Surveillance in our time is being transformed from a general and undifferentiated technique of governance to the active embodiment of governance itself. Surveillance is both the repository of governance norms and the discipline of those norms within any regulatory system. Surveillance is thus a bundle of assumptions, factors, assessments, and actions incarnated on the bodies of the regulated. Surveillance in its modern form represents another step in the perfection of social panopticism, of the creation of systems of social order that are self-regulating and internalized among those regulated. It represents a shifting of coercive power from the external-the state, the police, and the institution to the internal-the individual and the private. ("Global Panopticism: States, Corporations, and the Governance Effects of Monitoring Regimes," p. 112).
C. Conclusions of Law Related to Standing
The Court turns, then, to its conclusions of law with respect to standing. Plaintiffs advance at least five distinct theories of how they have been, or will be, injured due to the addition of a citizenship question on the 2020 census, namely: (1) diminished political representation, both between and within states; (2) loss in government funds, again both between and within states; (3) harm to the sovereign interests of state and local governments caused by degradation of the census data upon which they rely; (4) diversion of resources; and (5) loss of privacy. The Court begins by addressing the question of certain NGO Plaintiffs’ standing to pursue claims based on any of those theories on behalf of their individual members (in addition to any standing they may have to pursue such claims in their own right). The Court then explains its conclusions as to whether each of these forms of injury is legally cognizable and, if so, whether Plaintiffs proved by a preponderance of the evidence that they have been, or will be, injured in that manner. As the Court will explain, it concludes that all five forms of injury are legally cognizable and that, for all but one of the theories — namely, the loss of privacy — at least some Plaintiffs proved by a preponderance that they have suffered, or will suffer, injury in fact sufficient to support standing. The Court then turns to whether Plaintiffs proved that those forms of injury are fairly traceable to Secretary Ross’s decision to add the citizenship question to the 2020 census and whether they are redressable by a favorable decision. Answering both questions in the affirmative, the Court concludes that most, if not all, Plaintiffs have Article III standing to bring the claims in these cases.
The Court notes that, aside from challenging each of Plaintiffs’ theories of injury on “traceability” grounds, Defendants make only limited objections to each individual theory of
Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 156 of 277injury. Specifically, Defendants challenge Plaintiffs’ apportionment-loss and funding-loss theories only on the ground that they are insufficiently “imminent,” Defs.’ Post-Trial Br. 67-68, ¶¶ 11-14; the resource-expenditure theory only on the ground that Plaintiffs’ expenditures do not qualify as legally cognizable injuries-in-fact, id. at 69-70, ¶¶ 15-19; and the data-degradation theory only on the ground that it is not a sufficiently “concrete” and “tangible” injury for purposes of Article III, id. at 70, ¶¶ 20-21. Defendants do not explicitly address Plaintiffs’ theory of harm to their privacy interests anywhere in their Proposed Conclusions of Law. The Court cannot ignore the omitted objections given its independent obligation to assess its jurisdiction. But it is worth noting that Defendants themselves do not appear to believe that any of Plaintiffs’ injuries are not redressable or — apart from their ambitious objections to traceability, discussed below — fairly traceable to the citizenship question.
1. Associational Standing
Before turning to Plaintiffs’ theories of injury, however, the Court briefly addresses the issue of associational standing — namely, whether any of the NGO Plaintiffs have standing to bring claims on behalf of their individual members (in addition to any standing they may have to pursue such claims in their own right). An organization has “associational” standing to bring claims on behalf of its members if “(a) its members would otherwise have standing to sue in their own right; (b) the interests it seeks to protect are germane to the organization’s purpose; and (c) neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit.” Food and Commercial Workers v. Brown Grp., 517 U.S. 544, 553 (1996) (internal quotation marks omitted). The third of those requirements is only a prudential one — that is, it is not an element of Article III’s jurisdictional limitations on the power of the federal courts. See id. at 554-57.156Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 157 of 277Applying those standards here, the Court concludes — subject to its broader conclusions regarding standing below — that three of the NGO Plaintiffs, NYIC, MRNY, and ADC, have proved that they have associational standing to seek relief on behalf of certain of their members. The second and third prongs of the analysis can be swiftly addressed. Defendants make no argument with respect to the third, non-jurisdictional prong and, thus, have waived the issue. Cf. Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662, 670 n.2 (2010). In any event, concerns of “administrative convenience and efficiency” favor associational standing, as neither the claims asserted nor the relief requested in this litigation call for significant participation by individual members; at most, the claims call for proof of their residence, but that can be readily established without direct participation. Food and Commercial Workers, 517 U.S. at 556-57. As for the second prong of the analysis, which Defendants also fail to contest, the Second Circuit has explained that an interest is “germane” to an organization’s purpose if the lawsuit would “reasonably tend to further the general interests that individual members sought to vindicate in joining the association and . . . bears a reasonable connection to the association’s knowledge and experience.” Bldg. & Constr. Trades Council of Buffalo & Vicinity v. Downtown Dev., Inc., 448 F.3d 138, 149 (2d Cir. 2006). Here, as discussed above, ADC’s, MRNY’s, and NYIC’s missions involve obtaining government benefits for their communities, including — indeed, expressly by means of — ensuring a fair and accurate census count of those communities. Thus, the second prong of the associational standing test is satisfied.
That leaves the first — and only disputed — prong of the analysis: whether the NGO Plaintiffs’ members would otherwise have standing to sue in their own right. Although proof of a mere “statistical probability that some of [an organization’s] members are threatened with concrete injury” is not enough to satisfy the first prong of associational standing, an organization can meet the first prong by offering “specific allegations” (along, at this stage, with proof)157Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 158 of 277“establishing that at least one identified member had suffered or would suffer harm.” Summers v. Earth Island Institute, 555 U.S. 488, 497-98 (2009). Asserting that “a vague reference to unidentified members does not confer associational standing on an organization,” Defendants contend that the NGO Plaintiffs have “failed to identify any member who has standing to sue in his or her own right.” Defs.’ Post-Trial Br. 71, ¶¶ 25-26. While that is true of CASA, see generally Escobar Decl., it is not true of NYIC, MRNY, or ADC. NYIC has identified specific members who receive funds from federal programs that distribute those funds on the basis of census data. See Recitation of Facts ¶ 258. Similarly, MRNY has identified specific members who live in New York (and, more specifically, certain counties within the state) whose children attend Head Start programs or public schools that receive Title I funding. See Recitation of Facts ¶ 256. And ADC has identified specific members who reside in California, Florida, New York, the District of Columbia, and New Jersey. See Recitation of Facts ¶ 257. More generally, there is evidence that ADC has members residing in all fifty states. Id. That is sufficient to establish that those members, although unnamed, would have standing to the extent that standing depends only on the facts of their existence and residence in a particular jurisdiction. As will be discussed below, that is true with respect to some (albeit not all) of the theories of injury at issue here.48
48Defendants seem to argue that an organization must identify particular members by name in order to have associational standing to pursue claims on their behalf. See Defs.’ Post-Trial Br. 71, ¶ 25. But the cases they cite do not support that proposition. As discussed, an organization can satisfy the first prong of the associational standing analysis by offering proof “establishing that at least one identified member ha[s] suffered or would suffer harm.” Summers, 555 U.S. at 498. It would overread the word “identified” in that context to require an organization to name the member who might have standing in his or her own right. In the first place, such specific identifying information is often unnecessary to determine whether a person would have Article III standing. For example, as in this case, whether a person will suffer a loss of political representation depends on the facts of his or her existence and residence within a particular jurisdiction, but not on his or her name. Where those (or other relevant) facts are proved, a court need look no further to conclude that the organization has members who would have standing to158Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 159 of 277In short, three of the NGO Plaintiffs have established — in addition to whatever standing they may have in their own right — that they have standing to the same extent that certain of their individual members would have standing. More specifically, NYIC has standing to the extent that its members who receive funds from census-tied programs would have standing. MRNY has standing to the extent that its individual members residing in New York State generally, and Queens, Brooklyn, the Bronx, and Long Island specifically, would have standing and to the extent that its individual members whose children attend Head Start programs or public schools that receive Title I funding would have standing. And ADC has standing to the extent that its individual members residing in all fifty states and the District of Columbia — and certainly those residing in California, New York, Pennsylvania, Texas, Florida, Rhode Island, Washington State, Arizona, Illinois, Maryland, New Jersey, and the District of Columbia — would have standing.
