] for an increased supply of affordable housing for very low-income and low-income families of different size, as identified by objective measures of inadequate housing supply, substandard housing, the number of low-income families in housing likely to be in need of rehabilitation, the costs of producing housing, poverty, and the relative fiscal incapacity of the jurisdiction to carry out housing activities eligible under section 12742 of this title without Federal assistance. Allocation among units of general local government shall take into account the housing needs of metropolitan cities, urban counties, and approved consortia of units of general local government." 42 U.S.C. § 12747(b)(1)(A).

136. One of the programs that uses Bureau data is the CDBG, run by HUD, which uses Bureau data as part of its allocation formula under 42 U.S.C. § 5306(b). UF 55. This program provides funds to entitlement communities according to a set of formulas prescribed in law and that include data on population, poverty rates, and housing conditions. Reamer Decl. ¶ 71.

137. The statutory formula for CDBG grants considers "the average of the ratios between the population of that city and the population of all metropolitan areas; the extent of poverty in that city and the extent of poverty in all metropolitan areas; and the extent of housing overcrowding in that city and the extent of housing overcrowding in all metropolitan areas." 42 U.S.C. § 5306(b)(1)(A).

138. HUD awards the City of San Jose an annual allocation of CDBG and HOME funding; the amount of this funding is directly tied to data from the Census. Clements Aff. ¶ 10.

139. Of San Jose's 1,023,031 residents, 176,345, or 17.24%, are noncitizens. Judicially Noticed Facts ¶ 11. Thus, San Jose's percentage of noncitizens is nearly two-and-a-half times the national percentage of 6.96%. It follows that San Jose is likely to be undercounted relative to the population as a whole if the citizenship question is added to the Decennial Census. Tr. 546:6-17 (Barreto).

140. Because CDBG is one of the 18 programs that Dr. Reamer identified as sensitive to changes in population, and because Dr. Reamer concluded that any such program would provide less funding to geographic areas that are undercounted relative to the population as a whole, San Jose will receive less CDBG funding if is undercounted relative to the nation as a whole. Reamer Decl. ¶ 18; Tr. at 677:7-14. Therefore, it is more likely than not that San Jose will receive less CDBG funding if a citizenship question is added to the Census.

141. San Jose's Office of Emergency Management ("OEM") also faces a substantial risk of losing funding based on the addition of the citizenship question.

142. In his role as Director of OEM, and in prior positions, Raymond Riordan has applied for funding on behalf of San Jose *998from the Federal Emergency Management Agency ("FEMA"). Supp. Riordan Aff. ¶ 2.

143. When applying for funding from FEMA, Riordan completes a "Preliminary Damage Assessment" in accordance with FEMA guidelines as provided in its Damage Assessment Operations Manual and its Preliminary Damage Assessment for Individual Assistance Operations Manual. These manuals require Riordan to supply census data for areas affected by a disaster. Id. ¶¶ 4-5.

144. Riordan has personally worked on funding proposals to FEMA that were denied because FEMA determined not enough people lived in the affected area to qualify for funding. Id. ¶¶ 7-10. Because the total number of individuals affected by a disaster is a key factor in most applications for disaster funding, and because Riordan provides this number based on data from the Census Bureau, a net undercount of San Jose's population will impede the City of San Jose's ability to obtain adequate funding when the next disaster occurs. Id. ¶ 14. Indeed, San Jose is in a region prone to natural disasters, including earthquakes, floods, and fires. Id. ¶ 13.

6. Inclusion of the Citizenship Question on the 2020 Census Increases the Likelihood that California Will Lose Political Representation

145. Adding a citizenship question to the 2020 Census significantly increases the likelihood that California will lose at least one congressional seat. Dr. Bernard Fraga credibly testified that (1) California is expected to maintain its current level of congressional representation (53 seats) if the 2020 Census does not ask a citizenship question, and (2) adding a citizenship question to the 2020 Census increases the probability that California will, contrary to its actual population, lose a congressional seat.

146. To estimate the quantitative effect of adding a citizenship question to the 2020 Census, Dr. Fraga looked at four scenarios of nonresponse and NRFU-two based on Dr. Barreto's survey data, and two based on Census Bureau data. Id. ¶ 26. For each scenario, Dr. Fraga estimated how much of each state's population would not be counted in the 2020 Census because of the citizenship question. Id. ¶¶ 57, 58.

147. Dr. Fraga estimated that, based on Dr. Barreto's survey data and assuming NRFU will not remediate the differential in self-response rates (Scenario A), the citizenship question would cause 12.51 percent of Californians not to be reported in the census self-response. Id. ¶¶ 57-58.

148. Dr. Fraga performed the same calculation based on the Census Bureau's estimate of a decline in nonresponse by 5.8 percent for noncitizen households-again assuming NRFU will not have a mitigating affect (Scenario C). Id. ¶¶ 57, 60. Based on this estimate, the citizenship question would cause 1.68 percent of Californians not to be reported in the census self-response. Because California has a higher proportion of noncitizens than any other state, this was also the highest proportional undercount of all the states. Id. ¶¶ 57, 65. Using either the survey results or the Census Bureau's estimate, California will always have the highest proportional undercount, as long as the Census Bureau's follow-up efforts are anything less than 100 percent effective. Id. ¶ 65.

149. Dr. Fraga used these different undercount estimates to quantify the impact of adding a citizenship question on congressional apportionment, including the probability that apportionment would be affected by the question. Id. ¶ 66. In the baseline scenario with no citizenship question, California is projected to keep its current 53 seats in the House of Representatives.

*999Id. ¶ 75. Based on Dr. Barreto's estimated self-response rates, however, Dr. Fraga estimates that California is virtually certain to lose three seats if the citizenship question is included in the census, assuming the NRFU process does not mitigate the effects of this differential (Scenario A). Id. ¶¶ 73-74, 76.¹²

150. Dr. Fraga also concluded that, using the Census Bureau's 5.8 percent estimate of nonresponse by noncitizen households and assuming NRFU has no effect on the net differential undercount, the likelihood of California losing at least one seat nearly doubles to fifty percent probability (Scenario C). Id. ¶ 82. Even applying a NRFU success rate of 86.63%, Dr. Fraga testified that the likelihood of California losing at least one seat still increases by 15 percent (Scenario D). Id. ¶¶ 48, 82. The 86.63% NRFU success rate derives from Dr. O'Muircheartaigh's report and reflects the 2016 in person follow-up response rate during the 2016 ACS in census tracts with a higher than average share of noncitizen households. Id. ¶ 49.

7. Plaintiffs Have Had to Appropriate Funds to Mitigate the Harm of the Inclusion of the Citizenship Question on the 2020 Census

a. California Plaintiffs

i. State of California

151. The State of California has appropriated and will imminently spend increased funds on census-related community outreach due to the citizenship question. Former California Governor Jerry Brown initially proposed to the California Legislature for the FY 2018-19 state budget an appropriation of $ 40.3 million "to be spent over a three-year period for statewide outreach and other activities related to the 2020 Census count." UF 111. This budget proposal was made prior to Secretary Ross' issuance of the Decision Memo announcing the addition of the citizenship question. See PTX-502 at 3.

152. The final FY 2018-19 state budget that was enacted in the summer of 2018 included an appropriation of $ 90.3 million "to support the California Complete Count effort, which was established within the Government Operations Agency to perform outreach focusing on hard-to-count populations for the decennial census." UF 112.

153. Early on in the budget process, before Secretary Ross issued the Decision Memo, the Legislative Analyst's Office published an analysis of the census outreach budget item. PTX-502. The analysis observed that the potential introduction of a citizenship question to the 2020 Census could cause an undercount. Id. at 2-5 (noting that changes to the census, including "the potential for a question about citizenship[,] raise the possibility of an undercount in California in 2020").

154. This concern was echoed in legislative committee materials and at least one committee hearing. The legislative history of the FY 2018-19 state budget shows that one of the driving forces behind the increased appropriation was the citizenship question. PTX-504 at 140 (summary of FY 2018-19 state budget includes section devoted to census outreach to hard-to-count residents, and states that "[t]he Budget includes $ 90.3 million for statewide outreach and other efforts related to increasing the participation rate of Californians in the decennial census"); PTX-505 at 1 (description of FY 2018-19 state budget line *1000items references as a "major change" the $ 90.3 million allocated to "support the California Complete Count effort...to perform outreach focusing on hard-to-count populations for the decennial census"); PTX-506 at 8, 76 (Legislative Analyst's overview of FY 2018-19 state budget includes section describing $ 90.3 million allocated for outreach activities); PTX-509 at 23 (March 15, 2018 Senate Budget and Fiscal Review Subcommittee meeting staff report on the California Complete Count-Census 2020 agenda item notes that concerns about emphasis on Internet self-response, "in combination with the potential for a question about citizenship[,] raise the possibility of an undercount in California in 2020"); PTX-510 at 41-44 (April 24, 2018 Assembly Budget Subcommittee meeting staff report on 2020 Census Outreach agenda item states that one change to the 2020 Census is that "[t]he federal government has decided to include a citizenship question in the census, which is projected to reduce the rate of response," and identifies the citizenship question as one of the challenges that would justify "additional resources" for outreach); PTX-517 at 45:19-46:12 (statement by Assembly-member David Chiu at the April 24, 2018 Assembly Budget Subcommittee meeting that the citizenship question presents "a different world" that may justify doubling census outreach expenditures); PTX-512 at 24 (May 22, 2018 Senate Committee on Budget and Fiscal Review meeting staff report on California Complete Count-Census 2020 agenda item notes that "[d]ue to the significant changes to the census, providing state funding to target hard-to-count populations is reasonable," and "[d]ue to both the extreme importance of an accurate census to the state and the high cost of the necessary outreach, additional funding is warranted"); PTX-513 at 30-31 (May 24, 2018 Assembly Budget Subcommittee meeting staff report on the 2020 Census Outreach Funding agenda item proposed $ 113 million increase in census outreach funding, including $ 12 million for Los Angeles County's complete count efforts); PTX-514 at 79 (June 6, 2018 2018-19 Legislative Budget Conference Committee meeting staff report on 2020 Census Outreach cites $ 153.3 million request from Assembly and $ 135.3 million request from the Senate for outreach efforts); PTX-515 at 173 (June 8, 2018-19 Legislative Budget Conference Committee meeting staff report on 2020 Census Outreach recommends adopting compromise of $ 90.3 million allocation for census outreach).

155. Since the appropriation, the California Complete Count Committee, the body in charge of the outreach efforts, has submitted reports to the Governor and Legislature that underscore the challenge presented by the citizenship question to those outreach efforts. PTX-508 at 4-5, 13, 19, 27-28 (October 2, 2018 report to Governor acknowledges that citizenship question presents challenge to outreach efforts, describes formation of working group on citizenship matters, observes that the question will generate fear, and identifies possible difficulties of hiring trusted messengers); PTX-507 at 9-10 (October 1, 2018 report to Legislature states that in convenings with local partners, "[p]articipants identified the most significant barrier to achieving a complete count to be the Census citizenship question and the current political environment regarding immigrants").

156. The State's allocation of outreach funding to the County of Los Angeles also confirms increased expenditures due to the citizenship question. Baron Decl. ¶¶ 7-16. The State initially allocated to the County $ 8.7 million in census outreach funding. Id. ¶ 7, 12. In May 2018, the County requested an additional $ 3.3 million in funding *1001specifically due to the addition of the citizenship question on the Census. Id. ¶¶ 11-12 & Ex. A. The State met the County's request, in part. In November 2018, the State announced its County outreach allocations, allocating $ 9,393,090 to the County of Los Angeles for 2020 Census outreach to hard-to-count populations. UF 113; Baron Decl. ¶¶ 13-15.

i. City of San Jose

157. San Jose's population has been undercounted in prior censuses and San Jose is taking steps to mitigate the likely undercount of its population which will be caused by the inclusion of a citizenship question on the 2020 Census. Ruster Aff. ¶¶ 4, 11-19. San Jose, along with other cities, has partnered with the County of Santa Clara to form a "Complete Count Committee" to encourage participation in the Census by hard-to-count communities. Id. ¶ 3, 8. Such partnerships among localities are encouraged by the Bureau to ensure an accurate and complete count. Tr. 799:23-800:14 (Abowd). In fact, San Jose's preparations for the Census are being conducted in concert with the Bureau's integrated partnership and communication program. Id. ¶¶ 3-5, 11.

158. San Jose has dedicated approximately $ 300,000 in resources towards performing outreach and expects to allocate approximately $ 300,000 more before the Census is conducted. Id. ¶¶ 13, 15. San Jose has dedicated staff resources, including Jeff Ruster, the Assistant Director in the Office of Economic Development, and a full-time consultant, to prepare for the Census. Id. ¶¶ 3, 14, 19.

159. Consistent with recommendations from the Census Bureau and the Department of Commerce, these preparations include targeted outreach that is being performed specifically because Secretary Ross has decided to add a citizenship question to the Census. Id. ¶¶ 10-11, 16; Tr. at 1017:22-1018:17 (Abowd). Indeed, Dr. Abowd agreed that the addition of the citizenship question has made it reasonable for cities to increase their outreach expenditures to encourage participation in the census. Tr. at 979:16-25 (Abowd).

160. The targeted outreach being conducted by San Jose is designed specifically to mitigate the impact that adding the citizenship question to the Census will have on hard-to-reach populations in San Jose. Ruster Aff. ¶¶ 16, 17. This diversion of resources to address the citizenship question will limit San Jose's ability to use its resources to boost participation among other parts of the population. Id. ¶ 19.

161. In light of the substantial risk posed by the addition of the citizenship question, it is reasonable for San Jose to spend additional time and money on the outreach to address concerns about the addition of the citizenship question. Tr. at 979:16-25 (Abowd).

ii. BAJI

162. BAJI's Executive Director, Opal Tometi has reviewed and received feedback from impacted communities regarding the 2020 Census, formed coalitions with other immigration groups, spoken to members of BAJI, and participated in a number of events, panels, and town halls where concerns about the addition of a citizenship question to the 2020 Census were raised. Additionally, Tometi has worked with Black immigrant, refugee, and African American communities since 2010 and is familiar with the ways in which the census count affects the well-being of these communities. Tometi Aff. ¶ 3.

163. To ensure that BAJI's members are properly counted in the 2020 Census, BAJI plans to conduct additional outreach to these communities to encourage them to *1002participate in the 2020 Census questionnaire. Supp. Tometi Aff. ¶ 4. BAJI has determined that, due to the citizenship question, such outreach will require the expenditure of additional resources including money, staff time, and operational expenses. Id. To date, BAJI has expended considerable staff time educating constituents, and other community members, about the addition of a citizenship question to the Census. Tometi Decl. ¶ 20 (reaffirmed in Tometi Aff. ¶ 2).

164. Given the nature of the census taking process, BAJI is reserving the majority of the expenditure it will use to address the addition of the citizenship question-resources that will likely be diverted from its other essential services-for its efforts to bolster census participation among its members and other underrepresented minority communities who are fearful about responding to the citizenship question. Id. (reaffirmed in Tometi Aff. ¶ 2). Accordingly, BAJI expects to allocate at least an additional $ 200,000 in the next two years to addressing the addition of a citizenship question to the Census and attempting to mitigate its harmful effects. Id. (reaffirmed in Tometi Aff. ¶ 2).

165. BAJI has taken these actions because an undercount of noncitizens would disproportionately affect BAJI members in that BAJI's membership has a high proportion of immigrants who live in immigrant-rich metropolitan areas. Id. ¶¶ 9-10 (reaffirmed in Tometi Aff. ¶ 2). The impact of the addition of a citizenship question to the Census, and BAJI's diversion of its resources to address the same, has therefore impaired BAJI's ability to carry out its mission of fostering racial, economic, and social equality for Black immigrants and other historically underrepresented communities. Id. ¶ 19 (reaffirmed in Tometi Aff. ¶ 2).

C. Conclusions of Law Related to Standing

The California Plaintiffs and the San Jose Plaintiffs ultimately succeed in establishing standing with respect to the APA and the Enumeration Clause Claims.¹³ Once the Court determines that at least one of the San Jose Plaintiffs and one of the California Plaintiffs has standing, it need not consider the standing of the remaining parties. See Leonard v. Clark , 12 F.3d 885, 888 (9th Cir. 1994) ("The general rule applicable to federal court suits with multiple plaintiffs is that once [a] court determines that one of the plaintiffs has standing, it need not decide the standing of the others"). Here, several plaintiffs across both actions have standing.

1. Injury-in-Fact

To establish injury-in-fact, a plaintiff must demonstrate it "has sustained or is immediately in danger of sustaining a direct injury" as a result of the challenged action. Spokeo , 136 S.Ct. at 1552 (Thomas, J., concurring) (quoting Ex parte Levitt , 302 U.S. 633, 634, 58 S.Ct. 1, 82 L.Ed. 493 (1937) (per curiam) ). The injury or threat of injury must be "concrete and particularized" rather than conjectural or hypothetical. Lujan , 504 U.S. at 560, 112 S.Ct. 2130. Thus, where standing is based upon an alleged future injury, the plaintiff must demonstrate either that the harm is certainly impending, or that there *1003is a substantial risk the harm will occur. Susan B. Anthony List v. Driehaus , 573 U.S. 149, 158, 134 S.Ct. 2334, 189 L.Ed.2d 246 (2014). Injury-in-fact also exists where there is a "substantial risk" that harm will occur, which prompts plaintiffs reasonably to incur costs to mitigate or avoid that harm. Clapper v. Amnesty Int'l USA , 568 U.S. 398, 414 n.5, 133 S.Ct. 1138, 185 L.Ed.2d 264 (2013).

"For standing purposes, a loss of even a small amount of money is ordinarily an 'injury.' " Czyzewski v. Jevic Holding Corp. , --- U.S. ----, 137 S.Ct. 973, 983, 197 L.Ed.2d 398 (2017) ; see also Carpenters Indus. Council v. Zinke , 854 F.3d 1, 5 (D.C. Cir. 2017) ("A dollar of economic harm is still an injury-in-fact for standing purposes."); Council of Ins. Agents & Brokers v. Molasky-Arman , 522 F.3d 925, 932 (9th Cir. 2008) (noting that Supreme Court has found injury-in-fact even where magnitude of harm was only a few dollars).

The evidence establishes that Plaintiffs will be injured in at least three ways. These injuries include, (1) the loss of federal funding, (2) an increased risk of losing political representation, and (3) the expenditure of funds for census outreach to mitigate the substantial risk of harm flowing from the citizenship question.

a. Loss of Federal Funding

A plaintiff who is likely to lose federal funding due to a differential undercount of the population has suffered an injury-in-fact for the purposes of Article III standing. Carey v. Klutznick , 637 F.2d 834, 838 (2d Cir. 1980) ; City of Detroit v. Franklin , 4 F.3d 1367, 1373-1375 (6th Cir. 1993) (standing established where "census undercount will result in a loss of federal funds" to plaintiffs' city). Lost federal funding, no matter the magnitude, qualifies as an injury-in-fact. Czyzewski , 137 S.Ct. at 983.

A significant portion of federal domestic financial assistance is distributed based on census-derived data, including from 24 large federal financial assistance programs with geographic allocation formulas that rely in whole or in part on census-derived data. If, as Plaintiffs have shown, there is any measurable differential undercount of households containing noncitizens, California will lose federal funding, because California has a larger proportion of noncitizens relative to other states.

Similarly, some federal domestic financial assistance that is based on census-derived data is distributed among localities within the state. If, as Plaintiffs have shown, there is any measurable differential undercount of households containing noncitizens, LAUSD and San Jose will also lose federal funding, because these localities have a larger proportion of noncitizens relative to other localities. Having shown a "substantial risk" that the citizenship question will cause them to lose federal funding, the State of California and LAUSD (of the California Plaintiffs), and the City of San Jose (of the San Jose Plaintiffs) have each established injury-in-fact. See Susan B. Anthony List , 573 U.S. at 158, 134 S.Ct. 2334.

b. Loss of Political Representation (State of California Only)

A plaintiff's "expected loss of a Representative to the United States Congress undoubtedly satisfies the injury-in-fact requirement of Article III standing." Dep't of Commerce v. U.S. House of Representatives , 525 U.S. 316, 331-332, 119 S.Ct. 765, 142 L.Ed.2d 797 (1999) ; Carey , 637 F.2d at 838 (holding that a disproportionate undercount resulting in the loss of congressional representation confers standing); City of New York v. U.S. Dep't of Commerce , 713 F.Supp. 48, 50 (E.D.N.Y. 1989) (holding that a likely undercount *1004of populations disproportionately represented in plaintiff states conferred standing).

