If all lives matter, then all lives count.

This is the second in a series of posts regarding Secretary of Commerce Wilbur Ross’ March 26, 2018 memorandum (here) directing the Census Bureau to add a citizenship question to the 2020 decennial census form distributed to all households.  Administrative Record filed in New York v. Department of Commerce, Dkt. No. 18 Civ. 2921,  at 1313-20 (“Adm. Rec,”).  In his memorandum, the Secretary asserted that he had ordered the change, at the request of the Department of Justice, to enhance the information available to the parties litigating vote-dilution claims under the Voting Rights Act of 1965, Pub. L. 89-110, 79 Stat. 437, as amended by the Voting Rights Act Amendments of 1983, Pub. L. 97-205, 96 Stat. 131, §4 (codified at 52 U.S.C.A. § 10301).  Such suits involve allegations that state or local apportionment of legislative or other multi-member bodies have the effect of diluting minority communities’ voting power, in violation of section 2 of the Act.  This series of posts explores issues raised by two consolidated cases filed in the Southern District of New York challenging Secretary Ross’ decision, New York v. Department of Commerce, Dkt. No. 18 Civ. 2921 and New York Immigration Coalition v. Department of Commerce, Dkt. No. 18 Civ. 5025.  The first post (here) discussed the challenges made in the litigation and the District Judge’s disposition of the federal government’s motion to dismiss.

The plaintiffs in both lawsuits allege that the reason Secretary Ross offered for including the citizenship question was pretextual.  Indeed the New York Immigration Coalition plaintiffs allege that Ross acted with invidious intent, to harm minority communities.  (I identified three possible unstated motivations, policy, partisan, and racial, in my first post.)  The District Court can largely avoid an exploration of Secretary Ross’ motives if it finds his decision “arbitrary and capricious” under the Administrative Procedure Act (“APA”).

Pretextual decisions are often incoherent when subjected to close analysis.  Often the stated reason is not entirely consistent with the decision made or its underlying factual predicate.  Thus, it may be difficult for Secretary Ross’ decision to survive a “searching and careful review” if the Secretary’s motivations were partisan or invidious.  Accordingly, before turning to the question of motive, in the fourth post in this series, this post and the next will consider the vulnerability of Secretary Ross’s decision under the APA’s “arbitrary and capricious” standard of review.

The Arbitrary and Capricious Standard of Review

To paraphrase humorist Peter Finely Dunne’s Mr. Dooley, administrative agencies, even more than the Supreme Court, “follow the election returns.”  Clearly Secretary Ross wished to alter the approach of President Obama’s Secretary of Commerce to reflect the Trump Administration’s overall approach toward undocumented aliens and, possibly, its fundamentally different conception of legislative “representation.” (Granted, in this case Secretary Ross appears to have overturned a judgement consistently made over a number of Democratic and Republican administrations.)  As Justice Rehnquist noted in Motor Vehicle Manufacturers Ass’n v. State Farm Mutual Automobile Insurance, 463 U.S. 29  (1983)(“State Farm”):

A change in administration brought about by the people casting their votes is a perfectly reasonable basis for an executive agency’s reappraisal of [its decisions]. As long as the agency remains within the bounds established by Congress,^* it is entitled to assess administrative records and evaluate priorities in light of the philosophy of the administration.

Id. at 59 (Rehnquist, J., concurring in part and dissenting in part); accord, FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 189 (Breyer, J., dissenting); FTC. v.Wyndham Worldwide Corp., 799 F.3d 236, 252 & n. 15 (3d Cir. 2015); National Ass’n of Home Builders v. EPA, 682 F.3d 1032, 1043 (D.C. Cir. 2012)(Garland, J.); International Union v. Chao, 361 F.3d 249, 256 (3d Cir. 2004)(Pollack, J., concurring).  Indeed, Chevron deference is based on agency responsiveness to changes in governing philosophy brought about by presidential elections.  Chevron v. Natural Resources Defense Council, 467 U.S. 837, 865 (1984)(“an agency to which Congress has delegated policy-making responsibilities may, within the limits of that delegation, properly rely upon the incumbent administration’s views of wise policy to inform its judgments”).