2. Injury in Fact
With that, the Court turns to Plaintiffs’ five theories of injury: (1) loss of political representation, both between states (in congressional reapportionment) and within states (in redistricting); (2) loss of federal funds, also both between and within states; (3) harm to important sovereign interests caused by degradation of the census data on which state and local governments rely; (4) injury to organizations and local governments through the diversion of,
pursue a particular claim in their own right. Moreover, to hold that Article III requires an organization to name those of its members who would have standing would be in tension with one of the fundamental purposes of the associational standing doctrine — namely, protecting individuals who might prefer to remain anonymous. See NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 458-60 (1958); see also Food and Commercial Workers, 517 U.S. at 551-52. Dicta about “unidentified members” aside, the most that Defendants’ cases establish is that a plaintiff must prove “facts sufficient to establish that one or more of its members has suffered, or is threatened with, an injury.” Valley Forge Christian Coll. v. Ams. United for Separation of Church & State Inc., 454 U.S. 464, 487 n.23 (1982). For the reasons discussed above, ADC has done so, even as to those members whom it does not identify by name.159Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 160 of 277and strain inflicted upon, organizational resources; and (5) infringement of privacy interests in the information collected by the census. As noted, the Court concludes that all five forms of injury are legally cognizable and that, as to four of the five — namely, all except the privacy- infringement theory — at least some Plaintiffs have proved by a preponderance of the evidence that they have suffered, or will suffer, injury in fact sufficient to support standing.
a. Diminished Political Representation
First, Plaintiffs have proved that several states are likely to lose one or more seats in the next round of congressional redistricting if the citizenship question is added to the census. See Recitation of Facts ¶ 243-45. The Supreme Court has squarely held that the loss of a seat or seats in the House of Representatives “undoubtedly satisfies the injury-in-fact requirement of Article III standing” because of the dilution of political power that results from such an apportionment loss. Dep’t of Commerce, 525 U.S. at 331; accord Carey v. Klutznick, 637 F.2d 834, 836-38 (2d Cir. 1980); cf. Utah v. Evans, 536 U.S. 452, 458 (2002). Further, although the census is still months away, such an injury is sufficiently “imminent” to invoke federal-court jurisdiction now. See Dep’t of Commerce, 525 U.S. at 327, 332 (finding threatened vote dilution because of census modifications sufficiently “imminent” in two lawsuits brought twenty-five months before the census date).
In particular, Plaintiffs have proved that California’s prospective loss of a seat in the House of Representatives is “certainly impending.” See Recitation of Facts ¶¶ 243, 245. And although California is not a Plaintiff here, ADC is and, as discussed, it has individual members who reside in California. See id. ¶ 257. Those individual members’ “expected loss of a Representative to the United States Congress undoubtedly satisfies the injury-in-fact requirement160Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 161 of 277of Article III standing.” Dep’t of Commerce, 525 U.S. at 331.49 Plaintiffs have also proved that New York and Illinois face a “substantial risk” that they will lose at least one seat in the next Congressional reapportionment because of the citizenship question’s addition to the 2020 Census. See Recitation of Facts ¶¶ 243, 245. New York and Illinois are Plaintiffs here, and thus have standing to seek a remedy for that injury in their own right. Cf. Evans, 536 U.S. 452. Additionally, that provides an independent basis for ADC and MRNY to seek relief, as ADC has individual members who reside in New York and Illinois, see Recitation of Facts ¶ 257, and MRNY has individual members who reside in New York, see id. ¶ 256. ADC’s individual members who reside in Texas, Arizona, and Florida, see id. ¶ 257, likewise face a substantial risk of lost representation in the U.S. House of Representatives.
A closer question is whether Plaintiffs have also proved that any of the NGO Plaintiffs’ members will suffer an Article III injury in fact in the form of lost representation in intrastate
49Although Defendants do not raise the issue, the jurisdictional nature of the standing inquiry compels the Court to note that ADC’s standing does not depend on proof that its members residing in California and elsewhere are U.S. citizens or voters — matters on which the record is silent. While intrastate redistricting is governed by the Equal Protection Clause’s “one person, one vote” principle, interstate congressional apportionment is governed by Section 2 of the Fourteenth Amendment, which provides that “Representatives shall be apportioned among the several States according to their respective numbers, counting the whole number of persons in each State, excluding Indians not taxed.” U.S. Const. amend. XIV, § 2 (emphasis added). Thus, while a nonvoter might or might not have standing to challenge a district on a representational dilution theory under the Equal Protection Clause — an open question, however much common sense might suggest the answer is “yes,” see, e.g., Calvin v. Jefferson Cty. Bd. of Comm’rs, 172 F. Supp. 3d 1292, 1307 n.12 (N.D. Fla. 2016) — the “undoubted” Article III injury that a person suffers when his or her state loses a representative in congressional reapportionment, Department of Commerce, 525 U.S. at 331, ultimately traces to a legally protected interest that does not depend on that person’s citizenship status or eligibility to vote. See Evenwel v. Abbott, 136 S. Ct. 1120, 1127-28 & n.9 (2016) (describing the original understanding of population-based apportionment, which James Madison called a “fundamental principle of the proposed constitution,” as reflecting Alexander Hamilton’s “principled argument for allocating seats to protect the representational rights of every individual of the community at large,” and describing the understanding of the Fourteenth Amendment’s framers that “non- voting classes may have as vital an interest in the legislation of the country as those who actually deposit the ballot” (citations and internal quotation marks omitted)).161Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 162 of 277redistricting carried out on the basis of the federal decennial census. Intrastate vote dilution plainly qualifies as an injury in fact for purposes of Article III. Dep’t of Commerce, 525 U.S. at 332-33; Carey, 637 F.2d at 838. In Department of Commerce, for example, the Supreme Court held — given the fact that certain counties “were substantially likely to lose population” under the Census Bureau’s plan — that residents of those counties had satisfied Article III’s injury-in- fact requirement. 525 U.S. at 334. In other words, all the Court required was proof that (1) certain states relied on federal decennial census data for intrastate redistricting, (2) voters in certain counties in those states were “substantially likely . . . to suffer vote dilution as a result of the [Census Bureau’s] plan,” and (3) plaintiffs were among the voters who lived in those counties in those states. Id. at 332-34 (internal quotation marks omitted). As that language suggests, however, the Supreme Court has tended to describe the injury in the intrastate vote- dilution context as affecting only eligible voters. Here, Plaintiffs have proved that several of the NGO Plaintiffs’ individual members reside in counties that will lose statewide population shares because of the citizenship question, but the record is silent as to whether those members are also eligible voters. In this instance, the Court concludes that the interests of avoiding an unnecessary holding on a delicate constitutional question outweigh even the powerful interests favoring resolution of all open legal issues in this Opinion. See Ashwander, 297 U.S at 347 (Brandeis, J., concurring). The Court therefore declines to address whether the facts proved by Plaintiffs are legally sufficient to establish that any of the NGO Plaintiffs’ members will suffer an Article III injury when, at a minimum, they lose political representation in the next intrastate redistricting.
b. Loss of Government Funds
Second, given the Court’s factual findings, many Plaintiffs have proved that they will suffer a loss of funding from federal programs that distribute such funding on the basis of census data. Such a monetary loss is a classic form of Article III injury in fact. See Carey, 637 F.2d at162Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 163 of 277838 (holding that “citizens who challenge a census undercount on the basis . . . that improper enumeration will result in loss of funds to their city have established . . . an injury” for purposes of standing); accord City of Detroit v. Franklin, 4 F.3d 1367, 1374-75 (6th Cir. 1993); City of Philadelphia v. Klutznick, 503 F. Supp. 663, 671 (E.D. Pa. 1980); City of Camden v. Plotkin, 466 F. Supp. 44, 47-51 (D.N.J. 1978). More specifically, a state or locality that proves that it will lose funds under federal revenue-sharing programs satisfies Article III’s injury-in-fact requirement. See Carey, 637 F.2d at 838 (“New York City and New York State . . . have standing as recipients of federal funds under revenue sharing.”). So too, an individual who proves that he or she will suffer a loss of federal benefits has adequately proved an Article III injury. See id. (“The individual plaintiffs in this case have alleged concrete harm in the form of . . . decreased federal funds flowing to their city and state, thus establishing their standing.”).
Applying those principles here, the Court concludes as follows:
- Plaintiffs Illinois, Massachusetts, Maryland, New Jersey, New Mexico, New York, Oregon, Washington State, and the District of Columbia have proved that if the net undercount of noncitizens attributable to the citizenship question is as low as two percent — far lower than the Court has found to be the best estimate — they will lose funds from several federal programs. See Recitation of Facts ¶¶ 251-52. On that basis, the Court concludes that those Plaintiffs face “certainly impending” injuries.
- Plaintiffs have also proved that (1) MRNY’s individual members who reside in New York and (2) ADC’s individual members who reside in California, Texas, Florida, Nevada, and Hawaii, and in Plaintiff States New York and New Jersey, face a “certainly impending” Article III injury for the same reasons. See Recitation of Facts ¶¶ 256-67.
- Additionally, Plaintiffs have proved that MRNY’s individual members who reside in New York and benefit from Title I and Head Start funds will suffer an individualized harm that is “certainly impending,” in light of its propensity to occur in even the most implausibly low noncitizen net undercount scenarios. See Recitation of Facts ¶ 256.
- Finally, Plaintiffs have proved that NYIC’s members Chhaya, LSA, Korean Community Services of Metropolitan New York, Planned Parenthood of New York City, and Chinese-American Planning Council all receive funding through programs that allocate funds to states based on census data. See Recitation of Facts ¶ 258. Given the net undercount scenarios that the Court has found to be likely, those organizations have proved that they face a substantial risk of an Article III injury in the form of lost funding.
163Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 164 of 277Defendants argue that, even if Plaintiffs have proved these losses are likely to occur, they do not count as Article III injuries because they may be “offset” by gains that turn up elsewhere in the federal funding scheme. See Tr. 506; Defs.’ Post-Trial Br. 34-36, ¶¶ 220-38. But “the fact that an injury may be outweighed by other benefits, while often sufficient to defeat a claim for damages, does not negate standing.” Denney v. Deutsche Bank AG, 443 F.3d 253, 265 (2d Cir. 2006); see also Texas v. United States, 809 F.3d 134, 155-56 (5th Cir. 2015), aff’d by an equally divided Court, 136 S. Ct. 2271 (2016) (mem.); NCAA v. Governor of N.J., 730 F.3d 208, 223 (3d Cir. 2013) (“A plaintiff does not lose standing to challenge an otherwise injurious action simply because he may also derive some benefit from it. Our standing analysis is not an accounting exercise . . . .”), abrogated on other grounds by Murphy v. NCAA, 138 S. Ct. 1461 (2018); Alaska Elec. Pension Fund v. Bank of Am. Corp., 175 F. Supp. 3d 44, 53 (S.D.N.Y. 2016).