The California Plaintiffs have shown that adding a citizenship question to the 2020 Census will cause a differential undercount of the State of California's population relative to other states, creating a substantial risk that California will lose its fair share of political representation in Congress, and by extension, the Electoral College. Even using the most conservative data and assumptions (Scenario D), Dr. Fraga concluded that the addition of the citizenship question would increase the likelihood by 15 percent that California will lose a congressional seat. See Part III.B.6.

This estimate is conservative. As previously discussed, Scenario D is based on the Census Bureau's estimate that the self-response rate in noncitizen households will decline by 5.8 percent relative to all-citizen households. The Census Bureau has acknowledged, and the Plaintiff experts agree, that this is a conservative estimate. Accordingly, it is safe to conclude that the true differential decline in self-response rates will in fact be higher than 5.8 percent. Furthermore, the Census Bureau's NRFU efforts are unlikely to mitigate this differential decline in self-response rates. See Part III.B.2, supra .

In light of the considerable evidence in the record that the Census Bureau's NRFU will be differentially less effective in counting noncitizens and Latinos and therefore will not significantly mitigate the effects of the citizenship question, Dr. Fraga's estimates under Scenario C are the most probative. Under Scenario C, Dr. Fraga once again adopted the Census Bureau's conservative estimate that the differential decline in self-response rates for noncitizen households relative to all-citizen households will be 5.8 percent. Unlike Scenario D, however, Scenario C relies upon the assumption that the Census Bureau's NRFU processes will not remediate the differential in self-response rate caused by the citizenship question.

Under this scenario, the likelihood that the state will lose a congressional seat nearly doubles from 26.1 percent to a 49.9 percent, creating a near 50-50 chance that California will lose representation. Fraga Aff. ¶¶ 81-82. This evidence establishes that California faces a substantial risk of losing at least one seat as a result of the citizenship question. Such malapportionment of the State of California's congressional representation is a "threat of vote dilution" that "is 'concrete' and 'actual or imminent, not conjectural or hypothetical." Dep't of Commerce , 525 U.S. at 332, 119 S.Ct. 765 (quoting Whitmore v. Arkansas , 495 U.S. 149, 155, 110 S.Ct. 1717, 109 L.Ed.2d 135 (1990).

c. Expenditure of Funds to Mitigate Harm

i. The California Plaintiffs

The State of California has reasonably increased its expenditures on census outreach to attempt to mitigate the decline in self-response rates and the resulting differential undercount of Plaintiffs' residents caused by the citizenship question. These additional expenditures, which constitute a direct injury to the State of California, are sufficient to establish injury-in-fact for standing purposes. Clapper , 568 U.S. at 414 n.5, 133 S.Ct. 1138 (standing may be based on "reasonably incur[red] costs to mitigate or avoid" a "substantial risk" of harm); Monsanto Co. v. Geertson Seed Farms , 561 U.S. 139, 153-155, 130 S.Ct. 2743, 177 L.Ed.2d 461 (2010) (finding injury-in-fact where alfalfa growers increased administrative costs to minimize likelihood of potential contamination of their crops from genetically-altered alfalfa plant).

*1005The California Plaintiffs' expenditures were reasonably incurred because they face a substantial risk of a differential undercount of their residents. These plaintiffs have proven that the citizenship question will cause them to be differentially undercounted because they have a disproportionate share of noncitizens, immigrants, and Latino residents and that this differential undercount will cause them to lose federal funding and create a substantial risk that the State of California will lose political representation.

The legislative history of California's FY 2018-19 state budget and follow-up reports to the Governor and Legislature show that the State appropriated, and the California Plaintiffs are spending, additional funds on census outreach to attempt to mitigate these negative impacts caused by the citizenship question. Although it is not possible to pinpoint precisely how much the citizenship question drove the increase in the state budget's census outreach line item, the California Plaintiffs have shown that the Legislature's decision to boost the Governor's initial allocation for census outreach ($ 40.3 million) to a higher allocation in the enacted budget ($ 90.3 million) is due at least in part to the citizenship question. UF 111-112.

The California Plaintiffs also provided evidence that the citizenship question prompted the County of Los Angeles to request $ 3.3 million in additional funds to meet the County's need for funding for census outreach to the hard-to-count populations most likely not to respond to the 2020 Census because of the citizenship question. The State partially met this request by allocating hundreds of thousands of dollars of additional funding to the County for census outreach. Because any amount of costs incurred to mitigate harm is sufficient to confer standing, as long as such costs were reasonably incurred, the State of California's expenditures on census outreach constitute a direct injury to the state's budgets and resources that is sufficient to establish injury-in-fact for standing purposes.

ii. The San Jose Plaintiffs

The City of San Jose has also spent, and will continue to spend, additional money on outreach specifically because Secretary Ross directed that a citizenship question be added to the Census. Because there is a substantial risk that adding the citizenship question will cause a net differential undercount of noncitizens, which in turn would cause San Jose to lose federal funding, the City's diversion of resources to prevent that harm is reasonable. Accordingly, the City of San Jose has suffered injury-in-fact based on its additional outreach spending.

Because BAJI has similarly diverted resources to encourage its constituents to participate in the census and to counteract the chilling effects of the citizenship question, it too has suffered an injury-in-fact. An injury-in-fact exists where a nonprofit organization shows "a drain on its resources from both a diversion of its resources and frustration of its mission." Fair Hous. of Marin v. Combs , 285 F.3d 899, 905 (9th Cir. 2002) ; Havens Realty Corp. v. Coleman , 455 U.S. 363, 378-380, 102 S.Ct. 1114, 71 L.Ed.2d 214 (1982).

BAJI's Executive Director, Opal Tometi, testifies that BAJI has already expended additional staff time and related financial resources to educate its constituents about the addition of a citizenship question to the Census and encourage participation. BAJI intends to spend an additional $ 200,000 to counteract the chilling effect of the citizenship question on its members. BAJI's membership has a high proportion of immigrants and is concentrated in immigrant-rich metropolitan areas. Accordingly, an undercount of immigrant populations, *1006and any resulting loss of funding or political representation, would impair BAJI's ability to carry out its mission of fostering racial, economic, and social equality for Black immigrants and other historically underrepresented communities. In sum, BAJI's diversion of resources to mitigate the negative impact of the citizenship question on its organizational mission qualifies as an injury in fact.¹⁴

2. Traceability

The second standing element requires a plaintiff to demonstrate that the injury is "fairly traceable to the challenged action of the defendant, and not the result of the independent action of some third party not before the court." Mendia v. Garcia , 768 F.3d 1009, 1012 (9th Cir. 2014) (quoting Bennett v. Spear , 520 U.S. 154, 167, 117 S.Ct. 1154, 137 L.Ed.2d 281 (1997) ). This requirement is less demanding than a proximate cause standard. Lexmark Intern., Inc. v. Static Control Components, Inc. , 572 U.S. 118, 134 n.6, 134 S.Ct. 1377, 188 L.Ed.2d 392 (2014) ("Proximate causation is not a requirement of Article III standing, which requires only that the plaintiff's injury be fairly traceable to the defendant's conduct."). Indeed, "[c]ausation may be found even if there are multiple links in the chain connecting the defendant's unlawful conduct to the plaintiff's injury, and there's no requirement that the defendant's conduct comprise the last link in the chain." Mendia , 768 F.3d at 1012. When harm is caused by multiple actions, "what matters is not the length of the chain of causation, but rather the plausibility of the links that comprise the chain." Id. at 1012-13 (quotation and citation omitted).

While an injury is not fairly traceable it if is " 'th[e] result [of] the independent action of some third party not before the court,'...that does not exclude injury produced by determinative or coercive effect upon the action of someone else." Bennett , 520 U.S. at 169, 117 S.Ct. 1154 (emphasis in original) (quoting Lujan , 504 U.S. at 560, 112 S.Ct. 2130 ). The key question is whether the "government's unlawful conduct is at least a substantial factor motivating the third parties' actions." Mendia , 768 F.3d at 1013 (citations and quotations omitted). Even when some links in the chain of causation are illegal acts by third parties, the injury may still be "fairly traceable" to the original challenged action. See Attias v. Carefirst, Inc. , 865 F.3d 620, 629 (D.C. Cir. 2017) (harm caused by a data breach is "fairly traceable" to company's inadequate security standards even though data was stolen by third party hackers).

Plaintiffs have established that there will be a drop in self-response to the 2020 Census caused by the addition of the citizenship question. That nonrespondents have a legal duty to respond to the census does not alter this conclusion because the citizenship question is a "substantial factor" contributing to the nonresponse. Mendia , 768 F.3d at 1013. The Bureau and its top officials have concretely affirmed the predictable impact of adding a citizenship question to the 2020 Census on self-response rates. The harms Plaintiffs will suffer-namely the loss of federal funding, the risk to California of losing political representation, and the expenditure of resources to mitigate these harms-inevitably *1007flow from the disproportionate undercount of particular demographic groups that is likely to result from Secretary Ross's decision to include the citizenship question on the 2020 Census.

Defendants argue that any differential decline in self-response rates among noncitizens and Latinos is attributable to the hostile macro-environment surrounding issues of immigration, and that Plaintiffs fail to show that the citizenship question in particular caused this decline. Dr. Barreto testified persuasively, however, that a hostile macro-environment combined with the inclusion of sensitive questions on a survey can have a cumulative effect that is greater than either of these factors would have on their own. Tr. 386:21-25, 411:5-14 (Barreto). More to the point, both Dr. Barreto's survey and the Census Bureau's natural experiment attempted to isolate the effect of the citizenship question within the same macro-environment. Dr. Barreto's survey asked respondents about their willingness to participate in the 2020 Census both with and without the citizenship question. Tr. 568:5-569:1. The Census Bureau's natural experiment, and subsequent statistical analysis, compared response rates to the 2010 Census (no citizenship question) to those of the 2010 ACS (which included a citizenship question). PTX-103 at 6-7; PTX-148 at 6-7. Accordingly, the differential decline in response rates found in these experiments cannot be exclusively attributed to the current macro-environment surrounding immigration.

In sum, the injuries Plaintiffs have identified are fairly traceable to the Secretary's decision to include the citizenship question in the 2020 Census.

3. Redressability

Finally, to meet the third standing requirement, Plaintiffs must show that it is likely that a favorable decision will redress their injuries. Lujan , 504 U.S. at 561, 112 S.Ct. 2130. 48. A plaintiff need show only that "an injury" be redressed by a favorable decision. Larson v. Valente , 456 U.S. 228, 243 n.5, 102 S.Ct. 1673, 72 L.Ed.2d 33 (1982) (emphasis in original). A plaintiff "need not show that a favorable decision will relieve [ ] every injury." Id. (emphasis in original). A favorable decision vacating or enjoining the decision to add a citizenship question to the 2020 Census would redress Plaintiffs' injuries by diminishing the funds that they would need to be expended on census outreach and by ensuring that Plaintiffs do not lose federal funding or, in the case of California, political representation because of the citizenship question.

In sum, both the San Jose Plaintiffs and the California Plaintiffs have carried the burden of establishing all three of the standing requirements.

IV. APA CLAIM

Under section 706 of the APA, a reviewing court must "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; contrary to constitutional right, power, privilege, or immunity; in excess of statutory jurisdiction, authority, or limitations, or short of statutory right; [or] without observance of procedure required by law." 5 U.S.C. § 706(2)(A)-(D). Accordingly, the decision-making process that ultimately leads to the agency action must be "logical and rational." Allentown Mack Sales & Serv., Inc. v. NLRB , 522 U.S. 359, 374, 118 S.Ct. 818, 139 L.Ed.2d 797 (1998). Courts should be careful, however, not to substitute their own judgment for that of the agency.

*1008Suffolk Cty. v. Sec'y of Interior , 562 F.2d 1368, 1383 (2d Cir. 1977). Ultimately, a reviewing court may uphold agency action "only on the grounds that the agency invoked when it took the action." Michigan v. EPA , --- U.S. ----, 135 S.Ct. 2699, 2710, 192 L.Ed.2d 674 (2015). Post hoc rationalizations may not be considered. American Textile Mfrs. Inst., Inc. v. Donovan , 452 U.S. 490, 539, 101 S.Ct. 2478, 69 L.Ed.2d 185 (1981).

B. Scope of Review

In evaluating an APA claim, courts "typically" limit their review to the Administrative Record existing at the time of the decision. Sw. Ctr. for Biological Diversity v. U.S. Forest Service , 100 F.3d 1443, 1450 (9th Cir. 1996) ; accord Ranchers Cattlemen Action Legal Fund United Stockgrowers of Am. v. U.S. Dep't of Agric. , 499 F.3d 1108, 1117 (9th Cir. 2007). "Under limited circumstances, however, extra-record evidence can be admitted and considered." Ranchers Cattlemen , 499 F.3d at 1117. These exceptions include: (1) when plaintiffs make a showing of agency bad faith, or (2) when the agency failed to consider "all relevant factors" of the decision. Id.

1. Use of Extra-Record Evidence to Evaluate Plaintiffs' Pretext Argument

A court may consider extra-record evidence that is relevant to the reason for an agency action where there has been a strong showing of bad faith or improper behavior by the decision-makers. Public Power Council v. Johnson , 674 F.2d 791, 795 (9th Cir. 1982) ; Ranchers Cattlemen , 499 F.3d at 1117. In such circumstances, consideration of extra-record evidence is "necessary to a meaningful judicial review" of the agency's actual decision-making process. Tummino v. Torti , 603 F.Supp.2d 519, 543 (E.D.N.Y. 2009).

As discussed in greater detail in Part IV.C.12, infra , the Administrative Record alone provides a "strong showing" of bad faith which authorizes this Court to look outside the record for further evidence of pretext. That is particularly true here, where the Administrative Record includes pre-decision communications between Secretary Ross and his "point person" on the citizenship question issue expressing caution about what the Administrative Record would include. PTX-96, PTX-362. 2. Defendants proceeded to submit an initial Administrative Record that mischaracterized the Secretary's decision-making process and concealed important circumstances surrounding the DOJ's request for the addition of the citizenship question. Defendants supplemented the record only after being ordered to do so in the New York matter. For these reasons, and the reasons set forth in Part IV.C.12, infra , consideration of extra-record evidence is appropriate, but not necessary, for the purpose of determining whether Secretary Ross's decision was pretextual.

2. Use of Extra-Record Evidence to Evaluate Whether Defendants Considered All Relevant Factors

An agency's decision is arbitrary and capricious under the APA if, among other things, the agency failed to consider all "relevant factors," Ranchers Cattlemen , 499 F.3d at 1115 ; ignored "an important aspect of the problem," Motor Vehicle Mfrs. Ass'n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co. , 463 U.S. 29, 43-44, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) ; or made "an irrational departure from [settled] policy," INS v. Yueh-Shaio Yang , 519 U.S. 26, 32, 117 S.Ct. 350, 136 L.Ed.2d 288 (1996).

To apply these standards, a court must-as a threshold matter-understand what is "relevant," "important," or "settled policy" in the field where the challenged *1009agency decision was made. In many cases, the Administrative Record will provide the relevant benchmarks. But evidence outside the "bare record" may be required to determine "the applicable standard" to apply in evaluating the completeness of the agency's reasoning and in determining whether the agency ignored critical factors or information. See Citizens to Preserve Overton Park, Inc. v. Volpe , 401 U.S. 402, 420, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971) ; see also Nat'l Audubon Soc. v. U.S. Forest Service , 46 F.3d 1437, 1447 (9th Cir. 1993).

"It will often be impossible, especially when highly technical matters are involved, for the court to determine whether the agency took into consideration all relevant factors unless it looks outside the record to determine what matters the agency should have considered but did not." Asarco, Inc. v. EPA , 616 F.2d 1153, 1160 (9th Cir. 1980). The court cannot adequately discharge its duty to engage in a "substantial inquiry" if it is required to take the agency's word that it considered all relevant matters. Id. ; see also Nat'l Audubon Soc. v. Hoffman , 132 F.3d 7, 15 (2d Cir. 1997) ("[t]he omission of technical scientific information is often not obvious from the record itself").

Because this case involves complex technical issues related to survey methodology and census-related practices, meaningfully evaluating whether Defendants considered all relevant factors or irrationally departed from settled policy would be difficult on the Administrative Record alone. The Court cannot simply accept Defendants' assurances that they considered all relevant factors. Accordingly, extra-record evidence will be admitted for this limited purpose as well.

In the interest of facilitating review of this decision by higher courts, the findings of fact and conclusions of law that are based exclusively on the administrative record will be discussed separately from those that rely on extra-record evidence.

C. Findings of Fact Based Exclusively on the Administrative Record ¹⁵

1. Compilation of the Administrative Record

166. Defendants produced an initial Administrative Record, along with a certification and index on June 8, 2018, consisting of only 1,320 pages. See PTX-1 (AR 1-1320). These materials contain little documentation of internal discussions that took place before December 2017 and communications between the Departments of Commerce and Justice about the citizenship question. Id.

167. On June 21, 2018, Defendants filed a supplemental memorandum composed by Secretary Ross that revised the narrative of how a citizenship question came to be placed on the decennial census. PTX-2.

168. On July 3, 2018, and memorialized in a July 5, 2018 order, Defendants were ordered to supplement the Administrative Record in New York et al. v. Department of Commerce, et al. , No. 18-cv-2921 (S.D.N.Y.).¹⁶ In response to that order, Defendants produced extensive supplemental Administrative Record documents. The parties agreed that all documents bearing prefix-less Bates stamps between 000001 and 0013024 are part of the Administrative Record. Joint Pretrial Statement 11-13.

*10102. Secretary Ross's Deliberations Prior to Receiving the DOJ Letter

169. In March 2017, the Bureau reported to Congress the five "topics" that would be included on the 2020 Census, including gender, age, race, ethnicity, and homeownership status. The subjects did not include citizenship or immigration status. PTX-264 at 5-15. This report also listed topics limited to the ACS, including citizenship. Id. at 55.

170. That same month, Secretary Ross exchanged emails with his Deputy Chief of Staff and Director of Policy, Earl Comstock, regarding whether noncitizens are included in the census count for the purposes of congressional apportionment. PTX-30; PTX-55. Comstock specifically emailed Ross an answer to "Your Question on the Census," to confirm that noncitizens are indeed counted on the census. PTX-55.

171. On April 5, 2017, Ross's executive assistant wrote an email directed to Ross indicating that "Steve Bannon has asked the Secretary to talk to someone about the census." PTX-58.¹⁷ In their subsequent communications, Stephen Bannon connected Secretary Ross with Kris Kobach, former Vice Chair of the Presidential Commission on Election Integrity and Kansas Secretary of State, to discuss adding a citizenship question to the census. PTX-19; PTX-58. Kobach told Secretary Ross by phone that a citizenship question was necessary to address the "problem that aliens who do not actually 'reside' in the United States are still counted for congressional apportionment purposes." PTX-19 at 2.

172. A few days later, Mark Neuman, an outside advisor to Secretary Ross, emailed Comstock with the subject line "One of the Supreme Court cases that informs planning for the 2020 Census...." PTX-182. The email contained only a link to the Supreme Court's decision in LULAC v. Perry , id. , which considered citizen voting age population in assessing claims under Section 2 of the VRA. On April 13, Comstock emailed Neuman, asking when the Census Bureau would need to notify Congress of the questions that would appear on the 2020 Census. PTX-88; PTX-181. Neuman responded to Comstock on April 14, 2017, that the notification deadline for topics had already passed, and that "[t]here would be another opportunity next year." PTX-88.

173. On May 2, 2017, Secretary Ross emailed Comstock: "I am mystified why nothing [has] been done in response to my months old request that we include the citizenship question. Why not?" PTX-89. In reference to the statutory requirement that the topics for the Census be submitted by March 2017, Secretary Ross wrote to Comstock: "Worst of all they emphasize that they have settled with congress on the questions to be asked." Id. Comstock wrote back: "On the citizenship question we will get that in place....We need to work with Justice to get them to request that citizenship be added back as a census question." Id.

3. Comstock Seeks an Agency Willing to Request Addition of the Citizenship Question

174. Secretary Ross's May 2, 2017 email led to immediate action by Comstock. On May 3, 2017, Eric Branstad, a Senior White House Advisor, reached out on Comstock's behalf to Matthew J. Flynn, Senior Director of Cabinet Affairs at the Executive Office of the President, to find "the best counterpart to reach out to at DOJ regarding Census and Legislative issue?" PTX-85. Branstad subsequently referred Comstock to Mary Blanche Hankey *1011who was the White House liaison within DOJ. PTX-370.