Nevertheless, the abruptness of such changes is constrained by the “arbitrary and capricious” test. That standard is familiar to readers of this blog.  “A reviewing court must ‘consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.”  Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 416 (1971). In State Farm, the Court elaborated upon that basic standard as follows:

Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.

State Farm, 463 U.S. at 43.

Plaintiffs allege that Secretary Ross did not consider “the risk of an inaccurate count and the availability of alternative data that serves the federal government’s efforts to enforce section 2 [of the Voting Rights Act of 1965, as amended] equally well.”  First Amended Complaint, New York v. Department of Commerce, Dkt. No. 18 Civ. 2921 ¶191 (S.D.N.Y. April 30, 2018).  But Secretary Ross’ decision does address the first issue, namely the risk of an undercount, at some length.  Ross Memo at 3-4, 5-6.  Granted, that resolution appears to run counter to the judgment of the Census Bureau’s career staff (including its Chief Scientist), and may ultimately be found a clear error of judgment.

Secretary Ross also appears to address the information provided by the Census Bureau’s staff regarding alternative sources of information that could just as effectively provide the data to assist DOJ’s Voting Rights Act enforcement efforts.  Ross Memo at 2-3, 4.  (The judgment of the Census Bureau staff and the Chief Scientist appear in the Administrative Record. Adm. Rec. at 1278, 1284-85 (Memorandum from John M. Abowd, Chief Scientist, etc. to Wilbur Ross dated January 19, 2018); Adm. Rec. 1291-94 (Undated Response to Questions on the January 19[, 2018] Draft Census Memo, etc., Questions 12-23).  Again, one can certainly question whether the Secretary’s conclusion on this score ran counter to the facts or constituted a severe error in judgment.

Failure to Consider the March 31, 2017 Statutory Deadline for Submitting the Subjects for the Census

There is at least one important matter Secretary Ross appears not to have considered in reaching his decision — the statutory limitations on the timing of adding a question to the census.  Congress has prescribed a schedule for the preparation of the census questionnaire to ensure that the census is conducted in a timely and orderly manner.  In particular, 13 U.S.C. §141(f) provides that “a report containing the Secretary’s determination of the subjects proposed to be included, and the types of information to be compiled” should be submitted three years before the census date.  13 U.S.C. §141(f)(1).  “The questions proposed to be included in [the] census” are to be submitted two years before the census date.  13 U.S.C. §141(f)(2).  While Secretary Ross acknowledges the second deadline in his decision, Ross Memo at 2, he ignores the first.  Section 141(f)(3) provides that:

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

This well-established statutory limitation does not appear to have been amended since 1976.

In his decision, Secretary Ross failed to recognize and address this statutory limitation on the exercise of his discretion.  (Indeed, he did not acknowledge his departure from the elaborate schedule of consultations with the federal agencies and “stakeholders” on the subjects for the 2020 Census laid out in a April 29, 2016 Memorandum from the Associate Director of Decennial Census Programs.  Adm. Rec. at 1168-71.)  While the Justice Department “request” came well after the §141(f)(1) deadline, no change relevant to Voting Rights Act enforcement requirements occurred after March 31, 2017.  See, Letter from the U.S. Conference of Mayors to Wilbur Ross dated February 6, 2018 at 3 (Adm. Rec. at 1075).  Indeed, the principle of using citizen voting age population (“CVAP”) in a specialized subset of vote dilution cases, the nominal basis for adding the citizenship question, had been well-established by the late 1990’s.  In any event, Secretary Ross had decided to pursue “reinstating” the citizenship question before March 31, 2017.  Kravitz v. Department of Commerce, Dkt. No. 18 Civ. 1041, Supplemental Memorandum by Secretary of Commerce William Ross Regarding Administrative Record in Census Litigation (D. Md.); Brennan Center Documents at 1137.