Plaintiff Colorado will also suffer an injury for a related reason: It will lose the ability to spend funds because its own state constitution limits year-on-year increases in expenditures to a function of the state’s population growth as determined by federal census data. The Colorado “Taxpayer’s Bill of Rights,” ratified into the Colorado Constitution, limits the “maximum annual percentage change in state fiscal year spending” to “inflation plus the percentage change in state population in the prior calendar year, adjusted for revenue changes approved by voters after 1991,” and provides that “[p]opulation shall be determined” for that purpose “by annual federal census estimates and such number shall be adjusted every decade to match the federal census.” Colo. Const. art. X, § 20(7)(a); see Colo. Rev. Stat. § 24-77-103(1)(c)(III) (providing for a formula to recalculate the percentage change each year on the basis of new federal census estimates). Whether characterized as a loss in funding, or a loss in the ability to spend for the general welfare of its residents (and, thus, an infringement on Colorado’s sovereign interests), Colorado will suffer an Article III injury if its population is undercounted by even the smallest164Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 165 of 277amount. Given the Court’s findings as to the likely net undercount of Colorado’s population, see Recitation of Facts ¶ 249, the Court concludes that Colorado faces a “certainly impending” injury traceable to the addition of the citizenship question.
In sum, given the remarkably low net undercount of noncitizens that would prompt the foregoing losses of funding, Governmental Plaintiffs Colorado, Illinois, Massachusetts, Maryland, New Jersey, New Mexico, New York, Oregon, Washington State, and the District of Columbia; and NGO Plaintiffs ADC, MRNY, and NYIC have established an injury in fact that is “certainly impending.” At a minimum, there is no doubt that these Plaintiffs have demonstrated a “substantial risk” that they will suffer such injuries. In either case, these Plaintiffs have satisfied Article III’s injury-in-fact requirement.
c. Harm to the Quality and Accuracy of Data
Next, the Governmental Plaintiffs proved that they will suffer injury in fact from the degradation in data quality that would occur if the citizenship question appears on the 2020 census. At the outset, there is no reasonable dispute that Plaintiffs proved by a preponderance of the evidence that addition of the citizenship question would result in that harm. Indeed, the evidence at trial was undisputed that, regardless of how successful NRFU operations are in remedying a net differential undercount due to a differential decline in self-response rates, the addition of the citizenship question will result in harm to the quality of census data. First, as noted above, that was the original position of the Census Bureau in connection with its review of DOJ’s request. See Recitation of Facts ¶¶ 8, 19, 28-30, 34. Second, Defendants own expert, Dr. Abowd, agreed, testifying that the addition of a citizenship question would harm the overall quality of 2020 census data regardless of any net undercount scenario. See Tr. 882, 952-53, 1221-22, 1251. And third, Plaintiffs’ experts testified credibly in the same manner. Tr. 302-04. Additionally, Dr. Salvo, the Director of the New York City Department of City Planning’s165Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 166 of 277Population Division, explained credibly that the degradation in quality of the data would harm New York City’s ability to allocate educational and public health resources efficiently and effectively. See Recitation of Facts ¶ 263.50 Crucially, these harms will occur whether or not there is a net undercount — meaning that this theory of injury does not depend on the causal chain of events connecting the decline in self-response rates among noncitizen households to a net differential undercount of people who live in such households.
Understandably, then, Defendants raise no objections in their post-trial filings to the imminence, traceability, or redressability of this harm. See Defs.’ Post-Trial Br. 70, ¶¶ 20-21. Instead, the principal dispute between the parties is whether a degradation in the quality of census data is a legally cognizable harm. Compare id., with Pls.’ Proposed Conclusions ¶¶ 57- 59. The Court agrees with Plaintiffs that it is. For starters, the Supreme Court has consistently held that the “inability to obtain information” is a cognizable Article III injury. FEC v. Akins, 524 U.S. 11, 21 (1998); Pub. Citizen v. U.S. Dep’t of Justice, 491 U.S. 440, 449 (1989) (“[R]efusal to permit appellants to scrutinize the ABA Committee’s activities to the extent [the Federal Advisory Committee Act] allows constitutes a sufficiently distinct injury to provide standing to sue.”); id. (“Our decisions interpreting the Freedom of Information Act have never suggested that those requesting information under it need show more than that they sought and were denied specific agency records.”). If the inability to obtain information is a cognizable form of injury for Article III purposes, it follows a fortiori that the inability to obtain accurate
50Defendants contend that Dr. Salvo’s testimony about the citizenship question’s effects on data quality was too “conclusory” and “lacked key specifics,” including that he did not name the precise city services that might be affected by a decline in data quality. Defs.’ Post-Trial Br. 40, ¶ 264. At best, Defendants’ criticisms amount to little more than nitpicking. Dr. Salvo testified credibly that the City’s allocation of resources with respect to several important programs would be harmed by the degradation in data quality attributable to the citizenship question and, moreover, did name specific examples relating to the New York City Board of Education and Department of Health. See Recitation of Facts ¶ 263.166Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 167 of 277information would be as well. After all, there is no interest in obtaining false or faulty information.
To be sure, many of these cases involved statutory entitlements to certain information — the classic example of such an entitlement (although, interestingly, the example least discussed in the context of Article III standing) being the Freedom of Information Act. That is no objection to the conclusion that the Supreme Court has recognized the denial of information as an Article III injury, however, because if such an injury were not already “concrete” enough for Article III purposes, Congress could not make it so. As the Supreme Court recently clarified, “Congress’ role in identifying and elevating intangible harms” does not allow it to grant standing on the basis of pure statutory violations — instead, “Article III standing requires a concrete injury even in the context of a statutory violation.” Spokeo, 136 S. Ct. at 1540. Thus, if the claims of informational injury discussed in the cases amounted to nothing more than “bare procedural violation[s], divorced from any concrete harm,” they would not “satisfy the injury-in- fact requirement of Article III,” id. — but they are more than that, so they do.
Similarly, a state that relies on the information provided by the federal government under an existing statutory arrangement suffers a sufficiently “concrete” and “particularized” injury for purposes of Article III when the federal government degrades the quality of that information. States are sovereign entities with sovereign interests in the making and enforcement of their own laws. See Alfred L. Snapp & Son, Inc. v. Puerto Rico ex rel. Barez, 458 U.S. 592, 601 (1982); cf. Maryland v. King, 567 U.S. 1301, 133 S. Ct. 1, 3 (2012) (Roberts, C.J., in chambers) (concluding that Maryland suffered an injury to its “law enforcement and public safety interests” from a lower-court order preventing the state from utilizing DNA samples for law enforcement purposes pursuant to a state statute). But they frequently do so in collaboration with, or in reliance on, the federal government — such is the genius of the federal system, which has historically embraced167Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 168 of 277various creative models of “cooperative federalism.” See, e.g., New York v. United States, 505 U.S. 144, 167-69 (1992); Hodel v. Va. Surface Mining & Reclamation Assn., Inc., 452 U.S. 264, 286-89 (1981). Most relevant here, states have long relied on federal decennial census data for countless sovereign purposes, and many of the State Plaintiffs here even require the use of such data by law; in some instances, it is written into their state constitutions.
The most noteworthy examples of this reliance are the State Plaintiffs that mandate the use of federal decennial census data to apportion state representatives. See, e.g., Recitation of Facts ¶ 246; see also Georgia, 539 U.S. at 488 n.2; Dep’t of Commerce, 525 U.S. at 333 n.4, 334. But state reliance on federal census data for sovereign purposes is goes well beyond that, as the following examples — ranging from the profound to the mundane — make plain:
- States often seek to apportion governmental expenses equitably among local governments, for example, by requiring each county to contribute in proportion to its share of population as calculated by the latest federal decennial census. See, e.g., Colo. Rev. Stat. § 20-1-208 (requiring counties to contribute towards the salaries of employees in Colorado district attorneys’ offices according to the counties’ population share, as determined in the “last preceding decennial census” of the district in which each county sits); Conn. Gen. Stat. Ann. § 7-338 (requiring that the expenses of a charter commission for a new metropolitan fire, sewer, or water district be apportioned among “each town, city or borough” according to its proportion of the district’s total population, “as determined by the last-preceding federal census”).
- Other states provide for direct state aid to local governmental units on a per capita basis, relying on federal decennial census data to ensure a fair distribution of resources. See, e.g., 30 Ill. Comp. Stat. 115/2(a) (requiring monthly allocations of the Illinois “Local Government Redistributive Fund” and “Income Tax Surcharge Local Government Distributive Fund” in proportion to cities’ and counties’ population as determined by federal decennial census data); Minn. Stat. Ann. § 260.821(1)(b) (requiring that support grants to Indian tribes be apportioned, in part, on the basis of “the ratio of the tribe’s on- reservation population to the state’s total on-reservation population,” as determined by the “most recent federal census data”); R.I. Gen. Laws §§ 23-18.9-1, 23-18.9-3 (requiring that funds appropriated as grants in aid of local refuse disposal services be distributed according to a community’s share of the statewide population, according to federal census data); N.Y. State Fin. Law § 54(2) (McKinney) (providing that each fiscal year, “there shall be apportioned and paid to the several counties, cities, towns and villages, from moneys appropriated by the state, for the support of local government” amounts in proportion to their “population”); id. § 54(1)(a)(1) (defining “population” to mean “the population as shown by the latest preceding decennial federal census” or a “special
168Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 169 of 277population census”).
