Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 1 of 62 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO LATHAM & WATKINS LLP Steven M. Bauer (Bar No. 135067) steven.bauer@lw.com Sadik Huseny (Bar No. 224659) sadik.huseny@lw.com Shannon D. Lankenau (Bar No. 294263) shannon.lankenau@lw.com 505 Montgomery Street, Suite 2000 San Francisco, CA 94111 Telephone: 415.391.0600 Facsimile: 415.395.8095 LAWYERS’ COMMITTEE FOR CIVIL RIGHTS UNDER LAW Kristen Clarke (pro hac vice pending) kclarke@lawyerscommittee.org Jon M. Greenbaum (Bar No. 166733) jgreenbaum@lawyerscommittee.org Ezra D. Rosenberg (pro hac vice pending) erosenberg@lawyerscommittee.org Dorian L. Spence (pro hac vice pending) dspence@lawyerscommittee.org Ajay P. Saini (pro hac vice pending) asaini@lawyerscommittee.org Maryum Jordan (Bar No. 325447) mjordan@lawyerscommittee.org Pooja Chaudhuri (Bar No. 314847) pchaudhuri@lawyerscommittee.org 1500 K Street NW, Suite 900 Washington, D.C. 20005 Telephone: 202.662.8600 Facsimile: 202.783.0857 LATHAM & WATKINS LLP Richard P. Bress (pro hac vice pending) rick.bress@lw.com Melissa Arbus Sherry (pro hac vice pending) melissa.sherry@lw.com Anne W. Robinson (pro hac vice pending) anne.robinson@lw.com Tyce R. Walters (pro hac vice pending) tyce.walters@lw.com Genevieve P. Hoffman (pro hac vice pending) genevieve.hoffman@lw.com Gemma Donofrio (pro hac vice pending) gemma.donofrio@lw.com 555 Eleventh Street NW, Suite 1000 Washington, D.C. 20004 Telephone: 202.637.2200 Facsimile: 202.637.2201 Additional counsel and representation information listed in signature block UNITED STATES DISTRICT COURT FOR THE NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION NATIONAL URBAN LEAGUE; LEAGUE OF WOMEN VOTERS; BLACK ALLIANCE FOR JUST IMMIGRATION; HARRIS COUNTY, TEXAS; KING COUNTY, WASHINGTON; CITY OF LOS ANGELES, CALIFORNIA; CITY OF SALINAS, CALIFORNIA; CITY OF SAN JOSE, CALIFORNIA; RODNEY ELLIS; and ADRIAN GARCIA, Plaintiffs, v. CASE NO. 20-cv-5799 COMPLAINT FOR DECLARATORY AND INJUNCTIVE RELIEF WILBUR L. ROSS, JR., in his official capacity as Secretary of Commerce; U.S. DEPARTMENT OF COMMERCE; STEVEN DILLINGHAM, in his official capacity as Director of the U.S. Census Bureau; and U.S. CENSUS BUREAU, Defendants. 1 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 2 of 62 INTRODUCTION 1 2 1. This lawsuit challenges the unconstitutional and illegal decision by Secretary of 3 Commerce Wilbur Ross, and Census Bureau (the “Bureau”) Director Steven Dillingham, to 4 sacrifice the accuracy of the 2020 Census by forcing the Census Bureau to compress eight and a 5 half months of vital data-collection and data-processing into four and a half months, against the 6 judgment of the Bureau’s staff and in the midst of a once-in-a-century pandemic. 7 2. The Census Bureau’s staff spent most of the past decade developing a final 8 operational plan for the 2020 Census that reflected the Bureau’s understanding of the best 9 methods for counting everyone once and in the right place (the “Final Operational Plan”). In 10 April 2020, as the COVID-19 pandemic spread throughout the country, the Census Bureau 11 revised its plan to account for both the difficulties of census-taking during a pandemic and the 12 Bureau’s constitutional and statutory obligation to achieve a fair and accurate count (the 13 “COVID-19 Plan”). To achieve both ends, the Department of Commerce and the Census Bureau 14 delayed the counting process, shifted the timeframe for conducting and completing its data- 15 collection operation, and increased the time for conducting data-processing, while, crucially, 16 preserving the same amount of time for each step of those operations. 17 3. On August 3, 2020, the Department of Commerce and the Census Bureau 18 suddenly and without explanation reversed course and replaced the Bureau’s COVID-19 Plan 19 with a new one (the “Rush Plan”). The Bureau’s Rush Plan requires the Bureau to complete eight 20 and a half months of data-collection and data-processing in half the time. It ignores the multi- 21 month delay in census data-collection that the COVID-19 pandemic caused. It compels a final 22 date for delivering apportionment data to the President that Bureau officials have repeatedly 23 asserted they cannot meet. And it threatens a massive undercount of the country’s communities 24 of color and the municipalities, cities, counties, and states where they live. Under these 25 circumstances, the Bureau’s new plan to rush the 2020 Census violates, among other things, the 26 federal government’s legal obligations to secure an accurate count and statutory prohibitions on 27 arbitrary, capricious, and pretextual federal government action. 28 ATTORNEYS AT LAW SAN FRANCISCO 4. The federal government’s attempt to rush the census count poses a grave threat to 1 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 3 of 62 1 all the vital functions that rely on census data, from reapportioning the United States House of 2 Representatives and redrawing state and local electoral districts, to equitably distributing over 3 $1.5 trillion annually in federal funds that support basic needs such as food, health care, and 4 education. Undercounted cities, counties, and municipalities will lose representation in Congress 5 and tens of millions of dollars in funding. And communities of color will lose core political 6 power and vital services. In contrast to these dire stakes, the immediate solution to this problem 7 is simple: set aside and enjoin implementation of the impossibly-shortened Rush Plan, which is 8 based on an unexplained change of position, and allow the Census Bureau to implement the plan 9 that it had designed to fulfill its constitutional duties during the pandemic. 10 5. The COVID-19 pandemic upended all 2020 Census field operations, many of 11 which the Census Bureau designed to enumerate populations that it has long struggled to count, 12 including racial and ethnic minorities, non-English speakers, and undocumented persons. Among 13 the disrupted census operations was the largest, most time-consuming operation undertaken to 14 count the country’s hard-to-count communities—the “Non-Response Follow Up” operation. 15 During Non-Response Follow Up, the Bureau sends its employees to knock on the doors of 16 households that have not yet responded to the census and perform other vital data-collecting 17 functions. 18 6. The Bureau’s staff responded to the pandemic—and the impossibility of 19 conducting house visits during widespread lockdowns—by making necessary adjustments to the 20 timeline in the Final Operational Plan. This revised operational plan, the COVID-19 Plan issued 21 on April 13, 2020, was intended to ensure that hard-to-count communities would be enumerated 22 and the health and safety of Bureau employees and the public would be protected. This plan 23 adjusted the deadlines of, but did not shorten the time for, critical operations. Under this plan— 24 which experts and census stakeholders alike endorsed as a scientifically sound approach for 25 minimizing the pandemic’s potential damage to the accuracy of the count—the Bureau extended 26 its data-collection deadlines to October 31, 2020 and its data-processing deadlines into the 27 second quarter of 2021. Critically, the COVID-19 Plan delayed door-knocking by three months, 28 pushing it from May–July 2020 to August–October 2020. But the COVID-19 Plan ATTORNEYS AT LAW SAN FRANCISCO 2 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 4 of 62 1 acknowledged that the Bureau must spend the same amount of time—around eleven and a half 2 weeks—on door-knocking, just as it had planned to do before the pandemic. The COVID-19 3 Plan also incorporated the same methods and techniques contemplated in the Final Operational 4 Plan that the Bureau had spent years developing. Indeed, the only respect in which the COVID- 5 19 Plan altered the amount of time devoted to operations set out in the Final Operational Plan 6 was a requirement that the Bureau spend more time than originally planned processing the data it 7 collected—that is, performing the necessary work to transform over 100 million individual 8 census forms into high-quality, reliable, and legitimate data. This additional investment in data- 9 processing reflected daunting new challenges the COVID-19 pandemic posed to an accurate 10 count, including massive displacements of people that would introduce problems of duplicate 11 responses, responses without unique census identifiers, and other complex data issues. 12 7. The Department of Commerce and the Census Bureau also recognized that the 13 impact of COVID-19 had made it impossible to meet certain statutory deadlines for reporting 14 census results to Congress. Commerce Secretary Wilbur Ross and Census Bureau Director 15 Steven Dillingham announced that the Bureau was seeking relief from Congress to formally 16 extend two statutory deadlines: first, the deadline for reporting the state-population totals used to 17 calculate the congressional apportionment to the President, which Congress was asked to extend 18 from December 31, 2020 to April 30, 2021; and, second, the deadline for reporting redistricting 19 data to the states, which Congress was asked to extend from March 31, 2021, to July 31, 2021. 20 Commenting on the statutory-deadline extensions, President Trump publicly stated on April 13, 21 2020, “I don’t know that you even have to ask [Congress]. This is called an act of God. This is 22 called a situation that has to be. They have to give in. I think 120 days isn’t nearly enough.” 23 Hansi Lo Wang, Trump Officials Ask to Delay Census Data for Voting Districts, House Seats, 24 NPR (Apr. 13, 2020), https://www.npr.org/2020/04/13/833546675/trump-officials-ask-to-delay- 25 census-data-for-voting-districts-house-seats. 26 8. Recognizing that more time was necessary to complete an accurate census, and 27 consistent with the President’s statement, the Bureau proceeded immediately under its COVID- 28 19 Plan. The Bureau delayed its door-knocking operation to late summer, with the declared ATTORNEYS AT LAW SAN FRANCISCO 3 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 5 of 62 1 intention of completing it by October 31, 2020. And recognizing that a successful census is 2 dependent on all levels of government working together, the Bureau publicized this plan to the 3 public, as well as to government and non-profit partners involved in the years-long and multi- 4 million-dollar public education campaign to ensure public trust and encourage public 5 participation in the census. 6 9. Throughout the summer, Bureau officials repeatedly stated that the pandemic had 7 rendered it impossible for the Bureau to complete a reasonably accurate count by December 31, 8 2020. But to comply with its constitutional obligations, the Bureau continued collecting data on 9 the timelines set in the COVID-19 Plan, which itself extended the Bureau’s data-processing 10 11 timelines into 2021. 10. On August 3, 2020—in the face of a pandemic that has only grown worse and in 12 disregard of the Census Bureau’s constitutional and statutory duties to conduct an actual 13 enumeration of the entire population—Secretary Ross and Director Dillingham abruptly 14 abandoned the COVID-19 Plan. Without explanation, they announced the new Rush Plan for the 15 2020 Census, including shortening the Bureau’s data-collection operation by one month to 16 September 30, 2020, and requiring the Bureau to process and report the apportionment data to 17 President Trump by December 31, 2020. The Rush Plan cuts a crucial four weeks from the data- 18 collection operation. And it disregards the Bureau’s own prior conclusions that such rushed 19 processing renders it impossible to fulfil its constitutional obligation to ensure reasonable quality 20 and accuracy of 2020 Census data. 21 11. Defendants’ decision to abandon the COVID-19 Plan in favor of the Rush Plan 22 does not satisfy the Supreme Court’s clear command that any decision relating to the census bear 23 a “reasonable relationship” to producing an accurate count. See Wisconsin v. City of N.Y., 517 24 U.S. 1, 20 (1996). As demonstrated by Defendants' own prior statements, the challenged decision 25 cannot be justified by any legitimate interest in conducting an accurate census, and in fact will 26 introduce several inaccuracies in the count, chief among them major undercounts of communities 27 of color. 28 ATTORNEYS AT LAW SAN FRANCISCO 12. The reason for this abrupt change of position is not apparent on the face of the 4 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 6 of 62 1 press release announcing the Rush Plan or any other subsequently issued statements or 2 publications from the federal government. The Bureau has refused requests from Congress and at 3 least one Plaintiff in this action to provide one. 4 13. The announcement of the Rush Plan did reference two developments that 5 occurred between the adoption of the COVID-19 Plan and the announcement of the Bureau’s 6 intent to adopt the Rush Plan. But neither of these developments can justify Defendants' actions. 7 First, the announcement refers to the Secretary of Commerce’s direction to the Bureau to comply 8 with the statutory deadline of December 31, 2020 for completing the apportionment count. But 9 this statutory deadline cannot justify an unconstitutional decision to cut short crucial operations 10 and fail to satisfy its constitutional obligation. A statutory deadline, particularly one that was set 11 without a global pandemic in mind, cannot override the federal government’s constitutional duty 12 to accomplish an accurate census; there is “nothing sacred in the due date of the filing [of 13 apportionment data], especially when the work of the Census Bureau . . . is incomplete.” Carey 14 v. Klutznick, 637 F. 2d 834, 837 (2d Cir. 1980). Moreover, the Bureau was cognizant of this 15 deadline even as it designed and implemented the COVID-19 Plan, including delaying crucial 16 field operations by several months. And Bureau officials have repeatedly made clear that because 17 of the impediments introduced by COVID-19, together with the multi-month delay, it is already 18 too late to satisfy these pre-COVID-19 deadlines. 19 14. Second, both the text of the Rush Plan announcement and the timing of the 20 decision suggest that the federal government’s motivation for the Rush Plan is to facilitate 21 another illegal act: suppressing the political power of communities of color by excluding 22 undocumented people from the final apportionment count. On July 21, 2020—just a few weeks 23 earlier—President Trump issued a Presidential Order titled “Memorandum Excluding Illegal 24 Aliens From the Apportionment Base Following the 2020 Census” (the “Apportionment 25 Exclusion Order”)—which expressly stated the President’s determination to exclude 26 undocumented people from the population count used for apportionment. To increase the chance 27 that the President can fully effectuate the Apportionment Exclusion Order, he must receive the 28 population totals while he is still in office, and he ordered the Secretary of Commerce to provide ATTORNEYS AT LAW SAN FRANCISCO 5 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 7 of 62 1 him with 2020 decennial census information by December 31, 2020 to carry out his objective. 15. 2 The President’s Apportionment Exclusion Order (currently being challenged as 3 unconstitutional and unlawful in a number of lawsuits filed in jurisdictions around the country, 4 including in this District) represents only the most recent of Defendants' serial attempts to 5 manipulate the 2020 Census to suppress the political power of communities of color. These 6 attempts started with a campaign to introduce a historically unprecedented and untested 7 citizenship question onto the 2020 Census questionnaire to advantage—in the words of a 8 deceased Republican redistricting consultant—“Republicans and non-Hispanic whites.” Michael 9 Wines, Deceased G.O.P. Strategist’s Hard Drives Reveal New Details on the Census Citizenship 10 Question, N.Y. Times (May 30, 2019), https://www.nytimes.com/2019/05/30/us/census- 11 citizenship-question-hofeller.html. Since the Supreme Court blocked the question, Defendants 12 have looked for other means to achieve that same end, including collecting data on citizenship 13 from administrative records and, now, cutting the census short. 16. 14 Plaintiffs are local governments, civil rights and civic organizations, and 15 individuals whose communities will almost certainly be inaccurately represented and 16 underrepresented in the final census count if the administration succeeds in truncating census 17 data-collection and data-processing. 17. 18 Plaintiffs seek declaratory relief affirming that Defendants’ actions violate the 19 Enumeration Clause and the Administrative Procedure Act. Plaintiffs additionally seek to set 20 aside and enjoin implementation of the illegal Rush Plan, thereby permitting the Bureau to 21 implement the preexisting COVID-19 Plan it carefully designed to ensure a complete and 22 accurate count. This relief will allow the Bureau to conduct the 2020 Census on the timeline it 23 has repeatedly asserted is necessary to complete a full, fair, and accurate count. 18. 24 25 irreparable harm for at least another decade, until the next census is conducted. JURISDICTION AND VENUE 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO Without such relief, Plaintiffs and the communities they represent will suffer 19. This Court has subject matter jurisdiction under 28 U.S.C. §§ 1331, 1346(a), and 1361. 6 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 8 of 62 20. 1 Venue is proper in this judicial district under 28 U.S.C. § 1391(b)(2) and (e)(1). 2 Defendants are United States officers or agencies sued in their official capacities, a substantial 3 part of the events or omissions giving rise to this action have occurred or will occur in this 4 district, and one or more Plaintiffs reside in this district. 21. 5 6 and 2202. 22. 7 8 This Court may grant declaratory and injunctive relief under 28 U.S.C. §§ 2201 The proper intradistrict assignment for this action is the San Jose Division, in light of the location of Plaintiffs City of San Jose and members of the League of Women Voters. PARTIES 9 10 I. 11 Plaintiffs 23. The National Urban League (“Urban League”) is a civil-rights organization with 12 over 90 affiliates serving 300 communities in 37 states and the District of Columbia. Founded in 13 1910, the National Urban League is headquartered in New York City. The mission of the 14 National Urban League is to help African Americans and others in underserved communities 15 achieve their highest human potential and secure economic self-reliance, parity, power, and civil 16 rights. 