Section 141(f)(3) plainly requires the Secretary to have a special reason for adding a subject to the census after the March 31, 2017 deadline.  Some justifications for modifying the subjects of the census before the deadline cannot suffice to justify making such a change after the deadline.  Secretary Ross does conclude that the cost of preparing and adding the question will be minimal, but by itself the minimal cost of making the change does not constitute a “new circumstance” necessitating expansion of the subjects broached on the decennial census form.  The rarity of a request for a new question on the form sent to all households is worth noting — the DOJ “request” is virtually unprecedented in the last 30 years.  Adm. Rec. at 1296 (Undated Response to Questions on the January 19[, 2018] Draft Census Memo, etc., Questions 29, 31, 32).

A Clear Error of Judgment in Assessing Uncertain Risks

Secretary Ross makes a clear error of judgement in credulously accepting the Department of Justice’s statement of the need for such information, while adopting an extraordinarily skeptical view of the risk of an undercount noted by Census Bureau staff and the overwhelming number of “stakeholders” commenting on the issue.  The difference in scrutiny is particularly problematic given the severe and irreparable consequences of an error in underestimating the non-response rate and the more modest and remediable consequences of discounting the Justice Department’s asserted need for data.

Administrators’ often must act in circumstances of uncertainty – they must make predictive judgments about risks that cannot be fully quantified.  “It is not infrequent that the available data does not settle a regulatory issue and the agency must then exercise its judgment in moving from the facts and probabilities on the record to a policy conclusion.”  State Farm, 463 U.S. at 52.  “Policymaking in a complex society must account for uncertainty.”  Id.  Thus, “an agency confronted with scientific uncertainty has some leeway to resolve that uncertainty by means of more regulation or less.”  Cellular Phone Taskforce v. FCC, 205 F.3d 82, 91-92 (2d Cir. 2000); Center for Auto Safety v. Federal Highway Administration, 956 F.2d 309, 315-16 (D.C. Cir. 1999).  So Secretary Ross has some leeway in dealing with the uncertain consequences of his decision.

The Need for the Citizenship Question

The Department of Justice’s December 17, 2017 letter requesting that the citizenship question be added to the census was virtually the sole stated basis for Secretary Ross’s decision.  Indeed, he cites the letter in the very first sentence of his memorandum and discusses no other evidence of need.  Thus, he purports to base his decision on the interest of enhancing the ability of minority communities to achieve representation proportional to their size.

But the DOJ letter is curiously vague and short on specifics regarding actual litigation in which the absence of such data has proven troublesome.  (This is hardly surprising given that the Department of Commerce aggressively solicited the letter, from a seemingly indifferent DOJ, and even appears to have provided the cases DOJ could cite in establishing the legitimacy of its “request.” Brennan Center documents at 2386.)  And, strikingly, after sending the letter, “DOJ leadership” declined to meet with Census Bureau officials to elaborate on the points made in the Department’s letter.  Brennan Center documents at 2136.  (Note, since many of the Brennan Center documents were heavily redacted by the Department of Commerce, later communications with DOJ may have been redacted.)

The DOJ’s letter establishes that, as a general proposition, there is a subset of section 2 vote dilution cases, namely those where a jurisdiction has a large percentage of undocumented aliens and racially-polarized voting, in which “citizen voting-age population is the proper metric for determining whether a racial group could constitute a majority in a single member district.”  For example, the rule applied in Barnett v. City of Chicago, 141 F.3d 699 (7th Cir. 1998), because 40% of the Hispanics in the jurisdiction were aliens and a pattern of polarized voting was present, id. at 702.  How many cases involve such circumstances?  The DOJ letter leaves the reader to guess.  And note, the letter does not reference a single case litigated by DOJ in which it has confronted this problem, much less been unable to prove their case due to lack of citizenship data.  (Notably the DOJ website suggests that the Sessions Justice Department has brought no new section 2 cases. (See here.))  All of the cases cited in the letter are those litigated by private litigants.