- Some states also redistribute state tax receipts to cities on the basis of each city’s population share. See, e.g., Iowa Code Ann. § 312.3(2)(a) (providing that the state treasurer shall, on the first day of each month, “[a]pportion among the cities of the state, in the ratio which the population of each city, as shown by the latest available federal census, bears to the total population of all such cities in the state, the percentage of the road use tax funds which is credited to the street construction fund of the cities”); N.M. Stat. Ann. § 7-1-6.26(C) (providing that counties shall receive a share of the “county government road fund” calculated in part based on the county’s population “as shown by the most recent federal decennial census”); Or. Rev. Stat. § 221.770 (requiring that certain liquor revenues be apportioned among Oregon’s cities in part based on their population as determined by federal census data).
- North Carolina apportions state support to community colleges in part based on a formula that takes into account ratios of county and area populations as determined by the “latest United States census.” N.C. Gen. Stat. § 115D-31(a)(3).
- Pennsylvania sets the maximum salary for the mayor of any city with a population greater than 15,000 at “$500 per every thousand residents per year as determined by the most recent census data provided by the United States Census Bureau.” 11 Pa. Cons. Stat.
§ 11208(a), (b). It also sets the maximum “tapping fee” that a sewer or water authority may charge a property owner who connects to the authority’s sewer or water system as a function of the average household size “as established by the most recent census data provided by the United States Census Bureau.” 53 Pa. Cons. Stat. § 5607(d)(24) (i)(C)(V)(e).
- Some states more generally define “population” for the purpose of their laws to mean population as calculated in the most recent federal decennial census. See, e.g., Iowa Code Ann. § 9F.6 (“Whenever the population of any county, township or city is referred to in any law of this state, it shall be determined by the last preceding certified federal census unless otherwise provided.”); N.Y. Gen. Constr. Law § 37-b (McKinney) (“The term population when used in relation to this state . . . shall, unless otherwise provided in relation to such use, mean population as shown by the latest federal census published as a final population count by the United States bureau of the census.”); Va. Code Ann. § 1- 235 (“‘Population’ or ‘inhabitants’ means with reference to any county, city, town, political subdivision of the Commonwealth or any combination thereof, the natural persons in such county, city, town, political subdivision or combination as shown by the unadjusted United States decennial census . . . .”).
Meanwhile, it is, of course, the federal government’s job to collect and distribute accurate federal decennial census data. See U.S. Const. art. I, § 2, cl. 3; see also Utah, 536 U.S. at 478 (explaining that the Framers had a “strong constitutional interest in [the] accuracy” of the census); Wisconsin, 517 U.S. at 20 (holding that the conduct of the census must bear a
169Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 170 of 277“reasonable relationship to the accomplishment of an actual enumeration of the population, keeping in mind the constitutional purpose of the census,” namely, obtaining an accurate count of the population in each state); Pub. L. No. 105-119, § 209(a)(6), 111 Stat. at 2481 (“Congress finds that . . . [i]t is essential that the decennial enumeration of the population be as accurate as possible, consistent with the Constitution and laws of the United States.”). When the federal government degrades the quality of that data, it therefore inflicts a cognizable injury on the sovereign interests of reliant states.51
An example may be helpful in illustrating the point. Suppose a state were to premise certain of its policies on a person’s lawful presence in the United States — for example, suppose that it chose to deny certain benefits to undocumented immigrants or required its law- enforcement officials to inquire into the immigration status of any person detained in state custody for any reason. “The accepted way” for states “to perform [such] status checks” — and surely the most reliable — is to contact the DHS’ Immigration and Customs Enforcement (“ICE”), the federal agency that accepts and responds to such inquiries from interested states. Arizona v. United States, 567 U.S. 387, 411 (2012). Now suppose that ICE were to degrade the quality of its data set, thereby undermining its usefulness to the state as a tool for implementing its policy priorities. If this hypothetical state were to challenge the decisions causing the degradation in immigration-status data, the federal agency could certainly defend its actions on the grounds that they were lawful. But could it seriously deny that the state had suffered a cognizable injury for purposes of standing? Surely not.
51That does not mean that, in every case, a state will have a “right” to such data — or a right to data of a certain quality — sufficient to support a valid cause of action to obtain it. But it does mean that a state suffers a concrete and particularized injury when the federal government degrades important tools of sovereignty — or takes those tools away altogether.170Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 171 of 277Indeed, ample case law supports the proposition that a state has a strong sovereign interest in conducting its own policy, the burdening of which causes an injury in fact for Article III purposes. One such sovereign interest is a state’s “exercise of sovereign power over individuals and entities within [its] jurisdiction — this involves the power to create and enforce a legal code, both civil and criminal.” Alfred L. Snapp & Son, 458 U.S. at 601. Another such sovereign interest — which, in light of the frequent prohibition on parens patriae suits against the federal government, Massachusetts v. Mellon, 262 U.S. 447, 485-86 (1923), is “distinct from . . . the general well-being of its residents” — is a state’s “interest in securing observance of the terms under which it participates in the federal system,” Alfred L. Snapp & Son, 458 U.S. at 607- 08; cf. Maine v. Taylor, 477 U.S. 131, 137 (1986) (“[A] State clearly has a legitimate interest in the continued enforceability of its own statutes.”); Diamond v. Charles, 476 U.S. 54, 65 (1986) (“Because the State alone is entitled to create a legal code, only the State has the kind of ‘direct stake’ . . . in defending the standards embodied in that code.”).
The Fifth Circuit’s decision in Texas v. United States, 809 F.3d 134 (5th Cir. 2015), aff’d by an equally divided Court, 136 S. Ct. 2271 (2016), is instructive on this front. In that case, Texas led a coalition of states in a challenge to the Obama Administration’s Deferred Action for Parents of Americans and Lawful Permanent Residents program (“DAPA”). The Court held that the states had suffered a cognizable injury for purposes of standing because DAPA would have entitled its recipients to obtain driver’s licenses under existing state law and providing those licenses would have come at a financial cost to Texas. See id. at 155-56. In denying a stay of the district court’s preliminary injunction pending appeal, the Fifth Circuit cited Alfred L. Snapp & Son, explained that Texas possessed a sovereign interest in the maintenance of its own legal code, and held that “Texas’s forced choice between incurring costs and changing its laws is an injury because those laws exist for the administration of a state program, not to challenge federal171Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 172 of 277law, and Texas did not enact them merely to create standing.” 787 F.3d 733, 749 (5th Cir. 2015). The court reasoned that “if pressure to change state law in some substantial way were not injury, states would have no standing to challenge bona fide harms because they could offset most financial losses by raising taxes or fees.” Id. Several months later, the Fifth Circuit affirmed the preliminary injunction on the merits, reiterating and confirming its conclusions as to standing. The Circuit held that “states may have standing based on . . . federal interference with the enforcement of state law, at least where the state statute at issue regulates behavior or provides for the administration of a state program and does not simply purport to immunize state citizens from federal law.” 809 F.3d 134, 153 (alterations and internal quotation marks omitted). Such “intrusions,” the court explained, “are analogous to pressure to change state law.” Id.
Like the state plaintiffs in Texas v. United States, the State Plaintiffs here have enacted their reliance on federal census data into law — in some cases, as noted, even into their constitutions. Moreover, as in Texas v. United States, “there is no allegation,” let alone proof, that those jurisdictions enacted their laws or ratified their constitutions “to manufacture standing” in these cases. Texas, 809 F.3d at 159. If a citizenship question is added to the decennial census, these Plaintiffs will be subjected to a forced choice: They can use the degraded data, resulting in worse policy; they can spend money to compensate for the damage; or they can change their laws to relieve themselves of the legal obligation to use federal census data in making and enforcing their laws (which would presumably necessitate the expenditure of additional resources to collect data of their own anyway). Such “pressure to change state law constitutes an injury” within the meaning of Article III. Texas, 787 F.3d at 749; see Texas, 809 F.3d at 153.