17 24. For the 2020 Census, the Urban League has expended substantial resources 18 developing programs designed to encourage self-response and cooperation with Census Bureau 19 offices in historically undercounted communities. Specifically, the organization has engaged in 20 efforts to educate the public about the census through various methods, including virtual town 21 halls, production and distribution of toolkits, workshops for locally based get-out-the-count 22 organizations, and publication and upkeep of a website, www.MakeBlackCount.org, to 23 disseminate critical information about the census. The Urban League has also worked with 24 Census Bureau regional offices to encourage enumerator recruitment, and the organization uses 25 social media to encourage 2020 Census participation. 26 25. Plaintiff Black Alliance for Just Immigration (“BAJI”) is a nonprofit organization 27 organized and existing under the laws of California, with offices and members across the 28 country, including in Oakland, California, Miami, Florida, Atlanta, Georgia, and New York City. ATTORNEYS AT LAW SAN FRANCISCO 7 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 9 of 62 1 BAJI collaborates with African Americans and Black immigrants to organize and advocate for 2 equal and just laws in their communities. BAJI campaigns to advance racial justice and provides 3 partner organizations with varied assistance—particularly on immigration policy—and it spends 4 significant resources educating its partner organizations, individuals, and other constituents 5 through presentations, workshops, publications, technical assistance, and trainings. BAJI is a 6 membership organization, and its members either pay dues or volunteer their time to support the 7 organization. Members also actively participate in BAJI’s self-governance and decision-making 8 at the local level. 9 26. For the 2020 Census, BAJI has worked to ensure non-responsive households in 10 Black and immigrant communities are counted. BAJI has hired additional staff dedicated to 11 engaging local communities on the census, and has engaged in outreach using social media and 12 mailers to bolster self-response. In addition, since the outbreak of the COVID-19 pandemic, 13 BAJI staff regularly participate in webinars and virtual events to provide the public more 14 information about the census, with a specific focus on encouraging participation in Black and 15 immigrant communities. 16 27. The League of Women Voters is a nonprofit civic organization that encourages 17 informed and active participation in government. Founded in 1920, the League of Women Voters 18 is headquartered in Washington, D.C. The League of Women Voters has over 800 state and local 19 affiliates, located in all 50 states and in 764 specific communities, including affiliates with 20 members in San Francisco and Monterey County, California, Detroit, Michigan, Miami, Florida, 21 Philadelphia, Pennsylvania, and New York City. The League of Women Voters seeks to 22 empower voters and defend democracy. The League of Women Voters has over 65,000 members 23 nationwide, and its members either pay dues or volunteer their time to support the organization. 24 28. The League of Women Voters has engaged in significant efforts to ensure 25 historically undercounted communities are enumerated during the 2020 Non-Response Follow 26 Up operation. Prior to the outbreak of COVID-19 in the United States, the League of Women 27 Voters and its affiliates participated in public events across the country aimed at providing 28 information about the census to undercounted communities. Since March of this year, the League ATTORNEYS AT LAW SAN FRANCISCO 8 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 10 of 62 1 of Women Voters has shifted to a digital public-education campaign, encouraging education and 2 participation through social media, email listservs, webinars, and blog posts. Affiliates in 3 Kansas, South Carolina and Maine are also participating in state Complete Count Committees 4 that seek to increase awareness of the 2020 Census, improve participation, and coordinate with 5 Census Bureau officials. 6 29. Harris County, Texas is a political subdivision of the State of Texas. With over 7 4.7 million residents, Harris County is the third largest county in the United States. The county’s 8 population is over 43% Latino, 20% Black, over 7% Asian, and over 28% non-Hispanic White. 9 During the 2010 Census, 65.1% of households in Harris County self-responded to the census. As 10 of August 14, 2020, 58.3% of households in Harris County had self-responded to the 2020 11 Census. This response rate in Harris County was well below the national response rate on that 12 date, 63.6%. 13 30. For the 2020 Census, officials in Harris County engaged in extensive efforts to 14 encourage participation in the County. County officials formed a Complete Count Committee 15 with city officials in Houston that engaged in public education about the census, and built 16 partnerships with local Census Bureau officials to coordinate outreach efforts. In addition, in 17 2019, the County approved a budget of nearly $4 million dollars to conduct outreach during the 18 2020 Census. To that end, the County has contracted with vendors to conduct surveys about the 19 opinions and attitudes of non-responsive populations and develop a digital advertising campaign 20 on Facebook and Instagram to encourage 2020 Census participation. And the County receives 21 substantial federal funding tied to census data. 22 31. King County is a political subdivision of the State of Washington. Over 2.2 23 million people live in King County, making it the most populous county in Washington. As of 24 August 14, 2020, 26.1% of households in King County had not responded to the 2020 Census. 25 The county has large populations of historically undercounted communities. For instance, 26 according to the Department of Housing and Urban Development, King County had nearly 27 12,000 residents experiencing homelessness, the third highest total of any locale in the country. 28 The Seattle metro area, which includes King County, is estimated to have 140,000 ATTORNEYS AT LAW SAN FRANCISCO 9 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 11 of 62 1 2 undocumented immigrant residents. 32. King County worked in partnership with local cities to provide $1.17 million to 3 community-based organizations serving historically undercounted communities. Specifically, 4 King County sought to fund organizations that work with communities that are Limited English 5 Proficient. Through this funding, these organizations produced public education materials related 6 to the 2020 Census, and developed campaigns to get-out-the-count. And King County, too, 7 receives substantial federal funding tied to census data. 8 33. The City of Los Angeles, California is a municipal corporation organized and 9 existing under the laws of the State of California, and is a charter city pursuant to Article XI of 10 the California Constitution. The City is home to roughly 4 million people, and is located in the 11 county recognized by the Census Bureau as the hardest to count in the nation. The city’s 12 population is a large contributor to the County’s hard-to-count status as more than half of the 13 City’s residents live in census tracts that are hard to count. As of August 14, 2020, only 53.8% of 14 the City’s households had responded to the 2020 Census—well below the statewide average of 15 65.1% and even further below the City's own 2010 self-response rate of 68 percent. 16 34. As a result of its hard-to-count status, Los Angeles has engaged in years of 17 planning and devoted significant resources to developing a strategy for an accurate count, 18 tailored to the unique challenges of the City’s population. To fund these efforts, the City has 19 overseen distribution of roughly $2 million dollars to community-based organizations and the 20 investment of almost $1.5 million of both City general fund and grant money in its own efforts. 21 And the City of Los Angeles also receives substantial federal funding tied to census data. 22 35. The City of Salinas, California is a political subdivision of the State of California. 23 Salinas is the most populous city in and the government seat of the County of Monterey. The city 24 is home to more than 150,000 people, including 38.5% of the county’s “hard-to-count” 25 population. As of August 14, 2020, 57.2% of all households in Salinas have responded to the 26 2020 Census, which is 422nd out of all 482 California cities. The current response rate is 7.9 27 percentage points below California’s statewide average for self-responses and more than 10 28 percentage points below Salinas's self-response rate from the 2010 Census. ATTORNEYS AT LAW SAN FRANCISCO 10 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 12 of 62 1 36. Salinas has dedicated significant resources to funding and staffing its “Census 2 Action Team,” which is composed of city staff and representatives from the County of 3 Monterey’s “Complete Count Committee,” as well as community-based organizations, school 4 districts, and local businesses. The city’s population is more than 75% Latino, and more than 1 in 5 5 households have limited English-language proficiency. As part of its outreach, the Salinas 6 Census Action Team engages religious and community organizations, such as local food banks, 7 to assist with enumeration efforts in the Latino community and all communities of color as these 8 organizations are able to assist with trust and communication barriers that can make these groups 9 hard to count. The City of Salinas also receives substantial federal funding tied to census data. 10 37. The City of San Jose is a political subdivision of the State of California. San Jose 11 has over 1 million residents, making it the largest city in Northern California, and the tenth 12 largest city in the United States. San Jose’s population is 32% Latino, and 35% Asian, and nearly 13 40% of residents are foreign born. As of August 14, 2020, 28% of households in San Jose had 14 not responded to the census. San Jose has large populations of historically undercounted 15 communities. For instance, according to the Department of Housing and Urban Development, in 16 2019, San Jose had over 6,000 residents experiencing homeless. In addition, the San Jose metro 17 area is estimated to have over 150,000 undocumented immigrant residents. 18 38. The City of San Jose has engaged in extensive public-education and get-out-the- 19 count efforts during the 2020 Census. San Jose has formed a Complete Count Committee with 20 Santa Clara County, and nearly 90 community-based organizations. The Committee focuses on 21 raising awareness of the census in historically undercounted communities. San Jose also 22 disseminates information about the census to the public through city departments and offices. 23 San Jose also worked closely with the Census Bureau to recruit qualified bilingual enumerators. 24 The City of San Jose receives substantial federal funding tied to census data. 25 39. Plaintiff Rodney Ellis is the Commissioner for Precinct One on the Harris County 26 Commissioners Court. He is a resident and citizen of Harris County, where he is registered to 27 vote and regularly exercises his right to vote. Commissioner Ellis regularly drives on roads and 28 highways in Harris County. ATTORNEYS AT LAW SAN FRANCISCO 11 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 13 of 62 40. 1 Plaintiff Adrian Garcia is the Commissioner for Precinct Two on the Harris 2 County Commissioners Court. He is a life-long resident and citizen of Harris County, where he 3 is registered to vote and regularly exercises his right to vote. Commissioner Garcia also regularly 4 drives on roads and highways in Harris County. 5 II. Defendants 41. 6 Defendant Wilbur L. Ross is the Secretary of the U.S. Department of Commerce 7 and is sued in his official capacity. Secretary Ross oversees the U.S. Department of Commerce 8 and the Census Bureau. Congress has delegated the responsibility for carrying out the decennial 9 census to the Secretary of Commerce. 13 U.S.C. § 141(a). 42. 10 11 Executive Branch responsible for administering the decennial census. 43. 12 13 44. FACTUAL ALLEGATIONS I. 18 19 Defendant U.S. Census Bureau is an agency within the Department of Commerce responsible for planning and administering the decennial census. 13 U.S.C. § 2. 16 17 Defendant Steven Dillingham is the Director of the U.S. Census Bureau and is sued in his official capacity. 14 15 Defendant U.S. Department of Commerce is a cabinet agency within the Defendants’ Constitutional and Statutory Obligations. 45. Under the United States Constitution, the federal government must conduct an “actual Enumeration” of the population once every ten years. U.S. Const. art. I, § 2. 20 46. The population totals produced by the decennial enumeration are used to 21 apportion congressional representatives to the various states. Id. Census figures are also used in 22 state and local redistricting and in the distribution of federal funds to communities across the 23 United States. 24 47. The Enumeration Clause requires that decisions relating to the census bear a 25 “reasonable relationship” to the constitutional purpose of the enumeration. Wisconsin, 517 U.S. 26 at 20. 27 28 ATTORNEYS AT LAW SAN FRANCISCO 48. Similarly, the Census Act imposes a mandatory duty on the Secretary of Commerce to “conduct a census that is accurate and that fairly accounts for the crucial 12 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 14 of 62 1 representational rights that depend on the census and the apportionment.” Dep’t of Commerce v. 2 New York, 139 S. Ct. 2551, 2569 (2019) (citation omitted). 49. 3 Consequently, the Secretary of Commerce and the Census Bureau are 4 constitutionally obligated to make decisions in conducting the census that are reasonably related 5 to achieving a fair and accurate calculation of the population of the United States. 6 II. The Census Bureau’s Pre-COVID-19 Operational Plans for the 2020 Census. 50. 7 For the 2020 Census, the Census Bureau spent the better part of a decade 8 designing operations to fulfill its constitutional and statutory mandate, including: soliciting and 9 incorporating feedback from seasoned experts, advisors, and community groups; testing various 10 features of its data-collection and data-processing operations; and ensuring that its decisions for 11 conducting the census reflected sound, scientifically based judgment. 51. 12 To this end, the Bureau created an operational plan to guide its efforts, including 13 its efforts to collect data from census respondents and to process that data into usable forms for 14 constitutionally and statutorily mandated purposes, including reapportionment and redistricting. 52. 15 On December 31, 2018, the Bureau promulgated the final version of its 16 operational plan, which the Bureau called “Version 4.0” (hereinafter referred to as the “Final 17 Operational Plan”). See U.S. Census Bureau, Final Operational Plan (Dec. 2018), 18 https://www2.census.gov/programs-surveys/decennial/2020/program-management/planning- 19 docs/2020-oper-plan4.pdf. In the Final Operational Plan, the Census Bureau stated that its goal 20 for the 2020 Census is to “count everyone once, only once, and in the right place.” 53. 21 Under the Paperwork Reduction Act, the Office of Management and Budget must 22 review and approve the plans for any federal survey, including the decennial census, to ensure 23 that those surveys meet government standards, minimize respondent burden, and maximize the 24 utility of the collected information. 44 U.S.C. § 3504(c). 54. 25 The Office of Management and Budget formally reviewed and approved the 26 Census Bureau’s pre-COVID-19 plans for the decennial census, including the Final Operational 27 Plan. 28 ATTORNEYS AT LAW SAN FRANCISCO 55. The Final Operational Plan includes over 200 pages of detailed and transparent 13 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 15 of 62 1 2 3 4 conclusions for achieving the 2020 Census’s objective of an accurate count. 56. The Final Operational Plan reflects the conclusions of various experts including survey methodologists, statisticians, demographers, geographers, linguists, and mathematicians. 57. The Final Operational Plan states that it “reflects and supports evidence-based 5 decision-making” about the operations necessary to gather and process census responses from 6 every household in the country. 7 8 9 10 11 58. The Final Operational Plan states that it was “informed through research, testing, and analysis conducted from 2012 through 2018.” 59. The Bureau conducted at least fifteen tests between 2012 and December 31, 2018, when it published its Final Operational Plan. 60. Career Bureau staff developed the Final Operational Plan following substantial 12 consultation with outside experts and census stakeholders, including members of the Census 13 Scientific Advisory Committee and the National Advisory Committee. 14 61. The Census Bureau also produced a series of “detailed operational plans,” which 15 supplement the Final Operational Plan, and provide more parameters for the individual 16 operations that, together, comprise the 2020 Census. 17 18 19 62. The detailed operational plans likewise reflect the conclusions of various subject- matter experts regarding how to complete an accurate count. 63. The Bureau’s Final Operational Plan contains several major categories of 20 operations. Two of those categories are particularly important for purposes of this lawsuit: data- 21 collection and data-processing. 22 23 24 64. “Data-collection” refers to operations through which the Bureau obtains information from and about all the people living in the United States. 65. “Data-processing” refers to operations through which the Bureau fills in any gaps 25 in the personal information that it collects from people, transforms the resulting data into usable 26 forms, checks those results for accuracy and other aspects of data quality, and publishes those 27 results, among other things. 28 ATTORNEYS AT LAW SAN FRANCISCO 66. The Bureau must thoroughly, fully, and correctly perform both categories of 14 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 16 of 62 1 operations—collection and processing—to achieve its stated goal of counting everyone once, 2 only once, and in the right place. 3 A. Census Data Collection 4 5 6 67. During the census, the Bureau attempts both to determine the number of people in the country and their characteristics, such as their race and ethnicity. 