175. Hankey directed Comstock to speak with James McHenry, the director of the Executive Office of Immigration Review at DOJ. Id. Comstock spoke "several times" with McHenry of DOJ about adding a citizenship question to the census. Id. McHenry ultimately informed Comstock that the DOJ did not want to request the citizenship question "given the difficulties Justice was encountering in the press at the time (the whole Comey matter)." Id.

176. McHenry therefore referred Comstock to the Department of Homeland Security. Id. DHS, however, likewise declined to request the citizenship question. Id. Following his failed discussions with DHS, Comstock asked James Uthmeier, of the Office of General Counsel at the Department of Commerce, to investigate "how Commerce could add the question to the Census itself." Id.

177. On May 24, 2017, Secretary Ross and David Langdon, a policy advisor who reported to Comstock, met "all afternoon." PTX-151 at 2. During that meeting, Secretary Ross asked questions about the content of the decennial Census and "seemed...puzzled why citizenship is not included in the 2020" census. PTX-86 at 1. Later that afternoon, Burton Reist, Chief of Decennial Communications and Stakeholder Relations at the Census Bureau, emailed Langdon a 1988 internal DOJ memorandum that opined the Constitution does not mandate the counting of undocumented U.S. residents in the census apportionment count. PTX-448, PTX-449 at 1-2. That evening, Langdon requested further information from Census Bureau staff including Reist regarding "the criteria used to pick topics for 2020 versus ACS. Say, citizenship." PTX-151.

178. That same day, Langdon sent an email to Comstock entitled "Counting of illegal immigrants," which stated: "the counting of illegal immigrants (or of the larger group of noncitizens) has a solid and fairly long legal history...[there is] a Bush 41 era DOJ opinion that proposed legislation to exclude illegal aliens from the decennial census was illegal." PTX-397. Comstock responded to Langdon that day, asking for further information on the selection of questions for the census versus the ACS, and passing along the Supreme Court decision of LULAC v. Perry that Neuman had previously provided for the proposition that the government might have a use for citizenship data. Id. ; PTX-182.

179. On July 14, 2017, Kobach emailed Secretary Ross to remind him of their prior telephone discussion "a few months ago." PTX-19 at 2. Kobach wrote that during their earlier discussion, he and Secretary Ross "talked about the fact that the US census does not currently ask respondents about their citizenship," and further advised Secretary Ross that the absence of such a question "leads to the problem that aliens who do not actually 'reside' in the United States are still counted for congressional apportionment purposes." Id. Kobach further wrote that "it was essential that one simple question be added to the upcoming 2020 census" and that a variant of the question that appears on the ACS "needs to be added to the census." Id.

180. On July 21, 2017, Kobach called Wendy Teramoto, a Senior Advisor and Chief of Staff to Secretary Ross. Id. at 1. He also emailed her, forwarding his July 14 email to Secretary Ross stating that he had spoken to Secretary Ross about adding a citizenship question to the 2020 Census at the direction of Steve Bannon. Id. Teramoto subsequently arranged a call between Kobach and Secretary Ross for a few days later. Id.

*1012181. On August 8, 2017, Secretary Ross emailed Comstock, asking "where is DOJ in their analysis" of whether to add a citizenship question to the 2020 Census, and advising "[i]f they still have not come to a conclusion please let me know your contact person and I will call the AG." PTX-96; PTX-98. The very next day, Comstock emailed Secretary Ross, stating "we are preparing a memo and full briefing for you on a citizenship question. The memo will be ready by Friday...Since this issue will go to the Supreme Court we need to be diligent in preparing the administrative record." PTX-96, PTX-362. Secretary Ross responded to Comstock the next day, "I would like to be briefed on Friday by phone...we should be very careful about everything, whether or not it is likely to end up in the SC." PTX-96, PTX-362.

182. On August 11, Comstock and Uthmeier exchanged edits on briefing materials for Secretary Ross related to a citizenship question. During this exchange, Uthmeier wrote that he had "recommendations on execution," stating that he thought "our hook" was "ultimately, we do not make decisions on how the [citizenship] data will be used for apportionment, that is for Congress (or possibly the President) to decide." PTX-437 at 2. That same day, Comstock emailed Secretary Ross and Teramoto "a draft memo on the citizenship question" prepared by James Uthmeier. PTX-3 at 1140; PTX-147. The memorandum has not been produced.

183. On September 1, 2017, Secretary Ross complained to Comstock and Teramoto about a number of issues, including that he had "received no update [on] the issue of the census question" and Comstock responded, "Understood. Wendy and I are working on it." PTX-45; PTX-97.

184. On September 6, Secretary Ross and his senior staff had a meeting to discuss adding a citizenship question to the decennial census. PTX-31; PTX-35; PTX-36; PTX-46; PTX-47. The next day, Secretary Ross requested from his staff an update on "progress since the discussion yesterday regarding the citizenship question." PTX-37, PTX-49. A few days later, Uthmeier contacted Neuman to discuss "some Census legal questions for the Secretary." PTX-38. That same day, Comstock sent Secretary Ross a memo reporting on his efforts to identify someone who would request that a citizenship question be added to the 2020 Census, and advising that, as of that date, he had not been successful in locating any such agency. PTX-48; PTX-134. He also informed the Secretary that, once his discussions with DOJ and DHS fell through, he had enlisted Uthmeier's help in advancing the citizenship question. PTX-48; PTX-134.

185. In mid-September, John Gore of DOJ contacted and later spoke on the telephone with Teramoto "about a DOJ-DOC" issue. PTX-59; PTX-60. Following that conversation, Gore worked with an aide to Attorney General Sessions to set up a phone call between Secretary Ross and then Attorney General Sessions. PTX-63, PTX-67, PTX-68. On September 17, 2017, Danielle Cutrona in the Office of the Attorney General wrote to Wendy Teramoto, Ross's Chief of Staff, to say that "[t]he Attorney General is available on his cell," and that "it sounds like we can do whatever you all need us to do and the delay was due to a miscommunication....The AG is eager to assist." PTX-62 (emphasis added); see also PTX-67; PTX-68. Secretary Ross and Attorney General Sessions proceeded to speak on the phone regarding the subject of Gore and Teramoto's earlier conversation, presumably about adding the citizenship question to the census. PTX-57; PTX-61; PTX-62.

*1013186. The most reasonable inference to be drawn from these facts is that, on or before September 18, 2017, Secretary Ross spoke with Attorney General Sessions and asked that he instruct his subordinates at DOJ to request a question on citizenship be added to the Census.

187. On Sunday, October 8, Secretary Ross sent an email to staffer Peter Davidson of the Department of Commerce with the subject line, "Letter from DOJ" asking "what is its status." PTX-52. Davidson responded, "I'm on the phone with Mark Neuman right now...he is giving me a readout of his meeting last week." Id. On the evening of November 27, 2017, Secretary Ross emailed Davidson again, stating, "Census is about to begin translating the questions into multiple languages and has set the printing contract. We are out of time. Please set up a call for me tomorrow with whoever is the responsible person at Justice. We must get this resolved." PTX-144.

188. The Administrative Record shows that Secretary Ross issued a "request that we include the citizenship question," PTX-89, that Comstock assured him that "we will get that in place," id , that Secretary Ross's staff attempted to find a way for the Department of Commerce to add the citizenship question without an outside agency request when it appeared unlikely that any such request would materialize, and that Secretary Ross intervened with the Attorney General when career staff at DOJ and DHS refused to request the addition of the question. The reasonable inference to be drawn from these actions is that Secretary Ross had already made up his mind to include the citizenship question on the 2020 Census before he received the DOJ request letter.

189. There is no writing of any kind in the Administrative Record authored by the Secretary or anyone at the Commerce Department (or anyone else) that directly describes the reasons why the Secretary wanted to add a citizenship question prior to December 12, 2017.

190. The Administrative Record reveals, however, that at the same time the Secretary was discussing adding a citizenship question to the census in the spring and summer of 2017, he was asking Comstock questions about whether congressional apportionment based on the census included noncitizens, he was informed that including noncitizens in congressional apportionment was legally required, and he was advised by Kobach that the inclusion of noncitizens in the census was a "problem." PTX-19; PTX-55; PTX-58; PTX-86; PTX-89; PTX-151; PTX-397; PTX-437; PTX-444. Nowhere in the record does Secretary Ross explicitly state whether he subscribed to the same rationale for including the citizenship question as Kobach.

191. Nevertheless, it is clear that Secretary Ross's decision prior to December 12, 2017 to include the citizenship question was unrelated to the enforcement of the VRA. Secretary Ross's direction to his staff to find a way to include the citizenship question on the census, the fact that his staff solicited DHS as well as DOJ for a request, and the fact that Comstock's attempts to solicit a request were not limited to the Division of Civil Rights belie any notion that Secretary Ross's motivation was to meet DOJ's VRA enforcement needs. Rather, it was DOJ that was meeting his preferences. PTX-62; PTX-67; PTX-68.

192. The Administrative Record strongly supports the following inferences regarding the period from March 2017 through December 2017: (1) that before Secretary Ross received DOJ's request to add the citizenship question, he had already decided to include the citizenship question on the 2020 Census; (2) that his reasons for *1014wanting the question asked were unrelated to DOJ's alleged need for citizenship data at the census block level to enforce the VRA; (3) that he directed his staff to find a way to get the citizenship question added to the Census; (4) that, when he was told that he needed to have an agency ask him to add the question, he directed his staff to find an agency to ask him; and (5) that he did not care who asked for the question or the reasons for asking the question, so long as some agency asked.

4. DOJ Letter Requesting the Addition of the Citizenship Question

a. DOJ's Previous Decision Not to Request the Citizenship Question

193. In June 2014, Arthur Gary of DOJ wrote to the General Counsel of the Department of Commerce to indicate what census-derived data the DOJ uses and to confirm that it continued to use such data. PTX-1 at 278-83. Gary indicated that DOJ used citizenship data collected by the ACS to enforce the VRA, and that the "lowest geography" for which DOJ needed citizenship data was the "Census block group" level. Id. at 280. Gary did not indicate that DOJ needed citizenship data at the census block level, or that it needed a citizenship question on the 2020 Decennial Census.

194. In July 2016, Gary wrote to the Census Bureau to confirm that the DOJ "had no needs to amend the current content and uses or to request new content" for the 2020 Census. PTX-1 at 311. In November of 2016, Gary supplemented that letter to "formally request[ ] that the Census Bureau consider a new topic in the ACS relating to LGBT populations." Id. At no time prior to December 2017 did Gary express a need for more granular citizenship data or a citizenship question on the Census.

b. DOJ Changes Positions and Requests the Citizenship Question

195. On December 12, 2017, Arthur Gary of DOJ sent a formal letter to Dr. Ron Jarmin, Acting Director of the Census Bureau, requesting that a citizenship question be added to the 2020 Census (December 12 Letter). PTX-32. The December 12 Letter requests the addition of a citizenship question for purposes of VRA enforcement. Id.

196. The letter does not state that a citizenship question and/or block-level citizenship data is strictly necessary for the purposes of VRA enforcement. Rather, the letter contends, "the Department [of Justice] believes that decennial census questionnaire data regarding citizenship, if available, would be more appropriate for use in redistricting and in Section 2 litigation than the ACS citizenship estimates." Id. at 3. It further cites numerous published cases for the proposition that, "in order to assess and enforce compliance with Section 2's protection against discrimination in voting, the Department needs to be able to obtain citizen voting-age population data for census blocks…where potential Section 2 violations are alleged or suspected." Id. at 2.

197. The December 12 Letter states that one of the reasons decennial census data on citizenship would be preferable to ACS data concerns the margin of error: "The ACS estimates are reported at a ninety percent confidence level, and the margin of error increases as the sample size-and, thus, the geographic area-decreases...By contrast, decennial census data is a full count of the population." Id. at 3.

198. The letter does not state that any plaintiffs had lost any Section 2 enforcement actions due to insufficient CVAP data from the ACS. Nor does the letter suggest that DOJ has declined to bring *1015any Section 2 enforcement actions due to insufficient CVAP data from the ACS.

5. The Census Bureau's Analyses and Recommendations

199. Soon after the Census Bureau received the December 12 Letter, Dr. Abowd directed the SWAT Team, to formulate a response to the suggestion that a citizenship question be added, which Dr. Abowd managed and reviewed. PTX-4D at 284; PTX-75; PTX-148; UF 19. In a series of technical reports, responses to questions posed by Secretary Ross, and other briefing documents, the Census Bureau consistently recommended against adding a citizenship question to the 2020 Census. PTX-22; PTX-101; PTX-133; PTX-148.

200. The Census Bureau concluded that the stated goals of DOJ with respect to enforcement of the VRA could be accomplished, in a less costly and more effective manner, by linking Census responses to other administrative data sets available to the federal government. PTX-22; PTX-101; PTX-133; PTX-148. In other words, the Census Bureau determined that including a citizenship question on the decennial census was not necessary to provide complete and accurate data in response to DOJ's request. PTX-22; PTX-101; PTX-148. In fact, the Census Bureau advised Secretary Ross that relying exclusively on administrative records would result in more accurate citizenship data than adding a citizenship question to the census. PTX-22; PTX-101; PTX-148. The Bureau further concluded that relying exclusively on administrative records would produce more accurate citizenship data than attempting to use both administrative records and citizenship responses. PTX-24; PTX-25.

201. As part of its analysis, the Census Bureau examined the initial drop in self-response due to the citizenship question, as well as the potential costs of NRFU as a result of that drop. PTX-22; PTX-101; PTX-133; PTX-148. The Administrative Record does not, however, contain any information showing that the Census Bureau analyzed whether NRFU could or would mitigate the decrease in self-response rates, or whether the citizenship question would ultimately result in an undercount.

202. On or about December 15, 2017, Dr. Abowd directed senior executives and expert employees of the Census Bureau to evaluate alternative methods of providing estimates of the CVAP to support redistricting under Public Law 94-171 (P.L. 94-171) and Section 2 of the VRA. PTX-4D at 284; PTX-148 at 1. This evaluation is reflected in a memorandum dated December 22, 2017 and was provided to the Commerce Department (December 22 Memo). PTX-148.

203. In the December 22 Memo, the Bureau reported on the differential decline in self-response rates from the 2010 Decennial Census to the 2010 ACS in households with at least one noncitizen and found that there was a 5.1% greater decline among such households, which is "consistent with citizenship questions being more sensitive for households with noncitizens." Id. at 6-7.

204. Additionally, the Bureau found that "there is a tendency for noncitizen ACS respondents to report being U.S. citizens," and that "roughly 40 percent" of those who are legal resident noncitizens reported being citizens on the ACS. Id. at 7.

205. The memo recommended that that the citizenship data for DOJ VRA enforcement be obtained through the use of administrative records rather than the addition of a question to the decennial census instrument. Id. at 11. The Bureau based this recommendation on several factors, including its conclusion that the question *1016would create a differential increase in nonresponse rates of at least 5.1 percent in households with at least one noncitizen, and the fact that citizenship status is often inaccurately reported by noncitizens.

206. The Census Bureau identified eight administrative sources of citizenship information either already in use by the Census Bureau or available for acquisition by the Census Bureau. Id. at 3. Accordingly, the December 22 Memo advised that the best way to meet DOJ's stated need was to provide it with citizenship data from administrative records. Id. at 11.

207. The memo observed that, if the recommendation were followed, "[t]he 2020 Census questionnaire would not be altered, and the field operations would not have to be expanded to compensate for the lower rate of voluntary compliance predicted for a census that asks the citizenship question directly." Id. at 12.

b. January 3 Memo

208. Following the December 22 Memo, the Census Bureau further memorialized its research and analysis of potential sources of citizenship data in a January 3, 2018 memorandum from Dr. Abowd to Dr. Jarmin (January 3 Memo). PTX-101.

209. This document explained that the Census Bureau produces CVAP data in "two related data products: the P.L. 94-171 redistricting data produced by April 1st of the year following a decennial census under the authority of 13 U.S.C. Section 141, and the Citizen Voting Age Population by Race and Ethnicity (CVAP) tables produced every February from the most recent five-year ACS data." PTX-101 at 1. While the P.L. 94-171 data is released at the census block level, the CVAP data is released at the census block group level. Id.

210. The memo proceeded to analyze the cost and data quality implications of three alternative methods of meeting DOJ's request for census block-level estimates of CVAP. Id.

211. Alternative A was to "[m]aintain the status quo for data collection, preparation and publication," but then prepare a special product for DOJ that combines the P.L. 94-171 and CVAP tables to produce the Bureau's best estimate of block-level CVAP data. Id. Alternative A was estimated to cost $ 200,000. Id. at 2. The Census Bureau concluded this option was "not very costly and does not harm the quality of the census count." Id. at 3.

212. Alternative B was to "[a]dd a citizenship question to the 2020 Census questionnaire." The memo estimated that Alternative B would increase census nonresponse rates by at least 5.1 percent of all households with one or more noncitizens, or 700,000 households. Id. at 2. This would increase the NRFU workload and increase the cost of the 2020 Census by "at least $ 27.5 million." Id. While the memo stated that Alternative B suited DOJ's stated uses better than Alternative A, it noted that this option would be "very costly" and-because NRFU is less accurate than self-responses-would harm the accuracy of the census. Id. at 2, 3.

213. Alternative C would involve creating block-level citizenship data using administrative records without asking respondents about their citizenship status. Id. at 1. This option was estimated to cost less than $ 1 million. Id. at 2. The Bureau concluded that relying exclusively on administrative records would deliver higher quality citizenship data than Alternative B because the administrative records provide "very accurate" citizenship information. Id. at 3. The memo explained that this is because the administrative record data required proof of citizenship, whereas citizenship information is self-reported less accurately, and proxies report citizenship *1017even less accurately than self-responding individuals. Id.

214. The Census Bureau therefore expressly recommended Alternative C in the January 3 memo, reasoning that, compared to Alternatives A and B, Alternative C "even better meets DoJ's stated uses, is comparatively far less costly than Alternative B, and does not harm the quality of the census count." Id. On January 4, 2018, Dr. Abowd wrote to various census officials, including Dr. Jarmin, "Ron reports that he has discussed this with the Under Secretary, and she agrees with the recommendation of Alternative C, but Alternative A remains a possibility as well." PTX-122 at 1.

c. January 19 Memo

215. On January 19, 2018, Dr. Abowd sent another memorandum to Secretary Ross on the "Technical Review of the Department of Justice Request to Add Citizenship Question to the 2020 Census" (January 19 Memo). PTX-22. This memo presents the view of Dr. Abowd and the technical team that evaluated Alternatives A, B and C. Id. It contains the same recommendation and rationale as in the January 3 Memo, along with some additional details of their analysis. Id.

216. The memo specifically examined the issue of item nonresponse to the citizenship question, i.e. nonresponse to the particular question, rather than the whole questionnaire. Id. at 4. It noted that item nonresponse rates for the citizenship question "are much greater than the comparable rates for other demographic variables like sex, birthdate/age, and race/ethnicity." Id. Moreover, between 2013 and 2016, the item nonresponse rate of Hispanics to the citizenship question was approximately double that of non-Hispanic whites. Id.

217. The memo also examined the break-off rate for internet responses to the 2016 ACS, i.e. at what question people stopped taking the survey. Id. at 5. It found that "[b]ecause Hispanics and non-Hispanic non-whites breakoff much more often than non-Hispanic whites, especially on the citizenship-related questions, their survey response quality is differentially affected." Id.

218. The Census Bureau reiterated its estimate, based on the 2010 census and ACS, that Alternative B would cause a 5.1 percent drop in self response from households containing at least one noncitizen. Id. at 4-5. It explained that, while both citizen and noncitizen households responded to the ACS (which had a citizenship question) at lower rates than to the census (which did not), the decline between the two surveys was 5.1 percent greater for noncitizen households. Id. at 4.

219. The January 19 Memo further explained that lowered self-response rates would lower census data quality because data obtained in NRFU have greater rates of erroneous enumeration and whole-person imputation. Id. at 5-6. Erroneous enumerations are enumerations of a person who should not have been counted and whole-person imputations are enumerations of all characteristics of a person. Id. at 5.