Even so, in only one of the cited cases does the limitations of the data concerning citizenship seem to have mattered.  In Reyes v. City of Farmers Branch, 586 F.3d 1019 (5th Cir. 2009), plaintiffs were unable to prove their section 2 case, id. at 1021-23.  The remaining cases undermine the rationale for the request or are irrelevant to it.

At least two of the cited cases actually undermine DOJ’s request.  In Negron v. City of Miami Beach, 113 F.3d 1563 (11th Cir. 1997), the Court found the Census Bureau’s American Community Survey (“ACS”) data reliable, and rejected plaintiff’s claim due to plaintiff’s expert’s failure to use that data.  Id. at 1567, 1568-69.  The Court explained: “The use of sample data is a long-standing statistical technique, whose limits are known and measurable. We will not reject the citizenship statistics solely because they are based on sample data without some indication that the sample was tainted in some way.”  Id. at 1570.

In Barnett v. City of Chicago, 141 F.3d 699 (7th Cir. 1998), Judge Posner, writing for an appellate panel, explains why precise citizenship data is not critical.  He notes a standard rule of thumb for creating majority-minority districts given a multitude of causes for lower voter registration and lower voter turn-out in minority communities.  Generally, a majority-minority district for African-Americans requires that 65% of the populace be African-American and a majority-minority district for Hispanics requires that 70% of the population be Hispanic.  Id. at 702-03.  And, ironically, Judge Posner explains why the decennial census counts population, rather than voting age population: “The census counts the total population for the same reason that it counts rather than samples: in order to minimize controversy. To verify the age and citizenship of the population would enormously complicate the decennial census and open the census-takers to charges of manipulation.”  Id. at 704.

The two other cases cited also provide little support for the DOJ request.  In Romero v. City of Pomona, 883 F.2d 1418 (9th Cir. 1989), overruled in part on other grounds by, Townsend v. Holman Consulting Corp., 914 F.2d 1136, 1141 (9th Cir. 1990), there appears to have been no dispute between the parties with regard to citizenship statistics.   Id. at 1420-21, 1425-26.  Moreover, a majority-minority district could not be created because the court found no unity of interests between the African-American and the Hispanic communities.  Id. at 1426-27.  And in League of United Latin American Citizens v. Perry, 548 U.S. 399 (2006), plaintiffs prevailed on their vote dilution claim.

Perhaps the citizenship questions, if answered accurately, would lessen the complexity of such litigation.  However, it is not clear that information collected on a one-shot basis will necessarily definitively resolve such controversies.  The results may merely be one more data point over which parties to such litigation will argue.

So while the DOJ letter establishes a general legal proposition, the cited cases say almost nothing about the inadequacy of the Census Bureau’s ACS data in terms of the needs of private litigants bringing section 2 cases, much less the Department of Justice.

Note that DOJ could have sought information regarding felon disenfranchisement on the same basis as it requested for information on citizenship.  The presence of disenfranchised felons can have the same effect as the presence of non-citizens on the percentage majority needed to form a reliably majority-minority legislative district.  Just as a higher percentage of Hispanics are undocumented aliens, a higher percentage of African-Americans and Hispanics are disenfranchised due to felony convictions.    Obviously there are differences between disenfranchised citizens and non-citizens in terms of membership in the polity and entitlement to representation.  But in terms of the effect of reducing the voting power of minority communities, the result is similar.  Secretary Ross did not question that lack of a request for such a question on the census, nor did the DOJ letter explain why it failed to make such a request.  This is revealing; clearly the desire to ask the citizenship question has little to do with identifying vote dilution efforts and much more to do with the status of undocumented aliens.