68. Although the Census Bureau planned to deploy many methods during the 2020 7 Census to collect counts and characteristics from households around the country, the Bureau 8 contemplated, in both the Final Operational Plan, and in the supplemental detailed operational 9 plans, that three methods would account for the overwhelming majority of census responses: the 10 “Self-Response” method; the “Update Leave” method; and the “Non-Response Follow Up” 11 method. See U.S. Census Bureau, 2020 Census Detailed Operational Plan for: 18. Non- 12 Response Follow Up Operation (July 15, 2019), https://www2.census.gov/programs- 13 surveys/decennial/2020/program-management/planning-docs/NRFU-detailed-operational- 14 plan_v20.pdf. 15 69. The Self-Response method was the “primary methodology for the 2020 Census.” 16 Under this method, heads of households would provide their 2020 Census responses directly to 17 the Census Bureau by mailing back a paper census form, filling out a digital form on the 18 Bureau’s online census portal, or calling into telephone hotlines to provide their responses to 19 Bureau employees operating those hotlines. 20 70. The Update Leave method was the methodology for reaching housing units that 21 could not receive physical mail or did not have verifiable mailing addresses. Under this method, 22 Bureau employees would travel throughout both rural and urban areas, leaving invitations to 23 participate and paper census questionnaires at these housing units, so that the people living in 24 those locations could respond themselves. 25 71. The Self-Response method and the Update Leave method are crucial for obtaining 26 accurate information about the number of people in the country and their characteristics, because 27 data people report about themselves and the members of their housing units is the highest quality 28 data that the census collects. ATTORNEYS AT LAW SAN FRANCISCO 15 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 17 of 62 1 72. But for the tens of millions of households that do not report their personal data 2 through the Self-Response or Update Leave method, the Bureau’s next-best source of personal 3 data is data it collects directly from people through the Non-Response Follow Up method. 4 73. As part of the Non-Response Follow Up method, the Bureau sends its employee 5 enumerators directly to housing units so that they can attempt to speak with a person occupying 6 each unit and obtain information about everyone who should be counted in that unit. 7 74. The Bureau requires enumerators to record their responses for each household 8 through iPhones that the Bureau specifically contracted and customized for this purpose. The 9 enumerators’ iPhones include software designed to lead enumerators consistently and reliably to 10 solicit information from people at their doors. The enumerators’ iPhones also include software to 11 ensure that any data collected from housing units remains confidential as it is being transmitted 12 to the Bureau. The limited supply of these customized iPhones places a limit on the number of 13 enumerators that the Bureau can deploy in the field. 14 75. The Bureau’s Detailed Operational Plan for Non-Response Follow Up, which 15 supplements the Final Operational Plan, sets out a specific protocol for conducting Non- 16 Response Follow Up. 17 76. Under the Detailed Operational Plan, each housing unit assigned for a visit from 18 an enumerator was eligible for up to six “contact days.” A “contact day” could include more than 19 one attempted contact per day. 20 77. The Bureau concluded it could pursue less than six contact days only under 21 certain scenarios. 22 78. One scenario that would allow the Bureau to pursue fewer than six contact days 23 was the existence of high-quality administrative records for the housing unit. The Census Bureau 24 has collected data from federal administrative agencies, such as the Social Security 25 Administration, the Internal Revenue Service, and the Department of Housing and Urban 26 Renewal, among others, as well as data from states, which it uses to provide information about 27 the count and characteristics of non-responsive households. 28 ATTORNEYS AT LAW SAN FRANCISCO 79. If the Bureau had located administrative data from federal and/or state 16 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 18 of 62 1 administrative records and concluded that those records contained accurate demographic data for 2 the occupants of a housing unit, the Bureau’s enumerators would attempt only one contact with 3 that unit. If—during that contact attempt—the enumerator did not succeed in finding a live 4 person at the unit, then the Bureau would use the information in the administrative records to fill 5 in the census responses for that unit during the data-processing phase of the 2020 Census. 80. 6 A second scenario that would allow the Bureau to pursue less than six contact 7 days would arise if the Bureau identified a proxy—a person such as a neighbor or landlord that 8 the enumerator could ask for information about the occupants of the housing unit in question. 9 After a third failed contact attempt, a unit would become eligible for being counted through 10 11 proxy. 81. Proxies can produce many types of data. For instance, proxies are useful for 12 helping the Bureau identify whether a housing unit is vacant—and thus should be marked 13 “vacant” in the Master Address File that the Bureau uses to keep track of the overwhelming 14 majority of housing units that it must enumerate—or non-existent—and thus should be deleted 15 from the Master Address File. For the 2020 Census, the Bureau is planning to use administrative 16 records, such as the United States Postal Service’s directory of non-deliverable addresses, to 17 identify vacant housing, but proxies are generally more accurate for this purpose. Finally, proxies 18 provide vital data for other operations that the Bureau undertakes during its data-processing 19 phase, described further below. 20 82. If the Bureau is unable to enumerate a household after six contact days, in most 21 cases, it will resort to less accurate methods for determining the count and characteristics of the 22 household during its data-processing phase, described below. 23 83. The Bureau performs several other vital operations in addition to door-knocking 24 during the Non-Response Follow Up period, including a series of operations to ensure the quality 25 of the data that it collects in the field. 26 84. During the Non-Response Follow Up process, the Bureau: follows up with people 27 who self-responded to the census online but did not enter their unique census identification 28 number to ensure that they are counted in the right place (a process known as “Field ATTORNEYS AT LAW SAN FRANCISCO 17 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 19 of 62 1 Verification”); and corrects information reported erroneously or omitted from previously 2 submitted census forms (a process known as “Coverage Improvement”). 3 85. In addition, the Bureau re-collects census responses in select instances to ensure 4 that the original submissions were accurate (a process known as “Self-Response Quality 5 Assurance”). This operation protects against enumerators falsifying the information that they 6 provide to the Bureau. Specifically, the Bureau conducts quality control reinterviews of a sample 7 of households. This component is designed to deter and detect cases where enumerators have 8 provided false information about the housing units they are assigned to canvass. 9 86. Quality control reinterviews are part of a broader set of protocols that the Bureau 10 has developed to guard against factors that endanger the accuracy of the count. Non-Response 11 Follow Up is thus important not only for collecting information, but also for ensuring that the 12 information that is collected is accurate. These two components—gathering data and ensuring its 13 accuracy—must both occur for the Bureau to get a fair and accurate count. 14 87. The Bureau anticipated that approximately 60% of housing units nationally would 15 respond to the 2020 Census through Self-Response and Update Leave, potentially making up to 16 40% of housing units targets for Non-Response Follow Up. 17 18 19 20 21 88. A Non-Response Follow Up universe of 40% of the housing units in the country would have been the largest follow up universe on a percentage basis since at least 1970. 89. The Census Bureau did not anticipate that the Non-Response Follow Up universe in 2020 would mirror the demographic makeup of the nation’s population as a whole. 90. Instead, the Census Bureau anticipated that the Non-Response Follow Up 22 universe in 2020 would contain a disproportionate number of people who belong to communities 23 that the Bureau calls “hard-to-count.” 24 91. The Final Operational Plan describes hard-to-count populations as including, but 25 not limited to, the following populations: young children; highly mobile persons; racial and 26 ethnic minorities; non-English speakers; low-income persons; persons experiencing 27 homelessness; undocumented immigrants; persons who have distrust in the government; lesbian, 28 gay, bisexual, transgender, and questioning/queer (LGBTQ) persons; persons with mental and ATTORNEYS AT LAW SAN FRANCISCO 18 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 20 of 62 1 2 physical disabilities; and persons who do not live in traditional housing. 92. Historically, these populations have had low self-response rates and have, thus, 3 made up disproportionate shares of households that must receive contact days during Non- 4 Response Follow Up. 5 6 7 93. Consequently, the Final Operational Plan acknowledges, “[t]he NRFU Operation is entirely about hard-to-count populations.” 94. The Final Operational Plan also acknowledges that hard-to-count populations may 8 require more outreach than the Non-Response Follow Up method would normally provide, and 9 the Bureau designed its Final Operational Plan accordingly. 10 95. The Final Operational Plan states that “[w]hile most cases receive a maximum of 11 six attempts, cases in hard-to-count areas may receive more than six attempts to achieve a 12 consistent response rate for all geographic areas.” 13 96. Accurate data about the size, location, and characteristics of communities of color 14 is necessary to equitably distribute political power through congressional reapportionment and 15 redistricting at the state and local levels, enforce civil-rights laws that affect basic needs like 16 housing and employment, and conduct effective research, including on pressing issues like 17 public health. 18 B. Census Data-Processing 19 97. After collection activities are complete, the Census Bureau must process the data. 20 98. Census data-processing cannot begin until census data-collection concludes. 21 99. Census data is unusable for its intended purposes until it has been processed. 22 100. The Census Bureau’s data-processing operations transform tens of millions of 23 census responses into usable products, including the population totals used to reapportion seats in 24 the U.S. House of Representatives and to create electoral districts. 25 101. The Bureau uses its data-processing operations to, among other things, ensure that 26 data received from different data-collection methods are all in a single format allowing them to 27 be processed together. 28 ATTORNEYS AT LAW SAN FRANCISCO 102. The Bureau uses its data-processing operations to “unduplicate responses”— 19 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 21 of 62 1 meaning to resolve conflicts of information among multiple forms attributable to the same 2 housing unit. 3 103. The Bureau uses its data-processing operations to determine the final status of a 4 housing unit—such as vacant or inhabited—and determine the total number of people that should 5 be attributed to any apparently inhabited unit that was not counted through Self-Response, 6 Update Leave, or Non-Response Follow Up. 7 8 9 104. The Bureau also uses its data-processing operations to ensure that Bureau data products accurately report respondents’ characteristics, such as age, race, and ethnicity. 105. The Bureau uses administrative records and statistical imputation during the data- 10 processing phase to fill in both missing people and their characteristics. But administrative 11 records—especially low-quality administrative records—and statistical imputation are generally 12 less accurate than self-response data. 13 106. For many households, administrative data provides only low quality information, 14 replete with inaccuracies and incomplete information. This is especially the case for particular 15 communities that are underrepresented in administrative records, including communities of 16 color, immigrants, and low-income families. Use of this low-quality data to fill in missing 17 information for non-responsive households produces less accurate information. 18 107. Imputation involves the Bureau using information from surrounding responsive 19 households to infer the count and characteristics of a non-responsive household. Imputation thus 20 assumes the existence of other data points gathered through other data-collection methods—such 21 as self-response, proxies, and administrative records—and generates more accurate results when 22 it can be triangulated against those data points. The processes that the Bureau uses to collect and 23 process self-response data, proxy data, and administrative records are thus critical and 24 inextricably linked to the Bureau’s ability to impute data accurately. 25 108. At various phases of the Bureau’s data-processing operations, Census Bureau 26 personnel must review the quality of files in-process before those files can be sent to the 27 subsequent steps in the data-processing operation. These reviews include personnel with subject- 28 matter expertise from several different divisions of the Bureau. ATTORNEYS AT LAW SAN FRANCISCO 20 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 22 of 62 109. 1 The Bureau’s data-processing operations help ensure that people are not missed, 2 that other people are not counted multiple times, and that people’s characteristics are accurately 3 reported. These processes help eliminate or reduce undercounts, among other kinds of data- 4 quality issues. C. The Final Operational Plan’s Original Timeline for the 2020 Census 5 110. 6 7 January 21, 2020, to July 31, 2020, for a total of more than six months. 111. 8 9 10 The Bureau’s Final Operational Plan called for data-collection to run from In that window, the Self Response method was scheduled to run from March 12, 2020 to July 31, 2020, and the Update Leave method was scheduled for March 15, 2020 to April 17, 2020. 112. 11 The Bureau also scheduled several special operations to occur early in its census 12 taking process. The Service-Based Enumeration, which counts people experiencing 13 homelessness, was scheduled for March 30, 2020 to April 1, 2020, and Group Quarters 14 Enumeration, which counts people living in group housing such as nursing homes, was 15 scheduled from April 2, 2020 to June 5, 2020. 113. 16 17 2020 to July 31, 2020, for a total of approximately eleven and a half weeks. 114. 18 19 The Bureau scheduled up to five months—from July 31, 2020 to December 31, 2020—to process census data for the congressional reapportionment report. 115. 20 21 The Bureau scheduled the Non-Response Follow Up method to run from May 13, The Bureau also scheduled an additional three months—from January 1, 2021 to March 30, 2021—to process census data for redistricting. 116. 22 The Bureau’s timelines for implementing the Final Operational Plan reflect the 23 Bureau’s scientifically informed understanding of the time necessary to complete its operations 24 and generate an accurate count. 25 III. The Census Bureau’s COVID-19 Plan. 26 A. COVID-19 Disrupts the 2020 Census 27 117. On January 21, 2020, the Bureau began 2020 Census data-collection in remote 28 ATTORNEYS AT LAW SAN FRANCISCO Alaska. 21 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 23 of 62 1 118. On March 10, 2020, the Bureau began to accept self-responses on its website. 2 119. Shortly thereafter, many parts of the nation rapidly began to shut down due to the 3 4 COVID-19 pandemic. 120. The Census Bureau quickly concluded that it could not continue to engage in 5 operations safely. On March 18, 2020, the Bureau announced that it would suspend all field 6 operations for two weeks in order to “help protect the health and safety of the American public.” 7 Press Release, U.S. Census Bureau, U.S. Census Bureau Director Steven Dillingham on 8 Operational Updates (Mar. 18, 2020), https://www.census.gov/newsroom/press- 9 releases/2020/operational-update.html. 10 11 12 121. On March 28, 2020, the Bureau announced yet another two-week suspension until April 15, 2020, as the coronavirus pandemic made it impossible to engage in operations. 122. The suspension disrupted several field operations, including Update/Leave 13 method, the Service Based Enumeration counting people experiencing homelessness, and the 14 Group Quarters Enumeration counting people living in group housing. 15 123. In addition, the Bureau halted all hiring and training of the hundreds of thousands 16 of enumerators it needs to conduct Non-Response Follow Up. This included halting any and all 17 background checks and fingerprinting of enumerators that were conditionally hired at that time. 18 19 20 21 124. The Bureau also decreased office staff at regional centers responsible for processing mail-in self-response forms and at the Bureau’s call centers. B. Changes to the Final Operational Plan in the COVID-19 Plan 125. On April 13, 2020, the Bureau issued an adjustment to its Final Operational Plan 22 to account for the long-term impact of the COVID-19 pandemic. The new plan included a shifted 23 timeline for data-collection and data-processing operations that corresponded with the delays in 24 operations that the pandemic has caused (the “COVID-19 Plan”). 25 126. Adjustments to plans approved by the Office of Management and Budget under 26 the Paperwork Reduction Act must be re-submitted for approval. 44 U.S.C. § 3507(h)(3). The 27 Census Bureau submitted the COVID-19 Plan to the Office of Management and Budget on April 28 30, 2020. The changes were approved on May 11, 2020. ATTORNEYS AT LAW SAN FRANCISCO 22 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 24 of 62 127. 1 The COVID-19 Plan was designed to “[e]nsure a complete and accurate count of 2 all communities,” “[p]rotect the health and safety of the American public and Census Bureau 3 employees,” and “[i]mplement guidance from federal, state, and local authorities regarding 4 COVID-19.” Press Release, U.S. Census Bureau, 2020 Census Operational Adjustments Due to 5 COVID-19 Fact Sheet (Apr. 27, 2020). 128. 6 The COVID-19 Plan reflected the conclusions of various experts for how best to 7 proceed with completing an accurate count during the current pandemic. These experts include 8 survey methodologists, statisticians, demographers, geographers, linguists, and mathematicians. 129. 