220. The Census Bureau provided additional information about the shortcomings of attempting to ascertain citizenship status based on direct inquiries. It specifically explained that Alternative C would yield more accurate citizenship data than Alternative B because, based on historical census and ACS data, noncitizens misreport themselves as citizens "for no less than 23.8% of the cases, and often more than 30%." Id. at 7. Specifically, "[i]n 2010 and 2016, individuals for whom the administrative data indicate noncitizen respond citizen in 32.7% and 34.7% of the ACS questionnaires, respectively." Id. at 8. By *1018contrast, Alternative C would provide highly accurate citizenship data because administrative record citizenship data is "verified" based on proof of citizenship or legal resident alien status. Id. at 7.

221. The memo also addressed the increased burden on respondents associated with the addition of the citizenship question. The Census Bureau explained: "Survey methodologists consider burden to include both the direct time costs of responding and the indirect costs arising from nonresponse due to perceived sensitivity of the topic." Id. at 5. Thus, a citizenship question "would make the 2020 Census modestly more burdensome in the direct sense, and potentially much more burdensome in the indirect sense that it would lead to a larger decline in self-response for noncitizen households."Id.

222. Finally, the Census Bureau explained the $ 27.5 million increase in costs due to the addition of the citizenship question "is a conservative estimate because the other evidence cited in this report suggests that the differences between citizen and noncitizen response rates and data quality will be amplified during the 2020 Census compared to historic levels. Hence, the decrease in self-response for [non]citizen households in 2020 could be much greater than the 5.1 percentage points we observed during the 2010 Census." Id. at 6. Ultimately, the memo concluded that adding a citizenship question to the decennial census, "is very costly, harms the quality of the census count, and would use substantially less accurate citizenship status data than are available from administrative sources." Id. at 1.

223. Bureau staff had a single meeting with Secretary Ross on February 12, 2018, to discuss the January 19 Memo. The only record of the meeting in the Administrative Record is a February 12, 2018 email from James Treat of the Census Bureau to other Bureau employees, including Dr. Jarmin and Dr. Abowd, identifying five "actions from yesterday's meeting with the Secretary." PTX-128.

d. March 1 Memo

224. At some point after the February 12 meeting, Secretary Ross requested analysis of a fourth option, Alternative D, which would involve "combining Alternative B (asking the citizenship question of every household on the 2020 Census) and Alternative C (do not ask the question, link reliable administrative data on citizenship status instead)." PTX-133 at 2. On March 1, 2018, Dr. Abowd sent an additional recommendation memorandum to Secretary Ross performing an analysis of this fourth alternative. PTX-133. In this memo, the Census Bureau concluded that, "Alternative D would result in poorer quality citizenship data than Alternative C. It would still have all the negative cost and quality implications of Alternative B outlined in the draft January 19, 2018 memo to the Department of Commerce." Id. at 5.

225. The Census Bureau also identified additional problems with Alternative D in the March 1 Memo. First, census responses would be unreliable for filling in the data gaps for those who do not match to administrative records, because undocumented immigrants "have a strong incentive to provide an incorrect [citizenship] answer, if they answer at all." Id. at 4.

226. Second, lowered self-response rates due to the citizenship question would decrease the number of people who can be linked to administrative records, because NRFU personal identifying information is of lower quality than personal identifying information that is self-reported. Id.

227. Further, the March 1 Memorandum explains:

*1019Under Alternative C, there will be error in the administrative records, but we believe these to be relatively limited due to the procedure following by SSA, USCIS and State. In both Alternatives, the modeled cases will be subject to prediction error....Alternative D has an additional source or error, response error. This is where 2020 respondent give the incorrect status. Statisticians often hope these errors are random and cancel out. However, we know from prior research that citizenship status responses are systematically biased for a subset of noncitizens. Response error is only an issue in alternative D.

PTX-133 at 8.

228. The memo concludes that, under Alternative D, for the group of 22 million people for which the Bureau has both a census response and administrative records, but where they do not match, the citizenship data will be less accurate due to response errors. Id.

229. The Bureau also concluded that under Alternative D, it would not receive a response to the citizenship question from 35.4 million people and would likely be able to observe citizenship status for 21.5 million people and impute citizenship status for 13.8 million people, concluding that "there will be a need for imputing many cases across either alternative." Id. at 7-8.

230. As in previous Census Bureau memoranda on the subject, the March 1 Memo recommended against adding a citizenship question to the 2020 Census. Id. at 5.

6. The Set of 35 Questions Answered by the Census Bureau

231. Following the January 19 Memo, Comstock and Uthmeier developed and sent to the Census Bureau a set of 35 questions for the Census Bureau to answer about the analysis in the January 19 Memo. PTX-377. The Census Bureau's responses to the questions were submitted to the Commerce Department on March 1, 2018, along with Dr. Abowd's March 1 memorandum to Secretary Ross. PTX-133.

232. Question 31 asked, "What was the process that was used in the past to get questions added to the decennial Census or do we have something similar where a precedent was established?" Id. at 21. The Census Bureau provided the following response:

The Census Bureau follows a well-established process when adding or changing content on the census or ACS to ensure the data fulfill legal and regulatory requirements established by Congress. Adding a question or making a change to the Decennial Census or the ACS involves extensive testing, review, and evaluation. This process ensures the change is necessary and will produce quality, useful information for the nation.

The Census Bureau and the Office of Management and Budget (OMB) have laid out a formal process for making content changes.

• First, federal agencies evaluate their data needs and propose additions or changes to current questions through OMB.

• In order to be included, proposals must demonstrate a clear statutory or regulatory need for data at small geographies or for small populations.

• Final proposed questions result from extensive cognitive and field testing to ensure they result in proper data, with an integrity that meets the Census Bureau's high standards.

• This process includes several opportunities for public comment.

*1020• The final decision is made in consultation with OMB.

• If approved, the Census Bureau implements the change.

Id. at 21-22.

233. The description of the "well-established" process in the Census Bureau's response to Question 31 is consistent with other Census Bureau documents describing the process to add a question or change the content of the decennial census. PTX-4 at AR 3890, 3560, 9867; PTX-135; PTX-141.

234. The Administrative Record shows that, despite the fact that the questions were directed to the Census Bureau, Commerce Department Deputy General Counsel Michael Walsh drafted a different answer to Question 31. PTX-14. That answer states:

No new questions were added to the 2010 Decennial Census, so there is no recent precedent for considering a request to add questions to a Decennial Census. Consistent with longstanding practice for adding new questions to the ACS survey, the Census Bureau is working with relevant stakeholders to ensure that legal and regulatory requirements are fulfilled and that the question would produce quality, useful information for the nation. As you are aware, that process is ongoing. Upon its conclusion, you will have all of the relevant data at your disposal to make an informed decision about the pending request from the Department of Justice.

PTX-1 at 1296.

235. In Defendants' first production of documents in the Administrative Record, which they represented at the time constituted the complete administrative record, they included a version of the 35 questions and answers that included only Walsh's answer to Question 31. See PTX-1 at 1296. The Census Bureau's March 1 response to Question 31 was produced later and as a result of the New York court's order to supplement the record. See PTX-133 at 21-22.

236. Question 1 of the 35 questions asked, "With respect to Alternatives B and C, what is the difference, if any, between the time when the data collected under each alternative would be available to the public?" PTX-133 at 11. The Census Bureau answered Question 1 by stating that, between Alternatives B and C, there was no difference in the timing in which the citizenship data could be offered to the public. Id.

7. Memo Re: "Key Differences Between Alternative C and Alternative D"

237. The Administrative Record also includes a memorandum entitled "Summary Analysis of the Key Differences Between Alternative C and Alternative D." PTX-24. Like the March 1 Memo, this memorandum recommends using administrative data alone (Alternative C) and not adding a citizenship question. Id. at 1-2.

238. The Census Bureau explained that while both Alternative C and D will require the citizenship of a portion of the population to be imputed, or "modeled," Alternative D will suffer from accuracy issues because many noncitizens self-report as citizens, which will systematically bias the modeling in Alternative D. Id. at 2.

239. Neither this memorandum, or any of the memoranda previously discussed, analyzed whether NRFU would fully mitigate the nonresponse or whether it would ultimately result in an undercount of certain populations. See PTX-22; PTX-101; PTX-133; PTX-148.

*10218. DOJ's Refusal to Discuss Its Request with the Census Bureau

240. Following the Census Bureau's receipt of the December 12 Letter from DOJ, the Census Bureau sought to meet with DOJ to discuss its stated need for block-level citizenship data. PTX-4C at 1971; PTX-72. Dr. Jarmin emailed an initial response to the December 12 Letter on December 22. PTX-72. In that letter, he advised Gary that "the best way to provide P.L. 94-171 block-level data with citizen voting population by race and ethnicity would be through utilizing a linked file of administrative and survey data the Census Bureau already possesses. This would result in higher quality data produced at a lower cost." Id. In Dr. Jarmin's December 22 letter, he suggested to Gary a "meeting of Census and DOJ technical experts to discuss the details of this proposal." Id.

241. On January 2, 2018, Dr. Jarmin and Gary exchanged emails about meeting the following week. PTX-102. Although no date was set, Gary promised to "get back to" Dr. Jarmin about the timing. Id.

242. However, on February 6, 2018, Dr. Jarmin reported to Undersecretary Karen Dunn Kelley: "I spoke with Art Gary. He has spoken with DOJ leadership. They believe the letter requesting citizenship to be added to the 2020 Census fully describes their request. They do not want to meet." PTX-3 at 2136.

243. The Administrative Record contains no evidence that before Secretary Ross issued his March 26, 2018 Decision Memo, DOJ ever met with either the Census Bureau or the Commerce Department to discuss the request in DOJ's December 12 Letter.

9. Outside Stakeholders Weigh In

244. In early 2018, prior to issuing his Decision Memo, Secretary Ross conferred with multiple external stakeholders, including academics and representatives of interest groups, regarding the addition of a citizenship question to the Census. This input included communications from: (1) Six former Directors of the Census Bureau, PTX-1 at 1057 ("There is a well-proven multi-year process to suggest and test new questions. We strongly believe that adding an untested question on citizenship status at this late point in the decennial planning process would put the accuracy of the enumeration and success of the Census in all communities at grave risk."); (2) Members of the Census Bureau's Census Scientific Advisory Committee, PTX-1 at 794 ("We hold the strong opinion that including citizenship in the 2020 Census would be a serious mistake which would result in a substantial lowering of the response rate."); and (3) various state officials, professional associations, and other organizations expressing concern about the impact of the citizenship question on the integrity of the 2020 Census. See e.g. , PTX-1 at 787, 798, 1053, 1073, 1082, 1090, 1122, 1150, 1222, 1235, 1239, 1269; PTX-3 at 2281, 2284.

245. Numerous stakeholder letters also advised that a citizenship question was not necessary for Section 2 VRA enforcement. See e.g. , PTX-1 at 799 (letter from The Leadership Conference on Civil and Human Rights); id. at 1122 (letter from national Jewish organizations); PTX-3 at 2281 (letter from Constitutional Accountability Center).

246. In January and February 2018, before Secretary Ross issued the March 26 decision memo, the Leadership Conference on Civil and Human Rights, the Constitutional Accountability Center, and more than 15 other external stakeholders and voting rights experts, submitted letters to Secretary Ross explaining that that enforcement of Section 2 of the VRA has never once previously depended upon having *1022enumerated citizenship data since the statute's enactment in 1965. See, e.g. , PTX-1 at 799, 1122; PTX-3 at 2281.

247. Defendants attempted to enlist stakeholders to express support for the citizenship question, but had trouble doing so. PTX-71; PTX-4B at 137 ("We are trying to find someone who can give a professional expression of support for the proposal in contrast to the many folks we can find to give professional statements against the proposal").

248. On February 13, 2018, Dr. Jarmin wrote to an individual at the American Enterprise Institute (AEI): "We are trying to set up some meetings for Secretary Ross to discuss the proposed citizenship question on the 2020 Census with interested stakeholders. Most stakeholders will speak against the proposal. We're looking for someone thoughtful who can speak to the pros of adding such a question...." PTX-4B at 138. That same day, Michael Strain of the AEI responded: "None of my colleagues at AEI would speak favorably about the proposal." Id. Dr. Jarmin then wrote to Under Secretary Kelley, "Please see the thread below. Appears no one at AEI is willing to speak in favor of putting question on the 2020." Id. at 137.

249. Although they were in the minority, some groups and individuals expressed support for Secretary Ross's proposal. PTX-1 at 1198 (letter from Senator Tom Cotton); PTX-1 at 1199 (letter from Senator Ted Cruz); see also PTX-1 at 1203, 1206, 1261. Director Jarmin reported that certain individuals associated with the Center for Immigration Studies and the Heritage Foundation endorsed the addition of the citizenship question. PTX-1 at 1206, 1261; PTX-4C at 1645.

10. Secretary Ross's March 26 Decision Memorandum

250. On March 26, 2018, Secretary Ross issued his formal decision memorandum (Decision Memo) announcing and explaining his decision to adopt "Option D" (known in the Census Bureau memoranda as "Alternative D") and add a citizenship question to the decennial census. PTX-26. The Decision Memo states that Secretary Ross "set out to take a hard look" at the citizenship question "[f]ollowing receipt of the DOJ request" for it. PTX-26 at 1. Secretary Ross states that DOJ has requested CVAP data for census blocks "where potential Section 2 violations are alleged or suspected, and DOJ states that the current data collected under the ACS are insufficient in scope, detail, and certainty to meet its purpose under the VRA." Id.

251. The Decision Memo includes several statements that are inconsistent with the Census Bureau's unrefuted estimate that the citizenship question would cause a differential drop in self-response rates of at least 5.1 percent between households with at least one noncitizen and all-citizen households. PTX-26 at 3; see also PTX-101. These include:

• The statement that, with respect to "Option B" (the option of adding a citizenship question, referred to in the Census Bureau memoranda as "Alternative B") neither the Census Bureau nor the concerned stakeholders could document that the response rates would decline materially. PTX-26 at 3.

• The statement that a former Chief Operating Officer of the Census Bureau confirmed that to the best of his knowledge, "no empirical data existed on the impact of a citizenship question on responses." Id.

• Secretary Ross's conclusion that "while there is widespread belief *1023among many parties that adding a citizenship question could reduce response rates, the Census Bureau's analysis did not provide definitive, empirical support for that belief." Id. at 4.

• The statement that "there is no information available to determine the number of people who would in fact not respond due to a citizenship question being added, and no one has identified any mechanism for making such a determination." Id. at 5.

252. The Decision Memo sought to justify the selection of Option D by citing purported "[a]dditional empirical evidence about the impact of sensitive questions on survey response rates...from the SVP of Data Science at Nielsen." Id. at 6. The only evidence in the Administrative Record from this Nielsen representative are notes from a telephone call three days earlier between her (Christine Pierce) and Secretary Ross. Those notes indicate that "Ms. Pierce stated that her biggest concerns [sic] was that the reinstatement of a citizenship question could lead to a lower response rate." PTX-1 at 1276. During that call, Pierce also "noted the importance of testing questions." Id. There is no "empirical evidence" in the Administrative Record at all from Pierce regarding the citizenship question.

253. Secretary Ross also asserts in his memo that the citizenship question has been "well tested" because it has been on the ACS since 2005. PTX-26 at 2. The Secretary does not discuss the Census Bureau's usual process of testing new questions before placing them on the census. Nor does the memo identify any other testing standards that apply to addition of the citizenship question or discuss whether those standards have been met.

254. The Decision Memo asserts that asking the citizenship question of all people, "may eliminate the need for the Census Bureau to have to impute an answer for millions of people." Id. at 5. The Memo does not address the Census Bureau's estimate that, with a citizenship question on the census, it will have to impute the citizenship data of 13.8 million people. PTX-24 at 2; see also PTX-133 at 7-10.

255. Secretary Ross also states that Option D "would maximize the Census Bureau's ability to match the decennial census responses with administrative records," PTX-26 at 4, so as to allow for "more complete" citizenship data. The memo does not address the Census Bureau's analysis showing that adding a citizenship question will drive down the self-response rate and put more households into NRFU operations, thereby reducing the Census Bureau's ability to match survey responses with administrative records and resulting in less accurate citizenship data. PTX-25 at 4.

256. The Decision Memo ultimately states that the Department of Commerce prioritizes the goal of "obtaining complete and accurate [CVAP] data" over all else, and concludes that adding the citizenship question is the best way to achieve this goal. PTX-26 at 1, 7. The memo, however, points to no evidence to refute the Census Bureau's analysis which showed that use of a citizenship question would result in less accurate data than administrative records alone. PTX-22, PTX-25.

257. The Decision Memo does, however, make the following assertion:

[P]lacing the question on the decennial census and directing the Census Bureau to determine the best means to compare the decennial census responses to administrative records will permit the Census Bureau to determine the inaccurate response rate for citizens and noncitizens alike using the entire population.

*1024This will enable the Census Bureau to establish, to the best of its ability, the accurate ratio of citizen to noncitizen responses to impute for that small percentage of cases where it is necessary to do so.

PTX-26 at 5.

258. Nowhere in the Administrative Record does the Census Bureau state that adding a citizenship question would increase the accuracy of its estimate of inaccurate citizenship responses. Nor is it apparent why the inaccuracy rate of responders would help impute the citizenship data of nonresponders. If actual citizenship is benchmarked to administrative records, and the Bureau would use those records in any event, then adding a citizenship question to the decennial census would not assist in the imputation.

259. Finally, the Decision Memo does not discuss whether the Census Bureau's NRFU and imputation processes are likely substantially to mitigate any drop in self-response caused by the citizenship question.

11. Testing of the Citizenship Question

260. The Administrative Record contains no evidence that the cognitive and field testing that are part of the usual process for adding new questions to the census, as described in the Census Bureau's answer to Question 31 and in other documents in the Administrative Record, were performed. Moreover, the Administrative Record contains no evidence that Defendants considered any testing requirements from the Office of Management and Budget.

261. The Administrative Record does not include any information about what testing was performed, or how the citizenship question performed on that testing, before it was added to the ACS.

262. There is also no indication that Defendants considered obtaining any kind of waiver of any applicable agency guidelines regarding testing, or that the Census Bureau publicly noticed and provided a period for public comment about the citizenship question before Secretary Ross made the decision to add it to the Census, as required by the Census Bureau's "well-established" process described with the Census Bureau's answer to Question 31, and in other documents in the Administrative Record. See PTX-3 at 2236-37; PTX-4A at 155-56; PTX-4D at 812; PTX-135; PTX-141.

12. The Purpose and Timing of Secretary Ross's Decision

263. The weight of the evidence clearly shows that Secretary Ross made the decision to add a citizenship question before knowing whether DOJ had any need or even desire to add the question. In other words, Secretary Ross did not decide to add the citizenship question to the decennial census to aid in enforcement of Section 2 of the VRA.

264. There is no writing of any kind either in the Administrative Record authored by the Secretary or anyone at the Commerce Department (or anyone else) that expressly and directly describes the reasons why the Secretary wanted to add a citizenship question as early as the first quarter of 2017. The record reflects, however, that Secretary Ross was urged to include the citizenship question by Kobach, among others, to facilitate the exclusion of noncitizens from the population count for congressional apportionment. See, e.g. , PTX-19 at 2, PTX-55, PTX-437 at 2, PTX-448, PTX-449.

265. The following facts are particularly relevant: (1) Secretary Ross admits that he discussed a citizenship question with "senior administration officials" before he became *1025Secretary of Commerce; (2) Steve Bannon, the Chief Strategist and Senior Counselor to the President, then asked him to speak with Kobach, who wanted a citizenship question on the census to exclude noncitizens from the apportionment count rather than for VRA enforcement purposes; (3) on May 24, 2017, following Secretary Ross's meeting with David Langdon regarding the citizenship question, Langdon exchanged emails with the Census Bureau and received from them a 1988 DOJ memo about excluding "illegal immigrants" from the census count, and; (4) by August of 2017, when Secretary Ross's staff was preparing "a memo and full briefing…on a citizenship question" (which was not disclosed in the initially submitted administrative record), Commerce Department legal counsel emailed that their "hook" was that the Department "do[es] not make decisions on how the [citizenship] data will be used for apportionment," PTX-437 at 2.

266. By contrast, prior to the December 12 Letter, the Administrative Record includes no communications at all from Secretary Ross or any other Commerce Department officials indicating that they wanted to add the citizenship question for the purpose of aiding VRA enforcement. The only implicit reference to the VRA in the Administrative Record prior to December 12 is the LULAC v. Perry court opinion that Neuman emailed to Secretary Ross, and that Comstock emailed to Langdon in May 2017 as the Commerce Department was investigating whether it could request the citizenship question itself. There is no discussion in the record about the meaning of this case.