Accordingly, several Governmental Plaintiffs — including State Plaintiffs New York, Colorado, Connecticut, Delaware, Illinois, Iowa, Maryland, Massachusetts, Minnesota, New Jersey, New Mexico, North Carolina, Oregon, Pennsylvania, Rhode Island, Virginia, and172
Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 173 of 277
Washington, and Plaintiff New York City — have proved an imminent injury to their sovereign interests through the degradation in quality of federal census data.
d. Diversion of Resources
Next, the Court finds that the NGO Plaintiffs have established that they have already suffered, and will continue to suffer, injury in fact due to a diversion of their resources. In Havens Realty Corp. v. Coleman, 455 U.S. 363 (1982), the Supreme Court held that an organization can establish Article III injury in fact by proving “concrete and demonstrable injury to [its] activities — with the consequent drain on [its] resources.” Id. at 379; see id. at 379 n.21 (holding that an organization that proves it “has indeed suffered impaiment” in its activities has proved an Article III injury); Nnebe v. Daus, 644 F.3d 147, 157 (2d Cir. 2011) (requiring a showing of “perceptible impairment”); see also Ragin v. Harry Macklowe Real Estate Co., 6 F.3d 898, 904-06 (2d Cir. 1993). Although an organization may not inflict such an “impairment” on itself for purposes of creating standing — for example, by incurring litigation expenses in the very lawsuit at issue, see, e.g., Citizens for Responsibility and Ethics in Wash. v. Trump, 276 F. Supp. 3d 174, 189-93 (S.D.N.Y. 2017) (“CREW”) — the Havens Realty theory of organizational standing squarely covers a claim of injury from “purportedly illegal action [that] increases the resources the group must devote to programs independent of its suit challenging the action.” Spann v. Colonial Vill., Inc., 899 F.2d 24, 27 (D.C. Cir. 1990) (R.B. Ginsburg, J.); see CREW, 276 F. Supp. 3d at 190 (noting that Havens Realty applies where the “[d]efendant’s actions have impeded” an organization’s “ability to perform a particular mission-related activity, or [forced it] to expend resources to counteract and remedy the adverse consequences or harmful effects of [the] [d]efendant’s conduct”). This is exactly the kind of injury that NGO Plaintiffs allege here.
Defendants suggest that Havens Realty recognizes Article III injuries arising from organizational expenditures, but only where those expenditures are made in response to injuries173Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 174 of 277that are themselves sufficiently imminent and impending to satisfy Article III. See Defs.’ Post- Trial Br. 69, ¶ 15; Tr. 1486-89. Of course, that argument is beside the point because Plaintiffs have proved such injury, as discussed above. But the argument also makes no sense on its own terms. It would be illogical to recognize that organizations may be injured by expenditures made in response to future injuries, as the Supreme Court continues to do, see Clapper, 568 U.S. at 414 n.5, but to hold that that doctrine applies only in cases in which it would be superfluous. Indeed, that would render the category of plaintiffs that could establish standing under a Havens Realty theory a null set. Conspicuously, though, Defendants cite no Supreme Court case holding (or even hinting) that Havens Realty has been so cabined, much less overruled.
Applying Havens Realty here, the Court finds that the NGO Plaintiffs plainly have standing to challenge Secretary Ross’s decision. That is, all four NGO Plaintiffs have proved that the citizenship question will cause them — indeed, already is causing them — to divert organizational resources away from their core missions and towards combating the negative effects of the citizenship question. See Recitation of Facts ¶¶ 265-74.52 Defendants’ arguments to the contrary are somewhat ironic because the record makes clear that the Census Bureau itself relies on organizations like the NGO Plaintiffs to ensure a successful census — and will rely on them to counteract the indisputably negative effects of the citizenship question. See, e.g., Recitation of Facts ¶¶ 187-88, 274. In fact, Dr. Abowd, Defendants’ own expert, explicitly conceded that addition of the citizenship question on the census will make the “job[s]” of these
52Some of the NGO Plaintiffs cite expenses related to this litigation among the resources that they have expended because of the citizenship question. See, e.g., Altschuler Decl. ¶ 22. A claim of injury predicated on litigation expenses alone, however, would stand on shaky ground. See CREW, 276 F. Supp. 3d at 189-93. Given that the NGO Plaintiffs’ injuries involve diversion of resources other than mere litigation expense, see, e.g., Altschuler Decl., ¶¶ 19-21, the Court need not and does not rely on litigation expenses in reaching its conclusions that the NGO Plaintiffs have proved injury in fact.174Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 175 of 277organizations “harder.” Tr. 1303-05. In other words, Defendants’ own arguments against the NGO Plaintiffs’ injury depend on their prediction that any impending decline in self-response rates will be mitigated, in part by organizations such as the NGO Plaintiffs that will spend money and devote organizational resources to combat the citizenship question’s negative effects. Defendants’ own arguments against Plaintiffs’ other theories of standing therefore serve to confirm this one.
In contending that the NGO Plaintiffs have not proved this theory of injury, Defendants also fault the NGO Plaintiffs for the lack of any testimony “that their expenditure of resources took into account the extent to which the Census Bureau’s [NRFU] procedures would mitigate any differential net undercount attributable to the citizenship question.” Defs.’ Post-Trial Br. 40, ¶ 262. In other words, Defendants suggest that Plaintiffs should be denied standing to seek a remedy in federal court because they should have trusted the same parties who caused their injuries in the first place to fix them. That may be the kind of thing the fox would say to the henhouse, but it is not what one would expect the federal government to say to a coalition of civil rights organizations challenging what they believe to be unlawful governmental action. In any event, the argument is unpersuasive on its merits, too, as it has a circular quality: As noted above, Defendants’ own efforts to mitigate a decline in self-response depend in part on organizations like the NGO Plaintiffs expending resources to counteract such a decline. That is, to the extent that Defendants’ NRFU efforts are successful in any respect, it will be in part because the NGO Plaintiffs have expended resources in aid of those efforts. And in any event, the Court has already found, based on the evidence at trial, that Defendants’ NRFU efforts will not succeed in eliminating a net differential undercount as a result of the citizenship question.
Defendants’ only remaining objection to this theory of injury is without merit. They assert that the NGO Plaintiffs have “not met their burden of proving” a “direct conflict between175Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 176 of 277their missions and the reinstatement of a citizenship question on the census.” Defs.’ Post-Trial Br. 40, ¶ 263; see id. at 69-70, ¶ 18. In support of that conclusion, however, they rely on a misreading of the D.C. Circuit’s decision in National Law Center on Homelessness & Poverty v. Kantor, 91 F.3d 178 (D.C. Cir. 1996). In that case, an organization dedicated to helping the homeless asserted Article III standing to challenge an undercount of homeless individuals based on an anticipated injury to the organization’s “purpose of providing accurate information on homelessness and poverty.” Id. at 182. But the D.C. Circuit did not accept that characterization of the organization’s mission; pointing out that the National Law Center was not a “news reporting agency,” the Court observed instead that the “purpose of providing” such “accurate information” was “ancillary to [its] general approach of gaining governmental responses to improve the lot of the homeless.” Id. Given that definition of the organization’s mission, the “indirectness” that the court found to be problematic in that case is easy to see. It is equally easy to see that it is not present here.
More significantly, the D.C. Circuit went on to hold that the connection between the ability to disseminate accurate information about the homeless population and the probability of achieving tangible benefits based on the public’s reaction to that information was “at the far end of speculation” on the factual record before it. Id. Observing that “conjectural” connection, the Court held that it was “[f]or this reason” that the organization’s reliance on Havens Realty was “misplaced” — that is, that the group had simply failed to show, as a factual matter, that “a homeless undercount . . . impose[d] any . . . barriers to either the homeless or their advocates.” Id. (emphasis added); see Am. Soc’y for Prevention of Cruelty to Animals v. Feld Entm’t, 659 F.3d 13, 25 (D.C. Cir. 2011) (construing the “direct conflict” requirement to mean that “[i]f the challenged conduct affects an organization’s activities but is neutral with respect to its substantive mission,” it is “‘entirely speculative’ whether the challenged practice will actually176Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 177 of 277impair the organization’s activities”). Here, of course, Plaintiffs have proved, as a factual matter, that Defendants’ conduct will — absent an expenditure of resources — harm their core missions of advancing the interests, and enhancing the political power, of the communities they serve. There is thus a “direct conflict” between their missions and the conduct they challenge, and they have proved an Article III injury under Havens Realty.
Finally, Plaintiffs have also proved that New York City and Chicago have diverted limited resources towards counteracting the injurious effects of a citizenship question. That forced resource-diversion also qualifies as an Article III injury. It is long settled that a municipality that loses access to resources, thus “threatening its ability to bear the costs of local government and to provide services,” suffers an Article III injury in fact. Gladstone Realtors v. Vill. of Bellwood, 441 U.S. 91, 110-11 (1979). And it is of no import that New York City and Chicago could, theoretically, have foregone any efforts to remedy the harms caused by the citizenship question, as the “forced choice” between risking imminent harm and spending money to avoid it would constitute a cognizable Article III injury all by itself. Cf. Texas, 787 F.3d at 749.
Accordingly, Plaintiffs ADC, MRNY, NYIC, CASA, New York City, and the City of Chicago have each proved that the addition of a citizenship question to the 2020 census questionnaire will cause, and in some instances has already caused, cognizable Article III injury in the form of a diversion of valuable resources.
e. Loss of Privacy
Finally, the Court shares Plaintiffs’ view that any “invasion of privacy” that would be inflicted by the unlawful disclosure of confidential census data regarding individuals’ citizenship status would, it if were sufficiently imminent, constitute a cognizable Article III injury. A contrary view would fly in the face of “both the common law and the literal understandings of177Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 178 of 277privacy,” which “encompass the individual’s control of information concerning his or her person.” U.S. Dep’t of Justice v. Reporters Comm. for Freedom of Press, 489 U.S. 749, 763 (1989). That said, Plaintiffs did not prove that such an injury is sufficiently imminent to satisfy Article III in these cases. As Plaintiffs concede, it would be illegal for the Department of Commerce to “make any publication whereby the data furnished by any particular establishment or individual . . . can be identified.” 13 U.S.C. § 9(a)(2); Pls.’ Proposed Findings ¶ 1722. Consistent with that mandate, Dr. Abowd testified at trial that the Census Bureau will apply “disclosure avoidance techniques” to any data to ensure that information concerning particular respondents is not identifiable. Tr. 1033. To be sure, those techniques may reduce the fitness of the data for DOJ’s purposes — an issue addressed below. And, in theory, the statutory prohibition could be changed by a future Congress. But it is pure speculation to suggest that the Census Bureau will not comply with its legal obligations to ensure the privacy of respondents’ data or that those legal obligations will be amended. And given that, the fact that NGO Plaintiffs may be subjectively fearful that the government will misuse citizenship data obtained through the census, see Pls.’ Proposed Conclusions ¶ 55, however understandable such fears may be, is “insufficient to create standing,” Clapper, 566 U.S. at 417. Accordingly, the Court holds that Plaintiffs failed to prove that they have been, or will be, injured for Article III purposes through a loss of privacy.