9 Under the COVID-19 Plan, the Bureau suspended 2020 Census field operations 10 for several months, including those operations that were designed to ensure a full count of 11 traditionally undercounted communities. 130. 12 The COVID-19 Plan provided that the Bureau would start the nationwide Non- 13 Response Follow Up operation on August 11, 2020, and continue the door-knocking process 14 through October 31, 2020. 131. 15 Thus, the COVID-19 Plan delayed the start of most door-knocking by three 16 months while maintaining the same amount of time spent undertaking the process— 17 approximately eleven and a half weeks—as the Final Operational Plan had required. 132. 18 Under the COVID-19 Plan, the Bureau also delayed the start of other operations 19 that enumerate traditionally undercounted populations, including the enumeration of the 20 country’s homeless population, which the Bureau shifted from March 30, 2020 to September 22, 21 2020. 22 133. And the COVID-19 Plan permitted households to submit self-response data to the 23 Bureau until October 31, 2020, extending the deadline under which private persons were able to 24 submit their responses to be counted by more than one month. 25 134. The Bureau also granted itself one additional month to process data under its 26 COVID-19 Plan, extending the data-processing leg of its operations to nine months given the 27 pandemic. Under this plan, the Bureau would have up to six months to process the data for the 28 apportionment count (between October 31, 2020 and April 30, 2021) and three months to process ATTORNEYS AT LAW SAN FRANCISCO 23 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 25 of 62 1 2 the data for redistricting (between April 30, 2021 and July 31, 2021). 135. The Bureau’s timelines for implementing the COVID-19 Plan reflect a 3 scientifically informed understanding of the time necessary to appropriately and fully complete 4 its operations and generate an accurate count. 5 6 C. Expert and Stakeholder Response to the COVID-19 Plan 136. The Census Bureau solicited feedback on the COVID-19 Plan from relevant area 7 experts and interested stakeholders, including state and local governments and national and 8 community-based non-profit partners. 9 137. For instance, four former Census Bureau Directors—who served under both 10 Democratic and Republican administrations—issued a statement saying that they had “discussed 11 these operational and schedule adjustments with senior career leadership at the Census Bureau.” 12 Press Release, Vincent Barabba et al., Statement by Former U.S. Census Bureau Directors (Apr. 13 14, 2020), https://www.documentcloud.org/documents/6838166-Statement-by-Former-Census- 14 Bureau-Directors-04.html. 15 138. These four former Census Bureau Directors further asserted: “Based on (1) our 16 extensive experience in planning, executing, and often adjusting operations of previous decennial 17 censuses, and (2) our firm conclusion that the extension of the field operations reflect careful 18 analysis by the technical, scientific, and operational staff at the Census Bureau, we support the 19 decision and urge Congress to act in concert with it.” Press Release, Vincent Barabba et al., 20 Statement by Former U.S. Census Bureau Directors (Apr. 14, 2020), 21 https://www.documentcloud.org/documents/6838166-Statement-by-Former-Census-Bureau- 22 Directors-04.html. 23 139. Prominent civil-rights groups endorsed the COVID-19 Plan. Vanita Gupta, 24 President and CEO of The Leadership Conference on Civil and Human Rights and The 25 Leadership Conference Education Fund, stated that her organization “support[ed] the Census 26 Bureau’s updated timeline.” Press Release, Leadership Conference Education Fund, Census 27 Timeline Must Protect Health, Ensure Fair Count (Apr. 13, 2020), 28 https://civilrights.org/edfund/2020/04/13/census-timeline-must-protect-health-ensure-fair-count/. ATTORNEYS AT LAW SAN FRANCISCO 24 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 26 of 62 1 2 D. Implementation of the COVID-19 Plan 140. When announcing the COVID-19 Plan, Secretary Ross and Director Dillingham 3 issued a statement indicating that the Bureau requested that Congress extend by 120 days the 4 December 31, 2020 statutory deadline for reporting the state-population totals to the President 5 for purposes of calculating the state apportionments, and extend by 120 days the March 30, 2021 6 statutory deadline for delivering redistricting data to the states. 7 141. That same day, President Trump suggested this request was unnecessary, stating: 8 142. “I don’t know that you even have to ask them. This is called an act of God. This is 9 called a situation that has to be. They have to give in. I think 120 days isn’t nearly enough.” 10 Hansi Lo Wang, Trump Officials Ask to Delay Census Data for Voting Districts, House Seats, 11 NPR (Apr. 13, 2020), https://www.npr.org/2020/04/13/833546675/trump-officials-ask-to-delay- 12 census-data-for-voting-districts-house-seats. 13 143. Indeed, the Census Bureau did not wait for Congress to act before beginning 14 implementation of the COVID-19 Plan. And the Bureau continued implementation of the 15 COVID-19 Plan for over three months through the end of July 2020. 16 17 18 144. For instance, the Census Bureau field operations remained suspended through May 2020. 145. The Bureau only began re-opening a few limited operations, such as the Update 19 Leave method, on a phased basis through mid-June 2020, over two months after the operation 20 was originally planned to occur in the Final Operational Plan. 21 146. The Bureau did not undertake any Non-Response Follow Up operations in most 22 of the country between May 13, 2020 and July 31, 2020, the timeframe originally set out in the 23 Final Operational Plan. 24 147. Instead, while the Bureau “soft-launched” door-knocking in select regions of the 25 country in mid-July 2020, the COVID-19 Plan did not call for door-knocking across the country 26 until August 11, 2020, at the earliest. 27 28 ATTORNEYS AT LAW SAN FRANCISCO 148. The Bureau ultimately opened six area census offices for Non-Response Follow Up on July 16, 2020, six more on July 23, 2020, thirty-five on July 30, 2020, and forty additional 25 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 27 of 62 1 2 offices on August 6, 2020. 149. The remaining 161 stateside offices remained unopened until August 9, 2020, 3 including offices in many states and localities with relatively low response rates such as the 4 entire southeastern United States, Texas, New Mexico, Arizona, and Southern California. 5 150. All along the Bureau continually communicated to the public, and to important 6 local partners, including local governments and national and community based non-profit 7 organizations, that self-responses would be accepted until October 31, 2020, and that Non- 8 Response Follow Up would continue until at least that date. 9 10 11 151. Census partners, stakeholders, and state and local governments relied on the new deadlines set forth in the COVID-19 Plan to redirect their outreach efforts. 152. For example, Plaintiffs Urban League and BAJI, publicized the October 31, 2020 12 deadline, letting their constituents, members and local organizations know that households had 13 until that time to self-respond. Urban League representatives informed coalition partners 14 participating in the Black Census Roundtable of the new deadlines, and spoke of the deadlines on 15 webinars and other public events. Officials at BAJI publicized the deadlines at public events, 16 including webinars in July 2020, and as part of the organization’s social media campaign. 17 153. Similarly, officials in City of Los Angeles, Harris County, King County, City of 18 San Jose, and City of Salinas, publicized the new deadline while conducting 2020 Census 19 outreach efforts. 20 154. These public education efforts were significant because they were directed at the 21 general public and at local non-profits that do not primarily work on census issues. The latter 22 often rely on information about the census provided by Plaintiff national non-profits and local 23 governments when communicating with their constituents. Plaintiffs, by disseminating the 24 October 31, 2020 deadline for nearly three months to the public, were largely successful in 25 spreading the understanding that communities had until at least that time to complete the count. 26 155. For example, the City of Los Angeles announced this date on its own social media 27 platforms and in a social media toolkit that it developed for partner organizations. Los Angeles is 28 deeply concerned that residents have already received information about the October 31, 2020 ATTORNEYS AT LAW SAN FRANCISCO 26 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 28 of 62 1 self-response date and, as a result, will fail to respond before the newly shortened deadline, 2 especially given the Bureau’s own minimal efforts at explanation and outreach around the new 3 deadline. 4 156. Finally, the level of self-response during the 2020 Census, and the ongoing 5 COVID-19 pandemic, provided further evidence for the necessity of continued implementation 6 of the COVID-19 Plan. 7 157. Under its Final Operational Plan, for example, the Census Bureau had planned to 8 spend eleven and a half weeks canvassing a Non-Response Follow Up universe comprised of 9 39.5% of households nationally. 10 158. As of August 9, 2020, the first date of nationwide Non-Response Follow Up, the 11 national self-response rate was 63.2%, meaning that nearly 37% of households nationwide had 12 not yet responded to the census. 13 159. Several cities with large percentages of traditionally undercounted populations, 14 have even lower response rates. For instance, as of August 14, 2020, the response rate in the City 15 of Detroit was 48.9%, Miami was 49.9%, Philadelphia was 52.3%, Los Angeles was 53.8%, 16 Houston was 54.4%, and New York City was 55.6%. 17 160. The United States had 24,156 new coronavirus cases on April 13, 2020, the day 18 the Bureau announced its COVID-19 Plan. On August 3, 2020, the United States had 19 approximately 50,000 new coronavirus cases. 20 161. With COVID-19 limiting the willingness of people to apply for enumerator 21 positions, the areas where the Bureau can safely send enumerators to knock on doors, and the 22 willingness of the public to interact with enumerators, the Non-Response Follow Up operation 23 continues to face far more complications than the Final Operational Plan anticipated. 24 162. Given these conditions of low response rates and increased coronavirus spread, 25 the Bureau can reasonably expect that it will need to engage in a Non-Response Follow Up 26 operation at least as comprehensive and time-consuming as the operation laid out in the Final 27 Operational Plan. 28 ATTORNEYS AT LAW SAN FRANCISCO 163. Due to significant delays in operations resulting from the implementation of the 27 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 29 of 62 1 COVID-19 Plan, the Bureau itself has recognized that it would be impossible to produce fair and 2 accurate apportionment numbers to the President by December 31, 2020. 3 164. On May 27, 2020, Tim Olson, head of field operations for the 2020 Census, stated 4 during a May 26, 2020 webinar organized by the National Congress of American Indians that, 5 “[w]e have passed the point where we could even meet the current legislative requirement of 6 December 31st. We can’t do that anymore.” Nat’l Conf. of Am. Indians, 2020 Census Webinar: 7 American Indian/Alaska Native, YouTube (May 26, 2020), https://www.youtube.com/ 8 watch?v=F6IyJMtDDgY&feature=youtu.be&t=4689. 9 165. On July 8, 2020, Al Fontenot, Jr., Associate Director for Decennial Census 10 Programs and a top Census Bureau official, affirmed that the Bureau is “past the window of 11 being able to get” accurate counts to the President by December 31, 2020. U.S. Census Bureau, 12 Operational Press Briefing – 2020 Census Update at 21 (July 8, 2020), 13 https://www.census.gov/content/dam/Census/newsroom/press-kits/2020/news-briefing-program- 14 transcript-july8.pdf. 15 IV. The Census Bureau’s New Rush Plan. 16 A. The Announcement of the Rush Plan 17 166. On August 3, 2020, at the behest of the Secretary of Commerce, Director 18 Dillingham abruptly and without explanation abandoned the COVID-19 Plan and announced the 19 Rush Plan. 20 167. The Rush Plan drastically shortens the timelines for multiple operations set out in 21 the COVID-19 Plan. 22 168. The Rush Plan took the form of a short press release on the Census Bureau’s 23 website. The press release included a statement from Director Dillingham, which did not provide 24 an explanation for Defendants’ decision to suddenly abandon the COVID-19 Plan that the 25 Bureau had adopted and implemented for approximately three and a half months. Nor did it 26 provide any specifics as to why the Bureau no longer believed the timelines called for in the 27 COVID-19 Plan were necessary to ensure an accurate count. 28 ATTORNEYS AT LAW SAN FRANCISCO 169. The statement noted that the Bureau was taking this action at the direction of the 28 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 30 of 62 1 Secretary of Commerce. But the Secretary made no statement explaining his reason for giving 2 this directive. 3 170. The Director’s statement was largely silent on specific adjustments the Bureau 4 would need to make in order to reengineer its field operations to meet its new, artificially 5 compressed schedule. The statement included proposals for enumerator “awards” and 6 maximizing enumerators’ phone and tablet usage, but it did not provide any details about 7 adjustments to the detailed operations provided in the Final Operational Plan. 8 9 10 11 12 171. The only adjustments announced under the Rush Plan were severely truncated timelines for conducting data-collection and data-processing operations. 172. Under the Rush Plan, data-collection is now set to end on September 30, 2020, one month earlier than contemplated in the Bureau’s COVID-19 Plan. 173. While the Bureau’s pre-COVID-19 Final Operational Plan provided 79 days for 13 the nationwide door-knocking stage of the census, and the COVID-19 Plan provided 81 days, the 14 Rush Plan provides just 52 days of nationwide door-knocking. 15 174. The Rush Plan also cuts post-collection data processing for the apportionment 16 report from up to 6 months as provided in the COVID-19 Plan, and up to 5 months as originally 17 provided in the Final Operational Plan, to less than 3 months. 18 175. The Rush Plan also shortened the time under which households can self-respond, 19 providing that self-responses delivered after September 30, 2020—which previously would have 20 been timely under the October 31, 2020 deadline—will no longer be counted. 21 176. While the Rush Plan requires the Bureau to accelerate its operations to complete 22 the 2020 Census by the same deadline contemplated in the Final Operational Plan, it ignores the 23 multiple-month pause in operations, beginning in mid-March 2020, caused by the initial outbreak 24 of COVID-19 in the United States. 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO 177. The decision to rescind the COVID-19 Plan and adopt the Rush Plan was announced without consultation with important stakeholders. 178. As noted above, as late as July 8, 2020, senior Bureau officials were still confirming that it was impossible to complete an accurate count by December 31, 2020. 29 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 31 of 62 1 179. In addition, until July 30, 2020, just four days before the Bureau announced its 2 decision to abandon the COVID-19 Plan, the Bureau was informing respondents on its website 3 that it would engage in Non-Response Follow-Up until October 31, 2020 and that non- 4 responsive households would have until that date to self-respond. Those references were deleted 5 from the website on or about July 31, 2020 and were replaced with the shortened timeframe after 6 the August 3, 2020 announcement. 7 180. An official at the Government Accountability Office confirmed that Bureau 8 officials told his office that they were given “hours rather than days or weeks” to adjust their 9 plans to finish counting by September 2020. Hansi Lo Wang, ‘Not Enough Time’: Census 10 Workers Fear Rushing Count Could Botch Results, NPR (Aug. 11, 2020), 11 https://www.npr.org/2020/08/11/901202892/not-enough-time-census-workers-fear-rushing- 12 count-could-botch-results. 13 181. While the Census Bureau’s decisions, even during the COVID-19 emergency, 14 have often involved consultations with scientific advisory committees, the Committee on 15 National Statistics in the National Academies of Science, other external experts and local 16 government officials, and the thousands of organizations partnering with the Bureau to conduct 17 crucial outreach to historically undercounted communities, no such consultation was made 18 before the Bureau announced its abandonment of the COVID-19 Plan. 19 20 21 182. Census stakeholders immediately denounced the Rush Plan, including stakeholders who had endorsed the COVID-19 Plan. 183. The same four former Census Bureau Directors who endorsed the COVID-19 22 Plan issued a statement saying that “our expert opinion is that failing to extend the deadlines to 23 April 30, 2021 will result in seriously incomplete enumerations in many areas across our 24 country.” Press Release, Former Census Bureau Directors, On the Importance of Extending the 25 2020 Census Statutory Deadlines to Achieve a Fair and Accurate Enumeration of the United 26 States (Aug. 4, 2020), https://www.documentcloud.org/documents/7013550-Aug-4-2020- 27 Statement-By-Former-U-S-Census-Bureau.html. 28 ATTORNEYS AT LAW SAN FRANCISCO 184. These four former Census Bureau Directors further asserted: “The Census Bureau 30 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 32 of 62 1 will not be able to carry out the NRFU fully and will be forced to take steps such as fewer in- 2 person visits and rely instead on the use of administrative records or statistical techniques on a 3 much larger scale tha[n] in previous census. The end result will be under-representation of those 4 persons that NRFU was expected to reach and, at even greater rates for traditionally hard-to- 5 count populations and over-representation of all other populations with potentially extreme 6 differential undercounts.” Press Release, Former Census Bureau Directors, On the Importance 7 of Extending the 2020 Census Statutory Deadlines to Achieve a Fair and Accurate Enumeration 8 of the United States (Aug. 4, 2020), https://www.documentcloud.org/documents/7013550-Aug- 9 4-2020-Statement-By-Former-U-S-Census-Bureau.html. 10 185. The President of the American Statistical Association, the world’s largest 11 professional organization of statisticians, issued a statement saying “[t]here is no scientific 12 rationale to curtail the data-collection period for this constitutionally mandated activity, and the 13 premature cessation of census enumeration will produce flawed counts.” Letter from Rob 14 Santos, President of the American Statistical Association, to Mitch McConnell, U.S. Senate 15 Majority Leader (Apr. 5, 2020), https://www.amstat.org/asa/files/pdfs/POL- 16 CensusSenateAugust.pdf. 17 186. Nearly 450 nonpartisan philanthropic organizations who “rely on accurate census 18 data to help identify community needs and to prioritize grantmaking” issued a letter to Secretary 19 Ross and Director Dillingham urging the Bureau to revert to its COVID-19 Plan. Letter from 20 U.S. Philanthropy Leaders to Wilbur Ross, Secretary of the U.S. Dep’t of Commerce (Aug. 5, 21 2020), https://funderscommittee.org/wp-content/uploads/2020/08/Letter-Philanthropic-Leaders- 22 on-Census-Being-Cut-Short-8-5.pdf. 23 187. Prominent civil-rights groups condemned the Rush Plan. Vanita Gupta, President 24 and CEO of The Leadership Conference on Civil and Human Rights and The Leadership 25 Conference Education Fund, stated that “[c]urtailing operations is an obvious ploy to guarantee 26 the Census Bureau won’t be able to finish counting millions of people—especially those hit 27 hardest by the pandemic.” Press Release, Leadership Conference on Civil and Human Rights, 28 Trump Plans to Sabotage 2020 Census by Cutting Short Operations (July 31, 2020), ATTORNEYS AT LAW SAN FRANCISCO 31 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 33 of 62 1 https://civilrights.org/2020/07/31/trump-plans-to-sabotage-2020-census-by-cutting-short- 2 operations/. 