267. The fact that the Commerce Department began exploring ways to add the citizenship question without getting a request from another agency when it appeared no request was forthcoming strongly suggests that aiding DOJ's Section 2 enforcement was not the real reason behind the decision to add the question. PTX-370. The fact that Comstock attempted to solicit a request for the citizenship question from DHS after initially being turned away by DOJ further demonstrates that the Department of Commerce's reasons for wanting to add the citizenship question had nothing to do with VRA enforcement. Id.

268. The finding that Secretary Ross did not add the citizenship question for Section 2 enforcement is also supported by a dearth of evidence explaining why he would go to such lengths to persuade DOJ to request the question despite the agency's initial refusals.

269. Defendants' decision-making process further supports the finding that Secretary Ross was not motivated by VRA enforcement. The Administrative Record contains no disclosure by the Commerce Department to the Census Bureau about the intent to add a citizenship question until after the December 12 Letter was delivered, despite the fact that Secretary Ross and his staff spent the better part of 2017 communicating about the citizenship question and strategizing about how to elicit a request for it. While Secretary Ross is not obligated to inform the Census Bureau about all policy discussions he has with his staff, one would expect an open-minded decision-maker to consult with the experts at the Census Bureau early on to learn more about the consequences of adding the citizenship question to the census and to allow for the requisite testing to take place.

270. It is also highly implausible that Secretary Ross's true purpose was Section 2 enforcement because, as set forth in Part IV.C.10, supra , and discussed in greater detail in Part IV.E.2.a, infra , his Decision Memo includes statements plainly at odds with the evidence in the Administrative Record.

*1026271. Finally, Defendants' initial failure to disclose the full administrative record suggests bad faith in disclosing the true basis of Secretary Ross's decision. Defendants now concede that the Administrative Record consists of over 13,000 pages of documents, even though their initial submission contained only 1,320 pages. See Joint Pretrial Statement 11-13. Furthermore, the initial submission of the Administrative Record mischaracterized Secretary Ross's decision-making process and omitted critical information about the circumstances surrounding the DOJ request letter. It is also noteworthy that Defendants initially failed to produce the Census Bureau's answer to Question 31 (and produced only the Commerce Department's more favorable version regarding question testing), and that the Administrative Record includes correspondence between Ross and Comstock expressing caution about what ends up in the Administrative Record. PTX-362.

272. Together, this evidence establishes that Defendants intended to use the VRA enforcement as a pretext for adding the citizenship question when VRA enforcement was not, in fact, their true purpose. In sum, Plaintiffs have made a clear showing that (1) Secretary Ross acted in bad faith in disclosing the basis of his decision, and (2) Defendants acted in bad faith in compiling the Administrative Record.

D. Findings of Fact Based on Extra-Record Evidence

1. Secretary Ross's Decision-Making Process

a. Secretary Ross's Early Interest in the Citizenship Question

273. Extra-record evidence confirms that Secretary Ross talked to Steve Bannon about the citizenship question in the spring of 2017. Bannon asked Secretary Ross to speak with Kobach about adding a citizenship question to the decennial census. RFA response (ECF 146-6 at 3);¹⁸ RFA response (ECF 146-3 at 31). Secretary Ross's conversation with Kobach about adding a citizenship question to the 2020 Census came before any request from DOJ to add a citizenship question to the decennial census. RFA response (ECF 146-3 at 39); RFA Response (ECF 146-6 at 2-3).

274. Around that same time, Comstock set out to come up with a "legal rationale" to support the Secretary's request to add a citizenship question. Comstock Dep. 266:4-12 (ECF 175-3). Comstock believed that he needed another agency to request to add the question because OMB and the Paperwork Reduction Act required the Commerce Department to "justify" why a citizenship question was needed, and Comstock understood simply saying that the Secretary wanted to include the question would not "clear [the] legal thresholds." Id. at 153:6-154:11. He testified that it was his job to "help [the Secretary] find the best rationale" for adding the question, because "[t]hat's what a policy person does." Id. at 266:4-267:6.

275. According to Comstock, he did not "need to know what [the Secretary's] rationale might be, because it may or may not be one that is...legally-valid." Id. at 267:10-14. In fact, Comstock testified that the Secretary never told him why he wanted to add a citizenship question to the census. Id. at 251:1-254:17. Teramoto and Department of Commerce Undersecretary Kelley similarly testified that they had no knowledge of why the Secretary wanted to *1027add a citizenship question. Teramoto Dep. 32:5-23 (ECF 175-9); Kelley Dep. 39:3-40:1 (ECF 175-6).

276. Even outside of the Administrative Record, no writing produced in discovery authored by the Secretary or anyone else describes the reasons why the Secretary wanted to add a citizenship question as early as the first quarter of 2017.

277. When Comstock contacted DOJ on May 4, 2017 for the purpose of adding a citizenship question to the 2020 Census, he was not specifically seeking to promote enforcement of the VRA. See Comstock Dep. 167:8-172:5.

278. On July 25, 2017, Secretary Ross had a further telephone conversation with Kobach concerning the addition of the citizenship question to the 2020 Census. RFA Response (ECF 146-3 at 40).

279. Comstock, Teramoto, and Undersecretary Kelley have all denied having any recollection of their September 5 meeting with Secretary Ross to discuss adding the citizenship question. Comstock Dep. 221:1-16; Teramoto Dep. 60:6-61:11; Kelley Dep. 105:5-107:15.

280. Secretary Ross has publicly claimed that "what triggered the investigation, the real study, what triggered the process that led to the determination to [add the citizenship question] was the letter from the Department of Justice." PTX-472. Secretary Ross testified before the House Committee of Ways and Means on March 22, 2018. PTX-346. When asked whether the Department of Commerce planned to add a citizenship question to the 2020 Census, Secretary Ross began his response by stating, "Department of Justice, as you know, initiated the request for inclusion of the citizenship question." Id.

b. Role of DOJ in Requesting the Citizenship Question

281. The Census Bureau informed federal agencies that they were to submit requests for 2020 Census content by July 1, 2016. New York Tr. 995:18-996:3 (Abowd); see also PTX-214. DOJ, however, did not contact the Census Bureau about adding a citizenship question until December 2017. New York Tr. 996:4-10 (Abowd). Indeed, prior to the December 2017 request, DOJ had never communicated to the Census Bureau that ACS CVAP data was not ideal for DOJ's VRA enforcement purposes. Id. at 996:19-23 (Abowd).

282. John Gore, who was responsible for drafting the DOJ request letter, was the Acting Assistant Attorney General (AAAG) for Civil Rights at DOJ at the time. Gore Dep. 18:20-19:3 (ECF 175-4). One of the sections within the DOJ Civil Rights Division is the Voting Section, which enforced Section 2 of the VRA. Id. at 19:7-14.

283. Prior to becoming AAAG, Gore was previously an attorney in private practice. Id. at 14:20-22. As an attorney in private practice, he litigated numerous cases under the VRA on behalf of defendants. Id. at 14:20-16:1. The issue of the adequacy of CVAP data never came up in any of the VRA cases litigated by Gore and he never took the position a plaintiff's block-level CVAP data was insufficient because it was based on sample survey data rather than a "hard count" from the decennial census. Id. at 16:19-17:10.

284. Gore testified he does not personally believe that it is necessary for DOJ's VRA enforcement efforts to collect CVAP data through the census questionnaire. Id. at 300:8-11. He also stated that he has no experience drawing districts for the purposes of complying with the preconditions for VRA liability set forth in Thornburg v. Gingles , 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986), or using block-level data about the characteristics of populations. Gore Dep. 17:21-18:15.

*1028285. Gore became involved in the issue of the citizenship question through conversation with Hankey and then Attorney General Sessions. Id. at 73:2-75:14. On or around Labor Day of 2017, Gore had a conversation with Sessions regarding a citizenship question. Id. at 83:16-84:1.

286. Beginning roughly in mid-September, the Commerce Department initiated direct conversation with Gore. Id. at 91:18-92:6, 94:17-95:5. Throughout the autumn of 2017, Gore spoke with three individuals from the Commerce Department about a citizenship question: Peter Davidson, James Uthmeier, and Wendy Teramoto. Id. at 92:18-94:3. None of these conversations were initiated by Gore or anyone in DOJ's Civil Rights Division. Id. at 94:17-95:5.

287. Gore was first contacted by Davidson in mid-September, who called Gore to discuss adding the citizenship question to the 2020 Census. Id. at 92:18-93:6, 97:19-98:3. Davidson asked Gore to reach out to Teramoto. Id. at 96:4-97:14. The "DOJ-DOC" issue referred to in Gore's email to Teramoto was the addition of a citizenship question to the 2020 Census. Id. ; see also PTX-59. Gore and Teramoto spoke about the citizenship question on or about September 15, 2017, Gore Dep. 102:2-103:4, however Teramoto testified that she had no recollection of this conversation, Teramoto Dep. 74:22-77:20.

288. On or about September 22, 2017, Gore spoke on the phone with Uthmeier about adding a citizenship question to the 2020 Census. Gore Dep. 118:15-17. After their call, Gore was also provided with Uthmeier's August 11 memorandum discussing the addition of a citizenship question. Id. at 118:15-119:4. With the August 11 memorandum, Gore also received a handwritten note from Uthmeier. Id. This handwritten note from Uthmeier contained information that DOJ considered in drafting the final letter requesting a citizenship question. Id. at 123:16-124:2. That note was not produced in the Administrative Record and Defendants withheld it on the basis of privilege.

289. On or about November 1, 2017, Gore wrote the initial draft of the December 12 Letter. Id. at 126:2-127:2, 127:12-17. The only career staffer in the Voting Rights Section to provide input at any stage of drafting the December 12 Letter was Chris Herren, Chief of DOJ's Voting Section. Id. at 151:21-153:22. In early November, Gore emailed Herren, copying Ben Aguiñaga. Id. at 126:2-15. Gore attached the first draft of the letter that he had written requesting a citizenship question be added to the 2020 Census questionnaire and requested Herren's input. Id. at 126:13-127:2.

290. Herren, Aguiñaga, and Bethany Pickett provided substantive feedback on this first draft. Id. at 136:21-137:8. At the time, Aguiñaga and Pickett were political appointees in the front office of the Civil Rights Division who began working there in 2017, having graduated from law school around 2015. Id. at 133:9-134:11.

291. Gore does not recall sharing subsequent drafts of the December 12 Letter with Herren or receiving any additional feedback from him. Id. at 444:17-445:1. Gore has no recollection of receiving any other input from career Civil Rights Division staff. Id. at 152:12-153:5. The only other people in DOJ's Civil Rights Division from whom Gore can recall soliciting or receiving input on the draft December 12 Letter were Pickett and Aguiñaga. Id. at 136:21-137:8.

292. In mid-November Gore discussed adding a citizenship question with Rachael Tucker, then counsel in the Office of the Attorney General, and Robert Troester, then Associate Deputy Attorney General.

*1029Id. at 141:6-14. Neither Tucker nor Troester had experience as counsel in VRA cases, litigating Section 2 redistricting cases involving the use of CVAP data, or otherwise assessing the reliability of CVAP data used in VRA litigation. Id. at 140:1-17.

293. In late November 2017, again Gore solicited and received edits on the draft letter requesting a citizenship question from Tucker and Troester. Id. at 138:12-142:17. Gore did not receive substantive edits from anyone besides Tucker and Troester in the last few days before the December 12 Letter was sent. Id. at 146:7-11.

294. Final authorization to send the letter came from either Tucker or Troester on behalf of Attorney General Sessions in mid-December. Id. at 159:18-160:18. Attorney General Sessions ultimately made the decision for DOJ to request that the Census Bureau ask a citizenship question on the Census. Id. at 442:13-19.

295. Attorney General Sessions subsequently directed DOJ to refuse to meet with the Census Bureau to discuss DOJ's request in the December 12 Letter. Id. at 271:21-272:13.

296. Dr. Hermann Habermann described meetings with a requesting agency, such as the meeting Director Jarmin requested with DOJ on December 22, 2017, as "normal Census Bureau procedure. [Such a meeting] allows the technical experts to better understand how the Census Bureau can meet the needs of the proposers." PTX-821 at ¶¶ 28-29. Dr. Jarmin similarly testified that it was typical for the Census Bureau to meet with federal agencies requesting data in order to understand their needs and come up with the best way to meet those needs. Jarmin Dep. 33:1-15, 36:14-20 (ECF No. 175-5). Dr. Jarmin did not agree with DOJ's reasoning for refusing to meet because the Census Bureau would have liked additional information on how DOJ used CVAP data. Id. at 101:9-20.

297. Dr. Abowd also observed that it is "very unusual" for an agency to make a request to the Census Bureau to collect data through the census but then refuse to meet to discuss the technical aspect of that data request. Tr. 1055:8-12 (Abowd). He also stated that it is "very unusual" for the head of a cabinet agency personally to direct staff not to meet with the Census Bureau to discuss the Census Bureau's ideas for producing better quality data for that agency at a lower cost. Id. at 1055:13-17 (Abowd).

298. It is unknown whether the block-level CVAP data collected with a citizenship question on the 2020 Census will have a margin of error any more precise than the CVAP data on which the Department of Justice currently relies. New York Tr. 1045:19-1046:8 (Abowd). Gore is not aware of any communications between DOJ and the Census Bureau about whether or not adding a citizenship question to the census would in fact produce data that has smaller margins of error than the citizenship data currently used by DOJ, due to required disclosure avoidance techniques. Gore Dep. 228:11-20, 233:4-234:2.

c. Role of the Census Bureau in Evaluating the Citizenship Question

299. Dr. Abowd first learned about DOJ's request to add a citizenship question via email on December 15, 2017. New York Tr. 879:9-17 (Abowd). Following receipt of the December 12 Letter, Dr. Jarmin asked Dr. Abowd to assemble a team of technical experts, nicknamed the SWAT team, to discuss how to respond to the DOJ request. Id. at 878:23-880:5 (Abowd). The SWAT team specifically looked into using administrative records in lieu of a citizenship question on the census because *1030Dr. Jarmin believed that, under Title 13, the Bureau is supposed to use administrative records in lieu of direct collection when possible. Jarmin Dep. 59:9-60:7. A group of Census Bureau decision-makers in collaboration with Department of Commerce Undersecretary Kelley ultimately decided not to conduct a randomized controlled trial of the content of the citizenship question. New York Tr. 925:19-22 (Abowd).

d. January 19 Memo

300. The January 19 Memo to Secretary Ross summarized the opinions of the Census Bureau senior executive staff and was based on the SWAT team's work and other Census Bureau research. Id. at 880:10-18 (Abowd). The January 19 Memo memorializes the Census Bureau's credible, quantitative evidence, as well as its analysis, that adding a citizenship question to the 2020 Census could be expected to lower the self-response rate in households that may contain noncitizens. Id. at 881:4-10 (Abowd).

301. The views in the January 19 Memo are a summary of the technical work performed by the SWAT team and includes contributions made by other senior executives at the Census Bureau. Id. at 883:4-6 (Abowd). Dr. Abowd agrees with the conclusions in the January 19 Memo, id. at 883:7-9 (Abowd), which were ultimately reviewed and approved by Dr. Jarmin, id. at 883:10-12 (Abowd).

302. The analyses in section B(1), (2), and (3) of the memo all support the conclusion that the citizenship question would cause a lower self-response rate to the 2020 Census. Id. at 890:3-891:3 (Abowd). At the time the memo was released, 5.1 percent was the Census Bureau's best estimate of the effect of adding a citizenship question in terms of the question's differential impact of self-responses of noncitizen households as compared to citizen households. Id. at 893:12-22 (Abowd). A reduction in self response of 5.1 of noncitizen households would send more than a million additional people into NRFU. Id. at 894:1-16 (Abowd).

303. The lower self-response rates resulting from adding a citizenship question will increase the cost of conducting the 2020 Census. Id. at 950:10-13 (Abowd). This is because more people will have to be enumerated through NRFU, which costs money. Id. at 950: 15-20 (Abowd). The Census Bureau's estimate that NRFU costs will increase by $ 27.5 million is conservative because, among other things, the differences in self-response rates to the 2020 Census between citizen and noncitizen households is likely to be even greater than estimated in the memo. Id. at 951:11-19 (Abowd). Another reason this figure represents a lower-bound cost estimate is that it may take more NRFU visits to enumerate households that do not respond due to the citizenship question than assumed. Id. at 952:2-6 (Abowd). Moreover, this cost estimate does not incorporate any estimate about the effect of a citizenship question on reducing self-response rates from all-citizen households. Id. at 952:7-11 (Abowd). Finally, this figure does not capture increased communication campaign costs that may be needed as a result of the citizenship question. Id. at 952:12-16 (Abowd).

304. Citizenship status is a characteristic where administrative records tend to be more accurate than survey responses. Id. at 955:21-24 (Abowd). In other words, when an ACS response says that a person is a citizen, but the administrative records says the person is not a citizen, the most likely conclusion is that the person is, in fact, a noncitizen. Id. at 955:6-20 (Abowd). In fact, for more than 30 percent of noncitizens who provide a response to the ACS citizenship question, the response is incorrect. Id. at 956:16-21 (Abowd).

*1031305. The Bureau has no empirical basis to believe that noncitizens for whom a response is provided to the citizenship question on the census will have more accurate responses than they do to the citizenship question on the ACS. Id. at 956:22-957:2 (Abowd). Instead, the Census Bureau found strong indications that responses by noncitizens to a citizenship question on the 2020 Census will be even less accurate than they have historically been on the ACS. Id. at 957:3-7 (Abowd).

306. The Census Bureau still hasn't made any determination about how it will address disagreement between survey responses and the administrative records when producing block-level CVAP data used by the Department of Justice after the 2020 Census. Id. at 957:8-13 (Abowd). The Bureau has concluded, however, that using administrative records would deliver higher quality block-level CVAP data by race and ethnicity than including a citizenship question on the census. Id. at 958:19-22 (Abowd). The Census Bureau's proposal to generate such block-level CVAP data using administrative records rather than a citizenship question had the backing of the Census Bureau's redistricting office. Id. at 958:24-959:3 (Abowd).

307. In the January 19 Memo, the Census Bureau concluded that a citizenship question on the 2020 Census would be a sensitive one for Hispanics. Id. at 917:4-7 (Abowd). This memo also provided empirical support for the conclusion that adding a citizenship question will reduce self-response rates to the 2020 Census. Id. at 922:4-10 (Abowd). Dr. Jarmin agrees with the findings in the January 19 Memo, including that using administrative records would provide higher quality CVAP data at a lower cost than a citizenship question on the 2020 Census. Jarmin Dep. 65:22-67:2, 115:20-117:15.

e. The Set of 35 Questions Answered by the Census Bureau

308. After the Census Bureau communicated its views to Secretary Ross, the Commerce Department sent a list of 35 follow-up questions to the Census Bureau. New York Tr. 1004:19-25 (Abowd). Dr. Abowd was charged with making sure that the responses to the 35 questions were accurate. Id. at 1005:23-1006:1 (Abowd).

309. As of March 1, 2018, it was Dr. Abowd's understanding that adding a new question to the decennial census involves extensive testing, review, and evaluation. Id. at 1007:7-9 (Abowd). The answer to Question 31 found at AR 10900 summarized the Census Bureau and OMB's formal process for making content changes to the census. Id. at 1006:2-1008:8 (Abowd); Jarmin Dep. 137:6-138:21; see also PTX-4D at 1845-46; PTX-133 at 21-22.

310. Dr. Abowd did not write the answer to Question 31 that appeared in the initial administrative record, he does not know who wrote it, and it does not appear in the last version of the document in the possession of the Census Bureau. New York Tr. 1010:12-1011:3 (Abowd). The text in that document is not the text that the Census Bureau transmitted to the Commerce Department, id. at 1014:18-23 (Abowd), and does not represent the final Census Bureau version. Jarmin Dep. 137:6-16. Dr. Jarmin also does not know who wrote the answer to Question 31 that appears in Defendants' initial administrative record submission. Id. at 211:19-21.

f. February 12 Meeting Between Census Bureau and Secretary Ross

311. Dr. Abowd met with Secretary Ross to discuss the January 19 Memo on February 12, 2018. New York Tr. 883:17-19 (Abowd). In addition to Dr. Abowd, the Census Bureau staff attending the meeting were Dr. Jarmin, Dr. Llamas, Associate *1032Director for the 2020 Census Al Fontenot, Assistant Director for the 2020 Census Jim Treat, and Special Assistant to the Director Krista Jones. Tr. 824:10-18 (Abowd). The February 12 meeting also included several members of Secretary Ross's staff. Id.