3. Traceability and Redressability
In short, Plaintiffs proved that some have suffered, or will suffer, injury in fact in at least four ways: (1) diminished political representation, between and within states; (2) reductions in federal funding, again both between and within states; (3) harm to the accuracy and quality of census data; and (4) the diversion of resources. Thus, the Court turns to the second element of178Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 179 of 277standing, which requires that Plaintiffs prove a “causal connection” between their injuries and the conduct they challenge in this lawsuit. Lujan, 504 U.S. at 560.
Plaintiffs have done so here. First, the evidence at trial proved beyond any doubt, and certainly by a preponderance of the evidence, that the addition of a citizenship question will cause a disproportionate decline in self-response rates among households containing at least one noncitizen individual and that that, in turn, will force substantially more such households into the Census Bureau’s NRFU process. Right off the bat, that will cause a decline in the accuracy and quality of the data generated by the census, which will injure the Governmental Plaintiffs that rely on that data to make and enforce their laws. On top of that, the Court found, by a preponderance of the evidence, that the decline in self-response will translate into a net differential undercount of people who live in noncitizen households of at least 5.8%. That net differential undercount will, in turn, cause the harms set forth in the prior section of this Opinion. In arguing otherwise, Defendants attack the “traceability” of Plaintiffs’ injuries to the citizenship question in startlingly cursory fashion. After previewing these arguments with such fanfare, see Defs.’ Pre-Trial Br. 6 n.2; see generally id. at 5-20, Defendants devote a scant two paragraphs of their Proposed Conclusions of Law to contesting “traceability” on various grounds, each of which amounts to little more than a conclusory citation to the record, and each of which — even lending it the most charitable interpretation, despite Defendants’ near-abandonment of their supporting arguments — is ultimately unavailing. See Defs.’ Post-Trial Br. 70-71, ¶¶ 22-23.
The first such ground is the most familiar: Defendants argue that the connection between the citizenship question and Plaintiffs’ injuries is simply too “speculat[ive],” relying on too many “inferences” to satisfy Article III. Post-Trial Br. 67-68, ¶¶ 11-14.53 That argument, however,53
to the traceability — of only two types of injuries, namely lost political representation and lostNotably, Defendants raise this argument as an objection to the imminence — as opposed 179Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 180 of 277misunderstands both the law and the facts. Courts often dismiss claims for prospective relief because the plaintiffs’ claims that defendants will imminently cause them injury are too “speculative” to satisfy Article III. See, e.g., Clapper, 568 U.S. at 410-14. But, as the term suggests, the bar on “speculative inferences” in the standing analysis refers to speculation and inferences, as opposed to evidence and proof. To be sure, at the pleading stage, a plaintiff must plausibly allege facts connecting his or her injuries to the defendant’s conduct, and cannot substitute “speculative inferences” for plausible factual allegations of causation. See, e.g., Simon v. Eastern Ky. Welfare Rights Organization, 426 U.S. 26, 42-46 (1976) (ordering dismissal of a complaint that failed to allege that plaintiffs’ injuries “in fact result[ed] from” a challenged federal tax incentive through its effect on third parties, rendering it “purely speculative” whether those injuries were fairly traceable to the tax incentive “or instead result[ed] from decisions made” by the third parties “without regard” to the incentive (emphasis added)). But here, with a full trial record, the Court need not speculate: Plaintiffs proved each factual step in the causal chain and that each step is fairly traceable, at least in part, to the addition of a citizenship question. See, e.g., Mendia v. Garcia, 768 F.3d 1009, 1012-13 (9th Cir. 2014) (“[W]hat matters is not the length of the chain of causation, but rather the plausibility of the links that comprise the chain.” (internal quotation marks omitted)). So, while “highly attenuated chain[s] of possibilities” are not enough to survive a motion to dismiss, Clapper, 568 U.S. at 410 (emphasis added), here the Court deals with facts, not possibilities. Because the facts show both that Plaintiffs’ injuries are imminent and fairly traceable to Defendants’ conduct, this is not a case in
federal funding. Defs.’ Post-Trial Br. 67-68, ¶¶ 11-14. But for the Court’s independent obligation to assure itself of its own subject-matter jurisdiction, the Court would deem the objection abandoned as it relates to traceability — and perhaps Defendants meant to do just that, given the ultimate weakness of the argument. The Court addresses the “imminence” of Plaintiffs’ injuries in more detail below.180Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 181 of 277which “[s]peculative inferences are necessary to connect [Plaintiffs’] injury to the challenged actions of [Defendants].” Simon, 426 U.S. at 45. Indeed, the facts underlying the chain of causation connecting Defendants’ conduct to Plaintiffs’ injuries here is far stronger than those the Supreme Court found sufficient in Davis.
In fact, to the extent that either side in these cases invites the Court to substitute “speculation” for proof, it is Defendants, not Plaintiffs, who do so. Defendants assert that NRFU operations will remedy any decline in the self-response rate attributable to the citizenship question, Defs.’ Post-Trial Br. 16-20, ¶¶ 127-141, but, at the end of the day, they offer little more than a hope and prayer in support of that assertion. As discussed, NRFU operations have historically failed to remedy differential declines in self-response rates. See Recitation of Facts ¶¶ 211, 230. And Defendants offer no evidence in support of their claim that NRFU operations will do so this time; indeed, they have not even determined the imputation formulae that they will use in the 2020 census. See Tr. 1350-51. As a matter of fact, and as discussed at length above, there are several demonstrable reasons to believe that NRFU operations will do a worse job this time around than in past years in addressing any decline in response rate among Hispanic and noncitizen households. Considerable testimony supports the conclusion that NRFU will suffer from many of the same defects as the initial attempts to obtain answers through self- response. Dr. Hillygus testified that “all of the issues . . . with respect to confidentiality concerns associated with the citizenship question that the Census Bureau acknowledges and has shown to have an impact on the self-response, all matter for cooperation with a census enumerator” in NRFU, too. Tr. 97; see id. at 99-100 (discussing the likelihood that confidentiality concerns and the macroenvironment will hamper NRFU efforts more than in previous years). Dr. Barreto’s testimony supports the same conclusion, see Tr. 643-44, and also suggests that attempts to reassure potential NRFU respondents with confidentiality concerns will be disproportionately181Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 182 of 277less effective for Hispanic and immigrant populations. Tr. 688. Notably, these opinions were based in part on the Census Bureau’s own conclusions and studies. See Brown Memo at 43 n.60; see also PX-152; PX-662; PX-663. On top of that, it is undisputed that NRFU operations do nothing to address an undercount attributable to households that do self-respond but omit noncitizen members from those responses, Tr. 1309-10, a phenomenon that is substantially likely to rise in direct response to a citizenship question appearing on the census questionnaire, see Recitation of Facts ¶ 214. In short, Plaintiffs proved that they will imminently suffer a variety of harms that are fairly traceable to Secretary Ross’s decision; Defendants’ unsupported assertion that they will cure or mitigate those injuries before they materialize, supported by nothing more than a promise — and contradicted by both history and expert testimony about the conditions of the 2020 census — is not enough to render Plaintiffs’ injuries “speculative.” Cf. Duke Power Co. v. Carolina Envtl. Study Grp., Inc., 438 U.S. 59, 78 (1978) (noting, in the context of discussing traceability and redressability, that a plaintiff need not “negate . . . speculative and hypothetical possibilities in order to demonstrate the likely effectiveness of judicial relief”); Cent. Delta Water Agency v. United States, 306 F.3d 938, 950 (9th Cir. 2002) (“It would be inequitable in the extreme for us to permit one party to create a significantly increased risk of harm to another, and then avoid [sic] the aggrieved party from trying to prevent the potential harm because the party that created the risk promises that it will ensure that the harm is avoided[.]”).
Defendants’ second argument — and the one they come closest to abandoning (as it appears only in their proposed findings of fact) — is that the decline in self-response rates among Hispanic and noncitizen households is not “fairly traceable” to the addition of a citizenship question alone because it may also be traceable to an alternate cause, namely the “macroenvironment” of fear and distrust of government among the Hispanic and noncitizen182Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 183 of 277populations. See Defs.’ Post-Trial Br. 41-42, ¶¶ 272-76. But this is not how the “traceability” requirement works. A plaintiff must “demonstrate a causal nexus between the defendant’s conduct and the injury” to satisfy Article III’s traceability requirement — nothing more. Rothstein, 708 F.3d at 91 (emphasis added). And in these cases, overwhelming evidence (much of it from the Census Bureau itself and Defendants’ own expert witness, Dr. Abowd) supports the Court’s factual finding that Hispanic and noncitizen households will be less likely to respond to the 2020 census questionnaire if it includes such a question. That means that the addition of a citizenship question will — obviously — be a “but for” cause of the decline in self-response rates among those communities. It may well be true (and indeed, the trial evidence suggests) that adding a citizenship question in the current “macroenvironment” will lead to a greater decline than it would in another “macroenvironment.” But it would be perverse to suggest that merely because the background context for Secretary Ross’s decision will exacerbate its negative effects, that the decision is somehow not itself a cause of those effects. Even in a dry season, it is fair to trace the fire to the arsonist.