3 188. 4 this new timeline, explaining that the Rush Plan means that it will not be an accurate count for 5 the next 10 years. 6 B. The Rush Plan Fails to Appropriately Account for Key Factors Affecting the 2020 Census 189. The Rush Plan fails to account for several important factors that affect the 2020 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO And the Census Bureau’s own field workers have confirmed the impossibility of Census Non-Response Follow Up operation. 190. First, the Rush Plan does not adequately account for the large number of households in the Non-Response Follow Up universe. 191. Under the Rush Plan, the Census Bureau must attempt to count approximately the same number of households during Non-Response Follow Up as it anticipated counting in its pre-COVID-19 Final Operational Plan, but the Bureau will have four weeks less than provided in that plan to complete the operation. In other words, the Bureau must now try to complete the same amount of work in just 65% of the time it had originally scheduled to complete that work. 192. Over 37% of households nationwide are non-responsive, and several states have even higher percentages of households in the Non-Response Follow Up universe, including New Mexico (46.1%), South Carolina (42.4%), Texas (41.3%), and Georgia (40.8%). 193. While soft-launches of Non-Response Follow Up began in select locations in mid-July 2020, the operation did not begin in any of these states, with large amounts of nonresponsive households, until August 9, 2020. 194. Within states, and in particular cities and localities, there are even higher Non- Response Follow Up workloads. For instance, in Plaintiff Harris County, enumerators must still visit over 41% of households. In the City of Los Angeles, over 46% of households remain to be enumerated. The self-response rate in Los Angeles is approximately 14 percentage points below the final self-response rate the City attained during the 2010 Census. Counting in these jurisdictions also did not begin until August 9, 2020. 32 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 34 of 62 1 195. Moreover, given the time constraints placed by the Rush Plan, counting will need 2 to be conducted while these jurisdictions, in many places, struggle to control a surge in COVID- 3 19 cases. 4 196. While the Bureau announced on August 11, 2020 that it is now “training census 5 takers to follow up with households by phone” in light of the pandemic, that change in Non- 6 Response Follow Up operations was not anticipated in the Final Operational Plan. Press 7 Release, U.S. Census Bureau, Door-to-Door Visits Begin Nationwide for 2020 Census (Aug. 11, 8 2020), https://www.census.gov/newsroom/press-releases/2020/door-to-door-visits-begin- 9 nationwide.html. 10 11 12 197. Given the traditionally low response rates for phone surveys in the wireless era, following up by phone is unlikely to materially increase response rates. 198. A recent Census Bureau survey running in parallel with the 2020 Census 13 demonstrates the difficulty in obtaining responses via phone or email. This spring, the Bureau 14 began conducting a “Household Pulse Survey” to measure household experiences under the 15 COVID-19 pandemic. This survey solicited participation through emails and text messages. Over 16 the first twelve weeks of this survey, response rates were meager, ranging from 1.3% to 3.8%. 17 18 19 20 21 199. Second, the Rush Plan does not account for the staffing challenges that the Bureau is currently experiencing, many of which are directly related to the ongoing pandemic. 200. As demonstrated in the soft-launch of Non-Response Follow Up in select locales, the Bureau is already experiencing staffing shortages and retention problems with enumerators. 201. In the midst of the ongoing pandemic, prospective enumerators, many of whom 22 are elderly and at high risk of contracting a severe COVID-19 related illness, are less willing to 23 engage in the required door-to-door canvassing. 24 202. Indeed, Tim Olson, head of field operations for the 2020 Census, stated at a July 25 8, 2020 press briefing that “[a]bout a third of our [enumerator] applicants [are] older persons 26 considered high risk of the virus.” U.S. Census Bureau, Operational Press Briefing – 2020 27 Census Update at 21 (July 8, 2020), https://www.census.gov/content/dam/Census/ 28 newsroom/press-kits/2020/news-briefing-program-transcript-july8.pdf. ATTORNEYS AT LAW SAN FRANCISCO 33 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 35 of 62 1 203. And Deborah Stempowski, the Census Bureau’s Assistant Director for Decennial 2 Programs, noted the Bureau’s difficulty retaining enumerators in early August 2020, confirming 3 that potential enumerators were “a little hesitant because of the COVID environment.” Mike 4 Schneider, Census Bureau Drop-Outs Complicate Door-Knocking Efforts, Associated Press 5 (Aug. 8, 2020), https://www.usnews.com/news/us/articles/2020-08-08/census-bureau-drop-outs- 6 complicate-door-knocking-efforts. 7 204. In testimony before Congress on July 28, 2020, Director Dillingham confirmed 8 that the Bureau believed that “the pandemic is estimated to increase the number of no shows to 9 training sessions, as well as the number of employees who complete training but decline to show 10 11 up for work.” Id. 205. According to reports from census-operations staff working in the field, these 12 predictions have come to pass. One census field supervisor working in the mid-Atlantic noted 13 that, given the new rushed timeline and lack of sufficient staff, “[w]e’re just sending bodies out 14 regardless of whether they’re ready or not.” Hansi Lo Wang, ‘Not Enough Time’: Census 15 Workers Fear Rushing Count Could Botch Results, NPR (Aug. 11, 2020), https://www.npr.org/ 16 2020/08/11/901202892/not-enough-time-census-workers-fear-rushing-count-could-botch-results. 17 206. In addition to enumerator low-count and hesitancy, another source of staffing 18 issues involves delays in processing background checks on enumerator applicants and in 19 enumerator onboarding. 20 207. A June 2020 GAO report on the 2020 Census delays COVID-19 has caused, and 21 the risks the pandemic has exacerbated, noted that the Bureau “will have to quickly hire and 22 onboard sufficient staff to conduct its operations” to reach adequate staffing levels. U.S. Gov’t 23 Accountability Office, COVID-19 Presents Delays and Risks to Census Count (June 2020), 24 https://www.gao.gov/assets/710/707456.pdf. 25 208. That same report also noted that, once potential enumerators accept a job offer 26 from the Bureau, the new hires “must wait a minimum of 60 days before they can begin training, 27 a time period during which they must complete fingerprinting and a background check.” Id. 28 ATTORNEYS AT LAW SAN FRANCISCO 209. Reports from recently hired enumerators confirm that the Bureau is facing these 34 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 36 of 62 1 technical challenges as well, under the compressed timeline. One recent hire in Boulder, 2 Colorado noted that he lost six potential days of door-knocking because he was unable to 3 complete the Bureau’s online training module. 4 5 6 210. Thus, under the Rush Plan, the Bureau will not be able to hire and train sufficient enumerators. 211. Even if it were possible for the Bureau to hire all of the enumerators it will need, 7 the Bureau would also need time and funding to obtain additional equipment for any additional 8 enumerators it hires beyond its initial estimates of equipment. For example, the Bureau would 9 need more of the iPhones discussed above that the Bureau specifically contracted and 10 11 customized for 2020 Census enumerators. 212. With fewer enumerators in the field, in addition to training and equipment issues, 12 the Bureau cannot ensure that non-responsive households receive the requisite number of visits, 13 as contemplated in the Final Operational Plan. 14 213. While the Bureau had a $2 billion contingency fund prior to the existence of the 15 COVID-19 pandemic, it has already used $1.5 billion of that fund addressing pandemic-related 16 issues. The remaining $500 million will be needed to further respond to the pandemic, and, in 17 any event, is nowhere near the $1 billion that the administration claims that the Bureau would 18 need to conduct adequate Non-Response Follow Up operations under the Rush Plan. 19 214. Thus, instead of providing additional enumerators, the Bureau’s Rush Plan will 20 likely result in a smaller number of enumerators shouldering larger-than-planned workloads. 21 Increasing workloads for enumerators over a short period of time can result in errors and 22 inaccuracies in counting but it cannot make up for the time lost to the Rush Plan. 23 24 25 215. Third, the Rush Plan fails to account for factors relevant to efficient enumeration, such as the time when enumerators visit households. 216. For instance, under the Final Operational Plan, enumerators visit households at 26 specific times of day and on specific days of the week, depending on when residents are likely to 27 answer. 28 ATTORNEYS AT LAW SAN FRANCISCO 217. Under the Rush Plan, enumerators will be under pressure to complete their work 35 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 37 of 62 1 in a tightly constrained timeframe. As a result, ensuring that non-responsive households receive 2 the requisite number of enumerator visits at the most opportune times for enumeration may 3 become exceedingly difficult, if not impossible. Instead, the Rush Plan increases the likelihood 4 that households will either receive visits at less opportune times, or simply receive fewer visits 5 altogether. 6 218. Fourth, the Rush Plan fails to account for the additional crucial operations that 7 enumerators must conduct, as contemplated in the Bureau’s final plans for the 2020 Census. 8 Apart from visiting households upwards of six times, enumerators also engage in a host of 9 additional quality control activities. 10 219. As noted above, enumerators are expected to visit the households of persons that 11 self-responded to the census online but did not enter the unique identifier provided on census 12 mailers. This “non-ID processing” is necessary to verify the address information provided by 13 respondents. While this process only requires a single visit to a household, it nevertheless must 14 be completed in the compressed timeline provided for under the Rush Plan. 15 220. Similarly, the Bureau must conduct quality control reinterviews of a sample of 16 households during Non-Response Follow Up. This operation is designed to deter and detect 17 enumerator falsification. Detecting such falsifications will be especially important under the 18 Rush Plan where individual enumerators must shoulder a heavier workload. The use of 19 enumerators to conduct these reinterviews will, under the Rush Plan, place additional strain on 20 the Bureau’s already stretched labor resources. 21 221. Cutting any one of these functions will cause errors and inaccuracies to affect the 22 final 2020 Census data. By reversing the COVID-19 Plan and shortening the timeframe for 23 conducting Non-Response Follow Up by a month, the Bureau will likely need to make cuts to 24 one or more of these operations. 25 222. By reducing the amount of time and resources necessary to perform the kinds of 26 quality-control measures that the Bureau originally planned for Non-Response Follow Up, the 27 Rush Plan actively dismantles processes that the Bureau has specifically developed over the 28 course of time as checks against falsified census responses. The Rush Plan thus threatens census ATTORNEYS AT LAW SAN FRANCISCO 36 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 38 of 62 1 accuracy not only by reducing the Bureau’s time to collect data, but also by reducing the 2 Bureau’s time to ensure that the data it has collected has been collected properly and truthfully. 3 4 5 223. Fifth, the Rush Plan fails to account for the other field operations enumerators will need to conduct at the same time as they attempt to speed through door-knocking operations. 224. Under the Final Operational Plan, the Bureau planned to finish specialized 6 operations for counting people experiencing homelessness, and people living in group housing in 7 April 2020, before engaging in nationwide door-knocking. After suspending operations due to 8 COVID-19, the Bureau moved these operations to September 2020, well-before the October 31, 9 2020 deadline the Bureau set for completing the Non-Response Follow Up operation. 10 225. The new Rush Plan requires the Bureau to conduct these specialized operations at 11 the same time as it is scrambling to complete Non-Response Follow Up. This will further stretch 12 the Bureau’s limited resources and increase the likelihood of missing information. 13 C. The Rush Plan Also Fails to Appropriately Account for Factors that Will Affect 14 Post-Collection Data Processing 15 226. 16 17 The Rush Plan fails to account for the additional strain on data-processing operations resulting from the consequences of the COVID-19 pandemic. 227. Following the outbreak of COVID-19 in the United States in mid-March 2020, 18 colleges and universities across the country closed, and students moved out of campus and off- 19 campus housing. Similarly, many residents of cities, especially those living in COVID-19 20 hotspots, moved to locations where the virus was less prevalent. In a recent study, three percent 21 of people surveyed reported that they had moved permanently or temporarily as a result of the 22 pandemic. 23 24 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO 228. This significant movement of people coincided with Census Day, April 1, 2020, and will lead to confusion about what residence should be listed on responses. 229. It is likely that the Bureau will receive an increased amount of duplicate responses, which will, in turn, require more time and Bureau resources to review and correct. 230. The Rush Plan also fails to account for the Bureau’s inability to timely obtain and process all the administrative-records data crucial for completing an accurate count. 37 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 39 of 62 1 231. The Bureau relies principally on Title 26 data—that is, tax returns that individuals 2 file with the Internal Revenue Service (“IRS”)—for the administrative records it uses to fill in 3 missing people and their characteristics. 4 232. Because this year’s tax filing deadline was July 15, 2020, and the IRS generally 5 requires three months to transfer Title 26 data to the Census Bureau, the Bureau will not possess 6 all the Title 26 data it is planning to use until mid-October 2020, at the earliest. Once the Bureau 7 has possession of that Title 26 data, it will have to undertake a time-consuming round of 8 additional review and processing, further delaying its ability to use the data for its planned 9 purposes. These delays will compel the data-processing phase of 2020 Census operations to 10 11 proceed more slowly than the Rush Plan contemplates or would allow. 233. Ultimately, the solution to alleviate each of these problems was articulated in the 12 COVID-19 Plan: provide the Bureau’s limited number of enumerators with additional time to 13 conduct the data-collection operations necessary to ensure a complete and accurate census, and 14 provide Bureau staff with additional time to conduct the data-processing operations necessary to 15 ensuring the same. The Rush Plan fails to address these issues or explain why the Bureau's prior 16 conclusions were incorrect.. 17 18 D. The Rush Plan Does Not Account for Federal Statistical Guidelines 234. In replacing the COVID-19 Plan with the Rush Plan, Defendants departed from 19 federal government statistical standards that promote the accuracy of information collected and 20 disseminated by the agencies. 21 22 23 235. The Bureau’s failure to follow these standards further emphasizes its inability to conduct an adequate count in the time and under the conditions that the Rush Plan provides. 236. Under the Paperwork Reduction Act, the Office of Management and Budget is 24 responsible for coordinating the federal statistical system, including the development and 25 implementation of “Governmentwide policies, principles, standards, and guidelines” “concerning 26 [] statistical collection procedures and methods.” 44 U.S.C. § 3504(e)(3) (A). 27 28 ATTORNEYS AT LAW SAN FRANCISCO 237. The Office of Management and Budget is responsible for issuing guidelines that provide “procedural guidance to Federal agencies for ensuring and maximizing the quality, 38 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 40 of 62 1 objectivity, utility and integrity of information (including statistical information) disseminated by 2 Federal agencies.” Consolidated Appropriations Act, FY 2001, Pub. L. No. 106-554, § 515, 3 114 Stat. 2763 (2000). 4 238. One such guideline issued by the Office of Management and Budget provides 5 specific standards to agencies like the Census Bureau, in ensuring the quality and utility of 6 federal statistical surveys, such as the decennial census. Office of Mgmt. & Budget, Standards 7 and Guidelines for Statistical Surveys § 2 (2006). 