312. The February 12 meeting was the only meeting Dr. Abowd had with Secretary Ross to discuss the citizenship question before he issued the Decision Memo. New York Tr. 884:11-14 (Abowd). Prior to the meeting with Secretary Ross, Dr. Abowd had a pre-meeting with Undersecretary Kelley to discuss the memo. Id. at 883:20-884:1 (Abowd). During that meeting, she did not express any disagreements with the analysis in the January 19 Memo. Id. at 884:2-5 (Abowd).

313. At the February 12 meeting, Ross quickly dismissed Alternative A (not collecting block-level CVAP data) as a possibility. Tr. 826:20-25 (Abowd).

314. Dr. Abowd informed Secretary Ross during the February 12 meeting that the Census Bureau thought that the difference in self-response rates on the ACS and the census, when comparing citizen and noncitizen households, was related to the citizenship question on the ACS. New York Tr. 922:11-17 (Abowd). He also explained the Census Bureau's conclusion that using administrative records would better meet DOJ's stated uses than relying on a citizenship question on the census. Id. at 959:12-19 (Abowd).

315. After the February 12 meeting, Dr. Jarmin told Dr. Abowd that Secretary Ross and Undersecretary Kelley wanted Abowd to evaluate Alternative D (using both administrative records and citizenship question responses). Id. at 965:25-966:5 (Abowd).

g. March 1 Memo

316. The views in the March 1 memo are those of the senior executive staff at the Census Bureau. New York Tr. 966:23-25 (Abowd). The Census Bureau did not recommend Alternative D and still does not recommend Alternative D. Id. at 967:16-21 (Abowd). Instead, the memo concluded that using administrative records alone would be more accurate than attempting to combine administrative records and survey responses under Alternative D. Id. at 988:12-16 (Abowd).

317. As the memo explains, survey-collected citizenship data would not be reliable for many of the people falling in the gaps in the administrative records. Id. at 973:23-974:3 (Abowd). The memo specifically states that citizenship survey data gathered under Alternative D would be of "suspect quality" whereas administrative data on citizenship is "high quality." Id. at 974:12-20 (Abowd).

318. Accordingly, for the portion of the population that cannot be linked to administrative records, it would be more accurate to impute/model their citizenship status based on administrative records (Alternative C) than to obtain the information through their survey responses (Alternative D). Id. at 974:22-975:9 (Abowd). This is the view of both Dr. Abowd and the Census Bureau. Id. at 974:22-975:15.

319. Moreover, under Alternative D, due to the lower quality personal data on census responses from an increased number of households going through NRFU, there will be a reduction in the number of individuals whom the Census Bureau can link to administrative records. Id. at 969:2-23 (Abowd). In other words, the number of individuals that could be matched to administrative records that contain citizenship information would be higher under *1033Alternative D than under Alternative C. Id. at 975:17-21 (Abowd).

320. The Census Bureau also concluded in this memo that Alternative C is cheaper than Alternative D. Id. at 988:9-11 (Abowd).

321. These analyses and conclusions were communicated to Secretary Ross before he issued his Decision Memo. Id. at 988:17-19 (Abowd).

h. Secretary Ross's March 26 Decision Memorandum

322. Although Secretary Ross repeatedly claims in the Decision Memo that there was a lack of evidence that the citizenship question would lower self-response rates, Dr. Abowd provided Secretary Ross with credible, quantitative evidence that doing so would lower the self-response rate for households that contain a noncitizen. Tr. 1059:16-21 (Abowd).

323. The Decision Memo also states that Option D will provide more "complete and accurate" citizenship data than using administrative records alone. PTX-26 at 5, 7. However, the Census Bureau's analysis concluded that adding a citizenship question is not necessary to provide complete and accurate data in response to the Department of Justice's request. Tr. 1063:18-22 (Abowd).

324. The Decision Memo stated:

Finally, placing the question on the decennial census and directing the Census Bureau to determine the best means to compare the decennial census responses to administrative records will permit the Census Bureau to determine the inaccurate response rate for citizens and noncitizens alike using the entire population. This will enable the Census Bureau to establish, to the best of its ability, the accurate ratio of citizen to noncitizen responses to impute for that small percentage of cases where it is necessary to do so.

PTX-26 at 5. However, as of March 26, 2018, the Census Bureau had not analyzed these presumptions. New York Tr. 977:25-978:7 (Abowd). The presumptions were never discussed with Dr. Abowd and the Census Bureau does not agree with them. Id. at 976:18-977:24 (Abowd). In fact, adding a citizenship question will make it more difficult for the Census Bureau to establish the accurate ratio of citizen to noncitizen responses to impute. Tr. 1061:8-11 (Abowd).

325. The March 26 Decision Memo states that "no one has identified any mechanism" for determining whether the citizenship question would cause a drop in self-response. PTX-26 at 5. There were, however, several mechanisms available to determine whether the citizenship question would cause people not to participate in the census. Tr. 1061:23-1062:11 (Abowd). One such mechanism was the statistical analysis performed by the Census Bureau. Id. at 1061:23-1062:3 (Abowd). Another mechanism would have been an RCT, which the Census Bureau could have conducted, but did not. Id. at 1062:4-11 (Abowd). A group-of decision-makers in collaboration with Undersecretary Kelley decided not to conduct a randomized controlled trial of the content of the citizenship question. New York Tr. 925:19-22 (Abowd).

2. Testing Requirements for New Questions

a. The Census Bureau's Statistical Quality Standards

326. The Census Bureau's Statistical Quality Standards set forth the Bureau's internal standards, guidelines, and requirements on pretesting questionnaires and data collection instruments. PTX-205; Tr. 82:6-19 (O'Muircheartaigh), 832:1-833:8 *1034(Abowd). The Statistical Quality Standards apply when new questions are added to a data collection instrument or existing questions are revised. PTX-205 at 18.

327. Sub-Requirement A2-3.3 of the Statistical Quality Standards requires that "[d]ata collection instruments and supporting materials must be pretested with respondents to identify problems (e.g., problems related to content, order/context effects, skip instructions, formatting, navigation, and edits) and then refined, prior to implementation, based on the pretesting results."Id. at 18; Tr. 82:9-84:13 (O'Muircheartaigh), 833:11-18 (Abowd).

328. Under Sub-Requirements A2-3.3-1c and A2-3.3-1d, pretesting must be performed when "Review by cognitive experts reveals that adding pretested questions to an existing instrument may cause potential context effects" and when "An existing data collection instrument has substantive modifications (e.g., existing questions are revised or new questions added)." PTX-205 at 18.

329. One exception to the pretesting requirement of the Statistical Quality Standards is that, "On rare occasions, cost or schedule constraints may make it infeasible to perform complete pretesting. In such cases, subject matter and cognitive experts must discuss the need for and feasibility of pretesting. The program manager must document any decisions regarding such pretesting, including the reasons for the decision. If no acceptable options for pretesting can be identified, the program manager must apply for a waiver." Id. ; Tr. 833:19-834:7, 1046:18-1047:19 (Abowd).

330. Another exception is that, "Pretesting is not required for questions that performed adequately in another survey." PTX-205 at 18; Tr. 833:19-834:7, 1047:12-19 (Abowd). Those and other similar pretesting standards are used in the survey methodology and data collection profession more generally. Tr. 84:19-21 (O'Muircheartaigh).

b. OMB's Standards and Guidelines for Statistical Surveys

331. Under Congress's direction, the Office of Management and Budget (OMB) has also issued standards for designing, developing, and pretesting survey content. PTX-821 at ¶¶ 55-56; PTX-262; PTX-266; PTX-267; PTX-612. The OMB-promulgated standards for pretesting content on data collection instruments can be found in the OMB Standards and Guidelines for Statistical Surveys. PTX-266.

332. The Census Bureau must follow the OMB Standards and Guidelines for Statistical Surveys when preparing for and implementing the decennial census. Census Bureau 30(b)(6) Dep. Vol. I 321:14-17; New York Tr. 989:15-17 (Abowd); PTX-821 at ¶¶ 55-56; Tr. 88:22-89:12 (O'Muircheartaigh). These guidelines require that agencies conduct a pretest of all components of a survey, including by conducting a field test and full "dress rehearsal" for "highly influential surveys." PTX-266 at 14.

333. OMB Standard 1.4 requires that agencies "ensure that all components of a survey function as intended when implemented in the full-scale survey and that measurement error is controlled" prior to implementing the data collection instrument. Id. at 14; PTX-821 at ¶ 56. This is done either by "conducting a pretest of the survey components" or by "having successfully fielded the survey components on a previous occasion." PTX-266 at 14.

334. OMB Standard 2.3 states that "[a]gencies must design and administer their data collection instruments and methods in a manner that achieves the best balance between maximizing data quality and controlling measurement error while minimizing respondent burden and cost."

*1035Id. at 16. OMB Guideline 2.3.1 similarly demands that agencies "[d]esign the data collection instrument in a manner that minimizes respondent burden, while maximizing data quality." Id. at 16.

335. The Census Bureau has conceded that-within the meaning of OMB Standard 2.3-Alternative D would result in lower data quality, higher cost, and higher respondent burden than the Census Bureau's recommended Alternative C. Census Bureau 30(b)(6) Dep. Vol. I 321:18-322:19; New York Tr. 989:6-990:6 (Abowd). While Alternative C would comport with OMB Guideline 2.3.1, New York Tr. 990:7-991:1 (Abowd), Alternative D would not, given the degradation to data quality that would result.

c. Census Bureau Process for Adding or Modifying Census Content

336. In addition to abiding by the standards set forth above, the Census Bureau follows a well-established process for adding or modifying questions on the decennial census. This decade-long process involves multiple tests, including various randomized control trials. New York Tr. 994:18-22 (Abowd). It also involves extensive cognitive and field testing, ongoing research, and input from advisory committees. Id. at 996:24-997:14 (Abowd); PTX-214 at 4; Jarmin Dep. 47:13-48:17, 52:5-11, 138:16-139:19.

337. Another standard pretesting method is the randomized control trial (RCT), which tests an operation with an added element and compares that to a test of the operation without the element. Tr. 102:7-23 (O'Muircheartaigh). According to Dr. Abowd, the RCT is the "gold standard" for testing a proposed question's effect on the census count and data collection. Id. at 874:10-23, 1039:10-17 (Abowd); New York Tr. 923:16-924:9 (Abowd); Census Bureau 30(b)(6) Dep. Vol. II 426-430.

338. Based on the result of pretesting, the Census Bureau must finalize the actual 2020 Census questionnaires and then must submit them for OMB approval of the 2020 Census information collection. PTX-821 at ¶ 62.

d. Past Practices for Testing the Decennial Census Questionnaire

339. Past decennial census questionnaires-the complete 2010 Census questionnaire, for example-were subject to extensive cognitive testing and field testing. New York Tr. 997:11-23 (Abowd). After the 1990 Decennial Census, the Census Bureau investigated the possibility of adding a question concerning respondents' Social Security numbers on the decennial census short form questionnaire. Id. at 998:25-999:4 (Abowd).

340. To test that potential Social Security number question, the Census Bureau conducted an RCT comparing a version of the short form with the Social Security number question and one without. Id. at 999:5-8 (Abowd). That RCT allowed the Bureau to assess the impact on self-response rates of a Social Security number question. Id. at 999:9-11.

341. In that RCT, the self-response rate fell off in the group that had the Social Security number question by 3.4 percent. Id. at 999:12-15. The conclusion drawn from that RCT was that asking for a Social Security number would be sensitive. Id. at 999:16-18 (Abowd). As a result, the Census Bureau decided not to include a Social Security number question on the decennial census questionnaire. Id. at 999:19-24 (Abowd). The Census has never requested Social Security numbers on the census questionnaire, and one of the reasons is the effect of the question on self-response rates, as revealed by the RCT. Id.

342. The RCT to assess the impact of a Social Security number question was conducted *1036before any decision was made about whether to include a Social Security number question on the decennial census. Id. at 1000:8-13 (Abowd).

3. Whether ACS Data is Sufficient for VRA Enforcement Purposes

343. Plaintiffs offer the expert testimony of Dr. Lisa Handley and Professor Pamela S. Karlan regarding whether the inclusion of a question about citizenship status in the decennial census would assist in the enforcement Section 2 of the VRA.

344. Dr. Handley is a consultant in redistricting and in electoral district design. She has over thirty years of experience as an expert in redistricting, minority voting rights, and the use of census data for voting rights enforcement purposes, advising governments, non-profits, and NGOs on minority voting rights and redistricting-related issues and serving as an expert in dozens of voting rights cases, including five Section 2 redistricting cases on behalf of DOJ. PTX-819 at 788:22-796:3 (Handley). Based on her education, experience, and knowledge, Dr. Handley is well-qualified to offer reliable and credible opinions on Section 2 of the VRA, and the use of census data in Section 2 litigation and enforcement proceedings.

345. Dr. Handley testified that in her professional opinion "currently available census data has proven perfectly sufficient to ascertain whether an electoral system or redistricting plan dilutes minority votes." Id. at 796:22-797:12, 819:19-23 (Handley); PTX-650.

346. Citizenship data is most relevant to the first Gingles precondition, which is that the minority group in question be sufficiently large and geographically compact to constitute a majority in a single-member district. Id. at 798:6-799:19 (Handley); see also Karlan Trial Dep. 30:25-32:14 (ECF 145). Dr. Handley testified that, in her experience, CVAP estimates at the census tract or block group level are generally sufficient to satisfy the first Gingles precondition in Voting Rights Act cases. PTX-819 at 807:24-811:6 (Handley).

347. Moreover, Dr. Handley explained, where it would be helpful to present CVAP data at the block level, this information can be reliably and accurately estimated by applying the CVAP ratios from the census tract level to the block-level figures for total voting-age population. Id. at 808:10-815:5 (Handley). In Dr. Handley's expert opinion, block-level CVAP estimates derived from ACS data at the census tract or block group levels are reliable and accurate. Id. at 815:8-819:23, 855:23-856:10 (Handley).

348. Dr. Handley further explained that, in the district-specific functional analysis that she employs in VRA analysis, which is also used by DOJ, the outcome does not depend on a precise measurement of CVAP at the individual block level, but rather on an analysis of turnout rates and voting patterns within a district. The exact number of minority citizens of voting age at the block level is "essentially irrelevant" to the analysis. Id. at 820:2-823:11 (Handley).

349. Dr. Handley's work as a VRA expert has never been impeded by her use of 5-year ACS CVAP data, and she is also not aware of any VRA claim that failed due to a plaintiff's reliance upon 5-year ACS CVAP data. Id. at 832:14-21 (Handley).

350. Professor Karlan has served as Deputy Assistant Attorney General in DOJ's Civil Rights Division from January 2014 through September 2015, overseeing the work of the Voting Section, which enforces the VRA. She has also served as assistant counsel to the NAACP Legal Defense Fund, litigating voting rights cases *1037on behalf of plaintiffs and amici curiae , including numerous cases brought under Section 2 of the VRA before the U.S. Supreme Court. Karlan Trial Dep. 10:9-21, 12:18-14:7, 22:3-15. She has co-authored two casebooks which covers the VRA, among other topics, and has written approximately one dozen academic articles about the VRA. Id. at 18:2-20:3; 24:22-25:21.

351. Based on her education, experience, and knowledge, Professor Karlan is well-qualified to offer reliable and credible opinions on Section 2 of the VRA, and the use of census data in Section 2 litigation and enforcement proceedings.

352. Professor Karlan testified that, in her professional opinion, existing data sources from the ACS are sufficient for plaintiffs to bring and prevail in cases brought under Section 2 of the VRA. Id. at 29:14-23, 66:15-23. She specifically observed that no reported Section 2 case has ever failed on account of the purported inadequacy of ACS data (or, prior to the advent of the ACS, data from the longform census questionnaire) as a measure of CVAP. Id. at 52:14-53:18.

353. Nothing in the DOJ's December 12 Letter altered Professor Karlan's professional opinion that existing data sources are sufficient for plaintiffs to bring and prevail in litigation under Section 2 of the Voting Rights Act. Id. at 66:24-67:20. Indeed, the December 12 Letter did not identify any cases in which the inaccuracy or inadequacy of ACS data caused a plaintiff to lose a Section 2 case. Id. at 54:5-15.

E. Conclusions of Law

"A person suffering legal wrong because of agency action...is entitled to judicial review thereof." 5 U.S.C. § 702. Under the APA, "[t]he reviewing court shall...hold unlawful and set aside agency action, findings, and conclusions found to be," among other things, "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A).

1. "Contrary to Law" Review

An agency decision may also violate the APA if the agency action is contrary to law. 5 U.S.C. § 706(2)(A) ; Nat'l Wildlife Fed'n v. U.S. Army Corps of Eng'rs , 384 F.3d 1163, 1166 (9th Cir. 2004). Based on the Administrative Record alone, the decision to add the citizenship question violates the APA as contrary to law.

a. Section 6(c)

Section 6(c) of the Census Act requires the Secretary to use administrative records to address DOJ's data request rather than adding a citizenship question on the census. Although Congress delegated to the Secretary a degree of discretion in conducting the census, section 6(c), among other provisions, limits that discretion. Title 13, section 6 states in full:

(a) The Secretary, whenever he considers it advisable, may call upon any other department, agency, or establishment of the Federal Government, or of the government of the District of Columbia, for information pertinent to the work provided for in this title.

(b) The Secretary may acquire, by purchase or otherwise, from States, counties, cities, or other units of government, or their instrumentalities, or from private persons and agencies, such copies of records, reports, and other material as may be required for the efficient and economical conduct of the censuses and surveys provided for in this title.

(c) To the maximum extent possible and consistent with the kind, timeliness, quality and scope of the statistics required, the Secretary shall acquire and use information available from any source referred to in subsection (a) or *1038(b) of this section instead of conducting direct inquiries.

13 U.S.C. § 6. Subdivision (c) of section 6 was added to the statute in the 1976 Census Act. See 1976 Census Act § 5(a), 90 Stat. at 2460. So while the statute previously merely authorized the use of government records for census-related purposes, the amendment made the use of those records a mandatory alternative to "direct inquiries" in certain circumstances. Subdivision (c) of section 6 serves "the dual interests of economizing and reducing respondent burden." H.R. CONF. REP. No. 94-1719, at 10 (1976), reprinted in 1976 U.S.C.C.A.N. 5476, 5478.

The Administrative Record here demonstrates that it was "possible" to "acquire and use" administrative records from other government agencies that would produce data "consistent with the kind, timeliness, quality and scope of the statistics required." 13 U.S.C. § 6(c). The "kind" of data here would be the same regardless of whether it is gathered by administrative records or a census question-in both cases, the relevant data is simply block-level data on citizens versus noncitizens. Moreover, the Census Bureau advised Secretary Ross in the March Memo that regardless of whether administrative data or a citizenship question were used, there was no difference in timing on when the citizenship data would be available, PTX-133 at 11, and there is no evidence in the Administrative Record to the contrary.

Finally, the Census Bureau repeatedly advised Secretary Ross that the quality of the citizenship data would be higher from administrative records than from a citizenship question due to noncitizens' propensity to self-report as citizens. See Parts IV.C.5.d, IV.C.7, supra . The scope of the data would be the same, regardless of source, because data would be obtained for residents of the entire country (with a minority requiring imputation, regardless of the data source).

Thus, under every criterion set forth in section 6(c), with no evidence in the record to the contrary, using administrative records alone was superior to adding a citizenship question to the decennial census. Secretary Ross was therefore legally required to obtain citizenship data through administrative records rather than a citizenship question on the census. His decision to do otherwise violates the APA.

b. Section 141(f)

Adding a citizenship question to the 2020 Census is also contrary to law because Secretary Ross changed the subjects to be included on the Census after submitting his Section 141(f)(1) report despite the absence of any new circumstances justifying such a change. Section 141(f) provides:

(f) With respect to each decennial and mid-decade census conducted under subsection (a) or (d) of this section, the Secretary shall submit to the committees of Congress having legislative jurisdiction over the census-

(1) not later than 3 years before the appropriate census date, a report containing the Secretary's determination of the subjects proposed to be included, and the types of information to be compiled, in such census;

(2) not later than 2 years before the appropriate census date, a report containing the Secretary's determination of the questions proposed to be included in such census; and

(3) after submission of a report under paragraph (1) or (2) of this subsection and before the appropriate census date, if the Secretary finds new circumstances exist which necessitate that the subjects, types of information, or questions contained in reports *1039so submitted be modified, a report containing the Secretary's determination of the subjects, types of information, or questions as proposed to be modified.