Defendants’ contrary argument — that the decline in self-response will be traceable to the macroenvironment, not the citizenship question — implies that Article III permits only one legally responsible cause per injury. But that is not even how the concept of “proximate” cause works, much less how the “fairly traceable” requirement works. An event can have more than one proximate cause, see, e.g., Staub v. Proctor Hospital, 562 U.S. 411, 420 (2011), and each proximate cause need not always even be a sufficient cause.54 And in any event, “proximate cause” is “not a requirement of Article III standing.” Lexmark Int’l, 572 U.S. at 134 n.6. While54
of the strikeout.” Oral Arg. Tr. 69, No. 16-980, Husted v. A. Philip Randolph Institute (U.S. Jan.Thus, for example, “[s]trike one, strike two, [and] strike three” are “all proximate causes 10, 2018) (Kagan, J.), available at 2018 WL 353954.183Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 184 of 277the fact that there is another cause “of the plaintiff’s injury may foreclose a finding of proximate cause,” it “is not necessarily a basis for finding that the injury is not ‘fairly traceable’ to the acts of the defendant.” Rothstein, 708 F.3d at 92; see also Block, 793 F.2d at 1309. Indeed, the Supreme Court’s cases imply an even more expansive traceability bar than the one Plaintiffs have cleared here: that a defendant’s conduct need only be a “but-for” cause of a plaintiff’s injuries in the sense that its removal from the causal chain, through the relief sought in the action, will be likely to redress the injuries. See Watt v. Energy Action Ed. Foundation, 454 U.S. 151, 161 (1981) (stating that a plaintiff must “show that there is a ‘fairly traceable’ causal connection between the injury it claims and the conduct it challenges, so that if the relief sought is granted, the injury will be redressed” (emphasis added) (citation omitted)); Duke Power Co., 438 U.S. at 74-78 (concluding that an injury “fairly can be traced to the challenged action of the defendant” if it is a “but-for” cause of the injury and thus likely to be redressed by the relief sought (internal quotation marks omitted)).
Next, Defendants make much of the fact that some steps in the causal chains involve the actions of third parties — namely, those who choose not to self-respond to the census because of the citizenship question’s presence on the questionnaire. It is certainly true that traceability can be destroyed where the “independent actions of third parties” are responsible for a plaintiff’s injuries. Lujan, 504 U.S. at 562 (emphasis added). But as Justice Alito’s opinion in Davis makes clear, see 554 U.S. at 734-35, where record evidence, statistical analysis, or just plain common sense support a finding, as they do here, that third parties will respond to the challenged government conduct in a predictable way, “traceability” is not defeated. In the words of then- Judge Scalia, “[i]t is impossible to maintain, of course, that there is no standing to sue regarding action of a defendant which harms the plaintiff only through the reaction of third persons.” Block, 793 F.2d at 1309. If Defendants’ argument were true, he continued, it would be “difficult184Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 185 of 277to see how libel actions or suits for inducing breach of contract could be brought in federal court” or “how state threats and intimidation directed at the distributors of certain books could confer standing upon the publisher whose sales are affected.” Id. Put simply, the “fairly traceable” standard rules out injuries produced by the “independent choices of third parties” only where those choices are truly “unfettered,” Lujan, 504 U.S. at 562 — that is, causally independent from the challenged conduct. Article III “requires no more than de facto causality,” Block, 793 F.2d at 1309, which the presence of third parties in the causal chain does not necessarily undermine, see Glavin v. Clinton, 19 F. Supp. 2d 543, 550 (E.D. Va. 1998) (three- judge court) (permanently and universally enjoining the Secretary of Commerce’s decision to use statistical sampling in enumerating the population for apportionment purposes, despite the presence of third-parties’ “intervening” actions in the chain of causation leading to the plaintiffs’ injuries), aff’d, Dep’t of Commerce, 525 U.S. 316. The only question is whether, as a matter of fact, Plaintiffs’ injuries are “fairly traceable” up the causal chain to Defendants’ conduct. Here, for the reasons explained above, they are.
Perhaps recognizing the weakness of their general arguments regarding the intervening acts of third parties, Defendants renew an argument that they first pressed in their motion to dismiss: that Plaintiffs cannot prove traceability here because the chain of causation depends on the intervening acts of third parties that are unlawful. See Docket No. 155, at 19-21; Defs.’ Pre- Trial Br. 6 n.2; Tr. 1497, 1503. Defendants’ argument — for which they cite no supporting authority, from the Supreme Court or otherwise — may deserve points for creativity, but it gives way under the strain of only a little thought. For one thing, the purpose of standing doctrine is to “ensure that courts exercise power that is judicial in nature” — that is, the power to adjudicate “cases” and “controversies” that Article III confers on the judicial branch (and keeps away from the other branches). Gill, 138 S. Ct. at 1930 (internal quotation marks omitted); accord Spokeo185Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 186 of 277136 S. Ct. at 1547; see U.S. Const. art. III, § 2. Congress can instruct the courts not to exercise judicial power over certain cases or classes of cases, see, e.g., Patchak v. Zinke, 138 S. Ct. 897, 910 (2018), but it cannot change what the Constitution says the judicial power is. Thus, while declaring one of the intervening steps in a causal chain to be unlawful might affect the underlying merits of a given claim, it would not, and could not, change whether adjudicating that claim qualified as an exercise in the “judicial power” as defined by Article III.Defendants’ argument appears also to suffer from a common confusion between the standing and merits inquiries, which — of course — are conceptually distinct. See, e.g., Whitmore v. Arkansas, 495 U.S. 149, 155 (1990) (“Our threshold inquiry into standing in no way depends on the merits of [the plaintiff’s] contention that particular conduct is illegal.” (internal quotation marks omitted)). The question whether Plaintiffs’ injuries are “fairly traceable” to Defendants’ conduct is not a merits inquiry, as for example the question of proximate causation in tort liability would be; with respect to the latter, Defendants would surely have a colorable argument that they should not be held liable in damages for the unlawful intervening actions of another. See Restatement (Second) of Torts § 448 (noting that, absent certain circumstances, “[t]he act of a third person in committing an intentional tort or crime is a superseding cause of harm to another resulting therefrom”). But nobody is suggesting that Defendants should share in the criminal (or civil) liability that people who fail to respond to the census may incur. Cf. Block, 793 F.2d at 1309 (“That argument could be relevant to the merits of a tort action seeking to hold the government liable for damages as the legal cause of [the plaintiff’s] injury; but it is irrelevant to the question of core, constitutional injury-in-fact, which requires no more than de facto causality.”). The Article III standing inquiry is entirely different: It looks to the chain of causation connecting Defendants’ conduct to Plaintiffs’ injuries, and it asks whether Plaintiffs have shown that one is sufficiently traceable to the other so as to give Plaintiffs a sufficiently186Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 187 of 277“personal stake” in the outcome of their challenge to create the adverseness required by Article III. Here, Plaintiffs have plainly made that showing.Given the foregoing, it is not surprising that Defendants fail to cite any authority in support of their novel argument. Perhaps more surprising is the ample authority that contradicts their argument that they simply ignore. Thus, for instance, courts have found, in data breach cases, that customers have standing to bring claims against the companies that failed to safeguard their data — even though the hacker or “thief would be the most immediate cause of plaintiffs’ injuries.” Attias v. Carefirst, Inc., 865 F.3d 620, 629 (D.C. Cir. 2017), cert. denied, 138 S. Ct. 981 (2018); Galaria v. Nationwide Mut. Ins. Co., 663 F. App’x 384, 390 (6th Cir. 2016) (“Although hackers are the direct cause of Plaintiffs’ injuries, the hackers were able to access Plaintiffs’ data only because Nationwide allegedly failed to secure the sensitive personal information entrusted to its custody.”); Lambert v. Hartman, 517 F.3d 433, 437-38 (6th Cir. 2008) (rejecting the argument that the intervening “criminal act of a third party” defeated standing where the plaintiff “link[ed] the act of identity theft” to the personal information divulged by the defendant). So too, in terrorist financing cases, courts have found standing — even though the most immediate cause of injury is the terrorist, not the bank or financier. See, e.g., Rothstein, 708 F.3d at 93 (“[W]e cannot conclude that the Complaint failed to allege sufficiently that plaintiffs’ injuries in bombings and rocket attacks conducted by Hizbollah and Hamas were fairly traceable to UBS’s provision of U.S. currency to Iran.”); Goldberg v. UBS AG, 660 F. Supp. 2d 410, 417 (E.D.N.Y. 2009) (“Here, while a number of independent third parties were involved in the attack on Bus 19, plaintiffs have alleged a coherent and plausible causal nexus linking UBS’s alleged wire transfers for ASP to the bombing of Bus 19.”); accord Wultz v. Islamic Republic of Iran, 755 F. Supp. 2d 1, 23 (D.D.C. 2010); cf. Mendia, 768 F.3d at 1013 (describing the plaintiff’s theory of traceability, that “the government’s unlawful conduct,187Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 188 of 277while not directly causing his injury, nonetheless led third parties to act in a way that injured him,” as “perfectly viable”). Most relevant for present purposes, Defendants’ argument is incompatible with the many decisions, most notably the Second Circuit’s decision in Carey, holding that “citizens who challenge a census undercount on the basis, inter alia, that improper enumeration will result in a loss of funds to their city have established . . . an injury fairly traceable to the Census Bureau.” Carey, 637 F.2d at 838; accord City of Detroit, 4 F.3d at 1375; City of Philadelphia, 503 F. Supp. at 671; City of Camden, 466 F. Supp. at 50.