8 9 10 239. Under these standards, agencies are required to develop “realistic timetable[s]” for surveys. Id. § 1.2. 240. The Bureau failed to take this basic requirement into account when it decided to 11 implement the Rush Plan. The Rush Plan compresses the timeline for counting operations despite 12 evidence of staffing shortages and heavier workload. The Plan attempts to accomplish a task— 13 speedy delivery of results by December 31, 2020—that the Bureau has already deemed 14 “impossible.” 15 241. The standards also require agencies, including the Census Bureau, to “[e]ncourage 16 respondents to participate to maximize response rates and improve data quality.” Office of 17 Mgmt. & Budget, Standards and Guidelines for Statistical Surveys § 2.3.2. This standard 18 requires that the Census Bureau “[e]nsure that the data collection period is of adequate and 19 reasonable length.” 20 242. Again, the Rush Plan does not account for this standard. The Final Operational 21 Plan and the COVID-19 Plan provided for over eleven weeks of Non-Response Follow Up, and 22 up to five and six months, respectively, of post-collection data processing for the apportionment 23 report. The Rush Plan, on the other hand, cuts the time allotted for counting by four weeks, 24 without explaining how it will encourage more efficiency in collecting responses than the plan it 25 reversed. 26 243. The standards also require the Bureau to plan for “an adequate number of contact 27 attempts” to the respondent and to establish protocols for minimizing enumerator falsification, 28 including “reinterviewing respondents.” Office of Mgmt. & Budget, Standards and Guidelines ATTORNEYS AT LAW SAN FRANCISCO 39 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 41 of 62 1 2 for Statistical Surveys, Directive No. 2, § 2.3.3. 244. With the Rush Plan significantly cutting the time available to conduct Non- 3 Response Follow Up, it is expected that the Bureau will need to cut particular Non-Response 4 Follow Up processes. This includes reducing the number of housing unit visits it earlier deemed 5 necessary to enumerate a non-responsive household, or cutting back on enumerator reinterviews. 6 Either decision will conflict with the Bureau’s obligation to abide by federal statistical standards. 7 8 9 10 E. The Rush Plan Will Produce Low Quality and Inaccurate Data 245. Ultimately, Defendants’ decision to rush completion of the 2020 Census will produce a significantly less accurate census than the COVID-19 Plan. 246. By cutting down the time allotted for door-knocking, the Rush Plan will result in 11 fewer contact days by enumerators to non-responsive households, and less data collected by 12 enumerators about those households. 13 247. The concerns about inaccuracy resulting from shortening time for Non-Response 14 Follow Up are real and verified. A GAO review of the 2010 Non-Response Follow Up operation 15 determined that local census offices with “higher percentages” of “less complete house-hold 16 data” were more likely to have completed their Non-Response Follow Up in 53 days or less as 17 compared to those offices that took a longer period of time. U.S. Gov’t Accountability Office, 18 2010 Census: Data Collection Operations Were Generally Completed as Planned, but Long- 19 Standing Challenges Suggest Need for Fundamental Reforms (Dec. 2010), 20 https://www.gao.gov/new.items/d11193.pdf. 21 248. As noted above, after the Bureau exhausts attempts to enumerate households 22 through methods that render more accurate results, such as self-response and enumerator 23 interviews, the Bureau turns to less accurate sources of data and statistical methods as a last 24 resort to fill in missing information. 25 249. By curtailing Non-Response Follow Up, the Rush Plan will force the Bureau to 26 resort to less accurate methods of data collection, well before the exhaustion of more accurate 27 methods. Consequently, the Rush Plan will lead to the production of lower-quality information. 28 ATTORNEYS AT LAW SAN FRANCISCO 250. For instance, under the Final Operational Plan, the Bureau would not consider 40 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 42 of 62 1 low-quality administrative data before conducting the requisite number of contact days for a 2 particular type of housing unit. By reducing the number of enumerator contact days, the Rush 3 Plan will lead to reliance on these types of lower-quality data sources prior to exhausting the 4 more accurate methods contemplated in the Final Operational Plan. Consequently, the Rush Plan 5 will lead to more inaccuracies in the data. 6 251. Based upon past practices, the Bureau may also use whole-count imputation to 7 calculate missing household data but to an extent and in ways not used previously. Imputation 8 involves the Bureau using information from surrounding responsive households to infer the 9 count and characteristics of a non-responsive household. 10 252. In previous censuses, the Bureau imputed upwards of 2.0% of households left 11 over after exhausting its Non-Response Follow Up efforts. Under the time constraints of the 12 Rush Plan, the Bureau will need to turn to imputation before exhausting its in-person 13 enumeration efforts. One former Census Bureau Director estimates that, under the Rush Plan, the 14 Bureau may end up imputing up to 10% of households. 15 253. Since data produced through the Bureau’s current imputation methods are less 16 accurate than data collected from enumerator interviews, Defendants’ decision to rush 17 completion of the 2020 Census will result in significantly less accurate total-population data than 18 would have been produced under the COVID-19 Plan. This decline in accuracy will affect both 19 the census’s calculations of the total number of people living in the country and the census’s 20 recording of the characteristics of those people, and such inaccurate data will not meet the 21 constitutional minimum for conducting the decennial enumeration or satisfy the “strong 22 constitutional interest in accuracy” of the Census. Utah v. Evans, 536 U.S. 452, 478 (2002). 23 254. The Rush Plan will likely exacerbate the quality problems associated with 24 imputation by compromising the Bureau’s ability to collect the other kinds of data—such as self- 25 responses, proxies, and administrative records—that it requires to impute most accurately. As 26 noted above, imputation does not occur in isolation from the Bureau’s other data sources, but in 27 concert with them. With less data drawn from these other sources and less accurate data drawn 28 from these other sources, the quality of the Bureau’s imputation will decline. ATTORNEYS AT LAW SAN FRANCISCO 41 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 43 of 62 1 255. The Rush Plan will also disrupt the post-collection data processing operations, 2 described above. As noted by Secretary Ross and Director Dillingham in mid-April 2020, 3 following Non-Response Follow Up the Bureau engages in “lengthy, thorough and scientifically 4 rigorous” data processing, which is essential to ensuring an accurate census. 5 256. In announcing the new plan to rush the completion of the 2020 Census, Director 6 Dillingham stated that the Bureau would “streamline” these operations in order to meet the 7 December 31, 2020 deadline. 8 257. While the Director has not specified what this “streamlining” means for post- 9 collection operations, the bottom line is that the Bureau cannot fully engage in the operations as 10 contemplated in its Final Operational Plan on the shortened timeframe. As a result, the Bureau 11 will have to cut or reduce its efforts to review and process collected data to ensure accuracy. 12 F. The New “Rush” Plan Will Create Confusion that Plaintiffs Will Be Forced to 13 Spend Time and Money Counteracting 14 258. 15 16 The new plan to rush completion of the 2020 Census also creates additional confusion about census operations at a critical moment in the census-taking process. 259. The Census Bureau’s abrupt change will require groups and local governments 17 engaging in Get Out the Count campaigns, including Plaintiff localities and Plaintiff 18 organizations, to expend resources to correct confusion about the last date for counting in the 19 2020 Census. 20 21 22 260. As noted above, Plaintiff organizations and localities engaged in extensive public information campaigns that publicized the October 31, 2020 deadline. 261. The Rush Plan requires Plaintiffs to expend additional resources in order to 23 update existing public materials, distribute new materials, and engage in more public-facing 24 efforts to educate the public, their constituents, their members and/or constituents, and local 25 organizations that the self-response period for the census ends on September 30, 2020. 26 262. For instance, in Harris County, officials ordered a mailing to constituents 27 informing them that they had until October 31, 2020 to respond to the census. That order 28 occurred before the August 3, 2020 decision to implement the Rush Plan. In light of the new ATTORNEYS AT LAW SAN FRANCISCO 42 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 44 of 62 1 plan, the officials were forced to order stickers to cover the reference to October 31, 2020 on the 2 mailer and to dedicate office staff to spend time affixing those stickers and updating the mailer. 3 Similarly, Plaintiffs the City of Los Angeles, BAJI, and Urban League must update 4 advertisements on social media to correct previous communications that referenced the October 5 31, 2020 deadline. 6 263. Apart from correcting misinterpretations arising from earlier statements Plaintiffs 7 made in reliance on the Bureau’s COVID-19 Plan, Plaintiffs must now also engage in more, 8 unanticipated outreach to educate the public about the Census Bureau’s Rush Plan decision. With 9 one month less of counting, there is now increased urgency for non-responsive households to 10 self-respond. As a result, Plaintiffs are developing new plans to reach more households and 11 encourage more census participation. 12 13 14 15 G. The New “Rush” Plan Will Lead to Undercounting of Minorities 264. The new plan to rush completion of the 2020 Census will exacerbate undercounting of Black, Latino, and Native American communities. 265. As noted above, Non-Response Follow Up, is specifically designed to ensure that 16 traditionally hard-to-count communities, including Black, Latino, and Native American 17 communities are fully counted. By cutting Non-Response Follow Up short, the administration is 18 disrupting the operation most essential to ensuring an accurate count for these communities. 19 266. For the 2020 Census, Black, Latino and Native American populations make up a 20 disproportionate share of the population in tracts with the lowest self-response rates in the United 21 States. For instance, as of July 23, 2020, one in five residents living in census tracts with the 22 lowest self-response rates was Black, and one in four was Hispanic, far larger proportions than 23 Black and Hispanic shares of the general population. 24 25 26 267. Consequently, Black, Latino, and Native American households will make up a disproportionate share of the Non-Response Follow Up universe. 268. Given the challenges of the shortened Non-Response Follow Up timeline, Black, 27 Latino, and Native American households have a high likelihood of being missed, or inaccurately 28 enumerated through administrative records and imputation. As noted above, these alternative ATTORNEYS AT LAW SAN FRANCISCO 43 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 45 of 62 1 2 methods for enumeration will result in lower quality data for these groups. 269. The problem, however, is even more serious because it replicates and exacerbates 3 problems the Census Bureau has found in prior censuses and has striven to correct in subsequent 4 censuses. Data from previous censuses shows that Black, Latino, and Native Americans have 5 historically been undercounted. Over-reliance on alternative methods of data to enumerate a 6 disproportionate share of the population in these groups will further exacerbate potential 7 undercounting in these groups during the 2020 Census. 8 9 270. Accurate data about the size, location, and characteristics of communities of color is necessary to equitably distribute political power through congressional reapportionment and 10 redistricting at the state and local levels, enforce civil-rights laws that affect basic needs like 11 housing and employment, and conduct effective research, including on pressing issues like 12 public health. 13 271. Truncating Non-Response Follow Up will exacerbate undercounts of 14 communities of color in at least two ways: first, by missing members of those communities 15 entirely; or, second, by recording their characteristics incorrectly, such that the census results 16 will not register them as members of communities of color. In either instance, data regarding 17 communities of color will be inaccurate. This inaccuracy then deprives communities of color of 18 federal funding, all the material support that flows from federal funding, the protections of the 19 law, and political power at the federal, state, and local levels. 20 21 H. The New Rush Plan Has No Legitimate Justification 272. In announcing the Rush Plan, Defendants provided no express justification. 22 Defendants stated in passing, however, that reporting of apportionment data to the President by 23 December 31, 2020 is required by statute. 24 273. But there is “nothing sacred in the due date of the filing [of apportionment data], 25 especially when the work of the Census Bureau . . . is incomplete.” Carey v. Klutznick, 637 F. 2d 26 834, 837 (2d Cir. 1980). 27 28 ATTORNEYS AT LAW SAN FRANCISCO 274. The Supreme Court thus determined that the government can and should substitute apportionment counts that have already been filed and certified with “newer, more 44 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 46 of 62 1 2 accurate version[s].” Utah v. Evans, 536 U.S. 452, 462 (2002). 275. Defendants have also recognized that, in the event of a conflict between the two, 3 the constitutional requirement of a fair and accurate enumeration, rather than the statutory 4 deadline, is the controlling legal requirement. With the COVID-19 pandemic threatening the 5 health and safety of communities across the country, Defendants adjusted 2020 Census 6 operations in the COVID-19 Plan, shifting the timeline by several months. Defendants did not 7 wait for Congress to act to implement this plan, recognizing that the Plan was necessary to 8 protect enumerators and respondents, and to ensure an accurate count. 9 276. Because of those delays, as the Bureau itself recognized, it was no longer possible 10 for Defendants to produce data by December 31, 2020 that fulfilled their constitutional and 11 statutory mandate. Specifically, the Bureau could not simultaneously pursue an accurate 2020 12 Census, and speed through completion of census-taking in order to report numbers to the 13 President by the end of the year. 14 277. Several senior officials charged with actually conducting the 2020 Census 15 confirmed the impossibility of this task throughout the summer, including approximately four 16 weeks before Defendants’ abruptly announced their decision to adopt the Rush Plan. 17 278. The statutory deadline at issue is not mandated by the Constitution. Taking the 18 modest additional time necessary to ensure an accurate census, should not prevent a timely 19 reapportionment, as elections for congressional seats impacted by reapportionment will not occur 20 until 2022. 21 279. Ultimately, Defendants cannot sacrifice their mandatory constitutional obligation 22 to make decisions reasonably related to producing an accurate count in order to comply with a 23 pro forma statutory deadline. Congress clearly could not, for instance, satisfy its constitutional 24 obligations by providing the Census Bureau with a single week in which to conduct the census. 25 Strictly adhering to the December 31, 2020 deadline, as applied in extraordinary circumstances 26 of the ongoing pandemic, would be equally unconstitutional. 27 28 ATTORNEYS AT LAW SAN FRANCISCO I. Implementation of the Apportionment Exclusion Order 280. Defendants have not yet sought to justify their motivation for adopting the Rush 45 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 47 of 62 1 Plan, and it cannot be justified on the basis of artificial statutory deadlines. Instead, the timing of 2 the abandonment suggests that the decision was influenced by a desire to implement the 3 President’s Executive Memorandum excluding undocumented immigrants from the 4 apportionment count, thereby undercutting the contribution of communities of color to the 5 calculations for equal representation for purposes of congressional apportionment 6 (the “Apportionment Exclusion Order”). 7 281. In late June 2020, the White House took the unprecedented step of adding two 8 political appointees to Census Bureau staff with unspecified job duties. Neither appointee had an 9 expertise in statistics, and both had a demonstrated history of partisan activity. These unusual 10 appointees had previously engaged with the Census Bureau on questions about changing 11 operations and methodology. 12 282. In mid-July 2020, White House officials reportedly asked congressional 13 appropriators to include $1 billion in the next coronavirus stimulus bill for the purpose of 14 completing the 2020 Census by the December 31, 2020 deadline. 15 283. This abrupt change in policy coincided with and was motivated by the President’s 16 July 21, 2020 issuance of the unconstitutional Apportionment Exclusion Order declaring that it is 17 the policy of the United States to remove undocumented persons from the apportionment count, 18 and requiring the Secretary of Commerce to produce estimates of the number of undocumented 19 persons in the United States when reporting total population counts to the President. As noted, 20 the Apportionment Exclusion Order is currently being challenged as unconstitutional and 21 unlawful in a number of lawsuits filed in jurisdictions around the country, including in this 22 District. 23 284. Shortening the census timeline increases the likelihood that, regardless of the 24 outcome of the November 2020 election, this President will have the opportunity to implement 25 his Apportionment Exclusion Order. Delaying reporting until spring—as the COVID-19 Plan 26 issued by the Census Bureau and Department of Commerce previously did—leaves open the 27 possibility that the President will no longer be in office when data is provided, and thus will be 28 unable to effectuate the Apportionment Exclusion Order. ATTORNEYS AT LAW SAN FRANCISCO 46 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 48 of 62 1 285. Defendants did not justify their sudden, unexplained reversal of position with any 2 evidence that Bureau officials had been wrong in stating, repeatedly, that it would be impossible 3 to produce accurate counts by December 31, 2020. There is also no evidence that the decision to 4 cut short counting operations was driven by the scientifically based judgment of Bureau 5 personnel or external experts. 