13 U.S.C. § 141(f). Secretary Ross submitted his section 141(f)(1) report in March of 2017. PTX-264. That report did not include the subject of citizenship as a topic for the 2020 Census. Id. In March of the following year, Secretary Ross submitted his section 141(f)(2) report.¹⁹ Consistent with his Decision Memo, that report states a citizenship question will be included on the 2020 Census. Secretary Ross has not, however, submitted a report pursuant to subdivision (f)(3), nor did he identify the "new circumstances" that justified the addition of this new topic.

Defendants argue that only Congress can enforce section 141(f) because courts cannot redress any injury resulting from an inadequate report. Guerrero v. Clinton , 157 F.3d 1190, 1191 (9th Cir. 1998) ; Renee v. Duncan , 686 F.3d 1002, 1016-17 (9th Cir. 2012). In both Guerrero and Renee the Ninth Circuit held that the courts could not police the quality of congressional reports that were "purely informational" and from which no legal consequences flowed. Guerrero , 157 F.3d at 1194-1195 ; Renee , 686 F.3d at 1016-17.

While Defendants are correct that the primary function of section 141(f) is to impose reporting requirements, it also imposes substantive limitations on the Secretary's ability to modify the census. This section specifically prevents the Secretary of Commerce from changing the topics or questions to be included in the census after submission of the (f)(1) or (f)(2) reports absent "new circumstances" justifying the change. Therefore, while Defendants present a credible argument that the reporting requirement is generally not subject to judicial review, they do not address section 141(f)(3)'s substantive restrictions on the Secretary's authority to make last minute changes to the census without good cause, which is judicially reviewable.

Section 141(f)(3) requires that, "if the Secretary finds new circumstances exist which necessitate that the subjects, types of information, or questions" included in the census be modified, he submit a new report detailing these modifications. 13 U.S.C § 141(f)(3). The plain meaning of this provision is that the Secretary may not deviate from the subjects and questions outlined in his or her reports to Congress absent new circumstances. Construing the statute in any other manner would render the "new circumstances" clause utterly meaningless. This runs afoul of one of the most basic canons of construction, that a "statute should be construed so that effect is given to all its provisions, so that no part will be inoperative or superfluous, void or insignificant." Corley v. United States , 556 U.S. 303, 314, 129 S.Ct. 1558, 173 L.Ed.2d 443 (2009) (quoting Hibbs v. Winn , 542 U.S. 88, 101, 124 S.Ct. 2276, 159 L.Ed.2d 172 (2004) ).

The only interpretation of the statute that gives full effect to all provisions is that, once the Secretary has submitted either the (f)(1) or (f)(2) reports, he or she may not deviate from the contents set forth in those reports absent "new circumstances" that necessitate a late-stage change to the census. Here, this inquiry ultimately collapses with the pretext question discussed above. Indeed, the only new circumstances identified by Defendants is *1040the submission of the DOJ request letter. As previously explained, the VRA enforcement rationale is nothing more than a pretext designed to provide cover for the Secretary's unexplained desire to add the citizenship question to the census. Just as the DOJ letter does not provide the true basis for Secretary Ross's decision, neither may it qualify as a new circumstance that requires a change to the census. Accordingly, the Secretary's decision to add the citizenship question was contrary to law for this reason as well.

2. Arbitrary and Capricious Review

The standard for evaluating whether an agency's decision was arbitrary and capricious is whether the decision "was the product of reasoned decisionmaking." State Farm , 463 U.S. at 52, 103 S.Ct. 2856. This standard is deferential, Pac. Dawn LLC v. Pritzker , 831 F.3d 1166, 1173 (9th Cir. 2016), and does not permit the Court to "substitute its judgment for that of the agency," Ctr. for Bio. Diversity v. Zinke , 868 F.3d 1054, 1057 (9th Cir. 2017). The focus at all times must remain on whether the agency "considered the relevant factors and articulated a rational connection between the facts found and the choices made." Nw. Ecosys. All. v. U.S. Fish & Wildlife Serv. , 475 F.3d 1136, 1140, 1145 (9th Cir. 2007) (quotation omitted).

a. Review Based on the Administrative Record Alone

i. Pretext

The APA requires an agency decision-maker to "disclose the basis of its" decision and to "give clear indication that it has exercised the discretion with which Congress has empowered it." Burlington Truck Lines, Inc. v. United States , 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962) ; accord Federal Power Comm'n v. Texaco Inc. , 417 U.S. 380, 396, 94 S.Ct. 2315, 41 L.Ed.2d 141 (1974). Where the agency decision-maker fails to disclose the substance of relevant information that has been presented to it, the court "must treat the agency's justifications as a fictional account of the actual decisionmaking process and must perforce find its actions arbitrary." See Home Box Office, Inc. v. F.C.C. , 567 F.2d 9, 54-55 (D.C. Cir. 1977) ; see also U.S. Lines, Inc. v. Federal Maritime Comm'n , 584 F.2d 519, 534-535 (D.C. Cir. 1978) (explaining that the basis of an agency's decision must be disclosed, at the very latest, in the final decision to permit meaningful judicial review).

Secretary Ross violated the APA by failing to disclose the basis for his decision to add a citizenship question to the 2020 Census. As explained in the Findings of Fact Based Exclusively on the Administrative Record, the evidence overwhelmingly shows that Secretary Ross decided to add the citizenship question well before DOJ made the request in December of 2017 and that his reason for doing so was not to improve enforcement of Section 2 of the VRA. This purported purpose was a mere pretext.

While there is some evidence in the Administrative Record that Secretary Ross's interest in the citizenship question was related to the inclusion of noncitizens in the apportionment count, there is no need definitively to ascertain the Secretary's true purpose. For the purposes of the APA, it is relevant only that Section 2 enforcement did not supply the true basis of Secretary Ross's decision and that he disclosed no other basis. Secretary Ross has therefore violated the APA by failing to disclose the actual basis for his decision to add a citizenship question to the 2020 Census.

ii. Whether Secretary Ross Considered All Relevant Factors

An agency action must also be set aside as arbitrary and capricious if the *1041agency fails "to consider an important aspect of the problem." State Farm , 463 U.S. at 43, 103 S.Ct. 2856 ; see also SecurityPoint Holdings, Inc. v. Transp. Sec. Admin. , 769 F.3d 1184, 1188 (D.C. Cir. 2014) (vacating agency order where agency failed to consider potential harms of its changes to an airport advertising program); Stewart v. Azar , 313 F.Supp.3d 237, 263 (D.D.C. 2018) (vacating HHS Secretary's waiver of several requirements of expanded Medicaid because "[f]or starters, the Secretary never once mentions the estimated 95,000 people who would lose coverage, which gives the Court little reason to think that he seriously grappled with the bottom-line impact on healthcare" (emphasis in original) ).

Defendants failed to consider the potential harms the citizenship question could cause to the accuracy of the Census Bureau's final enumeration, and therefore to the allocation of federal funding and apportionment of congressional representation. The Decision Memo states that "[t]he citizenship data provided to DOJ will be more accurate with the question than without it, which is of greater importance than any adverse effect that may result from people violating their legal duty to respond." PTX-26 at 7. Secretary Ross does not, however, articulate the potential harms that he is allegedly weighing. In particular, he does not grapple in any meaningful way with the possibility of that a severe differential decline in self-response rates could ultimately affect congressional apportionment. Instead, the Secretary based his decision to add the citizenship question upon a purported lack of evidence about the effects of the citizenship question on the census-specifically self-response rates. Part IV.C.10.

The Secretary attempts to justify barreling ahead in the face of this alleged uncertainty by stating that "no one has identified any mechanism" for evaluating the impact of the citizenship question on the census. id. This statement is, frankly, perplexing. It is notable that Secretary Ross does not claim that the Census Bureau lacks the tools to test the citizenship question. Nor does Secretary Ross claim the Census Bureau advised him that testing the effects of the citizenship question was beyond the Bureau's ability. Ultimately, it is highly implausible that Secretary Ross was unaware that other testing mechanisms, beyond the Census Bureau's natural experiment, existed.

Secretary Ross's failure to investigate and consider the likely effects of the citizenship question on the accuracy of the Census Bureau's enumeration-and therefore on congressional apportionment and the allocation of federal funding-was an abdication of his duty to consider all relevant factors before making his decision. Therefore, the Secretary's decision to prioritize the inclusion of the citizenship question on the census over any harm that might result was necessarily arbitrary and capricious.

iii. Whether the Decision Ran Counter to the Evidence

An agency action is also arbitrary and capricious under the APA if the agency offers "an explanation for its decision that runs counter to the evidence before the agency." State Farm , 463 U.S. at 43, 103 S.Ct. 2856. An agency must "articulate a satisfactory explanation for its action including a 'rational connection between the facts found and the choice made.' " Id. (quoting Burlington Truck Lines , 371 U.S. at 168, 83 S.Ct. 239. Where a decision-maker adopts a "plainly inferior" course of action, that decision is arbitrary and capricious. Pub. Citizen, Inc. v. Mineta , 340 F.3d 39, 56 (2d Cir. 2003).

*1042Defendants' decision to add the citizenship question to the 2020 Census is contrary to the evidence. This decision is predicated on the assertion in the Decision Memo that adding a citizenship question on the 2020 Census will result in the "most complete and accurate" citizenship data for DOJ's stated purpose of VRA enforcement. PTX-26 at 1 ("The Department and Census Bureau's review of the DOJ request-as with all significant Census assessments- prioritized the goal of obtaining complete and accurate data " (emphasis in original); see also id. at 5 ("It is my judgment that Option D will provide DOJ with the most complete and accurate CVAP data in response to its request"), 7 ("[h]owever, even if there is some impact on responses, the value of more complete and accurate data derived from surveying the entire population outweighs such concerns"), 8 ("To conclude, after a thorough review of the legal, program, and policy considerations, as well as numerous discussions with the Census Bureau leadership and interested stakeholders, I have determined that reinstatement of a citizenship question on the 2020 Decennial Census is necessary to provide complete and accurate data in response to the DOJ request.").

In light of this stated goal, Defendants' decision to add a citizenship question to the 2020 Census runs counter to the evidence in the Administrative Record that shows adding a citizenship question will result in citizenship data less accurate and no more complete than citizenship data gathered through administrative records alone (Alternative C). This was the conclusion of every scientific analysis in the Administrative Record that addressed the issue, and this conclusion was repeatedly communicated to Secretary Ross. See, e.g. , PTX-22 at 1 (January 19 Memo) (explaining that adding the citizenship question would result in "substantially less accurate citizenship status data than are available from administrative sources"); PTX-25 at 5 (March 1 Memo) ("Alternative D would result in poorer quality citizenship data than Alternative C.").

The Census Bureau's analyses offered several explanations for the difference in accuracy between a census response and administrative records data. First, the citizenship data in administrative records is "very accurate" because that data point requires people to have provided proof of citizenship or legal resident alien status. PTX-101 at 3; PTX-22 at 7. Second, citizenship data that is self-reported in surveys is inaccurate for noncitizens. Historical census and ACS data show that noncitizens misreport themselves as citizens "for no less than 23.8% of the cases, and often more than 30%." PTX-22 at 7.

Third, the Census Bureau found that lowered self-response rates due to the citizenship question will decrease the number of people who can be linked to administrative records (which contain citizenship information), because the personal identifying information gathered in NRFU is lower quality than personal identifying information gathered through self-response. PTX-22 at 2; PTX-25 at 4. Fourth, imputation will be less accurate if based in part on self-reported citizenship data rather than administrative records alone. This is because many noncitizens inaccurately report that they are citizens, therefore the imputation model will be biased under Option D. In contrast, the imputation model under Alternative C would be benchmarked to accurate administrative records and therefore would not suffer from this bias.

Notably, there is no evidence in the Administrative Record supporting Secretary Ross's assertion that self-reported citizenship *1043data is more accurate than citizenship data from administrative records. Thus, all of the evidence shows that citizenship information gathered through a citizenship question on the Census (Alternative D) will be less accurate than citizenship information gathered through administrative records (Alternative C).

All of the evidence in the Administrative Record also shows that Option D will not yield more "complete" data than Alternative C. Secretary Ross implies in the Decision Memo that citizenship data from administrative records would be incomplete because using administrative records alone would require imputation of citizenship status for 10 percent of the population, whereas a citizenship question on the 2020 Census "may eliminate the need for the Census Bureau to have to impute an answer for millions of people." PTX-26 at 4, 5.

In fact, the Census Bureau estimates that under both alternatives millions of people would need their citizenship status imputed and the total number of people assigned a citizenship status would be approximately the same (330 million). PTX-24 at 1-4. In sum, all of the evidence in the Administrative Record shows that adding a citizenship question to the 2020 Census would yield citizenship data that is less accurate and no more complete than gathering that data using administrative records alone. The decision to add the citizenship question was therefore arbitrary and capricious in violation of the APA.

The decision was also counter to the evidence because it was replete with flawed assertions that are either not based on any evidence or contrary to the evidence in the Administrative Record. First, the Decision Memo repeatedly claimed that there was no evidence the citizenship question would cause a drop in self response, and that "no one has identified any mechanism" for obtaining such evidence. PTX-26 at 3-5. This assertion is contrary to the evidence in the Administrative Record. The Census Bureau performed a scientific analysis leading to an estimate that 5.1 percent of households with at least one noncitizen would not respond to the census due to the citizenship question. The Bureau repeatedly communicated that estimate to Secretary Ross. PTX-22 at 4; PTX-148 at 6-7. Nothing in the Administrative Record supports a contrary conclusion.

Second, the Decision Memo states that asking the citizenship question of all people "may eliminate the need for the Census Bureau to have to impute an answer for millions of people." PTX-26 at 5. However, the Census Bureau had estimated that with a citizenship question on the census, it would still have to impute the citizenship data of 13.8 million people. PTX-24 at 2. Nothing in the Administrative Record supports a contrary conclusion.

Third, the Decision Memo states that Option D "would maximize the Census Bureau's ability to match the decennial census responses with administrative records," PTX-26 at 4, so as to allow for "more complete" citizenship data. However, the Administrative Record reflects that because adding a citizenship question would drive down the self-response rate and put more households into NRFU operations, Option D actually reduces the Census Bureau's ability to match survey responses with administrative records. PTX-25 at 4.

Fourth, the Decision Memo also attempts to justify Option D by stating that adding the citizenship question to the census "will permit the Census Bureau to determine the inaccurate response rate" with respect to the citizenship question, and suggests that this would improve the *1044accuracy of the imputation model. PTX-26 at 5. However, nowhere in the Administrative Record, including in the Abowd memoranda, does the Census Bureau state that adding a citizenship question would increase the accuracy of its estimate of inaccurate citizenship responses. Nor is it apparent from the Administrative Record why the inaccuracy rate of responders would help impute the citizenship data of nonrespondents. If citizenship is benchmarked to administrative records, and the Bureau would be using those records in any event, then adding a census question would not assist in the imputation. Accordingly, Defendants' decision is arbitrary and capricious because these key statements in the Decision Memo, purportedly justifying the choice of "Option D," were counter to the evidence.

b. Extra-Record Evidence Confirms the Secretary's Decision was Arbitrary and Capricious

Extra-record evidence confirms that Secretary Ross's justification for his decision to add the citizenship question was pretextual. Several facts are particularly striking. First, Secretary Ross's senior officials at the Commerce Department all claim, rather implausibly, to be ignorant of why Secretary Ross wanted the citizenship question on the 2020 Census. Even Comstock, the Director of Policy in charge of soliciting the DOJ request for the question, claims that he never asked Secretary Ross to explain his reasoning. Comstock Dep. 171-72. This suggests either that, despite several months of discussion, Secretary Ross kept his senior staff in the dark about his reason for wanting to include the citizenship question or that his staff are dissembling in order to avoid revealing Secretary Ross's true purpose. Both of these conclusions suggest that Secretary Ross does not wish his reason for requesting the inclusion of the citizenship question on the 2020 Census to come to light.

Second, DOJ's December 12 Letter requesting the citizenship question was drafted with limited involvement from VRA career staff. Gore initially prepared the letter and sent an early draft to Chris Herren, the Chief of the Voting Rights Section, on November 1, requesting feedback. This was the only involvement by any career staff in the Voting Rights Section, even though the letter continued to be reviewed and revised by political appointees for more than a month afterwards. Id. The limited involvement of VRA staff in this process casts doubts on the true need for this data to support VRA enforcement. This bolster's the conclusion that the December 12 Letter was a mere pretext orchestrated by Secretary Ross, rather than a sincere request upon which the Secretary reasonably relied.

Third, the Administrative Record reveals that DOJ refused to meet with Census Bureau staff to discuss its request that a citizenship question be added to the 2020 Census. Extra-record evidence further establishes that DOJ's refusal to meet and discuss its data needs was both "very unusual" for a requesting agency and directed by Attorney General Sessions himself. Tr. 1055:5-9 (Abowd); Gore Dep. 271:21-272:13. That DOJ would choose not to pursue an offered alternative that the Census Bureau believed would better suit DOJ's stated needs is extraordinary-and all the more so given that the Attorney General himself made the decision. This evidence strongly suggests that VRA enforcement was not the true goal of either the DOJ officials who prepared the request or Secretary Ross, who solicited the request in the first place.

Extra-record evidence also confirms that Defendants failed to consider *1045several important factors when deciding whether to include the citizenship question on the census. First, Defendants failed to consider whether the question had been adequately tested under the applicable agency standards. While the Decision Memo states that the citizenship question is "well tested" because it has been on the ACS since 2005, PTX-26 at 2, Defendants point to no evidence in the Administrative Record demonstrating that the citizenship question was "performing adequately" on the ACS for the purposes of pretesting. See Part IV.D.2.a. Defendants also failed to consider the effect of documented changes in the macro-environment around issues of immigration on respondents' willingness to fill out a questionnaire that includes a citizenship question.

Second, Defendants failed to consider the effect of the Census Bureau's confidentiality obligations and disclosure avoidance practices on the fitness of citizenship data for DOJ's stated purpose, enforcement of Section 2 of the VRA. Under its disclosure avoidance protocols, the Census Bureau will apply disclosure avoidance techniques to data collected from every census block, meaning that even after adding a citizenship question, there will not be a single census block for which the reported citizenship data directly reflects the responses of the census block's inhabitants to the 2020 Census questionnaire, unless by random chance. New York Tr. 1033:16-21 (Abowd); Census Bureau 30(b)(6) Dep. Vol. I 53:12-17, 69:6-71:12. Therefore, even if the citizenship data is obtained through enumeration, some margin of error will unavoidably exist. Id. The Census Bureau does not know how that margin of error will compare to the margin of error for the ACS citizenship data currently in use. Id. The Bureau has not determined if, after disclosure avoidance, the error margins for block-level CVAP data based on information collected through the decennial enumeration will allow redistricting offices and the Department of Justice to use the data effectively. Census 30(b)(6) Dep. Vol. I 100:21-101:15.

This complete absence of certainty, confirmed by Dr. Abowd, of whether the Census Bureau's disclosure avoidance protocols will result in greater error margins than the ones associated with currently available CVAP data, or will generate a product that is "effective" for its intended use, directly controverts the Secretary's argument that "hard-count" citizenship data from the decennial enumeration will provide DOJ with the "most complete and accurate" data for its VRA enforcement efforts. PTX-26 at 5.

Third, extra-record evidence reveals that Defendants failed to analyze and consider whether hard-count citizenship data would aid in VRA enforcement. The unrebutted expert testimony of Plaintiffs' voting rights expert Dr. Lisa Handley confirms that a citizenship question on the census is not necessary to enforce the VRA. Dr. Handley testified that, in her experience, CVAP estimates at the census tract or block group level are generally sufficient to satisfy the first Gingles precondition in VRA cases. New York Tr. 807:24-811:6 (Handley). Dr. Handley further explained that, where it would be helpful to present CVAP data at the block level, this information can be reliably and accurately estimated using block-level CVAP data by applying the CVAP ratios from the census tract level to the block-level figures for total voting-age population. Id. at 808:10-815:5.

The unrebutted expert testimony of Plaintiffs' voting rights expert Professor Pamela Karlan also confirms that a citizenship question on the census is not "necessary" to enforce the VRA. Professor Karlan explained that no Section 2 case has *1046ever failed on account of the purported inadequacy of ACS data (or, prior to the advent of the ACS, data from the long-form census questionnaire) as a measure of CVAP. Karlan Trial Dep. 52:14-53:18.