Those basic standing principles are enough to turn away Defendants’ arguments against traceability here. The touchstone of any Article III standing injury is whether the plaintiff has suffered, or is likely to suffer, an injury that is fairly traceable to the challenged conduct and redressable by the relief sought. Although congressional action can obviously affect whether particular injuries are “redressable” by creating (or not creating) causes of action, the “injury in fact” and “traceability” inquiries are fundamentally practical inquiries, grounded in the real- world consequences of human interaction — the sorts of things that give rise to “concrete and particularized” injuries that give plaintiffs a “personal stake” in federal litigation. In both cases, Article III demands something less than an ultimate merits inquiry might require: only that the defendant’s conduct was a de facto cause of the plaintiff’s injury, not that it was the “legal” or “proximate” cause. Block, 793 F.2d at 1309. Thus, while all three strikes are proximate causes of the strikeout, the last domino to fall is fairly traceable to the first. On Defendants’ theory of traceability, Congress could abrogate standing to sue for that last domino falling by declaring it unlawful for the intervening dominoes to fall. But an injury’s traceability for purposes of Article III — like Plaintiffs’ injuries here — does not depend on whether the dominoes have congressional permission to fall; it depends only on whether, in fact, they will. Taking a practical look at the facts of these cases, the Court would have no trouble concluding that the188Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 189 of 277apportionment losses, funding losses, and harms to data quality that Plaintiffs have proved were proximately caused by Secretary Ross’s decision. It is even easier to conclude, as the Court does, that Plaintiffs’ injuries are fairly traceable to that decision.
That leaves only the element of redressability. Conspicuously, Defendants make no argument whatsoever concerning redressability — a tell, if there ever was one, that their arguments about traceability are themselves ultimately lacking. See, e.g., Watt, 454 U.S. at 161 (stating that a plaintiff must “show that there is a ‘fairly traceable’ causal connection between the injury it claims and the conduct it challenges, so that if the relief sought is granted, the injury will be redressed” (emphasis added) (citation omitted)). In any event, given the Court’s findings of fact and conclusions of law, it follows that Plaintiffs have proved that their injuries are “likely” to “be redressed by a favorable decision.” Lujan, 504 U.S. at 561 (internal quotation marks omitted). That is, if Secretary Ross’s decision to add a citizenship question to the 2020 census is set aside or enjoined, as Plaintiffs request in these cases and as the Court concludes it must be, it is likely that its effects on the net differential undercount will be mitigated to the point of relieving Plaintiffs’ injuries. Notably, to satisfy Article III’s redressability requirement, a plaintiff must show that the requested relief will remedy “an injury” to the plaintiff, not “every injury.” Larson v. Valente, 456 U.S. 228, 243 n.15 (1982). And here, although the data generated by the 2020 census may still be less than perfect (no census is perfect, after all), and the resulting political apportionment and funding allocations may not be accurate to the seat or dollar, Plaintiffs have proved that their injuries specifically caused by the citizenship question will be mitigated, if not wholly remedied, by its removal.
For the foregoing reasons, the Court concludes that most, if not all, Plaintiffs have standing to bring the claims pressed here — well more than the one Plaintiff with standing that189
Case 1:18-cv-02921-JMF Document 574 Filed 01/15/19 Page 190 of 277
the Court would need for jurisdiction.
There is no dispute that the Constitution, the Census Act, and the APA allow the Secretary of Commerce broad discretion over the design and administration of the decennial census. See Wisconsin, 517 U.S. at 18-20. Generally speaking, they do not preclude the Secretary from charting a new policy direction, even over the strenuous objections of career staff, or from recruiting other government officials to support such a change. Significantly, however, the discretion that they allow the Secretary is not unlimited. He must comply with the policy decisions that Congress — to which the Constitution gives authority over the census — has made and enshrined in statute, including but not limited to the preference for obtaining data through administrative records rather than through direct inquiries. He must follow the procedures mandated by law. And more broadly, the exercise of his statutory authority must “be reasonable and reasonably explained.” Mfrs. Ry. Co., 676 F.3d at 1096.
Measured against these standards, Secretary Ross’s decision to add a citizenship question to the 2020 census — even if it did not violate the Constitution itself — was unlawful for a multitude of independent reasons and must be set aside. To conclude otherwise and let Secretary Ross’s decision stand would undermine the proposition — central to the rule of law — that ours is a “government of laws, and not of men.” John Adams, Novanglus Papers, No. 7 (1775). And it would do so with respect to what Congress itself has described as “one of the most critical constitutional functions our Federal Government performs.” 1998 Appropriations Act,
§ 209(a)(5), 111 Stat. at 2480-81.
Accordingly, and for the reasons stated at length above, the Court vacates Secretary Ross’s decision to add a citizenship question to the 2020 census questionnaire, enjoins Defendants from implementing Secretary Ross’s March, 26, 2018 decision or from adding a question to the 2020 census questionnaire without curing the legal defects identified in this Opinion, and remands the matter to the Secretary of Commerce (to the extent that such a “remand” is even necessary) for further proceedings not inconsistent with the Court’s Order. (Opinion, pp. 276-277)
Judge rejects citizenship question for 2020 U.S. census
Nick Brown, Jonathan Stempel
5 Min Read
U.S. District Judge Jesse Furman in Manhattan said Commerce Secretary Wilbur Ross concealed his true motives in adding the question last March.
Ross had said the question - which has not appeared on the decennial census since 1950 - was necessary to enforce federal laws protecting eligible voters.
Furman’s decision will almost certainly be appealed, and could wind up before the Supreme Court this year.
The plaintiffs - 18 U.S. states, 15 cities and various civil rights groups - said that asking census respondents whether they are U.S. citizens will frighten immigrants and Latinos into abstaining from the count.
That could cost their mostly Democratic-leaning communities representation in the U.S. House of Representatives, as well as their share of some $800 billion a year in federal funding.
The plaintiffs alleged that was Ross’ plan all along, while he insisted the government needed citizenship data to better enforce the Voting Rights Act, which protects eligible voters from discrimination. Only American citizens can vote in federal elections.
Dale Ho, an attorney with the American Civil Liberties Union who argued the plaintiffs’ case, called Furman’s ruling “a forceful rebuke of the Trump administration’s attempt to weaponize the census.”
Kelly Laco, a Justice Department spokeswoman, said the administration was “disappointed,” adding that the “government is legally entitled to include a citizenship question on the census, and people in the United States have a legal obligation to answer.”
In a 277-page opinion, Furman called Ross’ Voting Rights Act rationale “pretextual.”
“He announced his decision in a manner that concealed its true basis rather than explaining it,” Furman said.
Ross said he added the question at the request of the Justice Department, but evidence at trial showed he independently pushed for it much earlier.
Ross also chose not to heed recommendations from experts - including from within the Census Bureau itself - who said adding the question would lead to an undercount and hurt data quality.
During a two-week trial in November, Justice Department lawyers argued Ross need not reveal every motivating factor, as long as his stated rationale was sound.
They also said he was under no obligation to take advice from experts. Ross “doesn’t have to choose the best option” as long as he considers all evidence in good faith, Justice Department lawyer Brett Shumate said at trial.
But Furman said Ross and his aides behaved “like people with something to hide,” leading to the “inescapable” conclusion that they “did have something to hide.”
The ruling means the Trump administration will have to keep litigating if it wants to preserve what has become one of the most controversial of its hawkish immigration policies.
Furman’s ruling bars Ross from re-adding the question unless he “cures the legal defects” in his rationale.
The case is thought likely to wind up before the Supreme Court, but time is short: the Census Bureau must print census forms sometime this spring.
Congress could also ban the question, which would eliminate the need for litigation. Democratic lawmakers have said they plan to use their newfound majority in the House to investigate the question.
Census consultant Terri Ann Lowenthal, a former staff director of the House census oversight subcommittee, said “lawmakers should overturn Secretary Ross’ decision, which the court correctly ruled was made in violation of legal requirements.”
Furman’s ruling muddies a census that is already among the most complicated and expensive in U.S. history.
The first-ever online census, set for April 2020, has been plagued by delays, missed deadlines and the cancellation of key tests, partly from legacy underfunding from Congress.
At least five other lawsuits seeking to quash the citizenship question remain pending.
Some advocates worry it may be too late to assuage the fears of immigrants in a climate of heightened immigration rhetoric.
“Hopefully a lot of damage hasn’t already been done, and immigrant communities ... will still come to the door when the census takers come,” said Doug Rand, a former Obama White House official who worked on immigration issues.
Reporting by Nick Brown and Jonathan Stempel in New York; additional reporting by Mica Rosenberg; Editing by Chizu Nomiyama and Jonathan Oatis
Our Standards:The Thomson Reuters Trust Principles.