The Risk of An Undercount

Meanwhile, Secretary Ross applied exacting scrutiny to the intuitive proposition underlying the concern about adding the citizenship question — that undocumented aliens will be less likely to return census forms if those forms ask about their citizenship. Secretary Ross asserts that the Census Bureau did not provide “definitive, empirical support” for their belief that adding the citizenship rate would reduce non-citizen’s response rates.  Id. at 4.  For example, though the Census Bureau provided statistics showing that the decline in the self-response rate from the short form (that lacked a citizenship question) to a the long form (that contained the citizenship question) was 3.3% greater for non-citizen households than for citizen households, Secretary Ross rejected the implications of the data because “Census was not able to isolate what percentage of the decline was caused” by the citizenship question rather than other questions on the survey.  Id.  Though stakeholders recounted the difficulty of obtaining information from undocumented aliens for purposes of fire relief efforts, Secretary Ross concluded that “no one provided evidence” that adding the citizenship question “would materially decrease response rates among those who generally distrusted the government and government information efforts, disliked the current administration, or feared law enforcement.”  Id. at 5.  Secretary Ross noted that while it is possible that there were citizens who would accurately respond to a census form without a citizenship question but would not respond if it included one, “no one has identified any mechanism for making such a determination.”  Id.  Ultimately, he concluded “[t]he Department of Commerce is not able to determine definitively how inclusion of a citizenship question on the decennial census will impact responsiveness.” Id. at 7 (emphasis added).

Similarly, Secretary Ross breezily dismissed evidence that using data from other agencies would produce a more accurate picture of the population of non-citizens than would a survey approach, using either the short form sent to all households or the long form sent to a sampling of households.  Ross Memo at 4; but see, Adm. Rec. at 1278, 1284-85 (Memorandum from John M. Abowd dated January 19, 2018); Adm. Rec. at 1291-94 (Undated Response to Questions on the January 19 Draft Memo); Adm. Rec. at 1311-12 (Memorandum from John M. Abowd to Wilbur Ross dated March 1, 2018)(concluding that survey data for those not covered by administrative records of other agencies are so unreliable that it calls into question whether such data will increase the accuracy of using data from other federal agencies).

The Relative Consequences of Error

The difference in the approach to assessing the DOJ’s need for the information and the risk of increasing the non-response rate among non-citizens is striking.  But more importantly, an assessment of the consequences of an erroneous judgment would have led a reasonable decision-maker to take exactly the opposite approach.  The consequences of an undercount resulting from inclusion of the citizenship question are more severe and irreparable than the consequences of the failure to provide DOJ with citizenship data, which could largely be counteracted.

The census of all households is conducted only once in ten years.  If there is a severe undercount due to insertion of the citizenship question, the undercount cannot be corrected until 2030.  Certainly its effect on the apportionment of Congress and the Electoral College is irreparable.  See, Pub. L. 105-119, Title II, §209, Nov. 26, 1997, 111 Stat. 2480 (“Congress finds that — (8) . . . if [the decennial] enumeration is conducted in a manner that does not comply with the requirements of the Constitution or laws of the United States, it would be impracticable for the States to obtain, and the courts of the United States to provide, meaningful relief after such enumeration has been conducted”).  And malapportionment of Congress and the Electoral College (not to mention distortion of the pattern of aid to states and localities) is a fundamental matter that could have profound effects.  Ross Memo at 1 (apportionment of Congress is a “foundational element of our democracy”).  On the other hand, any consequences from rejection of the DOJ request would be far have dramatic and could have been address by revising other census surveys done between decennial censuses (granted such surveys suffer from imperfections) and using data sources from other federal agencies to provide DOJ and section 2 plaintiffs with more precise data.