6 286. To the extent that Defendants’ are motivated by a desire to implement the 7 President’s Apportionment Exclusion Order, that motivation is improper. It bears no reasonable 8 relationship to the achievement of a fair and accurate census, and, under the circumstances 9 currently facing the count, implementing the Apportionment Exclusion Order will undermine 10 11 that goal. 287. Moreover, that Memorandum is just the latest attempt by the President and 12 Secretary Ross to manipulate the census along racial and ethnic lines. Beginning in 2017, 13 Secretary Ross attempted to add an untested citizenship question to the 2020 Census, claiming 14 that the question was necessary to better enforce the Voting Rights Act. In reality, the 15 administration was seeking block-level citizenship data so states could draw district lines in a 16 manner that would disadvantage Black and Latino communities. 17 288. Defendant Ross’s decision was litigated, and enjoined by three district courts. 18 One of those cases ultimately ended up before the Supreme Court. There the Court found that 19 Defendant Ross’s stated Voting Rights Act rationale to support the addition of a citizenship 20 question to the 2020 Census was “contrived” and vacated Defendant Ross’s decision. Dep’t of 21 Commerce v. New York, 139 S. Ct. 2551, 2575-76 (2019). 22 289. On July 5, 2019, following the Supreme Court’s decision, President Trump 23 confirmed the real rationale—and fully justified the Supreme Court’s holding that the 24 administration’s rationale for this census decision was pretextual—when he stated that the 25 administration sought a citizenship question, not to enforce the Voting Rights Act, but rather “for 26 districting” and “for appropriations.” Remarks by President Trump Before Marine One 27 Departure (July 5, 2019), https://www.whitehouse.gov/briefings-statements/remarks-president- 28 trump-marine-one-departure-51/. ATTORNEYS AT LAW SAN FRANCISCO 47 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 49 of 62 290. 1 Indeed, further evidence that Defendants’ actions were pretextual arose from files 2 of a prominent redistricting strategist, Thomas Hofeller. In 2015, Hofeller prepared a study titled 3 “The Use of Citizen Voting Age Population in Redistricting.” In the study, Hofeller 4 recommended adding a citizenship question to the census so that states could use citizen voting- 5 age population rather than total population to redistrict. This change in the redistricting base, in 6 Hofeller’s words, would be advantageous to “Non-Hispanic Whites” and would undercut the 7 political power of Hispanics. 291. 8 9 It was later revealed that Hofeller was involved in drafting portions of the memorandum from the Department of Justice to Defendant Commerce seeking addition of a 10 citizenship question on the 2020 Census, including sections relating to the pretextual reason for 11 requesting the question. See Ex. 8 to NYIC Pls.’ Mot. for Sanctions at 124-31, New York v. U.S. 12 Dep’t of Commerce, No. 1:18-cv-2921-JMF (S.D.N.Y. July 16, 2019), ECF No. 635-1; Defs.’ 13 Opp. to Letter Mot. to Compel at 3, New York v. U.S. Dep’t of Commerce, No. 1:18-cv-2921- 14 JMF (S.D.N.Y. Oct. 30, 2018), ECF No. 451. 292. 15 Shortly after the Supreme Court’s decision, President Trump issued an executive 16 order, demanding executive agencies provide the Census Bureau with administrative records 17 sufficient to allow the Bureau to determine “the number of citizens and noncitizens in the 18 country.” Exec. Order No. 13,880, § 1, 84 Fed. Reg. 33,821, 33,821 (July 16, 2019). The 19 Executive Order explicitly states that the reason this data is necessary is to design “legislative 20 districts based on the population of voter-eligible citizens,” instead of total population. Id. at 21 33,823-84. 293. 22 In light of that history, the Apportionment Exclusion Order, and the near- 23 contemporaneous decision to cut counting operations short represent yet another attempt by the 24 administration to manipulate the 2020 Census and potentially undercut the political power of 25 communities of color. Defendants cannot rely on this memorandum as justification to support 26 their decision to undermine the accuracy of the census. 27 V. 28 ATTORNEYS AT LAW SAN FRANCISCO Harm to Plaintiffs. 294. Plaintiffs and Plaintiff non-profits’ members and/or constituents reside in locales 48 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 50 of 62 1 that will suffer harm as a result of Defendants’ decision because that decision is very likely to 2 cause these locales to be more disproportionately undercounted in the 2020 Census than they 3 otherwise would have been. 4 295. On August 9, 2020, at the beginning of the Non-Response Follow Up operation, 5 Plaintiff City of Los Angeles, had a response rate of just 53.1%, which was significantly lower 6 than the 64.5% statewide response rate in California on that same date. 7 296. The Urban League, League of Women Voters, and BAJI have affiliates, 8 constituents, and members in major cities across the United States. This includes cities where 9 response rates were lower than their corresponding statewide response rates on the first day of 10 Non-Response Follow Up including San Francisco (61.4%) and Monterey (60.5%) as compared 11 to California (64.5%), Miami (49.6%) as compared to Florida (60.1%), Philadelphia (52%) as 12 compared to Pennsylvania (65.5%), Detroit (48.7%) as compared to Michigan (68.9%), and New 13 York City (54.9%) as compared to New York State (58.9%). 14 297. Plaintiffs Ellis and Garcia are residents of Houston, Texas. The response rate in 15 Houston at the beginning of Non-Response Follow Up was 54%, which was lower than the 16 statewide response rate for Texas on that date, 58.2%. 17 298. As noted above, Defendants’ decision will result in fewer enumerations through 18 Non-Response Follow Up, increased reliance on low-quality administrative data, and increased 19 imputation. Consequently, Defendants’ decision will result in cities’ with higher rates of non- 20 response (1) having less accurate data; and (2) experiencing higher rates of undercounting. 21 299. Because these cities have a higher proportion of households in the Non-Response 22 Follow Up universe than their corresponding states, these cities have a substantially higher 23 likelihood of being undercounted because of Defendants’ decision than surrounding communities 24 in their states. These disproportionate undercounts will cause Plaintiffs to suffer both fiscal and 25 representational harm. 26 A. Funding Harms 27 300. The Rush Plan will result in loss of federal funding for Plaintiffs Harris County, 28 ATTORNEYS AT LAW SAN FRANCISCO City of Salinas, and the City of Los Angeles and the communities where members of Plaintiff 49 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 51 of 62 1 2 non-profits reside, including Miami, Detroit, Philadelphia and New York. 301. Over 130 programs and 675 billion dollars are allocated to states and localities on 3 the basis of census-derived information. This includes funding to states for federal transportation 4 planning purposes, education, and healthcare. 5 302. Many important federal programs, including Title I Grants under the Every 6 Student Succeeds Act, require states to distribute funds to localities on the basis of census- 7 derived information. 8 303. 9 10 State Education Agencies must allocate Title I Grants, at least in part, on the number of children aged 5-17 living in poverty in a local education agency’s jurisdiction. 304. Given that members of Plaintiff non-profits reside in cities that are likely to be 11 more undercounted under the Rush Plan relative to surrounding communities in their states, 12 including San Francisco, Miami, Detroit, Philadelphia, and New York City, Defendants’ decision 13 will likely deprive the communities where these members reside of Title I Grant funding they 14 would have otherwise received. Similarly, Defendants’ decision places Plaintiffs Ellis and 15 Garcia’s community at higher risk of deprivation of Title I Grant funding. 16 305. Several additional federal programs require states to use census-derived 17 information to distribute funds directly to cities and counties, based on their share of a relevant 18 population. For instance, the Low Income Home Energy Assistance Program, the Workforce 19 Innovation and Opportunity Act program, and the Community Services Block Grant Program, all 20 require states to distribute funds to cities and counties, at least in part, on the proportion of a 21 state’s low-income residents living in those cities and counties. This data is derived from 22 information collected during the decennial census. 23 306. Both Harris County and the City of Los Angeles receive funds under these 24 programs. Consequently, disproportionate undercounting of Harris County and the City of Los 25 Angeles, as compared to their states, is likely to result in loss of funds under these and similar 26 programs. 27 28 ATTORNEYS AT LAW SAN FRANCISCO 307. Several federal funding programs provide funding directly to cities and counties based on census-derived information. For instance, the Community Development Block Grant 50 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 52 of 62 1 program, and the Emergency Solutions Grant, allocate funding to cities and counties based, at 2 least in part, on their share of the overall population count relative to other metropolitan areas. 3 308. Of cities with over 500,000 people, the City of Los Angeles had the fourth lowest 4 response rate in the country, just behind Detroit and Philadelphia. Consequently, Los Angeles 5 will likely lose Community Development Block Grant funds because of Defendants’ decision. 6 309. Similarly, members of Plaintiff non-profits live in major metropolitan areas with 7 some of the lowest response rates in the country, such as Miami, Detroit and Philadelphia. 8 Defendants’ decision will likely deprive these members’ communities of funding under the 9 Community Development Block Grant program. 10 310. Finally, the allocation of federal transportation including the Surface 11 Transportation Block Grant Program, and the Metropolitan and Statewide Nonmetropolitan 12 Transportation Planning Programs are based on the population of urbanized areas in a state 13 compared to those of other states, as determined by the decennial census. 14 311. Plaintiffs Ellis and Garcia regularly drive on highways and roads in Texas. 15 Disproportionate undercounting of urbanized areas in Texas during the 2020 Census will result 16 in reduced transportation funding for Texas under federal transportation programs. 17 18 19 20 B. Representational Harm 312. Defendants’ decision will also likely result in representational harm to individual Plaintiffs and to the members of Plaintiff organizations. 313. Plaintiffs Ellis and Garcia reside in Houston, Texas. In terms of self-response 21 rates, Texas ranks 39th in the United States. Approximately four million Texas households are in 22 the Non-Response Follow Up universe, which is more households than any state other than 23 California. 24 314. Consequently, Defendants’ decision will not only cause a substantial undercount 25 in Texas, but that undercount will likely be disproportionate as compared to other states. Texas 26 will likely be deprived of its fair share of representation in the next congressional apportionment. 27 28 ATTORNEYS AT LAW SAN FRANCISCO 315. As a result, Defendants’ decision is likely to result in reduction of voting power and representation for Plaintiffs Ellis and Garcia, because it will likely cause the loss of a seat in 51 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 53 of 62 1 2 Texas, and will result in fewer Representatives spread out over the state of Texas. 316. As for Plaintiff City of Los Angeles, at least one study has predicted that, were 3 California to lose a congressional seat because of the final census count, that seat is very likely to 4 come from a district that includes portions of South Los Angeles, thus reducing the city’s 5 representational delegation. 6 7 8 9 317. Defendants’ decision will also cause Plaintiff Ellis and members of Plaintiff non- profits to experience a loss of intrastate voting power. 318. By causing disproportionate undercounting of communities in Houston, Detroit, Philadelphia, and Miami, as compared to their corresponding states, Defendants’ decision will 10 result in drawing of district lines that do not accurately represent the population of the state, and 11 disadvantage Plaintiffs Ellis and Garcia, and members of Plaintiff organizations that live in 12 undercounted communities. 13 C. Inaccurate Data 14 319. Plaintiff local governments will suffer harm from the adverse impact Defendants’ 15 decision will have the accuracy of population counts produced by the Census Bureau. Plaintiff 16 local governments often rely on accurate information collected by the Census Bureau for crucial 17 public planning purposes, including planning for how to respond to emergencies. 18 320. For example, local governments often rely on a Social Vulnerability Index to 19 identify communities that are at high risk during a particular emergency. Government officials 20 rely on this index to determine where to allocate resources before and during emergencies. A 21 Social Vulnerability Index use census data to identify specific populations that may be 22 vulnerable to a particular emergency, including data relating to age, housing density, income 23 status, and race and ethnicity. Inaccurate census data would make disaster planning and 24 emergency response more difficult, and could disrupt important public programs. 25 321. In Harris County, officials used the Center for Disease Control’s Social 26 Vulnerability Index to inform decisions about proper distribution of COVID-19 Relief Funds. 27 The funds were allocated to provide relief to Harris County residents most impacted by the 28 global pandemic. That Social Vulnerability Index, which was based on census data, was used to ATTORNEYS AT LAW SAN FRANCISCO 52 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 54 of 62 1 identify census tracts with the most vulnerable residents, and applications from residents from 2 those tracts were prioritized and given higher chances of acceptance for funds. Without accurate 3 census data, Harris County would struggle to ensure that crucial relief funds were reaching the 4 communities most in need of them. 5 322. Similarly, King County relies on accurate census data to inform its public-policy 6 decision making. For instance, the county uses census data to plan public-transit service, and to 7 ensure priority populations have transit access, and to site public health clinics. 8 9 323. The low-quality data and undercounting that Defendants’ decision will cause will also harm Plaintiffs. For instance, undercounting of Black, Latino, Native American, and 10 immigrant communities will negatively affect the Urban League, League of Women Voters and 11 BAJI by undermining these organizations’ core missions of promoting equal and just laws and 12 empowering vulnerable communities through building coalitions and initiating campaigns with 13 African Americans and Black immigrants, and fostering racial, economic, and social equality for 14 the communities they serve. 15 16 D. Expending Additional Resources 324. Plaintiff organizations, the Urban League, the League of Women Voters, and 17 BAJI, and Plaintiff local governments, City of San Jose, Harris County, King County, City of 18 Salinas, and City of Los Angeles will need to expend additional resources and divert resources 19 from planned programs and projects in order to address the adverse consequences of Defendants’ 20 decision to abandon the COVID-19 Plan, and implement the Rush Plan. 21 325. Plaintiffs’ planned efforts to ensure the effective enumeration of historically 22 undercounted communities were based on the understanding that the Census Bureau would 23 implement the Non-Response Follow Up operation contemplated in the Final Operational Plan 24 and adjusted in the COVID-19 Plan. 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO 326. The abrupt reversal of the COVID-19 Plan, and the implementation of curtailed Non-Response Follow Up in the Bureau’s Rush Plan will adversely affect Plaintiffs’ plans. 327. Plaintiff organizations and local governments will likely need to adjust plans, and divert resources from other planned activities and programs in order to ensure the communities 53 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 55 of 62 1 they serve are adequately counted. Specifically, Plaintiffs will need to recruit and train staff to 2 engage in increased and expanded outreach to potential non-responsive households in order to 3 make up for fewer enumerator visits, or to other aspects of the Non-Response Follow Up 4 program, such as the reinterview process. 328. 5 For instance, Plaintiff BAJI is planning significant adjustments to its 2020 Census 6 outreach plans in light of Defendants’ decision, that include diversion of resources from other 7 sources, and significant expenditures. In order to engage in effective outreach, BAJI needs 8 organizing staff dedicated to civic engagement. With Non-Response Follow Up occurring from 9 August 11, 2020 through October 31, 2020, BAJI anticipated that it could spread its staffing 10 resources over that timeframe to ensure it was meeting its goals within the organization’s budget. 11 However, on a shorter timeframe, BAJI needs additional staff on a shorter timeframe, which will 12 require adjusting the organization’s budget and priorities for the next several months. 329. 13 The adjustment is also challenging for BAJI as the organization caters to 14 immigrant communities with a variety of language needs. Increasing staffing on a short 15 timeframe poses significant challenges for the organization, because it must locate staff that can 16 communicate with the particular community that the organization is targeting for outreach 17 efforts. CLAIMS FOR RELIEF 18 FIRST CLAIM FOR RELIEF Violation of the Enumeration Clause, and Fourteenth Amendment (U.S. Const. art. I, § 2; U.S. Const. amend. XIV, § 2) 19 20 21 330. 22 paragraphs. 23 331. Plaintiffs incorporate by reference the allegations set forth in the preceding Under the Enumeration Clause of the U.S. Constitution, Congress, and, by 24 delegation, the Secretary of Commerce, must conduct an “actual Enumeration” of the population. 