Defendants point to no evidence in the Administrative Record that suggests a contrary conclusion. Moreover, there is no indication that Secretary Ross independently weighed the value of the requested block level CVAP data, even after DOJ refused to meet with the Census Bureau to clarify its data needs. Ultimately, Secretary Ross's failure to evaluate for himself the value of the requested data renders his decision to prioritize the acquisition of this information over all else arbitrary and capricious.

In short, Defendants' failure to consider these additional factors further supports a finding that the Secretary' decision was arbitrary and capricious.

V. CONSTITUTIONAL CLAIMS

As previously explained, both sets of plaintiffs advance a claim for violation of the Enumeration Clause. The San Jose Plaintiffs also advance a claim under the Appropriations Clause; however, as discussed in Part III.C, supra , they do not adequately establish standing with respect to that claim. Accordingly, only the Enumeration Clause claim will be discussed here.

The United States Constitution mandates the "actual Enumeration" of the population every ten years for the purpose of apportioning congressional representatives among the states. U.S. Const. art. I, § 2, cl. 3. Congress has delegated the duty of taking the census to the Secretary of Commerce. 13 U.S.C. § 141(a). Although Congress and the states use census data for many purposes, including for allocating federal funding, the only constitutional purpose of the census is to apportion congressional representatives based on the "actual Enumeration" of the population of each state. U.S. Const. art. I, § 2, cl. 3 ; id. amend. XIV, § 2; see also Franklin v. Massachusetts , 505 U.S. 788, 806, 112 S.Ct. 2767, 120 L.Ed.2d 636 (1992) (reasoning that Secretary of Commerce's decision to include overseas federal employees in the apportionment count did not violate Enumeration Clause because the decision "does not hamper the underlying constitutional goal of equal representation"); Utah v. Evans , 536 U.S. 452, 500, 122 S.Ct. 2191, 153 L.Ed.2d 453 (2002) (Thomas, J., concurring in part and dissenting in part) (observing that the Framers' "[d]ebate about apportionment and the census...focused for the most part on creating a standard that would limit political chicanery").

The Census Bureau is not constitutionally required to perform an absolutely accurate count of the population. Wisconsin v. Cty. of New York , 517 U.S. 1, 6, 116 S.Ct. 1091, 134 L.Ed.2d 167 (1996). Nevertheless, there is a "strong constitutional interest in accuracy" of the census. Evans , 536 U.S. at 478, 122 S.Ct. 2191. The type of accuracy which most directly implicates the constitutional purpose of the census is distributive accuracy, as opposed to numerical accuracy. See Wisconsin , 517 U.S. at 20, 116 S.Ct. 1091 (explaining that "a preference for distributive accuracy...would seem to follow from the constitutional purpose of the census, viz., to determine the apportionment of the Representatives among the States"). Numerical accuracy refers to the accuracy of the overall count, whereas distributive accuracy refers to the accuracy of the proportions in which residents are counted in their proper locations. See id. at 11, 116 S.Ct. 1091 n.6.

*1047To promote distributive accuracy, the Enumeration Clause requires the Secretary's actions to bear "a reasonable relationship to the accomplishment of an actual enumeration of the population, keeping in mind the constitutional purpose of the census," which is to determine the apportionment of the Representatives among the States. Id. at 20, 116 S.Ct. 1091. While each and every question on the census need not be related to the goal of actual enumeration, a decision to alter the census in a way that affirmatively interferes with the actual enumeration, and does not fulfill any other reasonable governmental purpose, is subject to a challenge under the Enumeration Clause.

Defendants argue that the Enumeration Clause claims must be decided on the basis of the Administrative Record alone. See, e.g. , Bellion Spirits, LLC v. United States , 335 F.Supp.3d 32, 43 (D.D.C. 2018) ("[W]hen a constitutional challenge to agency action requires evaluating the substance of an agency's decision made on an administrative record, that challenge must be judged on the record before the agency"); Chiayu Chang v. U.S. Citizenship & Immigration Servs. , 254 F.Supp.3d 160, 161 (D.D.C. 2017) (summarizing cases holding that the assertion of constitutional claims does not remove a matter from the APA's procedural strictures); Jarita Mesa Livestock Grazing Ass'n v. U.S. Forest Serv. , 58 F.Supp.3d 1191, 1237 (D.N.M. 2014) (holding that the fact that a case "alleges constitutional violations as well as statutory ones does not take it outside of the APA").

As the district court in Bellion recognized, "caselaw on a plaintiff's ability to supplement an administrative record to support a constitutional cause of action is sparse and in some tension." 335 F.Supp.3d at 41. Some courts have held that such claims are subject to the strictures of the APA because section 706 specifically provides for review of agency action that is "contrary to constitutional right." See, e.g. , Jarita Mesa , 58 F.Supp.3d at 1237-38 (quoting 5 U.S.C. § 706(2)(B) ); see also Chiayu Chang , 254 F.Supp.3d at 161-162 (collecting cases). Other courts have reached a similar conclusion based on more practical concerns. See, e.g. , Alabama-Tombigbee Rivers Coal. v. Norton , No. CV-01-S-0194-S, 2002 WL 227032, at *3-6 (N.D. Ala. Jan. 29, 2002). Yet another group of courts has concluded that extra-record evidence may sometimes be considered in resolving such claims. See, e.g. , Puerto Rico Pub. Hous. Admin. v. U.S. Dep't of Hous. & Urban Dev. , 59 F.Supp.2d 310, 327-28 (D.P.R. 1999) ; Rydeen v. Quigg , 748 F.Supp. 900, 906 (D.D.C. 1990) ; see also Carlsson v. U.S. Citizenship & Immigration Servs. , No. 12-cv-07893, 2015 WL 1467174, at *13 (C.D. Cal. Mar. 23 2015).

Fortunately, there is no need to wade into the morass. Even assuming the usual APA strictures apply, Secretary Ross's failure to consider the impact of the citizenship question on the accuracy of the enumeration, and therefore on the apportionment of congressional representation, see Part E(1)(a)(ii), supra , empowers this Court to look beyond the Administrative Record to determine whether the Secretary's decision to include the citizenship question is consistent with his Constitutional obligation to carry out an accurate enumeration of the public. See Ranchers Cattlemen , 499 F.3d 1108 at 1117. To find otherwise would allow government agencies to insulate themselves from judicial review of the constitutionality of their actions by simply refusing to investigate pertinent factors during the decision-making process. Accordingly, the Findings of Fact Related to Standing set forth in Part *1048III.B, supra , form the basis of the Enumeration Clause claim, in addition to the findings of fact associated with the APA claim laid out in Part IV.C-D, supra .

C. Conclusions of Law

The Secretary's decision to add a citizenship question to the 2020 Census violates the Enumeration Clause of the Constitution because its inclusion will materially harm the accuracy of the census without advancing any legitimate governmental interest. This is no ordinary demographic inquiry. The record reveals that the inclusion of the citizenship question on the upcoming census will have a unique impact on the Census Bureau's ability to count the public, to the point where the inclusion of this question is akin to a mechanics-of-counting-type issue. In short, Secretary Ross's decision to add the citizenship question to the 2020 Census undermines the "strong constitutional interest in [the] accuracy" of the census, Utah , 536 U.S. at 478, 122 S.Ct. 2191, and does so despite the fact that adding this question does not advance any identifiable government purpose.

First, as previously discussed, the citizenship question will significantly impair the distributive accuracy of the census because it will uniquely and substantially impact specific demographic groups. Part III.B.1-3, supra . Specifically, the citizenship question will cause an undercount of noncitizens and Latinos and, by extension, localities where many such persons reside. Part III.B.3, supra . This, in turn, substantially increases the risk that California will lose a seat in the House of Representatives. Indeed, the likelihood that California will lose a congressional seat increases from 26.1 percent to at least 49.9 percent with the addition of the citizenship question. Parts III.B.6, III.C.1.b, supra . There is also credible evidence, based on Dr. Barreto's survey, that California could lose up to three congressional seats due to the citizenship question. Part III.B.6 Accordingly, the addition of the citizenship question implicates the "strong constitutional interest in accuracy." Evans , 536 U.S. at 478, 122 S.Ct. 2191 ; see also Wisconsin , 517 U.S. at 20, 116 S.Ct. 1091.

Second, Defendants fail to identify any countervailing governmental interest that could justify this harm to the census. It is clear from the Administrative Record that Secretary Ross's purported reliance on the DOJ letter was nothing more than a pretext designed to provide cover for Secretary Ross's unexplained desire to add the citizenship question to the census. See Part IV.C.12, supra . Extra-record evidence bolsters this conclusion. Evidence produced through discovery reveals that the DOJ letter was drafted at the Attorney General's direction, with minimal involvement from the Civil Rights Division's voting rights staff. Attorney General Sessions subsequently prevented DOJ officials from engaging in follow-up discussions with Census Bureau staff to clarify the request. Moreover, aside from DOJ's December 12 Letter, which was produced under curious circumstances and represented an abrupt change in DOJ's position, the record is devoid of any evidence to support the VRA enforcement rationale. Plaintiffs, on the other hand, adduce persuasive evidence that block-level CVAP data would have little effect on the prosecution of VRA actions.

Even assuming DOJ's VRA enforcement efforts would benefit from block-level citizenship data, there is no evidence to support the conclusion that surveying residents about their citizenship would increase the accuracy of available block-level citizenship data. Indeed, all evidence before this Court points to the opposite conclusion-that the inclusion of the citizenship *1049question would ultimately yield less citizenship accurate data than relying on administrative records and imputing the citizenship status of persons for whom such data does not exist. Part IV.C.5.d. In sum, the evidence shows that this question would impede the stated goal of obtaining more accurate block-level CVAP. Accordingly, Defendants have failed to point to any way in which including the citizenship question on the 2020 census will advance a legitimate governmental interest.

This is not to say that the Census Bureau may never ask about citizenship on future census questionnaires. It simply means that, where the inclusion of a particular question will degrade the accuracy of the Census to the point where the proper apportionment of representatives among the states is at risk, the government must identify a legitimate governmental purpose that is sufficiently weighty to justify this significant harm to the census. Defendants falls well short of this standard.

Moreover, finding that adding a citizenship question is unconstitutional does not automatically render all demographic questions on the census unconstitutional. Indeed, it is well established that each and every question on the census need not relate to the goal of enumeration. See Baldrige v. Shapiro , 455 U.S. 345, 353, 102 S.Ct. 1103, 71 L.Ed.2d 199 (1982) (acknowledging that the census "fulfills many important and valuable functions"). There is, however, no evidence that any other demographic question included on the census is likely to result in distributive inaccuracy that threatens to distort the apportionment of representatives, or that these questions were introduced despite the absence of any legitimate governmental interest in their being asked.

Defendants argue against an Enumeration Clause finding because it would suggest the constitutionality of a particular question could vary over time depending upon the social and political context. This argument is unpersuasive. As the Supreme Court has recognized, the constitutionality of a particular governmental action may depend on the larger social context in which that action occurs. See Shelby Cty., Ala. v. Holder , 570 U.S. 529, 553-557, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013) (striking down Section 4 of the Voting Rights Act because, inter alia , the preclearance formula set forth in that section was based on "decades-old data relevant to decades-old problems, rather than current data reflecting current needs"); Grutter v. Bollinger , 539 U.S. 306, 342-43, 123 S.Ct. 2325, 156 L.Ed.2d 304 (2003) (explaining that "race-conscious admissions policies must be limited in time" and predicting that 25 years later affirmative action would no longer be necessary or constitutionally justified). Ultimately, Secretary Ross's primary obligation under the Constitution is to ensure a reasonably accurate enumeration of the public, and to attempt to design a survey that will achieve that goal in the year 2020. The fact that the citizenship question may have been perfectly harmless in 1950, or that may be harmless again in the year 2050 is of little consequence to the Secretary's constitutional obligations with respect to the accuracy of the 2020 Census.

The addition of a citizenship question to the 2020 Census cannot be said to bear a "reasonable relationship to the accomplishment of an actual enumeration of the population." Wisconsin , 517 U.S. at 20, 116 S.Ct. 1091. Secretary Ross's decision thus violates the Enumeration Clause of the Constitution.

VI. REMEDIES

A. Vacatur and Remand

In light of this Court's ruling that Secretary Ross's decision to include the citizenship *1050question on the 2020 Census violated the APA, the Secretary's decision is hereby vacated and remanded to the Department of Commerce. 5 U.S.C. § 706(2) ("[A] reviewing court shall...hold unlawful and set aside agency action found to be" in violation of the APA).

B. Injunctive Relief ²⁰

A plaintiff seeking a permanent injunction must satisfy four requirements: "(1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction." eBay Inc. v. MercExchange, L.L.C. , 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006). The decision whether to grant injunctive relief ultimately falls within a court's equitable discretion. Id.

Here, each of the four elements is satisfied. It is simply beyond dispute that a loss of political representation to the state of California, among other harms identified by Plaintiffs, qualifies as irreparable harm that cannot be remedied at law. Moreover, given the likely harm to the quality of the census and the absence of any valid basis for Secretary Ross's decision, the balance of the hardships and the public interest both favor barring Defendants from including the citizenship question in the census questionnaire. Accordingly, an injunction is appropriate here. The proper scope of the injunction, however, requires additional analysis.

As previously noted, this Court is mindful of concerns regarding the authority and propriety of an individual district judge in one judicial district issuing an injunction of nationwide application. In light of the unitary nature of the question at issue, by definition no injunctive relief could be limited to only one geographic area or to only certain litigants. Put simply, the citizenship question is either on or off the 2020 Census, thereby practically requiring an injunction nationwide in scope.

The appropriate terms of this injunction depend on which claim forms the basis for relief. To the extent that the injunction relates solely to the APA claim, the reasoning in New York et al. v. Department of Commerce, et al. , No. 18-cv-2921 (S.D.N.Y.) is persuasive. As explained in that order, "an injunction is necessary to make the Court's vacatur effective, as it prevents Secretary Ross from arriving at the same decision without curing the problems identified in this Opinion." Id. at 272. Accordingly, with respect to the APA violation, Defendants are enjoined from including the citizenship question on the 2020 Census unless: (1) they establish that direct inquiries regarding citizenship are necessary given the "kind, timeliness, quality and scope of the statistics required" and that administrative records will not suffice, 13 U.S.C. § 6(c) ; (2) they identify new circumstances that necessitate the last minute addition of the citizenship question to the census, id. § 141(f)(3); and (3) Secretary Ross considers all relevant factors and evidence, and sets forth the actual basis for his decision.

The Enumeration Clause violation, however, requires a more expansive injunction. The record in this case has clearly established *1051that including the citizenship question on the 2020 Census is fundamentally counterproductive to the goal of obtaining accurate citizenship data about the public. This question is, however, quite effective at depressing self-response rates among immigrants and noncitizens, and poses a significant risk of distorting the apportionment of congressional representation among the states. In short, the inclusion of the citizenship question on the 2020 Census threatens the very foundation of our democratic system-and does so based on a self-defeating rationale. In light of these findings, Defendants do not get another bite at the apple. Defendants are hereby enjoined from including the citizenship question on the 2020 Census, regardless of any technical compliance with the APA.

IT IS SO ORDERED .

1 The Administrative Record comprises all documents identified in the parties' stipulation, specifically: AR 1 through AR 13024 (PTX-001 through PTX-014); PTX-016 through PTX-152; PTX-154; PTX-156; PTX-157; PTX-164 through PTX-170; PTX-172 through PTX-182; PTX-184; PTX-185; PTX-227; PTX-228; PTX-244; PTX-264; PTX-274; PTX-362; PTX-363; PTX-370; PTX-374 through PTX-390; PTX-397; PTX-399 through PTX-433; PTX-435 through PTX-452. Joint Pretrial Statement 11-12.

2 There is significant overlap in the arguments and evidence presented by the San Jose Plaintiffs and the California Plaintiffs. Accordingly, certain portions of this order will treat Plaintiffs arguments jointly.

3 The parties have stipulated to over one hundred undisputed facts, which can be found under Exhibit A to the Joint Pretrial Statement (ECF 144 in Case No. 18-cv-01865).

4 Several entities and individuals have filed motions for leave to file amicus briefs in this matter. Although these amicus briefs do not form the basis of this opinion, each of these motions is hereby granted.

5 Any conclusion of law that is mistakenly characterized as a finding of fact in this order may be recharacterized as a conclusion of law, and vice versa.

6 The parties have stipulated to the admission of Dr. Abowd's trial testimony in the New York matter, subject to Defendants' standing objection to the consideration of evidence outside the Administrative Record in adjudicating the merits of this action. Stip. & Order re Abowd Trial Transcript (ECF 172 in Case No. 18-cv-01865).

7 Removing California from the national self-response estimates reveals just how severe the differential decline in self-response rates are likely to be in California relative the rest of the country. Dr. Barreto estimates that the decline in self-response rates in the rest of the country, excluding California, is likely to be 6.5 percent, as compared to an estimated decline of 12.3 percent in California. PTX-871.

8 Question 2 of Dr. Barreto's survey reads: "In 2020, the federal government is adding a new question to require you to list whether you, and every person in your household is a U.S. citizen, or not a citizen. With the addition of a citizenship question, will you participate and submit your household information, or not?" Tr. 441:4-11 (Barreto).

9 Question 1 read: "The Census is an official population count that is conducted every 10 years by the federal government. It requires all households to list the name, age, and race or ethnicity of every person living in the home and provide that information to the Census Bureau either online, by mail, or in-person with a census taker. The Census is required to keep this information confidential, and every single household in the country is required to participate. In March 2020 you will receive an invitation from the U.S. Census to fill out the census form. Do you plan to participate and submit your household information?" Tr. 440:13-441:2.

10 The 18 state-share programs include: Federal Transit Formula Grants, Community Block Development Grants/Entitlement Grants, Crime Victim Assistance, Title I Grants to Local Educational Authorities (LEAs), Special Education Grants, Head Start, Supplemental Nutrition Program for Women, Infants, and Children (WIC), Child Care and Development Block Grants, Supporting Effective Instruction State Grants, Workforce Innovation and Opportunity Act (WIOA) Youth Activities, Rehabilitation Services: Vocational Rehabilitation Grants to the States, Unemployment Insurance administrative costs, Block Grants for Prevention and Treatment of Substance Abuse, Social Services Block Grants, Career and Technical Education-Basic Grants to States, WIOA Disclosed Worker Formula Grants, Special Programs for the Aging, Title III, Part C, Nutrition Services. Reamer Decl. ¶¶ 11, 17, Ex. D (PTX-245).

11 A summary of judicial noticed facts pertaining to the San Jose Plaintiffs can be found at ECF 180 in Case No. 18-cv-02279. The order granting their request for judicial notice may be found in ECF 176 of that same case.

12 Defendants argue persuasively that the "NRFU simulation" portion of Dr. Barreto's survey does not resemble the Census Bureau's actual NRFU process. Accordingly, the Court declines to rely on Scenario B, which was based in part on this NRFU simulation.

13 As previously explained, the San Jose Plaintiffs also advance a claim for relief under the Apportionment Clause. They do not, however, adequately establish standing with respect to the Apportionment Clause Claim as they have not shown how, even were California to lose congressional representation, that would impact San Jose in particular. This finding is of little practical consequence, however, given that the San Jose Plaintiffs successfully establish standing with respect to the Enumeration Clause Claim.

14 BAJI also advances two separate theories of associational standing. Hunt v. Washington State Apple Advertising Com'n , 432 U.S. 333, 343, 97 S.Ct. 2434, 53 L.Ed.2d 383 (1977) (setting for the requirements for associational standing). There is, however, no need to reach these theories in light of the several bases for standing for the San Jose Plaintiffs that have already been discussed.

15 The numbering of the findings of fact picks up where it left off in Part III.B, supra .

16 Judicial notice is taken of the order in New York et al. v. Department of Commerce, et al. , No. 18-cv-2921 (S.D.N.Y.) requiring Defendants to supplement the record.

17 At the time, Bannon was the White House Chief Strategist.

18 Where only one ECF docket number is listed, the docket refers to Case No. 18-cv-01865.

19 Judicial notice is taken of Secretary Ross's 141(f)(2) report, Questions Planned for the 2020 Census and American Community Survey, available at https://www.census.gov/library/publications/2018/dec/planned-questions-2020-acs.html.

20 Plaintiffs also request declaratory relief. In light of the vacatur of Secretary Ross's decision and the injunction barring Defendants from including the citizenship question on the 2020 Census, providing declaratory relief would serve little useful purpose and is therefore denied.