A Second Error: Discounting the Current Anti-Immigrant [Nativist] Climate

The Secretary made a second clear error in judgement when he discounted the effect of the policies of the Administration on undocumented aliens and their family members’ trust in the census.  Ross Memo at 5 (discussing concerns stakeholders raised about the effect of the “political climate” on the risk that adding the census question would reduce the response rate).  The figures for non-response rates Ross relied upon are all derived from results of surveys sent out during the Obama Administration (or earlier), namely the 2000 decennial census, the 2010 decennial census and the 2013-16 ACS surveys.  Id. at 3-4.  The Trump Administration has demonized undocumented aliens, aggressively sought to deport them, and sought to force sanctuary jurisdictions to provide information (jurisdictions that provide assurances of confidentiality not entirely different from those provided by the Census Bureau).  In such an atmosphere, it is unlikely that response rates will be similar to response rates during the Obama Administration (or the early days of the George W. Bush Administration).  While the Obama Administration engaged in significant efforts to deport undocumented aliens, that effort was not as aggressive as the present Administration’s efforts and did not give voice to the type of hostile nativist sentiments currently echoed by President Trump.  Indeed, between 2013 and 2016, the non-response rate to the citizenship question on the ACS questionnaire rose from 13% to 15.5% for Hispanic households.  Adm. Rec. at 1280.  There was no increase in the 6.2% response rate for non-Hispanic Whites during that same period.  Id.

Conclusion and the Subject of the Next Post

Secretary Ross’ asserted concern regarding protection minority voting rights, by ensuring that minority communities’ attain legislative representation commensurate with their size, is simply not credible.  Aside from the matters noted above, the strength of the civil rights credentials of the numerous elected officials and non-governmental group opposing inclusion of the citizenship question in comparison to those of the supporters of the move should have suggested to the Secretary that his analysis was problematic.  (See the Administrative Record for the contacts with the Secretary pro and con regarding addition of the citizenship question.  Adm. Rec. at 763-1276.)  And Secretary Ross does not appear to have reached out to any lawyers or groups that have actually litigated Section 2 Voting Rights Act cases.  (Instead, the Department reached out to a scholar at the American Enterprise Institute, who responded: “None of my colleagues would speak favorably about the proposal.  Is it important that the person actually be in favor of the proposal?”  Brennan Center Documents at 2083-84.)

While CVAP statistics can be useful in a subset of Voting Rights Act cases, such statistic would also be useful in advancing a project on which some conservatives have focused more attention, establishing CVAP as the primary means of implementing the Fourteenth Amendment’s one-person-one-vote requirement for multi-member bodies in all circumstances.  See, e.g., Evenwel v. Abbott, — U.S. —, 136 S.Ct. 1120 (2016)(rejecting the argument that use of CVAP for redistricting if required); Amicus Curiae Brief of Mountain States Legal Foundation in Support of Appellants at 2-4, 2015 WL 4747989, Evenwel v. Abbott, 136 S.Ct. 1120 (2016); Letter from Peter Kirsanow, Commissioner, Civil Rights Commission to Wilbur Ross dated March 15, 2018 [Adm. Rec. at 1217-19](if citizenship data is collected it will send an important message to elected representatives: “You are elected to represent American citizens”); Letter from Mike Hunter, Attorney General for the State of Oklahoma, et al. to Wilbur Ross dated March 13, 2018 [Adm. Rec. at 1210-11] (“Citizenship still matters. It is a privilege that is important and meaningful and should not be ignored”); Letter from Kris W. Kobach, Secretary of State for the State of Kansas to Wilbur Ross dated February 12, 2018 [Adm. Rec. at 1141-42](distinguishing undocumented aliens from legal aliens and aliens “on the cusp” of becoming U.S. citizens); Letter from Jeff Landry, Louisiana Attorney General to Wilbur Ross dated February 8, 2018 [Adm. Rec. at 1079-80](complaining that the voting power of sanctuary jurisdictions is increased by counting undocumented aliens in the census, thereby “undermining one of the most precious rights of citizenship,” namely the right to vote.)  Given the overall philosophy of the Trump Administration and the nature of the Administration’s support among the electorate, see, Brennan Center Documents at 2101, it is far more likely that Secretary Ross sought to advance this alternative more expansive agenda.  But seeking to uphold Secretary Ross’ decision on that basis raises three issues, which will be the subject of my next post in this series.