25 This clause requires that decisions relating to census-taking “bear a reasonable relationship to the 26 accomplishment of an actual enumeration of the population.” Wisconsin v. City of N.Y., 517 U.S. 27 1, 20 (1996). 28 ATTORNEYS AT LAW SAN FRANCISCO 332. The COVID-19 pandemic severely disrupted the 2020 Census, resulting in 54 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 56 of 62 1 months of suspended operations and significant delays in crucial counting processes. Moreover, 2 the public-health crisis continues to impact census operations, as the Bureau struggles to retain 3 enumerators and engage in door-knocking in communities experiencing surges of the virus. 4 333. To navigate this emergency, the Bureau took necessary action to adjust its 5 operational timelines in the COVID-19 Plan while seeking to maintain the operations and 6 processes included in the Final Operational Plan that had been designed to help ensure a 7 complete and accurate count. 8 9 334. Abruptly and without explanation, on August 3, 2020, Defendants abandoned the COVID-19 Plan and implemented the Rush Plan. The Rush Plan does not “bear a reasonable 10 relationship to the accomplishment of an actual enumeration of the population.” After delaying 11 all operations for months, the Bureau and its staff repeatedly recognized that it was impossible to 12 produce counts consistent with their duties to ensure a full, fair, and accurate count by December 13 31, 2020. Indeed, current conditions demonstrate that it is infeasible to obtain a fair and accurate 14 count by the end of the year. Nevertheless, the Defendants abandoned their constitutionally 15 mandated pursuit of fair and accurate data, in favor of the speed of the Rush Plan, and the 16 inaccurate data it will produce. 17 18 19 335. Under these circumstances, the decision to curtail crucial 2020 Census operations violates the Enumeration Clause of the United States Constitution. 336. These constitutional violations have caused, are causing, and will continue to 20 cause harm to Plaintiffs as alleged above, and there is a substantial likelihood that the requested 21 relief will redress this harm. 22 23 SECOND CLAIM FOR RELIEF Violation of Administrative Procedure Act—Arbitrary and Capricious (5 U.S.C. § 706(2)(A)) 24 337. 25 paragraphs. 26 338. Plaintiffs incorporate by reference the allegations set forth in the preceding The APA, 5 U.S.C. § 706(2), provides that a court shall hold unlawful and set 27 aside agency action found to be arbitrary, capricious, an abuse of discretion, or otherwise not in 28 accordance with law. The Rush Plan is final agency action because it marks the consummation ATTORNEYS AT LAW SAN FRANCISCO 55 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 57 of 62 1 of the agency’s decision-making process, and it is one by which rights or obligations have been 2 determined, or from which legal consequences will flow. Bennett v. Spear, 520 U.S. 154, 177-78 3 (1997). 339. 4 In determining whether an action violates the APA, courts consider whether the 5 agency examined relevant data and articulated a satisfactory explanation for its decision, 6 including formulating a rational connection between the facts found and the choice made. Motor 7 Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983). 8 Where an agency wishes to depart from an earlier decision, it must acknowledge that change and 9 any reliance interests its previous actions engendered. See Perez v. Mortg. Bankers Ass’n, 575 10 U.S. 92, 105-06 (2015). 340. 11 The Bureau spent several years developing its Final Operational Plan for the 2020 12 Census. That plan carefully determined the required length of each operation, including the 13 appropriate length for data-collection and data-processing. It also included details about the 14 implementation of the various operations. 341. 15 The COVID-19 pandemic disrupted census operations, and the Bureau responded 16 by adjusting its operations in its COVID-19 Plan. That plan involved retaining the details and the 17 length of time of various operations laid out in the Final Operational Plan, but shifting the 18 timeline for counting several months into the future to account for both the necessity of those 19 operations and the public-health emergency. 342. 20 The Bureau began implementing the plan, and critical operations were suspended 21 and delayed through the summer. Bureau officials publicly and expressly recognized that it was 22 no longer possible to comply with the December 31, 2020 deadline if the Bureau intended to 23 fulfill its constitutional and statutory obligation of producing reasonably accurate population 24 counts. 25 343. Without explanation and without citing any evidence, Defendants suddenly 26 changed their position and issued a new plan with shortened timelines. Among other things, that 27 change conclusively changed the legal rights and obligations of private households, who now 28 have substantially less time to respond if they wish to be counted in the 2020 Census. Defendants ATTORNEYS AT LAW SAN FRANCISCO 56 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 58 of 62 1 have provided no evidence to support rescinding the COVID-19 Plan, have failed to 2 acknowledge or explain their departure from their previous conclusions as to the length of time 3 necessary for an accurate census, and have cited no evidence that they could obtain accurate 4 counts on the shortened timeframe. Defendants’ unexplained and unjustifiable reversal is 5 precisely the sort of arbitrary and capricious agency action that the Administrative Procedure Act 6 forbids. 7 344. Defendants’ decision also fails to account for several factors relevant to the 8 decision, including the multiple-month long suspension in operations and delay of crucial census 9 operations, the staffing shortages facing the Bureau, the meticulously designed and tested 10 technical requirements for effective enumeration included in the Bureau’s Final Operational 11 Plan, and the various quality-control measures the Bureau must engage in to ensure that its 12 reported data is accurate. 13 345. Consequently, Defendants’ action is arbitrary and capricious. 14 346. This unlawful action has caused, is causing, and will continue to cause harm to 15 Plaintiffs as alleged above, and there is a substantial likelihood that the requested relief will 16 redress this harm. THIRD CLAIM FOR RELIEF Violation of Administrative Procedure Act—Pretext (5 U.S.C. § 706) 17 18 19 347. 20 paragraphs. 21 348. Plaintiffs incorporate by reference the allegations set forth in the preceding Under the Administrative Procedure Act, agencies are required to disclose the 22 “genuine justification[] for important decisions.” Dep’t of Commerce, 139 S. Ct. at 2569, 2575- 23 76. Courts will not accept “contrived reasons” provided by agencies as that would defeat the 24 purpose of judicial review. Id. at 2576. Moreover, agencies cannot simply avoid providing 25 reasoning for their decision-making altogether. 26 349. Defendants have decided to cut crucial operations in order to produce 2020 27 Census population results to the President by December 31, 2020. In announcing that decision, 28 Defendants provided no legitimate justification for abandoning the COVID-19 Plan and ATTORNEYS AT LAW SAN FRANCISCO 57 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 59 of 62 1 2 3 4 implementing the Rush Plan. 350. Any attempt by the Defendants to rely on the reporting deadline provided under the Census Act as justification for their decision is mere pretext. 13 U.S.C § 141(b). 351. For months, Defendants implemented the COVID-19 Plan, the timeline for which 5 necessarily assumed the statutory deadlines could not defeat the constitutional duty to conduct an 6 accurate enumeration, as applied to the extraordinary circumstances at hand. Defendants made 7 significant adjustments, including months-long delays of census operations, on the assumption 8 that the Bureau could and would conduct a full and robust count through the end of October 31, 9 2020. Since mid-April 2020, Defendants have expressly and publicly recognized that the Bureau 10 could not provide a complete and accurate count by December 31, 2020. And President Trump 11 maintained that the statutory deadlines need not be followed. 12 352. Defendants’ reversal of position on the 2020 Census timeline appears driven by 13 Defendants’ efforts to ensure implementation of the President’s unconstitutional Apportionment 14 Exclusion Order, which attempts to exclude undocumented persons from the apportionment 15 count and continues a long-running pattern of racially discriminatory and improperly politically 16 motivated conduct of the 2020 Census. 17 18 19 353. In light of these considerations, Defendants’ purported justification is pretextual and, thus, arbitrary and capricious under the Administrative Procedure Act. 354. Defendants’ unlawful action has caused, is causing, and will continue to cause 20 harm to Plaintiffs as alleged above, and there is a substantial likelihood that the requested relief 21 will redress this harm. PRAYER FOR RELIEF 22 23 355. Plaintiffs respectfully request that this Court: 24 356. Declare that Defendants’ promulgation of the Rush Plan, and corresponding 25 revocation of the COVID-19 Plan is unconstitutional under the Enumeration Clause, and 26 unlawful under the Administrative Procedure Act. 27 357. Vacate the Rush Plan, thereby reinstating the COVID-19 Plan. 28 358. Enjoin Defendants from implementing the Rush Plan or otherwise unlawfully ATTORNEYS AT LAW SAN FRANCISCO 58 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 60 of 62 1 interfering with the COVID-19 Plan. 2 359. Award Plaintiffs costs, expenses, and reasonable attorneys’ fees. 3 360. Award any other relief the Court deems just and proper. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Dated: August 18, 2020 LATHAM & WATKINS LLP By: /s/ Sadik Huseny Steven M. Bauer (Bar No. 135067) Sadik Huseny (Bar No. 224659) Shannon D. Lankenau (Bar. No. 294263) 505 Montgomery Street, Suite 2000 San Francisco, CA 94111 Telephone: 415.391.0600 Facsimile: 415.395.8095 Richard P. Bress (pro hac vice pending) Melissa Arbus Sherry (pro hac vice pending) Anne W. Robinson (pro hac vice pending) Tyce R. Walters (pro hac vice pending) Genevieve P. Hoffman (pro hac vice pending) Gemma Donofrio (pro hac vice pending) 555 Eleventh Street NW, Suite 1000 Washington, D.C. 20004 Telephone: 202.637.2200 Facsimile: 202.637.2201 Attorneys for Plaintiffs National Urban League; City of San Jose, California; Harris County, Texas; League of Women Voters; King County, Washington; Black Alliance for Just Immigration; Rodney Ellis and Adrian Garcia 21 22 23 24 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO 59 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 61 of 62 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 LAWYERS’ COMMITTEE FOR CIVIL RIGHTS UNDER LAW CITY ATTORNEY FOR THE CITY OF LOS ANGELES By: /s/ Jon M. Greenbaum Kristen Clarke (pro hac vice pending) Jon M. Greenbaum (Bar No. 166733) Ezra D. Rosenberg (pro hac vice pending) Dorian L. Spence (pro hac vice pending) Ajay P. Saini (pro hac vice pending) Maryum Jordan (Bar No. 325447) Pooja Chaudhuri (Bar No. 314847) 1500 K Street NW, Suite 900 Washington, DC 20005 Telephone: 202.662.8600 Facsimile: 202.783.0857 By: /s/ Danielle Goldstein Michael N. Feuer (Bar No. 111529) mike.feuer@lacity.org Kathleen Kenealy (Bar No. 212289) kathleen.kenealy@lacity.org Danielle Goldstein (Bar No. 257486) danielle.goldstein@lacity.org Michael Dundas (Bar No. 226930) mike.dundas@lacity.org 200 N. Main Street, 8th Floor Los Angeles, CA 90012 Telephone: 213.473.3231 Facsimile: 213.978.8312 Attorneys for Plaintiffs National Urban League; City of San Jose, California; Harris Attorneys for Plaintiff City of Los Angeles County, Texas; League of Women Voters; King County, Washington; Black Alliance for CITY OF SALINAS Just Immigration; Rodney Ellis, and Adrian By: /s/ Christopher A. Callihan Garcia Christopher A. Callihan (Bar No. 203010) legalwebmail@ci.salinas.ca.us BRENNAN CENTER FOR JUSTICE Michael Mutalipassi (Bar No. 274858) michaelmu@ci.salinas.ca.us By: /s/ Wendy R. Weiser 200 Lincoln Avenue Wendy R. Weiser (pro hac vice pending) Salinas, CA 93901 weiserw@brennan.law.nyu.edu Telephone: 831.758.7256 Thomas P. Wolf (pro hac vice pending) Facsimile: 831.758.7257 wolf@brennan.law.nyu.edu Kelly M. Percival (pro hac vice pending) percivalk@brennan.law.nyu.edu Attorneys for Plaintiff City of Salinas 120 Broadway, Suite 1750 New York, NY 10271 Telephone: 646.292.8310 PUBLIC COUNSEL Facsimile: 212.463.7308 By: /s/ Mark Rosenbaum Attorneys for Plaintiffs National Urban Mark Rosenbaum (Bar No. 59940) League; City of San Jose, California; Harris mrosenbaum@publiccounsel.org County, Texas; League of Women Voters; 610 South Ardmore Avenue King County, Washington; Black Alliance for Los Angeles, CA 90005 Just Immigration; Rodney Ellis, and Adrian Telephone: 213.385.2977 Garcia Facsimile: 213.385.9089 Attorneys for Plaintiff City of San Jose 28 ATTORNEYS AT LAW SAN FRANCISCO 60 COMPLAINT Case 5:20-cv-05799 Document 1 Filed 08/18/20 Page 62 of 62 1 ATTESTATION 2 I, Sadik Huseny, am the ECF user whose user ID and password authorized the filing of this 3 document. Under Civil L.R. 5-1(i)(3), I attest that all signatories to this document have concurred 4 in this filing. 5 6 7 8 Dated: August 18, 2020 LATHAM & WATKINS LLP By: /s/ Sadik Huseny Sadik Huseny 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 ATTORNEYS AT LAW SAN FRANCISCO 61 COMPLAINT Case 5:20-cv-05799 Document 1-1 Filed 08/18/20 Page 1 of 1 -6 &$1' 5HY &,9,/ &29(5 6+((7 7KH -6 &$1' FLYLO FRYHU VKHHW DQG WKH LQIRUPDWLRQ FRQWDLQHG KHUHLQ QHLWKHU UHSODFH QRU VXSSOHPHQW WKH ILOLQJ DQG VHUYLFH RI SOHDGLQJV RU RWKHU SDSHUV DV UHTXLUHG E\ ODZ H[FHSW DV SURYLGHG E\ ORFDO UXOHV RI FRXUW 7KLV IRUP DSSURYHG LQ LWV RULJLQDO IRUP E\ WKH -XGLFLDO &RQIHUHQFH RI WKH 8QLWHG 6WDWHV LQ 6HSWHPEHU LV UHTXLUHG IRU WKH &OHUN RI &RXUW WR LQLWLDWH WKH FLYLO GRFNHW VKHHW (SEE INSTRUCTIONS ON NEXT PAGE OF THIS FORM.) , NATIONAL D 3/$,17,))6 URBAN LEAGUE; BLACK ALLIANCE FOR JUST IMMIGRATION; LEAGUE OF WOMEN VOTERS; HARRIS COUNTY, TEXAS; KING COUNTY, WASHINGTON; CITY OF LOS ANGELES, CALIFORNIA; CITY OF SALINAS, CALIFORNIA; and CITY OF SAN JOSE, CALIFORNIA; RODNEY ELLIS, and ADRIAN GARCIA E &RXQW\ RI 5HVLGHQFH RI )LUVW /LVWHG 3ODLQWLII '()(1'$176 WILBUR L.ROSS, JR., in his official capacity as Secretary of Commerce; U.S. DEPARTMENT OF COMMERCE, U.S. CENSUS BUREAU, and STEVEN DILLINGHAM, in his official capacity as Director of the U.S. Census Bureau &RXQW\ RI 5HVLGHQFH RI )LUVW /LVWHG 'HIHQGDQW New York City, NY (IN U.S. PLAINTIFF CASES ONLY) (EXCEPT IN U.S. PLAINTIFF CASES) Washington County, D.C. 127( ,1 /$1' &21'(01$7,21 &$6(6 86( 7+( /2&$7,21 2) 7+( 75$&7 2) /$1' ,192/9(' F $WWRUQH\V (If Known) $WWRUQH\V (Firm Name, Address, and Telephone Number) Steven M. Bauer / Sadik Huseny Latham & Watkins LLP 505 Montgomery Street, Suite 2000, San Francisco, CA 94111-6538 ,, %$6,6 2) -85,6',&7,21 (Place an “X” in One Box Only) 8 6 *RYHUQPHQW 3ODLQWLII )HGHUDO 4XHVWLRQ (U.S. Government Not a Party) 8 6 *RYHUQPHQW 'HIHQGDQW 'LYHUVLW\ (Indicate Citizenship of Parties in Item III) ,9 &LWL]HQ RI 7KLV 6WDWH &LWL]HQ RI $QRWKHU 6WDWH &LWL]HQ RU 6XEMHFW RI D )RUHLJQ &RXQWU\ ,QFRUSRUDWHG or 3ULQFLSDO 3ODFH RI %XVLQHVV ,Q 7KLV 6WDWH ,QFRUSRUDWHG and 3ULQFLSDO 3ODFH RI %XVLQHVV ,Q $QRWKHU 6WDWH )RUHLJQ 1DWLRQ 1$785( 2) 68,7 (Place an “X” in One Box Only) &2175$&7 ,QVXUDQFH 0DULQH 0LOOHU $FW 1HJRWLDEOH ,QVWUXPHQW 5HFRYHU\ RI 2YHUSD\PHQW 2I 9HWHUDQ¶V %HQHILWV 0HGLFDUH $FW 5HFRYHU\ RI 'HIDXOWHG 6WXGHQW /RDQV ([FOXGHV 9HWHUDQV 5HFRYHU\ RI 2YHUSD\PHQW RI 9HWHUDQ¶V %HQHILWV 6WRFNKROGHUV¶ 6XLWV 2WKHU &RQWUDFW &RQWUDFW 3URGXFW /LDELOLW\ )UDQFKLVH 5($/ 3523(57< /DQG &RQGHPQDWLRQ )RUHFORVXUH 5HQW /HDVH (MHFWPHQW 7RUWV WR /DQG 7RUW 3URGXFW /LDELOLW\ $OO 2WKHU 5HDO 3URSHUW\ 9 9, 9,, 72576 3(5621$/ ,1-85< $LUSODQH $LUSODQH 3URGXFW /LDELOLW\ $VVDXOW /LEHO 6ODQGHU )HGHUDO (PSOR\HUV¶ /LDELOLW\ 0DULQH 0DULQH 3URGXFW /LDELOLW\ 0RWRU 9HKLFOH 0RWRU 9HKLFOH 3URGXFW /LDELOLW\ 2WKHU 3HUVRQDO ,QMXU\ 3HUVRQDO ,QMXU\ 0HGLFDO 0DOSUDFWLFH &$86( 2) $&7,21 5HPRYHG IURP 6WDWH &RXUW 3(5621$/ ,1-85< 3HUVRQDO ,QMXU\ ± 3URGXFW /LDELOLW\ +HDOWK &DUH 3KDUPDFHXWLFDO 3HUVRQDO ,QMXU\ 3URGXFW /LDELOLW\ $VEHVWRV 3HUVRQDO ,QMXU\ 3URGXFW /LDELOLW\ 3(5621$/ 3523(57< 2WKHU )UDXG 7UXWK LQ /HQGLQJ 2WKHU 3HUVRQDO 3URSHUW\ 'DPDJH 3URSHUW\ 'DPDJH 3URGXFW /LDELOLW\ 35,621(5 3(7,7,216 &,9,/ 5,*+76 2WKHU &LYLO 5LJKWV 9RWLQJ (PSOR\PHQW +RXVLQJ $FFRPPRGDWLRQV $PHU Z 'LVDELOLWLHV± (PSOR\PHQW $PHU Z 'LVDELOLWLHV±2WKHU (GXFDWLRQ 25,*,1 (Place an “X” in One Box Only) 2ULJLQDO 3URFHHGLQJ +$%($6 &25386 $OLHQ 'HWDLQHH 0RWLRQV WR 9DFDWH 6HQWHQFH *HQHUDO 'HDWK 3HQDOW\ %$1.5837&< $SSHDO 86& † :LWKGUDZDO 86& † /$%25 3523(57< 5,*+76 )DLU /DERU 6WDQGDUGV $FW /DERU 0DQDJHPHQW 5HODWLRQV 5DLOZD\ /DERU $FW )DPLO\ DQG 0HGLFDO /HDYH $FW 2WKHU /DERU /LWLJDWLRQ (PSOR\HH 5HWLUHPHQW ,QFRPH 6HFXULW\ $FW &RS\ULJKWV 3DWHQW 3DWHQWņ$EEUHYLDWHG 1HZ 'UXJ $SSOLFDWLRQ 7UDGHPDUN 62&,$/ 6(&85,7< +,$ II %ODFN /XQJ ',:& ',:: J 66,' 7LWOH ;9, 56, J ,00,*5$7,21 1DWXUDOL]DWLRQ $SSOLFDWLRQ 2WKHU ,PPLJUDWLRQ $FWLRQV )('(5$/ 7$; 68,76 7D[HV 8 6 3ODLQWLII RU 'HIHQGDQW ,56±7KLUG 3DUW\ 86& † 27+(5 0DQGDPXV 2WKHU &LYLO 5LJKWV 3ULVRQ &RQGLWLRQ &LYLO 'HWDLQHH± &RQGLWLRQV RI &RQILQHPHQW 5HPDQGHG IURP $SSHOODWH &RXUW 5HLQVWDWHG RU 5HRSHQHG 7UDQVIHUUHG IURP $QRWKHU 'LVWULFW (specify) 27+(5 67$787(6 )DOVH &ODLPV $FW 4XL 7DP 86& † D 6WDWH 5HDSSRUWLRQPHQW $QWLWUXVW %DQNV DQG %DQNLQJ &RPPHUFH 'HSRUWDWLRQ 5DFNHWHHU ,QIOXHQFHG &RUUXSW 2UJDQL]DWLRQV &RQVXPHU &UHGLW 7HOHSKRQH &RQVXPHU 3URWHFWLRQ $FW &DEOH 6DW 79 6HFXULWLHV &RPPRGLWLHV ([FKDQJH 2WKHU 6WDWXWRU\ $FWLRQV $JULFXOWXUDO $FWV (QYLURQPHQWDO 0DWWHUV )UHHGRP RI ,QIRUPDWLRQ $FW $UELWUDWLRQ $GPLQLVWUDWLYH 3URFHGXUH $FW 5HYLHZ RU $SSHDO RI $JHQF\ 'HFLVLRQ &RQVWLWXWLRQDOLW\ RI 6WDWH 6WDWXWHV 0XOWLGLVWULFW /LWLJDWLRQ±7UDQVIHU 0XOWLGLVWULFW /LWLJDWLRQ±'LUHFW )LOH Administrative Procedure Act, 5 U.S.C. 701 et seq. %ULHI GHVFULSWLRQ RI FDXVH Action to set aside unlawful agency action as arbitrary, capricious, and not in accordance with law, and declare Defendants’ actions unconstitutional. 5(48(67(' ,1 &203/$,17 &+(&. ,) 7+,6 ,6 $ &/$66 $&7,21 81'(5 58/( )HG 5 &LY 3 -8'*( '(0$1' Judge Lucy H. Koh ',9,6,21$/ $66,*10(17 &LYLO /RFDO 5XOH 6$1 )5$1&,6&2 2$./$1' 3ODFH DQ ³;´ LQ 2QH %R[ 2QO\ '$7( 08/18/2020 )25)(,785( 3(1$/7< 'UXJ 5HODWHG 6HL]XUH RI 3URSHUW\ 86& † 2WKHU &LWH WKH 8 6 &LYLO 6WDWXWH XQGHU ZKLFK \RX DUH ILOLQJ (Do not cite jurisdictional statutes unless diversity) 9,,, 5(/$7(' &$6( 6 ,) $1< (See instructions): ,; ,,, &,7,=(16+,3 2) 35,1&,3$/ 3$57,(6 (Place an “X” in One Box for Plaintiff (For Diversity Cases Only) and One Box for Defendant) 37) '() 37) '() &+(&. <(6 RQO\ LI GHPDQGHG LQ FRPSODLQW -85< '(0$1'